Do Electoral Systems Matter? Studying the Relationship between Electoral Institutions and Support in Italy and New Zealand

Material Information

Do Electoral Systems Matter? Studying the Relationship between Electoral Institutions and Support in Italy and New Zealand
Hamill, Jeffrey Scott
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (490 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Political Science
Committee Chair:
Moraski, Bryon
Committee Co-Chair:
Kreppel, Amie D
Committee Members:
Craig, Stephen C
Martinez, Michael D
Ardelt, Monika
Graduation Date:


Subjects / Keywords:
Democracy ( jstor )
Electoral systems ( jstor )
Government corruption ( jstor )
Parliaments ( jstor )
Political corruption ( jstor )
Political elections ( jstor )
Political parties ( jstor )
Political systems ( jstor )
Public policy ( jstor )
Voting ( jstor )
Political Science -- Dissertations, Academic -- UF
democracy -- elections -- majoritarian -- proportional -- support
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Political Science thesis, Ph.D.


This work is a study of the relationship between electoral systems and support. Support for governing institutions among populations is a very timely topic of study. The decline in the level of support for governing institutions has been offered as a reason for the mass protests that have been increasingly prevalent world-wide. I hold that an interesting and understudied area in political science is the scholarly research linking electoral systems and support. I draw, in this work, from the literature on electoral systems and support competing theoretical models linking different electoral systems and support. The theory is that elections can influence the decisions of voters and parties. That influence can help to determine the type of governments that are formed, the way they operate, and the ability of those governments to pass policies. The way governments operate and the types of policies they produce can influence support. The two competing models I draw link both increased majoritarianism and increase proportionality to higher levels of support. To test the competing models I offer New Zealand and Italy as excellent cases. New Zealand was a perfect example of a majoritarian system and changed its electoral system to a more proportional system. This presents a great opportunity to test the majoritarian model and the proportional model, while holding other political variables constant. Italy was a model of proportional government for a long period and attempted to increase majoritarianism through an electoral system change as well. They offer the same type of opportunity to test both the proportional and majoritarian models. Through a study of New Zealand and Italy before and after electoral system change, I hope to add the existing literature in several ways. Through the case study of New Zealand I believe we can learn how majoritarian electoral systems can work and also why they fail. In Italy, the reason proportional systems succeed and fail can also be learned as well. Both countries offer the chance to study whether electoral system changes can be successful raising the level of support among citizens. Ultimately, I hope to discern whether a credible link between electoral systems and supportive attitudes can be made. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis (Ph.D.)--University of Florida, 2012.
Adviser: Moraski, Bryon.
Co-adviser: Kreppel, Amie D.
Statement of Responsibility:
by Jeffrey Scott Hamill.

Record Information

Source Institution:
Rights Management:
Copyright Hamill, Jeffrey Scott. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
864880567 ( OCLC )
LD1780 2012 ( lcc )


This item has the following downloads:

Full Text




2 2012 Jeffrey S. Hamill


3 To Mom and Dad, my best friends


4 ACKNOWLEDGMENTS I thank my parents for their continued love and support I also would like to give special thanks to Bryon Moraski and Amie Kreppel for your patience, motivation, and advice throughout the years. The help of Stephen Craig, Michael Martinez, and Monica Ardelt for serving on my supervisory committee also deser ves special acknowledgment. Finally, I offer appreciation to the Department of Political Science and the Center for European Studies for the opportunity and financial assistance to pursue this project. With any of those listed, this work would not have b een possible.


5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 STUDYING SUPPORT ................................ ................................ ........................... 15 Introduction ................................ ................................ ................................ ............. 15 Defining Support ................................ ................................ ................................ ..... 19 Objects of Support ................................ ................................ ................................ .. 22 Why Support Matters ................................ ................................ .............................. 26 The Basis of Political Support ................................ ................................ ................. 31 erspectives ................................ ................................ ......................... 32 ................................ ................................ ......................... 35 Moving Forward: Where and What We Should Study ................................ ............. 42 Which Object? ................................ ................................ ................................ .. 42 Electoral System Change and Support ................................ ............................ 46 Measuring Support ................................ ................................ ................................ 50 Political Trust ................................ ................................ ................................ .... 51 Satisfaction with Democracy ................................ ................................ ............ 53 Preference for Alternatives ................................ ................................ ............... 57 Participation ................................ ................................ ................................ ...... 58 Moving Forward: Linking Electoral Systems and Support ................................ ...... 60 2 LINKING ELECTORAL SYSTEMS AND SUPPORT ................................ .............. 61 Introduction ................................ ................................ ................................ ............. 61 The Political Consequences of Electoral Systems ................................ .................. 66 Linking Proportionality to System Support ................................ .............................. 70 Possible Breaks in the Chain: The Links between PR and Support ........................ 80 Majoritarian Electoral Systems and Support: An Alternative Model ........................ 85 Possible Breaks in the Chain: The Link between Majoritarian Systems and Support ................................ ................................ ................................ ................ 89 The Best of Both Worlds? The Mixed Member Alternative ................................ .... 96 Defining and Surveying the Types of Mixed Member Electoral Systems ......... 96 Do Mixed Systems have Contamination Effects? ................................ ........... 103 Testing the Competing Models: New Zealand and Italy ................................ ........ 107


6 Conclusion ................................ ................................ ................................ ............ 113 3 THE RISE AND FALL OF MAJORITARIAN GOVERNMENT IN NEW ZEALAND 116 Introduction ................................ ................................ ................................ ........... 116 New Zealand: A Model of Majoritarian Government ................................ ............. 118 Making Majoritarianism Work in New Zealand ................................ ...................... 122 Producing a Two Party System: Disproportionality ................................ ........ 122 Fostering Accountability: Process in New Zealand Prior to 1984 ................... 126 Policy in New Zealand: Pre 1984 ................................ ................................ ... 132 Policy Changes ................................ ................................ .............................. 135 Procedural Change ................................ ................................ ........................ 138 New Zealand: A Picture of Declining Support ................................ ....................... 142 Party Membership ................................ ................................ .......................... 143 Turnout ................................ ................................ ................................ ........... 144 Testing Policy and Process Effects on Electoral System Choice .......................... 146 Dependent Variable ................................ ................................ ........................ 147 Independent Variables ................................ ................................ ................... 147 Expectations ................................ ................................ ................................ ... 153 Hypotheses ................................ ................................ ................................ .... 155 Results ................................ ................................ ................................ ........... 156 Control Variables ................................ ................................ ............................ 160 Conclusions ................................ ................................ ................................ .......... 161 4 T HE RISE AND FALL OF PROPORTIONAL GOVERNMENT IN ITALY .............. 169 Introduction ................................ ................................ ................................ ........... 169 Italy: A Model of Proportional Government ................................ ........................... 175 Proportional Elections and Political Parties: Italy 1948 1992 ................................ 179 The Process of Representation: Parties, Clientelism, and Coalitions in Italy, 1948 1979 ................................ ................................ ................................ ......... 182 The DC and PCI Representation: Associations and Clientelism, 1948 1979 186 The Formation of Governments an d Transformismo, 1948 1979 ................... 198 Policy in Italy: 1948 1979 ................................ ................................ ...................... 204 Committees ................................ ................................ ................................ .... 205 Policy Accomplishments, 1948 1979 ................................ ............................. 210 The Changing Nature of Representation in Italy ................................ ................... 214 Changes in Process: The Lo ss of Representation, 1979 1992 ...................... 217 Coalitions, Clientelism, and Corruption ................................ .......................... 223 Policy, 1980 1992 ................................ ................................ ................................ 234 Conclusion: The Proportional Model and Support in Italy, 1948 1992 .................. 237 Satisfaction with Democracy ................................ ................................ .......... 238 Turnout ................................ ................................ ................................ ........... 238 Party Membership ................................ ................................ .......................... 239 Moving forward ................................ ................................ ............................... 240


7 5 AN EXPERIMENT IN MIXED MEMBER PROPORTIONAL ELECTIONS: NEW ZEALAND 1996 2005 ................................ ................................ ........................... 249 Introduction ................................ ................................ ................................ ........... 249 The New Electoral Systems an d Expectations in New Zealand ............................ 252 MMP Rules ................................ ................................ ................................ ..... 252 Expectations ................................ ................................ ................................ ... 254 Elections under a New System: New Zealand, 1996 2005 ................................ ... 257 Elections under MMP 1996 2005: Proportionality and Ideological Representation ................................ ................................ ............................ 258 New Processes in New Zealand: Coalition Government and Party Hopping ........ 264 Coalition Formation/Government 1996 2005 ................................ .................. 265 Party Switching 1993 2002 ................................ ................................ ............ 278 Policy in New Zealand under MMP: 1996 2005 ................................ .................... 286 Reversing Past Policies and Reestablishing Some S ocial Welfare: 1996 2005 ................................ ................................ ................................ ............ 286 Policy Change Slows ................................ ................................ ...................... 289 Testing Policy and Process Effects: New Zealand 1999 2008 .............................. 292 Dependent Variable ................................ ................................ ........................ 293 Independent Variables ................................ ................................ ................... 294 Methodology ................................ ................................ ................................ ... 300 Expectations ................................ ................................ ................................ ... 300 Results ................................ ................................ ................................ .................. 303 Process Variables ................................ ................................ .......................... 303 Policy Variables ................................ ................................ .............................. 309 Discussion: Support for MMP in New Zealand ................................ ............... 315 Successes/Failure of the Proportional Model: Implications ............................ 317 6 TRYING MIXED MEMBER MAJORITARIAN ELECTIONS: ITALY, 1994 2005 ... 325 Introduction ................................ ................................ ................................ ........... 325 The New Electoral System in Italy ................................ ................................ ........ 330 Mixed Member Majoritarian Rules, Italy 1994 2005 ................................ ....... 330 Expectations ................................ ................................ ................................ ... 334 Direct Effects of MMM in Italy, 1994 2005 ................................ ............................ 338 Elections under MMM ................................ ................................ ..................... 340 MMM and the Majoritarian Model ................................ ................................ ... 348 2005 ........... 355 Mani Pulite, New Parties, and New Members ................................ ................ 355 Dealing with Corruption: The Center Right and Center Left .......................... 363 Public Attitudes toward Corruption: 1994 2005 ................................ .............. 374 Italian Policy Making: 1996 2005 ................................ ................................ .......... 380 Policy Making: Center Left and Center Right ................................ ................. 382 Discussion: Support in Italy under MMM 1994 2005 ................................ ...... 394 Conclusion: Thinking about Electoral Reform in Italy ................................ ............ 398 7 WHAT DO ELECTORAL SYSTEMS HAVE TO DO WITH SUPPORT? ............... 412


8 Introduction ................................ ................................ ................................ ........... 412 The Connection between Electoral Systems and Support ................................ .... 414 Can Electoral Systems Foster Supportive Attitudes? ................................ ..... 417 Le ssons from New Zealand and Italy ................................ ............................. 432 Contributing and Confounding Factors ................................ ................................ 437 Political Context as a Contributing and Confoundin g Factor .......................... 437 Institutional Rules as a Contributing and Confounding Factor ........................ 446 Agency as a Contributing and Confounding Factor ................................ ........ 454 Moving Forward: Thinking about Electoral Systems and Support ........................ 460 How to Increase Accountability and Representation ................................ ...... 461 Thoughts for Further Research ................................ ................................ ...... 466 LIST OF REFERENCES ................................ ................................ ............................. 4 68 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 490


9 LIST OF TABLES Table page 3 1 Two party dominance, disproportionality in New Zealand, 1943 1993 ............. 165 3 2 Third party support and disproportionality, New Zealand 1954 1993 ............... 166 3 3 Effects of policy and process variables on MMP versus FPP choice, New Zealand 1993 ................................ ................................ ................................ .... 168 4 1 Percentage of the vote and disproportionality, Italian Chamber of Deputies 1948 1992 ................................ ................................ ................................ ........ 242 4 2 Parties winning seats in Ita lian elections, 1948 1992 ................................ ....... 243 5 1 Summary of parliamentary election results: New Zealand 1996 2005 .............. 321 5 2 Effect of process a nd policy variables on MMP versus alternative system choice: New Zealand 1999 2008 ................................ ................................ ...... 322 6 1 Electoral Coalitons under MMM in the Chamber of Deputies, Italy 1994 2001 403 6 2 Electoral Coalitions under MMM in the Senate, Italy 1994 2005 ...................... 404 6 3 Italian Parliamentary Seniority by Group, Chamber of Deputies 1994 and 199 6 ................................ ................................ ................................ ................. 405 6 4 Italian Parliamentary Seniority by Group, Senate 1994 and 1996 .................... 406 6 5 Italian public perceptions of corruption in 2 004 comparted to 1992 .................. 409 6 6 Italian perceptions on the clean hands investigations After 12 Years ............... 410


10 LIST OF FIGURES Figure page 3 1 New Zealand turnout, 1954 1993 ................................ ................................ ..... 167 4 1 Partisan attachment, Italy 1978 1996 ................................ ............................... 244 4 2 Number of decrees issued versus decrees converted, Italy 1948 1996 ........... 245 4 3 C omparison of two corruption indexes, Italy 1980 1996 ................................ ... 246 4 4 Satisfaction with democracy, Italy 1973 1993 ................................ .................. 247 4 5 Turnout percentage, Italy 1948 1994 ................................ ................................ 248 5 1 Support for MMP versus an alternative system, New Zealand 1993 2008 ....... 324 6 1 Control of corruption index, Italy 1994 2005 ................................ ..................... 407 6 2 Corr uption perception index, Italy 1994 2005 ................................ ................... 408 6 3 Satisfaction with democracy, Italy 1994 2005 ................................ .................. 411


11 LIST OF ABBREVIATION S ACLI Association of Christian Wor kers ACT NZ ACT New Zealand Party AD Democratic Alliance AN National Alliance CCD Christian Democratic Center CDU Christian Democratic Union CISL Catholic Trade Union Association COC Control of Corruption Index CPI Perceptions of Corruption Index C S Socia l Christian Party DC Christian Democratic Party EC European Community EFIM Italian Machinery Trade Fares EMU European Monetary Union ENEA National Agency for New Technologies ENEL National Energy and Electric Entity ENI United Refineries (present day no lo nger an acronym) FI FPP First Past the Post GDP Gross Domestic Product IRI Institute for Industrial Reconstruction LSQ Least Squares Index M/E Party Membership as a Percentage of the Electorate MMM Mixed Member Majoritarian


12 MMP Mi xed Member Proportional MP Member of Parliament MSI Italian Socialist Movement NZ First New Zealand First Party NZES New Zealand Election Study PCI Italian Communist Party PD Italian Democratic Party PDS Democratic Party of the Left PLI Italian Liberal Par ty PPI PR Proportional Representation PRI Italian Republican Party PSDI Italian Social Democratic Party PSI Italian Socialist Party RAI Italian Radio and Audio RC Communist Refoundation RCES Royal Commission on the Electoral System R D Democratic Republicans RS Socialist Renewal SMD Single Member District SMP Single Member Plurality SNTV Single Non Transferable Vote STV Single transferable Vote United NZ United New Zealand Party


13 Abstract of Dissertation Presented to the Graduate Sch ool of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DO ELECTORAL SYSTEMS MATTER? STUDYING THE RELATIONSHIP BETWEEN ELECTORAL INSTITUTIONS AND SUPPORT IN ITALY AND NEW ZEALAND By Jeffrey S. Hamill May 2012 Chair: Bryon Moraski Cochair: Amie Kreppel Major: Political Science This work is a study of the relationship between elec toral systems and support. Support for governing institutions among populations is a very timely topi c of study. The decline in the level of support for governing institutions has been offered as a reason for the mass protests that have been increasingly prevalent world wide. I hold that an interesting and understudied area in political science is the s cholarly research linking electoral systems and support. I draw, in this work, from the literature on electoral systems and support competing theoretical models linking different electoral systems and support. The theory is that elections can influence th e decisions of voters and parties. That influence can help to determine the type of government s that are formed the way they operate, and the ability of those governments to pass policies. The way governments operate and the types of policies they produ ce can influence support. The two competing models I draw link both increased majoritarianism and increase proportionality to higher levels of support.


14 To test the competing models I offer New Zealand and Italy as excellent cases. New Zealand was a perfe ct example of a majoritarian system and changed its electoral system to a more proportional system. This presents a great opportunity to test the majoritarian model and the proportional model, while holding other political variables constant. Italy was a model of proportional government for a long period and attempted to increase majoritarianism through an electoral system change as well. They offer the same type of opportunity to test both the proportional and majoritarian models. Through a study of N ew Zealand and Italy before and after electoral system change, I hope to add the existing literature in several ways. Through the case study of New Zealand I believe we can learn how majoritarian electoral systems can work and also why they fail. In Ital y, the reason proportional systems succeed and fail can also be learned as well. Both countries offer the chance to study whether electoral system changes can be successful raising the level of support among citizens. Ultimately, I hope to discern whethe r a credible link between electoral systems and supportive attitudes can be made.


15 CHAPTER 1 STUDYING SUPPORT Introduction During the summer of 2011, riots broke out all over the country of England. This was less than a year after the sitting Labour Prim e Minister promised an immediate referendum on adopting an alternative vote electoral system after his government (Member of Parliament) In both 2009 and 2010 town meetings held by US House members were filled with the shouting of those opposed to the universal health care measure being debated in Congress. In 2010, the Republican Party swept the midterm elections, riding the wave of dissatisfaction with the Obama administration austerity measure s mandated by the European Union. While these events have been taking place, in other parts of the world citizens have risen up against authoritarian rulers. Widespread protests brought down a stagnant authoritarian regime in Egypt. Rebels in Libya mount ed a sustained campaign and toppled a dictator who was in power for thirty years Protests and political vi olence have been ongoing in Bah rain, Iran, Syria, and Iraq as well. Considering these events, it appears that there is world wide dissatisfaction for existing governing institutions. The relationship between citizens and their government, whether in a democracy or authori tarian government, is an important one. It is perhaps more important in a democratic government where polities are legitimized by the consent of the governed Support has been used as a dependent variable to explain why certain regimes maintain themselve s, are forced to change their organizational structures, or why they


16 are not able to make authoritative decisions. Considering this, and the examples cited above, support is one of the more interesting and debated variables in political science. While t h ere is plenty of reason to reconsider the concept of political support, the debate and interest have been ongoing in the American and comparative political s cience communities since the 195 0 s. During this time, American political leaders enjoyed high leve ls of support from the public. In the ensuing decades, Americans became increasingly dissatisfied, less trusting, and increasingly disengaged from the political process. This decline was often attributed to the perceived failures of Watergate and the war ehavior s. These types of behavior were especially evident among college students an d minorities who were lashing out over mistreatment in the inner cities. The existence of protests and riots as well as high disapproval led many in the political science community to believe that democracy in the United States was under a direct threat ( Miller 1974). This alarm was not isolated in the United States. Comparative political scientists noting these changes, began to argue that the decline in the United States was in fact part of a global decrease in support for democratic institutions in g eneral, one that has continued to present day ( Dalton 2004; Lipsett and Schneider 1983; Norris 1999 ) Crozier, Huntington, and Watanuki (1975) demonstrated that the growing gap between citizens and their government was not unique to the United States. We stern European nations and Japan experienced a simila s. Crozier, Huntington, and Watanuki raised the alarm that the democratic institutions


17 created in the 19 th century were unable to deal with 20 th century problems like the global recession and the oil shocks of the 1970 s. Protest behavior and riots were not isolated to the college campuses and inner cities of the United States. Major European cities like Paris and London also had severe responses to industrial problems an d their college campuses also had their share of radicalization (Norris 1999). These occurrences, as well as the happenings in the United States, led Crozier, Huntington, isis was based on the belief that democracy, especially the lively forms of democracy being practiced in the US and Western Europe, was incompatible with effective government (Brittan 1975; King 1975). Regardless of the beliefs that the democratic world was ready for a giant upheaval, democracy has persisted. Is it that that the declining levels of support had just been overstated or overemphasized? I do not believe that we should take these fact s or questions lightly. While, there is a rich research tra dition on support, the literature is fraught with disagreement and difficulty. Support as a concept has proved to be difficult to measure. For instance, researchers have faced the dual challenge of theorizing the elements of support and then finding spec ific survey questions that measure and match those theorizes. Also, support has not been as successful in explaining different political outputs. That is to say that support has not been very good at explaining, for example, why people engage in politica l protest behavior. There is a certain irony to this fact, since much of the literature on support was in reaction to the protests that were taking place in the United States and Europe in the 1960s and 1970s. Also, the decline in support in established democracies has not led to any sort of institutional change. This has led to the critique


18 of the support literature from scholars ( P r zeworski 1986 ) Resolving all of the theoretical, measurement, and explanatory problems that the support literature has is beyond the reach of this study. The goal of this project is to take a particular type of support, look at a particular object, only in particular contexts. By doing so I hope to make small improvements and heighten the understanding of political support. To be more specific in my intentions, the concepts that Chapter 1 will examine are the relationship betw een electoral systems and support. Electoral systems and their impact on political support present the most interesting, promising, and understudied elemen ts of the literature. I will suggest and provide evidence that simply studying established democrac ies is not the best strategy for understanding s upport. I contend that there is an understudied relationship between electoral systems and support and that advanced industrial democracies that have gone through an electoral system change present the most promising opportunity to study that relationship Chapter 1 is on the concept of support, and will pursue the following outline. After this brief introduction, I will define the concept of support, its components as well as the objects that support is dir ected toward. Next I will describe the difficulty and challenge that measuring support presents to researchers. Regardless of the challenges described and faced, I will describe why I believe support is a concept worth studying. Then, I will provide a l iterature review of the elements that affect political support, for example: individual and institutional factors. Next, I will focus in on a particular type of institution the electoral system. I will also describe why I feel scholars should pay partic ular attention to electoral systems and describe what I feel to be the ideal cases to study the relationship between electoral systems and support.


19 Defining Support Many definitions of support used in political science generally begin with the work of something by aid, countenance or adherence, the active promotion of the interests or cause o f an object; defending something as valid, right, just or authoritative, or giving it voting. When an individual votes for a party or individual, he or she is offerin g at least some type of supportive action. Support not only refers to the actions taken by an individual, but the underlying attitudes that precipitate that action. Take the example of voting. A person might vote for a candidate because they have some s ort of supportive say that A supports B either when A acts on behalf of B or when he orients himself (Easton 1965, 159). Support by citizens, whether actions or attitudes, may be thought of as an input into the political system. Ideally, t he government or elites try to foster such support by producing outputs that the voters or the public at large can evaluate. Outputs may include policy decisions and their implementation, public stances on issues, or constituency service. These outputs t government, and inputs into, the political system. Yet, support does not only exist in terms of an input/output orientation. David Easton (1965, 1975) defined two different types of support, specific and


20 evaluations of the political systems outputs. For example, a segment of the population advocates a given policy, and the legislature enacts that policy. These policies could be anything from a high er minimum wage to the end of a foreign war, or something as simple as the replacement of a broken street light. Diffuse support is something different. Rather than an evaluation of what an object does, diffuse support (Easton 1975, 444): Refers to evalua tions of what an object is or what it is or what it represents to the general meaning of what an object is or what it represents not that helps members accept or tolerate ou tputs to which they are opposed or the effects of which they see as damaging to their wants 1 Diffuse support is the collection of favorable or unfavorable attitudes that a country or regime has built up over time. To illustrate an example, a person may be completely dissatisfied with the performance of the government, while giving political leaders and parties failing grades for their handling of the economy or foreign policy, or more generally state that the country is heading down the wrong track. At the same time, the same individual might express pride in their country, their system of government, and not seek any fundamental change in the structures and processes of both. Specific and diffuse supports are, in most ways, separate phenomena. While, in most cases specific and diffuse supports are separate, there are instances where they are inter related. A political system relies on diffuse support during times of difficulty, especially when the performance of the government is viewed as sub par. For example, the government may ask for patience during times of 1 Easton is quoting himself from an earlier work (1965, 273)


21 attitudes have turned sour. Lack of specific support in the long run can carry over and impact the l egitimacy of whole institutional structures (diffuse support). While the erosion of diffuse support over time can happen if specific support is continually withheld, there remains s general belief among scholars that specific support is necessary if a par ticular government, administration, or incumbent is to stay in power. Diffuse support is necessary to uphold the entire s ystem of government (Dalton 2004 ). invaluable to th e topic, but they are by no means exhaustive or unchallenged. Many a starting point and that support itself has many components ( Klingeman 1999 ; Thomps on 1970; Weather ford 1987, 1992 ). For example, Weatherford (1987) suggested that diffuse and specific support merely represent end points on a continuum. Scholars have also noted that measuring the differences between specific and diffuse support is challenging, since m any of the indicators and measures that have been used to measure the two have proven to be highly correlated ( Kaase 1988 ; Loewenberg 1971 ). The difference between diffuse and specific support seems to make sense at the theoretical level, but they have pr oven difficult to measure empirically. Scholars have made repeated attempts to make them work at the empirical level and they have fail ed repeatedly (Kuetchler 1988). These difficulties make it important to spell out a precise and perhaps simple definit ion of support that deals with both specific and diffuse support while defining the term I define support as the individual belief that the government will produce good outcomes even if left untended (Gamson 1968; Heatherington 2005 ). I consider


22 support like Weatherford (1987) to be a continuum running from the most specific on one end to the most diffuse on the other. The definition I have given here has been used to identify spec ific support (Heatherington 2005 ), but like Weatherford (1987) and Easto n (1965) I believe that if citizens consistently lose specific support over time, or if there is a precipitous drop in specific support, then more diffus e attitudes will be impacted. Now that I have identified a working definition of support there is an im portant question to answer, support toward what? Objects of Support With a working definition of support it is important to answer the question: support for what? Support has to be given to something, so political scientists have created various analytic al frameworks for understanding the objects of political support. The first of these frameworks was presented by David Easton (1965, 1975) who identified the community, the regime, and the authorities as the objects of political support. This was an inte resting breakthrough because up until that point support was treated as an all encompassing enterprise. For instance support for a sitting politician and support for democratic values were treated in a similar manner although conceptually are clearly diff accepted them with two additions. Today five objects of support are generally accepted ranging from the most diffuse to the most concrete support for part icular politicians (N orris 1999, 10). The five are: the political community, regime principles, regime performance, regime institutions, and political actors. The political community refers broadly to any basic attachment to a community without consideration for its political leaders or institution. For example, a person may


23 characteristics of support for a political community. This type of support is particularly critical for democracies, since i t can be seen as a basis for the state. States are the first prerequisite for all democracies according to the oft 1 ) Political communities need not correspond along political lines (literall y, like a border dividing territory). Kurds in Iraq or Basques in Spain may express measures of support for a particular political community along ethnic lines which do not necessarily correspond with the political states in which they reside (Newton 1999 ). The following object of support originates from Easton who envisioned support for the regime (Easton 1965, 1975). This object has been subsequently divided into three parts. These three parts are: regime principles, regime performance, and regime inst itutions (Norris 1999). Regime principles represent the values of a political system. Rose, Shin, and Munro (1999) have person has about their political system. For democratic systems these might include fr eedom, participation, rational legal political rules, the rule of law, or tolerance and respect for alternative points of view (Beetham 1994) 2 These regime principles are separate since they represent what people envision their democratic system to be or what it should be. Regime performance and institutions tap support for the actual functioning and elements of a system. Regime performance gauges, not support for the values of democracy, but for the extent to which these values exist or are being deliver ed. It can be thought of as a 2 This is not to say that all conceptions of democracy are similar. Empirical studies have demonstrated e term itself comes with different understandings and there is a certain degree of fluidity in its definition, depending on who you ask in what place in what context (Norris 1999).


24 middle ground between lofty support for values and actual evaluations of the offices and institutions designed to deliver and guarantee these values. For example, in a democratic system, there may be support for and a belief in the value of fairness that every citizen should be treated the same. But, there may be a pervasive sense among others, and that elements of the political system are cl should be noted that I have not mentioned any particular institution. This is because the object (regime performance) taps what may be best described as an indistinct feeling about the system that remains formless in a because of this, is particularly arduous to measure, perhaps because it is so generalized. Since this is the case, scholars have had an especially difficult time locating methods to deal with this object (Fuchs 199 3 ; Klingemann 1999). Support for regime institutions is a bit more concrete since it examines what Rose ance, which is much a much more generally feeling of support regime instit utions lay ing blame, fault, or praise on specific institutions of government. These include executives, parliaments or legislatures, courts, military, police, or bureaucr acies (Lipset and Schneider 1983 ). To use and earlier example, an illustration of la ck of support for regime institutions would be the feeling that getting a fair trial requires money or connections, and that the courts themselves were slanted in favor of the rich. This is a much more specific grievance. Those who display less than supp ortive attitudes have a


25 toward regime institutions. What is interesting is that ther e is a separation from the actual individuals who make up that institution. Regime institutions tap a more generalized feeling than any of those that are directed toward a particular individual or particular institution. The objects of support that taps particular individuals or institutions are the political actors. Support for political actors refers to specific individuals or institutions. For example, people may not support the policies of President Barak Obama, the majority party in parliament, or t he rulings of a particular high court while at the same time believe that the president, parliament, or high court are the best institutions to uphold and defend their beliefs or values. A loss of specific support for a particular president does not endan ger the entire executive institution. Objects of political support can help scholars direct their research toward particular avenues. Studying support as a whole is a monumental undertaking, but by identifying particular objects of study the task become s somewhat more manageable. By classifying particular objects, scholars also have an instrument for categorizing the motivations for studying support at all. Now that I have given a working definition of support and what it is directed toward, it is the appropriate time to explain why support is a crucial variable that warrants study. To foreshadow, declining levels of support (whether trust or satisfaction with democracy) may not lead to the complete collapse of a political community or regime. T his is the high standard that P r zeworski (1986) sets. When only this level of change is satisfactory to scholars then it is clear that support has no real value as an explanatory variable. My contention is that declining levels of support need not lead to an im plosion


26 of democracy or a union, but it c ertainly can lead to intra regime changes. The inter regime change that I will focus on is electoral system change. Why Support Matters Having already defined support and described what support is directed toward ( objects of support) it is time to emphasize why scholars should care about support. There is a question that I ask ed myself before undertaking this c an not be used to explain other phenomenon, in my mind, it is not a crucial variable. That is not to say that purely academic pursuits aimed at general knowledge for its own sake are not important. But, I believe that if a dependent variable can be used to account for other occurrences, then its magnitude increases exponentially. To use support as an example, this section will seek to explain why support matters in a larger context. What consequences does support for or lack of support for an object hav e? In my estimation support has a bearing on matters of critical import to political science. One example of the importance of political support is its relationship to the survival of a political community or a regime. At the most extreme and obvious lev els are the creation and dissolution of states. While we might think of states (or political communities) as fixed entities, viewed over a long period of time states form, fall apart, and new states emerge. Support for political communities can be mainta ined, but even the most esteemed communities can rise and fall. Examples abound: the Roman Empire, the Pharaohs of Egypt, the Ottoman, Hapsburg, and Prussian Empires (Rose 2008: 5). More recent examples include the fracturing of Yugoslavia and the Soviet Union into multiple independent states. Most independence movements, the Tibetans, the Palestinians, the Kurds, Baathists, or Quebecois, exist within states but lack some


27 kind of support for the political community as a whole. The assertions that the s tability of political communities relies on support and those communities that can no longer cultivate support are doomed for dissolution presents problems that should interest researchers. I will explain i n the conclusion why I believe the political comm unity is perhaps not the most fruitful object for studying support. States (political communities) need not dissolve when support is lost. Even if a namely those that changes are much more frequent than whole sale dissolution of the political community. For example, political communities in Latin America have remained stable over time, but have ha d frequent changes in political regime between military dictatorships, authoritarian rulers, and popularly elected democratic civilian leaders. Among European nations Greece, Spain, and Portugal abandoned democratic governments for dictatorships. Germany Austria, Italy, and Finland changed regimes as well (albeit in the opposite direction) after WWII. When the Soviet Union collapsed, all of the political communities in Eastern Europe changed political regimes as well. To some extent regimes survive onl y when they are able to engender support from its people. Endurance of institutions depends on compliant behavior from the population (Rose 1969, 2008). There is evidence within the scholarly community that focuses on the relationship between support and different types of political behavior that might function to strengthen or weaken regimes. Behaviors of this type are usually placed into one of two groups of political behavior: conventional or unconvention al ( Coleman 1976; Craig


28 1980 1993 ; Easton 197 5; Gamson 1968; Kornberg and Clarke 1992; Norris 1999 Norris 2002 ; Rose, Shin, and Munro 1999 ). Among these scholars there is a widespread belief that those who are generally supportive of a political regime will engage in more conventional types of pol itical behavior. Conventional types of political behavior include acts within existing institutional rules and norms. All aspects of choosing a political party and voting would fall under this rubric. This process includes multiple elements. Theoretica lly a citizen would have to display an interest in the process and educate themselves as to who the parties are and what their platforms include. He then would have to actually turn out to vote, and follow and evaluate the extent to which the parties or c andidates do their job and keep their promises. Other types of behaviors might involve attending a rally or seeking out a representative for a face to face meeting or something as simple as writing a letter to a government office. All of these behaviors, because they operate within existing rules and system norms, help to buttress the very institutions that establish those rules and norms (Booth and Seligson 2009). If a collection of support can lead to conventional political behavior, then the opposite is also generally believed to be true. When support for regime institutions or political authorities fall, those holding negative evaluations are likely to engage in unconventional political behavior. This behavior can range from the least extreme, like acts of protest, to regime challenging acts, like failure to pay taxes, political violence, or other disregard for the rule of law. Some political scientists in America fixated on this hypothesis in the 1960 s and a few declared that the protest and anti government behavior that student demonstrations represented, urban disquiet, and the anti war


29 movement were the result of a loss of support for democratic institutions and that the country was perhaps on the verge of a revolution (Miller 1974a, 1974b). Ho wever, American scholars were unable to forge a credible link between declining levels of support and unconventional political behavior (Citron 1974; Craig 1980, 1990, 1991). This hypothesis stil l remains as Pippa Norris (1999, y believed comparative work done in Latin America in the 1990s and early turn of the 21 st century (Booth 1991 2006; Canache 2002; Foley 1996 ). These studies demonst rate some links between low levels of support and protest activity in Venezuela, Nicaragua, and El Salvador. While, t here is disagreement, what is clear is that the idea that political activity, whether conventional or unconventional, is linked conceptual ly if not empirically with levels of political support. Studying support is not only important in explaining different types of political behaviors, it is also important to the functioning of democracy itself. Political support is instrumental in the func tion of democracy in two main ways. It plays a vital role in the connection between citizens and their el ected officials one functional and the other more normative. The functional relationship relates to the extent to which citizens are able to rely on their authorities for information. The practice of democracy, the institutional mechanisms, and policy evaluations are all extremely difficult to process. Attempting to synthesize the amount of information is difficult even for the most highly trained p olitical scientist or economist. Average citizens who have their own day to day responsibilities and have neither the time nor experience to make these calculations instead rely on heuristics, shortcuts, and cues to form opinions about


30 politics and policy (Downs 1957 ; Popkin 1991 ). Some of these cues come from political parties, authority figures, or trusted news sources. If these relationships are broken, due to loss of support, then a critical connective component of democracy is lost. Pervasive attit how can we believe any of the information that we receive? Another functional reason for why we s hould support has to do with the recommendations that scholars make with regard to different institutions. We have a tendency to recommend constitutional orders (Lijphart 1984 199 4 ; Linz 1990 ; Powell 2000 ) or electoral systems ( Amy 1993; Lijp hart and Gro fman 1986 ; Reilly 2002) without real, concrete evidence with how citizens might react in terms of supportive or unsupportive attitudes and behaviors There is evidence that institutions impact support (an idea I return to later), but little evidence that citizens are more supportive of one institutional arrangement or another. Scholars make recommendations despite this lack of evidence, and normatively we probably should rethink that practice unless we find some link to more supportive attitudes. Finally there is another normative reason for why we should care about whether citizens support their democratic institutions, because we should. Scholars can quote reasons for why support matters, and I have just listed several but I would like to set those as consensus based on what citizens think and want. Citizens give democratic institutions their legitimacy and scholars should try to find out why. When large numbers of cit izens do not support their democratic institutions or political authorities then the entire


31 process itself is flawed 3 Many have taken up this call ( Booth and Seligson 2009 ; Dalton 2004; Norris 1999 ). It is the reason that I have chosen to study support and its relationship to specific democratic institutions, and it is with that spirit that I tackle this project. Support, to summarize, is a valuable topic of study. To this point, I have given a definition, supplied the objects support is directed toward and explained why I believe that the study of support has value. Any study of this nature also has to explain where support comes from. What are its bases? To determine whether certain electoral institutions affect support, a determination has to be m ade where else support may originate. Only then can proper control variables be identified and proper cases for study chosen. The Basis of Political Support Political scientists have identified many reasons why a person may choose to support or fail to s upport their political community, political institutions or leaders. Although, numerous reasons have been suggested, all can be placed into one of two broad categories. Weatherford (1992, the different levels of support for and participation in the political system. Almond and The Civic Culture (1963) would be an example of this particular level of an institutions that are in place and their relationship with political support. Put very 3 Not all scholars agree with this assessment. Some have said that democra cy and government is created to give citizens what they need and not what they want and that a healthy skepticism among citizens is a good thing (Dalton 2004). There is an argument to be made here it is one that I happen to disagree with however. Perhaps it is more of a normative point that I make.


32 generally, it is thought that those political institutions which engender higher amounts of particip ation, representation, and accountability will lead to higher levels of political support (Dahl 1971; Lijphart 1984). How citizens think about and behave toward the political system can be measured at the individual level. People di ffer in many respects: age, gender, race, occupation, education, and interest. All of these characteristics vary to a fairly large degree in different democratic societies. It is generally thought that these different sociodemographic characteristics can the political system. Generally speaking, scholars have demonstrated that those who are more involved in the political system (both psychologically and physically) are more likely to express higher level s of support for the system (Anderson and Gui llory 1997; Anderson et al 2005, 20). Early voting studies in the United States were the first to use social characteristics to explain political behavior. For example, the Columbia and Michigan models diffe red over how determinative social characteristics were to political behavior or the The Civic Culture (1963), was the first comparative study to argue that those with higher levels of education and i institutions 4 attitudinal correlates of system support has established that those with greater interest, 4 Needless to say, this notion has been challenged in recent literature, namely Dalton (2004) who argues that those who are more educated and involved in the political process are more likely to be demanding an d more cynical than other citizens.


33 i nvestment, and involvement in the existing political system are more supporti (Anderson et al 2005, 20) This is hardly surprising and frankly not particularly interesting. The more interesting question is how supportive attitudes develop and in which political or institutional contexts. of the general demographic variables like gender, race, education, or income. Interest general of these political variables, but there are other more specific and equally These include the effect that partisanship, perceived policy outcomes, and views about the political process have on political support. Partisanship refers to the emotional attachment an individual has toward a political party. Either the result of early socialization (Campbell et al 1960) or a heuristic based partisanship is a popular explanation for differing levels of political support. If an individual is attached to a political party they not on ly are more likely to vote for that party, but attachment may be related to how they process new information. Partisanship is a lens through which many view the world. The political arena includes a tremendous amount of information. People use their att achment with a political party as a mechanism for reducing information costs, in other words, it is used as a short cut (Downs 1957 ; Popkin 1991 ). Acquiring information about issues and policies is a costly endeavor in terms of time, and people (even poli tical scientists) have a difficult time processing all of the information that is available. People simply do not go through this


34 process, and scholarship has demonstrated that individuals who identify with a political party are more likely to adopt that on issues or policies (Campbell et al 1960; Jackson 1975; Jacoby 1988). Because individual attachment with a political party has such an influence about how individuals view policies, issues, and candidates, scholars have used partisanship as a variable to explain differing levels of support for the political system. Strong partisan attachments may be an indication that a person is engaged in and believes that the political process is valuable. Partisans clearly supp ort their political party, and by proxy may support the political system of which their party is an active and important piece. Political research has demonstrated that attachment to political parties is more likely to support the political system than th ose without partisan attachments (Dennis 1966; Holmberg 2003 ; Miller and List h aug 1999 ). Variables like partisanship, interest in politics, income, etc. have all been demonstrated to have some relationship to the variable political support, but they are by no means an exhaustive list. People interact with and learn about their government through various forms of institutions. We do not learn and participate in a vacuum. We do so in any number of different institutional structures. These may include th e type of government (democracy/dictatorship), the economic conditions (boom/bust period), the type of democratic rules (majoritiarian/consensual), along with any number of factors that could impact the ways in which we develop expectations for the governm ent. In influence how a citizen views and then develops an opinion about their government.


35 support generally fall into four categories: policy, process, institutions, and socialization. Approaches that support based on the effectiveness of these policies or th e extent to which these policies match citizen preferences. Process based studies minimize the importance of policy outputs. Instead they focus on the manner in which policies are negotiated and that political actors settle upon. Institutional approaches look at the various types of constitutional, legal, or electoral arrangements and whether or not these differing compositions effects support in distinctive ways. Finally, socialization examines the ways individuals learn about politics and their politic al institutions, both as children and throughout their lives as adults. This learning process has a bearing on the level of the relationship to policy outputs and polit ical support. Those studies that view policy as a means for explaining the level of citizen support base their beliefs in a relatively simple way. Governments are evaluated based (Easton 1965). The relationship between policy outputs and citizen support is theoretically straight forward. It makes perfect sense that those who are dissatisfied with the policy performance of the government would be less likely to support that same government. This has certainly been the belief of multiple political scientists ( C itri n 1974 ; Easton 1965, 1975; Gamson 1968; Mi ller 1974) Jack Citrin points d like those who


36 73) In fact a large number of scholars have claimed, in one form or another, that dissatisfaction with the policy outputs of governments coupled with the increasing demands of citizens has lead to a world wide decline in support for democratic institutions ( Craig 1991 ; Crozier, Huntington, and Watanuki 1975; Lipset and Schneider 1983 ). The relationship between policy and support may appear to be clear cut, but there are several complications when scholars have attempted to measure this relationship. The policy making process is often times a convoluted one. Scholars and experts have a difficult time evaluating the success and impact of particular policies. It is an even more daunting task for average citizens to evaluate the detail of policies. In fact most neither have any kind of detailed knowledge about po licy (Delcarpini and Keeter 1996 ), nor deep seated policy beliefs. Perhaps because of this, scholars have focused on the presumed secondary effects that good government policy might have. Studies have looked at characteristics like GDP, inflation, and unemployment and their relationship to support. The presumption is that good policies will produce higher GDP, lower inflation, and lower unemployment and when that happens, citizens will support th eir government at a higher clip. Various studies have taken up the impact of these indicators as well as individual level retrospective and prospective evaluations of the economy on supportive attitudes (Clarke, Dutt, and Ko rnberg 1993; Kornberg and Clark e 1992; Weatherford 1984, 1991, 1992). Opinion and studies have been divided on these points. Lipset and Schneider (1983) argue that there is evidence of a connection, while Miller and Listhaug (1999)


37 believe that there is not strong evidence of an asso ciation. Economic indicators and feelings about the economy seem to be a poor fit in several countries. After WWII, Japan and Italy experienced robust economic growth as well as an increase in political skepticism ( Morlino and Tarchi 1996 ; Pharr 1997 ). Studies have demonstrated that robust macro economic growth during the 1980s did nothing to impact the low levels of support in the United States (Lawrence 1997). Perhaps because of these conflicting accounts and the difficulty ordinary citizens have with recognizing and evaluating policy, scholars have looked for other explanations for varying levels of support. Miller and Listaug (1999) argue that economic evaluations need to be substituted for citizen evaluations of policy outcome fairness as well as t heir expectations of the government and their elected officials. While these suggestions are warranted, the data are lacking and difficult to gather (Norris 1999). Process based accounts of system support have been developed perhaps as a result. Citizens may not be deeply concerned with policies or economic indicators, so some scholars have attempted to dete rmine what they are concerned with Americans in particular are attuned to the way in which the government works rather than what it produces. Accor ding to Hibbing and Theisse Morse (2002) process trumps policy, People usually do not know what would change, say, drug and education policies for the better, but they do know that special interests, entrenched political parties, and careerist polit icians have a stranglehold on the political process. This sentiment is straight forward and manageable. It requires no mastery of the nuances of parliamentary procedure. In fact, it requires remarkably little information at all. (39) Americans are parti cularly adept at making accurate conclusions about their political leaders. They are not shy about declaring that the system is geared toward special interests, politicians are engrossed with partisan concerns, and that the agenda


38 is controlled without or dinary people in mind. Because most view this not to be the case, support in the United States has fallen preci pitously since the 1960 s. Americans do not advocate specific policies they only want their elected officials to compromise and treat each other in a civil manner while making decisions with average Americans as their primary worry. These conclusions are supported by several scholar s (Craig 1993, 1996; Popkin 1991 ). Process approaches are interesting, but seem to be limited to the United States where extensive interviewing and focus groups have taken place Any study of support should consider process as a key explanatory variable. Institutional appro aches consider both policy and process because institutions might help shape both. Institutiona l approaches have a long tradition in political science. Institutionalists believe that formal and informal institutional rules constrain our actions and decisions (March and Olsen 1984 ; North 1981 ). The most basic constraints on political behavior are c onstitutions. Constitutions define the range of political choice available to citizens. They also shape and delineate the political structures in which we develop and learn about politics. Anderson et al. (2005 ) describe their impact and importance to c omparative political science, political institutions at least partially determines how and what people think about politics. Politics and political contexts vary widely across i ndividual countries and types of political systems. As a result, party systems, political events, and power relationships, to name just a few examples, differ across countries and are themselves occasionally subject to change over time. Given that instit utional structures and political contexts vary across democratic systems, it is reasonable to conjecture that what and how people think about politics is affected by political institutions and varies across contexts as well. (21)


39 The impacts of different p olitical institutions are well established in the field of comparative politics. For instance, there have been multiple studies on the effectiveness of presidential versus parliamentary systems (Linz 1990 ). Also, studies have examined the differences in levels of support between authoritarian and democratic gover nments ( Geddes and Zaller 1989 ). In short, different political institutions help to produce different political attitudes across different contexts. This makes the connection between institution s and attitudes ripe for comparative study. Perhaps the foremost example of institutional study and its link to support is the work focused on the political consequences of electoral systems. Since they determine who gets to run and how winners are select ed, electoral systems are seen as the Electoral rules have wide ranging impacts including the number of political parties in a system, electoral volatility, candidate or party entry into elections, and the type of policies that are argued and adopted in the political a rena ( Lijphart 1984, 1994; Powell 2000 ; Rae 1967; Riker 1976 ). Even though there is a substantial body of literature dedicated to the effects of political institutions like electoral systems on political behavior, the link between institutions and political support has not spurred an equally rich research agenda (Anderson et al 2005, 22). This agenda has in fact been pursued, but in an insufficient manner. For example, Anderson and Guillory (1997) classified multiple countries in Western Europe by other factors (many of the variables I have mentioned previously), the work a ttempts to determine some differences between majoritarian and consensual systems. The study


40 suggests that the gap between winners and losers in an election is narrower, in terms of how satisfied respondents were with how democracy worked in their country in consensual systems than in majoritarian systems. Anderson (1998) built on this study by suggesting that overall levels of satisfaction with democracy were higher in consensual countries than in majoritarian countries. Later, the authors along with o thers return to the winner/loser gap in a more complete fashion ( Anderson et al. 2005) again concluding that the gap is smaller in consensual than majoritarian nations. There are several theoretical reasons for w hy this occurs. Scholars both accept and u se policy, process, and institutional bases for support. But these sources are not the immediate link in the causal chain. Even if policies in a given country improved drastically, this may not automatically lead to a rise in supportive feelings for poli cymakers. It is not a functional process between any of these variables and support where an improvement in one will necessarily lead to an improvement in supportive attitudes. What is obligatory is for citizens to learn about the changes, incorporated t hem into their attitudes and opinions, and change their supportive attitudes in accordance with modification in policy, process, or institutions. This learning process is known as socialization. Socialization is the process by which we learn about politic s and can be divided into two stages: childhood and lifetime learning. Childhood socialization occurs when a child tries to make sense of a confusing mass of information. In order to do so they compartmentalize and place information into some kind of ord er. Once this is done the conclusions that are drawn are difficult to change. Since, most members of the general public do not spend a considerable time thinking about politics, the tendency is to


41 accept elite leadership and government authori ty (Abramso n and Inglehart 1970, 421) 5 If all of what we learned about politics took place in during childhood then studying support would be a fruitless enterprise, since our levels of support would not change. Since I assume they do, socialization must continue into adulthood. This socialization that continues into adulthood is known as a lifetime learning model. This model summarizes the effects of regime support as the accumulation of old and new experiences (Mischler and Rose 2002; Rose and McAllister 1990 ; R o se, Mischler, and Haerpler 1998, 117). The idea that people can learn about politics over time makes socialization particularly relevant to the variable s that I have already discussed: policy, process, and institutions. For example, policies may change but up. Also, if a country were to change their institutions would we truly expect an immediate change in attitudes or an extended learning curve? Obviously, there woul d be a process of socialization that would take place. This makes socialization a particularly sticky variable that political scientists have to navigate. Any comprehensive and complete study of support must recognize these sources that I have just identi fied. It is not a matter of defining and identifying for the edification of doing so. The sources of support should be used to direct ou r future research on the subject. There is a great deal to be learned by doing so, but scholars have not done so adeq uately. By examining the gaps in literature that attempts to understand the sources of support, we can identify what scholars have missed. In short, we can learn where 5 Children may also associate executive leaders like the president with parental authority figures. They may also be compelled to support political institutions through positive or negative reinforcement. See also Ea ston (1965, 1975).


42 and how to study support. Through this process I will suggest a research program cent ered on the discovering and testing the theoretical link between support and electoral systems. Moving Forward: Where and What We Should Study The question that now presents itself is where and how scholars direct their research on support. I will present an argument that the most promising areas of study are in countries that have exp erienced some type of change in regime. More specifically, established democracies that have a long period of consolidated democracy, but whose electoral system was changed. I will offer New Zealand and Italy as example of such cases. Which Object? Research to date on support has taken place in what we might call Western developed nations. Examples of this type of research abound ( Anderson et al 2005; Anderson and Guillo ry 1997 ; D alton 2004; Norris 1999 ). The study of established democracies like the US, France, and England are prime cases for study because of their well established political science communities wh o have the resources and expertise to develop better surv ey methods and as a result have a rich history from which to draw. But research in these areas presents two problems: the problem of 6 As I stated earlier socialization has an impact on the level of s upport a person has. This presents a problem in older regimes because support is usually not in dispute. 6 there has been an absence of wide spread protest or political change. If support is to have any value as an explanatory variabl e, it is important to link the variability to political outcomes. Since democracy and democracies is problematic because of their continuity.


43 Everyone in these regimes learns to give their support through a continuous process of socialization 7 This begins in childhood and as parents, scho ols, peers, and the media continually bombard individuals with the hegemonic political values of the society adult, they will perhaps view the regime as the only country or government feasible (Rose 2008 6). Individuals can give unblinking support to a regime if they have been socialized to regard it as the only way in which the country could be governed. It is only half correct to say that these values are what citizen s want government to do for what people want is often what they have learned is normal in their society. We can explain support for a steady state regime as a consequence of the transmission of political values from one generation to the next over a centu ry or more. While the values and norms learned early in life may be modified according to their changing circumstances, this will not alter the underlying consensu s of the regime. (Eckstein 1988, 20) This is not to say that support does not rise or fall i n established democracies. But, what remains is the endurance of the political community and the regime. This is what I have labeled the continuity problem. revolution, civil war o r the l 1) As I stated earlier, the support is a key concept because polit ical scientists believe that it is a requirement for system stability Simple observation tells us that even in the face of acros s the board decline s in support Western democracies continue to chug along. In fact, none of these nations is even close to collapse and almost universally at least until very recent events, there is a lack of widespread pu blic outrage or protest (Booth and Se ligson 7 Thi s is not to say that there are not ruptures, but Rose (2008) only deals with socialization as a continuous process. I suppose when ruptures occur, a new process of socialization would begin.


44 2009). It may be that these countries have reached a level of economic success where their citizens are merely comfortable enough to grumble about the political system, but comment abo ut seccession aside, does any serious scholar actually believe that anger over health care legislation will lead to the Independent State of Texas? The instances of declining support without major consequence could lead one to ju In fact, some scholars have done just that. Adam Pr zeworski (1986 50 53) has gone as far as dismissing the value of gy; if a regime survives it has support, if it does not it is due to a lack of support. I am not ready to abandon the concept of political support, but I am worried about the tendency to study industrialized countries simply because they are rich data so urces. Doing so is similar to dropping your keys in a darkened area of your porch, only to walk across the street and look for them under a street lamp because that is where the light is (Booth and Seligson 2009) Scholars should accept the fact that som e industrialized nations have practiced democracy for centuries (Przewo ski 1986) and many for close to 50 years (Markoff 1995). These countries are also among the most wealthy in the world and are, as research has shown, so far above the minimum condition s for what democracy needs to develop that they are simple not at serious risk of b reaking down ( P r zeworski, Stokes, and Manin 1999 ; Seligson 1987 ). The solution appears clear, examine countries where democracy is new and very well could break down. Some examples would be post communist states, African


45 recent, economic conditions are far less favorable, and civil wars, insurrection, guerrilla warfare, and military coups h ave all played a role in recent histo and Seligson 2009, 222) The trouble with this type of research, to use my earlier metaphor, the keys are still in the dark. In other words, the scientifically conducted surveys done in these areas are frau ght with issues related to reliability, validity, and quality (Seligson 2005). The surveys that have been deemed reliable are either held in propriety by senior scholars or have required a massive grant fueled effort 8 Where does this leave scholars who w ant to study support? My suggestion is that we merely change the objects on which we would like to focus. It is absolutely correct to say that the likelihood of an established democracy like th e United States or England either falling apart into separate political communities or abandoning their democracies drastic examples. There are ways in which democracies, even established democracies, might change their political re gimes that does not require a wholesale abandoning of democracy. Much of Przeworsk losses of support will cause democracies to collapse, but other changes are certainly much more feasible. A much more likely result i s a country going through some kind of intra regime change. I suggest examining countries that have changed particular parts of their regime institutions. This could include the emergence of a third or fourth party in a traditionally two party system, who lesale changes to a constitutional order, or the scrapping of an 8 See Rose (2007) and Booth and Seligon (2009)


46 electoral system for replacement with another. These would be considered changes within a political regime rather than between political regimes. I contend that in established democratic sy stems, inter regime change is much more likely. P r zeworski and the other scholars I mentioned are perhaps setting the bar too high when evaluating the usefulness of variables measuring support. Countries need not abandon their political systems as a resu lt of a loss of support, and they certainly do not require large scale secession either. Moving forward, I believe there is ample reason to study the effects that electoral systems have on support. Electoral System Change and Support Electoral systems are an interesting object of study for several reasons. The first is that there is already a substantial body of literature on the political effects that different electoral systems can have. The second is there are cases of electoral system change in sever al established democratic regimes that present researchers with a laboratory type setting for which to stud y support In these cases of electoral system change it is possible to look at how support (or a lack thereof) precipitated the change and the exten t to which the change itself had on supportive attitudes. In the preceding paragraph I mentioned that there is a substantial body of literature that studies the effects of electoral systems on a host of political variables. It has been long understood t hat electoral systems define the context in which citizens participate in elections and help to condition their behavior (Duverger 1954). Numerous studies have found that there are systematic differences in election results due to different electoral rule s. Some of these differences include the number of political parties, the success and failure of certain kinds of parties and candidates, the level of participation, the types of policies that are enacted, the level of electoral volatility, and


47 other poli tical effects ( Lijphart 1984, 1994 ; Powell 1980 ; Rae 1971 ; Riker 1976 ). What is also interesting is that scholars of electoral systems have theorized that different electoral systems have an impact on the quality of policy and that different electoral sys tems produce different types of process. These are two crucial sources of support according to the literature that I cited early. It is not surprising then that some scholars have begun to examine the role that electoral systems play in differing levels of system support. The studies that are most often cited in this respect are Anderson and Guillory (1997) and Anderson (1998). These studies have moved the research program linking electoral systems to support in admirable ways. While they are praisewort hy, the studies themselves are lacking in several respects. Anderson and Guillory (1997) typology) and the level of satisfaction with democracy with democracy between winners an d losers in different European nations. The primary problem with this study is that it is done almost with a complete lack of theorizing. What is the link between certain types of constitutional orders and support? In their study, this question is surpr isingly left unanswered. The second problem has to do with where the study takes place. The cases are Western European nations, which suffer from both the problems of socialization and continuity. A later study done by these authors (Anderson et al. 200 5 123 ) corrects the problem of theorizing, but limiting the theory to the difference of support between winners and losers while stating specifically they are not making any claims between system support and overall levels of support This is the importa nt question I wish to tackle.


48 One study that indeed does make the claim that certain systems lead to higher corrects some of the problems that I have pointed out. Namely he includes in his study post communist democracies and he attempts to make a theoretical link between electoral systems and overall levels of support. He theorizes (1998, 575) Logically, the establishment of electoral procedures and institutions precedes the measurement of support for the system, this assumes a chain of causality running from institutions, party performance, and party system performance, on the one hand and satisfaction with democracy on the other earch electoral rules should result in particular kinds of party and party system performance, which in turn should affect citizen satisfaction with democratic governance. System forward but not a sufficient one in my estimation. I will detail in Chapter 2 a model that more accurately reflects a multi faceted relationship between electoral systems and suppor t. Does he mean that the process is better under certain electoral systems, or that they produce better policies? It is not clearly stated, so the theory needs improvement. Put simply I believe that Anderson is m issing a few steps and understand ing the gulf between the two variables requires a much more rigorous examination. results into question. Anderson attempts to draw correlations between aggregated scores of satisfacti on with democracy with levels of proportionality in those countries. In other words, he has system level independent variables (proportionality, effective number of political parties) so he aggregates his dependent variable (satisfaction with democracy). The trouble with this is that you are not able to draw conclusions about the


49 individual level effects of electoral systems from aggregated national attitudes. Since this is exactly the method that Anderson uses, the ecological fallacy 9 constantly hangs o ver his work. It is also a strange choice to use national level data to study satisfaction with democracy, which is an individual level measure. To truly understand the relationship between satisfaction with democracy (or trust) an individual level appro ach is necessary. I have identified two problems with the existing work on electoral systems and support. They are the reliance on established democracies and the lack of comprehensive theoriz ing. I will explain in the Chapter 2 what I believe to be the most accurate description of the relationship between electoral system and support. To correct the problems of examining the established democracies, I suggest a study of two countries that have changed their electoral systems: New Zealand and Italy. Thes e countries make interesting cases for several reasons. The problems with studying advanced industrial democracies, as I have mentioned, has to do with socialization. So, ideally researchers would look for cases with, as James Madison looking for cases without a history to fall back on or no history from which to draw for lessons. New Zealand and Italy present the interesting opportunity to study developed countries that both have a want of precedents and a rich history of data from which to tap. New Zealand, for example, had an uninterrupted period of 100 years during which 9 The ecological fallacy is an error in the interpretation of statistics. It occurs when inferences about the nature of individuals are drawn from aggregate statistics collected for the group to which those individuals belong. This error assumes that all members of a grou p exhibit characteristics of the group at large (Robinson 1950).


50 two political parties dominated. During this time one party would win a majority of seats and gov ern until the next election. Then in the period between 1993 and 1996, after the reform, they became a multi party system. With this multi party system the voters were introduced to the messy process of coalition building. New Zealand had a particularly confusing and frustrating coalition negotiation after the election of 1996, which took place in secret between National, Labour, and NZ First. Most citizens expected Labour (who won the most seats, but not a governing majority) to be in the coalition, bu t instead a coalition was formed between National and NZ First. Satisfaction with democracy fell to new lows in the country in 1997 (Karp and Banducci 1998 ). Italy also had a difficult time after reform. The reforms taken here were enacted against a ba ckdrop of scandal, as a result of which the predominant party for fifty years (Christian Democrats) was no longer part of the party system. Initially, voters did not know how to react and the party system actually fragmented to a greater extent under new, supposedly majoritarian rules (Sanchez 2002 ). It could be argued that New Zealand and Italy lost their institutional memories, and after reforms were forced to begin creating a new one. These countries present a setting that is fairly close to a laborat ory, and no study has taken advantage by comparing these two countries in conjunction with one another to determine the affect their different electoral systems had on support. Now that I have established where we should study, what remains is to determin e how support should be measured. Measuring Support The preceding sections define support, the objects it is directed toward, what impacts support, and why it matters for political science. What remains is to determine an appropriate and effect ive means t o measure support. In this study, I take a multi


51 faceted approach to measuring support. I consider popular survey items: trust and satisfaction with democracy, which measure support in different ways. It is my belief that the questions related to trust are an accurate measure of specific support, while satisfaction with democracy measures an element of support that falls between the two types of support Easton theorized (specific and diffuse). I also use various indicators that measure participation lik e voter turnout and party membership. The use of these variables paints a more complete picture of support in the cases of interest. There is disagreement over the meaning of trust and satisfaction with democracy, so it is best to address these issues be fore moving forward. This is followed by an explanation of why participatory variables also help to capture support. Political Trust Trust as a political variable has been consistently measured with some form of the popular National Elections Study survey were early questions about whether this question measured diffuse or specific support. Easton (1965, 1975) and Gamson (1968) imag ined that trust was a more durable form declining levels of trust in their elected officials signified a loss of diffuse support for American institutions, and that the cou ntry was at the verge of revolutionary change. Citrin (1974) disagreed. He believed that the trust measures and the responses citizens gave were simply indicators of citizen mistrust of specific political leaders (specific support) rather than a deep seat or political institutions (diffuse support). Responses to the trust survey index may have suggested that just because people might say that their government wasted money,


52 thought their government wasted mone sale replacement of their constitutional order (Citrin 1974, 975). Subsequent studies undertaken in European nations have asked about trust in a variety of different 10 Consistently, these studies have s hown that trust, in whatever form they are asked, is a measure of specific support (Craig 1980, Craig 1993; Craig, Niemi, and Mattei 1991; Craig, Niemi, and Silver 1990 ; Heatherington 2005). Easton (1975) generally disagreed with this assessment and argue d that trust was an indicator of diffuse support, and that the survey measures did not accurately tap into what he was describing. Craig (1991) correctly points out that this argument is tautological 11 In this study I will agree with the prevailing schol arship and treat trust as a measure of specific support. This fact need not hinder the use of the trust variable in a study such as mine. I will consider it as a measure of specific support in a given year. For instance, in New Zealand in 1993 a score of trust will be considered a measure of support for particular political circumstances in that year. I will do so while remaining mindful of the political context over multiple years. So, the study of New Zealand will look a trust not just in a 10 For a complete survey see Dalton (2004) and Holmberg (1999) 11 The argument that Easton offered is a tautology because he continually declared that trust was a measure of diffuse support reg ardless of the continued efforts to measure trust as such. Scholars attempted repeatedly to measure trust and their finding continually demonstrated a high correlation between trust and specific support. Easton suggested that any study that showed this p receding correlation was simply not accurately measuring trust. His contention that trust was diffuse support regardless of the evidence suggested makes it a tautology.


53 single yea r but over a 15 year period, which should provide a larger picture on the level of more diffuse attitudes of support. A study of a variable measuring specific support over time should give the researcher some hint as to the diffuse attitudes in the countr y. Craig 1993 ; Easton 1965, 1975 ). However, I believe that one measure of trust is not enough, but it should be studied in conjunction with another measure of support: satisfaction w ith democracy. Satisfaction with Democracy Satisfaction with democracy is a good general measure of system support for several reasons 12 28 0) and refers to the functioning of a democracy as a set of no rms (Anderson and Guillory 1997, 70). The most complete description of the variable was made by Kuechler who notes that, Satisfaction with democracy measures an evaluative rather than a purely emotional response, thus it measures neither, diffuse or specific support in the Eastonian sense; but rather a form of support not recognized or inadequately conceptualized by Easton (1991 279 ) Kornberg and Clarke n. They find evidence that satisfaction with democracy is correlated with both support for the regime and incumbent authorities. In later work Clarke, Dutt, and Kornberg (1993) discover that satisfaction with democracy is correlated equally with support for incumbent authorities, the regime, and the political community. Because of this, they consider satisfaction with 12 Many studies have used satisfaction with democracy as a dependent variable. These i nclude: Clark, Dutt, and Kornberg (1993), Fuchs (1993), Harmel and Robertson (1986), Kornberg and Clarke (1994), Kuecheler (1986, 1991), and Lockerbie (1993).


54 conditions in a democracy (Clark, Dutt, and Kornbe rg 1993 1003). This definition, and the overall use of satisfaction with democracy, has come under fire by scholars, namely Canache, Mondak, and Seligson (2001). These scholars critique the measure in 4 respects. First, they believe satisfaction with democracy may vary depending on the individual. For example, men and women might systematically view the question differently. Second, interpretations of the measure might vary across nations, making comparison not only problematic but an impossibility. Third, satisfaction with democracy may measure different things at different times. Finally, the constructs of the survey itself may change the constructs tapped by sa tisfaction with democracy ( Canache, Mondak, and Seligson 2001, 513 514). While these c oncerns are appropriate, there are a few reasons I believe that their outright rejection of the measure is overstated. their reliance on suspect survey data as evidence fo r their conclusions. Their research uses three separate surveys: a small sample of 222 respondents in Romania, a larger survey done in El Salvador, and a comparison of 17 countries surveyed in the Latin Barometer ( Canache, Mondak, and Seligson 2001, 517 5 22). The survey in Romania is alarmingly small, which the authors admit ( Canache, Mondak, and Seligson 2001, 518), so I hesitate to give full credence to their findings. The Latinbarometer is also a poor data source to dismiss satisfaction with democracy as a measure. I do not have any America, Africa, and Eastern Europe In one of his works Seligson (2005) declares that


55 scientific surveys in these areas lack reliabili ty, validity, and quality. The original work on satisfaction with democracy, and the use of the Latinbaromter, is published 4 years the use of the Latinbarometer after further study. If not, it would seem to be more than hypocritical to use the very surveys you find lacking as evidence for anything. from advanced industrial democracies. U nfortunately, the required data are not available. The SWD item has been asked numerous times on the Eurobarometer does 20 01, 516) Since their writing in 2001, survey question have been added to the Eurobarometer that might offer opportunities to run analysis on these alternative measures of support. These surveys ask respondents the extent to which they feel uestions that relate to the feeling of political community. Also, Canache, Mondak, and Seligson (2001) use confidence in the armed forces, judiciary, congress, and policy as evidence of system support. Comparable questions are available in the Eurobarome ter as well. The Eurobarometer also asks questions related to trust in different institutions like the European Union, the national parliament in their home country, and institutions like political parties. Further questions ask about the performance of the national government in particular policy areas and separate questions ask about the performance of the economy. These measures could be used as examples of indicators of support for incumbent authorities. This indicator of incumbent authorities is act ually lacking in the Latinbarometer data that Canache, Mondak, and Seligson (2001)


56 use to critique satisfaction with democracy. The only measure the Eurobarometer does not ask is the preference for authoritarian government over democracy. This should not leaving one object of support out of the analysis does not preclude its use. To summarize, the opportunity to test in advanced, industrial democracies the relationship be tween satisfaction with democracy is there, but as yet not taken advantage of. deploring conceptual and operational ambiguities in the wording of questions. Whatever the data measure, it is interesting to look at the results in l ongitudinal 120) I believe that the data can be used as a summary variable to measure support across multiple dimensions. This is the finding of Clarke, Dutt, and Kornberg w ho find that satisfaction with democracy is equally correlated with support for 2001, 512) Whil e I believe that the critiques of Canache, Mondak, and Seligson are overstated, I heed some of their warnings. The potential lack of comparability between countries is a warning that all scholars should heed. For this reason, I use the measure satisfacti on with democracy in Italy from 1974 to 2005. This limited context should allay the concerns related to comparability. I also intend to use the measure in conjunction with other data on support in order to bolster and improve my findings. One of these m easures is trust in the political system (an item I describe earlier), others include: the preference for alternatives and participation.


57 Preference for Alternatives Support as an explanatory variable has been devalued by authors like P r zeworsk i (1986) who say that it is a tautology to suggest that countries who continue as democracies (or political communities) have the support from their populations, while those that fail do not. I only have two comments to make about this contention. The first is that system change may not have anything to do with system support. There are certainly international factors to consider. Countries defeated in war, like Germany and Japan, have new political system foisted on them or in the case of Germany, two new politica l communities. The second criticism I have is that just because an argument is simple, does not preclude it from being true. But, Pr taken, for support to have any discernable value as a variable then declining levels should ac tually mean something. Other scholars have taken already begun to address this question. Theoretically, the existence and feasibility of alternatives is necessary for support to have real meaning. Richard Rose makes this claim in several of his works ( Ro se 2007 ; Rose and Mischler 1996 ). He app the worst form of government, save all o 30) to post communist countries. Rose argues that support for democracy has n o meaning in contexts where the population has conception of an alternative system. In other words, support for the current system only matters when there is a different system to compare it with. This makes a lot of theoretical sense. If anyone were go ing out to search for the opinion about a movie, we would not take the word of someone who had seen only one movie in their lifetime as a valuable source. Why would we do the same with a political system? Rose takes this viewpoint and applies it to post communist cases by


58 comparing feelings about their new democracies versus their old author itarian regimes. In Chapter 1, I already state that Western d emocracies are not a threat to become authoritarian, scholars should look to countries that have changed electoral system and then look at how citizens evaluate the new system versus the old one. New Zealand presents an opportunity to study feelings about different electoral systems. There have been a number of surveys commissioned entitled the New Zealand E lection Study. These surveys ask respondents, in the year of the electoral reform referendum, whether they voted for a change or not. Following surveys ask what dep endent variable, it is possible to ascertain a progress report for the existing electoral system as well as determining what variables are related to a preference to an alternative means of elections. Unfortunately, these types of survey question are not available in Italy, so I use a further measure of support: participation. Participation Political participation, both conventional and unconventional, has been used as evidence of both a supportive and unsupportive population. I describe this l iterature e arlier in Chapter 1 so I will no t belabor it by repeating. Generally speaking, citizens with 20) In this study I concentrate on types of p articipation that indicate supportive attitudes toward the system: voter turnout and share high legitimacy norms, most citizens would likely engage within institution al channels (voting, contacting officials, party activism). Their behavior thus would tend to reinforce the sy 22 23) Why do citizens turnout to


59 vote and why is that related to support? Voter turnout could be explained as support for incumbent authorities because of satisfaction with the way of things, or it may be that the intent, voting in and of itself is a demonstration of support for something. It could be that a citizen wants to affirm their support for a set of ideas or declare rejection of others. Whatever the case, the act of voting demonstrates the active support of a certain process of democracy. Non voting is evidence th at the person is either apathetic or does not believe in voting, neither of which indicate support. I do not make higher. What I will say is that when voter turnout dec lines sharply or gradually over time, it is an indication that voting, the most basic practice in a democracy has diminished in value. When this happens, it is an indication that supportive attitudes have diminished as well. Another way to determine if su pport exists or has declined is to look at party membership. Joining organizations and groups, like voting, is one way to affirm support not only for that group but the larger system of democracy as well. The formation of political organizations and grou ps as a sign of a healthy, active, and supportive democratic citizenry is one of the most frequent claims made by political scientists (Putnam 1993). Political party membership is an example of one type of group directly tied to existing political systems levels of party membership are. What I do claim is when party membership declines, either rapidly or over time, is evidence of a citizenry that is disengaging from the political system and become less supportive.


60 There are four different means that I will approach measuring support: trust, satisfaction with democracy, preference for alternatives, and participation. These measures are not perfect by any means, but I believe taken together they provide a more complete picture of supportive attitudes in a country. Moving Forward: Linking Electoral Systems and Support Chapter 1 summarizes the literature on support and makes a case for re examining the relationship between electoral institutions and supp ort in contexts where electoral systems have been changed. I suggest New Zealand and Italy as cases because they are countries where this change has taken place. They are also Western, deve loped nations so this answers Pr (1986) critique that t he study of support is meaningless in contexts where there is no change. Before moving forward to these cases, it is important to discuss the theoretical links between electoral systems and support. I contend in Chapter 1 that these theoretical links hav e been underspecified. Electoral systems do not create support, but they might create the conditions where certain policies and processes then le ad to support. The Chapter 2 attempts to add precision in this area by drawing a theoretical link between bot h majoritarian and proportional electoral systems, the types of party systems they promote, the policies and processes that are more attributed to those systems, and support.


61 CHAPTER 2 LINKING ELECTORAL SYSTEMS AND SUPPORT Introduction Chapter 1 dea ls with the concept of political support: its meaning, the objects it is directed toward, the consequences for the varying levels of support and why political scientists should care ab out the variable. Chapter 2 provides a theoretical link between elector al systems and support, using this review of the literature as a guide. Before discussing this link a brief review of the literature on support is necessary. Support is impacted by several factors including policies, process, and institutions. Support m ay be influence d by policy in the following manner. Citizens elect representatives (whether individuals or parties) on the basis of the types of policies they would like to see enacted. Support is given or taken away based on the extent to which policies are enacted and successful. Process based approaches to support note that policy is often tisanship, the parsing of language, corruption, or gridlock are more likely to move citizens than policies they may not be aware of or understand. Any evaluations of the performance of elected officials or government in general, whether based on policy or process, are made within a particular context. Because of this, it is theorized that institutions themselves can impact support as well. Explaining the theoretical relationship between institutions, types of policies and processes, and the level of supp ort is the main goal of Chapter 2 I argue that scholars should look at the relationships between certain types of institutions, namely electoral systems, in particular contexts, in order to mo re fully understand what affects support.


62 The review of the literature on support suggests several ways to move the support to matter it has to be studied where there is a real potential for political change, or where politica l change has actually taken place. In my view, one of the real weaknesses of the literature is that Western, developed nations that have undergone no major political changes are the main topics of study. The study of these countries makes some sense, sin ce they are rich data sources, but absent any political change as a result of rising or declining levels of support saps the weight of its use as an independent variable 1 Part of the problem is that political scientists have come to expect political chan ge at too high a level. It is practically inconceivable that the United States, England, or France would abandon democracy for authoritarian governments or split up into separate nations. What is conceivable is that these countries might go through some sort of lesser institutional change, like an alteration of their electoral systems, as a result of declining levels of support. These are the types of changes that occurred in New Zealand and Italy during the early 1990s and because of this they make exc ellent cases for study. Before a study of the relationship between electoral systems and support can be made, regardless of context, there needs to be an explicit theoretical accounting of the theoretical link between electoral systems and support. This particular area of theory has relied on untested hypotheses and supposition s rather than clear theoretical reasoning. To summarize, I suggest a research agenda that examines how institutions, specifically electoral systems, can impact support both positiv ely and negatively in countries that have gone through some kind of institutional 1 This is opposed to developing nations where surveys have demonstrated problems with their metho dology and reliability (Booth and Seligson 2009).


63 change. Thro ugh this the aim Chapter 2 is to help bring some clarity to the theoretical gap between elections and support among the population. Electoral systems, or any ins titutions for that matter, do not cause directly support to rise or fall. Institutions instead provide context. Political scientists are tasked with the job of determining what the theoretical steps are in between. Unfortunately, there has been a real l ack of theorizing in this direction. This is puzzling because there is enough theory to develop an inclusive connection between electoral systems and support. For instance, there are studies that contend that particular electoral systems foster produce m ore consensual legislative processes (Baylis 1989; Powell 2000) and better, more widely respected policies (Baylis 1989; Lijphart 1999 ; Lizzeria and Perisco 2001 ; Powell 2000; Snyder 1989 ). There are also studies that examine the relationship between vari ables that are attached to some measure of support like participation ( Blais and Dobrzynska 1998 ; Ladner and Milner 1990 ; Norris 1997; Powell 1980 ) or satisfaction with democracy and trust ( Anderson 1998 ; Anderson and Guillory 1997; Anderson et al. 2005 ; B anducci, Donovan, and Karp 1999 ). Policy and process are explanatory variables for changing levels of support, thus the natural fit for the study of the correlation between electoral systems and support. The trouble I have with existing studies on elector al systems is two fold: there has been a proliferation of research that links electoral systems to a seemingly increasing level of variables and outputs and the lack of clear, concise theorizing ab out what electoral systems are responsible for and what th ey just might provide the context for if other, more direct outputs are produced. Examples of such studies include those that link particular electoral systems to levels of corruption in particular countries ( Chang and


64 Chu 2006 ; Chang and Golden 2003 ; Mai nwairing 1991 ), higher levels of GDP (Rogowski and Kayser 2002), as well as higher numbers of minority and women candidates (Duverger 1954; Lakeman 1970, 1983; Lijphart 1999 ; Norris 1985, 1987 ). I do not suggest that these relationships do not exist, but these studies are evidence of the larger second complaint, because often studies like these treat the relationships as if they are direct, without understanding and examining the crucial theoretical steps between them. All institutional rules have direct a nd indirect effects and scholars that do not distinguish between the two do so at their own peril. Electoral rules are no different. based on their particular rules. That support (Rae 1971 68) are more tenuous and require careful theoretical examination. This is the reasoning that is lac king in my estimation. Take for example the variable cor ruption, which has been presumed to be impacted by the type of electoral system. Giovanni Sartori in different works (1976, 1994) describes how proportional electoral systems might lead to higher le situation where ideologically incompatible parties all receive seats. When a situation like this occurred, Sartor i theorized that the government would have a difficult time operating the existing institutions and meeting citizen demands. If that occurred, then elites and individuals might seek extra legal remedies, like corruption to meet their goals (Sartori 1976, 1994). Scholars have used this reasoning to test the relationship


65 between proportional electoral systems and levels of corruption in countries ( Chang and Chu 2006 ; Chang and Golden 2003). The trouble I have with this is that there is no real examination of the theoretical depth that Sartori suggests. He suggests a particular case where proportionality leads to a particular party system, which is then followed by a specific reaction and a precise n to be correct scholars must demonstrate that proportional systems lead to corruption, and all of those theoretical links must be in place. When scholars merely test these theories, they leave out the theoretical underpinnings that brought the link betwe en electoral systems and corruption together in the first place. This also prevents examination of alternative theories that are equally plausible. There are alterative plausible explanations for why corruption may or may not be as pervasive in a proport ional system as well as reasons why majoritarian systems might lead to higher levels 2 Without an examination of all the links, a positive confirmation does little to add to our theoretical understandings, and a negative result does little more. This les son should apply to the study of the relationship between support and electoral systems as well. If it is erroneous to say that electoral systems lead to more/less corruption, then it is equally incorrect to say that electoral systems lead to more/less su pport. Electoral systems may help to create conditions that foster support, but only through indirect means. It is the goal of Chapter 2 is to 2 For instance in a majoritarian electoral system: the electoral system reduces the number of political parties down to two and these parties maintain holds on political power for a long period of time. Me mbers are increasingly re elected and parliamentary turnover goes down as seats become safer. Representatives become patrons, doling out goods to the clients in their districts. These patron/client relationships coupled with the safety in office leads to increasing ties between moneyed interests and the political class, leading to higher levels of corruption.


66 determine what steps are included between electoral systems and support, using the literature on support and el ectoral systems as a guide. The literature on electoral systems contains enough theorizing to create a theoretical chain between types of electoral systems and differing levels of support. I draw this chain with a constant eye on the variables that impac t support: policy and process. The theoretical models drawn use the existing literature on electoral systems, especially the work of Maurice Duverger, on the political consequences of different electoral systems. These models will construct paths from bo th proportional electoral systems and majoritarian electoral systems to system support. The proportional model will look familiar to most scholars, since it is the system that most advocate. I will also include a model for how more majoritarian system mi ght lead to higher levels of support. Once these models are drawn, I present potential counter arguments that might confound the theoretical models. Finally, I set the stage for my case studies of New Zealand and Italy. The Political Consequences of Elec toral Systems Maurice Duverger inspired much of the work done on electoral systems (1954). This work made claims about the relationship between electoral systems and party systems 3 ballot rule tends to party 4 ballot system and proportional 3 parties; coordinating with other like m inded parties or groups; persuading voters of their merits; and mobilizing their supporters. Each of these activities has been used previously to define differences 4 Later nam


67 5 (Duverger 1954, 1972, 1986) Although Duverger originally claimed that these propositions occurred with law like regularity, he weakened these claims in subsequent work, admitting that the relationship was merely probabilistic in nature (Duverger 1986). These two simple statements ultimately spawned a host of 6 Eventually that work would lead to drawing connections between different electoral systems and differing levels of system support. Duverger used a set of simple examples to demonstrate his law and hypothesis. He considered a single member plurality district 7 in whic h 100,000 moderate voters were opposed by 80,000 communist voters. Duverger reasoned that if the 100,000 moderate voters were split into two parties (with one having the support of at least 20,000 voters) the communist candidate would win the election. I n the following election, Duverger reasoned that the two moderate parties would unite. If they did not, Elections determined by a majority v ote on one ballot literally pulverize third parties (and would do worse to fourth or fifth parties, if there were any; but do not exist for this very reason). Even when a single ballot system operates with only two parties, the one that wins is favored, a nd the other suffers. The first one is over represented its proportion of seats is smaller than its percentage of the v otes. (Duverger 1972, 24) 5 6 Rae (1971), Riker (1976, 1986), and Lijphart (1 984, 1994). 7 In this example a single member district, contains one seat where the candidate or party that receives the most votes, wins the seat.


68 An example of how one party is over represented against a single alternative party in a single member district system is if a winner in a single member district receives 51% of the vote, they will receive 100% of the seats in the district, and conversely the party that receives 49% of the vote receives 0% of the seats. The party that loses that election is grossl y under represented. Duverger called this phenomenon of over and under representation the mechanical effect of the electoral system. The mechanical effect is the mathematical impact the electoral system has on the party system. This effect is direct, de pends on electoral rules, and impacts the translation of votes to seats. The vote hurdles that single member districts create for third parties who have limited support exemplify this type of effect (Duverger 1954, 1972). There are also indirect impacts Psychological effects can cause a voter to cast a strategic vote. Strategic voting is when a voter consciously votes for a party or candidate that is not his or her first choice. A pe rson may behave this way because of a belief that doing so would insure the best possible outcome 8 belief by a voter that a vote for a third party under most circumstances would be wasted understanding of the disadvantages that third parties face against unfavorable electoral rules (Cox 1997 ; Downs 1954 ) 9 In systems of proportional representation (PR ), Duverger believed, the influence of the electoral system is very different. List proportional representation differs from 8 An example would be a Nader supporter voting for Al Gore in the 2000 election out of fear of a George W. B ush presidency. 9 Generally the concept of strategic voting has been confirmed by other scholars including: Bensel and Sanders (1979), Cain (1978), Fisher (1973), and Spafford (1973).


69 plurality systems in three respects. First, instead of voting for an individual, votes are cast for a list of candidates 10 Secon d, electoral districts allocate more than one seat. Third, PR distributes the available seats proportionally according to the percentage of votes received by each party (this is under pure PR). Duverger used another theoretical example to demonstrate t his claim. He reasoned that any minority party, no matter how small, could be assured, absent legal thresholds, of representation in the legislature under pure PR. Because of this there would be nothing to stop the formation of splinter parties, even if they were only separated by slight differences of opinion. He pictured a legislature where, If the conservative party has 6 million votes in the country, corresponding to 300 seats in parliament, and if it splits into three groups about equal in numbers, proportional representation will give each of these about a hundred deputies, and the conservative family will have the same strength in parliament. (Duverger 1972, 25) So, the kinds of barriers and incentives under single member district systems are not p resent under proportional representation. For instance, under a single member district system, the conservative parties would perhaps form a coalition, rather than split their strength. Under PR, these barriers would not be there. This is how Duverger described the direct impact of proportional and majoritarian electoral systems 11 He also had theories about how these mechanical effects would have more indirect effects as well. Almost all of the work linking electoral systems to any other variable outs ide of the direct, mechanical effect (number of parties, types of policies enacted, or support for the political structure) rely on theories of what the 10 This is not the case for single transferable voting systems (STV). 11 It should be noted that Duverger also included double ballots elections along with proportional representation as electoral systems that favored multi partism.


70 psychological or indirect effects cause. Electoral systems do not directly cause people to support the ir government, but they may create a series of conditions that result from a series of indirect effects. Understanding these indirect effects and how they connect together is crucial to drawing a theoretical link between electoral systems and support. Th ese ties will be drawn from the particular type of electoral system to the number of political parties, to their impacts on the type of governments created their behavior and the policies they enact as well as the level of participation among the citizenry I will first show how all of these variables are connected under proportional systems. It should be said that most scholars believe that proportional systems lead to higher levels of system support, so I will begin by developing a theoretical model bet ween increasing levels of proportionality and increased levels of system support. Linking Proportionality to System Support In order to link proportionality and support, it is necessary to l ook back at Chapter 1 to determine what influences support. More specifically, what influences support and could those variables conceivably be impacted by electoral systems? Immediately the nothing to do with age, income, education, or g ender. But, it is possible that a link can be drawn, albeit indirectly, from electoral systems to policy and process. This section details a connection between proportional electoral systems and support by detailing the intervening variables including bo th policy and process. Then I discuss how policy and process variables here are inter related. In the preceding section on the mechanical effects of electoral systems I detail the differing mechanical effects of electoral systems. Since the mechanical ef fects differ under majoritarian and proportional electoral systems, the psychological effects are


71 voting for a third party do not exist under proportional representation 12 He reasons that because votes are translated into seats almost continuously under proportional representation, two things would occur. First, voters would have less of a reason to vote strategically; second, parties would have no reason to form pre elec tion coalitions. Because of these two suppositions, he reasoned that the number of political parties would be higher under proportional systems. This relationship between proportional systems or majoritarian systems (a topic I deal with subsequently) and the number of political parties is a very important first step member district systems should produce two party systems 13 They do so through a mechanical effect, which is bi ased against third parties and a psychological effect that electoral system creating a specific type of party system, one that does not and is not designed to re Duverger contends that proportional systems offer neither the mechanical hurdles for third parties, nor the psychological effects for voters. Since there are fewer barriers, the limits on political parties are also r educed, leading to multiparty systems. Duverger 12 Cox (1997) describes situations where voters under proportional, list systems might vote st rategically. threshold as slim, so they might defect. This is similar to the voting behavior under single member districts where voters vote for a least wors t alternative rather than cast a sincere vote. 13 This production is only at the district level, at the national level or in the aggregate the party system may include more than one party. There are a number of reasons why this might occur. For example, a country could have parties with significant regional strength that allowed them to compete in particular districts and win seats (Sartori 1986). Generally speaking, within those types of districts there still are Law may appear to be a fallacy when only examining the national level, in actuality at the district level, it is still performing well.


72 has been challenged by other scholars (Lijphart 1984; Taagepera and Shugart 1993 ) s important because it suggests that there is an opportunity for more voices to be allowed into the political system. The result is moment), but for Duverger and Lijph art this meant more political parties. deals with the number of political parties in a manner more conducive to drawing a theoretical model by offering a proposition for the im pact of increased proportionality on can expect the number of parties to be (Lijphart 1984, 159). He argued that the number of parties would increase from lowest in p lurality systems, to somewhat higher in majority systems, higher still in semi proportional systems, and highest under proportional representation (Lijphart 1984, 159). To summarize what has been stated to this point. Electoral systems have both mechani mechanical effects lessening the psychological effects on voters, leading to increased number of political parties. To spell out this progression, I present several models that de monstrate the step by step relationship between electoral systems and citizen support for government. These models include all of the intervening variables that I believe have been understudied. The first of these models includes the work of Duverger and Lijphart, and explains the relationship between proportionality and the number of political parties:


73 (1) political parties This model states that that as the proportionality of an electoral system increases, decrease, leading to an increased number of political parties. This model is a representation of both the work of Duverger (1954) and Li jphart (1984) However, the model I have drawn to this point has a few more steps that need to be added to get from the type of electoral system to citizen support for their government. One of these steps is the impact that electoral systems have on, not only the number of political parties, but the type of political parties that exist in the party system. Gary Cox theorized that as the number of political parties increased, so would the number of issue dimensions that the party system included (Cox 1997, 1999). There is a relatively simple idea behind this theory. Cox stated what Duverger had implied in his earlier work. Voters were not the only participants subject to the psychological effects of electoral systems, but parties were also impacted. Par ties recognize when they have a poor chance of gaining seats in a given district and they may not choose to contest a seat that they know they have little chance of winning. This is known as strategic entry or non entry (Cox 1997). If parties feel that t hey have political standing in a country whether ideological or regional, then they will run for office. This affords voters a chance not only to cast their ballot sincerely, but also to cast their vote for a party that is closer to their own ideological beliefs. When the electoral system is more permissive, more parties representing a larger ideological spectrum have a greater chance at gaining seats. So, when this idea is included the model looks like this:


74 (2) ) mechanical/psycholo This model states that as proportionality increases, the mechanical and psychological effects for voters and parties decrease, leading to increased multipartism, and as the number of p arties increases so do the number of ideologically diverse political parties. Increased representation is not limited to ideology. Scholars have theorized and demonstrated that proportional systems tend to have increased representation of groups including : women, various ethnic groups, or religious factions (D uverger 1954; Lakeman 1970, 1984 ; Lijphart 1999 ; Norris 1985, 1987 ). There are several reasons why this might occur and they result for the different psychological effects attributed to proportional systems. Strategic entry/non entry is not the only psychological impact for parties. Parties also realize, under proportional systems that they need to maximize their vote share and one of the strategies for doing so is to increase the number of women or minority candidates on their list. Ethnic or religious parties receive the benefit of being able to form and field their own lists under proportional systems rather than being subsumed by one of the larger two parties in a majoritarian system. For those who favor proportional representation, this type of matching, whether ideological or based on identifying characteristics, represents a democratic ideal. In other works, democracy is created to mirror society. Dahl (1971) asks the question These scholars ( Amy 199 3 ; Lake man 1984 ) clearly fall on the side of the delegate in the classic Burkian debate concerning whether elected officials should represent their or individual characteristics (delegates) as closely as possible or govern


75 as they see fit (trustees). As such, PR advocates, it would seem, prefer that electoral systems act like a photograph where the population is represented as closely to reality as p ossible. The root that links proportional electoral systems to support depends on this incr eased representation. So, I replace another link to the theoretical chain, ide ological, gender, ethnic, or religious differences in a society. When this is done the model looks like so: (3) This model states that as proportiona lity increases, the mechanical and psychological effects decrease. Without the fear of wasting their ballot, voters feel freer to vote sincerely. When this happens, more political parties, representing a wider cross section of the population (both ideolog ically and demographically), have a chance to be elected. This leads generally, to higher levels of representation. Representation is an important process in the theoretical chain, but pure representation is not enough to reach the end of the theoretical chain (support). Rather than start at the beginning of the model, let us think about what might result from a party system that has multiple parties that represent a broad, cross section of the population. The activities of a governing coalition are imp acted in a number of respects. Multi party system, most times, have to go through a process of coalition building after the election in order to form a government. This is a much different occurrence than what takes place in a majoritarian government tha t is generally dominated by one of two main parties. Many governments in proportional contexts are made up of several parties. When there are multi party coalitions the parties have to find a way to work out their


76 differences with one another. The acrim ony associated with two party competition where there is an in party and an out party seeking to replace them is replaced by a more consensual, decision making process (Lijphart 1984; Powell 2000). Any type of decision making when there are multiple inter ests involved are going to be more considered, done by consensus, and simply take a longer amount of time. Scholars have stated that this slower, more considered process will be more acceptable to a larger segment of the population (Baylis 1989). Just be cause there are more parties representing a broader section of the population winning seats in the legislature does not mean that they will be a part of the government. This does not matter to some scholars who believe that as long as more people feel lik e the process is fair are willing to accept the decisions, even if they are contrary to their interests (Katz 1997). All of these procedural considerations might lead to higher levels of support. Representation here goes further than just taking a photo graph of the population the election brings representative agents of all the factions in the society into the policy making arena. These agents then bargain with eac h other in a flexible and accom 6) Representation contains two separate, though related components: the representation of ideas and groups in societies and the weighing of those ideas and interests in an equal and accommodat ing fashion. This is the proportional ideal of representation, so I keep the model the same, with the added These processes alone might lead to higher support. Here is a model that represents how those steps might l ook:


77 (4) To summarize, an increase in proportionality leads to a decr ease in the mechanical and psychological effects and an increase in political parties. When political parties are increased, representation is increased not only in terms of the increased ion making process (Powell 2000, to higher levels of support. In my estimation, this explanation is a bit dissatisfying. Citizens are not going to be content to watch their elected officials bargain and debate unless they come to policy decisions that the public can get behind and actually support. This means that the process of representation has to produce better policies. In a democracy the preferences of all citizens, not ju st an electoral majority should be taken into account in the making of policies Even if they represent the citizen majority positions on all the issues, a majority of representatives should not ride roughshod over the preferences (especially if intense) of the minority. The best guarantee that the majority will take account of minority preferences is to give the minority some valuable policy making power [emphasis added]. (Powell 2000, 6) P roportional systems, to be successful, must combine representativ e processes with good policy outcomes. Quantitative, cross national research suggests that proportional systems do perform better. These works show that countries that use proportional representation have higher gross national products, lower unemploymen t, lower inflation, policies are more egalitarian, and public goods are distribute d more efficiently (Baylis 1989 ; Lijphart 1999; Lizzeria and Perisco 2001 ; Powell 2000 ; Snyder 1989 ). The process of representation here is closely connected with policy out comes. A representation of this model looks like this:


78 (5) (+) Proportionality sion To s ummarize, increased representation precedes a more consensual a nd inclusive decision making process. This leads to policies that are constructed with a broad, cross Increased levels of support are a result of these better, more widely accepte d policies. As we rem ember from Chapter 1 the literature that deals with the relationship between electoral systems and support (and more specifically proportional representation and support) has not developed enough rigorous theories in these areas. Thi s is not to say that the literature does not hint at or use elements of the models that I have drawn, but those have attempted to test the impact proportional representation have not laid out a step by step theoretical model that describes all of the links in the chain. That said, there are studies that show that proportional systems have higher levels of satisfaction with democracy, fairness, and efficacy ( Anderson 1998; Anderson et al 2005 ; Anderson and Guillory 1997 ; B anducci, Donovan, and Karp 1999 ). While these studies are certainly interesting, without developed theories we have not broadened our understanding. It is important to remember that electoral systems do not cause support to rise or fall. They may only create the conditions for increased support and it is our job as scholars to understand when those conditions are in place This is a level of theoretical precision that scholarship has yet to reach. I present two separate models linking proportional representation to increased feelings of support. The two models are rooted in the literature concerned with the sources of support. The key explanatory variables in models (4) and (5) are process


79 (representation) and policy. Model (4) treats process as a separate independent variable that ex plains varying of levels of support by themselves, but I believe that the more interesting connection between electoral systems and support is model (5) that treats process as an intervening variable. Governments, whether elected through proportional or m ajoritarian means, have to govern, which means they have to produce something (Sartori 1994). When examining the cases, the relationship between process and policy in proportional settings is something I am going to keep a particularly close eye on. There is another idea that needs examining. It is a step forward to specify the theory concerning the connection between electoral systems and support, but it does add a measure of complexity. This complexity certainly adds to the difficulty of studying this question, but there are other consequences that have to be considered as well. Now that there is a theoretical chain, there are more links where the theory might fall apart. Put another way, there are direct consequences of electoral systems, for example the mechanical effects. All other effects can be considered psychological or distal effects. As we move down the theoretical change, the less certain we can be about the more distal effects. For instance, what happens if the increased representation do es not lead to a more consensual decision making processes? What if the policies are just lowest common denominator compromises rather than a reflection The next section deals with some of these questions and examines where, looking forward, the connection between electoral systems and support might be lost.


80 Possible Breaks in the Chain: The Links between PR and Support Since I am going to examine my cases studies using the theoretical models, (4) and (5) that I have just draw n, it is important to stop for a moment and see where the models might break down. Those who advocate for proportional representation have a particular vision for the ways that democracies should work. They focus on the onal systems favor. I try to keep these ideals in mind when I draw models (4) and (5). But, it is fair to say that the models I draw paint a rather rosy picture well, rather than an accur ate picture of how they actually or could work. These actualities might present a much different picture. Even if proportional systems reach seats, this does not mean t hat the rest of what follows is guaranteed. What happens after proportionality is conditional, so several other plausible results are just as likely. This section presents some of these other results and how they might be confounding for models (4) and ( 5). The model presents a situation where a proportional electoral system delivers a parties then bargain in good faith and create policies that everyone will accept. Bu t, PR, There is no longer any one group which extends its influence over the entire political spectrum. Parties settle anywhere on the political map, and they enter cam paigns (as their leaders, in the German case, used to say in the days of the Wiemar Republic) not just represent differences as they are. They escalate them, making it to form a coalition. (Hermens 1984 22 23)


81 Sartori (1976) disagrees with Duverger (1954, 1984), Hermens (1984), and Lijphart (1984) suggesting that proportional systems do not automatically lead to increased fragmentati on of the party system. He believes instead that proportional rty proliferation (Sartori 1976, 173). What There are several mechanical consequences that are liable to follo w somewhat like a chain reaction more the parties, the more the party pattern will belong to the rigid ideological non aggregative variety rather than to the pragmatic brokerage variety; the more the parties, the less they can all expect to share governmental responsibility, and the more this creates the conditions for irresponsibility, and the more this creates the conditions for irresponsible opposition; the more the parties, the more the conditions for the polit ics of outbidding and outflanking. (Sartori 1976, 173) What these authors present is a very different view of what happens under proportional elections. There is no guarantee that parties are going to get along or consider the best interests of the public before making decisions. The reason this happens is that there is no direct control over the government formation following the election. The separation of the election of parties to the formation of the government is part of the proportional ideal. Pow suggests that it is mere voting is not a good way to determine the consensus position of the public. It is better, in his description, to elect representatives and allow them to reason and bargain among themselves to find the appropriate median positions and the government to represent them. As Hermens and Sartori suggest, proportional systems do not provide incentives for parties to behave responsibly. One scholar has offered the example of how coalitions are formed under proportional systems (LaPalombara 1987). In this movie 5 d ifferent individuals witness an event and


82 the entire movie is a retelling of this one event from the varying perspectives of those 5 individuals. T he 5 individuals have vastly dissimilar recollections of what happened. Now consider an election under proportional representation. There may not be a clear coalition follo wing the election and we have already read the claims of scholars who make the argument that proportional systems do not necessarily encourage parties to behave responsibly. The government is now in the hands of parties who are predisposed to behave irresponsibly, horse trade, or jockey for position in the government. The government that i s formed, like the events in Roshomon, may not be reflective of what citizens had in mind when they voted for their parties during the interpretations of meaning. Voters may im agine one government, end up with one that they do not want, feel that their voice is not being heard, and the process of representation is not met. Like the proverbial sausage making process, the building of a coalition is neither n eat nor tidy and may i mpact the way a citizen offers support. This calls into serious question the expected behavior under model (4). The building of a coalition is not the only process that can be difficult under proportional representation. The actual maintenance of a coali tion can be difficult as well. Sartori (1976) and Hermens (1984) have negative views of the type of party system that proportional representation allows and the behavior that the parties are trade and bargain, but to behave irresponsibly, outbid one another, and maintain strict ideology. These are not the behaviors that are conducive to maintaining a stable government coalition. The critics of proportional systems suggest that coalitions ar e more fragile and stability is a


83 continual problem. The fear is that proportionality is a natural precursor to chaos like in the classically used examples the Germa n Weimar Republic, the Italian First Republic, and t he French Forth Republic. In Weimar G ermany, a fragmented party system was incapable of governing and led to the rise of the Nazi Socialist Party. Italy did not have that type of disaster, but governments fell repeatedly after WWII with a governmenta l months. So, if the processes in the theoretical model (4) are a best case scenario of how proportional systems behave, model (5) which emphasizes the policy benefits of proportional representation look s even less and less likely Parties run for office not only to gain seats, but to enact the policies that they support (Aldrich 1996 ). Sartori (1994) states something similar. He suggests that the point of all elections is to elect representatives who are capable of governing. The just detailed do not make enacting significant, widely accepted policies possible. If parties are inclined to horse trade and bargain following the election without keeping the intentions of the electors in mind, the government that is formed is not truly representative, why should we believe that the policies they enact will be better, more widely accepted? Also, if governments can not hold themselves together then no meaningful policy can be p assed. Good policy requires some stability in office, and if the electoral system does not provide stability in office, then it can not be connected to good policy. Without processes that involve consensual decision making and coalition formation, then a ny good policies that are happened upon can not be attributed to the electoral system.


84 Considering that point, there is a further point that I would like to make about making policy in general. Any theoretical design, like model (5), can not take into acc ount all of the variables that account for policy making. The making of policy requires a lot of variables, some of which model (5) represents, but a series of others that it does not. Constructing policy requires talented elected officials who are willi ng to experiment, set narrow, partisan interests or ideological preferences aside and support policy that is good for the entire country. Does any sort of electoral system guarantee that this will happen? No institution of democracy can create that kind of assurance. Institutions might provide certain conditions, but a degree of agency is also necessary. Electoral systems do not create first rate bureaucracies or excellent elected officials, which are very important to policy making. To summarize, there are several events that might confound both models (4) and (5). Proportional representation, according to its critics, does nothing to promote or provide consensual decision making processes or representative governing coalitions. In fact, these critics suggest they do the opposite. Coalition building may require log rolling, horse trading, and political infighting. Much of this could be done in secret, type of barga ining is more likely when parties have real, incompatible ideological differences. If parties are more inclined to behave irresponsibly, governing coalitions might be difficult to maintain, making effective policy impossible. These are not the processes that lend themselves to higher levels of support so perhaps we should take some of the more positive distal effects of PR with a grain of salt Duverger (1984) presents an appropriate critique of proportional representation in his work.


85 By dispensing the voters among numerous independent parties, PR prevents the citizens from expressing a clear choice for a governmental team. It transfers this choice to the party leaders. For instance, after most of them fragile and divided. Which one is preferred by the citizens? It is impossible to know this. Only the new deputies can choose between the different potential combinations, with the freedom to make a different choice without reference to the voters. (32) Considering this opinion, political scientists should be realistic and ask the question, is it likely that political parties will treat each other collegially and openly, while quietly accepting roles outside of government even if they feel as though they deserve them? Or are parties likely to fight, trade, and bargain for political power and influence? Certainly the second is just as likely, if not more likely than the first under a proportional system. Much of the literature, save the sev eral I mention, favors proportional representation. There is a theoretical case for majoritarian electoral rules as well, just not as forcefully represented. The potential for models (4) and (5) to fall apart, should force the consideration of alternativ e explanations. It is for this reason that the work of Duverger (1954, 1984) should be reconsidered. There is real potential that majoritarian electoral systems can be successful if the right conditions are met. The next section considers a theory of ho w support can be raised by reducing the number of political parties, manufacturing a majority government cable of passing effective policy efficiently, while remaining accountable to the citizenry through annual elections. Majoritarian Electoral Systems and Support: An Alternative Model The f ollowing theoretical models use essential ly the same literature that I draw upon to construct the proportional model, so I do not belabor repeating the history of electoral systems research. Essentially the model lin king majoritarian rules to support


86 work in the opposite direction than the proportional model. Proportional systems minimize the psychological effects that reduce the number of political parties. Majoritarian systems, on the other hand, have mechanical ef fects that punish third parties, leading to psychological effects that encourage voters to vote for parties that have a reasonable chance of gaining seats. These circumstances, in many cases, lead to 2 party systems 14 A rendition of this model at this po int looks like this: (6) ) number of political parties In the majoritarian model process and policy are inextricably linked. This is very different than the proportional model, where there is a conceivable link between process and support, without policy. In model (4), I propose a situation where support is increased through representation. When all elements in society are represented, it is conceivable that more people will support the govern ment even if they disagree with the outcomes. In majoritarian systems, process and policy ar e inextricably linked, and they look government. Schattschneider was part of a consortium of political scientists who were attempting to improve the deficiencies in the two party system in the United States. They believed that several elements needed to be in place for a party system to be responsible. These include a two party system where t he two political parties offer the public a consequential choice. The public should expect that their choice will be informed based on information provided by the political parties. When a political party wins office the party then implements the program that it promised prior to the election. 14 There are exceptions where a majoritarian rules can still allow for more than 2 effective parties at the national leve l, I discuss these exceptions in the section on the breakdown of the majoritarian model.


87 Finally, in the subsequent election the party in power stands for reelection based on the extent to which they enacted their pre election promises and the degree to which they were successful 15 (Schattschneider 1950 and differs from the proportional vision in several respects. The main process in proportional systems is representation in majoritarian systems (like those described by Schattschneider) the key process is acc ountability. to policy found in proportional systems. PR properly expresses the allow them to choose a concrete set of policies and a team to execute them. In contrast with this representation of opinions, what may be called the representation of wills enables the voters to choose those who will lead them during the entire life of the legislature. The citizens can not govern themselves, but the delegation of power may be reduced to one step (citizens government) instead of being stretched out to two steps (citizens deputies government) in which the governing team may be accomplished by the voters themselves instead of being left to the deputies. (Duverger 1984, 32 33) In majoritarian systems, the goal of decreasing the number of parties is to have one party in control following the electio n, so that party may form a governmen t. This government enacts policies promised before the elections, and voters judge their success or failure accordingly Accountability and policy outcomes go hand in ha nd. One authors concurs. [In majoritarian syste ms] the elected officeholders are able to make and implement policies. Responsibility for policy is obvious to everyone. From one perspective the citizens use elections to choose between prospective teams of policymakers. From another perspective the citizens use elections to reward or punis h the incumbents. (Powell 2000, 5) 15 The opposition party also puts forward an agenda for if they were to be elected to office.


88 Accountability and policymaking are both crucial to the majoritarian model. The ability to vote a government out of office (accountability) is stripped of its meaning without alt ernative policy making by the out party. Is it possible to imagine a situation where people offer more support for a system where parties alternate in office, but offer no differences in policy? I do not believe that citizens would. Because of this, the model for majoritarian systems is complete, running from the reduction of parties to the creation of governments who are judged on coherent policy agendas. A representation of this model looks like this: (7) ) number of government. This model states as majo ritarian tendencies increase, the mechanical and psychological effects that are biased toward third parties increase a s well This leads to a decrease in the number of parties and the election of a government that is accountable to the voters. These governments produce the policies that they were elected to pursue and are judged accordingly. When governments are succes sful in those policy pursuits support among the population goes up. Like proportional systems, there are several places where this model could break down. The next section details where this might happen. To foreshadow, in the quest to manufacture a maj ority, parties might be able to form a government with a very low amount of electoral support. For example, t otal control of the government can be achieved with 40% of the vote. Majoritarian systems may also be confounded by a third party that has region al strength, turning a two party system into a three party system. Also, just because governments are elected does not guarantee that they will follow


89 their pre election promises. Finally, alternation in government does not necessarily mean a change in p olicy direction. All of these could have a potential drag on support. Possible Breaks in the Chain: The Link between Majoritarian Systems and Support Most scholars today do not advocate for majoritarian electoral systems. There is a host of reasons why t his is the case. The main reason is that majority governments are created out of pluralities, not majorities. In New Zealand, one of the countries I examine in this study, rarely did the single party government that was formed with a majority of seats wi n a majority of all votes cast. For example, in 1984 the fourth Labour Government was formed with 59% of seats after winning only 43% of the aggregate vote (Nagel 1994) 16 Because of this, some scholars have advocating calling these types of systems plura litarian (Nagel 1994; Shugart and Wattenberg 2001). The fact that governments are formed out of pluralities, not majorities, is problematic for several reasons. Main among these is the under representation of interests, parties, and demographic segments of the electorate. One of the main reasons has to do with the plural nature of these systems. When governments are formed with less than a majority of votes, there is necessarily a majority that wants something other than what that government offered befo re the election. Take the example of New Zealand I cite in the previous paragraph. The Labour government formed after the election only received 43% of the vote, but had a majority of seats in parliament. This means that there were actually more New Zea landers who wanted something other than what the Labour party was offering. 16 This is only one of the more extreme examples, but it was routine for the winning party to receive a plurality of votes, while winning a majority of seats (Bale and Roberts 2002).


90 This is the main problem with majoritarian systems. They treat majority interests like minority interests. In search of a governing majority, large numbers of citizens and opinio ns can be ignored. Those who seek to have their voices heard by voting for an alternative also run the risk of having their votes wasted. Most of the presumed ills of the FPP system can be summarized by stating preferences. Many voters do not even have their first preferences taken into account. Those in constituencies where their candidate wins, waste that fraction of the total that is in excess of the total needed to win the seat. Those in constituencies where their candidate loses waste all of their vote it is irrelevant to the final outcome f the election (in terms of the allocation of seats). In any one constituency, the number of effective votes may be relatively small. (Johnson 1984, 63) Majoritarian systems simply deny representation to a variety of interests. Johnson (1984) argues that many voter preferences are ignored or marginalized, but there are other ways that interest s are underrepresented. The above point that I make is that majoritiarian systems turn majority interests into minority interests. There is also the problem of representing true minority interests. Take a third party with a national basis for support th at reaches 20%. This party is going to be seriously disadvantaged under majoritarian rules. This is not merely a hypothetical, it has been a persistent issue in Britain, when the Liberal/Social Democratic alliance (now merely the Social Democrats) rose i n the early 1980s. In the 1983 election, this alliance polled 26% of the total vote, but only won twenty three seats, a mere 3.5% of the total. The Labour party in the same election, received only 2% more of the total vote, but a full nine times th e numb er of seats (Lakeman 1984, 43). Another issue with majoritarian systems is the problems that can be created by smaller, regional parties that are able to gain significant representation like in Canada or England. Elections where regional parties gain


91 sig nificant support lead to a similar situation where the government could be elected not apply to the aggregate, but that its effects are apparent at the district level ( Duverger 1954; Sartori 1986; Taagapera and Shugart 1993 ). The question is, how can any system justify denying representation in such a manner? A third problem with majoritarian systems is the under representation of minorities and women. Proportional sys tems have to maximize the number of votes, so they actively seek out women and minorities to fill out their lists. Also, parties that solely represent a minority group have a lower threshold to win seats, so are encouraged by the rules to run for office. All of this adds up to higher levels of women, minorities, religious, and ethnic groups represented under proportional representation (Duv erger 1954; Lakeman 1970, 1984 ; Lijphart 1999 ; Norris 1985, 1987) Majoritarian systems simply do not offer the same kinds of opportunities. The goal of these systems is to manufacture majorities, not represent everyone so the groups I name are not represented at the same level. These are the problems associated with majoritarian systems in terms of how they represent the country as a whole, but there are additional problems with how those governments then pursue a package of policies. One of the presumed benefits of majoritarian systems and one of the justifications for over representing one of two main political part ies is the effectiveness that governments can pursue their package of policies in an efficient manner. The above arguments should highlight the problematic nature of such assumptions. When a government is elected with less than 50% of the vote, this mean s that there is not a majority mandate in the country for the policies they want to pursue. In fact, their policy


92 agenda has been implicitly rejected by a majority of the country. The only alternative voters are left with in this type of system is to vot e for the alternative. But, this raises another question. Do majoritarian systems guarantee that parties will be accountable? Accountability requires that parties offer different policy agendas from one another and that if elected the government then pur sue that policy agenda. Parties are then judged based on their success or failure and are rewarded or replaced depending on their actions. This is the ideal situation in a majoritarian system, but it does not necessarily have to come to fruition. Altern ation in governments does not guarantee that parties will pursue policies that are appreciably different than their counterparts. This has been a main complaint of third party candidates in the United States. George Wallace famously declared that there w the Republicans and Democrats, while Ralph Nader labeled George W. Bush and Al dee and Tweedle reflections of one another, with no real difference in policy programs or agendas. When this is the case, replacing one party for another does not represent a real change is policy direction, it becomes for of a cosmetic switch between two relatively similar alternatives. There is also the real possibility that a party could promise a series of policies or a policy direction then change and set an alternative policy course when elected. This is a danger in majoritarian systems where winners are granted control of the governing institutions. For accountabil ity to exist parties must heed their electoral mandates and there is not guarantee that they will once in office. Even when parties present a platform to the voters and are elected, pursuing real policy change is not guaranteed under majoritarian systems.


93 One of the reasons that real policy change is hard under majoritarian elections is the nature of the elections. Since parties that are elected with a plurality not a majority, pursuing a major policy change might create real discord in the country. Sign ificant changes in policy direction (implementation of national health care, the elimination of farm subsides, or tax increases for example) requires consensus and majoritarian the tendency to think in terms of party warfare rather than of government in the interests of the whole nation. That again is encouraged by the majority system of election. In each (Lakeman 1984 49) Majoritarian elections are a zero sum game in this respect. This can create an adversarial situation where partisan bickering is more important than consensus building and compromise. For the out party in the legislature, the goal becomes winning a majority of seats and taking over the government. This means that there is little incentive to help the government pursue its policy agenda. I have laid out potential models that link majoritarian and proportional systems to support and the potential pitfalls that each of these systems has that could confound those expectations. To summarize, proportional systems seek to increase the amount of ideological representation in the decision making processes of the legislature. They do s o by offering a variety of interests the chance for representation through proportional elections. Governments operate by consensus and pass a policy agenda that is widely accepted by the population leading to higher levels of support. Majoritarian syste ms seek accountability and efficiency by creating pressure on voters to choose between one of two large political parties. The winning party is


94 disproportionally represented and given control of the government, where they are free to pass their policy age nda. Voters hold parties and governments accountable for their actions and elect or remove them from office based on their successes or failures. Accountable, effective government leads to higher levels of support. Both systems have potential pitfalls, o r breaks in the chain between the election and support. Proportional systems do not necessarily guarantee that parties will get along or make decisions by consensus. In fact, there might be more incentive for the parties to horse trade, logroll, or barga in rather than make decisions with the overall good of the country in mind. There is also a lack of accountability in these systems. The election occurs then parties form a government. That government may or may not be what the electorate pictured when they were voting. When parties are engaged in a constant process of bargaining and negotiation, the stability of the system is at risk. If governments can not hold themselves together or are made up of pieces based on political convenience rather than tr ue policy or ideological consensus, there is difficulty passing a coherent set of policies. When both process and policy suffer, we should not expect high levels of support among the population. Majoritarian systems do not necessarily lead to higher level s of support among the population either. One reason is that majoritarian systems elect governments through pluralities, not majorities of the vote. It is not at all uncommon for a winning party in a majoritarian system to win a majority of seats with on ly 38% to 45% of the vote (Shugart and Wattenberg 2001). Under this type of system many interests, parties, and minorities and women are often shut out of the system. The trade off is that governments are supposed to be able to pass effective policies wh ile remaining


95 accountable to the population. There are a number of reasons why this might not happen. There is no guarantee that an elected government will follow the policy agenda that they lay out prior to the election, or that the out party will pursu e an appreciably different set of policies. Due to the plural nature of the elections, even when governments win control, their policy agenda has been rejected by a majority of the voters. Finally, the zero sum nature of majoritarian elections can lead t o bitter infighting and partisanship among the parties in the system. If these elements manifest themselves, it is difficult to see how majoritarian systems garner support. These problems with majoritarian and proportional systems have led some scholars t conditions that generate refo 27). It is very rare that a More countries are turning to mixed choosing an elected representative for their localities, but also provide for some e lement 1) There are differences and variation in mixed member systems in terms of whether the proportional or majoritarian elements are more or less prominently featured. A brief review of th ese differences of is important to this study, because New Zealand and Italy, the cases I mixed alternative. I believe the models, drawn from literature on proportiona l and majoritarian systems are still appropriate to study New Zealand and Italy because of the directional nature of the respective changes and the way my models are written. This is


96 a point I return to when discussing why I believe New Zealand and Italy make good cases for study. First, I engage in a brief discussion of the different types of mixed members systems and the potential confounding elements they present to existing electoral system theory. The Best of Both Worlds? The Mixed Member Alternativ e Much of Chapter 2 has been dedicated to drawing a link between increased proportionality and support as well as increased majoritarianism and support. The theoretical bases of these links are derived from literature on electoral systems that concentrate on the historical alternatives: proportional and majoritarian electoral systems. Since the early 1990 s, more than 30 countries have chosen an alternative to these two choices, preferring instead a mixed electoral system. These systems combine some form of a single member district tier with proportional representation in a second tier (Hainmueller and Kern 2008, 213). This section surveys the different types of mixed systems and discusses what potential confounding impacts that these types of electoral systems can have. Defining and Survey ing the Types of Mixed Member Electoral Systems Mixed member electoral systems are based on mixed principles of representation. That is, these systems employ some form of proportionality and majoritarianism at the sa me time. Shugart and Wattenberg (2001) classify mixed member systems as a type of multiple tier electoral systems. Multiple tier systems allocate seats in two or more overlapping sets of districts where each voter casts one or more vote that are used to assign seats in more than one tier (Shugart and Wattenberg 2001 10). Mixed member systems include one tier that distributes seats nominally, while the other tier allots seats by lists. The difference between the two tiers is based on the type of vote ca st by the


97 citizen and how it is used to assign seats. Under nominal voting, voters cast a ballot for a candidate by name and seats are awarded to those candidates based on the number of votes they receive. List votes are collected and allocated among sev eral candidates placed on a list determined prior to the election by some kind of political party, organization, or coaliti on (Shugart and Wattenberg 2001, 10 11). Generally, in a mixed member electoral system a voter is provided the opportunity to cast s eparate votes in each tier 17 The nominal vote and the list vote may be cast by voters and allocate seats in a number of different ways. The nominal vote is usually cast in a single seats are most often allocated throu gh a plurality vote, although some systems use run off elections 18 Electoral formulas in nominal tiers are usually majoritarian, although there are examples of countries that use the single nontransferable vote (SNTV) 19 Systems that use SNTV are not the norm and most scholars define a nominal tier as also contain a list vote, a tier that overlaps the nominal tier. Nearly all list systems use a form of PR, although there a re cases where the list tier is either totally or partially majoritarian 20 Closed lists are the primary choice for list tiers that use PR. In this type 17 There are cases, like in the Italian Senate from 1994 to 2005, where a voter only casts a nominal vote. 18 Run offs are where a second round of e lections between the top two vote getters are held if no candidate wins a majority of votes in the first round. Albania, Georgia, Hungary, and Lithuania have all used this type of election in their nominal tier (Shugart and Wattenberg 2001, 11). 19 Example 20 systems. Countries that use complete majoritarian system in their lis t tier are the US electoral college, while countries that use a partial form of majoritarian lists include the former systems of Mexico and South Korea and the current systems of Chad and Cameroon (Shugart and Wattenberg 2001, 12).


98 of list, parties, not voters, determine the order of party representatives 21 There are alternatives t o the closed list, some systems order their lists on the basis of which (Shugart and Wattenberg 2001 12). Best loser systems do not have an actual candidate list in the li st tier. Instead, parties nominate candidates in the nominal tier. After determining the winning candidate from the nominal tier, the seats that a party may obtain from the list are taken from its pool of nominated candidates who were not able to win the ir district. Best loser systems actually use the nominal tier as nominating districts for their list tier. Seats are then allocated to parties in a multi seat district, but are awarded to candidates within the parties based on the success they had in the ir respective nominating distric ts (Shugart and Wattenberg 2001, 12 13) 22 In terms of voter choice, the best loser systems are the same as a closed list since voters do not have the opportunity to determine the order of the list. Best loser systems do pr ovide incentives for candidates to maximize their popularity within their districts, even in districts were their likelihood of winning is small, because the most popular candidates will be placed on the list tier ahead of their less popular co partisans. These are examples of some of the variation that exists between nominal and list tiers and how those tiers allocate seats to parties or candidates. Mixed systems do combine some form of majoritarian and proportional principles in the same electoral syst em, but they also tend to favor either proportionality or majoritarianism in their overall impact. Based on this, scholars have identified two main subtypes of mixed 21 There is no reason tha t list tiers could not include open lists where voters are allowed a preference vote among candidates on the list, but there are not current examples of countries where this takes place. 22 Variations of the best loser have been used by Mexico (1964 1976) the Italian Senate (1994 2001),


99 systems: mixed member majoritiarian (MMM) and mixed member proportional (MMP) (Shugart an d Wattenberg 2001 13). The main difference between the two subtypes is whether there is a connection between the nominal and list tiers in the electoral system. If there is no link between the tiers, than the typical majoritiarian enhancement that the l argest party receives in the nominal tier is not going to be impacted by the proportional allotment from the list tier. In these MMM systems the largest parties retain their advantage, which is a key feature of majoritarian systems in general. MMP system s instead make the list PR tier a priority, so large parties either do not receive augmentation in overall seat allotment or a smaller one than they would under MMM. Linkage can happen in a number of different ways. Some systems have no linkage whatsoev er. For systems that have nominal/list linkage, one way is for votes to be transferred from the nominal tier to the list tier. Another way is for the number of list seats a party wins is in someway based on the number of nominal seats it wins. Those sys tems with no linkage are perhaps the easiest to explain. Here the parties seats are not adjusted at all based on the votes cast or seats won in the nominal tier. A party takes whatever seats it wins in the nominal tier and adds that to the number it wins in the list tier to reach its overall total. These are clear cut examples of MMM systems. MMP systems and some MMM system have some form of linkage between the nominal and list tiers either through seats or votes. If seats are linked, the number of se ats a party wins in the nominal tier directly impacts the number of seats that party is able to win in the list tier. These systems apply a formula to the two tiers, either in regional sub districts or through an entire jurisdiction, which establishes a p

PAGE 100

100 tot al (Shugart and Wattenberg 2001, 14). Examples of these are compensatory type systems. Here, a party wins a percentage of seats nationally that is determined by a PR formula. The party reaches its entitlement of seats by claming the seats it won in the single member districts and supplements them with the number of candidates off the list that is necessary to acquire its overall entitlement. The Ge rman and New Zealand elections use this type of MMP. If it is the votes (rather than seats) that are linked, then votes for the party list are not the sole determinant of how list seats are allocated. Here votes are adjusted by the shift of votes from th e nominal ti er (Shugart and Wattenberg 2001, 15). This linkage may use a positive or negative transfer of votes. For example, in Hungary the votes of candidates who lose SMD races in the he other hand, the nominal 23 There are three different linkages between list and nominal tiers: no linkage or separate list and nominal tiers, seat linkages between nominal and list tiers, or a linkage, whether positive or negative, between votes cast in the nominal and list tiers. Shugart and Wattenberg (2001) classify different types of mixed systems based on these 3 types. MMM systems have no linkage between the list and no minal tiers. MMP systems have some sort of seat linkage between the two tiers. Finally, these scholars identify those systems with a vote linkage between tiers as MMM with a partial compensation (Shugart and Wattenberg 2001 15). These are broadly defin ed sub types, but any country that adopts a mixed member system has a significant number of 23 2001). Their system is something that I discuss in a l ater section and in a Chapter 6 of this work.

PAGE 101

101 choices to make at the outset. It is important to survey some of these choices before moving forward with the potential effects of mixed systems. If a democratizin g country or a country reforming their electoral system chooses to adopt a mixed member system that country has to make a few decisions. Shugart and Wattenberg (2001) label several different important variables that need consideration. The linkage betwee n the tiers is identified as the most important variable for determining the level of proportionality, because it largely determines whether a system is going to be MMM or MMP. A second important choice is the percentage of seats set aside in the list tie r In MMP systems the number of seats set aside must be an adequate percentage of the total to guarantee a high level of proportionality 24 The percentage of seats set aside in the list tier is even more important under MMM than under MMP. This is becaus e the two tiers are separate and therefore can not rely on any additional PR allocation. A third important consideration is the magnitude of the tier list For instance, in New Zealand there is a single district nationwide containing 120 seats, while in Germany the lists are divided into 16 districts with an average magnitude of 41 (Shugart and Wattenberg 2001, 19). In MMM systems, the magnitude of the list tiers often contain a large, nationwide district as in Russia (225 seats) and South Korea (46 seat s). Some countries use regional allocation, as in Japan that has parallel PR allocation of three fifths of the whole number of representatives in 11 regional districts with a magnitude ranging from 7 to 33 (Shugart and Wattenberg 2001, 22). The threshold a party needs to acquire seats in the list tier is another consideration in mixed 24 nominal tier seats allocation is likely to be, the great the share of list tier seats that needs to be set a side

PAGE 102

102 systems. 5% of the national vote is the most widely used threshold among countries using mixed systems. A final variable to contemplate is the number of seats and formula of the nominal tier. those, plurality formula is the most common, although some countries use a form of run off if no candidate receives a majority in the first round of balloting (Shugart and Wa ttenberg 2001 18 24). As the previous survey demonstrates, there are a wide range of choices that countries can make when adopting mixed electoral rules. Scholars suggest that these s can certainly choose how much of each world they want. Mixed member systems lend themselves to myriad variations in their details, but as a class they all offer the combination of both majoritarian and proportional principles 001, 24) While mixed member systems remain a popular choice among electoral systems scholars, some have suggested that they are not quite sure why that is the case (Bowler and Farrell 2006). This might be because the research is divided on the political consequences of such systems. Scholars disagree about the impacts of having both some form of PR and SMD electoral systems combined in a mixed system. Some, like Shugart and Wattenberg, believe that mixed (2001) S till other scholars hold that the combination of PR and SMD transforms the effects of each system due to spillover or contamination effects (Cox and Schoppa 2002; Ferrara, Herron, and Nishikawa 2005; Hainmueller and Kern 2008 ; Nishikawa

PAGE 103

103 and Herron 2001 ) 25 The potential contamination effects are important to this study for several reasons. Do Mixed Systems have Contamination Effects? P revious sections describe the mechanical and psychological effects associated with majoritarian and proportional electoral systems. This literature suggests that electoral rules, through their mechanical effects, shape the psychology of voters and parties encouraging them to behave in a strategic manner. Mixed systems have different rules and different mechanical effects, so it is reasonable to expect that there will be different psychological effects associated with them as well. Parties and voters maybe encouraged to behave strategically in a number of different ways. There may be no contamination effects, where scholars can treat the SMD and PR segments of the electoral system as separate, with no contamination effects. Another potential effect is for the SMD portion of the system to influence the PR portion. Alternatively, some hold that the PR effects might spillover into the SMD section. electoral systems represent a social laboratory in which effe cts of different types of electoral systems can be studied in isola 576) Mixed systems seem to offer these scholars the advantage of testing these different rules, while holding constant difficult control variables like cult ure, political history, or socioeconomic development (Moser 1999; Moser 2001; Moser an d Scheiner 2005). 25 level, when the behavior of voter, a party, a candidate, or a legislator in one tier of the election is demonstrably affected by the institutional rules employed in the other tier. At the aggregate level, contamination is observed when a particular outcome produced in one tier (like the number of political parties) is affected by the institutional features of the other (Ferrara, Herron, and Nishikawa 2005 8 9).

PAGE 104

104 Ferrara, Herron, and Nishikawa (2005) argue that these beliefs are only correct if the two tiers are independent from one another, where each tier of t he electoral system behaves as expected without being influenced by the other. This means that the mechanical and psychological effects associated with such systems should be readily apart in the respective nominal and list tiers. The SMD tier of the ele ctoral systems should produce two party systems at the district level, with multiparty systems in the PR tier. Psychologically a voter might split their ticket, voting strategically for one of the smaller party in the PR tier. limited opportunity to win, preferring the more permissive PR tier (Moser and Scheiner 2005). These scholars generally believe in the indepe ndence of the tiers in mixed electoral systems with a two addendums. The first is that while the SMD tier of mixed systems encourages two party competition at the district level, spillover from the PR tier can help explain why the effective number of part ies is greater at the national level than at the district level. The promotion of smaller parties with a geographic concentration as viable national parties happens because of the PR tier. This PR projection, scholars claim, is the reason that the two pa rty competition is evident at the district, but not the national level (Moser and Scheiner 2005) 26 A second addendum is that political context matters a great deal when discussing the impact of mixed systems (and electoral systems in general). The most d istinct variation in the political impacts of mixed systems is due to the degree of party institutionalization, rather than the 26 This is a point that Nishikawa and Herron (2001) disagree with, suggesting instead that there is indeed

PAGE 105

105 differences in electoral institutions the mselves (Moser and Scheiner 2005 ). This is especially true in post communist democraci es, where institutions consistently produced unexpected consequences, in contrast to the products of more consolidated democracies with highly institutionalized party systems (Moser 1999, 2001). Post communist states, like Russia and the Ukraine, experien ced high levels of party fragmentation in the SMD tiers. Scholars that hold there are minimal contamination effects from mixed systems believe that the low level of party institutionalization is to h more established parties contagion from the combination of PR and SMD rules was not the primary cause for the fragmenta 598) The belief that there are no, or limited, contamination effects in mixed member systems is not universally held. Others argue that contamination effects can be apparent in two main ways: The SMD tiers contaminating the PR tier or the PR tier contaminating the SMD t ier. the PR tier is a minority view among scholars who study such effects. Duverger (1986) member system, hypothesized tha t the two party competition typical of the SMD tier would create, parties would persist despite the existence of a more permissive proportional tier. Still other scholars believe the opposite Cox and Schoppa (2002) argue that the tendency to two party They theorize this is the case because small parties have an incentive to run candidates

PAGE 106

106 in districts nation findings are support ed by works from Nishikawa and Herron (2001) and Ferrara, Herron, and Nishikawa ( 2005). Nishikawa and Herron note that the number of parties in ure SMD systems. Ferrara, Herron, and Nishikawa find similar results after examining a larger number of mixed membe Hypothesis is questionable. Consequently, new democracies considering the mixed system should recognize that it is not a simple combination of the best features of SMD and PR, but a hybrid system with uncertain consequences for the development of political parti 2001, 83) The existence of contamination effects is by no means unive rsally accepted. Hainmuller and Kern (2008) call findings with evidence of contamination effects, that parties might places candidates in districts where they expect success or because they have a degree of organization strength. If this were true, the positive correlation between the presence of district candidates and PR vote shares would be question able (Hainmueller and Kern 2008, 215). The same authors also poin t out that estimating contamination effects caused by the presence of district candidates is only possible if the parties are selective and choose some districts to run in but not others. In most countries, most small parties routinely stand for office in all or almost all districts. Identifying treatment effects is impossible without variation in the treatment variable

PAGE 107

107 (Hainmueller and Kern 2008 215). The overall impression that scholars give about mixed member systems is hazy. An interesting study by Bowler and Farrell (2006) polled electoral system scholars and their data revealed a strong preference for mixed systems (especially MMP), but no clear reason why. The kinds of things that drive us to prefer these systems are, moreover, not well establish ed in the literature. We can show this is the case of MMP, where to the extent that we can actually identify any values associated with support for this system, it is not entirely clear how these should fit. For instance, one reason for liking MMP is that we value consensus over clear Another value that drives support for MMP seems to be a dislike of asking voters for preference rankings per se and a liking for one representative per distr ict. But these may not be well short, then, the preference for MMP among our sample of experts is (Bowler and Farre ll 2006, 455) Since two of the ca ses I choose to study in this work chose mixed member alternatives, it is important to bear these potential consequences of mixed systems in mind. Testing the Competing Models: New Zealand and Italy New Zealand and Italy make for interesting cases to res earch for several reasons. First, the two countries have a special place in the political science lexicon. Prior to style majoritarian 1984 16) Italy, in almost direct contrast, was considered prior to 1993 to be the quintessential proportional system (Lijphart 1984). their respective types. It is generally thought that the inherent problems with the systems le d to their collapse and there is some truth to those claims. But, New Zealand operated successfully for over 100 years with majoritarian government and Italy for 50

PAGE 108

108 years with a proportional system. This sec tion details what the problems attributed to both of the systems, were how they were able to overcome those problems, and what precipitated the change s in the electoral system s unicame ral legislature, single member districts, and a strong two party system. New representation of the two major political parties, Labour and National, and the under representation of third parties. Thes majoritarian system. representation of the two major political parties, and the under representation of thi rd parties. Between 1935 and 1993, Labour and National combined to win 1494 of the 1516 seats, roughly 98.5% of all seats contested (Bale and Roberts 2002, 11). This type of majoritarianism also led, in several instances, to one of the two major parties winning majorities with less than a majority share of the national vote. For example, in 1978 National received 39.8% of the votes and 55.4% percent of the seats contested. National again had minority support in 1981, with 38.8% of the vote share, but 5 1.1% of seats. Labour had this distinction in 1984, when they won only 43% of votes 27 and 6 0% of seats (Vowles 1995 97 ). New Zealand, historically, also had severely under represented third parties. There were numerous occasions where third parties rec eived significant support, but 27 It should be noted that when I cite percentages of the national vote, I am referring to aggregated percentages across multiple single member districts. There is no national percentage that had an importance other than as a gauge of national attitudes. When I say importance, I mean that there were wide.

PAGE 109

109 were dramatically under represented in parliament. The first push from a third party for electoral system change came in 1954 when the Social Credit Party formed and picked up 11% of the national vote, but no seats in the pa rliament (Bale and Roberts 2002, 11). As support for Nation al and Labour waned in the 1970 s, other third parties would form, win support, but no seats in parliament. The Social Credit Party and the Values Party (a forerunner to the Green Party) received varying levels of support throughout the 1970 s but no seats. Ironically, the reforms were not initially suggested by third parties, but by one of the two major parties (Labour), after National won a majority of seats, with a minority of the national vote. National joined their major party counterpart in 1984, after Labour also had a government with a minority of support from the people. The two parties created the Royal Commission on the Electoral System in 1986. This Commission issued a report in the s ame year that suggested that New Zealand adopt a German style mixed member system (MMP) to fix the apparent problems of over and under representation. Growing voter diss atisfaction throughout the 1980 s came to a head in 1990, when third parties received 1 7.7% of the national vote, but only 1% of the seats and in 1993 when third parties received 30% of the vote and only 4% of seats (Banducci, Donovan, and Karp 1999, 537). First past the post democracy in New Zealand ended in 1993, with 85% of the electora te rejecting majoritarianism in a referendum. Disproportionality is certainly part of the reason that New Zealand changed its electoral system, but it is not the whole story. If disproportionality was the main reason, then it would have changed systems a long time before 1993. At that point, the country had been using the same electoral system successfully for 100 years. It is my

PAGE 110

110 contention that New Zealand was able to follow model (7) which details how a majoritarian system can actually facilitate suppo rt. Successive parties in government increased accountability through processes like the party manifesto, group consultation, and a caucus system. With this accountability came generous social policies that the public enjoyed. After 1984, consecutive La bour governments followed by a National government disbanded many of the processes that promoted accountability. Exacerbating this was an effort by both parties to dismantle the social safety net that citizens relied upon. Only when this happened did dis proportionality become problematic and become a major reason for the electoral system change (Mulgan 1995) Italy, as a case for st udy, is interesting for different reasons. Any political scientist could identify with the following statement, The casual observer of Italian politics may be forgiven if he or she judges the oft quoted fact that Italy has had about 58 governments in about as many years and the chief problem with Italian politics is government instability. More precisely, Italian governments have lasted an average of no more than 11 months in office, a figure without parallel in the established democracies of the modern world. (Sanchez 2002, 259) Here Italy seems to be a perfect example of what the critics of proportional systems warned of. G iovanni Sartori, in several writings between 1947 and 1992, described what he saw as the problems with the Italian party system. Sartori stressed one significant factor above all others, polarization 28 The polarization that existed in Italy was described as tri polar. The three poles were represented by the extreme right (the MSI, Movemento Sociale Italiano, a neo fascist group), the extreme left 28 Poles are the political space occupied by a party or group of parties that are closely allied. (Pasquino 2004, 3)

PAGE 111

111 (represented by the Parito Communisto Italiano, PCI), and the center which was occupied by the Christian Demo crats (DC) (Passquino 2004, 3). The tri blocked because of the ideological distance between the three poles. Coalitions were able to form, but not without the Christian Democrats. The cause of this was that they were positioned ideologically in the center and they were the largest of the political parties. The Neo fascists and the Communists could only agree on bringing down the government, but could never get together to form an a lternative government because in no system do the extreme left and right join a coalition t ogether against the center The pole made them indispensable to formation of any an all government coalitions. Also, they had to prevent either of the two extreme parties from entering a coalition out of fear of a regime crisis (Pasquino 2004 3). Many problems associated with the Ital ian government occurred as a result of this every government since the 1940 s, which caused a lack of alteration in power. A lack of alteration in power led to several problems, inc luding no turnover in the political class, few new public policy ideas, and a hardened relationship between members of government and interest groups and therefore increased s ystem corruption (Mershon and Pa squ ino 1997 ; Sanchez 2002 261). Early discussio ns about changing the ele ctoral system began in the 1970 s, because of these factors, but actual reform did not begin for several reasons.

PAGE 112

112 The main reason that parties opposed the reform was because they saw it as detrimental to their interests. To adopt s ome form of majoritarianism meant that some of the smaller political parties would literally have to vote themselves out of existence (Sanchez 2002, 261). The larger parties had their own interests to protect as well. The Italian Communist party was esse ntially an unelectable party under majoritarian rules considering the context of the Cold War, but it did not want to give up its status as an opposition party to the government. The Italian Socialists determined that majoritarian reform would force them to relinquish their status as a power broker, and require them to give up the benefits of such a position and choose a coalition partner prior to the election (Sanchez 2002, 262). Because of these internal misgivings, scholars have suggested that an exter nal shock to the Italian political system was required to implement electoral system change. Su ch shocks did occur in the 1990 s. The first of these was the fall of the Berlin Wall and the end of the Cold War. With the end of the E ast/ W est conflict, the I talian Communist party became a viable electoral party. As its electoral prospects increased, their opposition to reforms declined. Secondly, there was a large judicial operation on within the Italian political system. This operation was the driving forced behind the destruction of the largest political party, the Christian Democrats (Sanchez 2002, 262). Ultimately, in 1993, the voters forced a change from the previous h ighly pro portional system to MMM through a referendum. Again, like New Zealand Italy was able to overcome these problems through processes of representation. Political parties operated and dominated politics in Italy

PAGE 113

113 and were able to ingrain themselves into publi c life to such an extent that representation that belied an inherent stability where parties received the same percentages of votes election after election, and the gove rnments while fragile, were just a reshuffling of the same parts. Policy was not the ideal, grand bargain policies imagined by the models. Rather, policies were micro clientelistic servicing small constituencies, thereby increasing representation. The c hanges that occurred had to do with the process of representation. Parties began to lose hold o f their control during the 1970 s, and the clientelistic relationships exploded into outright corruption during the 1980 s. The corruption led to a decline in su pport for the existing party system and caused its collapse during the early 1990 s. Conclusion Both New Zealand and Italy chose a mixed member electoral system as an the discussion of mixed member systems is important to any study that attempts to link the products of electoral systems to support, it especially important to mine. I frame the theories as what we might expect from increasing proportionality on the one h and, and increasing majoritarianism on the other. I contend that the outputs of electoral systems impact the number of political parties, the type of governments that are formed, the type of policies they produce, and as a result the level of support citi zens offer. Mixed systems, if their political impacts can be determined, might be a middle ground between complete PR and total majoritarianism. For instance, in the cases I select, New Zealand and Italy adopted different mixed systems to address the dif ferent problems facing their respective democracies. New Zealand had a single house legislature,

PAGE 114

114 party system. They switched to a MMP system in order to increase the level of proportionality to allow more parties into the system i n order to create coalition governments that would produce more widely accepted policies, while Their system was flooded with political parties, a blocked political syste m led by a corrupt dominant party, and a government seemingly unable to pass decisive and effective policy. They adopted a system of MMM in order to create a system of bipolar competition, reduce the number of political parties, and cohesive governments t hat could pass effective policy while maintaining some type of PR. I hold that these cases are examples of countries moving toward more proportionality or more majoritarian systems and are therefore good cases for testing the models I draw. The potential contamination effects of mixed systems are important to keep in mind, because the models themselves are drawn from a different literature on pure proportional and majoritarian electoral syst ems. Considering this, Chapter 5 and Chapter 6 on New Zealand an d Italy after the change of electoral system will pay additional attention to the potentially confounding effects of mixed systems. Chapter 3 and Chapter 4 detail these changes in both New Zealand and Italy Chapter 3 uses the majoritiarian model to descr ibe how New Zealand was able to make majoritiarian government work for close to 100 years. Chapter 3 also details the changes in process and policy that took place during the 1980 s and how that led to a drop in support among the population preceding the e lec toral system changes. Chapter 4 does the same for Italy and uses the proportional model to describe how Italy was able to overcome the problems of their polarized party system and what changed during

PAGE 115

115 the 1980 s. These changes led to a loss in support a nd a change in electoral system as well. Chapters 5 and Chapter 6 discuss the specific changes in electoral systems in Italy and New Zealand respectively. Chapter 3 and Chapter 4 also detail the expectations of the changes, the products of the changes, a nd the resulting impacts on support.

PAGE 116

116 CHAPTER 3 THE RISE AND FALL OF MAJORITARIAN GOVERNMENT IN NEW ZEALAND Introduction In Chapter 2, I draw a model that represents how majoritarian electoral designs can lead to higher levels of support among the popul ation. To summarize, majoritarian electoral systems have mechanical effects that penalize third parties. These mechanical effects increase the psychological pressure on voters to steer their votes toward one of two major political parties and away from t hird parties who have little chance of winning seats. The purpose of this, in the majoritarian ideal, is to manufacture a majority for one party who would then form a government capable of effective action. Majoritarian systems are not designed to be pro portional, so representation is not as important. The key process in this type of system is accountability. Parties run for office based on a series of promises, are elected, and then held accountable by the voters based on their actions. Process and po licy are inextricably linked in this model with governments promising policy agendas and given the ability to act on them through manufactured majorities. Governments either pass those policies or are held accountable and replaced. If governments keep th eir promises, then citizens reward them with further terms, and support for the system rises. There are numerous places where this model could break down. In situations where third parties can win seats, like in instances of regional strength (Sartori 198 6), the winning party is able to form single party governments with low levels of electoral This might not be problematic in one election, but if this circumstance is con tinually

PAGE 117

117 repeated, majoritarian governments could have difficulty cultivating support. Another r elated potential set back would be if the process of accountability failed. It is conceivable that alteration in office might not be enough to change the polic y directions in a country. If successive governments continued with unpopular policies then accountability is lost and support could potentially suffer. New Zealand makes an excellent case for study for several reasons. The country is an example of both how majoritarian government can succeed and how it can fail. Prior to the 1980 s New Zealand was well known for its stable democracy, a solid two party system, and high levels of political support in areas like political participation, voter turnout, and s atisfaction among citizens (Mi tchell 1969; Vowles 2002). The country w as able to maintain this high level of support by keeping to the model I draw in Chapter 3. The majoritarian electoral system, aided by majoritarian constitutional designs, continually produced single party governments. Parties also maintained unique processes of accountability and generous social policies that voters enjoyed. Support was lost beginning in 1984, when the fourth Labour Government reversed decades old processes and bega approach to public policy. These programs were extremely unpopular, but when Labour was replaced by National in 1990, the new government de epened and exacerbated the roll back of public spendin g. This loss of accountability led to a decline in support and provided the context for the change in the electoral system. Chapter 3 details these changes to provide evidence for how the models I draw in Chap ter 2 can work and how the models can fail and is broken down in the following manner. The first section details why New Zealand is an excellent case to test the

PAGE 118

118 majoritarian model and describes their constitutional stru cture. The rest of Chapter 3 follows the path of the model and shows how two par ty governments were created and how those governments were accountable to the citizens New Zealand: A Model of Majoritarian Government 1 model 16) The majoritiaran model of democracy includes the following facets: The concentration of executive power: one party and bare majority cabinets Fusion of power and cabinet dominance Unicameralism A two party system A one dimensional party system Plural syste m of elections Unitary and centralized government Unwritten constitution and parliamentary sovereignty Representat ive democracy formulations laid out by Lijphart, but they do (wit h the noted exceptions) include the following characteristics. The first is the concentration of executive power through one party and bare majority cabinets. From 1935 to 1993, New Zealand politics has been dominated by two major political parties: Labo ur and National. Between 1935 and 1993, Labour and National combined to win 1494 of the 1516 seats, roughly 98.5% of all seats co ntested (Bale and Roberts 2002, 11) 2 All cabinets have been made up of one of these two main political parties, who have bee n supported by clear majorities in 1 Lijphart (1984 2 The Social Credit party was able to win a single seat in 1966 and 1978. The same party claimed two seats in 1981.

PAGE 119

119 Parliament. The majority party has been able to boast between 50 and 60% suppo rt of Parliament (Lijphart 1984, 17). Secondly, New Zealand cabinets simply dominate politics and governing. Parties are disciplined within parliament and concentrate power within the cabinet, and although they rely on the votes of confidence of the members, for various reasons (which I will explain) cabinet power in New Zealand is supreme ( Levine 1979 ; Lijphart 1984 ). is also unicameral, lacking an upper house since 1950. As I mentioned earlier, New Zealand is also a two party system where, since 1935 3 only Labour (1935 1949, 1957 1960, 1972 1975, 1984 1990) and National (1949 1957, 1960 1972, 1975 1984, and 1990 199 3) formed cabinet governments. Like most two dimensional or a centripetal party system. The country lacks ethnic cleavages (short of a very small Maori minority population), and its politics are domina ted by a center right party (National) and a center left party (Labour). These parties are elected through a series of plurality elections. New Zealand had 95 single member districts elected through a first past the post electoral system until the electo ral system change in 1993. The single member districts have made disproportionality a recurrent feature of New Zealand democracy. government. New Zealand has no federalism to s peak of and all major powers rest with the national government 4 These powers are enhanced by the fact that New Zealand 3 This lasted until the electoral systems c hange in 1993. 4 for 6 provinces that had significant regional powers. These w ere abolished in 1875 (Lijphart1984, 19)

PAGE 120

120 series of acts dating back to the original Const itutional Act of 1952. This feature essentially grants parliament (and in this case the majority party in parliament) total control over all aspects of the country since there are literally no legal hurdles to avoid. One scholar went so far as to say tha most part are passed through parliament although the parliament has made use of the referendum over the years to put to the public various issues from liquor laws to compulsory military service 5 order in a manuscr ipt about electoral systems and support. But, the make up of New aid out in Chapter 2 Consider the model here: (7) (+) Majorita ) number of support for government. The items which may be thought of as under the rubric of process or procedural elements relate d to support are accountability, efficiency, and stability. These are naturally going to be affected by the constitutional order of a given country. For example, there are a number of constitutional designs that could affect accountability like a second House (Senate, House of Lords, Upper Chamber) or a Supreme Court. I 5 Also, the electoral reform act of 1993 was passed due to a referendum. The use of referendum is still well within majoritarian rules, since it is the parliament that determines when and if it can be used. There is no extra parliamentary procedure for placing binding referenda on the approval.

PAGE 121

121 like to think that accountability is weakened anytime someone else can be blamed. During the policymaking process, when there are two houses, who do you blame when things go wrong? The same can be said of efficiency. When another step is added to the policy making process it is more difficult, thus less efficient, to get anything passed (Tsebelis 1990). Additional variables like these make efforts to link electoral systems to support m uch more difficult. New Zealand does not have any extra institutional structures to speak of. They have single member districts, one house, and no written constitution. They literally have nothing in the way, and that makes for a wonderful case to study the theoretical links between electoral systems and support. It may seem natural to use New Zealand as a case to study for why majoritarian governments fail. Many scholars have done just that (Aimer and Miller 2002; Renwick 2010 ; Siaroff; 2003; Scheine r 2008; Vowles 2000 reasons for why it failed. This is not all that surprising since there is an obvious contradiction between calling a syst representative democracy in consecutive sentences. The fact that New Zealand so over represents two major political parties would seem to aid those that argue for a more proportional model of democracy (li ke the model I drew in Chapter 2 ). But, New Zealand did practice the majoritarian model for over 100 years, something must have worked. The paradox between total parliamentary control, over representation of major parties, and representative democracy is not so difficult to fathom if the model that I drew is considered.

PAGE 122

122 Making Majoritarianism Work in New Zealand This section will examine the institutions that made majoritarian electoral systems work in New Zealand. These informal arrangements helped New Zealand (for a time) government 6 As a reminder, Schattschneider was part of a consortium of political scientists who were attempting to improve the deficiencies in the two party sy stem in the United States. They believed that several elements needed to be in place for a party system to be responsible. These include a two party system where the two political parties offer the public a consequential choice. The public should expect that their choice will be informed based on information provided by the political parties. When a political party wins office the party then implements the program that it promised prior to the election. Finally, in the subsequent election the party in power stands for reelection based on the extent to which they enacted their pre election promises and the degree to which they were successful 7 (Schattschneider 1950). For a period of over 50 years, New Zealand was able to make this model work and practic ed government that produced something closely resembling the model I drew earli er. This portion will demonstrate how they did so. Producing a Two Party System: Dis p roportionality It is the nature of every democracy that uses single member plurality electi ons to none are cr 75) New Zealand is a perfect example of 6 Ironically, this party model was designed to improve party performance in the United States, but New disciplined t wo party system and one house that the United States does not have. 7 The opposition party also puts forward an agenda for if they were to be elected to office.

PAGE 123

123 this quote. Between the years 1943 and 1993, New Zealand conducted 18 ele ctions and never failed to produce a majority party in parliament 8 even though in only three of those elections did one party acquire more than 50% of the vote 9 This type of arrangement favored the two largest parties, Labour and National, to such an ex tent that between 1935 and 1993, they combined to win 1494 of the 1516 seats. This is roughly 98.5% of all seats contested (Bale and Roberts 2002 11). The supremacy of the two part ies is displayed in Table 3 1 The first conclusion that can be drawn fro m this chart is that the electoral system worked. First past the post electoral systems are designed to manufacture majorities where they might not otherwise exist. Because of this, Labour and National dominated the political landscape in New Zealand for over fifty years, and were the benefactors of disproportional results. The years in which third parties contested elections nation wide resulted in much more d isproportional results. Table 3 1 display s several years where the disproportionality eclipses 10% for the winning party. These include 1954 (11.95%), 1966 (11.4%), 1975 (15.61%), 1978 (15.63%), 1981 (12.28%), and 1993 10 (15.4%) in favor of majority National governments. The Labour party was the beneficiary in 1972 (14.81%), 1984 (15.94%), and 198 7 (10.76%). It should not be surprising that these years also coincide with significant efforts by third parties and there seems to be a 8 A majority party in parliament has more than 50% of the total seats and is able to form a sin gle party majority government. 9 Since New Zealand had no party vote until 1996, the votes for these parties are aggregate vote totals from the single member districts. 10 It should be noted that this is the only year in which there was more than 10% dispro portionality for both major political parties when the Labour party finished second in seats won, but was the beneficiary of 10.85% disproportionality result.

PAGE 124

124 correlation between third party success (in terms of %, not necessarily seats) and higher levels of disproportionality Table 3 2 displays this relationship. From 1954 1993, third parties were only able to win nine seats while creating very high levels of disproporitonality in several elections. Model (8 ) accounts for this occurrence because majoritarian electoral syste ms are designed to systematically over represent the two largest parties at the expense of smaller, third parties. The elections in New Zealand created single party governments consisting of a Prime Minister and Ministers from one of the two largest polit ical parties, Labour or National. From 1960 1993, either Labour or National accounted for 100% of all cabinet positions, and one party maintained a majority in Parliament throughout their 3 year term (Vowles 2002 411). At this point let us consider the first part of the majoritarian model: (8) ) number of New Zealand clearly meets the first two conditions, with more information necessary to confirm the third. The electoral system was clearly disproportional, with mechanical effects that punished third parties. This is evidenced by the ability that Labour and National had winning seats to parliament. This was a design of the electoral system and t he distortions themselves were not enough to directly cause a decline in support. It may have been that, Electoral fairness, it may be noted, was not the issue. New Zealanders, like citizens in other Westminster democracies, were generally tolerant of the distortions of nationwide party vote to which first past the post gave rise. They were prepared to accept a system which gave a parliamentary party which did not have a majority to a majority of the votes, so long as that party behaved according to accep ted democratic norms. (Mulgan 1995 83) Yet, these distortions were enough, however, to raise sporadic grumblings from third parties, like Social Credit, who were persistent challengers to the locked in major

PAGE 125

125 parties. This alone was not enough to put majo r reforms on the agenda, until the late 1970s. This time period also happens to coincide with two elections where National spurious majorities 11 3) an outcome that is in at least in part responsible for electoral system reform. At this time, Labour added to its manifesto a promise to create a Royal Commission to study the efficacy of the first past the post electo ral system. watershed moment for electoral reform in New Zealand, since this was the first time since 1920 that serious discussion of changing the system was undertaken (Vowl es 12 during these two election cycles and the minor policy promises made afterward, first past the post elections continued in New Zealand continued until 1993. Electoral system reform di d fairness, simply did not motivate the public at large. When the Royal Commission issued its report in 1986, it largely fell on deaf ears. However, b y 1992 New Zealand ers voted overwhelmingly in favor of changing the electoral system from the then current arrangements. Majoritiaran electoral systems are designed to produce disproportional results, for 100 years. This is because prior to 1978, New Zealand was able to attain governance that was 11 12 Geoffrey Palmer was newly elected to Parliament in 1978, and wrote a book called Unbridled Power (1979). In this book Palmer sharply criticized the New Zealand Constitution and advocated that one third of Parliamentary seats be selected by proportio na l representation (Palmer 1979, 155 157).

PAGE 126

126 accountable, efficient, and representative. In other words, they succeeded in both policy and process. As I stated earlier, elections do not produce support they can only create the conditions for support, because theoretically they are only the first link in the chain. Disproportionality alone is not enough to lose support among the population, because there is a trade off, between less proportional repres entation, for more accountable, stable, and efficient government. By examining what worked in New Zealand (in terms of elections, policy, process, and support), what changed, and the extent to which that change led to a different electoral system will hel p shed light on the role that support has on inter regime change. The reasons that majoritiarian government continued without change and the reasons for why the public mood rapidly changed course must be examined. The previous section demonstrates that th e first several conditions of the majoritarian model have been met. The electoral system created mechanical and psychological effects that provided the conditions for stable alteration between one of two political parties. These parties were able to prov ide all of the Prime Ministers and cabinet positions for their full three year terms. This is not enough evidence to demonstrate that these governments were accountable. Accountability in New Zealand was maintained through a series of processes including : use of the manifesto, group consultation, and a caucus system. These processes aided vertical accountability. The next section describes the uses of these processes and is followed by a section that describes how the use of those processes led to widel y accepted policies. Fostering Accountability: Process in New Zealand Prior to 1984 At first glance, New Zealand onstitutional order does not appear to present an ideal situation for accountability and the representation of po licy demands. With no

PAGE 127

127 writ ten c onstitution, an electoral system that produces single party governments and no formal restraints on political power once in office, New Zealand governments would appear to be free to do whatever they wish. But, New Zealand political parties followed a series of informal institutions that made the process of doing business much more accountable, representative, and stable. These processes contributed to supportive attitudes in New Zealand prior to elect oral system change in the 1990 s. Accountability i s cited in my models as the key beneficial process of majoritarian governments, where a government is either held responsible for its failed policies, or re helped aid accountabi lity. One practice that helped to increase accountability was the use of extensive and detailed manifestos. Normally a manifesto, or election policy, is a collection of party principles and policy statements that they intend to put into action if elected to office. The individual policies usually cover all areas of government purview, including most if not all departments and min isterial positions (Mulgan 1990, 1995). Traditionally, in New Zealand, the party policy is drawn up by members of the party t hrough commit tees. For National it was the J oint Policy Committee, and for Labour the Policy Council. Both parties drew ideas and interest from members of the party, whether in government or outside. Once the party manifesto was agreed upon, there was a general understanding that it would command the support of all party members, since it provided a basic framework for what the party was and what it stood for. The party manifesto or election policy is thus a lynchpin of democratic party government. As an electoral mandate, it helps to articulate the choice offered to the electorate, providing an obligation on elected government to enact the policies they have committed themselves to enact. At the same

PAGE 128

128 time, by placing an obligation on all party member s, it reinforces the guarantee to the voters that a party in power will keep its commitments through a disciplined parli amentary majority. (Mulgan 1990, 13) The use of the manifesto was seen as a promise to the voters, a way of keeping the government accou ntable. The voters were handing over a tremendous amount of power to one political party to the other during the election with the understanding that the government would carry out what it promised to do. This was a commitment that Prime Ministers in New Zealand took very seriously, and is perhaps best represented by the actions of Prime Ministers during the 1970 s. s to make sure that they were in line. This practice was a nuisance to ministers who saw their policy advice overlooked in favor of the, 14) 13 Robert Muldoon, who followed Kirk as National Prime Minister in 1975, also made a point to follow the explicit promises explicit pledges that his party made in the manifesto. The first, an assurance to abolish Dam project in 1981, although many in the party felt that the economic situation in the country was too perilous. This led to several prominent resignations among members 13 Kirk was especially sensitive to the power of the manifesto and the public mandate because of circumstances over the planned 1972 South Africa Springboks Rugby Tour. Kirk (and shadow Prime Minister Robert Muldoon ) made no electoral commitments in the mandate, but both promised during the campaign not to interfere with the tour. When Kirk canceled the tour after the election, he received heaps of public criticism from the opposition and the public who considered h is promise to be a binding commitment. Rugby is the national sport in New Zealand with South Africa and Australia their main rivals. Canceling the game would be akin to calling off a World Cup Soccer match between England and Germany or canceling the Sup er Bowl. Kirk took great lengths to explain this change, and made a point never to break an explicit promise made in the manifesto (Mulgan 1990).

PAGE 129

129 of parliament from the National party. But, what it did demonstrate is that, to Muldoon, th e authority of the voters super seded all else. The use of the manifesto was not the only way in which informal instit utional practices fostered accountability, New Zealand parties also made use of a caucus system that guaranteed the input of the back benches into policy making. As I stated earlier, the p arliamentary party in power in New Zealand dominated the constitut ional order. One of the ways that the party in power kept from becoming an elected dictatorship was the nature of the caucus, and the frequent consultation with the backbench outside of the government. New Zealand arliament is small (ranging from 80 9 9 over the years) and the parties governed with strict party discipline. Because of these two aspects the governing caucus, which usually numbered from 40 55 members, was particularly close knit. Backbenchers were routinely in contact with members of the government and ministers, whether formally or informally. Leaders that were successful were able to foresee what their colleagues would accept and support. Generally, caucuses functioned according to the values of other important New Zealand institution s, rugby and the military, which valued teamwork and loyalty above all else (Mulgan 1995). The caucus served as a powerful instrument of democratic accountability. It provided significant political input for backbenchers who were more in touch than minist ers with party opinion and with the views of ordinary voters. Caucus thus helped to keep ministers responsive to the public and provided a useful antidote to the influence of the bureaucracy. (Mulgan 1995 85) Routine communication between the back bench and cabinet was an important factor in increasing both accountability and representation. Another process that

PAGE 130

130 increased representation was the expectation that all groups that had a sectional concern would be routinely consulted on policy making. After World War II, interest groups became a prominent component of political life in New Zealand. Political parties and successive governments developed a set of connections with these sectional interests similar to those in other W estern democracies ( Mulgan 1995 ; Vowles 1992 ). Various interest groups could expect that they would be consulted on any particular policy that might affect them directly. This happened to sense t hat some groups became officially integrated into the state. For example, groups concerned with the industrial sector, like labor groups, were incorporated into the national system of conciliation and arbitration. The producer boards that had statutory c ontrol over marketing and sales were a result of interests within the agriculture sector. The influence of labor and agriculture can be seen in the influence they had over post and une mployment insurance as well as binding arbitration procedures and mandatory union membership, reflect the influence of that group. Agricultural interests were also the beneficiaries of favorable policies considering that New Zealand had one of the most pr otected import markets in the world, placing heavy tariffs on imports especially in the agriculture sector. This protection for farmers was coupled with generous subsides from the government that protected and supported the agriculture based export econom y (Mulgan 1990) consolidate power in the hands of a single party based on disproportional results.

PAGE 131

131 Governments, when elected with a smaller percentage of the population, like there are in New Zealand are necessarily not as popular among the citizenry. When coupled with consolidated power, it is easy to see why scholars would criticize these types of n its systems based on a series of processes that considered and involved citizens in the manners of government. Through the presentation of a detailed manifesto that governments felt obligated to follow, citizens could feel that their elected representat ives made promises and could be held accountable if they did not. Accountability was also fostered through a caucus system where Members of Parliament in the back bench were in close contact with the government, and could expect to have their opinions hea rd helped to increase representation. Finally, interest groups were considered in the making of policy, and particular interests groups who had a traditional standing, like farm and labor, were actually absorbed into the state apparatus. All of these pro cesses helped to make a system, whose rules were perhaps not inclined to good governance, more accountable and representative. Labour and National governments, prior to 1984, used these processes to good effect but as I state in the majoritarian model a ccountability in majoritarian systems requires results. To put another way, parties can communicate with the public as often as they like, but unless they deliver policies that the public wants, then they are going to lose support. In New Zealand citizen s wanted an activist government and generous social policies. The next section details how Labour and National delivered on those policy demands and is followed by a description of how this led to higher support in New Zealand.

PAGE 132

132 Policy in New Zealand: Pre 1984 Policy in New Zealand, prior to 1984, played a significant role in the maintenance of first past the post electoral system. Welfare and social policy in New Zealand has a long history, dating all the way back to the 19 th century. Besides Germany, Ne w Zealand was one of the first countries in the world to enact comprehensive social legislation. For example, women were given the right to vote in 1893. 1894 saw some of the first labor market reforms including mandatory arbitration and a pension progra m that was put in place for the poor. New Zealand had mandatory secular public education all the way back to 1877. Public hospitals were first constructed in 1846, and the Hospital Act of 1885 used taxpayer money to help subsidize health care for the poo r. By the turn of the 20 th century a feeling among the general population developed 14 (Sautet 2006 573). From the late 19 th century to the 1920 s, New Zealand was one of th e five wealthiest countries in the world 15 New Zealand developed this economy on the back of an abundance of fertile land, a smaller population, and agricultural exports based on a favorable trade relationship with England. This successful economic perfo rmance policies and the overall economy as well. In the 1930 s New Zealand extended both the welfare state and state controls over the economy as a result of the crumbing w orld economy due to the Great Depression. Between 1935 and 1949, the first Labour government enacted a series of micro 14 Additional social policies included a minimum wage policy for families in 1908. 15 Measured by gross domesti c policy (GDP) per capita (Sautet 2006).

PAGE 133

133 policies including farm price guarantees, high level s of industry regulation, and import restrictions. The goal was to create and maintain a high level of economic growth and full employment (Mc Clintock 1998; Silverstone, Bollard, and Lattimore 1996). The social welfare system was expanded in all areas in cluding housing, education, health 499) through the Social Security Act of 1938 16 The 1950s and s were a period where the welfare programs were deepened and strengthened. Af ter 1949, despite seemingly rightward leaning inclinations, successive National governments took a moderate approach balancing the demands between the market and the economy. Rather than eliminating the welfare and state interventionist policies of the pr evious Labour governments, National instead added to or strengthened existing state activities (McClintock 1998). By the 1970 s, the state played a major role and had enormous influence on all sectors of the economy including output, investment, and employ ment (Duncan 1996 393). During most of the twentieth century, This economy devoted to the values of security and equality succeeded remarkably well. Aided by preferential access to the British market, rich natural resources and a highly literature popul ation, New Zealand generally Second World War up to 1976, involuntary unemployment was virtually unknown. With further advantages of a mild climate, marvelous scenic beauty, low popu lation density and a lifestyle centered on outdoor that they enjoyed the best quality of life on the globe. (Nagel 1998 228 229) 16 For more details on this act see Silverstone. Bollard, and Lattimore (1996, 454 )

PAGE 134

134 Because of these advantages, the main job of the maj or two political parties, when in government, was to divide the rich economic spoils to their population. They did so, with generous welfare, health insurance, education, and retirement benefits. Two major changes in the world wide economy would threaten position, and the incoming National government of 1978 turned to state intervention to solve these difficulties. The international community underwent two major changes in the earl y 1970 s. In 1972, Britain joined the European Ec onomic Community. This changed the favorable the first of several oil shocks occurred when OPEC increased the cost of oil significantly by reducing the supply of oil pu t into the world market. In 1978, the newly installed National government pursued a series of policies designed to reduce its balance of own energy supply. The major late 1970 s. Financed largely by massive overseas borrowing these projects were These projects represent, in a way, The default reaction to most problems was decisive, if not effective, government action. In the 1970s and early 1980 s, the government increased income tax rates and devaluated the currency several times. D espite these efforts, inflation remained over 10% from 1974 to 1982. Unemployment increased steadily over this time period, but GDP rose only 0.2% a year, lagging well b ehind the world wide average (Nagel 1998

PAGE 135

135 229). The height of government intervention reached an apex, when the Muldoon government enacted wage and price controls in 1982. As stated, the normal way of conducting policy in New Zealand had reached a turning point by 1984. At this time the country was suffering from the worst economic conditions since the Great Depression, and its active, interventionist governing philosophy was being questioned by its economic and political elite. New Zealand by this time had one of the most activist governments in world, which had produced an expansive welfare state, and the most controlled economy among democratic countries. The incoming Labour party was going to change that in 1984, by radically reshaping the role that government had played for the previous 50 years. In short, they were going to decision was a drastic departure from what New Zealanders had come to expect of their govern ment in terms of policy, and would be a major reason for the loss of support that had been built on the back of 50 years of consistent interventionist policy. Policy Changes New Zealand, prior to 1984, followed a particular pattern of public policy. New Z ealand was governed with certain values that became deeply embedded in the culture of politics in New Zealand. These values included the expected relationship between citizens and their government. To put it briefly, New Zealanders came to expect that th e owned enterprises, and regulating the private economy through import contols and other industrial sub 326) The extent to which successive Labor and National governments changed these values and reversed these policies after the election of 1984 can not be overstated.

PAGE 136

136 The economic reforms taken up by the Labor governments of 1984 and 1987 and the National government of 1990 were in five areas: tax reforms, labor market reforms, trade reforms, monetary policy reforms, and fiscal reforms (Sautet 2006, 575). Tax policy was changed in 1985 from a progressive income tax to one of the flattest income tax rates in the world supplemented by a value added tax (Darwall 2003). The top marginal income tax rate was cut in half from 66% to 33 % between 1985 and 1988 (Sautet, 576). A value added tax called the goods and services tax (GST) of 10% was also implemented. The high regulation of the labor markets was also drastically changed from a system that required union membership and made arbitration between workers and owners compulsory. The Employment Contract Act of 1991 eliminated both and replaced centralized bargaining with individual bargaining. It gave employees and employers a choice between individual or collective employment contracts and no third party arbiter needed to represent parties in contract or employment disputes (Sautet 2006, 577). Trade was also dramatically liberalized after 1984 as th e Labor government abolished policies that promoted exports and reduced tariffs (Challies and Murray 2008 236). One of the most striking examples was the drastic reduction of government subsides to industry and agriculture from 16% of government sp ending to just 4% (Darwall 2003, 64). Farm subsides received the largest cuts which fell from NZ$1.2 billion in 1983 to NZ$206 million in 1990 (about 670 million to 115 million adjusted US exchange rate) (Darwall 2003, 204). New Zealand is really the only coun try in the world to successfully scrap farm supports.

PAGE 137

137 Monetary and fiscal reforms were also undertaken during this period. In 1989 the Reserve Bank Act gave the Reserve Bank of New Zealand independence from the control of the government. Monetary policy in the past could be adjusted to suit whatever political problems the government thought it had at the moment. The 1989 Act took control of the money supply out of the hands of the politicians with the intention of taking a longer view when it came to th e monetary health of the country. The Reserve Bank used this independence to drastically reduce interest rates through the tightening of the money supply (Sautet 2006, 578). The Labor government of 1984 also started several policies with the goal of chan ging the fiscal direction of the nation by balancing the budget. This was done through deep cuts in social services like welfare, health, and unemployment spending as well as through the privatization of formally state owned enterprises. The tight contro l over budgets was highlighted by the 1989 Public Finance Act. This law forced government departments to show that spending should be increased because cost increases outweigh productivity gains rather than the old formula of indexing spending increases ( Sautet 2006, 580). The Fourth Labour government was summarily dismissed from office during the 1990 elections, but this did not end the cut back of the social welfare state. National, rather than softening these policy trends actually exacerbated the roll backs. The neo liberal policies of the 4 th finance minister Roger Douglas, and the reforms taken under his counterpart in the reforms could be seen as c omple mentary to the Labour reforms of the previous 6 years. National continued the deep cuts in social spending especially in the areas of housing,

PAGE 138

138 health, welfare, and pensions and created new fees for health and educat ion services which previously had been free. National maintained the tax on superannuation, or old age pensions, a direct violation of a pre campaign promise. They passed the Employment Contracts Act in 1991, which stripped unions of compulsory membershi p heir pre election manifesto (Silverstone, Bollard, and Lattimore 1996). These are only some examples of the policies enacted between 1984 and 1993. T here were in fact literally hundreds of new policies that were designed to drastically reshape the relationship between the government, the market, and the people 17 and New Zealand should not be considered just one of many counties that also enacted reforms during this time period. Economic restructuring was far more radical in New Zealand than in other democracies. Globally, reforms in a free market direction were hardly unusual in the 1980 s. Besides Thatcherism in Britain and Reaganism in the United St ates, many smaller democracies moved to deregulate, liberalize and privatize their economies. But New Zealand accomplished a more extreme change than any other democracy, not only I relation to where it started but also with respect to where it ended up. (Nagel 1998, 224) Simply put, New Zealand transformed from one of the most highly regulated economies in the world to one of the most open in less than 10 years. Procedural Change New Zealand practiced several informal procedures that helped foster accou ntability and representation and tempered its majoritarian electoral system. These processes were the use of the manifesto to increase accountability, and the employ of the caucus system and group consultation to foster representation. The 17 For a detailed list of laws passed and economic r eforms undertaken see Silverstone, Bollard, and Lattimore (1996, 24 28) who li st over 100 such laws.

PAGE 139

139 incoming fourt h Labor government of 1984 tossed those traditions aside and the National government of 1990 failed to re implement them. As a result, those institutions and confidence i n the existing system vanished. The voters saw the mandate and manifesto as contracts. The negative reaction many New Zealanders had to the economic reforms of the mid to late 1980 s was that there was no warning or mention of the policies in the Labor gov election manifesto. Some saw the party manifesto of 1984 as a deliberate deception, where reformers inside the Labor party felt that the public would not support the party if they was hazy on economic policy and lacked specific proposals. They did not come out and say that they would undertake such a program of reform, but they did not come out and say they would not ensuring that investment flows into the areas of economic activity which are of most ny believed the government would continue the interventionist policies of the previous N ational government (Mulgan 1990, 15). Fueled largely by contributions from business elites who loved the new economic policies and the unwillingness of the National Par ty to rollback reforms if elected, the Labor party was returned to power in 1987. It is here that they showed true contempt for the mandate. In this election cycle the party did not bother to publish a manifesto until two weeks after the election. This document was a brief booklet rather than the

PAGE 140

140 lengthy manuscript that was historically appropriate. The Labor government did not even bother to follow the scant details in the booklet, and broke numerous pre election promises (one was to hold a referendum on electoral reform). National was swept into fast pace of reform initiated by Labor in the previous six years. After the election they proceeded to make deep cuts in social services including unemployment insurance, while New Zealand had their highest levels of joblessness since the Great Depression. They also broke an explicit promise to eliminate the unpopular tax on superannuation imposed by Labour. [These th ree election cycles] challenged previously existing conventions about election policies and manifestos. Lengthy manifestos, which had previously been treated as binding on an elected government in every detail, were said to be irrelevant. A government no w claimed to govern as it saw fit, regardless of specific election promises, and subject only to the ultimate verdict of the voters at the next election. (Mulgan 1990, 17) The manifesto was not the only casualty of the fourth Labor government. They also a bandoned the caucus system and group consultation that existed previously. One ministers supported cabinet decisions. This eliminated the consultation between the cabinet and backbench and was designed to eliminate dissent. During the Labor government of 1984 to 1990 out of a caucus of 56 or 57, 20 were members of the cabinet; in addition another 8 held offices (like party whip) who were obligated to vote with the ministry. This meant that a majority of 11 to 14 members of the ministry controlled the entire caucus and could pass laws with a very minimum of support in parliament (Mulgan 1995; Nagel 1998).

PAGE 141

141 Labor also did away with the practice of group consultatio n. The market reforms undertaken during this period were not a reaction to any sort of demand made by the public or interested group. Many saw the swiftness of reforms as anti democratic, a style that market liberals favored. Reformers believed (support ed by noted economists like Mancur Olsen) that interest groups used process of democratic consultation to secure benefits for themselves. When governments followed those narrow interests, ed. It is not a coincidence that governments the business elite pointed to as models were the non demcratic regimes of Chile, Hong Kong, and Singapore. The Labor government took the position that they were going to pursue policies regardless of the polit ical opposition from groups they normally consulted. Roger Douglass, the father of the economic Government is responsible and answerable to the nation as a whole; not to particu lar 19) All of these processes that aided accountability were changed, but the actions of Labour and National hurt accountability in a different manner as well. Accountability in a majoritarian system is about being able to replace a policy deficient government with an alternative who is offering something different. When National continued the cut backs in government spending and protection, they merely continued the policies th at Labour began. Voters were left with this alternative: they could vote for the existing Labour government or they could vote for a National government who criticized the Labor policies of the past. When National broke their pre election promises and me rely accelerated the changes started under the previous government, voters were left with

PAGE 142

142 nowhere to turn. The main process of accountability, which majoritarian systems are precipitated, was lost. The majoritarian model predicts that increased majorita rianism leads to a two party system and the creatio n of more accountable governments that are able to pass a package of policies that the public demands, leading to higher support. The previous three sections deal s with how the electoral system led to bot h policies and process the public supported and the changes successive governments took to change both policy group consultation, and the caucus system led to policies tha t were widely supported by New Zealand citizens. At this point, the next step of the model would be to demonstrate that there were higher levels of support. Unfortunately, measures that indicate support existing data show is a downward trend in supportive behaviors like participation that culminate in very low levels of trust and satisfaction in the early 19 90 s. This is in line with what I describe as a change in policy and process New Zealand: A Picture of Declining Support I use a series of indicators to demonstrate that supportive behaviors in New Zealand declined over time. Many of these indicators have to do with the level of participation. Chapter 2 describes how participat ion is tacit evidence of support. For instance if you join the membership of a political party, you are demonstrating some level of support for that positions and performance of that political party, but also a commitment to existing democratic practices as well. When a citizen goes to the polling place and votes they may do so out of anger or a feeling that the government is not performing well, but they are also affirming a belief in the process of voting, thereby

PAGE 143

143 supporting democratic practices. Final ly, a person may confirm through surveys that they support the government or the practice of democracy in their country. New Zealand demonstrated a decline in all of these indicators over time. Party Membership Party membership is a good indicator of the level of support that the public has for the two party system. Joining a party and maintaining membership in that party demonstrate suppo rt for the system. In C hapter 1, I describe how much of the end to support what we gh membership rates in the 1950 s and 1960 s. In 1954 party membership stood at 250,000 a full 51.4% ratio of party members/voters. In 1960, these numbers stood at 246,000 and 44.2% resp ectively. These decades were a high water mark for the party, but from 1972 to 1981, the National Party increased its membership in every election year reporting: 1972 (150,000), 1975 (196,000), 1978 (200,000), and 1981 (200,000). National lost significa nt numbers over the next 6 years reporting 133,000 party members in 1984 a nd 100,000 in 1987 (Vowles 2002, 416). This represents staggering decline, representing half of the party membership. The Labour party never maintained the total numbers of party m embers a s the N ational party, but unlike their counterparts they actually improved their member ship from the highs of the 1960 s. The years 1963 to 1981 saw an upward trend in numbers of Labour party members. In 1963 party members stood at 13,122, and imp roved in the years following: 1972 (12,313), 1975 (13,997), 1978 (45,838), and 1981 (49,837). Similar to National, Labour saw significant membership flight in the ensuing years: 1984 (38, 337), 1987 (19,203), and 1990 (19,203) (Vowles 2002: 416). Labour, like National, saw half of their party rolls disappear during the

PAGE 144

144 1980s. Total party membership follows a similar pattern. While never reaching the highs in the 1950s and 1960 s 18 party membership increased from 1972 to 1981 only to decline in the years following. The rise of party membership and M/E ratios increased in the following years: 1972 (164,313/9.3), 1975 (209,997/10.5), 1978 (245,838/11.8), 1981 (271,837/12.5). After 1981, party membership and M/E ratios fell in both 1984 (182,337/8.1) and 19 87 ( 130,701/5.5) (Vowles 2002, 416). Party membership both overall and among the two largest parties were in decline following 1981. I argue this is due to the changes in process and policy that occurred following these years. When accountability and so cial welfare policies were in place, party membership was high only to drop when they were replaced. Labour and National also lost their command of the vote sh are over time. During the 1970 s Labour and National commanded and average of 69.1% of the total voter. That n umber was 67.9% during the 1980 s, and fell precipitously to 58.25% during the 1990 s (Vowles 2002, 416). While there is some evidence of a long term dealignment, the near 10% drop in mean sc ores from the 1980s to the 1990 s provides evidence that something else was a contributing factor outside just a long term trend. I have held that support fell because of the actions taken to undermine accountability taken by successive Labour and National governments when they passed a series of policies that the public did not support. Turnout Another indicator of declining support is the falling levels of voter turnout. Turnout follows party membership in that there was a decline in the early 60 s, followed by a rise 18 Total party membership was 280,760 a full 21.9 M/E ratio (% of the total electorate who are party members) in 1954 and 279,020 and 20.1 M/E ratio in 1963.

PAGE 145

145 throughout the 1970s into the 1980 s, only to decline after the Fourth Labour go vernment was elected in 1984. Figure 3 1 displays these results. Turnout was 85.5% in 1984 and dropped to 76% in 1990 and 1993 (76.6%). There are a few reasons that I believe this is important. 76% and 76.6% re present historic lows in turnout where the closest result was 78.6% in 1966. The turnout % in 1984 represented almost a 30 ye ar high with 85.5% (Vowles 2002, 416). The close to 10 year drop, I believe, has to do with the changes in policy and process in the years between 1984 and 1993. It is not coincidental that a rise in turnout was halted then reversed at the same time that the changes in policy and process were taking place. I believe the decline represents an indication of a loss of support for the majoritarian system. Further evidence includes the mean scores per decade that show a dramatic dec line in turnout during the 1990 s. Me an turnout scores include: 1960s (81.5%), 1970s (80.7%), 1980 s (82.1%), and the 1990 s (76.3%) (Vowles 2002, 416). Ther e is no evidence here, unlike the vote percentage scores of the two largest parties, of any long term tur nout decline. In fact the 1980 s saw the highest mean scor es of any decade since the 1950 s. The 6% drop in turnout from the s to the s is f urthe r evidence that early 1990 s were a turning point in support in New Zealand. These results are displayed in Figure 3 1. Survey data in New Zealand was virtually non existent prior to the 1990 s. The only data on supportive attitudes are in the reports of s cholars. These schol ars report that during the 1960 s and s over 60% of New Zealanders reported satisfaction, confidence, and trust in their government, institutions, and elected officials (Lamare 1991; Lamare and Vowles 1996; Levine and Robinson 1976; Vowles 1994, 104; Vowles

PAGE 146

146 2002). From these reports, I consider 60% to be a high water mark for support in New Zealand. A brief examination of the 1993 New Zealand Election Study shows that support for governing institutions fell across the board. Trust in Labour and National would be better if there were more available survey data in order to create a complete picture of attitudes in New Zealand. Unfortunately the data available are only sporadic and incomplete. But, from the data that do exist, part icipation in elections, party membership, % of the vote for the two major parties, and the episodic survey data, a picture of an increasingly dissatisfied and disaffected New Zealand citizen can be drawn. Changes in policy and the loss of accountability we re the causes of declining support in New Zealand. No connection has been made, however, between this falling support and the change of electoral system that took place through referendum in 1993. What remains is to test empirically the extent to which t hese changes in policy and keep the old first past the post system. The following section uses survey data from the 1993 New Zealand Election study to measure which eleme nts of policy and process have relationships with electoral system choice. Testing Policy and Process Effects on Electoral System Choice In order to test the historical data I lay out in the previous sections empirically I use data from the 1993 New Zealan d Election Study. This study was a round of surveys that asked New Zealanders their opinions on a range of political issues. There are

PAGE 147

147 many questions included in the study that are relevant to my analysis. These queries include how a person voted in the referendum on the electoral system in 1993, some survey items that can be used to capture elements of policy and process, as well as measures of support. I employ these measures to determine how support, policies, and on on the choice of electoral system through the use of a binomial logit model. Dependent Variable The dependent variable in the model that I run is a survey item that asks the respondent how they voted in the 1993 referendum on the electoral system. The as dummy variables where a 0 signifies a vote for FPP and a 1 indicated a desire for MMP. The ques tion captures how a voter feels about the status quo. Voters who are satisfied with the current conditions would most likely continue to support FPP, while a vote cast for MMP signifies a desire for change. In this model I test the impa ct that various policy, process, support, and demographic variables have on the choice between electoral systems. Independent Variables In order to test the effect that policy has on electoral system choice, I use a series of survey questions designed to c apture citizen feelings about the social welfare state. following services. Th ese include: a decent living standard for all old people, decent standards of living for the unemployed, a job for everyone who wants one, housing for

PAGE 148

148 those that can not afford it, free health care for all, and a free education from pre school to polytechn ic and university levels 19 Those who agree that the government should supply these services are generally supportive of a social state. This grants an opportunity to test the relationship between support for a social welfare state and the desire for elec toral system change. There are other survey questions that could be used as possible alternative independent variables. They are questions that attempt to determine the level of interest in similar issues (health, education, or welfare), as well as survey items that seek to discover support for additional spending in the same areas. I did not use these questions is their consistent use throughout the surveys, they appear in ever NZES from 1993 2008, which allows for true comparability over time. The spending and interest questions are not included in all surveys. I also hesitate to use questions related to interest and spending due to the possibility of measurement error For instance, why such divergent opinions as those who advocated the complete abolition of the government service, or those who believe that it is a right that shou ld be guaranteed. the word spending might bring among responders. Frankly, citizens might instant recoil g of government sponsored programs might be lost. 19 Responses have been recoded so that a 1 represents the opinion that the go non response

PAGE 149

149 The questions I do use have the advantage of capturing the feelings and beliefs that New Zealanders have about the social state in general. I argue in p revious sections that MMP was adopted in part bec provided in New Zealand was dramatically and quickly taken away. The policy variables, represented by these questions, gets that the general feeling that the New Zealanders have about the role that the state has in the areas of housing, health, welfare, social security, unemployment, and housing. Those respondents who believe that the state has a responsibility in those areas might be more inclined to vote for MMP. There are another series of independent var iables that I use to capture policy: where respondents place themselves on the left/right scale and support for lower taxes. dimension upon which party competition takes place (Dow ns 1957). Voters, in the estimation of researchers, have solid understanding of what the terms mean and are able to place themselves accurately on the s cale ( Dalton 2010 ; McAllister and White 2007 ). This does not mean that individuals have a clear unders tanding of complicated theoretical concepts like socialism, communism, or liberalism. But, the scale can be used as a general summary of the particular issue cleavages that define the political left/right scale has been used ns (Dalton 1985; Klingeman, Hoffbert, and Budge 1994) and as a meta issue that embodies the, (Inglehart 1990, 273) Scholarship has also shown that where individuals place themselves on the left/right

PAGE 150

150 scale is strongly correlated with their positions on the important political issues facing their country (Dalton 2006). I consider the left/right sc would you put your v iews on the scale below where 1 is most left and seven is most 20 A summary variable like the left/ right scale offers a nice comple ment to the education, pensions, an d housing. Between these variables, I believe an accurate ascertained. The final policy variable I have is a measure that gauges the respondents support for reducing indicate whether you would support such a policy or oppose it, and if you feel strongly about it: 21 I add this variable for two reasons. The first has to do with the one of the main policy goals of the Labour/National governments from 19840 1993, to drastically reduce tax rates across the board. These tax cuts were financed by drastically reducing social services. The second reason is that support for higher taxes could be seen as support for a government that is more active and has to pay for more 20 responses are coded as non response. 21 response.

PAGE 151

151 generous social welfare policies. It might be easier to say that you support total government funding of education or welfare benefits, but to say that you strongly disagree with reducing taxes, this means that you are willing to pay for those benefits. A ll total there are three main components of the policy related independent var iables. There are a series of six survey questions from the 1993 NZES that estimate whether or not a respondent believes that the government should be responsible for providing social services. These services include benefits for the elderly, unemployed, those with difficulty affording housing, education, providing jobs, and universal health care. The next is a left/right scale variable I use as a summary variable that encapsul taxes, indicates both the level of commitment a person has for paying for the social programs they want as well as a rejection of the tax cutting policies of previous governm ents. Between all of these variables the model will include both specific and positions impact how they feel about getting rid of FPP in favor of an alternative system. In addition to these policy variables, I also include measures that represent the processes that fostered accountability under FPP in New Zealand. I present hi storical evidence earlier that certain processes were tossed aside by the Fourth Labour Governmen t and the ensuing National government. The processes that were practiced prior to 1984 fostered accountability and prevented wide spread beliefs among the population that their extreme majoritarian system would not descend These included the government respecting their mandate by making pre election promises presented in a manifesto and sticking with them

PAGE 152

152 following the election. Governments also consulted frequently with interest groups and with their own caucus, especiall y the back bench, to maintain the connection with citizens between elections. These processes created a general sense among the population that their government was upfront, listened to, and connected with them. When these processes were jettisoned, this connection was shattered contributing to the loss of support and a change in electoral system. To measure what I have labeled process variables, I use a series of survey questions that ask about whom respondents think the government should listen to befor Parliament when they are considering major new laws. How important do you The NZES the n offers a series of different factors, but I only include three of these in the 22 I include these particular questions because they represent the processes that I describe in the historical narrative. The and relates to accountability of electoral processes in majoritarian systems. I detail how majoritarian systems, in the ideal, provide for vertical accountability. An important component of vertical accountability is a party stating prior to the election what th ey are going to do and then bench 22 These questions have been recoded so that numerical increases indicate strengthening beliefs that response.

PAGE 153

153 forward indicator of the practice of consultation that took place between elected officials and interest groups who were to be affected by potential legislation. These questions, There are also a few other independent variabl es that I use in the model. One key please indicate how much you agree or disagree with the opinions expressed in the following statements: You can trust the governmen Responses range from 1 to 4, with a 1 representing strong disagreement and a 4 strong agreement. Trust is my indicator current issues in politics and their f eelings about the current government and politics. The remaining independent variables are standard controls variables including: age, education, income, and gender. I have multiple expectations for the effect that these policy, process, and control varia The next sections detail these expectations, report the results, and then discuss what the results tell us about the majoritarian model and how it can be confounded. Expect ations The expectation that I have for the policy, process, and independent variables mirrors th e historical case that I offer in p revious sections In terms of policy, broadly speaking, I anticipate that those who are generally more supportive of a socia l welfare state will be more likely to reject FPP for MMP. The reason for this belief is these programs were essentially gutted under successive administrations from 1984 1993. New Zealanders could no longer count on the type of care and security from th eir

PAGE 154

15 4 government that they had received in the past. Programs that were pillars for decades were suddenly yanked away and I believe that this contributed to the desire for some type of change. In this case, that meant support for a different electoral syst em. In terms of self placement on the left/right scale, I believe similar patterns will hold. I consider the left/right scale to be a summary indicator of policy positions. Since the policies passed by the Labour and National government from 1984 to 19 93 could be considered conservative in nature. I believe this incr easing conservatism will correlate for a preference for maintaining the status quo Favoring the status quo will be signified with a preference for FPP. more a respondent is opposed to lower taxes, the more they will favor MMP. This is due to the tax cutting policies that precipitated cutting social services under Labour and National government from 1984 to 1993. The majoritarian model requires that gov ernments pursue policies that the public demands. When governments fail to do this, a link in the majoritarian theoretical model is broken. I argue in Chapter 3 that policy and process are inextricably linked in majoritarian systems. Here governments ar e held accountable to the voting public based on the policies they promise prior to the election. Scholarship on New Zealand also argues that governments relied on other processes, between the elections, that also fostered accountability. The transformati on of processes also helped generate a context where electora l reform could occur. I expect and relevant interest groups will be more apt to vote for MMP. This is due to the argument that the ca ucus in general and the back bench more specifically help the

PAGE 155

155 government to stay in touch with the feelings of their constituents. This could be confounded a bit due to the unpopular nature of the National government at the time. Since parties und er FPP began to ignore interest groups, my general expectation is that as a respondent feels more strongly that an MP should listen to interest groups will be more inclined to vote for MMP. Finally, I expect that as individuals feel more strongly s should keep their pre election promises (listen to their mandates or follow their manifestos), the more likely the will be to choose MMP. The reason that I expect these results has to do with the changes in processes that occurred after 1984. Prior to this date, governments used these processes in order to foster accountability, which is a key component in the majoritarian model. When successive government discarded these processes, support for the existing order fell. When accountability is lost, the majoritarian model fails. Among the other key control variables, I have one particular expectation when it comes to the variable measuring trust. It has been a consistent contention of mine that low levels of support can have an impact on institutional m aintenance. I expect that if a respondent does not trust the government then the odds that they will vote to maintain the status quo will decrease. In this case, that means a vote for MMP over FPP. The next section displays these expectations in a serie s of hypothe se s. Hypotheses This is a brief section that summarizes my expectations in simple terms. My hypotheses for the logit models and the post estimation odds ratios are as follows:

PAGE 156

156 1. As a respondent reports increasingly conservatism (left/right scale ), the odds of voting for MMP decrease. 2. If an individual is increasingly opposed to reducing taxes, the odds of voting for MMP will increase. 3. As feelings that the government should be responsible for certain social services (jobs, pensions, welfare, unempl oyment, housing, and education) the odds of voting for MMP will increase. 4. should listen to relevant interest groups, and should listen to their mandates, the odds of voting for MMP increase. 5. When trust in the government increases, the odds of voting for MMP decrease. The following section describes the results of the models and the accuracy of the hypotheses. Results Results for the model testing process and policy variables on ele ctoral syst em choice are found in Table 3 3 Hypothesis 1 states: as a respondent reports increasingly conservatism (left/right scale), the odds of voting for MMP decrease. The logit model indicates s trong evidence to support H ypothesis 1 The odds of vo ting for MMP decrease 36.5% with every unit increase in conservatism. This result is significant at .01. This is statistical evidence that helps to confirm some of the presumptions th at I make earlier I hold that from 1984 to 1993, Labour and National governments dramatically altered relationship between society and the state. This included: the drastic reduction in social services, flattening tax rates, cutting farm subsides, and removing worker, environmental, and trade protects that existed for over 50 years. Speaking broadly, this package of changes can be considered conservative in nature. If the left/right scale is conceived as who hold more conservative beliefs would be happy with the changes that took place in the recent past. Why would a person who held conservative values be anything but

PAGE 157

157 satisfied with the policies that were enacted? Keeping this in consideration, a person holding conservative policy positi ons would have no reason to look for a change in the status quo. The results from this model bear out this expectation as an increase in conservatism increased odds of keeping the existing electoral system. This means that conservative voters generally f avored FPP over MMP. Hypothesis 2 states that if an individual is increasingly opposed to reducing taxes, the odds of voting for MMP will increase. The results from the log it model bear evidence that H ypothesis 2 is valid. The odds that a respondent woul d vote from MMP increased 20.4% per one unit increase in the opposition to raising taxes. This result is significant at .05. Again, I believe that this finding is due to the change in government policies from 1984 to 1993. During this time tax rates wer e dramatically lowered and flattened. Those that did not support the reduction of tax rates, in my view, would be more inclined to look for change. This supposition is supported by the statistical evidence, where those who were opposed to lower tax rates were more likely to vote for a change in electoral system to MMP. Hypothesis 3 states that as feelings that the government should be responsible for certain social services (jobs, pensions, welfare, unemployment, housing, and education) the odds of voting for MMP will escalate. Here the evidence is more mixed, but with som e indication of the accuracy of Hypothesis 3 Evidence that H ypothesis 3 is valid include the positive relationship between the following variables that measure the extent to which resp ondents believe the government should: provide jobs, provide for the elderly, assist unemployed, provide housing, provide universal health care, provide elementary to college education and a vote for MMP. The odds that a respondent will

PAGE 158

158 voter for MMP incr eased 13.7% for a one unit increase in belief the government should provide jobs. For those who increasingly believe that the government should provide for the elderly, the odds of voting for MMP increase 10.9%. The odds of a vote for changing the electo ral system increase 31.9% as feelings that the government should assist the unemployed escalate. The same relationship exists between the belief that government should provide housing and a vote for MMP with 40.3% increase in odds. Also, as feelings that the government should provide universal health care are increase, the odds for voting for MMP also increase 3.4%. Finally those that hold the government should provide elementary to college education improve the odds of voting for MMP 26.2% per one unit increase. While all of these variables provide the directional relationships that I expect, one of the concerns that I have is that only two of the variables, government should provide housing and education, reach the standard levels of statistical signi ficance. Both are significant at .1. There are no obvious theoretical reasons for why these two particular variables are statistically significant while the others are not. One potential reason for this relates to the why I include this variables in the model in the first place. I argue that the social welfare state was dismantled in the years between 1984 and 1993. I use these variables, taken together, to measure support for this social welfare state. Because of this, multicollinearity could be an i ssue. Overall, I am encouraged by the production of these variables and the evidence in support of H ypothesis 3 Prior to 1984, New Zealand had some of the most generous social welfare benefits in the world and one of the most activist governments. Leadi ng up to the referendum on electoral system change in 1993, these benefits were all but

PAGE 159

159 eliminated. I hold that the elimination of these services precipitated the electoral system change because of the unpopularity of the policy direction of Labour and Na tional governments. Closely related to policy change are the processes of accountability that were also discarded. Hypothesis 4 deals with a package of these process changes. listen to their caucus, should listen to relevant interest groups, and should listen to their mandates, the odds of voting for MMP increase. Here, like H ypothesis three, the directional relationships meet expectations. A one unit increase in belief that should listen to their caucus, increases the odds of voting for MMP 9.2%. Those who hould follow should follow their mandates is statistically significant at .1. Unlike the policy variables, there is a theoretical reason for why this might be the case. I argue in Chapter 3, that policy and process are inextricably linked in the majoritarian model. This models states that majoritarian rules reduce the number of political parties, which allows one party to form a government based on a series of promises. This government is evaluated based on their ability to keep those promises. The main process here is accountability. There are numerous ways that a government, in any system, can increase accountability. For instance, in New Zealand the historical data suggest three ways: the manifesto, interest group consultation, and the caucus system. Only one of these processes relates directly to the majoritarian model, the use of the manifesto. Here there is a direct link to the elections, and by extension the

PAGE 160

160 ma joritarian model. Parties make promises and then are held accountable based on result speaks a great deal for the efficacy of the majoritarian model and by extensio n how it can fail. The other two processes relate to inter election behavior and have no real connection to the election. They might help increase accountability, but they have little to do with majoritarian elections. So, the fact that the separate var iables related to group consult ation and the caucus system does not yield significant results is not devastating to the model. The H ypothesis predic ts that increasing trust reduces the odds of voting for MMP. There is strong evidence to confirm this suppo sition. For a one unit increase in trust in the government, the odds of voting for MMP decrease 36.3%. This finding is statistically significant at .01. Trust is a measure of specific support meaning it signifies support for the current practices of gov ernment. Those who believe that elected officials and the government are functioning as they should when they voted in the referendum of 1993 really had no reason to vote for MMP. The results of this logit model demonstrate the validity of the previous system. Control Variables Of the remaining control variables, age, education, gender, and household income, there are no surprising results. One interesting finding is the role education plays in FPP versus MMP choice. Here there is a strong, significant (.01) positive relationship. For every one unit of increased education the odds of voting for MMP increase 12.9%. I believe a potential reason for this is the complicated nature of electoral system politics. The political consequences of electoral syst ems are difficult for electoral scholars to identify, for those without a background in political research the

PAGE 161

161 task might seem even more daunting. Those with increased education may pay closer attention to the nuances of the electoral system debate in New Zealand, and the statistical relationship might indicate a genuine belief that MMP would improve the political circumstances. None of the other variables demonstrate significant results and the directional relationships raise no particular concerns. Her e increased income reduces the odds of voting for MMP, while men and older individuals favor MMP. Conclusions The majoritarian model offers a theoretical link between electoral systems and support. The beginning of this chain is a majoritarian electoral s ystem, which provides mechanical and psychological pressures against voting for third parties. The net result is a party system where two parties compete for support. Voters judge these parties based on the policies they offer before the election and hol d them accountable when they fail to follow through on their mandate. The electoral system is designed to create government who can then act on their policy prescriptions, so as to be judged in future elections based on their record in office. Citizens g ive support to the system as a whole based on the ability of the governments to remain accountable, while passing productive policies. New Zealand, until 1984, presents a case where this model worked. The majoritarian electoral system reduced the number o f effective parties down to two: Labour and National. These parties won nearly all of the seats contested in elections and one party or the other made up 100% of the government following the election. Parties remained accountable to the voters by publish ing a pre election manifesto and sticking to it if they were elected to office. Both parties typically supported generous social welfare policies. The fact that the model performed its function led to widespread

PAGE 162

162 support for the system. This support is e videnced by the high levels of public participation in the system. Political parties had active memberships and individuals turned out to vote at a high rate. There is also episodic survey evidence that suggests satisfaction, trust, and confi dence were o ver 60% in the 1960s and 1970 s. Perhaps the greatest evidence of support was how precipitously it fell from 1984 to 1993. This is the time period where the links between majoritiaran elections and support broke down. There were no changes in the performa nce of the electoral system elections still produce disproportional results in favor of Labour and National. What changed were the behaviors of the governments that followed. When Labour won election in 1984, their pre election manifesto had no mention o f the dramatic change in economic and social policy that was to follow. In 1987, they did not even bother to publish a manifesto. The National government promised a change in policies in its pre election manifesto only to continue the program slashing th at typified the previous largest parties and in politicians generally also declined, whic h led to a change in electoral system. The statistical evidence in the model I run supports this contention. Respondents who were more conservative tended to support the existing system, while those who wanted more liberal social welfare policies preferre d MMP. There is also a link between election and a vote for changing to MMP. The connection between accountability (pre election promises) and policy (belief in social welfare) state are keys to understanding why New Zealand changed to MMP.

PAGE 163

163 There are a few other points that I would like to make before moving forward. The first has to do with the performance of the electoral system in New Zealand. It should be noted that the electoral rules performed their immedi ate effects as they should. Disproportionality is built into majoritarian electoral systems, and in New Zealand that is exactly what they produced. The evid ence presented in Chapter 3 s uggests that varying levels of support existed over time. What diffe red was not the performance of the electoral system, but what happened after the elections were held that the respective systems succeeded or failed. Electoral system performance can neither be blamed for the Fourth their electoral mandate, to make the system work for them and their citizens at times, and alienate those same citizens in other periods. The larger question movi ng forward is, do certain electoral systems sufficiently influence party system behavior to suggest generalizable tendencies? This is a question that is better answered after a discussion of further cases. A second point that I would like to make has to do with the nature and purpose of elections themselves. The problems in New Zealand were exacerbated by the inability of the voters to address the performance of the parties through elections. This is a facet of accountability that is covered only gener ally in the statistical models that I run, but has larger theoretical importance. When Labour and National both broke pre election promises, this led to the connection between the belief that pre election promises were important and a vote for MMP in the statistical model. Democracy does not work if voters are not allowed to remove poorly performing parties with alternatives offering a

PAGE 164

164 different path. In New Zealand, voters were faced with the choice between two parties who were essentially offering the same policy prescriptions. These parties were made it clear to parties that they were unhappy with the procedural choices of the parties, and were not even given the chance to give their consent. Parties essentially determined a policy path, hid it from the voters, and then continued down the path as if they had consent. Elections did not have the chance to succeed because the parties refused to provide meaningful al ternatives. Because of this the gap between citizens and their elected officials grew progressively wider. Elections were essential blamed for change the behavior of the parties. Chapter 4 tests the efficacy of the two proportional models I draw in Chapter 3. Italy, like New Zealand, was able to make proportional government work for a time, only to hav e the system collapse. Chapter 4 details what made the system wo rk and what precipitated the changes with the proportional models as a guide.

PAGE 165

165 Table 3 1. Two party dominance, d isproportionality in New Zealand, 1943 1993 National Labour Year Seats 1 % Seats % Votes % Dis 2 Seats % Seats % Votes % Dis 1943 34 42. 5 42.8 .3 45 56.25 47.6 8.65 1946 38 47.5 48.4 .9 42 52.5 51.3 1.2 1949 46 57.5 51.9 5.6 34 42.5 47.2 4.7 1951 50 62.5 54 8.5 30 37.5 45.8 8.3 1954 45 56.25 44.3 11.95 35 43.75 44.1 .35 1957 39 48.75 44.2 4.55 41 51.25 48.3 2.95 1960 46 57.5 47. 6 9.9 34 42.5 43.4 .9 1963 45 56.25 47.1 9.15 35 43.75 43.7 .05 1966 44 55 43.6 11.4 35 43.75 41.4 2.35 1969 45 53.57 45.2 8.37 39 46.42 44.2 2.22 1972 32 36.78 41.5 4.72 55 63.21 48.4 14.81 1975 55 63.21 47.6 15.61 32 36.78 39.6 2.82 1978 51 55.4 3 39.8 15.63 40 43.47 40.4 3.07 1981 47 51.08 38.8 12.28 43 46.73 39 7.73 1984 37 38.94 35.9 3.04 56 58.94 43 15.94 1987 40 41.23 44 2.77 57 58.76 48 10.76 1990 67 69.07 47.8 21.27 29 29.89 35.1 5.21 1993 50 50.5 35.1 15.4 45 45.45 34.6 10.85 *Resu lts obtained from New Zealand Ministry of Justice 3 1 The total number of seats in Parliament changed over time. From 1943 1966, there were 80 seats contested. There were 84 in 1969, 87 in 1972 and 1975, 92 in 1978 and 1981, 95 in 1984, 97 in 1987 and 1990, and 99 in 1993. 2 The percentage of disproportionality is calculated by subtracting the percentage of aggregated votes cast for the party from the percentage of seats obtained in Parliame nt. 3 The New Zealand Ministry of Justice posts its historical election results online. These are found here: www.electionresults.

PAGE 166

166 Table 3 2: Third party support and d isproportionality, New Zealand 1954 1993 Year Disproportionality % Third Party %* 4 1954 11.95 11.1 1966 11.4 14.5 1972 14.81 8.7 1975 15.61 12.8 1978 15.63 18.5 1981 12.28 20.7 1984 15.94 19.6 1987 10.76 5.7 1990 21.27 12 1993 15.4 26.6 *Results obtained from New Zealand Ministry of Justice 4 The third party percentage is calculated by adding together the two largest third parties, who at least attained 2 % aggregated national support. The Social Credit Party was responsible for all of the perce ntages in the elections of 1954, 1966, 1981, and 1987 (renamed the Democrats in 1987). The remaining years 1972 (Social Credit/Values), 1978 (Social Credit/Values) 1984 (Social Credit/NZ Party), 1990 (Greens/New Labour), and 1993 (Alliance/NZ First) are calculated using the two largest third parties.

PAGE 167

167 Figure 3 1. New Zealand t urnout, 1954 1993

PAGE 168

168 Table 3 3. Effects of policy and process variables on MMP versus FPP c hoice, New Zealand 1993 Independent Variables N = 725 LR Chi2 = 2005.67 Prob > Chi2 = 0.00 Pseudo R2 = 0.2084 B (SE) % Odds Ratio Left / Right Scale .454*** (.081) 36.5 Government Should Reduce Taxes .186** (.087) 20.4 Government Should Provide Jo bs .128 (.128) 13.7 Government Should Provide for Elderly .103 (.173) 10.9 Government Should Assist Unemployed .276 (.194) 31.9 Government Should Provide Housing .338* (.211) 40.3 Government Should Provide Universal Health Care .033 (.147) 3.4 Govern ment Should Provide Elementary to College Education .232* (.139) 26.2 .088 (.107) 9.2 Groups .141 (.097) 15.2 .139* (.080) 15.0 You can Trus t Government to do what is Right .451*** (.094) 36.3 Education .120*** (.042) 12.9 Age .011 (.006) 1.1 Gender .151 (.183) 16.4 Household Income .032 (.045) 3.2 P < .1 P < .05 ** P < .01 ***

PAGE 169

169 C HAPTER 4 THE RISE AND FALL OF PROPORTIONAL GOVERNMEN T IN ITALY Introduction Chapter 3 uses New Zealand, and its majoritarian electoral system, to test the merits of the majoritaria n model of support. Chapter 4 describes how, through historical data, majoritarian electoral systems and support co existed for more than 100 years. It also demonstrates what changed, how that led to a decline in the levels of support for electoral institutions, which led to electoral system change. To summarize, single member districts in New Zealand produced, time and again, g overnments formed by one of the two largest parties: Labour and National. These parties made use of a pre electoral manifesto, as well as other post election processes, to cultivate the process of accountability. Labour and National, by maintaining a str ong sense of accountability, reacted to the needs and desires of their constituents and pursued generous social welfare policies and protectionist economic policies. The links in the majoritarian model were met: a majoritarian electoral system led to a tw o party system and successive one party governments, these governments followed the process of accountability through a pre election manifesto, and enacted policies that the public wanted. New Zealanders felt like they had a very high quality of life and this led to high levels of support, demonstrated by the high level of voter turnout, party membership, and trust in government. But, beginning in 1984, the majoritarian model that had worked so well for the previous 100 years began to break down. The elec toral system continued to favor Labour and National and there were one party governments in 1984, 1987, and 1990. What changed drastically were the commitments to accountability and social welfare

PAGE 170

170 and protectionist policies. Labour and National governmen ts dramatically cut the social welfare state, eliminated many of the protections that labor groups and farming interests depended, and completely revamped the tax code in nine short years. They did this without signifying to the public beforehand what the ir plans were. Labour did not even bother to publish a pre election manifesto in 1987, ending a practice that endured for more than 50 years. National promised in 1990 to slow or reverse the policies of Labour, then extended and strengthened them. Accou ntability was shattered, and the policies and process that made majoritarian government work were gone. Support as a result fell, demonstrated by the decline in party membership, turnout, and trust. The logit models that I run demonstrate that those who favored conservative policies were more likely to favor keeping first their mandates (a representation of accountability) believed in a switch to mixed member proportional. I make a few conclu sions f rom the data. The first is that majoritarian elections did their job. For 100 years, the electoral system produced disproportional results favoring one of the two largest parties, Labour and National, at the expense of smaller parties. One of those part ies, Labour or National, retained enough seats following the election to create a single party government with a majority of seats. The variation occurs afterward, where theoretically New Zealand moves down the majoritarian chain to support. Parties in N ew Zealand pursued both policies and processes that cultivated support and different policies and processes that led to a decline in support. This leads me to believe that it is the party system behavior, not the electoral system that is responsible for v ariation in support. It also directs me toward a conclusion that the

PAGE 171

171 models I draw are merely courses of action for parties that would make majoritarian elections work, rather than a suggestion of theoretical likelihood. This is a point I return to in th e conclusion, but what remains is to test the connection between proportional elections and support. To do so, I examine elections and the party system in Italy from 1948 to 1992. Italy presents a very interesting countervailing case. If New Zealand was a model of majoritarian system, Italy was the model of a proportional system. The Italian First Republic (1948 1992) is often used as an example of how proportional systems do not work. There is certainly evidence to support those contentions. The Itali an party system was polarized, with communists, fascists, and centrist parties competing for electoral support. Governments rose and fell routinely and during this time period as Italy averaged a new government every 11 months. The general consensus was that the coalition governments that were produced were not able to deal with the problems of the day and relied on extensive networks of clientelism and corruption to ma intain their status. I try to look at Italy mething must levels of participation in their democratic process evidenced by high levels of voter turnout and party membership. The Italian economy also grew at very high rates in the 1950 s and s as Italy eventually became the seventh largest economy in the world. T hese successes existed alongside high levels of public dis content There were also high levels of corruption discovered in t he early 1990 s. In shor t there are a lot of moving parts in the story of Italian democracy. It is certainly a more difficult country to explain than New Zeala nd, but the project of Chapter 4 is to try and describe the

PAGE 172

172 connection between proportional elections, party systems, p rocess and policy, and support. I argue that Italy, like New Zealand, was able to make its proportional system work, then when changes and policy and process occurred the efficacy of the system was lost. The proportional model I tes t states that as prop ortionality increases so do the number of ideologically diverse political parties represented in the legislature. Governments are then formulated from these ideologically diverse parties, which increases the process of representation. These governments t hen engage in consensual decision making, where it is thought that better policies will result. In order to test this model Chapter 4 unfolds in the following manner. st udy. If New Zealand is an example of concentrated and centralized power, Italy is the opposite. Here the constitutional structures are designed for disaggregated and diffuse power, more akin to the proportional model. The sections following go through t he proportional model step by step. The fi r st step shows how the proportional rules did lead to a multi party system with ideologically diverse political parties. The subsequent section details how the parties were able to represent and articulate the in terests of their constituents, albeit imperfectly, which satisfied the process requirement of the proportional model. More specifically parties like the Christian Democrats (DC) and Communists (PCI), used ancillary organizations to represent their respec tive sub cultures. The DC also developed organizations in the southern part of the country they maintained with clientelism. The DC also installed their members into all aspects of the sottogoverno (sub government), large, technically non governmental or ganizations that

PAGE 173

173 had tremendous influence in the country. These unique representative structures were necessary because of another important feature of the Italian political system: the permanent exclusion of the left and right parties from government. T his impacted the way governments were formed and their stability. The DC developed factions that constantly fought over control of the sub governmental positions causing governments to collapse and reform routinely. The way governments formed impacted policy, which to a large extent ended up supporting those clientelist ends. Italian governments, and parties outside of government, were also able to pass a significant number of clientelist policies that kept the country moving. Some of these policies and sub governmental operations were tremendously successful as Italy experienced high levels of economic growth during the 1950 s and s. Also at times the legislature was able to deal with significant national issues. For example, during the Historic Compromise from 1976 to 1979 the legislature was able to deal with significant issues like the mounting public debt and the pressure from Europe. While this could not be considered the norm, prior to 1979, significant policies were achieved. Proporti onal government in Italy was far from perfect. Despite this, Italians displayed high levels of participation during this time period both in the number of people who joined political parties and voted in elections. This high participation, paradoxically, was coupled with public attitudes that were highly critical of the functioning of democra tic organizations. Chapter 4 also discusses why this was the case.

PAGE 174

174 Like New Zealand, there were changes that precipitated the fall of the proportional electoral sy stem in Italy. The electoral system continued to produce a multi party system with ideological diversity. The main change was the loss of r epresentation. During the 1970 s the cultural ties that bound people to political parties, like Catholicism, were w eakened. This led to more variable electoral results and loosened the g rip the DC had on the strong clientelist structures they had built. As the DC lost influence and vi ewed politics as a means for achieving personal gain. The loss of electoral support for the DC meant that more parties were necessary to form governments. This increased o ffice. The clientelist structures were still in place, but the level of c orrupt behavior of elected officials and appointed members of the sub governments expanded exponentially. Because of this, the features that had made the party system Generally corruption is viewed as one of the major reasons for the loss of support in Italy, but there are other factors that need to be considered as well. International events, and the Italian response to them, mattered a great deal. When the Cold War ended, the PCI abandoned its call for a dramatic reorganization of the Italian state and economy, eventually splitting into two parties: the Democratic Party of the Left (PDS) and the Communist Refoundation (RC). One of the main reasons for voting for the DC, anti communism, was now gone. Another party the Northern League, fed up with the funneling of money to the South, challenged DC strong holds in the north. All of this was happeni ng while the existing parties were under increasing pressure from Europe to rein

PAGE 175

175 in the large public debt and budge t deficit. Unlike the mid 1970 s, there was no government of national solidarity to be found. Policy could no longer be used to relieve pres sure, and this also led to a decline in support. The combination of the change in processes, namely representation structures, and the lack of any kind of policy led to a loss of support and a change in electoral system. The following section shows how It 1948 to 1979 both through the process of representation and policies that reflected that representation. The ensuing discussion concerns the changes in processes and the lack of policy that led to a loss of sup port and a change in electoral system. Ultimately, I argue similarly to New Zealand, that the electoral system in Italy functioned as it was designed to function. The party system behavior is what determined the varying levels of support. The next secti on describes the institutional arrangements in Italy and how the electoral system led to a multi party, ideologically diverse party system. Italy: A Model of Proportional Government Italy is not a perfect representation of the proportional or consensual mo del, but it has many features which make it a good fit for the category (Lijphart 1984; Powell 2000). Lijphart identified eight elements that a consensual system of democracy should have: Executive power sharing: grand coalitions Informal and formal separ ation of powers Balanced bicameralism and minority representation Multiparty system Multidimensional party system Proportional representation Federalism and decentralization Written constitution and minority veto

PAGE 176

176 ructure closely fit these characteristics. Lijphart does not start with the type of electoral system, but since it is the root cause of many of the unique features of the Italian political system, I start there. During the First Republic (1946 1993), Ita ly had one of the most purely proportional systems in the world. Seats were allocated on a proportional basis with two almost inconsequential thresholds. First, in order to be eligible for seats a party had to aggregate at least 300,000 votes nationally to be considered for seats Additionally, Italy was divided into 32 circonscrizioni (districts) that were awarded seats ba sed on population. Members of p arliament needed to be elected within those regions through a preferential list system (roughly 60,00 0 votes would win a seat). 90% of seats were allocated this way. The remaining votes where a winning candidate did not receive the minimum number of votes would be thrown into a national pool, which was then divided proportionally among parties and candi dates not directly elected. (Koff and Koff 2000). Because the system of electoral rules was so permissive, smaller parties benefited and the party system was fragmented to a large degree. From 1946 to 1993, 10 national parties regularly competed for seats as well as 2 or 3 regional parties. These parties not only competed but were able to win seats on a regular basis, and it was not a rare occurrence to have upwards of 10 to 15 parties with some representation in Parliament 1 (Spotts and Wieser 1986 ). Pa rties were ideologically diverse, covering a full range of ideological beliefs from left to right 2 1 For complete list of parties winning seats and their share of the vote, see appendix. 2 This is a point I return t o in a subsequent section that discusses the relationship between proportional elections, the number of political parties, and the ideological diversity of those parties.

PAGE 177

177 governments from 1946 have been multi party coalitions, although the Christian Democrats have formed minority governments from time to time. The unique featu re of coalition governments was the permanent exclusions of the left and right, forcing the 3 The govern ments that were formed were weak, due to the reliance on smaller parties. These parties generally had somewhere between 3 or 4% and gave them a veto that far The strength of smaller parties in this regard and the factions within the Christian Democratic Party led to frequent government collapse, with a new government needing to be formed, on average, every 11 months. If this were not a difficult enough situation, governments also had to navigate a bicameral legislature. The legislature is made up of two houses: the chamber of deputies and the s enate, who have identical law making powers. There are a number of institutional rules that aid decentralization, howeve r. The first is the use of the secret ballot for legislative votes. This made it extremely difficult for governments to pass a coherent legislative program, since they had no mechanism to determine how their members were voting. Another form of decentra lization was the powers of the legislative committees. Committees can pass laws through private consent without the approval of both the Chamber and Senate. These features encourage decision making that include the broadest bases of support possible with in the legislature, a key feature 3 This is a point I return to in a later section on representation and coalitions.

PAGE 178

178 All of these components make Italy a proportional system, and a clear understanding of the features is vital before using Italy to test the following model: (+) Proportionality decision government 4 Italy, at first (and maybe second) glance would not appear to make the most favorable case to study this relationship. As I mention in Chapter 2, Italy is often described as either an aberrant case or a country that exhibits all of the negative attributes of proportional government. A cur sory glance at the polarized party system, the instability of governments, and the high levels of corruption might lead someone to that conclusion. Despite these conditions, Italy consolidated its democracy, practiced it for over 50 years, and created the seventh largest economy in the world. It may be a simple assumption, but I believe something must have worked for this to have happened. Considering this presumption, something must have changed for the existing party system to fall apart and for the p roportional electoral rules to change. I believe the above model can shed light on the answers. Specifically, the processes of representation changed and the legislature was no longer able to pass policies that dealt with the changing circumstances in th e country. While the country found processes and policies that work ed and when those policies and processes changed, 4 As I mention in Chapter 2, there are actually two different models represented here. There is an open question whether process alone can engender support or if an additional step including policy is necessary. For the proportional process model, adequat decision making process in proportional is enough to increase support. For the proportional policy model, t he purpose of brevity I have included the one model as a reminder to readers. It is my belief that policy and process are inextricably linked, a point I return to later.

PAGE 179

179 the electoral system continued to function as intended. The subsequent section deals with the first part of the model and whether propor tional elections in Italy led to an ideologically diverse, multi party system. The rest of C hapter 4 follows the model from the re to determine whether ideologically diverse parties were able to incorporate more representative agents into the decision maki ng processes. If more representative agents were indeed incorporated, the next step is to determine if those agents were able to pass widely accepted policies. Proportional Elections and Political Parties: Italy 1948 1992 The proportional electoral sys tem of the First Republic offered virtually no barriers to representation and smaller third parties were able to take advantage and consistently win seats. Results were consistently proportional and a brief examination of electoral results demonstrates th is fact. The highest levels of disproportionality ranged from 2 to 4% and usually favored the Christian Democrats 5 It was a much more regular results of less than 1%. Table 4 1 displays these results The proportionality of the electoral system led to a multiparty system where up to 10 parties consistently competed for seats and support i n national politics. Table 4 2 is a list of the number of political parties t hat won seats in the Cha mber of Deputies and the Senate. The party system itself, besides being quite large, was also ideologically diverse. The largest party the DC occupied the center position in the ideological spectrum. The party was founded in the e arly 1940 s by anti fascist Catholics and the pre fascist Italian 5 Disproportionality was measured by subtracting the %of seats from the %of votes.

PAGE 180

180 across classes and ideologies, a catch all strategy, designed to provide a centrist base and a safeguard against communism. Backed strongly by Catholics and church organizations, the DC was the largest political party in every election and a part of every government during the First Republic (Newell 2010). The largest party on the left was Communist party (P largest party and the largest and most electorally successful communist party in the democratic world. The party was nominally committed to reform while behaving quite responsibly in the legislature. Despite this responsible a ction, due to the politics of the Cold War, the PCI was subject to the con v entio ad excludendum. This meant that they were permanently excluded from government by the DC and smaller lay parties (Newell 2010). The reason for the exclusion was that the par ty was perceived, as a party that wanted to not only modify the state or provide it with a different s ort of personnel, but to replace the state with somethi ng different. The PCI was not the only party that was excluded from governing coalitions. Until 1 962, the Italian Socialist Party (PSI) was also left out of government. The PSI was a party formed around working c lass interests in the late 1890 s. They were continually placed in an awkward ideological and political position since they occupied the spa ce between the DC and the PC I. Unlike the PCI, entered into a coalition with the DC in 1962. They would enter and leave the government in ensuing years, eventually gaining the premiership in the ear ly 1 980 s. The PCI and PSI made up the main parties on the left, but there was another party on the right excluded from government coalition, the Italian Socialist Movement (MSI).

PAGE 181

181 generals. The party cautiously avoided any public calls for a return to fascism, since a reformulation of the fascist party was banned by the post nationalism and authoritarian bent placed it firmly on the right ideologicall y and were Three of the largest parties, the PCI, MSI, and PSI (temporarily), were left out of the governing coalitions. The proportionality of the electoral system gave opportunities for smaller, non secu lar lay parties to gain seats and take part in the government. One class. The PLI generally supported neo liberal economic policies and a strong, authoritarian cent ral government. The PLI provided the DC with a partner who could appeal to the urban bourgeoisie and big business. The PLI averaged about 3.5% of the vote from 1948 to 1992 and was one of three smaller parties that joined with the DC led governments. On e of the others was the Ital ian Republican Party (PRI). It was an anti monarchy, anti clerical, anti fascist, pro market, transatlantic party. This party gave more conservative voters who were not business owners or partic ularly religious a party to suppo rt The PRI averaged around 2.8% of the vote from 1948 1992. The PLI and PRI occupied the political space on the center right. The last of the DC satellite parties was the Italian Social Democratic Party (PSDI). This party split from the PSI over its communist party, the PSDI had a left leaning socialist ideol ogy. Between 1948 and 1994 the party claimed an average o f 4.4% of the vote (Newell 2010, 27 28).

PAGE 182

182 Now is a good time to stop and think about the model I present in the previous pages. This proportional model suggest s that the more proportional a system the more ideologically diverse parties would be in the party system 6 Altogether, the Italian party system contained right wing (MSI), left w ing (PCI, PSI), centrist (DC), center right (PLI, PRI), and center left (PSDI) in Parliament. There were also a number of smaller regional parties that won seats as well. The electoral system was certainly a determining factor in the formation of this sy stem. For instance, Duverger (1954) argues that there would be no incentive for smaller parties to join or be subsumed by larger parties with similar ideologies. For example, the PSDI had virtually identical ideological p ositions with the PSI, but a d isa greement over party strategy they broke away and ran their own list of candidates. With 4.4% of support, thanks to the permissive proportional rules, they were able to consistently win seats. The first conditions of the proportional model have been met. What remains to be seen is if constituents and the types of coalition governments that were formed. The next section details these two processes of representation. The Process of Representation: Parties, Clientelism, and Coalitions in Italy, 1948 1979 The previous section describes why Italy is a good case to test the proportional models I draw in Chapter 2. The model suggests that proportional rules will lead to a multiparty system containing ideologically diverse parties. Elections in Italy did indeed lead to a party system that had ideological diverse parties. From 1948 to 1979, 6 This is due to the more permissive mechanical effects and the resulting relaxation of psychological effects. For a more complete discussion, see Chapter 3 and the works of Duverger (1954), Rae (1971), and Lijphart (1984).

PAGE 183

183 between 9 and 11 partie s were represented in p arliament ranging from communists on t he far left and reformed fascists on the right. Disproportionality was low with highs of between 2% and 4%, generally in favor of the Christian Democrats (DC). Usually smaller parties had experienced disprop ortional results of less than 1% At the very least this created at least some degree of ideological representation in the Italian party system. Parties were the key representative institution in Italy, and they were the most dominant feature of Italian democracy. This section goes further than just looking at election results and how the electoral system allowed for an expression of the ideological differences in Italy. In Chapter 2 I argue that ideological expressions are not enough for true representation. Citizens need to see their ideas and in terests expressed and e m bodied in their elected officials. Parties in Italy developed representative institutions in unique ways. They cultivated deep social and communal ties with their constituents through different means. One of these means was the u se of sub governmental organizations to cultivate clientelist ties to various populations. Chapter 2 also describes how coalitions are supposed to be representations of how the population votes. Italy developed its own means for forming coalitions howeve r. Because of their stated revolutionary beliefs the second largest party in the system, the Communist Party (PCI), was barred from governing. This cause d a situation where the DC was obliged to lead the formations of governments. This section describes the processes of representation in Italy before 1979 including: the deep communal ties, clientelism, and the formation of governments. In terms of process, the proportional model suggests that more proportional electoral systems allow for more political parties, higher levels of representation of interests and ideology by

PAGE 184

184 those parties, and the representation of those ideas and interests in the formation of government coalitions. Ultimately, al l of these proportional processes are determined by the polit ical parties and these parties in Italy played that vital role. The strength of the political parties in Italy, prior to reform, can not be overstated. Their power dates back to the formation of the new Italian Republic after WWII. The p arties construct ed the Italian C onstitution between 1945 and 1948. They purposefully designed institutions that would advantage them over other institutions of government. By the time Italy had its first parliamentary elections in 1948, the political parties h ad been ru ling the country for four years. Other countries have strong parliamentary parties (France, England, Germany), but it would be absurd for Prime Ministers in those countries to check with their party organizations or party chairmen before pursuing a legisl ative agenda. In Italy, that is exactly what happened and may help to explain why many ambitious politicians preferred to be party secretary rather than Prime Minister. Parties had a favored position in Italian political life, and it is important to deta il the means that they used this position to represent their constituents. R epresentation is the most commonly cited advantage of PR electoral systems as described in the proportional model, but to just stop with the ideological representation fostered by of support). Parties not only represented their voters ideologically, they also be haved like octopi whose tentacles reached into all social, economic, and cultural areas of the nation. Partitocrazia (rule by parties) is a term that is used to describe the level of

PAGE 185

185 control and importance parties had in all aspects of Italian life 7 To explain how parties represented Italians, I use the Christian Democrats (DC) and the Communist Party (PCI) as examples. There are several reasons for only examining the DC and the PCI rather than looking at all 9 to 11 parties that won seats during this t ime. One reason is that only (LaPalombara 1987) and other than the DC or PCI, only the Socialist Party (PSI) ever reached more than 10% of the vote. Another reason for exam ining only the DC and the PCI is their central importance to the political system. From 1948 to 1976 the two parties garnered an average of 67.98% of votes cast and won an average of 72.77% of all seats in each election 8 Not only did the dominate the nu mber of votes and seats won, they also represented the two predominant sub cultures in Italian politics, [The DC and PCI] were the main political expressions of a profound strong f eeling of identity with the working class and a conviction that the latter was destined to play a central role in the process of social transformation, and, on the other side, a strong attachment to the Church and to the ecclesiastical hierarchy, an accept in personal, social, and political matters and, hence, a fie rce anti communism (Newell 2000, 15) The DC and PCI represented their constituents in differing fashions. The DC used its ties to church organizations in the North a nd created strong clientelist ties through a 7 Parties had a large degree of social control, but this did not translate into policy making strength. Italian parties, especially the Christian Democrats, were hindered by factions within their own ranks. Votes in parliament were also, until 1991, conducted by secret ballot. This m eant that there were no mechanisms to keep members in line. Parties in Italy had deep social roots, but thin policy making abilities. I discuss this irony in a later section. 8 The yearly percentages for the two parties for % of vote/ % of seats for each election are: 1948 (79.5/85), 1953 (62.7/68.8), 1958 (65.1/69.3), 1963 (63.6/67.6), 1968 (66/70.3), 1972 (65.9/70.6), 1976 (73.1/77.8) (Bull and Newell 2005, 46, 66).

PAGE 186

186 series of sub governmental offices in the south (and in the north as well). The PCI became a mass based party that used various grass roots organizations to develop a sense of community and partisan attachments the center of the country (Ginsborg 1990). The following section describes how these two main parties used various party organizations to increase the levels of representation throughout the country. The section also details how the parties were keys to the political system in other ways as well. The Communists, even while receiving more than 25% of the vote in every election, were excluded from being a part of any coalition government. This was due to their pu blicly held anti system beliefs and commitments to transformative change. The DC, as the systems largest party, was obli ged to govern the country. The party did so as the head of every post war government until 1980. This section describes how the DC an d PCI represented their constituencies, how the DC used clientelism and state The DC and PCI Representation: Associations and Clientelism, 1948 1979 The DC and the PCI we re the largest two political parties in Italy and they represented broadly the two pre dominant sub cultures within the country: Catholicism and communism. Both parties were mass based organizations, who serviced their members through several means. One main mechanism was the cultivation of both respective subcultures. To be more precise, the DC formed close ties with Catholic organizations, while the PCI had its own commun ity and worker establishments. The two parties used these affiliations to represent their membership and strengthen their ties to the community.

PAGE 187

187 The DC was organized in 1943 by members of the pre fascist Catholic Popular Party, Catholic trade unionists, a nd members of Catholic professional organizations. The first party conference in 194 6 cobbled together these different groups into a coalition supporting policies that their varying members could support. Their members included every social category from socialistic workers to more conservative landowners. This made their ideological commitments broad, centrist, while also including representation on the left and the right (Spotts an d Weis er 1986, 21). Prior to the first elections held in 1948, the DC f aced opponents on the left (PCI and the PSI) who had strong national organizations. The party had no national organization of its own, so it turned to its likely ally, the Catholic Church, for assistance. For its part the church was determined not only to combat Communism but mold Catholics into a political bloc and tie that bloc to the party. At election time t he church, pronounced it a religious duty not only to vote, but to vote for the Christian Democrats. These were the days when parishioners were d 244) T o support these feelings, the church used the organizational resources at its disposal create a DC party structure and membership. With their help, the party had close to 500,000 members, 7,000 local offices, 9 daily newspapers, and 8 weekly papers by the time of the first election of 1948 (Spotts and Weis er 1986, 22). The party structure of the DC was never as centralized as their counterparts of the left, so prior to the 1960s, they relied heavily on Catholic community organizations like the Catholic Ac tion and its 9 9 Catholic Action was not the only church organization that the DC continually relied up on during this time. Examples of other organizations include the Coldiretti, which was organized to support the interests of peasant families. This group boasted 1,600,000 members during the 1950s (Ginsborg 1990, 171). Other

PAGE 188

188 Catholic Action was led by its politically active President Luigi Gedda. This was a national group that attempted to organize every social activity of the family around the church and Catholic teachings. The male youth section of this organization set as a proselytizing amongst youths; the preparation of young men for family and social life; the furtherance of a healthy intellectual, physical and recreati 1990, 169) Children were brought into these associations at an early age, with Bible lessons on Sunday afternoons, prayer meetings, summer camps, and sporting associations. As they aged into teenagers, they graduated to summer holiday camps where scout leaders preached the dangers of industrialization and the importance of the communion of man and nature. Catholic Action also was active among adults where, se of public morality, and on the Christian educati 169) Cultural activities were also important. Catholic Action was heavily invested in a network of cinemas, over 4,000 countrywide, where church approved m ovies were show n (Ginsborg 1990, 169). well. Families were mobilized at the street level as well. For instance, in 1948 (the year of the Madonna) and in 1950 (the Holy Year) the Mad onna pelligrina was brought to neighborhoods around the country. This community ritual saw the statue of Madonna socialize and pray. The statue was then moved to another then another street, and eventually to another town or village (Ginsborg 1990, 177). organizations included the C atholic Trade Union Association (CISL) and the Association of Christian Workers (ACLI) (Hine 1993, 117). These groups were political active in favor of the DC.

PAGE 189

189 The activities of Catholic Action were not limited to developing a C atholic based civil society. They were also overtly political, when electi llot for the DC During the mid 1950 s, Catholic Action could give this task to its roughly 2, 700,000 members, a va st number of which were in the N orth. Veneto had the largest membership (240,000), followed by Lombardy (220,000 ) and Piedmont (105,000 ). Southern membership lagged considerably with its largest membership hosted in Sicily with roles numbering 64,000 (Ginsborg 1990, 169). These numb ers and the reliance on c hurch backed organizations belied several problems for the DC. The first of these problems was that the party really had no centralized organization of its own to depend upon. By relying on the church for support, they also ceded that the church would have an active role in government. The second problem was that the associational character of the N orthern parts of the country, which the DC mined for electoral support, did not exist in the southern parts of the country. The DC h ad to come up with different representative strategies for these areas of the country. The strategy that emerged was the use of extensive sub governmental agencies to develop clientelist relationships in both the south and the north. By doing so, they co uld simultaneously separate themselves from church control and reach a segment of voters in the south that the party needed to maintain their standing as the largest party. familis nization of Chiaramonte in the S outhern province of Potenza. Chiaramonte, a town of 3,400 farm labors, had no hospital or school system. Banfield noted that residents of the town

PAGE 190

190 were both bitter about the lack of education and health services and unwilling to collectively organize to do anything about it. Banfield asked why this deprivation did not lead to any sort of political action that would lead to the town addressing its co llective material, short run advantage of the nuclear family; [and] assume that all others will d o outhern town came to be the primary mechanism for understanding the political motivations of Southern Italians. meant that Southern Italians were not willing to sacrifice short term benefits in favor of long term advantage. Because of this, the DC needed to develop and use mechanisms that Cl ientelism is a term that describes the condition where the political support is exchanged for the unequal, or non universalistic, distribution of public resources is low, people have difficulty in protecting and advancing their interests on an organized basis and are therefore obliged to seek the necessary protection on an (Newell 2010, 157) In the S outhern parts of Italy these levels of social capi tal existed and the inability to organize to pursue collective interests was pervasive, so clientelism clientelistic fashion in order to create a mass base tied material ly to the Christian Democratic Party. In the South, clientelism was more important than the bishops, local government more than Catholic a

PAGE 191

191 strategy for developing and maintaining clientelist ties in these are as was to install party s sottogoverno (sub government ). These public corporations are a distinct feature of the Italian public administration. Made up of some 45,000 entities these corporations are of the government but not technically in it. These bodies have private shareholders, but receive state funds and the major b anks and credit institutions, social service agencies (pensions and welfare), sport, culture, and recreational organizations, theater, the opera, museums, TV and radio stations, universities, and ho spitals (Spotts and Weis er 1986, 137). The most important of these public entities, are the large state industrial corporations. These state holding companies control over 1,000 sub entities. The largest is the Institute for Industrial Reconstruction (IRI), which by the 1 980 n with over 600 holding companies containing many of the aforemen tioned institutions including: three s largest banks, most highways several airports, most of the ly one state holding company. There is also the National Corporation for Hydrocarbons (ENI), chemical firms. Outside of the state industrial corporations is the banking system which is another important component of the sottogoverno. All the major commercial, savings, and credit institutions are part of the public corporations. Included also, outside of banks and state holding companies, are the pension and welfare agencies which manage all the various social services including welfare, social security, unemployment

PAGE 192

192 insurance, and health benefits. All of these public corporations were designed to be autonomous organizations. Personnel were supposed to be recruited without p olitical considerations and the operation of the corporations was to be free of governmental interference. But, since the government maintained the ability to appoint top managers of these corporations the ability to influence their operation was always p ossible (Spotts and Wei s er 1986, 141 143). The DC used the availability of this option to set up vast clientelist n etworks. Beginning in the 1950 s, the party slowly installed loyalists to positions in the non commercial firms. They then moved into the in dustrial firms and banks and by the late 1960 s IRI and ENI (and their subsidiaries), most of the banks, the social security and control (Spotts and Wei s er 1986, 145). Contro l of these agencies was used for a variety of purposes including: spending, access to credit, and discretionary powers (Ginsborg 1990 178). Money could be spent on various public works projects. At a more personal level, resources could be committed to hiring party member or creating posts for party members. In the southern parts of the country (and to a certain extent, the entire country) the awarding of jobs was, Jobbery pure and simple. Every country has its spoils system where at least some top gov ernment posts in the US 3,000 federal positions go to party supporters. But in Italy the political selection process was extended from the top to the bottom of the government and then spread into the public agencies, where it influenced virtually every appointment, including secretaries, messengers, and clerks. In this way the sottogoverno, with its thousands of public entities and hundreds of thousands of jobs, was transformed by the Christian Democrats into the biggest pool of patronage in any de mocra tic state. (Spotts and Weis er 1986 144) Another means of handing out patronage through public spending was the awarding of pensions. These became a very important source of income, especially in the South. It

PAGE 193

193 was not uncommon for a party member to have two or three pensions, doled out due through the aid of a party patron (Ginsborg 1990). Loans were another powerful mechanism for guaranteeing party loyalty. Any person who needed a loan, whether a large industrial firm, a small shopkeeper, or an indivi dual applying for a mortgage, had their odds of success increased if they were a loyal party member. This was particularly important in the agrarian South. Agricultural credits were vital to small farmers, but unless the farmer was a member of the DC or its ancillary group the Confederation of Small Fa rmers acquiring said loan became a difficult proposition. If the person is a member, he may find his path to a credit eased, or given a favorable interest rate in exchange for continued political support fo r the The public administration in Italy was (and is) noto riously unwieldy. Italians found themselves in a situation where the most routine of behaviors required some form of government sanction. At the same time, attaining this permission required an inordinate amount of procedural hurdles, like a multitude of government documents and several approvals from multiple civil service offices. The DC found a way to take a dvantage of this by creating recomendazione (recommendation) forms that would speed up the movement of papers through the bureaucracy. For instance an Italian could go to his local party patron who would sign a recommendation form for him that would speed up the awarding of licenses (for building permits, permission for restaurants to keep outdoor tables, etc) and allow that client to avoid a very painful experience with the bureaucracy (Ginsborg 1990).

PAGE 194

194 The party expected returns for the favors it gave out and received several from clients in the South. At the most basic level, this was support on Election Day. Even this was not just a simple matter of casting a vote for the DC. Italy, until 1991, used preference voting in its elections. This meant that an individual could cast a vote not only for the DC, but rank the candidates in order of preference 10 This gave the elections a personal element, which helped divide the Christian Democratic Party in to factions that fought over access to state resources 11 Southerners who received favors were expected to be members of the DC or one of its affiliate organizations as well. At higher levels, the businessman or property owner looking for more elaborate political patron. This behavior would become much more important during the 1980 s as a much more systematic and pervasive practice. Its root s were developed by the Chri stian Democrats during the 1960 s (Ginsborg 1990) It may seem contradictory to suggest that the clientelism that I have just described could be remotely considered to be a form of representation. Both Piattoni (1998) and Zuckerman (1997) disagree. Piattoni describes the form of clientelism she observes in tire community. Here, the patrons extract resources from the center and distribute them to 10 For more on how preference voting aided clientelism in the South see Golden (2003). This aut hor makes the case that the preference vote was a key component of clientelism in Southern Italy, where individuals cast preferences at a much higher rate than in the North. 11 This is a point that I return to later when discussing the formation of coalitio ns and the stability of governments.

PAGE 195

195 the periphery. Clientelism can be a successful representative strategy that aids both patron and client when it is used to build hospitals, schools, roads, museums shops, or to provide jobs, aid to the elderly, or medical care to the sick. Not all clientelism in Southern Italy was a representation of these ideals, but there is a case to be made that it was successful. This is a point that I return to when I discu ss policy making in a following section. Representation was not only confined to the Christian Democratic party, but the second largest party, the PCI, also developed mechanisms to represent their interests even without the aid of state resources. Unlike the DC that heavily relied on the existing organizations of the church and clientelist ties cultivated by a control of the sottogoverno, the PCI had autonomous party organizations tied to communities where they had political strength. While the DC relied extensively on church organizations and a weak, factionalized party organization boosted by the availability of state res ources, the PCI during the 1950 s boasted the largest communist party in the democratic world with 2,100,000 members (Ginsbor g 1990: 194) and held a membership of at least 1,500, 000 members until the late 1980 s. This is r emarkable, considering that the party was left out of every governing coalition and had virtually no access to the clientelist state system. The party support ed their membership instead, with a network of organizations that were designed to bring civil society. One of the largest of these organizations was the Case del Popolo ( were located mainly in the center of the country and in the surrounding smaller communities. These were central meeting points for communities and became a place where a number of activities took place

PAGE 196

196 including: deb arranged, and some even had pu blic bath houses (Ginsborg 1990, 195). When many of ct the communist strength), PCI sup and defiance (Ginsborg 1990). The PCI also had significant strength in other mass organizations controlled by the ociation of organizations p erformed a variety of functions. F or instance Ginsborg describes the d ballet gave a performance and the trousseaus of the members who were to be married that year were put on display. Other activities included petitions for public housing and for peace, the selling of the journal Noi Donne assistance to the older and sic k women during the winter months, solidarity with women workers sacked at the local shoe s ea. (Ginsborg 1990, 196) Other than the organizations, the PCI maintained an extensive communication and the party published several journals (Rinascita, Politica ed Economia, or Studi Storici), which receive d critical praise. The activities of the DC and PCI described here are examples of how the Italian party system went much further than pure ideological representation. Both parties (and to a lesser extent, all parties) developed close communal ties to their supporters. But, descriptions of these activities are not sufficient to understand the extent to which the parties permeated all aspects of society. I have already mentioned that the DC firmly implanted itself in the sottogoverno and that the PCI developed extensive associational

PAGE 197

197 ties with the ir supporters, but the importance of parties needs to be stressed further. Identification with a party meant participation in particular political or social groups, recreational clubs and also helped to define where you got your information. Membership i n either the DC or PCI could conceivably mean that an individual woke up read the party newspaper or watched the news on the party controlled TV station, went to work at a job they received from a party recommendation, had coffee at the party controlled ba r, or might take a vacation on a party sponsored and organized trip. Voting and joining a particular party was more than a signaling of support for a particular policy onging (Parisi and Pasquino 1980 ) 12 At this point the first elements of the proportional model have been met. The electoral system in Italy provides for a proportional distribution of votes to seats. This is evidenced by the low levels of disproportionality reported in the election res ults from 1948 to 1992. The next step in the model is the relationship between proportionality and the number of ideological diverse political parties in the system. This condition has also been met as the Italian party system included parties on the rig ht and center right (MSI, PRI, and PLI), the center and center left (DC, PSDI), and the left (PSI, PCI). DC had a dual strategy of using church sponsored organizations in the North and use of clientelism and the state apparatus, developed close relationships with workers groups and civic organizations in order to cultivate strong repre sentative relationships. 12 Voting and party membership is a point I return to later when I discuss declining levels of representation in Italy.

PAGE 198

198 The next step in the model is to determine whether representation extended into the actual decision making processes of government. In most cases, the logical place to look is the make up of governing coalitions. A cursory glanc e at how governments formed and collapsed in Italy might lead to the premature conclusion that this is where the model falls apart. But, there are several reasons that I believe that the proportional model is still successful. The main reason is that dec ision making did include a number government. The Formation of Governments and Transformismo, 1948 1979 The most oft repeated criticism of Italian democracy is the complete lack of government stability. Governments rose and fell, on average, every 11 months from 1948 1992. This would seem to be a problematic occurrence for the models that I draw and a boon to the critics of proportional electoral systems. As a reminder, th e proportional model suggests that increases proportionality in the electoral system leads to an increase in ideologically diverse political parties as well as an increase in the making process. Critics of proportional elections point out that such ideological diversity leads to unmanageable coalition governments that can not hold themselves together or make decisions with all interests included. This would be a concern if the coalition governments during the First Republic were central in the decision making processes, but they were not. Decision that this is the case. These reasons lie in the nature of different elements o f the party system, the institutional rules, and the internal divisiveness of the DC. The logic behind

PAGE 199

199 the process of representation in decision making without understanding these conditions is impossible. The party system itself contributed due to the a ctual and perceived commitments of the PCI and MSI (as well as the PSI pre 1962). Both of these parties were alian democratic and economic order (Newell 2000, 47). The MSI was considered to be supportive of a return to the more authoritarian practices of the fascist regime since the party was originally formed by Mus s that garnered up to 40% support on Election Day were subject to the convention ad excludendum and were left out of ever y post war government. This left the DC, as the largest party, in a position where they had to be the center of every government, eith er as a minority coalition. G overnments formed in this manner, but describing how decisions were reached requires more explanation There are several institutional fe atures of Italian democracy that prevented the government from being the central, representational figure in decision making. The constitution adopted in 1948 was highly influenced by the historical legacies that preceded it and by the interests of the par ties that drafted the document. The historical legacy of fascism and the prevention of a strong leader like Mussolini can be seen in the constitution. The DC and the PCI, the two largest parties, also were concerned with what would happen if the other re ceived a majority of seats in both chambers. The result was a founding document filled with institutional rules that sought

PAGE 200

200 to restrain centralized power and guarantee a government subordinate to parliament. One of these rules was the vote of no confiden ce granted to parliament. This type of confidence vote allowed parliament to dismiss the government without offering a confidence 13 Another was the liberal use of secret ballots for all manner of legisla tive votes (actual legislation or procedural votes). This meant that the government had no mechanism for whipping votes among its members (Hine 1993). Also, individual members had fairly considerable powers to slow down the legislative process. It took relatively few members to prevent the stream lining of the legislative process 14 Most legislation was passed through one of the stream lined procedures, due to the inordinately di fficultly passing laws through normal channels 15 The institutional rules an d nature of the party system left governments in a we ak position to take the lead in the decision making processes. Another reason was the nature of the DC. The oft critiqued instability of Italian governments was due to the nature of the party respon sible for forming governments in the first place. The DC was rife with factions that originally reflected the ideological and social differences within the party. 13 confidence is an exception, comparatively speaking. I mention it here as an institutional rule that might have provided for additional stability. 14 One of these stream lined procedures is the ability of the chambers to grant the permanent committees the ability to pass laws without returning the legislation to the who le house for approval. Others are the various delegations that can be given to the government to pass legislation, again without the final approval of both chambers. For more see Di Palma (1976), Della Salla (1993), and Kreppel (2009). Further discussion of the types of policies passed through these procedures can be found in a following section. 15 The conventional legislative procedure is long and complex. To describe briefly, the constitution requires, any bill be examined by a committee, a vote taken on every article, and then a final vote on all bills submitted to either of the chambers. The process of transferring of a proposal from one chamber to another has to be repeated as many times as necessary until identical texts are passed by a formal conc lusive vote of both parliamentary branches. For a complete description and model of this procedure see Cotta and Verzichelli (2007, 145).

PAGE 201

201 This phenomenon was largely set in motion du ring the 1950 s with the m and in particular the rapid expansion of the state industries, which enticed party groups to develop their own centers of economic power. There have been least seven major factions, at times as many as a dozen. With their own organization and funds, fa ction leaders have their own supporters, territorial bases, and links to organizational interests. (Spotts and Weiser 1986, 29) The preference vote aided and exacerbated the behavior of the factions within the DC. positions 16 Because of this, factions constantly fought over control of governmental posts, positions in the sottogoverno, and state resources. It was the factions that selected the party leaders that served in government and they did by a precise formula that allocated positions based on factional streng th within the party. This was like organisms, the factions themselves have been in a constant state of change growing, dividing, dissolving, and er 1986, 29) When a faction increased its relative strength within the party, it expected that their control of state resources would simila rly increase. The frequent collapse of means for different factions within the DC to take control of the state apparatus based on their relative strength at the time. G overnments fell apart and reformed with regularity, but with a total lack of any kind of renewal. Essentially, the governmental collapses were a reshuffling of the same deck of cards. All of the se conditions including : the nature of the party system, the institutional rules to prevent consolidated power, and the intern al characteristics of the weak DC led

PAGE 202

202 governments. These governments certainly were not in a position to develop and enact a coherent legislative agenda. This meant that the important d ecisions in parliament were made outside of government. For anything to get accomplished in this environment there had to be broad based consensus among the political parties. In order to do this the parties turned to a tradition established in during th e liberal regime of the late 18 th and early 19 th century called transformismo (transformism). This is the practice of setting aside ideologies in order to create broad based, temporary coalitions between elites in order to deliver clientelistic benefits t o the areas they represented. From 1948 1979, transformismo was the principle governing philosophy of Italian political parties. This was especially true for the PCI since they had no real chance of entering the government in an official capacity and bec ause of that had no incentive to behave as a true opposition party. They made a calculation that the only way to the institutional rules I describe above, the PCI could have effectively blocked much of times this number reached 90%) ( Di Palma 19 76; Spotts and Weis er 1986, 113). In return, the parties in government assisted the PCI in passing a number of their private member bills. Through this type of cooperation, the two largest parties were able to service their respective constituencies. E vidence of transformismo occurred from the creation of t he republic until the late 1970 s There were three periods. The first occurred in the 1950 s, where the DC was the sole party in a minority government or formed governments with the smaller lay

PAGE 203

203 parti es (PRI, PLI, or PSDI). During this time, the PCI and DC engaged in fierce ideological fights in public, but this did not prevent transformismo, and ultimately the PCI supported both the existence of the minority government, but as mentioned 70% of the go slative program (Spotts and Weis er 1986: 113). In fact, the opinions of the PCI were more important than those of the lay parties that were actually in the coalition. The second period was during the 1960s when the PSI entered into a coali tion with the DC to form the first center left government of the First Republic. In 1963, the PSI went from an illegitimate party to a member of the government, without Comp and the sottogoverno in exchange for supporting the DC led minority government. This is the purest example of transformismo, where an agreement was reached with an ideolog ically unacceptable party that allowed the PCI to serve in the government, without actually serving in the government. These are just examples of opposition parties being persuaded to participate in the governing of the country. It also established the D C as the sole mediator between the different parties. While not perfect, I believe this behavior encapsulates the representative process identified in the proportional model since, t ransformismo has had the advantage of generating governments well calibra ted to the political mood of the moment and the balance of power within and among the political parties. But, because the majority can not or will no t govern without the acquiescence of the minority, Italy moves only when there is massive consensus somet imes quantified as around 80 percent of the parliamentary forces. (Spotts an d Weis er 1986, 115) Ideally, a proportional electoral system would lead to an ideologically diverse coalition government. But, the model that I drew does not require it, what it d oes

PAGE 204

204 require are representative agents involved in the decision making process. Since governments were weak and often fell apart and reformed, due to the nature of the party system, the institutional rules in place, and the fragmented nature of the DC the Italian legislature had to find alternative means to cooperate. The answer was forward any programs. At this point, I believe the conditions of the proportional model have been met. Italy had a proportional electoral system that delivered proportional results and an ideological diverse party system. These parties found means to r constituents, the DC through C atholic association and cl ientelism and the PCI through ancillary organizations. In the legislature, parties used transformismo to consider the broadest possible coalition to pass any legislative program. I have discussed how parties behaved procedurally, but the proportional mod The next section details the types of policies the Italian legislature produced from 1948 1979. Policy in Italy: 1948 1979 At first glance, it might seem that the Italian political system would have an impossible time functioning. The combination of ideologically diverse parties, weak governments, instability, and the amount of support needed to forward legislation would not seem to be the most conducive to effective polic y making 17 Most scholars criticize the effectiveness of the Italian legislature. The main complaints refer to the clientelistic nature of the policies passed (Di Palma 1976; Golden 2003) or the tendency of elites to 17 This is exactly why Duverger (1984) argues against adopting proportional style electoral systems and advocates ma joritarian rules. For more information see Chapter 2.

PAGE 205

205 punt on the major issues of the day, k icking the can down the road rather than dealing with pressing issues of society (Bull and Newell 2005; Newell 2000; Newell 2010). This section takes a different approach and argues that while imperfect, the Italian legislature was able to actually make s ignificant policy achievements. One of the ways for doing this was a government through consensus and the use of decentralized procedures like bases. This legislat ion might have been clientelistic, but that does not mean that it was not effective. Even without the committees, the Italian legislature passed significant measures in the areas of education, infrastructure, social services, and even economic austerity. These achievements were done (out of necessity) with wide spread agreement among the political parties. This section details how the Italian legislature was able to use committees as well as transformismo to pass legislation before 1979. Committees Befor e explaining how parties used the committee system in Italy, a brief discussion of the Italian constitutional and political institutions is a necessity. The Italian Republic was created in the aftermath of WWII and sought, above all things, to prevent a r eturn to fascism. The institutions and structures that were designed had the twin goals of operating efficiently, while stopping the rise of an authoritarian leader. Because of this, the Italian Constitution includes contradictory characteristics especia lly in the legislative process. The Constitution provides different legislative procedures which favor decentralization namely the passage of laws through the committees, and the more centralized decision making by the government and executive through dec ree laws (both legislative and executive) (Kreppel 2009). This section deals with committees. I argue that the committees, since their use requires wide spread

PAGE 206

206 agreement, can be used as an example of cooperation among the political parties. Later I use the rise of the use of the emergency decrees as an example that cooperation in the policy making proc ess, especially during the 1980 s, was lost. The most widely used of the decentralized procedures in the Italian Constitution is the use of committees in law making. Article 72 of the Constitution requires all legislative proposals be first considered and approved in committee before going to the full Chamber for a vote. The legislature can also decide to give the committees the power to bass binding leg islation without sending the final bill to the whole floor for a vote ( Della Salla 1993; Di Palma 1976; Kreppel 2009). More centralized procedures are included in Articles 76 and 77 of the Constitution. The first of these is the ability of the legislatur e to delegate power to the government with clear, explicit guidelines for the purpose, scope, and length of such delegations (Kreppel 2009, 190). These emergency decrees go into effect for 60 days, but must be presented to the entire legislature the same day for approval. If they are not converted within that time frame the decrees lose all authority from the initial date of declaration. These are some examples of how the Italian Constitution impacts policy making. Some argue that the use of these cons titutional procedures are an indication that effective, coherent policy creating legislature did not exist (Di Palma 1976). Governments, critics suggest, were simply unable to present a series of policies to the full legislature for an up and down vote. Forced to use the committees, all the meaning that committees were used to pass clientelistic bills that all parties could agree on. I forward a different view. Yes, comm ittees were used to pass clientelistic bills, but

PAGE 207

207 that does not mean that these bills were not effective or that the normal legislative procedure could be used in the alternative. The primary mechanism for passing legislation from 1948 to 1972 was through committees 18 As I mention earlier, committees could be empowered to pass legislation without the full approval of the Chambers. During this time period 75% of all legislation was passed through committees without going to the full floo r for a full vote ( Kreppel 2009, 192). This procedure allowed the Italian legislature to pass a staggering number of laws over 8,000 were passed from 1948 to 1972. This begs the question, how do we reconcile the image of the Italian legislature as an ineffectual body incap able of action? One of the most common explanations is that the legislation that is passed through committees are small, clientelistic laws that have a narrow, targeted interest. These private interests. Individual members pursue leggine to satisfy their constituency and are usually allowed through committee through something akin to Se natorial courtesy (Di Palma 1976 ; Golden 2003). These scholars attribute the necessity of these laws to the features of the system. Since governments are so weak, they need a grand coalition to pass legislation throug h the normal procedure. This requires agreement across vast ideological divides. Major legislative programs need the support of both the PCI and DC. An alternative to this is to merely pass these smaller, target laws that all sides can agree upon easily rather than take up the more difficult issues that require compromise. 18 This may seem to place Italian committees on par with other legislatures with strong committees like the United States. To say that committees in Italy are strong is a mistake. Committee offices ar e cramped, understaffed, and short of resources. Committee assignments are not usually sought after by fulltime professional staff, and draconian se niority system that embody the US legislative committees, the Italian committee system pales in comparison and lags well behind (Della Salla 1993)

PAGE 208

208 There is divided opinion about the nature of these laws, however. Kreppel (1997) challenges the prevailing wisdom by suggesting that leggine are merely low cost pieces of legislation which is why they pass through committee without dissent. These bills, she argues, may very well deal with substantive issues rather than clientelistic ones. critique, while taking the traditional view, does not offer any new analysis of the laws either. The position I take in this work is that clientelism is a large part of the Italian political culture of representation (especially in the South). Legislation passed in committees may very well be low cost, clientelistic, and substantive as well. Take for instance, in the US context, the yearly passage of transportation and highway funds. since they service particular states. But, while the improvement of Interstate 95 in Florida does benefit Floridians, it also benefits anyone else who wants to drive to Miami or Disney world. The idea that clientelism and universal benefits are mutually exclusive is false in my view. The idea that clientelism can be combined with universal benefits is not only my view it is the opinion of Italian scholars as well (Piattoni 1998 ; Zuckerberg 199 7 ). describe in a previous section, policies in Basilica, the author details how clientelism was combined with welfare politics to create something positive for the region. Leggine were used to extract resources from Rome and use them in the region to build highways, hospitals, schools, and to

PAGE 209

209 provide welfare and jobs. Piattoni demonstrates how something similar happened in Abruzzo. Further evidence is provided by Ginsborg (1990) who describes how the DC used the state agencies I describe earlier as a means to improve co nditions in the country. One in particular, t he Cassa (or fund), founded in 1950, was an extensive public works program, concentrated on the rural areas of the South. In the first 10 years of the Cassa, extensive funding for the building of irrigation infrastructure, roads, and aqueducts and drains was authorized. Some of the achievements were notable, building programme carried out by the Cassa [were judged] to have been the construction of the first decades after 162) During the same time period one of the state enterprises, ENI, built an unparalleled industrial force within the state sector. ENI was a holding company that eventually controlled, through 5 major operating companies a staggering number of activities. These included control of petrochemicals and construction, texti les, nuclear po wer and research (Ginsborg 1990, 163). The success of some of the state sector enterprises coupled with the need for sector jobs to political supporters created a vicious circle. Legislators and burea ucrats had an incentive to produce government services in an excessively bureaucratized manner. To do so raised the demand for facilitation services. The electoral process did not check this among candidates. If voters disapprove of excessive bureaucratization, electing a less effective facilitator. As the public bureaucracy grows larger, the importance of facilitation gr ows, and a legislator who is a good facilitator will be increasingly likely to be re elected. (Fiorina and Noll 1978, 257)

PAGE 210

210 As a result, the public bureaucracy became larger and larger. When this was controlled by the DC and the economy was growing, this w as no t problematic. During the 1980 s, when the DC lost control over the state apparatus, the politicization of the bureaucracy led to devastating problems for the existing political order. Over time, the clientelistic policies pursued by the Italian le gislature were damaging, but in the 1950 s and s they supported the process of representation. Because of this, immediately rejecting all of the leggine out of hand because of their flaws is misguided. Yes there were abuses, but there were successes as well. Through a combination of leggine and normal legislation, the Italians were able to accomplish a great deal. Parties in Italy were also able to accomplish these things through consensus, certainly imperfect compromises, but accomplishments none the less. It is my contention that a broader look at policy in Italy before the 1980 s is necessary before we can understand what changed and why electoral system took place when it did (and not years before). Policy Accomplishments, 1948 1979 At the end of WWII, Italy stood as an occupied nation facing an economic collapse. By the 1980 s the Italians could boast the seventh largest economy in the world (Newell 2010) 19 At the same time, the per capita income was right near the bottom alon gside lesser developed countries like Spain, Portugal, or Greece By the 1980 s, Italians ranked higher than France and Belgium, trailing only England (LaPalombara 1987). I hold that this could not be accomplished without some serious 19 the 1 980s was somewhere closer to 5 (LaPalombara 1987; Newell 2010).

PAGE 211

211 policy achievements. This sections detail some of those achievements and how they run counter to the general conclusions often made about the state of the country. One of the great achievements made during this 30 year time period was in the areas of education, health, welfa re, and housing. After WWII, some areas of the country (especially in the South) had over a 90% illiteracy rate. By 1980, illiteracy was all but eliminated. Much of this improvement can be attributed to the development of mandatory universal education f or the young as well as highly successful literacy programs for those older members of the population. Du ring the 1970 s alone, Italian u niversities opened up at a remarkable pace as the number of students enrolled in higher education incre ased 12 times (L aPalombara 1987, 163). In the area of health, the parliament passed universal health insurance in 1978. Although in the US the welfare state is considered to be a pejorative, Italians were out in front on providing services to the unempl oyed and poor. During the 1960 s, the Italian legislature put a floor on poverty levels. It also established an interesting means to increase the wages of av erage workers. In the mid 1970 s, parliament reconfigured the scala mobile (escalator clause), which turned out to be one of the most effective means for redistributing income. The result of the law was to reduce the gap between rich and poor (LaPalombara 1987, 164). In the area of housing, Italians established some of the renter friendly policies in the world. Re nts were controlled allowing for more affordable housing prices and renter standards were set at a very high level. All of these policies were achieved in a political environment that has been nstrates how parties, at times, could set aside their differences and achieve agreement on major issues of the

PAGE 212

212 electoral support fell to 37.7% while support for the PCI incr eased to 34.4%. These elections were preceded by a vote in local elections in 1975 where PCI support increased 6% to 33.4% only 2% behind the DC and a vote in European elections where the PCI vote trumped the DC 33.4% to 33% (McCarthy 1996, 103 104). The se election results coincided with the country facing serious economic problems including the rising national debt and a flagging economic growth due to the recent world wide oil shocks. The country needed action and the DC and PCI decided to form a gover nment of government, without actually serving in the government. Since the PCI was officially barred from governing, they had no ministers in the government, but supp legislative program. with some of the economic crises facing the nation. The Andreotti government, supported by the PCI implemented a series of austerity measures. These included raising interest rates from 12 to 15%, tighter limits were placed on the acquisition of foreign currency, prices were increased on government controlled items like tobacco, petrol, telephone services, and electricity, as well as modificati ons made in the wage indexing system (which I ha ve mentioned) (McCarthy 1997, 112). By 1977, the balance from 1973 to 1976 were paid off, and investments country wide inc this time for the good of the country rather than the delivery of spoils.

PAGE 213

213 Some might suggest that I paint a rosy picture of the accomplishments of I talian policy makers and perhaps there is some truth to that claim. But, considering the, these: fragmentation, helplessness, paralysis, immobilism, confusion, lack o f direction, polarization, indecision, scandal, corruption, apathy, alienation, venality, bad judgment, [and] 161) a little balance might be in order. The Italian legislature had to deal with a rapidly modernizing, seculari zing country that grew economically 5% annually, and was able to deal with a remarkable number of issues despite political institutions that were not ideal and a divided party system (Newell 2010). Despite these hindrances, Italians have dealt with matter s of serious import to the country. Some of these have included difficult social questions pertaining to labor relations, divorce, abortion, university and educational reform, creation of a national health service, rent control, as well as the creation o f public financing for political parties. Parliament was also a place where national issues were discussed, and provided a forum for the communist and fascists, who in other contexts have become violent anti system forces that could threaten the existing democratic order rather than participating in it. Rare ly have scholars pointed to these facts to remind others that the which political decisions can be produced in responses to demands originated in other parts of the political 1978, 168) To summarize, the Italian p roportional electoral system performed as it was intended. Proportional elections led to proportional e lectoral outcomes, evidenced by the difference between the percentages of seats to the percentage of votes was

PAGE 214

214 relatively small. These outcomes allowed for an ideologically diverse party system, which meant that the process of representation in terms of d ifferent point of views was satisfied. Parties also developed alternate means of representation. For instance, the DC and PCI used ancillary organizations like Catholic and labor groups to build their memberships. The DC also used different representati onal strategies in the South where the organizational norms were different. Governments were consistently formed around the DC, but the strategy of governing and policy making followed the practice o f transformismo in which parties form broad alliances, a cross ideologically lines to govern. The legislature was able to use committees to service their clientelist bases. Also, the legislature was able to pass significant legislation, contrary to popular beliefs about the efficacy of policy making in Italy. At this point what remains is to describe how the steps in the model led to higher levels of support. Like Chapter 3 on New Zealand, I demonstrate how support declined when there were process and policy changes follow ing the 1970 s. To foreshadow, the c haracter of representation changed in two different ways. The first was the decline the organizational groups that the parties had come to rely. The second had to do with the type of coalitions were formed and the amount of cross party cooperation that e xisted in the past. Finally, I examine briefly the inability the legislature had dealing with the pressing issues of the day. The Changing Nature of Representation in Italy The previous section demonstrates how Italians were represented by their political parties. Specifically, I use the DC and the PCI as example of how parties represented their constituents both socially and through policies, whether clientelistic or substantive.

PAGE 215

215 This section describes what changed in the process of representation and w hy that was the case. Before going into detail about representation, I would like to examine some data that give some indication that something has changed with regard to the relationship between political parties and the public. The first is the over dec line in electoral support for the two largest parties, the DC and the PCI. One of the more interesting features of Italian voting behavior is that there was little volatility. People tended to go to the polls and vote for the same parties over and over a support totaled 38.3% in 1963, 39.1% in 1968, 38.7% in 1972, and 38.3% in 1976. place, this is a remarkably consistent result The PCI was more successful in bringing in new voters during this time. The party increased its vote share in every election, rising from 25.3% in 1963 to a high of 34.4% in 1976. After this election both parties, who represented the two largest subcu ltures and dominated Italian politics for close to 20 For comple te election results see Table 4 1. The decline in support for the two largest parties w as a boon for the non cartel parties. These are the parties out side of the traditional cartel parties (DC, PCI, PSI, PSDI, PLI, and PRI) who were the effective parties of government. These were the parties that could serve in government or the PCI that were vital to its functioning. 20 By the 1992 election, the PCI split into two separate parties, the PDS (Democratic Party of the Left) and the RC (Communist Refoundation). These two parties won 16.1% and 5.6% respectively in the 1992 previous election.

PAGE 216

216 status quo. On this list of parties includes the Neo fascists, Radicals, Monarchists, the N orthern League and La Rete to name several. Some of the ideological positions held by these parties included: undermining any legislative actions (the Radicals), secession for the Northern regions (Northern League), a return to a constitutional monarchy (M onarchists), and an outspoken anti mafia party (La Rete). What all the parties shared was distaste for the clientelistic practices and bargaining of the cartel parties. In the a total of 10% in 1979, to 12.8% in 1983, 14.0% in 1987, and 19. 9% in 1992 (Bardi 2002, 56). The willingness effective political parties from 3.5 in 1976 to 6.4 in 1992 (Bardi 2002, 52 ). The rise in the number of parties and the willingness to vote for them are indications that the votes of Italians were increasingly up for grabs. Further evidence that this was the case is demonstrated in the rise in voter volatility which increased f rom 4.9 in 1972 to 14.2 in 1992 (Bardi 200 2, 52). These are merely rough sketches of an electorate that was changing, especially with regard to the relationship between voters and political parties. Although long t erm party identification data are virtual ly non existent, the patchwork data that d o exist all suggest the same thing: party identification severely diminished a fter 1976 (Mannheimer 1989 ; Schmitt and Homberg 1995). One of the few consistent survey items is from the Eurobarometer which measures the level of attachment to one of the political parties. A respondent

PAGE 217

217 has declined signifi cantly since the 1970 s. A depiction of these r esults is displayed in Figure 4 1 This reason that I list the decline in support for the DC and PCI, the rise in support falling levels of partisan attachment is to demonstrate that there was a change i n representation in Italy after the mid 1970 s. What remains is to explain what this change was and why it happened. It is my belief that while the electoral system continued to do its job, the representative features of the system changed. They changed in three main ways: the DC and PCI could no longer count on their traditional bases of support, corruption replaced clientelism as the prevailing process which undermined representation, and the types of coalitions that were formed were different than in t he the day. The following sections detail this changes and how they related to the decline in support and a change in electoral system. Changes in Process: The Loss of Representation, 1979 1992 In an earlier section I explain how the DC and the PCI, as the two largest parties representing the two largest sub cultures, used ancillary organizations to represent their respective constituencies. The DC relied on Catholi c groups in the North and extensive clientelist relationships in the South to maintain consistent electoral support before 1976. For various reasons, the strength of the Catholic associations waned, which led to a loss of control of clientelistic structur es, which led to an overall decline in electoral support. The PCI was always in a difficult political situation as a party permanently barred from governing. Unlike the DC, the groups associated with the PCI did not lose membership, but the party itself lost membership over time. The reason for this is the

PAGE 218

218 gained membership and electoral support, but by entering into an agreement with the DC, the party alienated many of its core constituents. The following years saw a decline separation of both the DC and the PCI from their respective bases of representation. The DC relied on three th ings for its electoral support: Catholicism, clientelism, and anti communism. The original basis for DC electoral success was the sustained support of Catholics. One of the most distinguishing features of DC voters was their commitment to Catholic religi ous values. As church attendance went up individuals were more likely to vote for the DC. The party itself did not have a strong organizational basis and instead i t relied on ancillary Catholic o rganizations like Catholic Action and the Civic Committees for assistance during election season The strength of these org anizations declined by the 1970 s due to a rapid period of modernization and the ensuing secul arization of society. The 1950 s and s were a time of economic growth, technological growth, an d migration. Younger Italians were afforded educational opportunities, which dramatically increased literacy. The television became a fixture in most Italian homes during this time period as well. Also, Italians were increasingly mobile as millions of S outherners moved to the N orth for employment opportunities. Traditional, religious ties were being replaced by a more modern, secular society. The strength of the DC was eroded because of two major components of this modernization and secularization proce ss: the decline in the number of Italians who considered themselves religious and the decline of religious practices even among those who considered the mselves Catholic (Pasquino 1980 92). Overall by the 1970 s, fewer

PAGE 219

219 Italians attended church which dimini shed the influence that the Catholic hierarchy had This process of secularization and the decline in importance of religious attitudes had a visible effect on the membership of important Catholic organizations. For instance, Catholic Action a group that the DC relied on heavily for electoral support as well as the recruitment of party figures had almost 3,000,000 members in the 1950s, but by the late 70s, that number dropped to close to 600,00 0 (Pasquino 1980 93). The It members to 400,000 in the same period. One of the fundamental pillars of DC support was now eroded. With this came a loss of electoral support and therefore a loosening of the par 21 The party still relied on staunch anti communism during the 1980 s to hold on what remained of its electoral support. The transformat ion of the PCI in the late 1980 s would make this last component irrelevant. The PCI becam e two sep arate parties in the early 1990 s. One party, the Democratic Party of the Left ( PDS ) wanted a break from the communist past. The more hard line communists formed the Communist Refoundation ( RC ) The transformation of the party was in response to the end of the Cold War, but also due to larger issues of secularization and modernization that hurt the DC, proved to be a boon to the PCI, whose main basis of support was from intellectuals and members of the working class. From 1970 to 1976, the party increased its membership about 300,000 members to 21 This is a point that I return to in a later section on coalitions, clientelism, and corruption.

PAGE 220

220 1,800,000 members (Hine 199 3, 114). Between 1972 and 1976 the party also a strong presence in the trade union movement, especially in the three largest confederations. It has a major stake in the cooperative movemen t, and in various other commercial enterprises. It mounts an ambitious program of summer festivals which are cross between popular cultural gatherings and political propaganda. And it takes very seriously its role of political education and stimulant to social and economic policy research, running regular party schools, research institutes, con ferences, and so on. (Hine 1993, 115) Their electoral success can be attributed to their strong organization and support from their traditional work class base. Bu t, the party always faced an organizational dilem ma. They were criticized by the DC as an anti system party committed to fundamental change of the Italian state, while their legislative behavior could be described as cooperative and responsible (Newell 20 10). The party seemed stuck between its ideological branding as being support ive of change and its desire to be taken seriously in the legislature. This contradiction continually plagued the party and is suffered by the party in the ensuing years. After the 1976 election, the DC and PCI came to an agreement where the PCI would support a DC led government as a member of the government outside of the government. The Comm unists justified this decision by saying that the country was facing a crisis that could topple the entire constitutional order. The economic situation in the country was dire, with mounting public debt, increasing inflation, and low economic growth. Thi s economic situation, they believed, required reformist policies and a program of austerity. The party felt that the country could not afford to wait for an

PAGE 221

221 national unit y coalition, the Communist leadership argued, could expect to gain broad consent from diverse strata of the population necessary to undertake the social policies th 127) The interest in the party leadership in entering int o agreement with the DC did not rece ive wholesale support from the P arty G eneral or the grassroots. As the national party asked for traditional bases of support this disagreement became more obvious. Complaints by party l eaders echoed the increasing and more intense arguments seen at local party meetings. Although the party might be commended by a neutral observer for helping to get the country through a difficult economic period, they did little to satisfy their own base After a decade of party membership growth and electoral success, both began to wane. During the 1980 s the party lost 400,000 members and 13% of the vote. Part of this can be attributed to The 1980 s were almost a decade of soul searching for the PCI. The DC was not the only party facing a changing electorate due to modernization. The par ty, with its commitment to centralized planning, seemed to be consistently at odds with what some typified the country (Hine 1993, 115). Hine argues that during the 1980 s the Italian working class was becomin g increasingly successful and diversified, which caused workplace ties and solidarity to diminish. The party membership roles, which in the past were full of the working class, day laborers, and share croppers were gone. Their replacement by new members from the white collar middle strata brought into the party recruits with a more tenuous commitment to repetitive

PAGE 222

222 organizational party tasks like selling newspapers and making pasta for party rallies, and far less need for the networks of social support (wo rking operatives, etc.) traditionally pr ovided by the party. (Hine 1993, 116) The changing and declining membership coupled with the political miscalculations ld have led to a slow death for the PCI if the downward trajectory of electoral results continued. Hist ory intervened in the late 1980 s with the revolutions in Eastern Europe and the eventual fall of communism. Party leaders within the PCI, most notably Achille Ochetto who became party secretary symbols (if not the ideological essence) of positions now ostentatiously abandoned by the political s 117) His solution was to re brand the party under a new name, the Democratic Party of the Left (PDS), with a new commitment to social democratic values where progress ives and leftists could find a political home. The PDS contested its first election in 1992 winning a mere 16.1% of the vote 22 With the end of the Cold War and the reconstitution of the PCI, the DC lost another of its justifications for holding onto poli tical power. This is a point that I return to later, but what is worth noting now is the remarkable decline of the two pillars of Italian democracy since the beginning of the Republic. At their highest point, the DC and the PCI occupied close to 75% of t he vote. By the elections in 1992 that number was closer to 45%. The traditional mechanisms for representation, the ancillary organizations, 22 A more traditional commun ist splinter group the Communist Refoundation (RC) polled at a little over 5%.

PAGE 223

223 were gone as well. This is a key component to understanding how the proportional model of democracy broke down i n Italy. Parties, as the DC and PCI exemplified, had deep social roots. Parties cultivated these ties with newspapers, social clubs, or religious and workers o rganization. By the early 1990 s these ties were all but gone. The decline in social ties is o ne major process change, but there was also a change in clientelism as well. Clientelism served as a main representational characteristic between the DC and its southern base. As the party lost electoral support, it had to increasingly rely on other part ies to hold governing coalitions together, and as a result it lost its stranglehold on state resources. This increasing marketplace for clientelism led to another drastic change in representative processes. The following section describes these changes. Coalitions, Clientelism, and Corruption DC was still a member of every coalition g ov ernment, but was joined by the three lay parties of the center and the PSI. The coalition government was made up of the DC, PSI, PLI, PRI, and PSDI. It was also the first time that the DC gave up the Prime ni Spandolini, in 1980 and to a Socialist, Bett ino Craxi, in 1983. The coalition was problematic for several reasons. One was the ideological differences between the parties, for example the PLI and the PSI had never consented to join a government togeth er. A second problem had to do with the number of parties needed to form the government in the first place. The consequence of the growing fragmentation of government was to increase the number of veto players. That is, the number of parties necessary t o make any decision was increased (Tsebelis

PAGE 224

224 1999). This impacted the processes of representation in two different ways: it made the sort of consensual decision making processes (characterized by transformismo) more difficult and it opened up the vast stat e sector and the spoils of office. Previous governments, while unstable, were able to pass a significant amount of legislation either through committees or through the normal legislative process with the support of large majorities that cut across party li nes. Governments of the 1980 s were increasingly unable to get the entire parliament to agree on passing their agenda or point for cooperation when 85.1% of government bills pa ssed In each successive legislature, this percentage fell: 72.4 % between 1979 and 1983, 72 % from 1983 to 1987, and 69.1 % from 1987 to 1993 (Cotta and Verzichelli 2007 154). The Craxi administ ration was the most successful in terms of longevity serving from 1983 1987, but faced the on the government back benche s. These snipers us procedures to block legislation to which factions in the DC were opposed. One example was the defeat of the government on a proposal to reform local government finance in 1986. The represented the 160 th such defeat of a gove rnment proposal in less than three years of office (Gilbert 1995, 12). The PSI was able to maintain the confidence of parliament with the tacit support of the PCI and left leaning factions within the DC (as well as some support from the lay parties). Wha t they could not count on was the cooperation of these parties to pass their program or refer their bills to committee 23 23 For a numerical breakdown of the decline of leggine passed through different Italian Legislatures see Kreppel (2009, 193).

PAGE 225

225 The lack of cooperation was a large prob lem for governments in the 1980 s Since governments could not rely on the cooperation of parli ament, they had to find alternative means to service their respective clientelist bases. The use of emergency decrees controlling the legislative agenda. Emergen cy decrees guaranteed that the bills would be placed first on the legislative calendar, but they did not ensure their conversion by the entire legislature. The parliament, especially those in opposition, used a number of stalling tactics to make sure that the 60 day deadline for conversion was not met. Governments responded by continually reiterating the same decrees (Kreppel 2009 ) Unconverted decrees are an indication that the level of cooperation declined. In earlier legislatures before 1972, 95% of decrees were converted into law. In the 6 th and 7 th Legislatures (1972 1979), this numbe r fell to 85%. During the 1980 s the number of unconverted decrees increased significantly. During the 8 th legislature (1979 1983) this number fell to 38% and fell ag ain to 17% in the 12 th legislature (1994 1996). As the number of unconverted decrees rose, the rate of re iteration increased from 25.2% to 49.7% (Kreppel 2009: 198). From 1948 to 1972, only 11 of 300 decrees were reiterated. In the following 24 years t his number ros e 955 out of 2537 (Kreppel 2009, 197). By the early 1990 egregious that it could no lo 198) A representation of the rise in unconverted decrees is di splayed i n Figure 4 2 F igure 4 2 shows that the number of decrees issued increased as the number of decrees converted decreased. In the legislative period of 1976 1979 the parliament converted 136 of 167 decrees. When the Socialists claimed the premiership in 19 83,

PAGE 226

226 the numbers and disparity increased in every legislature. In the period of 1983 to 1987, 459 decrees were issued against only 187 converted. From 1987 to 1992 these numbers stood at 493 to 123, with an astounding jump from 1992 to 1994 where 718 decr ees were issued, while only 122 were converted by the full parl iament (Kreppel 2009, 196). In short, the 1980 s saw the rapid rise of both unconverted and reiterated decrees, which represented a real change from the more cooperative representative processe s of the prec eding years. This interests and represented the increased distance between the Italian people and their representatives. This is just a representation that processes were changing and I am not ma king the claim that the Italian people were up in arms over the number of had changed about what constituted appropriate behavior. The other major representationa l process change was the increased marketplace for clientelism and outright corrupt behavior. Governments in the past were dominated by the DC. This meant that the party controlled much of the sottogoverno and was responsible for much of the clientelism that occurred in previous decades. As they lost electoral stre ngth beginning in the late 1970 s, they had to increasingly turn to other parties to hold a government together, 80 s. The consequence of the pentapartito coalition government was not limited to the decline in cooperative behavior as I describe in previous passages Perhaps the more important was the opportunities it presented other parties to demand an outsized sha re of the spoils in exchange for their support for the government. The DC was in a diminished

PAGE 227

227 position and was forced to come to an agreement with parties that had been shut out of the state system for decades. The lay parties (PLI, PRI, PSDI) were frequ ently given the same number of ministerial seats as the DC even though their total electoral support can Spandolini whose party only received 3% of the vote in the 1979 election. Bett ino Craxi and the PSI held on to the Premiership for nearly four years with only 14% of the national vote. Parties within the coalition fought over control of the positions in the sottogoverno as well. The PSI had the most success, since it was the largest of the coalition partners (and perhaps because they were a potential part ner for the PCI). In 1980, it took over the Ministry of State Participation, which the DC had co ntrolled since 1956. The Socialists smaller savings banks, and 25% of the board positions in other state en terprises. The PSDI took control of several banks and the top position of EFIM. The Liberals and Republicans split Vice Presidencies of ENI and IRI as well as seats on the boards of ENEL, ENEA, and variou s banks (Spotts and Weiser 1986, 148). The reason that this competition that took place over the state spoi ls was two fold. First, i t was a time when the voting public was less inclined to identify with particular sub cultures that the parties had represented in the past. Scholars argue that in the [si c] cast this type of vote consider their act as the affirmation of subjective identification with a political force which is seen as having an organic liaison with the social group to which the voters isi and Pasquino 1980 17) These types of connections wer e all but gone

PAGE 228

228 by the late 1970s. By the 1980 off, for the satisfaction of a need or the achievement of an interest of the vo 17) Two pieces of evidence support this conclusion. The first is the increase in voter volatility duri ng the 1980 s, which suggests an electorate that was more willing to change their vote. The second is the dramatic i ncrease in the use of the preference vote, especially in the South where vote rs used their preference votes three times more than they had in the past (Golden 2003). Parties simply had to have some sort of resources to distribute to their clients or they would lose their support. The second reason that competition over the spoils took place was the rise of a different class of party members and politicians and the opportunities that these politicians h ad by the end of the 1970 s. By this time there was a generational change from the earlier party members who represented the social groups and ideologies of society. Della Porta and Vannucci (1999) describe this evolution, A new political class that formed in the wake of the first experiences of government replaced the prewar political class, which had maintained its power within the party through traditional political resources (such as charisma). This new political class consolidated its power through the occupation of party appointed positions in the pub lic sector, positions that appropriation of public resources. Business politicians, in fact, lined their pockets through bribery and the exploitation of their political power in other activities, particularly where that power furnished them with an edge over their competitors. (73) The business politician viewed the party as a means for personal ga in through corrupt practices. Business politicians used different institutional me chanisms, divergent political offices, with varying degrees of success. Those who believed politics was just

PAGE 229

229 another business and means for upward mobility generally demonstrated three main characteristics. Th e first characteristic of business politicians was their lack of any traditional political resources, namely money, power, or prestige when they entered politics (Della Porta and Van nucci 1999, 73). These politicians started literally at the bottom and were irrespective of their slogans or ideas, is actual 241 ; Della Porta and Vannucci 1999, 73) Although these politicians were not bound by ideology, they all displayed a secondary characteristic, that is, they used the party apparatus as a means o f upward mobility. Careers in public service depended on membership in the party rather than a certain set of skills or administrative capabilities. Due to this, the hierarchy of political posts also became somewhat confused. Being a driver f or a powerful political boss might be more attractive than being a low level elected of ficial. This is because being just a driver was no impediment to running a hospital or being placed at the head o f an important public entity. Business politician s are concerned with the potential for personal enrichment above all else, so the importance of a post was determined by the po tential for gain. Finally, politics becam e a means to achieve political success. That is, this new class of politicians entered politics solely for the purpose of gaining a political position. This may seem circular, but consider traditional politics. In many countries a lawyer, businessperson, even priest may use their professional qualifications and connections to run for office. In Italy, individuals join the

PAGE 230

230 political party in order to gain a different political position, one that hopefully leads t o some kind of private gain (Della Porta and Vannucci 1999, 74 75). By the beginning of the 1980 s many things were coming together to make corruption a problem that had severe consequences for Italian democracy. One of these is the increase in the numbe r of available jobs in the state sector and civil service. A study by R obert Putnam (1973) in the 1970 s discovered that most of the public sector and state service employees had been in the same jobs since the fascists were in power. As these workers wer e due to retire, a fierce political struggle over which party woul d get to replace them ensued. Business politicians took advantage of this vacuum and found their way to extract resources from these jobs. As public documents related to the Mani Pu lite in vestigations of the 1990 s demonstrate, these resources were larger and larger sums of money. Several authors agree that the 1980s were a time where large scale bribes for public goods increased exponentially ( Del Monte and Papagni 2007 ; Golden 2003; Moss 1995 ). Perhaps no party better represented the rise in corrupt beha vior than the PSI, led by Bettin o Craxi. He is credited, through court documents, with creating a complex set of strategies for controlling the release of public information, jobs, and g oods for bribes and kickbacks. He used trust lieutenants to check on PSI members in elected and eceived by the party (Moss 1995, 64). Craxi was also responsible for preventing and fighting the magistr ates who attempted to prosecute this illegal behavior. In 1981, Craxi fought the prosecution and jailing of Roberto Calvi, a noted mob banker and PSI contributor, later securing his release. When Craxi was elevated to Prime Minister, he issued a decree s hielding his

PAGE 231

231 dealings. There are additional examples, but the important point is that the PSI represented a party rife with business politicians who changed the natur e of politics by demanding fees/bribes for public services. The question might be asked, what is the difference between the kinds of bribes that the DC demanded in earlier decades to the kind of corruption represented by the behavior of business politici ans ? One answer is the sheer sums involved and the frequency of the accounts of corruption uncovered by the Mani Pulite investigations. One scholar suggests that the level of bribes that existed i n the 1960 s was more likely to be small, in the 30 to 60 d ollar range (Sassoon 1995). Evidence of larger bribes proved more idiosyn cratic; where by the early 1990 s the exchange of large sums appeared, to ordinary Italians, more systematic. The Mani Pulite investigations provided Italians with daily reminders of the corrupt behavior of the political class. These investi gations began in the early 1990 s with the arrest of Mario Chiesa, a socialist manager of a Milanese charitable organization. Chiesa was caught in the act of accepting a bribe of about 4,000 dolla provided. Chiesa was th e textbook business politician He commanded a salary of 50,000 dollars, but was found to have over 9, 000,000 dollars in assets in property. This story made the front page of every paper and due to the nature of Italian legal 24 Faced with the harsh reality of 24 Prosecutors in Italy have the unusual power of kee ping those accused of crimes in jail indefinitely if there is a reasonable assumption that they might destroy or hide evidence against them if they are given ecutors

PAGE 232

232 prison, Chiesa named names and figures. When t his happened, the extensive web began to unravel and the indictments for extortion by a public official and corruption of a public official exploded 25 Those incl uded in the indictments were former prime ministers and prominent members of the business comm unity 26 The sheer sums involved were staggering as well. For instance, the stories included Raoul Giardini the 95 million dollars in bribes. 60 million of these wer e found in secret accounts in different countries, and the remaining was kicked back to Craxi, Forlani, and various Socialists and Christian Democrats. The PSI, from 1987 to 1992, collected a total of 119 million dollars in funds from variou s private enti ties (Sasoon 1995, 138). The public certainly felt that corruption was going through a d ramatic upswing during the 1980s and 1990 s. To measure these feelings I use two indices of corruption from a study by Del Monte and Papagni (2007). The first of these indices is the number of reported corruption crimes country wide 27 The second measure is from were able to arrest and hold other implicated parties, like second tier politicians. This led to more confessions and eventually prosecutors were able to implicate higher political figures like Bettino Craxi. 25 Number of individuals indicted of Ex tortion by a Public Official: 1992 (588), 1993 (1043), 1994 (1644), 1995 (1764). Number of individuals indicted of Corruption of a Public Official: 1992 (431), 1993 (1178), 1994 (2407), 1995 (1764) (Fabbrini 2000, 177). 26 The web of corruption uncovered b y the investigation is complex and a proper description requires more space than this chapter provides. For a general explanation of the webs of corruption see Della Porta and Vannucci (1999, 2007). For a more detailed explanation of the actual events of the Mani Pulite investigations see Gilbert (1995, 126 155) and McCarthy (1995, 139 167). 27 It should be noted that this index has some drawbacks. The first is that this is not an adequate reflection of the actual amount of corruption, since these are onl y reported crimes. But, this is perhaps not problematic because the index shows an increase of corruption crimes over time. This could be due to an increased willingness to prosecute these crimes or in the reporting of these crimes. There is also the is sue of temporality, reported crimes do not happen in the year in which they are reported, they happen at t 1. This would be a larger issue of the laws on corruption changed over time, but in Italy there were no discernable changes. I believe the validity of the measure is increased by including the second perceived corruption index as well (Del Monte and Papagni 2007, 384).

PAGE 233

233 Transparency International, which used a collection of in country surveys to gauge the level of perceived corruption in the country. Results are displa yed in F igure 4 3 : What these measures both show is something similar to what I have already argued. Both the number of corruption crimes reported and the perceived feeling of corruption rise, startin g in the late 1980 s and continuing until the first election und er new electoral rules for the Chamber and Senate in 1994. The sharpest of the upturns is following the beginning of the Mani Pulite investigations that began in 1992. Here we see a dramatic rise in both the number of corruption crimes reported and the p erceived level of corruption. The per iod beginning in the early 1980 s introduced new representation processes to Italian democracy. The electoral system continued to produce proportional results, but the parties did not represent their constituents in the manners that they had in the past. Modernization and secularization led to a decline in the traditional sub cultures that the DC and PCI had represented so successfully in previous years. The two main parties that had dominated Italian elections began t o lose support and members as the 1990 decision making processes that had reigned in the past. The use of committees and transformismo were replaced by reiterated emergency decr ees and the pentapartito. Clientelism, which was used to build a support system in the South and grease the wheels of the bureaucracy, degenerated into outright corrupt practices by the political parties. The behavior of the party system was both represe ntative and not representative depending on the time period examined. Policy did not go through such

PAGE 234

234 a dramatic change, but the ability of the Italian legislature to deal with the pressing issues of the day (albeit imperfectly ) was lost during the late 19 80 s. Policy, 1980 1992 Policy making in Italy had always been a product of the processes of the party system. The processes that w ere predominant before the 1980 s were transformismo and clientelism. Policies reflected these processes, as parties cooperat ed to pass legislation through committees that would service their respective bases. There was actually some success and the country experienced rapid economic growth during the s and s. Substantive policies were also passed during this time as the legislature helped create national systems of health, transportation, and education. Also, when there were pressing issues that required attention, the legislature was able to address those issues. An example of this was the rising public debt and econo mic crises in the mid 1970 s. The DC and PCI were able to come together and pass austerity measures that righted the course of the coun s these legislative accomplishments came to an end. Clientelism was still important, but the tendency to steer projects toward those who paid the required bribe to the parties were more likely to receive public contracts. The country also faced the familiar problem of rising public debt, but was unable to find the answers as it had in the pas t. Policy in the 1980 s was used to support a vast clientelist system. Most of the public spending was centered in the increasingly bloated public sector. As I describe in an earlier section, parties were under increasing pressure to provide jobs and bene fits to their clients who were more willing (due to the lack of ideology or social roots) to behave as entrepreneurs than party members. By the 1980s the Italian state employed 4,000,000 people, a full 1 in 7 of all working Italians (Gilbert 1995, 17). T he pentapartio

PAGE 235

235 government had no solutions to streamlining this bloated public bureaucracy. Instead, they increasingly used state resources to enrich themselves. Some estimate that the cost of this behavior from 1980 to 1990 reached 1 trillion dollars. Some of this went directly to politicians in the form of bribes, but the real waste, in economic terms, was the billions wasted on unnecessary publ ic works projects (Gilbert 1995, 130). Public works for tens of billions of dollars were co ntracted every y ear in the 1980 s, by provincial and regional governments, as well as the national government, and the political parties creamed off 5 15% of many or most such contracts, is an undoubted fact. The parties were paying for their glamorous congresses, subsidi zed newspapers and intellectual magazines, lavish receptions and extravagant electoral campaigns with what amounted to a d isguised tax. (Gilbert 1995, 130) This behavior was not only corrosive to the representative processes I describe in the previous sect ion. It had a real effect on the economy and budget. In 1980, the reached 92%, and 10 1% of GDP by 1992 (Gilbert 1995, 17; McCarthy 1997, 129). Of countries in the European Community only Belgium had a higher level of public debt at a faster rate (Gilbert 1995, 17). This was a pressing issue to the country, since Italy was one of the original members of the EC and would not be allowed into the co mmunity with such a high level of debt. The government was both unable and unwilling to make cuts to the bloated public and civil services. The solution they tried was to increase tax rates across the board. In 1980 the government took in 3 2% of GDP in taxes, by 1990 it was 40 % (Gilbert 1995, 18). This was the highest increase in all of Europe. In true Italian fashion, rather than apply the law universally, the government targeted certain categories, forcing some to adopt intricate bookkeeping practice s. For example, restaurant owners and shopkeepers were required to purchase special cash registers (or shoulder the cost of modifying the old

PAGE 236

236 ones) to submit to state controls. These were not the only controls imposed on businesses. Italian coffee bars, for instance, have to provide customers with a special distance from the bar. Until that time, they are required to show the receipt to a police officer to demonstr ate that the bar has paid the proper tax on coffee sales. Italian business owners had to choose between costly alternatives around the new laws, or deal with a maddening state bureaucracy. Self employed Italians generally tend to avoid paying taxes or hi de part of their incomes, so the state was forced to target employers and their employees. Due to the high rate of tax evasion, the state had to increasingly rely on indirect methods (like those mentioned above) as well as t ax increases on those who were f orced to pay tax (Sassoon 1995, 137). There were two consequences to these new tax policies. First, they did not work. Taxes went up, but the public debt continued to climb. The solutions and cooperation that put a halt to the rise in public debt durin g the mid 1970 s was gone. Secondly, there were increasing segments of the Italian population that grew disconnected from the state. Individuals, who do not like to pay taxes under ordinary circumstances, became bitter at the prospect of being singled out by a state they did not much care for. The Italian legislature was succeeding, through its policies, to separate itself from their citizens and the citizens from one another. These changes were exacerbated by the rampant corruption that the parties in I taly practiced and condoned following the and policies changed in Italy and how the models help explain the rise and fall in support in the country.

PAGE 237

237 Conclusion: The Propor tional Model and Support in Italy, 1948 1992 The argument of Chapter 4 is that proportional election in Italy performed their functions for close to 50 years. What changed was the behavior of the p arty system. Prior to the 1980 s, the parties performed th e process of representation well. Parties represented no t only a vast array of ideological positions, their activities extended into developed, the parties used clientelis went a bit further, due to the difficulty of passing legislation through normal procedures, parties routinely cooperated to pass legislation. Much of this legislation was small, clientelistic laws called leggine that serviced a particular constituency. This was not the only way the legislature operated however and the parliament deserves credit for the many social problems that it was able to deal with. This (imperfect) effectiveness disappeared followin After this date, longer term trends that undermined representation, were becoming apparent. The parties saw the particular subcultures they represented disappear as modernization and secularization over took the country. Parties were left only with their clientelistic bases and found themselves in an increasingly competitive marketplace for personal gain and the process of representation was replaced with corruption. The legislature was also unable to garner the kind of cooperation that it had in the past. Governments had to rely more and more on emergency decrees in order to deliver publi c goods to their constituencies. The government was also unable to pursue effective policies to curb the level of public debt. Where the legislature had succeeded in passing legislation when it became absolutely

PAGE 238

238 necessary previous ly, the governments of t he 1980s and early 1990 s were completely devoid of effectiveness. I expect that the variations in support will be indicative of these changes. I have already mentioned some measures of support like party membership and the support for alternative parties. The following sections detail further evidence such as satisfaction with democracy, voter turnout, and further measures of party membership. Satisfaction with Democracy Satisfaction with democracy in Italy has always been low. For most of the 1970s and 1980 s, those dissatisfied with the way democracy worked remained somewhere between 70 % and 80%. What is inter esting is that in the late 1980 s there is a precipitous drop in the level of those satisfied with democracy. In 1989 31% of people expressed sat isfaction, by 1993 that number dropped to 10%. These fi gures are displayed in Figure 4 4 Turnout Turnout levels, unlike satisfaction with democracy, had always been high in Italy. In fact, turnout was among the highest in the world (generally trailing o nly Austria ). Italians have mandatory voting but do not have a process of voter registration (like Americans). Instead, they are registered from birth in the town or city in which they were born. Aided by discounted train tickets, vast majorities of ci tizens traveled (sometimes hundreds of miles) to their home towns to cast a ballot. What does this behavior signify? polling booth is a place where the average Italians is able to as sert, to reaffirm to give testimony voting is not as public as the evening stroll, the ubiquitous passeggiata, but it serves a similar purpose. Both acts are intended to establish or reaffirm

PAGE 239

239 for onesel f as well as for others, who one is. This identity who and what rson belongs. ( LaPalombara 1987, 136) chance to reaffirm their id entities. This is what makes the decline in turnout even more interesting. When the subcultures became less impo rt ant in the middle of the 1970 s parties might enact, and stayed home. From 1976 to 1994, turnout fell 8% This is precisely the time when the changes in policy and proce ss happened. Italians were less and less willing to participate in a system that neither represent ed them nor provided policies the same way they had in the past For a representation of turnout in Italy see Figure 4 5 Both satisfaction with democracy and turnout fell during the years that policy and process changed. Further evidence that support fell are in the level of party membership. Party Membership Parties, which had been the strong point of Italian democracy before 1979, saw t heir influence wan e in the 1980 s. Italians were less and less willing to join political parties. Every one of the major parties lost drastic number of members during this tim e. In 1987 the DC had 1,800,000 members by 1991 they had lost a full 600,000 mem bers. The PSI we nt from 631,000 members in 1987 to 51,000 in 1992. The PCI, as I note earlier, reformulated itself into the PDS (and a splinter group the RC). The P CI had 1,500,000 members in 1987, in 1992 the PDS only managed to keep 770,000 of these membe rs. The RC h ad 119,000 members, so all tog ether the former PCI lost 700,000 members. The lay parties involved in the pentapartito also lost significant membership.

PAGE 240

240 From 1987 to 1992 the P SDI went from 215,000 to 133,000 th e PLI went from 59,000 to 40,000 and the P RI went fro m 108,000 to 72,000 (Bardi 2002, 55 57). Membership as a percentage of the electorate went d own considerably after the 1970 s. This dec line is represented in Figure 4 6 Again, what these results display is an increasing tendency for Italians to disengage from the political system. Overall there are three main pieces of evidence that suggest that the level of support in Italy fell dramatically d uring the 1980 s. The first is an 18% drop in the level of satisfaction with democracy from 1987 to 1 992. The second is a downward trend in the number of Italians turning out to vote, an 8% drop from 1976 to 1992. The final piece of evidence is a precipitous decline in the number of Italians who were members of political parties. It is not coincidental that all of these decreases took place when the Italian party system was dramatically changing its behavior. Parties were no longer representing their constituents as they had in the past and they were no longer able to deal successfully with pressing is sues of the day through effective policy. Moving forward Electoral system change in Italy and New Zealand both happened in the vacuum of support created by dramatic changes in policy and process. What is interesting is that both New Zealand and Italy t he electoral systems behaved exactly as intended. New Zealand overrepresented the two largest political parties, while Italy provided opportunities for smaller parties to have a place in the legislature. Both countries were able to make their respective electoral systems work when the parties respected the processes of accountability and representation. By doing so, governments were able to pass policies that the public wanted and support for the system was high. Fundamental changes in these policies an d processes led to a decline in public support. Electoral

PAGE 241

241 system change happened when these changes oc curred. The remaining sections of this work discuss whether different electoral systems in Italy and New Zealand were able to correct the issues of the past by changing electoral rules.

PAGE 242

242 Table 4 1. Perc entage of the vote and d isproportionality, Italian Chamber of Deputies 1948 1992 Parties 1948 1953 1958 1963 1968 1972 1976 1979 1983 1987 1992 % of the Vote (Disproportionality) DC 48.5 (4.6) 40.1 (4.5) 42.4 (3.4) 38.3 (3.0) 39.1 (3.1) 38.7 (3.5) 38.3 (3.5) 32.9 (8.7) 32.9 (2.8) 34.3 (2.8) 29.7 (3.0) PCI 31.0 (0.9) 22.6 (1.6) 22.7 (0.8) 25.3 (1.0) 26.9 (1.2) 27.2 (1.2) 34.4 (1.6) 30.4 (1.6) 29.9 (1.5) 26.6 (1.5) 16.1 (0.9) PSI 12.7 (0.0) 14.2 ( 0.1) 13.8 (0.0) 14.5 ( 0.1) 9.6 (0.1) 9.7 ( 0.7) 9.8 (0.0) 11.5 (0.1) 14.3 (0.6) 13.6 (1.0) PSDI 7.1 ( 1.4) 4.5 ( 1.3) 4.6 ( 0.9) 6.1 ( 0.9) 5.1 ( 0.5) 3.4 ( 1.0) 3.8 ( 0.6) 4.1 ( 0.4) 3.0 ( 0.3) 2.7 ( 0.2) PRI 2.5 ( 0.9) 1.6 ( 0.8) 1.4 ( 0.4) 1.4 ( 0.4) 2.0 ( 0.6) 2.9 ( 0.5) 3.1 ( 0.9) 3.0 ( 0.5) 5.1 ( 0.5) 3.7 ( 0.4) 4.4 ( 0.1) PLI 3.8 ( 0.5) 3.0 ( 0.8) 3.5 ( 0.6) 7.0 ( 0.8) 5.8 ( 0.9) 3.9 ( 1.0) 1.3 ( 0.5) 1.9 ( 0.5) 2.9 ( 0.4) 2.1 ( 0.4) 2.9 ( 0.2) MSI 2.0 ( 1.0) 5.8 ( 0.9) 4.8 ( 0.9) 5.1 ( 0.8) 4.5 ( 0.7) 8.7 (0.2) 6.1 ( 0.5) 5.3 ( 0.5) 6.8 ( 0.1) 5.9 ( 0.3) 5.4 (0.0) Others 5.1 ( 1.7) 9.7 ( 2.4) 6.4 ( 1.4) 3.0 ( 1.1) 7.2 ( 2.0) 3.9 ( 1.9) 3.3 ( 1.1) 6.9 ( 2.2) 6.8 ( 3.0) 10.1 ( 3.5) 19.6 ( 4.4) Results obtained from Bull and Newell (2005, 46, 66)

PAGE 243

243 Table 4 2. Parties winning seats in Italian e lections, 1948 1992 Year Chamber Senate 1948 10 10 1953 9 10 1958 12 11 1963 10 12 1968 9 8 1972 9 10 1976 11 11 1979 12 10 1983 13 13 1987 14 17 1992 16 18

PAGE 244

244 Figure 4 1. Partisan a ttachment, Italy 1978 1996

PAGE 245

245 Figure 4 2. Number of d ecre es issued versus decrees c onverted, Italy 1948 1996

PAGE 246

246 Figure 4 3. C omparison of two corruption i ndexes, Italy 1980 1996

PAGE 247

247 Figure 4 4. Satisfaction with d emocracy, Italy 1973 1993

PAGE 248

248 F igure 4 5. Turnout p ercentage, Italy 1948 1994

PAGE 249

249 CHAPTER 5 AN EXPERIMENT IN MIXED MEMBER PROPORTIONAL ELECTIONS: NEW ZEALAND 1996 2005 Introduction Chapter 3 and Chapter 4 discuss the underlying changes to policy and process and the resulting decline in supp ort that precipitated electoral system change in both New Zealand and Italy. In Chapter 3, I describe how in New Zealand, majoritarian elections performed as they were designed. The two largest political parties, Labour and National, were overrepresented in every election. It was the behavior of the parties that changed, which led to variability in support. Prior to 1984, both parties practiced the key process of majoritarian elections, accountability. They did this by publishing a pre election manifes to, which leaders followed if they were elected to office. National and Labour continually supported generous social welfare and protectionist policies as elections led to disproportional results favoring one of the two main political parties. One of the parties formed a government and fostered accountability through a pre election manifesto (as well as the caucus system and interest group consultation). Governments pr oduced generous social welfare policies as well. Because of this, New Zealanders displayed high levels of support for their syst em e videnced by high levels of voter turnout, party membership, an d widespread trust and satisfaction. Starting in 1984, succes sive Labour and National governments dramatically changed both processes and policies. Both governments either ignored the promises made in their manifestos or did not publish one at all. This major change in process was coupled with an equally striking change in policy as well. In a short nine years, governments in New Zealand completely dismantled the social structures they bui ld in

PAGE 250

250 the previous seventy For example, governments dramatically reduced spending on health care, pensions, farm subsidies, a nd welfare. These changes in both policy and process represent breaks in the chain between majoritarian elections and support. When those breaks occurred support declined as well. Party membership, turnout, and satisfaction and trust all deteriorated by 1993. The preference for an alternative electoral system was also impacted by the changes in policy, process, and support. Logit models in Chapter 3 demonstrate that those who felt politicians did not follow their mandates, more liberal respondents, and those less trusting of the government favored a change to MMP. This is evidence that changes in policy, process, and support precipitated electoral system change. Majoritarian elections were blamed for the behavior of the parties. The thought was that c hanging to a more proportional electoral system would prevent one party governments and unaccountable actions by elected officials. C hapter 5 considers how the new electoral system performed in New Zealand between 1996 and 2005. I discuss the nature of t he change and the expectations and goals of the reforms. Electoral system change did not produce higher levels of support in New Zealand, but the elections did perform their function. The more proportional electoral system did allow for more parties to g ain representation in parliament. MMP elections also led to coalition governments. But, increased representation in parliament and government did not automatically lead to needs for success One of the reasons that New Zealanders did not feel represented was the continued difficulty that parties had forming coalitions. Parties were seemingly unwilling

PAGE 251

251 to state before the election the parties they would be likely to be partners with after th e elections. Even if they did suggest a potential partner, parties were happy to toss those promises aside for the sake of expediency. The first coalition formed after the election of 1996 was formed with great difficulty and drama and produced a result that New Zealanders were not ha ppy with. Subsequent coalitions were formed with less difficulty but voters never felt that parties were adequately stating who they would form a government with after the election. Another strange process that became a la rge issue for MMP New Zealand was party hopping. After 1996, there was a rash of list elected Coalition dynamics and party hopping seriously undermined the process of representation and contributed to declin ing levels of support for MMP. Another contributing factor to the decline in support was the failure of governments in New Zealand to return to the generous social welfare policies of the past. Since policy was a contributing factor i n electoral system ch ange I expect that New Zealanders would evaluate the new electoral system based on the types of policies passed. There was some movement by the Labour led coalition government in 1999 2000, but these efforts stalled after initial improvements. Policy d id not substantially change under MMP, and elites failed to provide the type of representation necessary. Taken together, there was no improvement in process or policy, so support continued to lag. C hapter 5 unfolds in the following manner. First is a di scussion of the new, MMP electoral system adopted in New Zealand. This is followed by a description of the type of electoral results produced under MMP. I also detail the difficulty New Zealand

PAGE 252

252 political parties had in forming and maintaining coalition g overnments. Next, I describe the process of party hopping which also did nothing to improve representation. A discussion of the types of policies passed by New Zealand coalition governments follows. I then run several logit models to determine the conne ction between the processes and policies described and a desire to keep MMP or return to FPP. I conclude with a discussion of these results and what they say about support in New Zealand under the new electoral system. The New Electoral Systems and Expect ations in New Zealand As I state previously, New Zealand makes an i nteresting cases for study for two ideal type as wholly majoritarian. The second reason is that th ey sought to improve the perceived flaws of the old system by replacing it with a newer, more proportional electoral system. This section describes the new electoral system in New Zealand and what the changes were expected to produce. MMP Rules The MMP re presentation system adopted after a 1993 referendum in New Zealand closely resembles the electoral system used in Germany. Under the new system citizens cast two votes on a split ballot. One of these votes is for a constituency seat, contested under modi fied singe member distracts, and is known as the electoral vote. The party vote, which appears on the left of the ballot, is a closed list of pre selected and ranked candidates on a party list (Denemark 2001). The party vote is a top up vote that determi nes the allocation of seats for parties across the entire parliament. The list vote is used to add totals to the constituency seats, with the top up designed to increase te in

PAGE 253

253 New Zealand (Denemark 2001, 95; Vowles 2005, 297). For example, in the 2002 general election Labour won the most seats in the electoral tier garnering 45 seats. The party received 41.26% of the vote on the list ballot, so it was only awarded 7 addi tional seat s, bringing its total to 52 (T able 5 1 for results). 52 seats out of a 120 seat legislature correspond to roughly 43% of the total and demonstrate how the top up process works to achieve proportionality. There are complications in New Zealand There is a threshold that may be crossed in two different ways: by gaining 5% or more in the party vote, or by winning at least one single member district. When one of those two thresholds is met, proportional allo cation applies based on the party vote (Vowles 2005, 297). For instance, in 1999, New Zealand First failed to reach the 5% threshold receiving only 4.25% of the vote on the party list. The leader of the party, Winston Peters, was able to secure victory i n his single member district, so the party was awarded an additional 4 seats based on its proportion of the party vote, bringing its total up to 5. Another complication arises if a party wins more constituency seats than its party vote entitlement. When this occurs the size of the legislature temporarily overall proportionality (Vowles 2005, 300). The number of constituency seats is set at 16 for the South Island. Popula tion increases in the North Island has had an effect on the total number of constituency seats, increasing their number fr om 65 in 1996 to 69 in 2005 ( Table 5 1). Separate Maori seats are preserved with the exact number varying depending on the number o f Maori deciding to register on the Maori role rather than their geographical role. As their number has grown, the number of Maori constituencies

PAGE 254

254 has grown from five in 1996, and six in 1999, to seven in 2002 and 2005. The existence of a Maori party in t rules. This party won 4 Maori constituencies (3.3% of the total seats), but only 2% of the party vote, so parliament was temporarily expanded to 121 seats. I hold in Chapter 2 that New Zeala nd is a good example to test the proportional model because of efforts to increase proportionality after the country changed to a system of MMP in 1996. The previous paragraphs describe how the system itself seeks to improve the level of proportionality. Additionally, the expectation was that the new electoral system would increase proportionality as well as bring some of the other features of the proportional model. These include: increased ideological representation, coalition governments, better, more widely accepted policies, and higher levels of support. The following passage details these expectations. Expectations MMP, as an alternative, was first placed on the public agenda by a report drafted by the New Zealand Royal Commission on the Electo ral System (RCES) in 1986. The New Zealand and the extent to which the existing system of FPP allowed for fair and n and McRobie 1998). Detailed in representation of minorities and any advantages gained by over representing the two largest parties like accountability or efficiency was outweighed by the treatment of minorities and minority interests (Denemark 2001). As an alternative to FPP, the Commission suggested a change in electoral system to MMP based on the presumed benefits of such a system. A survey

PAGE 255

255 of these expectations demonstrates that expectations for an MMP electoral system closely match the expectations for increased proportionality laid out in the proportional model detailed in Chapter 2. The Royal Commission high lighted several of the potential advantages of mixed systems. While these advantages are not laid out in a theoretical chain, many of the for its high levels of proportio allow minor parties a chance to win seats, improving the level of ideological representation in parliament. The Commission also believed coalition governments and consensus decision making would lead to more widely accepted policies (Jackson and McRobie 1998, 120). Summarizing these findings the Commission noted that, MMP is clearly superior. It is fairer to supporters of significant political minority and special interest groups. It is likely to provide a more effective parliament and also has advantages in terms of voter participa tion and legitimacy. (RCES 1986, 63) Most scholars agree with the conclusions of the RCES and cite increased proportionality, more representation, less psychological pressure, more effective government, and increased support as the benefi ts of MMP (Aimer and Miller 1998 ; Banducci, Donovan, and Karp 1999; Sakamoto 1999; Scheiner 2008; Vowles 2000). It is clear that the expected results of MMP coincide with the expectations laid out in the proportional model described in Chapter 2. The RCES also describes some of the potential negative consequences of MMP. Some of these I identify as potential breaks in the theoretical link between proportionality and support. Under MMP, parties not voters determine the composition

PAGE 256

256 of governments. This may lead to small er parties having an undue influence on the formation of governments. I note in Chapter 2 that having proportional governments does not guarantee representative governments. If that were to happen, the link between proportionality and support would be br oken. The RCES also highlights the cause (Jackson and McRobie 1998, 120). This is something that I highlight as well in Chapter 2 during the discussion of the potential contamination effects of mixed me mber electoral system. Duverger points out contamination effects. He postulates that the electoral tier and the domination of larger parties, could contaminate the list tier, and f avor the largest political parties in the system (Duverger 1986). If something like this occurred, a change in electoral system would not have the intended consequences. The increase in representation would be muted, with less political parties and ideol ogical interests incorporated into the system. Additional voices would not be included into decision making processes, which could have an effect on the types of policies that are forwarded. The level of support might be affected as a result. The potent ial confounding effects contain two elements: the negatives associated with proportional systems (parties not voters selecting governments for example) or the contamination effects associated with mixed member electoral systems. In addition to the expecta tions that the RCES and scholars had for MMP in New Zealand, the lessons learned from Chapter 3 suggests a few additional elements to consider. Chapter 3 argues that New Zealand changed its electoral system for two main reasons. The first was the dramat ic changes in policies pursued by successive

PAGE 257

257 Labour and National governments beginning in 1984. Citizens grew used to a social safety net and a protective set of government policies that were quickly reversed. Elites pursued these drastic changes without the consultation that New Zealanders were used to. As a result, citizen felt that their elected officials were doing one thing and saying another. If these are the circumstances under which the public in New Zealand rejected their electoral institutions for new ones, it stands to reason that the change in electoral system would help to correct those issues. The remainder C hapter 5 examines the change in electoral system in New Zealand and whether it meets the criteria laid out in the proportional model. This model suggests that an increase in proportionality will correct the problems that existed in New Zealand under FPP. Successful transition to MMP would include increased proportionality leading to more representation for ideologically diverse politic al parties. being considered in decision making processes. With the increased representation in decision making processes the goal is to produce policies that are bette r and more widely accepted. If all of these steps are in place, the end result is higher levels of support. What follows is a description of what occurred in New Zealand, beginning with the elections under the new system of MMP. Elections under a New S ystem: New Zealand, 1996 2005 The new electoral system in New Zealand was designed to correct some of the perceived flaws of the old system. The model I draw in Chapter 2 expresses how theoretically a new electoral system could make those improvement s O ne of the key components was to break the two party duopoly of National and Labour that existed for the previous 100 years. The old system of FPP was mechanically biased against third

PAGE 258

258 parties so smaller parties had a difficult time winning seats. One pa rty, dominant government was the norm, which allowed for a drastic shift in policy beginning in 1984. Voters felt that they were not consulted on these shifts and that processes that increased accountability were not followed. MMP was adopted to allow mo re political parties into the system by decreasing the pressure to vote for either Labour or National. By allowing more parties into the system the hope was that policies would become more universally acceptable and the connection between voters and their elected officials would be restored. I ar gue here and in previously that electoral systems have certain direct consequences and only perhaps some indirect effects. This section details the impact the new electoral system had on the p arty system in New Ze aland MMP electoral system definitely increased proportionality, broke the two party hold of Labour and National and allowed for third parties to gain footholds into the system. The following segment describes the more immediate impact s of the new system of MMP. These impacts are level of proportionality, the psychological impacts of the new system (along with the potential contamination effects of the mixed rules), and the ideological make up of the new legislature. Elections under M MP 1996 2005: Proportionality and Ideological Representation The first step of the proportional model requires that there is the existence of proportionality (or increased proportionality). New Zealand elections under MMP certainly succeeded in this regar d. An examination of the least squares index (LSQ), which measures disproportionality between the distributions of votes and seats,

PAGE 259

259 demonstrates the improvement in proportionality 1 Disproportionality stood at 15.4 (1984), 8.89 (1987), 17.24 (1990), and 18.19 (1993) in the four elections held under FPP before the change in electoral system. There was dramatic improvement in LSQ under MMP with 3.43 (1996), 2.97 (1999), 2.37 (2002), and 1.13 (2005) (Gallagher 1991; Gallagher and Mitchell 2005). These figu res certainly suggest an increase in proportionality under MMP. Political parties took advantage of the more proportional rules and were able to gain representation at a much higher rate than they had in the past. The effective number of parliamentary par ties increased from 2.16 in 1993 under to 3.76 (1996), 3.45 (1999), and 3.76 (2002) under MMP elections (Gallagher and Mitchell 2005) 2 Parties that actually won seats rose from 6 in 1996, to 7 in 1999 and 2002, to 8 in 2005. The results of the elections are displayed in Table 5 requirement for increased proportionality and increased opportunities for parties to gain seats seems to have been met. This rise in representation also led to more ideological diverse political parti es entering the system. Among the parties that won seats were the two dominant political parties under FPP, Labour and National, that represented the center left and center right ideologically. These two parties continued to be the two largest political parties in the system. Two parties that had run previously were also represented in parliament. New Zealand First, a centrist party, actually won one electoral district under FPP in 1993. 1 LSQ is measured by taking the square root of half the sum of the squares of the difference between percent of the vote and percent of the seats for each of the political parties. Low scores indicate the highest level of proportionality (Gallagher and Mitchell 2005). 2 To see how the effective number of parliamentary parties is calculated see Laakso and Taageparera (1979).

PAGE 260

260 That seat was acquired by Winston Peters, a former National MP, who left the party after a disagreement in 1991 (Aimer and Miller 2002). The alliance was a left wing political party that consolidated several of the third parties, including the Social Credit and Green Parties that had unsuccessful run for office under FPP. Led by former Labour MP Jim Anderton the Alliance believed in free health care, education, and other generous social welfare policies and had garnered 18% of the vote in 1993 under FPP, but only two seats (Aimer and Miller 2002). The Green Party, co ncerned with environmental issues, left the Alliance and ran on its own in 1999, 2002, and 2005. The Alliance held together in 1999, but split into the Progressives and Maori party in 2002 and 2005. All other parties were new political entities on the sc ene having never run for office previous. They were formed by interest groups previously left out of the political system or disaffected members of one of the two largest political parties. These new political parties were diverse ideologically as well. Joining New Zealand First in the center of the ideological spectrum was United New Zealand. This itself in the same ideological space as the more successful New Zealand First Party and only won one seat in both the 1996 and 1999 elections. Another party, the Christian Coalition was a religious party based on Christian teachings and supported conservative social policies. This party was not able to cross the 5% threshold or win an electoral seat, so it was never represented in parliament (Aimer and Miller 2002). Due to their lack of success, the Christian Coalition and United New Zealand formed United Future, a center right coalition committed to conservative social prin ciples and moderate fiscal policies. This coalition had more success winning 8 and 3 seats in 2002

PAGE 261

261 and 2005. The right of the ideological spectrum also included a fiscally conservative party as well. ACT New Zealand was formed outside of parliament in 1 994 and positioned itself at the opposite of the ideological spectrum as the Alliance. ACT supported the policies advocated by former Labour Finance Minister Roger Douglass including flatter taxes, the selling of state assets, smaller government, and mark et liberalism (Aimer and Miller 2002). ACT was able to win 8, 9, 9, and 2 seats in the elections held under MMP from 1996 to 2005. Electoral results for all of these political parties are provided in Table 5 1. After a cursory glance at the elections under MMP, it is apparent that some increased representation has occurred at least in terms of the first few steps of the proportional model. The goal of adopting a new electoral system was to allow more parties, with divergent ideologies into the system. The first four elections did just that. Between 6 and 8 parties won seats after 1996. These parties ranged from liberal, environmental parties (Greens) to conservative, neo liberal parties (ACT) with parties represented on the center left ( Labour), center (NZ First), and center right (National) as well It is clear that proportionality was increased and that the party system included more ideologically diverse political parties. What is left to determine is whether mixed electoral rules ha d the type of psychological effects that are expected under the proportional model. The proportional model suggests that as proportionality is increased voters will feel freer to vote sincerely rather than strategically (Duverger 1954; Cox 1997). MMP intr oduced an increase in proportionality, but also potential contamination effects due to the mixed nature of the rules. The potential contamination effect that I identify as most

PAGE 262

262 likely is the possibility of the electoral tier and its tendency to promote a two party system to contaminate the electoral tier, reducing the number of political parties overall (Duverger 1986). For scholars who do not see evidence of contamination effects, mixed rules should promote both strategic and sincere voting (Moser 1999, 2001). In New Zealand, this type of behavior would be evidenced by split ticket voting between the electoral and list tiers. This would mean a voter would vote strategically, perhaps for a larger political party like Labour or National or a competitive s maller party, in the electoral tier. This same voter would recognize the overall importance of the party tier and the lower threshold for representation and cast a sincere vote for a party closer to their ideological interests. The evidence from New Zeal and suggests that voters are neither confused by the electoral rules nor contaminated by the electoral tier. New Zealanders split their tickets at a higher rate than other countries 3 In the 1996 election 37% of voters split their tickets (Johnson and Pa ttie 2002, 586 587). This number was 35% in 1999 an d 39% in 2002 (Aimer and Vowles 200 4 23). Work done after the 1996 election, based on data collected by the New Zealand Election Study, report that 89% of the voters who expressed preferences for a part y, measured by a likes and dislikes scale, cast a party vote consistent with that party (Karp et al. 2002: 5). In contrast, the relationship between party preference and votes in the electoral tier are much stronger for Labour and National than they are f or smaller parties (Karp et al. 2002). These results tended to hold over the 1999, 2002, and 2005 election as well (Karp 2006). Due 3 Voters i n Scotland and Wales, in their first elections under MMP rules, split their tickets at a rate of 20% and 17% respectively. In Germany, the number of ticket splitters hovered around 6% in the sixties and seventies, and rose to an all time high of 22% in th e elections of 2002 (Gschwend and van der Kolk 2006, 164).

PAGE 263

263 split ticket voting that ha New Zealand still had to cast at least one strategic vote, but the fact that they could make at least one sincere vote on the party ballot is certainly an improvement over the types of choices that had to be made under FPP. MMP in New Zealand satisfies the first few steps of the proportional model. Proportionality was increased and more ideologically diverse political parties had a chance to win seats and were successful in winning seats. Voters also were able to navigate the mixed rules, and could cast a sincere party vote, while voting strategically with their electoral vote. This is particularly important because it is the party vote percentage that determines the overall seat allocation. It seems that the more immediate effects of increased proportionality (increased number of parties, eased psychological effects, and increased ideological representation) have been met. What is left to determine is if the more distal effects of the electora l system followed suit as the proportional models suggests they will. After parties are elected, the proportional model suggests that MMP would produce co alition governments. This occurred after every election under MMP. So, in terms of the mor e direct effects of the more proportional electoral system, the model appears to be accurate. But, as I suggest earlier, this would not be enough to raise the level of support in the country. D rastic reformulations in policy led to the electoral system c hange, so the success o r failure of the new system should be judged based on how those policies were reversed. New Zealand citizens also expected that there would be an increased connection between themselves and their elected officials. MMP was supposed to produce this improved representation by incorporating more

PAGE 264

264 making process. The following section demonstrates how the party system in New Zealand failed to produce that outcome. While there was increased repr esentation ideologically, parties and elites continued to behave as if their actions were not connected to their constituents. This distance is evident in the processes of coalition building and the strange practice of party switching. These two actions by parties and elites un dermined representation, and led to declining levels of support for the new electoral system. New Processes in New Zealand: Coalition Government and Party Hopping The popularity and success of the new MMP system in New Zealand was i mpacted negatively by several new processes that became prominent after the electoral system change. The 1996 election did not produce a clear winner, with National and Labour lacking the necessary seats to form a government. NZ First, the third largest party, became the king maker. After a lengthy, costly, and secretive negotiation between all three parties, the public was left with a National/NZ First coalition. This was a partnership that the public neither expected nor wanted (Aimer and Miller 2002) The behavior of the coalition also did nothing to engender a positive feeling about the new system among the population. Coalition government got off to a shaky star t, a not altogether unexpected consequence of increased proportionality The RCES note s that this was a possibility and it is one of the potential breaks in the proportional model that I describe in Chapter 2. This was not the only process that undermined representation in New Zealand, party switching became a tense political issue followi ng the first election held under MMP. Immediately following the referendum in 1993, pol iticians saw that there would be increased opportunity to form their own parties and win seats and a s a result a rash of

PAGE 265

265 party switching occurred. This was perhaps ex pected due to the nature of the changes involved. Single member districts were eliminated and safe spots on the party lists were limited. Politicians formed new parties or joined others if they determined that the switch improved their electoral prospect s (Vowles 1998) Party switching continued after the first election under MMP in 1996 however. Multiple politicians, elected through a part y list, became independents or formed alternative parties. These moves called the legitimacy of MMP into question and had a profound impact on how citizens judged the practice of their new electoral institutions (Geddis 2002, 2006) Both coalition government and party switching were two new processes that negatively impacted the some connection with the processes that came before the change in electoral system. The concern previously was that there was disconnect between what the public wanted and what politicians and elites did. Elected officials, following the reforms, contin ued to behave in a manner of their choosing rather than the way the public thought the y ought to behave. In the following passage I describe the impact of coalition formation and party MMP in much greater detail To do so, I use a historical narrative as well as a series of logit models that use data collected from the New Zealand Election Study. Coalition Formation/Government 1996 2005 The 1996 election produced a new, albeit not completely unexpected result. No party received a majority of seats in the parliament. Under the guidelines of MMP either National (44 seats) or Labour (37 seats) could have formed a minority government. All efforts in that direction were halted when Winston Peters, the leader of the NZ First Party, declared that his party wanted to be a part of any government formed. National and Labour simply did not have enough votes within the existing parliament to form a

PAGE 266

266 minority government and a grand coalition had no support within either party or among their respective electorates (Aimer and Miller 2002). The Alliance, a left wing party who would support Labour on votes of no confidence, had 13 seats. This, hypothetically, y. National was in a similar situation, they won the most seats during the election (44), but the other right leaning party ACT NZ only garnered 8 seats during the election. This left the right leaning parties at 52 votes, still 8 short on votes of no co nfidence. NZ First was therefore necessary to form any government whether minority or majority 4 The party declared immediately following the election their intentions to negotiate with both Labour and National for spots in the government. From the outs ide looking in, this might have seemed like a normal occurrence since NZ First presented itself to the electorate as a centrist party between Labour and National, and as such it may seem that they would be a natural fit for any center left or center right government. This assumption did not meet the expectations of the New Zealand people, however, and the ensuing negotiations and eventual coalition government that was formed had serious consequences for the first coalition government under the new electora l system and During the election campaign, Winston Peters generally dodged the question 4 National could form a majority government with NZ First or a minority government with their support on no confidence votes. Labour needed both the Alliance and NZ First, whether as part of a maj ority government or as allies on no confidence votes. While it is possible that a minority government could have formed despite these seat totals, governing would have been a different matter. Passing legislation for a Labour led or National led minority government would have been extremely difficult if not impossible. Also, when NZ First declared that it wished to form a government with either National or Labour, both nces of taking over the government.

PAGE 267

267 As the election season progressed, comm entators and news organizations grew 121) Despite the efforts by Peters to simultaneously advance and dampen the prospects of a coalition with Labour there were plenty of signals and ind ications that NZ First would not choose National as its eventual partner. While Peters did not come out and explicitly state his favored coalition partner prior to the election he did continually signal the direction in which he was leaning. He was quote will not win the election, [and] that they will not form part of any post ele ction 120) In a meeting with voters three months prior to the election, one political analyst reporte d that Peters categorically stated there would be no coalition with National, an assertion Peters later denied (Jesson 1997). The night before the election, in a last appeal to voters, Peters declared that the only way to change governments (National was a single party minority government at this point) was to vote for NZ First (Miller 1998 121). While all of these comments seem to suggest that NZ First would not choose National as a coalition partner, the comments attributed to Peters only implicitly su ggest that conclusion. Although an explicit, blanket statement from the NZ First leader was absent, voters had other reasons to believe that NZ First would choose Labour over National. Since NZ First was considered to be a centrist party there was cause to believe that the more natural fit was with Labour rather than National. Prior to the election, the Secretary of the Ca binet for NZ First, identified three principles from which a coalition would be formed: policy compatibility, procedural concerns (ho w well the potential coalition partners would be able to govern together), and the make up of the cabinet

PAGE 268

268 (who gets what job). Based on the first two criteria, it would seem that NZ First was a closer fit with Labour than with National. On issues of poli cy, the platform of NZ First seemed almost antagonistic to National and much more congruent with Labour (Boston market ideas of the past 11 years (started by a Labour government in 1 984 5 ). National wanted to continue privatizing state industry and increase foreign investment into New Zealand owned industries (sell state assets to foreign companies or states), continue to lower taxes, limit spending on social programs, and liberali ze immigration. Labour and NZ F irst were generally opposed to all of the policy positions. The policy incongruence ran (Miller 1998 124 125). Policy differences were not th e only reasons that voters did not expect a NZ First/National coalition. The second criterion, that the coalition partners were able to work together in a cooperative manner, also suggested that National and NZ First were not a good fit. The acrimonious n ature between NZ First leaders and National party leaders was well known in New Zealand. Winston Peters formed NZ First after a well publicized split from National in the early 1990 s. This political divorce was mainly over policy and combative personal r elationships between Peters and senior National politicians. He spent the better part of the years between 1993 and 1996 firing broadside attacks against his former party, at times accusing National and its prominent business supporters of corruption and ocratic principles (Miller 1998, 121). These rancorous feelings were prominent at the NZ First 1996 party 5 For details, see Chapter 3.

PAGE 269

269 conference where Peters and his deputies used the public forum to criticize their counterparts in the National Party. Pr ime Minister Jim Bolger, Finance Minister Bill Birch, and the Minister in charge of Welfare, Jenny Shipley (who later became National Party leader and Prime Minister) received the bulk of the attacks. Tau Henere, NZ would never serve in a government with Bolger, Birch, and Shipley (Miller 1998 believe that the kind of politicians depicted by Bolger, Birch, and Shipley, is not to be promoted into cabinet. As a consequence, we will not have truck w (Miller 1998, 124) Considering this history and statements, the public at large could not be reproached for believing that NZ First and National did not meet the second criteria for forming a coaliti on. It seemed that the leaders of National and would not be able to work well together. What remained, the third criteria, that the coalition would be formed based upon how the government was to be divided (who got what job). This third criterion is per haps the least democratic of the three. Forming a coalition based on policy distance, where center right parties join with further right parties, is more representative of how the public saw the election and probably more democratic. The same could be sa id for forming a coalition based on working relationships. The less acrimonious the relationship, the better the policy making process would be and perhaps the better the policies would be as a result. These first two principles fit with the models I dra w in Chapter 2, and were the reasons that New Zealanders adopted MMP in the first place. They wanted a more representative electoral system that would promote more consensual processes and better policies. When NZ First and National struck a deal to form a government after the 1996 election,

PAGE 270

270 the public was confused. These two parties apparently had little relationship in terms of policy and did not appear to have a good working relationship. The only explanation m National, eschewing the more democratic principles that MMP was based. The negotiations between NZ First, National, and Labour did nothing to dispel these feelings. The course of negotiations was secretive, long, and generally disliked by the public. The actual formation of the coalition did little to endear the public to its new electoral system. There was an expectation among New Zealanders that a coalition, in all likelihood, would have to be formed following the first election under MMP in 1996 6 It was thought that this process would happen quickly and openly. These expectations were shattered when Winston Peters, rather than enter into negotiations with either party, decided to go sailing for five days, and did not meet with either party until 10 days after the election took place. When negotiations moved forward, there was rampant decision would be reached in days. The final negotiations did not end until two m onths National and NZ First and Labour took place simultaneously. Peters mentioned that each party had a fifty (Miller 1998, 125). These comments did nothing to inspire confidence in the process, and increasing numbers of New Zealanders were skeptical that NZ First had anything but the interests 6 This is evidenced by the numerous polls that took place prior to the election demonstrating that voters had an opinion about their favored coalition partners (See New Zealand Election St udy 1996).

PAGE 271

271 of the party in mind 7 It seemed to the public that whatever party sold out its interests to the highest degree would be able to join in the coalition with New Zealand First. confidence in the new system of MMP. Peters demanded that the tal ks take place in secret, and reports leaked that negotiation rooms had to be swept for bugs before meetings could start. NZ First also spent a reported 1.3 million dollars on the s an advisor. Peters also missed a meeting with the Prime Minister after getting into a late night argument with supporters inside a New Zealand restaurant the night before. This behavior led to both National and Labour voters reporting little confidence that NZ First was acting in the best intere sts of the country (Miller 1998, 125). National and NZ First reach ed a coalition agreement after three months of negotiation. The 74 page document awarded 5 of the 20 cabinet positions with a promise to receive 3 m ore by October 1998 (Aimer and Miller 2002, 7). Winston Peters was awarded the post of Treasurer, the powerful head of the Finance Ministry. While there is no concrete, explicit proof that NZ First took the best deal (rather than making the coalition decision based on policy similarity or the ability to work together), the number of cabinet positions they did receive is good circumstantial evidence that they did just that. The terms of the agreement were met by a less than enthusiastic public who neit her wanted nor expected this agreement and partnership. 7 negotiations, it appeared that NZ First was simply holding out for the best deal. In terms of pure rational calculation, this might make sense, but ran counter to the ideals and expectations of MMP in general.

PAGE 272

272 These feelings were especially true among those who supported National and NZ First. Before the election, only 13% of National voters preferred NZ First as a coalition partner (Remarkably, a simila r number preferred Labour) (Miller 1998: 129). NZ First voters also ranked National last among preferred partners. The first choice was a coalition involving Labour, who was favored over 2 to 1 against a coalition with National. NZ First voters desired a coalition with Labour and the Alliance (20%) or only Labour (24%) over a coalition with National (13%) ( Miller 1998, 129). Perhaps the reason that NZ First was not favored were the feelings about National voters toward the party in general and Winston P eters specifically. Among National voters, 47% expressed either strong dislike or dislike for NZ First, against only 12% who liked or strong like the party (against 40% neutral). National voters expressed even more negative feelings toward Peters when 53 .2% stated strong dislike/dislike and only 14.4% like/strong like (31.7% neutral) (Miller 1998: 130). NZ First voters found National and their party leader Jim Bolger equally objectionable. 46% expressed dislike/strong dislike of the National Party and 5 8% felt the sam e about Jim Bolger (Miller 1998, 131). Considering these feelings, the new coalition appeared to be off to a shaky start, and the behavior and performance of the first government did little to improve the situation. From the outset of the new government, NZ First proved itself to be incapable of governing responsibly. Since the party itself was relatively new it was continually plagued by inexperience, poor party discipline, and disorganization. The only cabinet member with any experienc e inside the government was Winston Peters. In fact, 5 of the 9 New Zealand First Ministers were new to parliament al together (Aimer and Miller

PAGE 273

273 2002, 7). Almost immediately, NZ First members were embroiled in a series of high profile embarrassments that attracted significant media attention. Morgan, was implicated by media reports that he spent large amounts of public money from the Aoteroa Television Network on designer clothing after had resigned as director of the company. Rather than apologize and move forward, Morgan was caught trying to sell his story to television networks while in office. Another NZ First Minister, Robyn McDonald, was criticized for taking a lavish and unnecessary trip France using public finances. Later in 1998, Tau Henare was replaced as Deputy Leader of NZ First after he took a similar trip to France, this time to collect shrunken Maori heads ( Boston, Church, and Pearse 2004, 588). Perhaps the most outrageous of behavior was performed by the party leader himself. Winston Peters was found guilty of assaulting a National MP, John Banks, when a debate on the parliamentary floor spilled out into the lobby. Peters was also embarrassed by the findings of which found that his claims that several prominent businessmen and National supporter s were frequent tax evaders, were false. Peters refused to apologize for either event, which further antagonized his coalition partners (Bost on, Church, and Pearse 2004). These are singular events, but typified the difficulty the first coalition government faced. The public responded to the performance of the coalition by blaming NZ First, whose favorability among the voters fell. In July 199 6, the party received 29% approval from the public, its highest ever. One year later in July 1997, after the scandals I detail above, party approval stood at 4%. By October, party ap proval was only 2% (Miller

PAGE 274

274 1998, 133). These continued issues, and decl ining support caused embarrassment for partners. The public embarrassments did nothing to improve relations among the parties in government. The coalition eventually cr umbled in late 1998 over issues related to the selling of several state owned assets. The first coalition government under MMP was an unmitigated disaster. Parties, to a certain extent, learned their lesson of the National/NZ First government and attemp ted to come to coalition agreements prior to the elections. But, the uncertain nature of election results made this type of government planning difficult. Labour was also rel uctant to adamantly say who it would prefer to form a government with before the elections. This was especially during the election of 2002 after the debacle concerning their coalition with the Alliance which came to a disastrous end in 2001. Also, the governments themselves were made up of parties that often times had disparate v iews on issues. One of the goals of MMP was to create more ideologically representative coalition governments, but these governments were more difficult to navigate and hold together. What was introduced into the system was a degree of uncertainty about who would be a part of the government and whether that government could hold itself together. Before the election of 1999, there was some movement toward making post election coalition plans. Labour and National benefited from a pre arranged coalition agr eement in several respects. After the NZ First/National post election bargaining, the public was strongly in favor of prearranged pre election coalition agreements. The Alliance also benefited by reducing the risk that they would be outmaneuvered by eith er

PAGE 275

275 the Green Party or NZ First after the election. Because of this the Alliance was flexible on some of their dearly held policy beliefs. They compromised with Labour on issues related to taxation and free market policies. Despite this Labour leader Hel en Clark criticized Alliance positions like the abolition of tertiary fees, free health care, and an 117) Due to genuine policy disagreements and a degree of uncertainty about the electoral pr ospects of NZ First and the Greens, Clark continually refused to rule out the possibility of dealing with Winston Peters if NZ First held the balance of power (Miller 2002). When the initial election results came in it appeared that Labour and Alliance ha d 63 seats a majority in parliament. When special votes were counted, the Green Party crossed the 5% threshold and won 7 seats, denying seats from Labour/Alliance and reduced their share to 59. The Greens came to an agreement on confidence and supply sho rtly after the election and the 5 th Labour government took office. The agreement reached between Labour and the Alliance was only 1.5 pages in length, a stark contrast from the 74 page agreement made between National and NZ First in 1996. It avoided any d etailed policy agreements and focused instead on 6 ). The vague nature of the agreement led to dissatisfaction among Alliance supporters. Alliance supporters had diffi culty achieving any sort of policy gains and the coalition eventually collapsed when Labour pursued a free trade agreement with Singapore and when the government decided to commit troops to Afghanistan. The

PAGE 276

276 latter issue actually fractured the Alliance and was one of the reasons Labour called early elections in 2002 8 The other reason for the early elections had to do with the Green Party which, in the run up to the 2002 election, began to harden its policy demands. Although the Greens had maintained a goo d working relationship with the government, they staged a Greens withdrew support from Labour and threatened that if Labour did not reverse its position on the issue, t hey would use their parliamentary strength to block a Labour led government. Labour used this to criticize coalition government and urged voters to return them to office with a parliamentary majority. The election left Labour and the Progressives (the re maining members of the Alliance, who held on to 2 seats) 7 seats short of the necessary 61 to pass any parts of their agenda. They decided to enter into ty, United Future (Bowler, Karp, and Donovan 2010 4 ). This was a surprise to the voters of United Future, who saw their support drop from a high of 6.7% after the 2002 election to below 2% before the 2005 election. The Labour/Progressive government also came to a surprising agreement with the Green party on matters of support. This government managed to hold together through their entire 3 year term, but once again the public was left with a coalition agreement that they did not expect. This pattern was repeated during the election o f 2005. During the 2005 election Helen Clark the leader of the Labour party, campaigned on a Labour/Progressive/Green coalition ticket, but abandoned that ticket once the 8 I describe how the Alliance fell apart in the following section on party switching.

PAGE 277

277 s eat added another, and the Greens held 6 seats, leaving the coalition 4 seats short of a governing majority. Once again, Winston Peters would be in the center of coalition negotiations. He declared before the election that he would seek an agreement with the (Vowles 2006, 1217) Following the election NZ First and United Future both declared they would be prepared to enter into an agreement with Labour, but only if the Green Party was not a part of the government. Labour/Progressive/NZ First/United Future commanded 61 votes, and the Greens were left out. The arrangements made were Fir st and United Future received significant positions in the government outside of cabinet. Peter Dunne, leader of United Future, became the Minister of Revenue. Winton Peters, in a more controversial move, became Minister of Foreign Affairs. This was an unusually high position for someone who was not a mem ber of the cabinet (Vowles 2006, 1218). Peters had again promised the electorate one thing, only to reverse himself and choose political gain. The formation of the first four governments under MMP int roduced a degree of uncertainty. 1996 gave the public a government it neither wanted nor expected and the government collapsed before the end of its term. 1999 offered some improvement, but parties were still reluctant to enter into pre election agreemen ts, and the uncertainty of election results made this process difficult. The government coalition was also n ot able to make it through the three year term and elections were called early. In 2002 and 2005, unusual governments were formed after the electi ons. Coalition governments

PAGE 278

278 were one of the products expected from MMP, but the actual formation was continually problematic and the change from the previous system was stark. The reality under FPP was very different. The leader of one or other of the cen trist parties (Labour or National) which won a majority of the seats on election day claimed the mandate to govern. Usually the incoming or returned Prime Minister was known within a short while of the polls f virtual direct election of government. Voters voted for their preferred party knowing who would be Prime Minister if their party won the election. However, all that changed under MMP, which vest the prerogative of government formation in the political parties that the people elect to parliament. (Joseph 2008, 8) New Zealanders cast their vote with conflicting, mixed, or no signals from the leaders of the political parties about who would be in the government after the election. This is one of the proce sses that I believe could have a negative impact on support for the new system. However, c oalition government was not the only new process under to switch parties, a pract ice that continued after MMP elections began. In t he following section I describe this new process. Party Switching 1993 2002 MMP was adopted in 1993 by a referendum, but the first election under the new rules did not take place until 1996. This left a three year window for new 3 rd parties to form and organize. A s I describe earlier, some third parties already existed like New Zealand First and the Alliance. Other parties, like ACT New Zealand, were formed outside of parliament by political activists not holding any elected o ffice. Another group left their former parties to form new parties or join existing parties. The reason behind these defections is varied, but either had to do with ideological difference with s desire for self pr eservation

PAGE 279

279 The first of these rationales is easy to understand. For a long period of time Labour and National dominated the political scene as center left and center right parties. Essentially, in order to get elected, politicians joined one of these two political parties even though they might not have been a pur e, ideological fit. This is the nature of a catch all party, but now that there was increased opportunity to form or join a party Some did just that including National MP Grame Lee. Lee was a strong Christian conservative wh o left N ational to form the right wing Christian Democratic Party. One j unior National Cabinet Minister joined the newly formed centrist United New Zealand Par ty after concerns with some of Nationals more neo liberal economic policies (Aimer and Miller 1998, 4). preservation. These are members whose districts were to be eliminated under the new electoral ru les and who could not guarantee a favorable list position by either National or Labour. Rather than retire or risk losing their seat by taking a disadvantageous position on the proportional list, some decided to form their own pa rties. National MP Pauline Gardiner would have faced a toug h race in her new district decided instead to join United New Zealand. Vowles (1998) dubbed the remaining three members of the Between 1993 and 1996, thirteen political parties or join ed existing ones so party hopping was a new process introduced to New Zealanders prior to the fir st election under MMP in 1996. Some degree of

PAGE 280

280 FPP and the first under MMP, but what neither voters nor experts expected was the continuation of party hopping among elites a fter 1996. These switches had consequences for how voters felt about their new system as well. After the initial rush of party hopping, the number of switches died down until 1997. After this date, several high profile and controversial party defections made the practice a national issue. The first and most controversial of these changes in party occurred in July of 1997 when an Alliance list MP, Alamein Kopu, resigned from he r party and declared herself an independent. She justified this move on the ba sis that the 9) There seemed to be little evidence of her complaints or outward disagreements with the party hierarchy. In fact, less than a year before, Kopu was pla ced in a favorable list spot which was the reason she won her seat. While it is impossible to know her true motivations, obse rvers were baffled by her move switch] made little or no sense to anyone but herself, her famil y, and those members of 2002, 687) members switching parties. Another round of party hopping occurred over the tense relationship betwe en National and NZ First (more evidence that the two parties had difficulty operating together). Prior to, and perhaps foreshadowing, the collapse of the coalition, Neal Kirton a NZ First list MP, left the party to become an independent. His stated reaso n was that National was not living up to the coalition agreement. When the independents where they split 4 ways. Ann Batten, Jack Elder, Tuku Morgan, and Rana

PAGE 281

281 Witai cre ated the Mauri Pacific Party. Tauriki Delemare joined the Te Tawharau party, while Peter McCardle remained an independent. Perhaps one of the reasons for the who switche d parties and named themselves the head of a new party (even an I ndependent party) could claim a higher salary in parliament, as well as funding f or their new party (Geddis 2006, received a singl e vote from the electorate and could theoretically have only one member, as in the case of Kopu (Geddis 2006). As puzzling as some of the moves might have been and the anger they stoked, they raised serious questions about the legitimacy of the list port ion of MMP. Since was based solely from their membership and support from a political party. When the with the voters. Kopu, for example, ran in an e lectoral district in 1996 (along with being included in the Alliance list), where she garnered a mere 5.7% of the vote (Aimer and Miller 2002 9). Kopu owed her entire electoral legitimacy to her political party, without it she would not have been an MP a t all. When she left the party, her connection to the voters was severed. Jim Anderton, the leader of the Progressive Alliance became one of the most outspoken critic s of party switching saying the act y standard of ethics

PAGE 282

282 MP who switched parties to resign, which many in the Labour Party vigorously endorsed. Much of the public agreed. To many it was simply intolerable that those who switched parties, appar ently out of personal gain, could remain in parliament (Geddis 2006). As a result, two political parties, Labour and the Alliance who formed a minority government in 1999, pledged to address the situation through legislation. Although the principle behin d the bill had wide support among the public, the parties did not pass a bill until 2001 in the form of the Electoral Integrity Amendment Act (Geddis 2006). This party switching law required that the Speaker of the House declare a seat vacant if the MP wh which he or she was elected (Geddis 2006, 38). This law would appear to have solved the problem of party switching, but a series of events involving (ironically) the Allia nce made a farce of the new law. The Alliance Party joined the Labour party in a formal governing coalition following the election of 1999. But, like the NZ First Party before it, the party had little practical experience in actual governing. Many i n the party felt boxed in by the trappings of government. These feelings came to a head when the government decided to send troops to Afghanistan in support of the United States. This issue divided the party. Eventually, the Alliance split into two fact divide between the two factions was too great leading to a complete rupture of the party. In actuality, the factions began operating as separate parties, eventually meeting in separate caucuses. The factions never formally declared themselves as separate parties, thereby getting around the Electoral Integrity Act. The situation became even

PAGE 283

283 parliamentary governing council original pa rliamentary leader Jim Anderton. S o the recognized leader of the Alliance lli 2006, 42). The members of the factions never officially declared themselves separate parties, which made a joke of the party hopping law. The time spent on the nically fell under the party hopping law was one of the reasons that Labour called early elections in 2002 [Parliamen [ing] its unfortunate image as a n institution 43) view of the appropriate MMP. This view suggests tha a team member w hose conduct sticks to party or adjust their view if they do not. If an individual is unable to do this and decides to formally leave the party to which they owe their electoral success, then their right to continue holding a seat in parliament is called into question (Geddis 2006). Party switching fell into direct contrast with this view and was another process that impacted the publi (Geddis 2002, 2006) Before continuing on to a discussion of the policies that the coalition governments pursued, I would like to explain what the preceding historical narrative suggests about the proportional model I draw in Chapter 3. T o reiterate the model suggests that

PAGE 284

284 increased proportionality leads to more ideologically diverse political parties. When decision making (th rough coalitions or another mechanism) then the process of representation is met. In New Zealand, more ideologically diverse political parties were incorporated into the system. In the past, only two centrist parties, Labour and National, had success gai ning seats. Under MMP, the party system included neo liberal, right wing economic parties (ACT), environmental parties (Greens), and a religious based, Christian party (United Future), among others. The inclusion of these ideological di fferences means th at the first two steps of the model are successful. The next step, making proves to be more difficult. The proportional model fails to predict the difficult coalition formation process or party hopping. N ew Zealand parties, from 1996 to 2005, never seemed to manage forming pre election coalition agreements or post election bargaining in a satisfactory manner. The first National/NZ First coalition was an unmitigated disaster. Then the 1999 and 2002 coalit ion process es were somewhat improved, but parties were still hesitant to clearly state before the election what other parties they were willing or able to work with. The 1999 coalition between Labour and the Alliance fell apart before the 2002 election. The 2005 coalition once again demonstrated that individual politicians were willing to place their own interests first. Labour discarded its pre election promises and brokered a deal with Winston Peters (who was once again, like in 1996, in the middle of controversial coalition bargaining) to form a government without the Green Party. Parties engaging in post election bargaining and log rolling without considering the will of the public was something that I discuss as a potential break in the

PAGE 285

285 proportional demonstrates those concerns were warranted. Party switching presents an interesting break in representative processes. There is no theoretical reason that one electoral system or another might lead to party hopping, so it was not a practice I spent any time with while developing theories. The practice of party hopping does show is that individuals, regardless of institutional context, have a profound impact on representative processes. When an individual, list MP decides to leave their party, especially when the party is responsible for their seat in the first place, a key component of representation is beca use of a party vote, but does not cause the problem. Individuals make this choice because of their own volition. What is clear is that the processes of representation because of difficulty in coalition formation and the con tinued practice of party switch ing were not able to meet the conditions laid out in the proportional model. At this point, due to difficulty with coalitions and party hopping, the proportional model shows signs of breakage. In a similar vein, successive governments in New Zealand wer e slow to rollback the neo liberal policies that caused the change of electoral system in the first place. The first National/NZ First government was not able to continue the aggressive program of privatization, tax cuts, and spending cuts of the past (pe rhaps because of the generally free spending NZ First Members). However, 1984, FPP New Zealand. Successive Labour led governments that followed had more success returning some of the social and economic benefits of the past. I believe that over time, the changes in policy led to more favorable attitudes toward MMP. The following

PAGE 286

286 section discusses the policy direction of the governments formed under MMP from 1996 2005. P olicy in New Zealand under MMP: 1996 2005 As I describe in Chapter 3, successive governments (Labour 1984 1990) and National (1990 1996) pursued policies that seriously cut back on social spending and the protectionist policies that characterized the previ ous 100 years. It is my belief that these changes in policy were the driving force behind the loss in legitimacy for FPP and precipitated the change to the new electoral system of MMP. I argue that MMP would only be deemed a success if citizens of New Ze aland felt that the new electoral system reversed some of these policies. Policy changes were not forthcoming after the first election under MMP, when a surprising coalition of National and NZ First formed a government. This government did little to reve rse the policies of the past. The Labour led governments actually returned some of the services that had been taken away from New Zealanders. But, after a rush of activity from 1999 2002, the pace of policy change slowed. Labour was reluctant to return New Zealander to its free spending roots of the pre 1984 period. This section describes this change in policy direction and in later sections is used to explain the changes in support for MMP Reversing Past Policies and Reestablishing Some Social Welfar e: 1996 2005 National nearly lost the election of 1996 and was forced to go into a coalition with New Zealand First. In order to remain in government, National had to make policy concessions to the more free spending NZ First Party. This put a halt, for the most part, to the aggressive neo liberal policies of the past. Even with the roadblocks put up by New Zealand First, the coalition still passed several policies that further reduced social benefits to New Zealanders.

PAGE 287

287 One of the major policies passed b y the National/NZ First government was a scheme known as the community wage. Passed in 1998, this act merged the Department of Social Welfare and the New Zealand Employment Service together. The goal behind this merger, as stated by the government, was t for the unemployed where they would receive job training and placement as well as unemployment assistance (Peters 1998). Essentially the move was an attempt to ers, whose work would then be subsidized by the government (McTaggert 2005). National and NZ First wanted to streamline the process of work training to work in order to reduce the overall payment of unemployment benefits. The National/NZ First government also pushed two different plans to change the payments of superannuation to retired New Zealanders. The government tri ed to reduce pension payments by lowering the floor of wages to which it was tied (Mclelland and St. John 2006). In 1998, the governmen t backed a referendum on a NZ First plan that would have created private accounts for superannuation in an effort to move retirement funds away from complete public financing. The referendum failed 92% to 8%, in what was an early sign of the satisfaction with the conservative policies that dominated the pre vious decade (Vowles et al. 2002 ). When the Labour/Alliance coalition (aided by the Green party) took over the government in 1999, they quickly moved to reverse some of the more draconian, conservative p olicies of the past, passing a rash of legisl ation in their first two year s in government. In the first three weeks of their term, the government announced an increase in the minimum wage, eliminated the interest on student loans for full time and

PAGE 288

288 low inc ome students, and introduced new legislation that would increase the tax rate on those with the highest incomes (Sinclair 2003). In the first year alone, the center left restored the wage related floor for state pensions, reversed the privatization of acc ident insurance, as well as increasing the superannuation payments for married couples. In early 2000, the welfare to work scheme was eliminated and health care reforms designed to increase access to help were also introduced. In the area of housing, the Housing Restructuring Amendment Bill was passed. It set rents at 25% of household income, making them much more affordable than the market based approaches of the past (McClelland and St. John 2006 ). The government also set out to improve education by i ncreasing funding for early childhood education, increasing subsides for tertiary education, and also reversed a decision by the National government to increase the rate of student loan repayment (McTaggert 2005). One of the most drastic reversals of past policy was in aid to children and families. of th e 1990 s and an unacceptable level of poverty had emerged. John 2006, 182) The government passed t hree separate tax credits aimed at assisting families with children: The Parental Tax Credit, A Child Tax Credit, and a Family Tax Credit (McTaggert 2005) 9 The government also recognized an increasing problem in education and passed several policies to h elp at risk students. The center left committed itself to opening new schools in lower income areas and placing more social workers in these schools to deal with the problem of attrition among these students (McTaggert 2005). In 2001, an additional 23 mi llion dollars was allocated to reducing 9 These credits would later be replaced in 2004 by the Working for Families package.

PAGE 289

289 the attrition rate in schools. Funding for safe environments for low income students to the ideal that opportunity in early life should not be determined by who or how well of 104) Commitments to increase education by 300 million a year were also made in 2001. Policy Change Slows After the rush of activity in the first two years of center left government, the policy changes slowed. But, this did not mean that left leaning policies had been abandoned. The commitment to poorer families with children was co ntinued in 2004 with an aggressive program called Working for Families. The program began in 2005 and continues to the present day. The main objective of Working for Families is to provide additional assistance to New Zealanders with families. This is d one through four tax credits: the Family T ax C redit, the In Work Tax Credit, the Minimum Family Tax Credit, and Parental Tax Credit. The Parental Tax Credit provided a 150 dollar a week for the The Minimum Family Tax Credit guaranteed a minimum family income for lower income families with dependent children provided parents work the required number of hours a week. The In Work Tax Credit is similar but provides smaller levels of assistance to all families regardless of income as long as the families work a certain number of hours a week. Finally, the Family Tax credit provides assistance to all families with dependent children without a work requirement (Miller 2005 ). Several other key polici es were also passed after 2004. One of these changes in policy was the reversal of the Employment Contracts Act, passed in the early 19 90 s. This extremely controversial bill stripped New Zealanders of

PAGE 290

290 their collective bargaining rights. Employment law was changed to give trade unions more Northern League l support and helped facilitate national bargaini ng (Vowles 2005, 1137). The government also granted civil union rights to same sex couples, a bill that passed on a conscience vote through support from the center left members. Overall, the shift in policy did not reclaim the generous social policies of the past. Now is a good time to pause and detail how the changes (or lack thereof) in policy and the lack of an increase in the process of representati on due to the trouble of coalition formation and part y switching relate to the proportional model To reiterate, the proportional model states that as proportionality increases, the number of ideologically diverse parties also increases (due to the reduce d impacts of mechanical/psychological effects of the electoral system). Ideologically diverse parties are not enough to encapsulate the process of representation. In order for citizens to be truly represented, lved in the decision making process. reached will be a better reflection of what the people want. This means overall better policy for the country. When citizens are both r epresented (through agents in the decision making process) and the system produces better policies the expectation is that overall support will be higher. How can the historical data from New Zealand on policy and process under a new MMP electoral system i nform this model? What is clear is that the model is less than successful in predicting what occurred in New Zealand after the change to a more proportional system. When the electoral system became more permissive, more political parties were allowed int o the system. Under FPP, Labour and National

PAGE 291

291 dominated elections; under MMP, more parties were able to win seats, which increased the amount of ideological representation (parties winning seats included those on the left (Greens and Progressives), right ( ACT and United Future), and center (NZ First). After this step, the proportional model becomes less of an explanatory tool. In terms of making process, New Zealand fell short. Parties were never really able to cope with coalition formation, at least in a way that was satisfactory to the voters. The first coalition formed under MMP was an unmitigated disaster, no one wanted a National/NZ First government, but that is what party elites d elivered. While coalition formation never reached the level of disaster as the 1996 government, parties still were unwilling or unable to state clearly the parties that they would be likely to work with following the election. Even when parties did state clearly, like in 2005, they were willing to throw aside their promises for political expediency. Overall, in short, New Zealanders did not receive the kind of representation from the coalitions that the proportional model might suggest. There was also t he additional, new process of party hopping that had a negative effect on ted through a proportional list took the initiative to switch party lis t have no independent standing in the electorate, but through the party that supports them. When the practice of party hopping became more frequent (at least in the minds of the voters), the representative character of MMP was also undermined. The new e lectoral system overall, did not provide the type of increased representation that the proportional model suggests. When this link in the chain is broken, the likelihood of good policy is lessened. There were moments, like in 1999,

PAGE 292

292 when parties did stat e prior to the election the parties that they would like to vote for following the election. I do not find it coincidental that this is the time when policies that more closely represented what New Zealanders wanted were enacted. Other than that brief pe riod, New Zealand governments were unwilling to commit the country to the generous social policies of the past although they did improve the social character of legislation on a piece meal basis. The historical narrative makes it clear that the transition to MMP was not a smooth one 10 What is left to determine is the extent that these process and policy difficulties had an impact on support for the existing electoral regime. In order to make this determination, I use a series of logit models to test diff erent policy and process variables on support for alternative electoral systems. Testing Policy and Process Effects: New Zealand 1999 2008 In this section I test the extent to which the polici es and processes of the first four governments in New Zealand under the new electoral system affected whether citizens of that country preferred MMP or an alternative system. To do so, I use data from the 1999, 2002, 2005, and 2008 New Zealand Election Studies, which is continuing series of surveys that have traced reaction from New Zealanders since the referendum on the adoption of MMP was passed in 1993 (with the first election under MMP taking place in 1996). I use data from the studies to test the relationship between policy, process, support, and demographic va riables (age, sex, education, and income) to determine whether they have an impact on MMP versus an al ternative system. This section is divided into three sections. First, I describe the variables that I use as representations of policy and process as we ll as the control variables that are used in the models. 10 New Zeala nd did maintain an electoral tier, but because of the increased proportionality I focus on the representational character of New Zealand election results.

PAGE 293

293 Second are descriptions of the statistical models and the expectations of the results that these models will yield. Finally, I present the results and discuss what the results mean for both the suc cess and failure of MMP and how they offer evidence of the theoretical links I draw in Chapter 3. Dependent Variable The dependent variable that I use in the statistical model is one that measures whether or not New Zealanders prefer MMP to alternative ele Election Studies (NZES) that asks the public how they would vote in a hypothetical referendum if it were held today. Voters were given the choi ce between MMP, FPP, an alternative system, a combination of FPP and MMP, a corporate model, single transferable vote (STV), or single member plurality (SMP). I recode these choices to create a dummy variable where a 1 represents a desire to keep MMP, whi le a 0 signifies support for an alternative electoral system. There are several theoretical reasons for using this measure as a dependent variable. Theoretically, the existence and feasibility of alternatives is necessary for support to have any real me aning. Richard Rose in his works (Rose and Mischler 1996; 30) to post communist co untries. Rose argues that support for democracy has no meaning in contexts where the population has no conception of an alternative system. In other words, support for the current system only matters when there is a different system to compare it with. In the post communist cases Rose compares feelings about the new democracies versus their alternative authoritarian regimes. Since, I argue

PAGE 294

294 previ ously (Chapter 2) that Western d emocracies were not a threat to become authoritarian, scholars should look to countries that changed electoral systems and then evaluate how the new system was viewed by the population. It is my view that the success or failure of new systems is based on how the country and corrected the problems of the past. In a similar fashion, Rose and Mishler (1996) argue that those worse off economically under democracy would be more likely to be more supportive of authoritarianism. I take a different tack by determining how citizens evaluate how the new system has corrected the policies and processes that were deemed to be problematic under FPP. Specifically, I argue that MMP will only be deemed successful policy wise if voters feel that the social structures that were d ismantled under FPP were returning. Process wise, MMP w ill be success if representation and a connection to voters and their desires are embraced by elites. By MMP by determining the extent of the relationship between MMP and a series of variables representing policy and process. Independent Variables There are two groups of independent variables that have particular note, policy and process variables. For poli cy, I use a series of variables that measure whether or not people felt the government should be responsible to provide certain social services. These services include (asked in separate questions and variables) a decent standard of living for the elderly a decent standard for the unemployed, jobs for anyone who needs one, housing for those who can not afford it, universal health insurance, and free

PAGE 295

295 primary to college education 11 There are other survey questions that could possibly be used as alternative independent variables. They are questions that attempt to determine the level of interest in similar issues ( health, education, or welfare) a s well as survey items that seek to determine the support for additional spending in the same areas. I did not u se these queries for a few reasons. One of the advantages of the appear in every NZES from 1993 2008, which allows for true comparability over time. The spending and interest questions are not included in all surveys. I also hesitate to use question s related to interest and spending due to the possibility of measurement welfare migh t be spurred by such divergent opinions as those who advocate the complete abolition of welfare, or those who believe that it is a right that should be that the word spending might bring among responders. Frankly, citizens might instantly sponsored programs might be lost. The spending questions are also not asked on all of the surveys from 19 93 to 2008, so some degree of comparability would be lost. The questions I use have the advantage of capturing the beliefs and feelings that New Zealanders have about the social state in general. I argue in Chapter 4 that MMP was adopted in part because t Zealanders was dramatically and quickly taken away. The policy variables, represented by these questions, gets at general feelings that New Zealanders have about the role 11 For exact wording of the survey question see appendix.

PAGE 296

296 that the state has in the areas of health, welfare, and social security, unemployment, and housing. Those respondents who believe that the state has a responsibility in those policy areas might be more favorable to MMP or look to the past fondly (when the state was more active) un der FPP depending on the performance of the new government s under MMP. I use another ser ies of independent variabl es to capture policy, support for lower taxes and where respondents place themselves on the left/right scale. In Chapter 3, I use to conceptualize the dimension upon which party competition takes place (Downs 1957). To reiterate, v oters, in the estimation of researchers, have a solid understanding of what the terms mean and are able to place themselves accurately o n this scale (McAllister and White 2007; Dalton 2009). This does not mean that individuals have an understanding of theoretical concepts like socialism, communism, or liberalism. But, the scale can be used as a general summary of the particular issue cle avages that define the political competition in 1994) and as a meta issue t political syst 273) Scholarship has also shown that where individuals place themselves on the left/right scale is strongly correlated with their positions on the important pol itical issues facing their country (Dalton 2006). issues of the day. A summary of policy issues is appropriate for the questions that I am asking. In Chapter 4 I argue tha t those who supported more government involvement in

PAGE 297

297 all areas favored a switch to MMP. In Chapter 5 I use the same measures to see if that relationship remained the same after the change. Also, a summary variable like the left/ right scale offers a nice comple health, welfare, unemployment, old age, and housing. Between these variables, I areas of policy can be ascertaine d. In addition to these variables, I include one more variable in the policy area: support for lower taxes. The final policy variable I have is a measure that gauges the respondents support for reducing taxes. I add this variable for two reasons. The fi rst has to do with one of the main policy goals of the Labour/National governments from 1984 1996, to drastically reduce tax rates across the board. These tax cuts were financed by drastically reducing social services. The second reason is that support f or higher taxes could be seen as support for a government that is more active and has to pay for more generous social welfare policies. It might be easier to say that you support generous education or gree with reducing taxes, this means that you are willing to pay for those benefits. All total there are three main components of the policy related independent variables. There are a series of six survey questions from the 1999 2008 NZES that estimate wh ether or not a respondent believes that the government should be responsible for providing certain social services. These services include benefits for the elderly, unemployed, those with difficulty affording housing, education, providing jobs, and univer sal health care. The next, support for lower taxes, indicates both the level of commitment a person has for paying for the social programs they want as well as a

PAGE 298

298 rejection of the tax cutting policies of previous governments. Finally, I include a left/ri ght on a range of issues. Between all of these variables the model will include both specific and general policy positions. This should give a good indication of how y positions affect how they feel about the success or failure of MMP. I have also described pa rticular processes that affected how people felt about the new electoral system and I include those in the model as well. In Chapter 3 I use a ser ies of process variables that will not be inclu ded in the model Chapter 5 These process related survey questions asked the respondent how important they believed certain factors were to the way their Member of Parliament made decisions. Respondents were election promises ere to the party caucus, interest groups, or their electoral mandate (pre election promises). These are the processes that helped garner support for FPP. I argue that when these processes were abandoned (especially the electoral mandate), support f or FPP dropped. In Chapter 5, I do not use this series of questions because they were not asked in the surveys after 1993. Since the focus here is on the whether there was an increase in representation the t wo new processes under MMP that I consider those effe cts that will affect success: coalition formation and party hopping. I have already described how learning, among elites, played a role in the process of forming a coalition. The first coalition between National and New Zea land First was

PAGE 299

299 no t a popular one and the actual process of forming was even less so. There was a degree of learning that took place afterward with the Labour led coalitions that followed. I expect that this learning would impact the way people judged MMP In order to measure this process, I use a survey question that asks respondents whether or not they believe parties should signal, prior to the election, what other parties they would potentially enter into a post election coalition. A second process t hat could also impact how New Zealanders felt about MMP was the continued practice of party hopping. legitimacy of their seats was called into question. There is a survey q uestion in the switched parties should resign. Both of these new processes might have an impact on how a citizen views the new system of MMP. The policy and process variables tha t I have just described are my main independent variables of interest. There are also other independent variables that will act as important control variables. I use standard control variables in my models: age, education level, income, and gender. Outs ide of these demographic variables, I also use a survey question that determines the level of trust a person has in their government. The trust variable asks respondents whether they agree with the following what is right most of the tim one of my indicator variables for specific support while preference for alternative systems is indicative of more diffuse attitudes I have already described how I feel the relationship between support and insti tutions (and institutional change) is expressed, so

PAGE 300

300 I will not repeat it here. These are the variables that are used in my models, in the next section I describe what type of statistical techniques are used. Methodology The statistical tec hniques t hat I use in C hapter 5 are a series of binomial logit models. These are appropriate procedure s because of the nature of the dependent and independent variables. The dependent variable is a dummy variable where a response of 1 indicates a willingness to v ote for MMP in a hypothetical referendum, while a 0 represents a desire to vote for another system. The independent variables are also taken from survey questions and create ordinal variables. The use of a dummy, dependent variable and independent, ordin al variables calls for the use of logit models. The post estimation techniques are odds ratios, which help to explain the relationships between dependent and independent variables more clearly. I also use descriptive statistics as well to help interpret some of the findings of the logit models. expectations for what the models will show. These expectations are rooted in the historical analysis that I described in th e early sec tions I separate these expectations by different NZES, and include a brief explanation of why I believe those results will bear out. Expectations I have several expectations for what data the models will bear. Speaking generally, I believe the impacts d emonstrated in the logit models will follow the historical narrative that I lay out in the earlier section. This section details the expectations that I have for the process, policy, and independent variables that I chose for the statistical tests.

PAGE 301

301 In ter ms of process, there are two main variables of interest: coalition formation and party switching. To briefly summarize, the process of coalition formation was new to New Zealanders under the new system of MMP. Previously, governments were formed on elect ion day with one of the two major parties winning the election and forming a government. This was changed under MMP, with no elections granting a single party a majority of seats. From the outset, coalition formation became problematic. Twice, in 1996 a nd 2005, coalitions were formed contrary to what leaders of parties declared during the election. Several of the coalitions were not able to hold together for the duration of their terms as well. Because of this, I argue that New Zealanders had reason to believe that they were not being informed properly prior to the election about what the composition of the government post election would be. As a result, I expect that as respondents feel more strongly that each party ought to state explicitly before th e election their preferred coalition partner, the more they would also favor an alternative system (representing a loss of support for MMP). The other process the received significant attention was the increasing number of s practice seriously undermined those who were elected whose election was through a party list, rather than through a single member district, lost the legitimacy t o hold office if they left that party. This was such a contested issue after the passage of this law, members of the Progressive Alliance shattered its spirit when they op erated in parliament as two separate groups. Because of this I anticipate

PAGE 302

302 likely to support an alternative electoral system. The other section of independent variables of interest has to do with policy. I contend that as policies moved away from the neo liberal reforms that came under FPP the more favorable people would view the new electoral system. I describe in the policy section several policies the Labour led, coa lition governments passed that s l owly chipped away at the neo liberal reforms of the past. In a series of legislative measures, they returned some of the social benefits that New Zealanders had lost. Because of this, I have several suppositions regarding the specific variables. I use the variable left/right scale as a summary variable that absorbs the policy positions of respondents. I presume that those who are more conservative would favor FPP. As far as the questions dealing with the specific issues of government involvement in the area of taxes, housing, education, unemployment, old age, health care, and jobs I believe the results will be mixed. I expect this since social welfare policies were dealt with on a piece meal basis over time. By 2008, s ocial welfare policy in New Zealand had definitely turned much more liberal. Because of this, I assume that those in favor of increased government action will display support for MMP. There are several independent variables of interest in the models as well. One of these is the variable measuring trust in elected officials. I expect that as trust increases the more people will favor the existing electoral system of MMP 12 I include a series of demographic, control variables including: age, gender, educ ation, and income. I do not expect significance from the gender variable. The only possible significance I can 12 The survey question asks respondents wheth

PAGE 303

303 imagine is if there is a gender gap where men prove to be more conservative than women. If this occurs, then I expect men will favor FPP. I b elieve similarly about the other main control variables. For instance, if older voters are more conservative in general, they will favor FPP. I predict something similar with education and income as well. If the higher educated prove more liberal (or le ft), then they will favor MMP. Also, if those with higher incomes are more conservative, they will favor FPP. The following section displays the results of the logit models and also discusses, with the aid of some descriptive statistics, what the results mean. Results This section reports the findings in the logit models. I present the findings in passages dealing with policy and process variables separately in order to analyze their impacts on support for MMP. For a complete list of findings see T abl e 5 2 Process Variables New Zealanders were presented with new processes following the adoption of MMP in 1996: coalition formation and party switching. I argue in Chapter 4 that the connection between citizens and their elected officials was lost when s uccessive Labour and National governments dramatically changed the relationship between the state, the economy, and society. I present evidence in Chapter 4 that there was a loss of connection between voters and their representatives. Voters felt like the ir elected officials no longer paid attention to their mandates, which had the ancillary effect of bringing about the change to MMP 13 The goal of MMP was to improve the connection 13 The dramatic changes in policy were the driving force behind the electoral system change. Voters also felt that their elected officials were not h onest with them about their policy intentions. But, it was the policy changes that preceded these feelings and are therefore the root cause. For a complete explanation, see Chapter 4.

PAGE 304

304 between citizens and the elected. But, following the change to MMP histori cal evidence shows that coalition formation and party switching undermined that connection. An examination of the logit models provides further evidence that this is the case. The main finding from these results is that the new processes had a negative effect on support for MMP. As respondents increasingly believed that parties should state before the election wh ich parties they would form a coalition with, the more they favored another type of electoral system. This is a consistent, significant findi ng from 1999 to 2008. In 1999, for every unit increase in belief that coalitions should be named prior to the election, the odds that they supported MMP decreased by 15.6%. This negative relationship remained consistent in the other years as well. The o dds of choosing MMP decreased 31.4% (2002), 24.1% (2005), and 17.0% (2008) as well. This relationship, considering the historical evidence, is not unexpected. Leading in to the 1996 election, New Zealanders had an expectation for the type of coalition that they might receive after the results were tabulated. The parties, especially NZ First, offered mixed signals about who they would prefer to work with 14 Voters, on the other hand, developed a clear sense of what parties they wanted in the coalition. Vir tually none of the National or NZ First voters wanted to form a coalition with the other party. Most expected a center left government following the election, but instead they received the National NZ First coalition. As I describe in an earlier section, the process of forming this coalition raised serious questions about the intentions of both parties, especially with regard to how the two parties regarded the views of their constituents. 14 It should be noted that Winston Peters, the NZ First party leader, o ffered some clear signals that National would not be in government following the election. This led people to believe that his party would not enter into negotiations with National.

PAGE 305

305 T he government stumbled through two years before collapsing in 19 98. Ensuing governments learned somewhat from this experience and tried to negotiate agreements prior to the election, but this was only marginally successful. The main reason for the lack of success was the uncertainty of the results of the MMP elections The Labour party and the Alliance worked out a pre election agreement and were set to form a majority government following the election in 1999, but late election results handed seats to the Greens. The center left government now needed the Greens to h elp pass legislation. This sort of uncertainty was unheard of when elections were held under FPP. When the election was completed under those rules, one party won the most seats and formed a government the next day 15 The government that was formed was a lso stable, and would manage their responsibi lities until the next election three years later. The first two governments under MMP did not last through their whole term. The element of uncertainty was new, and even when parties tried to come to agreemen ts afterward, their plans might be thrown array. This happened again after the 2005 election, where pre election promises were thrown aside. The Labour party campaigned on a promised Labour Progressive Green alliance, unceremoniously dropped the Greens f ollowing the election. They did so when they realized they could make a better deal by offering important positions to two other party leaders. One of these leaders was Winston Peters, NZ First leader, who again brokered his electoral position into a pos ition of power. When coalitions were formed against the expectations of the electorate, accountability was lost. The connection between voters and their 15 Generally, there was ample evidence about which party would win the ele ction as well.

PAGE 306

306 representatives was something that was supposed to be improved under MMP, but the process of coalitio n formation did little to increase that connection. The fact that those who felt strongly that parties should name their coalition partners before the election were more likely to support an alternative electoral system is a reflection of this reality. On ly in 2002 did the Labour party say that it would work with the Progressives before the election. But this came on the heels of a governmental showdown with the Alliance, who fell apart, and the Greens who walked out of parliament. Coalition government i ntroduced an element of uncertainty that had never existed before. Stable, one party governments were replaced by unstable coalition governments, whose make up was not clear prior to the election. This had a negative impact on support for MMP, but was no t the only process that had this effect. The persistence of party switching after MMP was enacted also harmed support for MMP. When New Zealanders voted for MMP through a referendum in 1993, there was a rash of entrepreneurial, party building behavior. T his included a number of members of Labour and National leaving their respective parties to form others. Party switching was done for a number of reasons ranging from ideological differences with their old party to an electoral calculation. When the prac tice continued after MMP was implemented, there were serious questions about the legitimacy of switches. Party switching became such a large national issue that it spawned legislation in 1999 requiring those who changed parties resign their office. Becau se of this I expect that those who felt that ively about MMP. The resu lts of the logit models confirm my expectations that as respondents ld resign, their support for MMP

PAGE 307

307 dropped. The question on party switching was only asked in the 1999 and 2002 New Zealand Elections Studies 16 so the examination of its impacts is limited to those years. In 1999, with every unit increase in feelings that resign led to a 12.3% decrease in the odds that the respondent would favor MMP over resign when switching parties, the odds of supportin g a system other than MMP increased 14.4%. Both results were significant at .01. These results also confirm my expectations and demonstrate that several events had an impact on support for the electoral system. The practice of party switching became an i ssue for several reasons. There was a parties, and continued to sit in parliament under as an independent or the leader of a st are not directly elected in a single member district, so there is no direct citizen connection. The legitimacy of their office comes to hold that seat could be ques tioned. This became a larger issue when it w as discovered that as the head of a new political party, the party switcher was entitled to a practice was so unpopular that a law was passed in 1999 requiring that members resign if they left their parties. The logit analysis shows that party switching was still a drag on support for MMP in 2002 a full three years after legislation was passed. That is because parties merely 16 This is a curious choice since party switches periodically occurred well after this date.

PAGE 308

308 fou nd ways around the law. Specifically, the Alliance fractured into two separate entities in late 2001 and continued to sit in parliament and as part of the government. These two groups operated as two separate political parties, and the leader of the Alli ance party Jim Anderton was cast out of the Alliance by the party apparatus outside of parliament. But, because the party formally split, but the switch was already clear to the public. This is why party switching continued to have a negative impact on support for MMP. MMP was adopted, in part, to improve the relationship between the public and elected officials. Under FPP, elites behav ed as if the will of the public did not matter. People felt like thei r politicians were selling them one thing before the election and delivering something completely different afterward. MMP was supposed to improve those conditions. The way governments went about the process of forming coalitions demonstrated that MMP ha d not improved the situation at all. The uncertainty of election results made coalition agreements difficult to plan, and even when pre election plans were made, they were tossed aside when politically expedient. The practice of party switching was also another process that demonstrated that the gap between elites and citizens remained wide. While the practice took place under FPP as well, the fact that list caused a drag on su pport for MMP 17 I describe in the discussion section how in the absence of major changes in policy, these processes were damaging to support for 17 electoral connection to the voters outsid e their membership in a party, left the parties that were responsible for their election in the first place.

PAGE 309

309 MMP. The next section deals with how policy impacted support for MMP over this same period. Policy Variables While the process variables, coalition formation and party switching, had a clearly negative impact on support for MMP, the policy variables are not as clear. Two variables left/right scale and willingness to pay taxes yield the most compelling results. The remaining policy variables do not present as clear a picture, but still provide complimentary evidence that policy had a role to play in increased support for MMP electoral system. There is a clear connection between the left/right scale and willingnes s to pay taxes and support for MMP. Self placement on the left/right scale has the most consistent and clear relationship with support for MMP. Over every year in the study this variable has demonstrated that the odds of supporting MMP decrease as the re spondent is more conservative. Results are highly significant in all years. In 1999, with a one unit increase in conservatism, the odds of supporting MMP decrease 18.5%. This relationship remains the same in 2002 with a one unit in crease in conservatism lowering the odds of supporting MMP by 15%. This connection also exists in 2005, with the odds of MMP support decreasing 20.6% as the respondents become more conservative. 2008 is no different with an 18% decrease in the odds for a one unit increase in conservatism (all results are highly significant). believe the results that this variable yields have important conclusions for understanding support for MMP in New Zea land. Overall, MMP has been a loss for conservatives. They simply did better in elections and their policy goals were met under FPP

PAGE 310

310 Electorally, National (the right leaning of the major two parties) performed much better under FPP. From 1975 to 1996, National controlled the government in 15 of the 21 years. Following the first elections under MMP, they still formed a government, but in a coalition that their voters neither wanted nor expected 18 In the election of 1999, 2002, and 2005 National was re N orthern League ted to an opposition party. I believe that part of these results can be explained by electoral self interest. National voters were simply not winning elections under MMP. Electoral self interest is only part of the story though because con servative voters also did much better under FPP policy wise than they did under MMP. The entire package of conservative, neo liberal policies that precipitated electoral system change were passed under the old system of FPP. Taxes were cut across the boar d, farm subsides were eliminated, social services were slashed, and government involvement in the economy was curtailed, from 1984 1996. After the elections in 1996, these policies were slowly rolled back. Taxes were raised, welfare to work policies were eliminated, and programs providing government funds for families were increased. New Zealand did not return to the gener ous level of social spending that preceded 1984, but they did move in that direction. Conservatives certainly are reacting to that ch ange in policy direction from 1999 2008 and evidence is apparent in the logit results. The other policy item that demonstrated clear results was the variable that measured willingness to pay hi gher taxes. The post estimation odds ratios show a positive relationship between willingness to pay higher taxes and support for MMP. In 1999, with a one unit increase in the willingness to pay higher taxes, the odds of 18 For a full explanation of this, see earlier sections in this chapter.

PAGE 311

311 supporting MMP increased 26.8%. The increases in odds were: 10% in 2002, 21.6% in 2005, and 19 .4% in 2005 for one unit increases in willingness to pay taxes (all are statistically significant). This is not a surprising result. Generally speaking, the desire and willingness to pay higher taxes could be considered a l iberal position. The National P art y and those to the right of it consistently made lower, flatter tax rates a part of their platform. Also, the neo liberal reforms of the FPP era included large cuts in rates as well as a flatter tax structure. According to the logit results, those wh o wanted a return to the more progressive policies and tax rates of the past favored MMP. The remainder of the policy variables, particularly those that I use to encapsulate support for a more interventionist, social state, does not clearly indicate suppo rt for or against MMP. My feeling for this that center left governments took a piecemeal approach to social policy from 1999 to 2008 One of the clearest relationships that the model demonstrates is the connection between housing and support for MMP. Un like the variables that I have already described, the relationship between the beliefs that government should help provide housing and support for MMP is not statistically significant across the four years of interest. However, there is a consistent, posi tive relationship even if the coefficients do not reach the accepted levels of significance in every year. There are significant results in 2002 and 2005 where with a one unit increase in the conviction that government should provide housing there is a 30 .8% (2002) and a 24.7% (2008) increase in the odds of supporting MMP. The reason that I believe that this relationship is consistently positive is that the Labour led coalitions did make housing a priority. Early in their first term, the government passe d the Housing Restructuring Amendment Bill, which reestablished state control on rents. This Bill

PAGE 312

312 This is the kind of state intervention that existed during the FPP ye ars, and those that favored this positive government action were more likely to support the new electoral system. Another variable that yielded positive, although not consistent ly positive results was education. The belief that government should provide c hildhood to college education also had the relationship with support for MMP that I expe cted. In 1999 and 2002, with a one unit increase in the feeling that government should provide education the odds of supporting MMP increased 19.7% and 14.7% respectiv el y. Afterward, the relationship remain s positive, but do not reach levels of statistical significance. The reason for this is the immediate push by the Labour led coalition to reverse some of the policies of the previous governments concerning education The center left acted early in 2000 to increase funding, upwards of several hundred million dollars, in several different areas of education. The government moved immediately to cut interest rates on student loans. The previous National/NZ First gover nment had increased the interest and moved the scheduled payments for college loans, a move the Labour led government committing to building new schools across the cou ntry, as well as ending block funding for school distri cts. These policies would help to explain why the significant relationships exist in 1999 and 2002. The reason the variable fails to attain significance, to really pass any significant reforms after 2004. There was a commitment to increase funding for education, but no moves to provide the kind of educational benefits that existed before 1984. The failure of the model to produce

PAGE 313

313 significant results in 2005 and 2008 coul d be a case of lost momentum on the issue of education. Two other policy variables that have mixed results are feelings about the level of government responsibility for providing jobs and help for the unemployed. These are variables that seem to be si milar in nature and produce similar results. In 1999, both have positive relationship with support for MMP with increasing feelings that the government should be involved in aiding the unemployed and helping provide jobs leading to increased support for M MP. Only the jobs variable yields significant results with a 26.8% increase in the odds of supporting MMP for every one unit increase in the belief of government intervention. This is not altogether surprising, since I believe that in 1999 the more liber al positions on issues would be related to MMP. What is interesting is that in 2005, the relationship switches. The two variables have significant, negative relationships with support for MMP. With one until increases in the determination that the gover nment should provide jobs and should help the unemployed, the odds of supporting MMP decrease 15.1% and 18.1% respectively. A possible explanation for this was the criticism that the Working for Families package of tax incentives received from liberals af ter its passage in 2004. The main thrust behind the criticism was the work requirements needed to collect the benefits. Many were insulted by the logic behind the bill, namely the belief that all the unemployed needed to improve their circumstances was g overnmental motivation. Also, there was criticism of the delayed payment of benefits. The bill was passed in 2004, but partial, initial payments were not given until 2005, with no increase in 2006, and a small increase in 2007 (Miller 2006). Full benefi ts were given in 2007 and these increased payments and

PAGE 314

314 the attention they received could be a reason for the change in direction of the relationships. There is a statistically significant relationship between those who increasingly felt that government sh ould provide jobs and support for MMP. The final policy variable, government should provide for the elderly, yielded no positive results. The coefficients and odds ratios are also very slight, suggesting that the variable has virtually no theoretical sig nificance. It seems likely that since the policy battles were generally waged in the areas of health, welfare, education, and taxes perhaps the issue of care for the elderly was overwhelmed by the more dominant issues of the day The most important of the control variables is the item measuring the amount of trust a person has in the government. Trust is an indicator of specific support, which may have a connection to more diffuse attitudes related to the preference for MMP or an alternative. I expect th at since trust is a variable that measures specific support, the more a person feels that the existing government is doing well at the present time, the more they will support the current electoral arrangements. This relationship bears significant results in all of the models across all of the years. In 1999, a one unit increase in trust leads to a 9.3% increase in the odds of supporting MMP. The increase in odds in the subsequent years is: 14.2% (2002), 12.8% (2005), and 20.7% (2008). Other control var iables also yield interesting, although not unexpected results. Men had significantly higher odds of supporting a system other than MMP in 2005 (10.1%) and 2008 (20.1%). Those who were higher educated had increased odds of supporting MMP with significant results in all years. Those with higher incomes have greater odds of preferring a system other than MMP, with significant results in 2005 and 2008.

PAGE 315

315 What is left is to determine how policy and process impacted overall support for MMP. Public support for MMP fluctuated between 1996 and 2008, the question is why? I have laid out a case using historical data that policies were slow to change after the electoral reform and the government never committed to provide the services that it had in the past. Also, the historical record shows that coalition formation and party switching were two processes that became very important after the switch to MMP as well. The statistical models show that policy and process had an impact on support for MMP. The following s ection concludes with a discussion of how policy and process impacted the fluctuation of support for the electoral system from 1996 to 2008. Discussion: Support for MMP in New Zealand With only brief periods of improvement, support for MMP never reached th e level of public approval that existed before the system was enacted. What follows is a description of why levels of support have fluctuated and why MMP, for the most part, was consistently outpolled with a brief period of higher support from 2001 and 2 002, by alternative choices. Figure 5 1 is a representation of the changes in support between MMP versus and alternative system. The two highest points of public support for MMP were actually before and during the first election. In the 1993 referendum t he public voted for MMP over FPP by 54 % to 46 % Prior to the first election under MMP in 1996, support for MMP was at it highest point with 57% of New Zealanders preferring MMP to 43% favoring FPP. In 1998, MMP support dropped precipitously, with only 37 % of New Zealanders offering it as their preference, and 63% favoring an alternative. Support rebounded somewhat in the following three years. In 1999 support for MMP stood at 45%, improved to 52% in 2001, and maintained that level in 2002, making the tw o years a post electoral system

PAGE 316

316 change high for support for MMP. Alternative systems polled higher than MMP in both 2005 and 2008, with the same scores, 52% (Alternative) against 48% (MMP). The question is why support fluctuated with lows in 1998, a rebo und in 2001 and 2002, and lost support again in 2005 and 2008? The historical narrative I draw and the statistical models I run offer evidence for why this occurred. 1998, when support for MMP hit a low point, is perhaps the easiest to explain. The gover nment that controlled parliament was made up of two political parties: National and NZ First. This was an unlikely partnership that the public neither expected switching, from list elected par ty seats to independent seats, without any direct electoral legitimacy from the voters (without the parties). If MMP was to be judged, as I theorize, by improving the connection between voters and their representatives, the process of coalition formation and the persistent cases of party switching shattered this connection. The pre MMP governments they also did little to reverse them. Since New Zealanders as a whole s upported more government involvement in areas like housing, education, welfare, unemployment, and help for the elderly the failure of the National/NZ First The improvement that came followi ng 1999, in my estimation, has to do with changes in policy. The Labour/Alliance government of 1999 2002 had a rash of policy activity that partially rolled back some of the more conservative policies of the past. They raised taxes, increased spending on education, and eliminated some of the work for welfare policies enacted by National in the previous government. The policy action, I

PAGE 317

317 believe, led to higher levels of support for MMP. The reason I feel that policy is the reason for the improvement in sup port is that party switching and coalition formation, according to the logit models, continued to be a drag on support for MMP. The action of the government was enough to override those processes that weakened support for MMP. It is when the Labour led c oalition slowed its policy action after 2002 that support began to wane once more. Labour after 2002, took the policy position (perhaps because of their electoral the y had staked in the center, preferring instead to pick away at the edges of social policy. They did make some efforts in the area of education, but passed a series of policies designed to provide further assistance to working families that was criticized on the left for its work provisions. After 2004, there were virtually no major changes in social policy to speak of. Without changes in policy, process became an issue once again. The 2005 government was formed contrary to the campaign pledges of Labour who ditched the Green party after the election. The coalition was formed with two parties who were not even mentioned by Labour. When the leaders of those parties were offered plum positions outside of cabinet in exchange for their support, there were parallels drawn to the unpopular coalition process of 1996 (one of the main benefactors was Winston Peters in both coalition bargains). In the absence of good policy or process, support for MMP dropped. Success es /Failure of the Proportional Model: Implica tions Both the historical narrative and the logit models demonstrate that the proportional model was less than successful in predicting what would happen in New Zealand following a change to a more proportional system in 1996. Several factors kept s upport

PAGE 318

318 from increasing : the failure of representative processes including coalition formation and party hopping and the inability of governments to reenact generous social policies. In this section I would like to detail briefly the general conclusions I draw f rom these failures. The first conclusion I believe to be of some import is the confounding role that individuals can have on the most carefully designed institutional models. This is especially clear in the process of representation in New Zealand. Coali tions were clearly affected by the personal characteristics of some political elites. Not to lay all the blame for the negative feelings New Zealanders had about coalitions (and as a result, the negative feelings for the system as a whole) in the lap of W inston Peters, but he clearly had a large role. Popkin (1991) suggests that singular events can shape how coalition building process received so much attention, and his actions were so egregious, that they plagued the first coalition formed under MMP. His, and the behavior of others, could be blamed on old political hands attempting to operate within a new institutional framework. But, this does not explain why a similar circumstance, with virtually the same political figures happened in 2005. Winston Peters was again in the center of what was considered to be suspect political negotiations when he received a upport for the new Labour led Zealand to leave their political parties. Alamein Kopu, left the Alliance to form her own

PAGE 319

319 independent political party despite the fact that she was elected because of a political party and when she ran in a single member district o nly garnered 5% of the vote. By forming this political party, she received a higher parliamentary salary. This switch, and others, led Alliance leader Jim Anderton to make party hopping a political issue. Anderton, himself, ended up splitting the Allian ce into two separate parties without resigning as the bill he supported demanded. What institutional rules or models can explain why individuals engage in these types of behaviors? It seems, at least from these examples, that the concerns I raise in Chapt er 2 are well founded. I posed the questions, are parties and leaders willing to cooperate and come to decisions by consensus, or are they more likely to horse trade, bargain, and seek their own benefit above the good of others? It seems from New Zealand that the electoral system was unable to change the basic nature of these elites. Individuals, when given latitude within institutional contexts, are free to pursue their interests as they see fit. This is why I would like to forward the proposition that electoral systems impact the type of party system that is created, but the behavior of that system depends on what is permissible under the rules. There was nothing in New Zealand election law that demanded a particular coalition after the 1996 election, so elites (especially NZ First) pursued the best deal for themselves. When Alamein Kopu saw an opportunity to improve her personal and political situation by switching parties, she took it. When the Progressives wanted to split into separate parties wit hout resigning their office, they did so, finding a loophole in the newly passed law that suggested that they should resign. These are decisions and actions that had dramatic impacts on representative processes, but they have nothing or little to do with the type of elections that were held.

PAGE 320

320 To put another way, the electoral system can only shape or control so much, the behavior of parties and elites are beyond their capabilities. Something similar can be said about the types of policies that are enacted. Here, it is important to distinguish between what the electoral system is responsible for and what the party system is responsible for. If the types of coalitions that are formed have more to do with the actions of individuals and parties than elections then it stands to reason that the types of policies that they enact follow the same logic. The adoption of MMP, the increase in the number of ideologically diverse political parties did not lead to more representative coalitions and did not lead to the adoption of more generous social policies. But should substantially different policies be expected in the first place when all that changed was the electoral system? This is a point that I explore further in the conclusion, but before that I would like to make a comparison of what happened in New Zealand following electoral system rela tive ly unprecedented heights after six years with mixed member majoritarian, these initial positive reac tions were lost in the ensuing five years. Chapter 6 details the processes and policies in Italy under a new, more majoritarian system and their relati onship to the levels of support in the country.

PAGE 321

321 Table 5 1. Summary of parliamentary election r esults: New Zealand 1996 2005 Parties 1996 1999 2002 2005 % Seats % Seats % Seats % Seats E L T E L T E L T E L T National 33.8 30 14 44 30.5 22 17 39 20.9 21 6 27 39.1 31 17 48 Labour 28.1 26 11 37 38.7 41 8 49 41.2 45 7 52 41.1 31 19 50 NZ First 13.3 6 11 17 4.26 1 4 5 10.3 1 12 13 5.72 0 7 7 Alliance 10.1 1 12 13 7.74 1 9 10 ACT 6.10 1 7 8 7.04 0 9 9 7.14 0 9 9 1.51 1 1 2 Un ited NZ 0.88 1 0 1 0.54 1 0 1 Greens 5.16 1 6 7 7.0 0 9 9 5.30 0 6 6 United Future 6.69 1 7 8 2.67 1 2 3 Progressives 1.70 1 1 2 1.16 1 0 1 Others 7.54 0 0 0 6.03 0 0 0 4.89 0 0 0 3.35 4 0 4 Tota l 100 65 55 120 100 67 53 120 100 69 51 120 100 69 52 121

PAGE 322

322 Table 5 2. Effect of process and policy v ariables on MMP versus alternative system c hoice: New Zealand 1999 2008 Independent Variables 1999 2002 2005 2008 B (SE) % Odds Ratio b (SE) % Odds R atio b (SE) % Odds Ratio b (SE) % Odds Ratio Left/Right Scale .204*** (.024) 18.5 .162*** (.023) 15.0 .230*** (.026) 20.6 .198*** (.033) 18.0 Support Higher Taxes .237*** (.046) 26.8 .094** (.047) 10.0 .195*** (.0613) 21.6 .177** (.073) 19.4 G overnment Should Provide Jobs .233*** (.068) 26.3 .082 (.065) 8.6 .164** (.076) 15.1 .182** (.090) 20.0 Government Should Provide Care for Elderly .058 (.090) 5.7 .052 (.089) 5.1 .005 (.114) 0.5 .158 (.132) 17.2 Government Should Provide for Unem ployed .072 (.094) 7.5 .106 (.089) 11.2 .199* (.106) 18.1 .100 (.117) 10.6 Government Should Help Provide Housing .129 (.102) 13.8 .268*** (.094) 30.8 .078 (.116) 8.1 .220* (.128) 24.7 Government Should Provide Universal Health Coverage .060 (.070) 5 .8 .116* (.069) 11.0 .018 (.085) 1.8 .074 (.104) 7.7 Government Should Provide Education from Childhood to College .180*** (.069) 19.7 .138** (.067) 14.9 .047 (.086) 4.8 .001 (.101) 0.1 Parties Should State Who they Would Form a Coalition With .170 *** (.083) 15.6 .376*** (.063) 31.4 .275*** (.059) 24.1 .186*** (.069) 17.0 Parties Should Resign .131** (.054) 12.3 .115*** (.049) 14.4

PAGE 323

323 Table 5 2. Continued Independent Variables 1999 2002 2005 2008 B (SE) % Odds Ra tio b (SE) % Odds Ratio b (SE) % Odds Ratio b (SE) % Odds Ratio You Can Trust the Government to Do What is Right .089* (.052) 9.3 .133*** (.055) 14.2 .120** (.063) 12.8 .187*** (.075) 20.7 Gender .114 (.097) 12.2 .013 (.094) 1.3 .106 (.114) 10.1 .223* (.112) 20.1 Education .182*** (.044) 20.0 .122*** (.029) 13.0 .114*** (.033) 12.1 .172*** (.036) 18.8 Household Income .025 (.029) 2.5 .032 (.029) 3.2 .104*** (.034) 9.9 .118*** (.039) 11.2 Age .011*** (.003) 1.1 .021*** (.003) 2.1 .033*** (.004) 3.3 .016*** (.004) 1.6 P < .01*** p < .05** p < .1*

PAGE 324

324 Figure 5 1. Support for MMP versus an alternative s ystem, New Zealand 1993 2008

PAGE 325

3 25 CHAPTER 6 TRYING MIXED MEMBER MAJORITARIAN ELECTIONS: ITALY, 1994 2005 Introduction Chapter 5 describes the difficulty that New Zealand had with its new electoral system following a switch to MMP in 1996. This new system was designed to correct some of the perceived problems attributed to FPP. The old system became the scapegoat for the changes i n policy and the lack of accountability that began under a Labor government in 1984 and continued by a National led government in 1990. The logic behind the switch was that a more proportional electoral system would break the hold Labour and National had over politics and allow smaller, more ideologically diverse making process, the hope was that better, more widely accepted policies wou ld be enacted (in this case, a reversal of the more draconian neo liberal policies adopted in the pr evious nine years). Higher levels of support among the population were supposed to come with this increase in representation and better policies. There wa s a gap between the expectations of the new system and what actually occurred. The more permissive electoral system did allow increased ideologically representation, with smaller third parties winning seats on a consistent basis. Where the proportional m odel (liking proportionality to support) began to fall apart was in the process of representation. Parties and elites in New Zealand never adequately adjusted to the new electoral rules, especially in the area of government formation. The first coalition government was formed under intense public scrutiny and criticism. Even after the National/NZ First government, which was considered an unmitigated disaster,

PAGE 326

326 parties were still hesitant and evasive about which parties they preferred to work with followin g the election. The process of representation was also undermined by the odd, continued practice of party elected and formed another or joined an existing party. Voters in New Zealand were able to vote for more parties that had a chance to win seats, but post election coalition bargaining and party hopping, eroded representation. MMP was not able to produce the type of representation that was expected. It should not come as a surprise that there was n o return to the social welfare policies of the past. Governments were willing to provide some benefits, but did so in a piecemeal fashion rather than a wholesale return to the generous policies of the past. In fact, after a rash of initial policies follo wing the election of a Labour/Progressive government in 1999, there was little movement on the policy front. Support for proportional government in New Zealand lagged with these disappointing results. The lowest levels of support, came in 1998 when an ext remely unpopular National/NZ First government was about to implode. Higher levels of support came with the election of a new Labour/Alliance government. This was the only time under MMP that parties agreed to form a coalition before the election. This c oalition also delivered on some social policies and passed a law against the unpopular practice of party hopping. The momentum was lost however, when the coalition fell apart before the election of 2002 and the Alliance split into two separate parties wi thout the Northern League lly required resignations. Party switching became less of an issue, but s upport for MMP plunged in the ensuing years as parties fell back into old habits.

PAGE 327

327 Parties made little effort to pass substantive social policy and remained coy about what coalitions they would join following the election. In Italy, the problems prior to reform were a bit different. If policy changes drove electoral system reform (with an assist from process), process changes were front and center in Italy. Here, the partitocrazia that distinguished the First Republic collapsed when parties, which had deep social ties (if weak governing ability), lost control over their previously reliable voters. The clientelist system that marked the First Republic was als o threatened when volatility increased in the mid 1970 s. When the Christian Democrats lost control over the government around this time, there was an increasing market for votes and parties went to greater and greater lengths to maintain cli entelist ties. During the 1980 s the clientelist system exploded into outright corruption, where parties required massive sums of money for most public services. This corruption was unearthed during the Clean Hands i nvestigations of the early 1990 s and succeeded in dec imating the existing party system. There were no real dramatic changes in policy in Italy, but as the corruption scandals unfolded Italian government also embarked on a series of austerity measures designed to curb public debt and placate the demands of the European Community. These policies were not the driving force behind electoral system change, but they became representations of the growing gap between citizens and elites in Italy. By the early 1990 s it was clear that Italy needed a new direction to correct these problems. Rather than adding proportionality, like in New Zealand, Italians sought to reduce the proportionality of the electoral system and replace it with a mixed member

PAGE 328

328 majoritarian (MMM) alternative 1 The majoritarian model I draw in C hapter 2 represents something fairly similar to what Italians wanted when adopting MMM. Italians were seemingly plagued by a multitude of ideologically diverse political parties that formed governments that were weak, fragile, and plagued with corruption. Parties simply did not represent Italian citizens anymore. Governments, due in part to the impacts of process, were completely unable to develop coherent policies to deal with the pressing issues of the day. By adopting a more majoritarian system, the goal was to reduce the number of political parties, create a system of bi polar competition between coalitions who presented competing policy visions for the country. By doing this the hope was to introduce accountability where voters could remove politic ians who were corrupt or not keeping their promises. Through this in creased accountability, the belief was that voters would be able to hold parties to their pre election promises and judge them on the effectiveness of the policies that they passed. Italy successful. The majoritarian portion of the electoral system did encourage parties to form pre election agreements and this created center right/center left coalitions. But, bipol arity did not come coupled with a reduction of political parties I n fact the number of parties winning seats actually increased after the adoption of MMM. Although coalition agreements were sorted out before the election, instability and bargaining stil l occurred and this led to continued instability. The first two governments elected under MMM lasted only eight mon ths and two years respectively. Despite these problems, there were some positive improvements. The electoral system did punish coalitions that 1 A full description of the mixed member majoritarian electoral system us ed in Italy from 1994 to 2001 follows this introduction.

PAGE 329

329 were not unified. This happened with the center right in 1996 and the center left in 2001. Both were penalized and lost the election when they could not come to an agreement before the election. There were other successes as well. The first cent er left government presented to the voters a series of policies and did a fairly good job of enacting them following the election. This government was able to rein in the rising public debt and was able to secure entry into the European Monetary Union. T his period of effective policy also coincided with a general feeling among the population that the level of corruption in the country was declining. This was due to efforts toward reducing corruption through policy, but also because voters were able to vo te out member districts. As a result, support in Italy rose to all time highs in 2001. The gains made under MMM initially, were lost when the center right took office in 2001. Silvio Berluscon i came to power after making a promise to deal with the issue of conflicts of interest and continuing to improve the Italian economy. The center right, plagued by its own internal divisions, was unable to del iver on many of the pledges the coalition made. The government also, rather t han dealing with the issue of corruption, passed legislation to shield its supports from prosecution and railed against prosecutors who were pros ecuting those crimes. Italians increasingly felt that corruption was not being dealt with in a satisfactory ma nner. Because of the lack of accountability and failure to enact coherent policies, support fell from 2001 to 2005. C hapter 5 Italy from 1994 to 2005 The first section outlines the new electoral rules, what the expecta tions for those rules were, and how they relate to the majoritarian model I draw in Chapter 2. The following section

PAGE 330

330 describes how the new MMM electoral system produced mixed results, namely how the party system remained fragmented while encouraging a sys tem of bipolar competition. After, I compare the actions of the first center left government from 1996 to 2001 to the center right government of 2001 to 2006 in the areas of policy and process. Finally, I detail the changes in support in Italy and what t he implications are for the majoritarian model. The New Electoral System in Italy As I state previously, Italy makes an interesting case for study for two main an al often used Italy as an example for what could go wrong with an overly proportional electoral system 2 The second reason Italy makes a good case for study is that Italy change d its electoral system in order to correct some of the perceived problems in the country attributed to the overly proportional rules. Italy, unlike New Zealand, has two chambers with different (albeit similar) electoral rules, which require a bit more des cription. Mixed Member Majoritari an Rules, Italy 1994 2005 The reforms of 1993 actually introduced two different electoral systems to Italian under the same rules save one v ery important feature. In both cases 75% of the seats 2 I argue that these claims about the ineffectiveness of the Italian proportional electoral system are overstated. For a complete description of proportional elections in Italy during the First Repu blic see Chapter 4.

PAGE 331

331 distributed by PR. Both the elections for the chamber and for the senate are classified as mixed member majoritarian (MMM) systems (Shugart and Wattenberg 2001) 3 This distribution of SMD to PR seats is the only similarity between the two. the remaining 155 are elected in 26 multimember constit uencies through party lists. The magnitude of these constituencies ranges from one and eleven. PR seats are allocated at the national level and parties must receive at least 4% of the national vote to be awarded seats, although cartels between parties ma y be formed before the election to improve chances of passing the threshold. Voters cast two ballots for the Chamber, the first for a candidate within their electoral district, with the other for a list in their constituency. All SMD candidates must be a ffiliated with a PR list, but the inverse is not the same. Parties may put forward lists without having any candidates in the SMD tier. The configuration of association must be the same throughout each multimember constituency, but may vary across consti tuencies. Candidates seeking office in the SMD tier may run identifying themselves with different symbols or affiliations (between one and five). This means that if a party or group of parties arranges a stand down agreement within a district, that commo n candidate may identify themselves with the symbols of the party they belong to and (up to five) of each party in 257). The SMD and PR tiers are linked as well. The most important linkage is mechanism of negative vo 3 member majoritarian with instead of compensatory, but the

PAGE 332

332 scorporo was designed to limit the disproportionality of the plurality elections by is that winning parties 2005, 257). The scorporo works in the following manner. Before assigning PR seats, accomplished by sub votes for the second determine the n umber of PR seats awarded to each list. Once the effective votes have been determined, the next step is to ascertain which lists have received more than 4% of PR votes at the national level regardless of the scorporo. All lists crossing the threshold wil l be awarded seats. The scorporo is not the only thing that connects the two tiers of the electoral system. There is also the provision that awards list seats, under certain circumstances, possibility when a candidate who lost his SMD election may be awarded one of the PR seats won by the list that they are affiliated with. This might occur when the list wins more seats than the number of candidates it forwards on t he ballot. Lists are allowed a number of candidates no more than one third of the total seats assigned in a constituency. The largest constituency contains 11 seats, which means that no list have more than 4 names associated with them. If a list in a pa rticular constituency wins more

PAGE 333

333 that list will be awarded the PR seats that could not be assigned to PR candidates The elections for the Senate ar e conducted under a different set of rules. In the Senate, the SMD and PR tiers are not separate (or parallel), instead they are merged together. The Chamber has two separate ballots, but in the Senate elections the only candidates are those running in t candidates and the 83 remaining PR seats are allocated by a repechage mechanism. losers among the SMD candidates. There are eighteen regional PR constituencies with a district magnitude ranging from one to twelve. Candidates may run alone or affiliated with one label and all candidates affiliated with the groups needing to have the same symbol. Groups are formed by at least three SMD candidates running in the same constituency under the same label. Only groups are allocated PR seats. There is a scorporo that operates in the Senate elections as well. It is calculated by subtracting votes garnered by those candidates associated to the Parties and voters face different electoral rules and therefore have different options and strategies available to them. The Se nate elections present the greatest challenge to smaller parties, since there is not separate PR tier. Smaller parties are forced to run candidates in all SMD districts, even in those the parties have no chance of winning, in order to collect the votes ne cessary to win PR seats. In the Chamber a party does not have to run in every district, it can join a coalition or enter into a stand down agreement with other parties while maintaining viability in each SMD thanks to the

PAGE 334

334 PR tier. In the Senate this opti on is not available. The same can be said for the options of voters. Since the Chamber has two ballots, it is possible that the voter to split their ballot. They might vote strategically on the SMD portion of the ballot and sincerely on the PR section. Voters do not have that option when voting in the Senate. There is only one vote cast, so voters have a decision to make. They can vote strategically for or against a favored or disliked party that has a chance of winning, or they can vote sincerely for a smaller party with no chance of winning the SMD tier with the hope that The previous paragraph is only a quick summary of some of the potential consequences that the MMM system could have on voters and parties. It is a subject I return to when discussing the effect that the new rules had on the party system. Before I detail those results, it is important to take a look at the expectations that policy makers and scholars had leading i nto the reforms. I hold that these expectations make Italy a good case for evaluating the efforts to increase majoritarianism. Because of this, I use I outline in Chapter 2. Expectations were optimistic declared that Italy was beginning a Second Republic (Pasquino 1994). The thought was that Italy was at a crossroads and about to e mbark on a clear departure from the previous fifty years (the aptly named First Republic). Mershon and Pasquino put made this point succinctly (1995, 42), parties govern, and which insti models of governance contribute to a fourth: a redefinition of the

PAGE 335

335 sociopolitical system, that is, a change in the nature of the relationships between civil society, on the one hand, and political forces and inst itutions, on the other. features of the new system. One was to improve the excessive proportionality that characterized proportional elections held during the First Republic. The thought was that the party system was plagued with a multitude of political parties. The electoral system change, it was hoped, would reduce the number of political parties and create a system of bipolar competition. This type of competiti first time and end the instability that characterized the First Republic. By reducing the number of political parties, creating bipolar competition, and increasing stability of government s scholars believed that effective legislative programs could be debated during elections and passed through parliament more effectively and expeditiously than in the past (Katz 2001; Newell and Bull 1997; Sakamoto 1999 ). This is the type of outcome that the majoritarian model expects 4 substantial SMD tier, but was not purely majoritarian. Since Italy had a mixed system, it may not be the best test of the majoritarian model 5 Some scholars suggested at the time tha t the reduction of political parties and creation of bipolar competition through alternation in clear alternatives was not likely. This was due to the high level of fragmentation that existed in the country as well as the tendency 4 See Chapter 2 for political consequences of electoral systems. 5 The majoritarian model does not need completely majoritiarian elections, since the first step in the mode the model would be an appropriate fit. For more detail on the role of mixed systems and their relation to both the majoritarian and proportional mod els see Chapter 2.

PAGE 336

336 of the mixed rules to cr eate unexpected outcomes (Moser 1999, 2001; Katz 2001). Italy had a history of fragmentation and was going through a massive reordering of the party system. There was a fear at the time that these conditions would not allow the SMD tier to reduce the num parties and voters would have to adjust and learn about the effects of the new rules. Smaller parti es could use that to their advantage by blackmailing larger parties into standing aside in safe seats out of fear of losing the seat altogether. The result would reason contamination effects that could come from having both SMD and PR tiers in the same election. An example of how this could happen is if third parties, in an effort to maximize th eir vote share in the PR tier, would ignore Duvergerian pressures and run and Schoppa 2002). Larger parties, worried about losing precious votes to small parties, might en ter into stand down agreements as a result. These two separate causes, the newness of the party system and electoral system or contamination effects, would cause the same thing: no reduction in the number of political parties. The potential contamination effects described above and the potential for voters to split their vote really only applies to the elections for the Chamber. Elections for the Chamber have two ballots, so the ability of voters to split their vote offers them the opportunity to vote str ategically and sincerely at the same time. With the Senate elections there is only one ballot so voters and parties face different choices. Smaller

PAGE 337

337 parties, if they want to take advantage of the PR component of the electoral system, have to run in as man the PR tier. This is different than the logic in the Chamber elections where running in separate PR b allot. Both large and small parties face a similar decision however. They can try to reach stand down agreements, where large parties cede some safe seats to ideologically similar small parties, or they can risk the small parties bleeding off valuable vo tes and lose the seat to an ideologically different party. Voters in Senate elections face different choices as well. There is only one vote for a candidate, so voters may cast a strategic vote or a since vote. Voters have to make a decision to vote for one of the largest parties with a chance of winning the SMD seat or they can vote sincerely for a smaller, more ideologically compatible party with the hopes that the party will acquire enough votes nationally to qualify for PR seats. These alternate cho ices, especially if parties choose stand down agreements, might also lead to a fragmentation of the party system and therefore problems for the majoritarian model. If the numbers of political parties are not reduced or if, at the very least, no bi po lar competition emerges then the majoritarian model will be a failure in explaining any behavior in Italy. It is difficult to see how a fragmented party system could produce cohesive, accountable governments that are able to pass comprehensive or effectiv e policy agendas. Without those improvements in policy and process, support would not improve among the population. Beyond these concerns, Italy was also facing a particular political context that needed address by whatever governments were elected under the new system, regardless of their composition. C hapter 4 argues that the

PAGE 338

338 electoral system changes took place for particular policy and process reasons It was the rise in corruption and the failures of governments to deal with the pressing policy dema nds of the day (rise in public debt and pressure from Europe) so citizens lost support for the existing electoral order because of those changes. The success or failures of the change to MMM were going to be judged by how the changes facilitate improvemen ts in those areas as well The remainder of Chapter 6 deals with that question. Creating a bipolar system of competition in Italy, with less political parties, and more stable governments is an important part of the electoral system change. But, Italians will only judge the changes successful (evidenced by an increase in support) if the corruption is weeded out and corrected and if policy making improves. If these changes occur I expect support to rise. Chapter 3 discusses how a majoritarian electoral s ystem could conceivably be used to manufacture more support. The following discusses whether Italy follows those steps. Direct Effects of MMM in Italy, 1994 2005 As a summary, the old proportional electoral system was replaced for several reasons. The pr oportionality of the old electoral system led to an increase in the number of parties, allowing smaller, third parties outsized influence. This made forming and maintaining cohesive governments near impossible. Exacerbating this fact was the presence of anti system parties barred from government, which forced the Christian Democrats to be a part of every post war coalition. Without stable governments or any means of alternating power, parties in Italy were unable to pass coherent policy programs and were forced to look to alternative means to maintain support. Their main means of doing so was through a vast, complex network of clientelistic relationships

PAGE 339

339 which led eventually to fu ll blown corruption in the 1980 s. A mixed member majoritiarian system was adopted in 1994 in order to correct some of these issues. The goal was to create a system of bipolar competition by either eliminating or forcing smaller parties to join with larger parties through more majoritiarian rules. With small parties not holding government hostage, the hope was that stability could be increased and more coherent legislative programs could be passed. Since parties would be judged on their legislative achievements rather than their ability to deliver in patron/client relations, red ucing the practice of corruption. The expectation of this new more majoritarian electoral system closely mirrors the majoritarian model that I draw in C hapter 2 This model states that as proportionality is decreased, the mechanical and resulting psycholo gical effects act in conjunction to reduce the number of political parties by eliminating smaller parties or forcing parties into pre election coalitions. When the number of parties is reduced, voters have a clear choice between governments that present c learly defined programs. The voters then judge the governments on their success or failure and either reelect them or vote them out of office. This is the process of accountability (which was decidedly lacking in First Republic Italy). Governments are f ree to act on the programs since they have a mandate from the public and this decisive action leads to better policies. The combination of accountability and better policy leads to higher levels of support. Italy adopted more majoritarian rules, so what is left is to determine whether they produced the desired results. I argue he re, and previously that electoral systems have both direct and indirect effects. This section deals with the impact the new electoral system had on the party system. To foresh adow, the results are mixed. The electoral

PAGE 340

340 system was able to create pressure on parties to form pre election coalitions and did establish bipolar competition. In this respect the electoral system performed its function. But, ultimately the number of po litical parties was not reduced. Thi s has to do with the nature of the changes in the party system as well as the strategic choices made by parties Elections under MMM The first elections under t he new MMM rules introduced Italian voters to some new poli tical parties as well as some reconstituted parties salvaged from the wreckage of the Mani Pulite investigations that ensnared many members of the parliament elected in 1992. For the first time, in 1994, Italians had the opportunity to choose between dif ferent coalitions of parties on the center left and center right. The next three elections held under MMM (the electoral system was replaced with a proportional bonus system in 2006) would be contested by varying coalitions of the center left and center r ight. This section includes descriptions of those coalitions and a report of how they faired in the elections of 1994, 1996, and 2001. The center right coalition was led by Silvio Berlusconi, the populist head of a vast business empire that included the o wnership of television networks and the soccer club the oft repeated cheer heard at AC Milan soccer games. Berlusconi presented his new party, and himself, as e ntrepreneurial and uniquely suited to create jobs and fix the economy (McCarthy 2001, 163 of specific policy platforms (Gilbert 1995, 170). What Berlusconi proved particular adept at was bringing to gether two divergent parties on the right, with whom he formed two separate alliances, as well as part of the r econstituted DC. In the North, FI joined the

PAGE 341

341 Northern League to form the Liberty Pole. In the center of the country they allied with the recons tituted Fascist (MSI) party n ow called the National Alliance (AN) to create the Pole of Good Government. The last party allied with the FI was the Christian Democratic Center (CCD) who represented the right wing of the old DC 6 This center right coalition was opposed, in 1994, by a fragmented amalgam of parties of the center left. The two largest parties in this coa lition, called the Alliance of Progressives in 1994, were the Democratic Party of the Left (PDS) and the Communist Refoundation Party (RC). T he PDS was the wing of the former Communist Party (PCI) that repudiated its communist past by giving up its search for major structural changes. The RC was less willing to give up its communist roots. The coalition was joined by former members of the Ita lian Socialist Party (PSI), who split to form the PSI and the Socialist Renewal (RS), and the Republican Party (PRI) now called the Democratic Alliance (AD). The PDS as the largest party was faced with a difficult job. The RC wanted an alliance that excl uded the center parties like the AD, while the AD wanted a coalition of the center that left out the RC. The PDS tried to solve this problem by creating the largest coalition possible and in the end include d a multitude of parties: the PDS, PSI, RS, AD, R C, CS (Social Christians, made up of former DC and PCI), the Greens, and the Rete (a new par ty formed in 1992) ( Bull 1996; Newell 2000 ; Rhodes 1994 ). The election of 1994 was contested between the Pole of Freedom/Pole of Good government on the center right and the Alliance of Progressives on the center left. The 6 Pre electoral allegiances in this case refers to agreements over pre election bargains over what single member district seats the various members of the coalition would run in, thereby maximizing the single member seats in both house s of parliament (Bull and Newell 2005, 52; McCarthy 1996, 142).

PAGE 342

342 Italian Popular Party and Pact for Italy (a group of former DC coalesced around Mario Segni), the two largest DC splinter parties, ran in the center as the Pact for National Renewal chose not to al ly itself with either of the coalitions. The results were a resounding win for the center right coalition who gained a majority of seats in the Chamber of Deputies and a plurality in the Senate. In the single member district elections for the Chamber of Deputies, the center right parties together garnered 46.1% of the vote and 302 seats. In the proportional ballot FI won 21% of the vote and 30 seats, the AN gained 13.5% and 23 seats, and the Northern League came in at 8.4% and 11 seats which gave the coa lition another 64 seats for a total of 366. On the left, the Alliance of Progressives totaled 33% of the single member district votes for 164 seats. Only two parties of the Alliance were able to gain list seats, the PDS with 20.4% of the proportional vot e for 38 seats and the RC gained 6.1% and 11 seats. All total the Alliance won 213 seats in the Chamber of Deputies. The Pact for Italy, as the lone centrist coalition, won 15.6% of the total SMD vote for 4 seats. The PPI won 29 seats from 11.1% of the vote and the Pact for National Renewal picked up 13.7% of the vote and 13 seats. For a complete breakdown of election results in the Chamber see T ab le 6 1. The center right coalition had similar success winning a plurality in the Senate. The coalition w on a total of 42.5% of the total vote and 128 seats in the SMD tiers with an additional 28 seats in the PR tier. The Alliance of progressives garnered 33.4% of the total SMD vote and 97 seats, while adding 26 seats in the PR tier. The only other coalitio n to win seats was the Pact for Italy. The coalitions picked up 16.7% of the vote for only 3 total seats, with an additional 28 coming in the PR tier. Complete Senate

PAGE 343

343 results are displayed in Table 6 2. The failure of the PPI in the Chamber and the Sena te to win enough seats to become a decisive center block caused a dissolution with the resulting components joining the center left and center right. The new center right government formed under Berlusconi only lasted 8 months, and after 14 months of a tec hnical government, new elections were called in 1996 7 After losing the previous election the Alliance of Progressives made a concerted effort to absorb the center leaning parties at the expense of the RC 8 and formed a new coalition called the Olive Tree This coalition was led by Romano Prodi, the former head of the state holding group IRI, and had a reputation of honesty among Italians (Newell 2000). The Olive Tree was made up of 4 main components. The largest component was the PDS, who made up appro candidates in the 1996 election 9 The second major component was a group of center parties, the largest being the PPI. Third, was a group organized around the most recent Prime Minister Dini, and finally the Greens round ed out the coalition. The Olive Tree had particular historic important for the allegiance between the left (PDS) and the center (PPI). The two parties came together, not out of ideological agreement, but an understanding that electorally, one could not h ope to win office without the other (Bull and Newell 2005, 53). This type of mutually beneficial agreement was not evident on 7 I detail this collapse when I discuss the stability of governments in later in the chapter. 8 ree which gave them a free run at 27 seats in the Chamber and 17 in the Senate in exchange for not placing candidates where they might hurt the Olive Tree (Newell 2000, 38). 9 The PDS also allied with the European Left a group of 6 parties who were also kn who were close to the PDS but were not willing to be completely absorbed by it (Newell 2000, 37).

PAGE 344

344 the center right, however and they put forward a fractured coalition for the 1996 elections. In the 1994 election Silvio Berlusco ni was able to cobble together (albeit briefly following the election) a coalition between the AN and the Northern League. But, the League was buoyed by its show of support in local elections and decided to forgo any pre election agreements with the cente r (Ignazi 1997) 10 The Pole of Freedoms also suffered by a split in National Alliance. is was the symbol of the party). When the party congress approved his recommendations a splinter group calling itself the Tricoloured F lame left the party (Newell 2000 35). The two major defections of the Northern League and the Tricoloured Flame hobble d the Pole of Freedoms in the 1996 election. The Northern League used its popularity and took valuable votes away from FI in the North. The Tricoloured Flame did not garner much support country wide (about 1.7%) but by running in several single member di stricts in the southern and central districts was able to take away valuable support from the AN. These percentages contributed to the Pole of Freedoms defeat in as many as 52 center south single member districts (Ignazi 1997 422). The fractured center r ight helped provide a small plurality victory for the Olive Tree in both the Chamber of Deputies and the Senate In the 1996 Chamber election the Olive Tree won 45.7% of the FPP vote for 247 seats. All total in the center left parties 10 It should also be noted that the Northern League was a party that championed the devolution of power to the regions and pushed for inc reased federalism, especially in Milan their base of support. This put them in direct contrast with the Rome centralization (Newell 2000).

PAGE 345

345 took 34.8% of the p roportional vote for an additional 38 seats for a total of 285 seats. The Pole of Freedoms lost by a close percentage in the FPP section of the election with 43.2% (a difference of 2.5%), but were defeated by a margin of 78, claiming only 169 seats. The Pole had much greater success on the proportional ballot, trouncing the Olive Tree by 7.4% with 42.1%. But, because of the nature of the electoral system with only 25% of seats being awarded by proportional vote, the Pole of Freedoms only picked up 77 sea ts. This left them behind the Olive Tree 246 to 285. The unaligned parties the Northern League (59 seats) and the Communist Refoundation (35 seats) also picked up considerable support. For a summary of results se e T able 6 1 With the support of the RC, the Olive Tree formed a government that lasted in various forms until the parliamentary term was up in 2001 In the Senate, the Olive Tree was able to win a plurality as well. They acquired tional 23 seats in the PR tier. The House of Freedoms polled 37.3% of the vote for 67 SMD seats and 49 PR seats. The Northern League had marginal success in the Northern parts of the country with 10.4% of the total SMD vote and 18 seats with an additiona l 9 PR seats. The Communist Refoundation was the only other group that acquired substantial nation wide support with 2.9% of the overall vote for 10 seats with no PR seats to add to their total. With their support, the Olive Tree was able to cobble toget her a majority in the Senate to go along with its majority in the Chamber. For complete Senate results see Table 6 2. The Olive Tree, through its own shortcomings, was not in a good position to defend its position as the government leading coalition going into the elections of 2001.

PAGE 346

346 The Prodi government collapsed in 1998 when the RC refused to support the government over matters of policy (Pasquino 2002). The RC entered the opposition and decided not to enter into any pre election agreement with the Oliv e Tree, reversing its position from 1996. The Olive Tree also failed to recruit a new list of candidates who rallied around former Mani Pulite prosecutor Antonio Di Pietro who called themselves the Di Pietro List The coaltion still maintained the suppor t from several former PSI and and a small splinter faction of the RC, the Party of Italian Communists, rounded out the coalition. The center right, on the other hand, recon stituted its 1994 coalition. They presented a unified list to the voters behind the leadership of Silvio Berlusconi that included the Northern League whose poor showing in the recent regional elections brought them back into the House of Freedoms. The ce nter right also included Northern League and a small group of center catholic parties. This united list won a tremendous electoral victory in the 2001 election (Ignazi 2001; Pasquino 2002). Like the 1996 election, the FPP section of the electoral system was closely contested between the two coalitions. The aggregated difference between the two was only 2.1% (House of Freedoms with 45.5% and the Olive Tree with 43.5%), but the total seat difference was 93, in favor of the House of Freedoms. The Olive Tree had to compete with the Di Pietro list in the single member districts. The opposition group did not gain any seats, but accrued a critical 1,500,000 aggregate votes, which no doubt cost the Olive Tree valuable seats (Pasquino 200 2). In the proportional list, it appeared that the House of Freedoms won a landslide with 49.5% of the vote (64 seats) against

PAGE 347

347 35% (49 seats) for the Olive Tree. The center left coalition had to compete against both the RC and the Di Pietro list who won 9% of the vote between them. The RC won 11 seats through the proportional vote, with the Di Pietro list falling just short of the 4% threshold. Altogether, the House of Freedoms won 368 seats and the Olive Tree 247. For complete results see Table 6 1. Affected by the same electoral burdens, the Olive Tree lost control of the government to the House of Freedoms. The results in the Senate again mirrored those in the Chamber. The House of Freedoms claimed 42.5% of the vote which translated to 152 SMD se ats and 24 PR seats. The Olive Tree lost support from the 1996 elections, down to only 39.2% of the vote and won only 77 SMD seats with 51 PR seats. The Communist Refoundation ran g more seats the party was able to improve its total vote from 2.9% in 1996 to 5% in 2001. This increased vote share kept the party from being shut out completely and it won 4 seats through the PR tier. The combined Chamber and Senate results allowed Sil vio Berlusconi to return to the Premeirship after a seven year absence and this would be the last election contested under MMM. These election results demonstrate some insights into the success o f the models I draw in Chapter 2. The goal of the electora l change was to create bipolar competition that would provide for alteration in office and to presumably reduce the number of political actors 11 Italy clearly reorganized itself into two coalitions on the center left and the center right, but the number o f political parties in the system remained high. The 11 This is a highly contested subject. There is disag reement over whether or not the system itself was designed to actually reduce the number of political parties. For a discussion of this debate see (Bartolini,

PAGE 348

348 following passage details the effect that the new MMM rules had on the party system in Italy and whether the majoritarian model holds in this context. MMM and the Majoritarian Model The most noticeable development in Italy after the electoral system reforms has been the establishment of two competitive, clearly identifiable, pre electoral coalitions that dominated the electoral arena from 1994 to 2001. The bipolarity that has emerged exists alongside hi become a two party system. Italy has become a two coaltion system and while these coalitions have not r eplaced parties, they are more important than parties for winning question is how did this happen? The answer lies in the complex electoral coordination that took place among parties due to the ability of smaller parties to negotiate favorable stand down agreements with the larger parties. The coalition that was best able to negotiate these agreements was the one to win majorities in the Chamber and the Senate. This coordinat ion also suggests that there were limited contamination effects due to the mixed nature of the electoral system. Electoral coordination in Italy happens in the form of negotiated stand down agreements among political parties that unite under a single coali tion banner. This is due to a provision in the electoral law that allows SMD candidates to be listed on the ballots next to the coalition they are allied with such as Olive Tree or House of Freedoms. In elections for the Chamber that candidate may also i ndicate to their district which party they belong to as well, but in the Senate candidates are only allowed to list one affiliation.

PAGE 349

349 T hese arrangements are difficult to make and parties over time developed sophisticated strategies to divvy out seats to parties within the coalition. Key to these agreements was determining both the number and quality of the districts assigned to coalition members. Districts had to be classified by their relative degree of safety then assigned to parties based on their relative strength within the coalition. The goal was to match the percentage of seats each party won in the SMD tier with their percentage of got better at making thes e arrangements. A discussion of this learning process will help to explain both the mechanical effects of the new system as well as the psychological effects that encouraged large parties to form coalitions with small parties. In 1994, there were actuall y four coalitions that won seats in the single member districts and PR tier in both the Chamber and Senate. On the center right FI formed two separate alliances with the Northern League and the AN in the North and Center of the country respectively. The Alliance of Progressives ran nation wide as a center left coalition and the Pact for Italy attempted to create a center pole. The 1994 elections demonstrated that only parties with substantial regional support could afford to remain outside of one of the largest coalitions. For instance, the Pact for Italy carried over 15% wide in both the Chamber and the Senate, but only 4 and 3 seats each in both chambers, with only an additional 31 and 46 coming from the PR tiers. This is a relativ ely insignificant number considering the large assembly size in Italy. The nation wide for the Chamber and Senate running in only in the regions of South Tyrol and won 3 seat s each for both Chambers. This is evidence that the single member

PAGE 350

350 districts were doing their job and that coalitions could not rely on the strategy of running in every SMD to improve their chances in the PR tier (at least in the Chamber, since there was n o second ballot in the Senate). The 1994 elections had several repercussions leading into the 1996 elections. The Pact for Italy disbanded. The newly formed Olive Tree took on many of these former parties, leaving out the RC. The parties did come to a stand down agreement however. In twenty the Senate (6%) the Olive Tree did not run candidates and agreed to support the RC in the remaining districts right had more difficulty in 1996. The Northern League decided that due to its regional strength in the North that it would not enter into stand down agreements with the House of Freedoms (made u p now of FI and the AN as a nationwide coalition). The Northern League did particularly well in northern regions winning 25% in most districts and as high as 30% in certain constituencies. This explains why the Northern League was able to win a substanti al number of seats in the effects of the new electoral rules. The Pact, running nation wide, had more than 6,000,000 votes but could only win 4 SMD seats in the Chamber and 3 in the Senate. The League won only 4,000,000 votes and 39 seats in the Chamber and 18 seats in the pread out nation wide, able to secure a pivotal role in parliament but it did cost the center right several crucial

PAGE 351

351 seats in the North allowing the Olive Tree to pick up seats that it otherwise would have lost against a united center great deal from the first two elections and ran united fronts for the 2001 election. Bipolarity seemed to have taken hold in Italy. Th e two major coalitions increased their share of the votes and seats from the 1994 to the 2001 elections. In 1994, the two coalitions won 80% of votes and 92% of seats in the Chamber and 67% and 85% respectively in the Senate. In 2001, those numbers incre ased to 89% and 98% in the Chamber and 82% and 97% in the Senate. In 2001, only in six districts in the Chamber and three districts in the Senate did either the center right or center left not win the most are treated as parties, looking at the effective number of electoral parties at the SMD district level were 3.07, 2.47, and 2.41 for the 1994, 1996, and 2001 elections. Cox and Schoppa (2002) use this as evidence that a degree of Duvergerian learning took place among voters and the coalitions in Italy. But, the coalition are not parties, but collection of parties and while the existence of single member districts might have provided enough incentive for parties to form pre election coalitions the number o f parties was not reduced. When disaggregated from the coalitions, the number of parties winning seats in the Chamber numbered 20, 14, and 19 in 1994, 1996, and 2001 and in the Senate the numbers were 8, 9, and 9 (Bardi 2007, 723). The fact that coalitio ns developed extensive stand down agreements with parties explains why the number of political parties remains high. The question that remains is why small parties do not run in all the single member districts to boost their share in the PR tiers? This i s especially puzzling since the only way to win PR seats in

PAGE 352

352 the Senate is to run in the SMD tiers. What else needs explanation is why large parties would choose to cede SMD seats to smaller parties in the first place? The expectation that smaller p arties in mixed system would run candidates in all of the single member districts in order to raise their profile in the PR tiers comes from the literature on the contamination effects of mixed system (Cox and Schoppa 2002). These scholars notice that the se systems do not conform to Duvergerian tendencies in the single member districts. Italy shows a distinct lack of an interaction effects between the SMD and PR tier in this regard. Unlike other mixed systems, it is in the best interest of small parties in Italy to coordinate with larger parties rather than go it alone in the SMD tiers. In fact, what they gain in terms of SMD seats is far more than the extra seats they might gain by boosting their performance in the PR tier by running candidates in ever y SMD. For most small parties the chance of reaching the 4% threshold is so remote that electoral coordination is their This does not answer why larger parties would cede seats to the smaller parties in the first place. Larger parties at the head of the center right and center left coalitions had a The first has to do with the blackmail power of small parties. An example of the damage a small party can have is the Tricoloured Flame an AN splinter party that they House of Freedoms refused to sign a stand Chamber and Senate, mainly in the South. Within these districts, the average vote for Flame candidates was 5% for both chambers. As it turned out the Flame had little many districts it cost the

PAGE 353

353 House of Freedoms and estimated 36 seats in the Chamber and 26 in the Senate winning an absolute majority in both chambers. This type of blackmail po tential is enough for the large parties to grant them seats in an effort to win the election, rather than risk losing a district to the opposite coalition. Ultimately it is the voters that gave the small parties blackmail power, since they were still willi hard to explain why this is the case due to the lack of survey data that might help explain why voters made the choices they did, although a number of theories have been forwarde take effect. The fact that voters were coalescing around the two coalitions suggests that some learning was taking place and that if the elections had just been given more time vo A related notion is that the party system was in such flux from the Clean Hands Investigations, that a new party system would have to be consolidated before the effect of majori tarian rules could take hold. In this way, Italy following electoral system change was a lot like post communist democracies. Elections in these countries had unexpected outcomes because of the lack of institutionalization in the party system (Moser 1999 ) Another theory suggests that even though the national elections were held under mixed rules, there were still regional and European elections that were held under proportional rules. This might have created an atmosphere of proportionality where third p arties could form, organize, and consolidate for those elections and

PAGE 354

354 that Italians had a long history of supporting third parties and those traditions would not bed easi ly set aside, despite new electoral rules (Moser 2001). MMM in Italy did provide incentives for parties to join pre election coalitions, but these coalitions contained a fragmented party system within. Coalitions did offer voters a clear choice between alternatives and the opportunity to vote out non performing governments. In that respect, the model is successful. But, the majoritarian model also suggests that the alternatives will be able to govern effectively and pass effective policy. Here the fra gmented nature of the coalitions in Italy had an effect. Parties could generally agree to join pre election coalitions, but when the election was over agreements were not so reached as readily. What follows is a description of the performance of Italian governments specifically in the areas of corruption and policy. I pay particular attention to the Olive Tree government elected in 1996 and the House of Freedoms government elected in 2001. These were the two longest serving governments and their actions and differing levels of success affected the level of support country wide. T he center left government had some success dealing with the issue of corruption. They steered much needed reforms through parliament and as a result, Italians felt that the lev el of corruption was falling and support for the system was improved. This government was also aided by the Clean Hands investigations and the high levels of parliamentary turnover that took place in 1994 and 1996. By 2001, a new political class began to emerge. In the same year the center right Berlusconi government took power. Corruption legislation was almost immediately taken off the table and policies promised by the new Prime Minister were not enacted. As a result,

PAGE 355

355 Italians increasingly felt tha t corruption was a major issue after 2001. When this was for the system fell. This section describes the changes, successes, and failures that took place following electoral reform in 1994. Is it Still a Corrupt System? Italians eliefs about Corruption 199 4 2005 One of the driving forces behind electoral system reform in Italy was the expl osion of corruption in the 1980 s. I make the claim is that in order for the new electoral system to be deemed a success, Italians would have to view corruption as being reduced or dealt with by their elected officials. After the new electoral system was adopted in 1994, view s about corruption were affected in my view, by two th ings. First, the Clean Hands investigations decimated the existing political class and there was a tremendous amount of parliamentary turnover in both 1994 and 1996. Also, the Olive Tree made a concerted effort to attack the issue of corruption through a series of policy initiatives. Together these contributed to Italians feeling that their system was less corrupt. Following the election of 2001, and the return of Berlusconi to the Premiership, efforts to curb corruption stalled. Feelings about the nat well. This section describes the rise and fall of these feelings. Mani Pulite, New Parties, and New Members The Clean Hands investigations had a tremendous impact on the existing political class in Italy During the 1990 s there was a sharp rise in the number of corruption related prosecutions. Between 1984 and 1991 average number of crimes was 252 with 352 people charged in these crimes. From 1992 to 1995 that number more than tripled with 1,095 separate charges were brought, accusing 2,084 peopl e with crimes (Della Porta 2001, 12). In the 1990 s the number of people sentenced for crimes also

PAGE 356

356 multiplied: 159 in 1991; 185 in 1992; 263 in 1993; 369 in 1994; 549 in 1995; t o 856 in 1996 (Della Porta 2001, 12). By 19 98, the total number of people investigated was 4,000 with 2,970 requests by prosecutors to the investigating judges (Della Porta 2001 12). These investigations included many giants of the Italian political class including: Bettino Craxi, the leader of t he PSI and a former President of the Council of Ministers, Giulio Andreotti a DC leader and former Prime Minister, Georgio La Malfa the leader of the Republican Party, Arnaldo Forlani a DC leader and former Prime Minister, Gianni De Michelis a member of th e PSI and a former Minister for Foreign Affairs, as well as many others (Gilbert 1995). Because of this, Italian parties had undergone various more or less radical transformations. Some had undergone a major facelift; others were in crisis and their tran sformations had already begun; still others had split up or disintegrated and virtually disappeared; [and] brand new parties had also been established. (Morlino 1996, 6) The party system that emerged after electoral system change was a reorganized edition of the one that dominated election after election prior to 1994. There were two predominant themes in relation to the political parties contesting the 1994 election under new electoral rules: the reorganization and fracture of the parties associated with political parties. The party that perhaps represented the old system was the Christian Democrats that had been a part of every post war government. The DC was not on the ballot in 1994, a fairly dramatic circumstance in and of itself, having spit into several pieces. The largest schism formed the new PPI made up of those in the center, but the party also lost members on the left and right. On the right, some members joined the new Forza Italia, others the AN, still others formed the new Christian Democratic Center and joined with the Pole of Freedoms for the 1994 election. On the left, some left to

PAGE 357

357 follow a former DC leader Mario Segni, others departed to form La Rete, an anti mafia party. The DC was not the only major party to go through a drastic reordering. The other two dominant parties during the First Republic, the PCI and the PSI, were also dismantled and dispersed (Morlino 1996). The PSI was also impacted by its ties to the Christian De mocrats and the scandals of the Clean Hands investigations. Its leader Bettino Craxi was the face of the c orrupt politician who represented the First Republic. The PSI did not have a main successor party like the Christian Democrats (to the PPI), they sp lit into several small parties who had little electoral impact, with other leaders joining the Democratic Alliance on the left or Forza Italia on the right. The party that was least touched by the scandals was the PCI, it decided in 1991 to soften its vie ws on how they would attempt to restructure the state apparatus. The main force of those behind this softening created the Democratic Party of the Left (PDS). Others, dissatisfied with the break from the communist roots, joined instead a new party the Co mmunist Refoundation (Morlino 1996). The three main political parties that represented the First Republic, the DC, the PCI, and the PSI, were now gone replaced or absorbed by other groups. The center right was represented in this change as well with two new parties, Forza Italia and the Northern League, and a restructured Fascist Party, the National Alliance. Silvio Berlusconi formed the Forza Italia in late 1993 in the run up to the election in 1994. His stated goals were to create a pro business Italy and save the country fr om an inevitable communist take recruited mainly from the business and financial community, took advantage of the vacuum left by the dissolution of the DC, PCI, and PSI. Forza Ita lia was joined in the

PAGE 358

358 new center right coalition by the Northern League. Fronted by their charismatic leader rn regions of the country. It support ed more devolution to the regions as a re action to the clientelistic relationships between the center of the county and the south that represented the post war system. Finally, the center right was joined by a refashioned MSI, whose leader ist roots. When Italians went to the polls in 1994, they were presented with a series of new choices. These choices would not be real or important if the new party system were merely a reshuffling of old players, who put down their old banners, picked u p new ones, and simply carried on with business as usual. This is not what occurred. While there was some reorganization, many leaders from the past were simply unable to carry on because of their implication in the corruption investigations that were st ill ongoing. Many of these elites were simply no longer to carry on as elected officials, and were replaced en masse, starting in 1994. In 1994, there was a sweeping renewal of parliamentary personnel. During the elec tions held in 1992 more than one th ird of the members of the Chamber of Deputies had already served three or more terms in the parliament and the overall turnover rate was close to 44%. In the elections of 1994, those numbers were reduced to 12% of or more terms a nd over 71% who had no parliamenta ry experience at all (Katz 1996, 47). The number of newly elected officials wa s greatest among the parties that had no previous connection with the old political system. Forza Italia had the highest such amounts with 90. 5% newly elected in the Chamber of Deputies and 94.4% in the Senate (Verzichelli 1997 149). Over 80% of these members

PAGE 359

359 from Forza Italia report that they had no experience in electoral politics befor e the 1994 elections (Katz 1996, 47). Overall, 30% of M political experience (Katz 1996, 47). The Allenza Nationale and Northern League, parties that also made large gains, contributed to the high level of parliamentary turnover as well. The AN had 74.5% new members in the Chamber and 6 4.6% in the Senate, while the Northern League had 69.5% and 64.6% respectively (Verzichelli 1997, 149). Even the parties mostly closely associated with the old political system the PDS, the PPI, and RC had over 50% new members in both the Chamber and Sena te (Katz 1996, 47). A changed parliamentary class emerged from the election of 1996 as well. Although not as high as the breaks that took place during the 1994 election, the percentage of parliamentarians who were elected for the first time remained high. The parties that experienced the highest levels of turnover in 1994 did the same in 1996. The FI (47.2% in the Chamber and 58.3% in the Senate), the Northern League (50.8% in the Chamber and 48.1% in the Senate), and the AN (32.3% in the Chamber and 32. 6 % in the Senate) suffered losses in the election, but sent large number s of new Members to the legislature. On the left, parties returned more parliamentarians from the 1994 election. The RC experienced some of the lowest levels of turnover, with only 9 .1% with new members in the Senate, but a higher number, 40.0% in the Chamber. The winning coalition Olive Tree experienced smaller levels, but still significant. The PDS had 39.9% new members in the Chamber and 27.6% in the Senate. The PPI experienced high turnover in the Chamber (70.1%) as well as the Senate (40.7%). These numbers compared quite favorably with past parliamentary turnover, where numbers seldom reached 30%. In 1979 the number of new deputies made up 27%,

PAGE 360

360 32.1% in 1983, and 28.2% in 198 7 (Morlino 1996, 17). Despite these figures, the fact that the winning coalition experienced low levels of turnover led some to argue that the reorganization of t he new parliament had stalled. One of the critiques leveled by scholars is that the 1994 ele ction was an aberration returning under new banner s in 1996 ( Della Porta and Vannucci 2007 ; Newell 2000 ). The concerns of these scholars are perhaps no t without reason. I n the past crises of Italian politics were rarely accompanied with any significant change. Governments collapsed, but were just followed by a reorganization of existing personnel. This was not the case in the elections of 1994 and 1996. In these case s th ere was evidence of actual l number of p arliamentarians returning to the legislature was 40, with 23 in the Chamber of Deputies and 17 in the Senate. Of the largest numbers of returnees: 12 of these were from the PDS, 7 from the PPI, 4 from RC, and 5 from CCD CDU (Verzichell 1997, 147 148). More evidence that these feelings were overstated is evident after an examination of the length of parliamentary experience. Overall the l evel of experience electe d before 1990 (Verzichelli 1997, parliamentarians is evident in the p ercentages of those serving in thr ee or more Deputies served in the 3 or more legislatures, while the number in the Senate was 7.9%. In 1996, there were marginal increases, where 8.9% (Chambe r) and 9.5% ( Senate) served in three or more legislatures. The average seniority of (in legislatures)

PAGE 361

361 is also small, with an average of 1.6 and 1.7 in the Chamber and Senate in 1994. These were only increased slightly in 1996 with 1.9 and 2.0, respectively. See Tabl es 6 3 and 6 4 for results To summarize, the legislatures that were elected in 1994 and 1996 were different than those that came before them. The Clean Hands investigations essential destroyed the existing political class. Political parties broke apart, were snuffed out of existence, or were replaced by new parties. The individuals who were giants of the First Republic were embroiled in these investigations as well. As more and more information was released to the public, leading politicians were now u nder investigation, indicted, or merely embarrassed by what was learned. Consequently, many did not stand for who did stand for election wer e voted out at a high level. Because of this, parliamentary turnover was very high in both 1994 and 1996 compared to historical embers came down. hat views about the level of corruption in the country improved. This newness was also coupled with governmental efforts to reduce the levels of corru ption. By 2001, however a new political class began to emerge and this emergence of a new professionali zed might have had an effect on how Italians viewed the levels of corruption in their country. The 2001 election brought a halt to the rapid change in the make up of the parliamentary class. One indicator of this was the decline in the numbers of parliame ntarians elected for the first time. 2001 marked the third straight election where the number of first time parliamentarians fell. Another indication of the

PAGE 362

362 institutionalization of a new political class was length of seniority. The level of seniority fo (Verzichelli and Zucchini 2002, 223). The increase in the average duration of p when the entrepreneurial behavior of some politicians is considered. Overall, a re professionalization of the parliament has certainly taken hold. The percentages of political actors with a prior background in politics had returned to pre 1992 levels. s (Verzichelli and Zucchini 2002, 225). better chance of winning seats. In 2001 60 incumbent deputies and 28 incumbent senators switched districts. In both cases, 60% for the Chamber in 1996 and ran in an SMD for the Senate in 2001 or vice versa. 45 deputies ran for Senate seats and 73% won and 62% of the 34 Senators who ran in Sen ate is the effect I believe it will have on feelings toward the level of corruption among the population. The First Republic was characterized by an entrenched political class that used its political connections to enrich themselves. Italians might feel more

PAGE 363

363 inclined to feel that the level of corruption was going down while judicial investigations and new elections removed those political elite from power and replaced them with something new. By 2001, there were signs that what was once new was now becom ing familiar. Efforts at legislation to deal with corruption had marginal success from 1994 to 2001, but might have been masked by the changes that were occurring in terms of the replacement of the old political class. But, when the Berlusconi led govern ment of 2001 took office, the political climate was very different. I just offer this as a backdrop to the discussion of efforts of both the center right and center left to deal with corruption. Dealing with Corruption: The Center Right and Center Left T he center right and center left coalitions both got a chance, after the electoral lived government passed no bills preventing either corruption or conflicts of interests. While his government did not have much of an impact, the judicial system filled the void, ramping up the corruption investigations including those involving the Prime Minister himself. In 1996, after one year of technical government, the Olive Tree coalition took power with a renewed vigor to deal with the p roblem of corruption. While its legislative achievements w ere not sweeping in nature, it did pass several important reforms that would help make the operation of public offi ces and bodies more open. The coalit further efforts to pass conflict of interest legislation died out as the 2001 election approached. But, the efforts toward resolving the problem of corruption coupled with the increasing number of cases prosecuted and investigated by the judiciary i mproved the way that Italians felt about the level of corruption in their country. When the center right regained control of both houses, corruption legislation went by the wayside. The momentum of the Clean Hands investigation was lost and public attitu des soon turned

PAGE 364

364 pessimistic on the issue of corruption. This section deals with these issues and concludes with a discussion of the rise and fall of public feelings about the level of corruption in the country. When the new electoral system was adopted i n 1994, changes in the party system were already underway. The elections produced a new central political party and politician, Silvio Berlusconi. But, the Forza Italia led center to deal with and combat the issue of corruption was put to the te st immediately. Berlusconi raise d a business empire in the 1980 s based, in large part, on public contracts awarded based on his close, personal relationship with the then Prime Minister Bettino Craxi. During the election, there was a ce rtain irony to the leader of the Forza Italia railing against the actions of the parties during the First Republic considering his media empire was built in the midst of all the activities he now opposed. candidacy sparked questions about whether he was running for office out of a genuine commitment to improving the country or protecting his own personal, financial interests (Bull and Newell 2005). Berlusconi brought to the Prime Ministership a serious con flict of interest. Both while he was seeking office and when he took office, he controlled Finivest, a financial holding company, made up several pieces. It composed a number of companies include insurance and banking, a publishing house, A.C. Milan, and most importantly Mediaset. This company was (and still is) the largest private entertainment company in Italy. It owns three channels: Canale 5, Italia 1, and Rete 4, as well as channels in Spain, Endemol (a distribution company operating in 23 states), and several other companies related to TV broadcasting (Hibb e rd 2007). All total, he owned 90% of all

PAGE 365

365 private television networks in the country, with 43% of all Italians as viewe rs (Bull and Newell 2005, 17). Few observers believed that Berlusconi woul d refrain from using this massive media empire to his advantage. Studies demonstrate the Forza Italia used different, targeted media strategies on his various networks to appeal to the different audiences that tuned in to those channels. The same studies also demonstrate that Forza Italia received overwhelming positive coverage as well as air time on several of his channels as well (Hibberd 2007). Given this evidence, it is understandable how some commentators were concerned about Berlusconi becoming Pri me Minister. Shortly after winning the election, Berlusconi offered several proposals to deal with conflicts of interest. These included: requiring politicians to sell their businesses when elected to office or to transfer control of their holdings to a third party. Many of their proposals were adopted in a legislative proposal presented by the Berlusconi government in 1994. This proposal was approved in the Senate, but not approved in the Chamber of Deputies due to the dissolution of government later t hat year (Della Porta and Vannucci 2007, 838) Passage of this resolution would have prevented people from simultaneously holding government office when they held controlling interests in public bodies or businesses. This was the one serious effor t made by the first Berlusconi G overnment made to prevent conflicts of int erest, but by the end of their eight months in office it had not enacted any substantive legislation designed to prevent corruption. The failure of the government to act and actually stren gthened the resolve of the judiciary and investigating magistrates. Many of these investigations involved the business activities of the Prime Minister or those close to him. These inquires discovered 330 million lira in

PAGE 366

366 bribes paid to the Financial Poli (Della Porta and Vannucci 2007, authorized the payments, but declared that he was extorted. As 1994 drew to a close, Berlusconi was informed that he was being investigated for corruption for his activities involving his companies Videotime, Mondadori, and Mediolanum. He tried to stave off the ensuing public outcry by blaming the whole affair on a conspiracy of communist judges. But, his government s oon fell when an agreement was reached between the PPI and the center left parties. The Northern League withdrew its support for the government and Berlusconi was ousted at the end of 1994 (Della Porta and Vannucci 2007). Lamberto Dini lasted a year before calling elections. That time period was notable, in a legislative sense, for not dealing at all with the problem of corruption 12 In 1996, the Olive Tree won a close election over 13 Led by Romano Prodi, a product of the First Republic known for his honesty, the center left immediately place corruption legislation back on the agenda. Less than two weeks after the election, the government ed the release of some 500 mafiosi accused action, but the ef forts to curb corruption nation wide were often difficult to realize. The 12 Wh ile the Dini government did not deal at all with the issue of corruption, there was more activity on the judicial front, especially in regards to the continuing Northern League troubles of Silvio Berlusconi. During 1995, he was charged with a series of cr imes related to alleged corrupt activities including false accounting at Finivest, corruption of officials in the Ministry of Finance, fiscal fraud and the accounting of land, and ilNorthern Leaguel campaign funding of the PSI. (Della Porta and Vannucci 2 007, 839) 13 It was in this election that the Northern League ran as a separate party, refusing pre election coalitions with the center right.

PAGE 367

367 Olive Tree government did organize a p arliamentary commission to inquire into corruption. The commission met for several months, often arguing over whether abuses by the judic iary should be included in its study, and finally presented legislative text to the Chamber of Deputies. This legisla tion never passed, but it demonstrated that the Olive Tree was serious about spen ding time on the issue, and it would reaffirm this commitment over their 5 year term. Since 1994 when Berlusconi first gained the office of Prime Minister the issue regarding the Chamber of Deputies to come up with some solutions for conflicts of interest was approved in April, 1998. The group created by the bill affirmed a number of issues a nd designed legi slation to deal with them. It first endorsed the idea that there were certain incompatibilities between private, outside activities and holding public office. Their solution included a number of measures. Main among them were provisions requiring major assets to be placed in a blind trust, a strict timetable for setting up this trust, the selection of a trustee from a list of registered individuals approved by the Chamber of Deputies, and severe fines or the removal of state concessions i f the law was not followed (Hine 2002, 266 267). The bill immediately passed the Chamber but languished in the Senate for 2 years. Efforts to resurrect the bill failed in 2000, and the elections of 2001 were held without major legislation passed dealing with corruption directly. That is not to say that there were no reforms enacted that made corruption less likely. Many of their substantive gains were indirect, and were directed at improving the inefficient public bureaucracy where a lot of corrupti on was born. As I write in Chapter

PAGE 368

368 4, the public administration was filled with red tape, and it was a very difficult process to get a permit. In a context where permits were required for almost everything, people looked to speed up the process in a numb er of ways. There were recommendations given by politicians, who would be reelected based on the number of favors they could way for public transactions. During the 19 80 s, the exchange of money for services that were otherwise universal became so egregious, they helped launch the Clean Hands investigations and bring about electoral system change. By reforming the public administration, the government also eliminated a breeding ground for corrupt behavior. The major efforts in this area were the Bassanini Laws. Named after the Minister of Local Government, Franco Bassanini, the Bassanini Laws outlined new roles for the regional and local governments and their administra tion (Gilbert 1999). The main force of the legislation was contained in two laws: Bassanini I and II, as well as several complementary decrees. Bassanini I, law 59/97, begins with a general statement that all functions and administrative tasks concerning the care of e vel by any organ or bodies (Gilbert 1999 141) 14 The regions were handed the task of transferring administration down to the lower levels of government who were charged with assigning authority based on, t 1999 142). The purpose of this law was to, 14 The central government held back on several functions including foreign affairs, trade, immigration, justice, sc ience and technology, and public universities (Gilbert 1999).

PAGE 369

369 Devolve significant administrative and policy making powers to the regions, who would, in turn, devolve the new powers down the line to the provinces, the municipalities and even semi public voluntary associations. At no point in the chai n of administration should there be two layers of government doing the same thing. (Gilbert 1999 142) The 59/97 law was quickly followed by the Bassanini law II, law 127/97. The goal administrative practices and procedures of decision mak 142) The through this hugely complex and lengthy piece of legislation reveals a viv id picture of the bureaucratic miseries that Italians have been accustomed to suffer at the counters of the various branches of all the public 143) The law had the effect of slashing through much of the red tape involved in any government action. For instance, individuals no longer had to get many of the certificates required in the past to request even the simplest government action. The center left government might have had a mixed record on the issue of corruption, but t hey deserve credit for placing it very high on th eir agenda. It was not successful at passing significant anti corrup tion legislation, but while it was pursuing those policies the Clean Hands investigations were successfully prosecuting corruption crimes. When this is coupled with their reforms of the public administration, a place where corruption tended to originate, I believe that Italians believed that the country was beginning to clean itself up. When the Berlusconi, center right coalition won the e lection in 2001, the forces against corruption appeared, according to the historical data, to have momentum. Whatever forward progress was made from 1996 2001 was blunted by the House of Freedoms. The government spent much of its time in office ignoring the

PAGE 370

370 members of the government or coalition from being prosecuted. I believe this had disastrous effects on how Italians viewed corruption in their country and ultimately how much they supported the system. The second Berlusconi government showed its commitment to legislation curtailing corruption almost immediately upon taking office. The government passed law 61/2002 which dealt with false accounting. False accounting statu tes were the mechanisms many in the judiciary used to attack the problem of bribery that plagued the Italian First Republic. Businesses would pay bribes to state agents, but use accounting practices nt categories or line items. The law passed by the center right shortened sentences, reduced the statute of limitations, made it much more difficult to prosecute these types of crimes, and actually decriminalized certain forms of false accounting. The eff ect of this law was of particular important to the Prime Minister. Thanks to this law, several ongoing investigations into other charges despite initial guilty verdict s (Della Porta and Vannucci 2007, 842). Other laws made it more difficult to prosecute those in the business community with crimes related to corruption. These included an amnesty for funds il lega lly hidden in overseas accounts, where the owner of these accounts could now return the money to Italy with only a minimal 2.5% tax and a law that demanded all evidence obtained outside of the country follow strict page by page accounting procedures. These two laws resulted in greater difficulty prosecuting busi nesspeople of crimes related to corruption. This was not surprising considering the large number of people Forza Italia recruited from the

PAGE 371

371 business community. While this government was making it more difficult to prosecute cr imes related to corruption, i t simultaneou s made it easier to conduct its own business. Two laws designed to streamli ne and regulate public projects had the consequence of increasing the likeliho od of corruption. The de lega ted law on infrastructure passed in 2001 and 2002 raised many questions about the opportunities for abuse. These laws de legat could either manage these projects or sub contract the wo rk out to other beneficiaries, exchange to emerge in a public market, especially one that is traditionally closed to (Della Port a and Vannucci 2007, 842) Rather than opening up the process of allocating public money, these new laws increased the opportunities for private exchanges based on political favoritism. The laws passed on public works were not the only laws passed which f avored the business class of beneficiary of favorable policy making by his center right government. The government reconfigured the Italian television system when it passed t he Gasparri law in 2004. The law was initially slowed by the opposition and the President 15 s passage. When the law was eventually passed it impacted all Italian media: radio television, and communications. Specifically, the law had two effects: it allowed for it 15 ceremonial position (Cotta and Verzichelli 2007).

PAGE 372

372 control over the Italian public television channel RAI. The first part of the law al lowed television networks to own newspapers and permitted companies to control a larger piece of the advertising market than was previously allowed. This freed Mediaset to expand beyond its current limits and create an even larger media conglomerate. The Retequattro, to move to satellite transmission. This would have led to a plunge in viewership and ad revenue (Ardizzoni 2007, 36). Another section of the law deals with the privatization of the public broadcast company RAI. The previous center left Olive Tree government had, from 1998 2001, consistently tried to sell part of RAI to private investors 16 The idea was to make the network more profitable and to reduce public exp enses. The sale to an American company was immediately blocked by the incoming center right government and its Minister of Communications, Gasparri. The Gasparri law allowed the privatization of only 1% of the network, which basically left the RAI in the hands of the Italian state and controlled by the government of Silvio Berlusconi. Many criticized this move because of the inherent conflict of interest between an owner of a major private media company and his political position as the de facto head of his major competition (Arizzoni 2007). The issue was clearly something that the center right government had no interest in dealing with conflicts of interest or legislating against their own interests on the matter. Conflicts of interest were an on going issue with Silvio Berlusconi since his emergence on the political scene in 1994. Since the center left government was unsuccessful in passing any sort of legislation limiting or dealing with conflicts of 16 For more on the history of the public sale of the RAI and potential conflicts of interest with Silvio Berlusconi see Hibberd (2004, 157 161).

PAGE 373

373 interest, when Berlusconi reemerged on the po litical scene in 2001 the issue was still unresolved. During the election campaign, Berlusconi repeatedly reassured voters that legislation would be passed within the first 100 days that would deal with the issue once and for all. Strangely, and perhaps foreshadowing the next five years, the center right never detailed any specific plans for what this legislation would contain. Berlusconi had no details, eve n though he had spent the past seven years in opposition, and the center left government had made two unsuccessful attempts at dealing with the previous government, now opposition, reintroduced conflict of interest legislation in the Chamber of Deputies early in 2001 and o ffered the government the chance to use a fast track procedure. If Berlusconi so chose, he could have met his pre campaign promise and shepherded the legislation through the two houses within 100 days. But instead, the government chose to delay (Hine 200 2). Rather than push a bill through in the early part of their term, the center right waited until 2004, where it passed a bill that had less that satisfactory results. The law was designed to address the tensions that existed between a new class of legis lators who came almost entirely from the business commun ity. The trouble with conflict of interest is that it is difficult to legislate. Ethics are something a person either has or does not have. This is evidenced by the type of legislation that was pas sed despite the law. The center right has passed bills making it more difficult to prosecute corruption based crimes, made it easier for those who broke the law by hiding money to return it to the country, made it easier for political favored groups to ac quire public

PAGE 374

374 against a backdrop of almost constant criticism of the Clean Hands judges by the government and Silvio Berlusconi in particular, who was quoted in La Repubblica and secondly they are mad anyway. To do that job you must be mentally disturbed, you must have psychological problems. If they do that job they must e anthropolog ically (Della Porta and Vannucci 2007, 847) There is little doubt that their attitudes and actions toward the issue of corruption had an s the rise and fall of public perceptions of corruption in Italy. Public Attitudes toward Corruption: 1994 2005 The separate trajectories of the public debate and government acti on on the issue of corruption are described in the previous section. To summa rize, I argue that the first government formed after the first election under MMM was a profound disappointment in the area of corruption legislation or control. The first center left government took positive steps to prevent conflicts of interest and tri ed to reduce opportunities (or necessity) for bribery by making the public administration more efficient. Their efforts were not always successful, but the Olive Tree prevention and elimination of corruption a major part of their agenda. The moves against corruption in Italy w ere aided by the Clean Hands investigations that until 1998, were still making headlines and prosecuting major political and business figures from the First Republic. So even without major legislation (outside of improving the public administration), the period from 1996 2001 could be considered a high water mark for public efforts to reverse the corrupt behavior of the First Republic. After 2001, the

PAGE 375

375 momentum t hat was gained in the previous five year s was lost, as the second Berlusconi government generally ignored and exacerbated the issue. Upon entering office for the second time in 2001, Silvio Berlusconi promised legislative action on conflicts of interest in the first 100 days. Despite the center willingness to fast track the legislation, nothing was passed until 2004, a full three years later 17 the Gasparri laws. It was widely believed that these laws were a b oon to Silvio that a potential competitor, the RAI television network s remained in governmental control. Rather than passing conflict of interest legislation, they exacer bated the issue by their questionable actions. The center right government also tried to make it more difficult for corruption crimes to be prosecuted, especially against business people (the occupational group that made up a large part of Forza Italia). There was also the curious choice of changing the way public contracts were handed out, which some believed would increase the likelihood of bribery or politically motivated behavior in the allocation of contracts. The differing policy aims of the center left and center right also came during differing political contexts. The center left only had modest gains in the area of corruption legislation, but while they were aided by a number of factors while trying to achieve those goals. From 1994 to 2001, jud icial investigations decimated the old political order and a new parliamentary class replaced the entrenched elite. This might 17 This legislation did little to address the problems of the center conflicting with their legislative duties. For insta nce, Silvio Berlusconi did not sell any part of his massive business conglomerate. The only change of note was his resigning as president of the AC Milan football club, although he continued to own the team (Della Porta and Vannucci 2007).

PAGE 376

376 have led to citizens giving the center left the benefit of the doubt in their efforts toward corruption. By 2001, there were si gns that the new political class was becoming a professionalized political class and the public was receiving signs that nothing had really changed. When the Berlusconi government made no effort to legislate against corruption they might not have been aff orded the same benefits from the public that the center left enjoyed. I expect that public evaluations of corruption in Italy to follow this historical narrative. Public views about corruption should improve during the period from 1996 to 2001, while decl ining the years that followed. In order to measure public feelings surveys that demonstrate how Italians felt about corruption during the second Berlusconi government. The Control of Corruption Index (COC) encapsulates the perceptions of how much public power is utilized for private benefit and includes both smaller and larger scal e corrupt activities. The measure also attempts to determine the extent to which the state 18 The index ranges from 2.5 (for the worst performing countries) and 2.5 (for excellent performance) 19 This m easure follows the historical narrative that I lay out in the earlier section. The COC shows that the control of corruption improved from 1996 to 2001, where it reached a high point. Scores were .42 in 1996/1997, .59 in 1998/1999, and finally reached a h igh in 18 For a complet e list of concepts measured and sources used see: 19 For more, see

PAGE 377

377 2000/2001 with .89 20 After this point, scores begin to drop from .89 in 2001 to .67 in 2002, falling every year afterward. 2006 rebounded slightly from 2005, but still came in below the previous low score in 1996. Fo r complete results see F igure 6.1 The Perceptions of Corruption Index (POC) is a means of ranking corruption by the degree that corruption is perceived to exist among public officials and politicians. It is a composite score from 0 to 10, where 10 represents the least amount of perce ived corruption, while 0 signifies the most. The POC draws on corruption related data from expert, business, and citizen surveys 21 The results of the POC demonstrate a similar, although not quite as dramatic, upside down U At the beginning of 1995, aft er the Berlusconi government fell and was replaced by a technical government the POC was at its lowest point with a score of 2.99. It rose to 3.42 in 1996 and again in 1997 to 5.03. Scores traded between 4.6 and 4.7 from 1998 and 2000 before reaching a h igh in 2001 of 5.5. Since 2001, the scores have fluctuated but have fallen since their high in 2001. Fo r complete results see F igure 6 2 : The two corruption indices demonstrate that there was clearly an improvement concerning the way corruption was perce ived by the public and how the government dealt with the issue leading up to 2001. Both measures reach their highest point in 2001. There are some mixed results after that date. The COC data provides a clear bell shaped curve. This is right in line wit h how I describe government efforts toward corruption under center left and center slower to drop after 2001. There could be an element of lag in the measure considering 20 From 1996 to 2001, the World Bank only tabulated scores every other year. Starting in 2002 the scores were reported yearly. 21 For complete methodology and data see

PAGE 378

378 the numbers fall significantly by 2009 (4.3) and 2010 (3.9) when the center right returned to power again. These figures are outside of the temporal period that I have selected for this study, so I use some survey data that was collected after 2001, which helps illustrate my argument. A surv ey taken by SWG Conferescenti in 2003 provides some evidence that views about corruption improved leading into 2001, but then declined following. In this opinion survey, 55% of businessmen confirmed that corruption was as prevalent as in the past, against 20% who felt that it was less so. In 1995, in this same survey, only 13% of respondents believed that corruption was destined to increase, while in 2003 that number increased to 33%. Also, in 1995, 56% of respondents believed that in order to receive a contract from the public administration it was necessary to pay a bribe. The numbers increased to 75% in 2003 (Della Porta and Vannucci 2007, 834). Transparency International also held a survey in 2005, measuring feelings about corruption in Italy. 50% of Italians reported that the level of corruption had increased in the previous 3 years, while 38% believed it had remained unchanged. In the same survey, 41% believed the situation would worsen in the following three years, while 38% believed the problem would remain unchanged (Della Porta and Vannucci 2007 834). A survey commissioned by Euripes offers similarly discouraging results. The survey attempted to determine how Italians felt about corruption in 2004, a full 12 years after the Clean Hands inves tigations began (and at least 6 years since the height of the investigations). Compared with 1992, 46.5% of Italians felt that corruption had remained unchanged, with roughly equal numbers believing that it had gone up (24.7%) or gone down (24%). For a c o mplete breakdown, see Table 6 5

PAGE 379

379 In the same survey, respondents were asked a series of more specific questions on the issue of corruption. Here, 70.4% believed that Tangentopoli was never stopped and the system of corruption continued to operate as it u sed to. 56.9% declared that corruption had always existed, will always exist, and there is nothing that can be done about it. Perhaps the reason for this pessimism was the beliefs of 61.2% of Italians who agreed that the Clean Hands judges did not have t he chance to complete their work because politicians blocked them. For full results see Table 6 6 Taken together, the COC, POC, and the surveys from SWG Confesercenti, Transparency International, and Euripes demonstrate how Italians felt about corruption following the change in electoral system. It is my view that parliamentary turnover, efforts to deal with corruption and conflicts of interest, the reforms of the public administration, and the ongoing Clean Hands investigations during the period between 1996 and 2001 led to improvements in how the pub lic view corruption. When the second Berlusconi government entered office in 2001, they not only failed to capture the momentum that had been built over the previous 5 years, but completely reverse it. The y did so by making corruption crimes (especially against businesspeople) tougher to prosecute, failing to treat the issue of conflicts of interest seriously by passing bills e allocation of public contracts more likely. This change is reflected in the data that I present where, generally speaking Italian attitudes about corruption improved from 1996 to 2001, and fell from 2001 to 2006. I believe that the change in these beli efs had an impact on how much Italians supported their new democratic rules.

PAGE 380

380 Before I describe this relationship, I examine whether policy making improved in Italy after electoral system change. I believe that the ability of governments to keep their pr omises after the election and pass effective policy could be a likely alternative making was one of the theorized outcomes of the new electoral system, by scholars and policy makers in Italy. It is also something that I theorize in the majoritarian model. Effective policy, in this model, is a product of electing a governing majority through increased majoritarian elections. The following passages are a description of the successes and difficulties that fractured coalitions had passing effective policies following electoral system change. Italian Policy Making: 1996 2005 In Chapter 4, I describe how the policy making pr ocess functioned in Italy prior to electoral system change. Policie s were generally made through grand bargaining between the political parties, rather than any coherent legislative package presented by a unified government. This was a product of several things, but the party system was first among them. During the Firs t Republic, the elections meant very little in terms of selecting a government based on the policy proposals they offered during a campaign. public about what policies they preferred. One of the goals of the electoral reforms was election coalitions, where voters would be presented with alternative governments who had different policy visions. T he MMM electoral system produced mixed results. There was some success in forcing the parties to form pre electoral coalitions. Voters were presented with a clear

PAGE 381

381 choice between center left and center right coalitions and their respective potential Prime Ministers before the election 22 The electoral system performed its function here reasonably well. Where the system did not perform as well was in the translation of with pre electoral coalitions formed out of parties with disparate policy goals. These policy differences did not go away when a coalition won the election, and the stability that was sought through a change in electoral system, did not come to fruition. Fractured coalitions were the result of parties choosing to enter into stand down agreements with smaller parties rather than risk losing the seat altogether to the opposite coalition. The center right and center left dominated elections, but were made up of many component pieces which were not always in agreement over policy direction. This section describes the successes that the center left government had at the outset of its term, only to fall victim to infighting and two govern mental crises after an initial two year run. They were replaced by the center right, who promised a series of policies in the first 100 days of taking office. This optimism was short lived as Silvio Berlusconi had a difficult time bringing together the opposing parties, the A N and Northern League together. Policy making in Italy follows a similar pattern as corruption. After initial success, the parties fell back into old patterns and failed to build on their successes. This section describes the initial successes in polic y making, the 22 Both coalitions had difficulty holding together at various points in different elections, the center left in 1994 and 2001 and the center right in 1996. When the coalitions fragmented it cost them the election, so fashion in a different section of this present chapter.

PAGE 382

382 subsequent failures, and is followed by a description of how they impacted support for the system in Italy. Policy Making: Center Left and Center Right Romano Prodi, as the head of the Olive Tree was appointed President of the Council of Min isters on May 16, 1996 and received a confidence vote in the Senate two weeks later 23 He would remain Prime Minister for 886 days, from 1996 to 1998, which was the second longest ory (Fabbrini 2000, 121) 24 The coalition government was the first based on a pre election commitment given to the Italian voters 25 They presented this agreement in the form of the Olive Tree (Thesis for the Definition o f the Programmatic Platform of the Olive Tree ), which was intended to serve as a foundation for the legislative program for the government. The document contained 72 specific policy pledges that covered a range of issues from bioethics to judicial reform (Newell 2000 173; Parker 1997) 26 Most of the pledges that were made by the Olive Tree involved the areas of the economy and taxation, (15 pledges) and the government, administration, and Northern 23 The Olive Tree was comprised of several parties: the PDS DS, PPI, PD, RD, and Verdi, and relied on parliamentary support from the RC, who did not join the government in parliament, but did agree to a pre election bargain on electoral seats (Fabbrini 2000). 24 The longest tenured government was led by Bettino Craxi an d lasted for 1,093 days from 1983 to 1986 (Fabbrini 2000, 121) 25 lived government formed after the election. Its pre election, electoral coalitions were based in North (Pole of Freedoms) and the South (Pole of Good Government), so there was no unifying document or manifesto that incorporated the ent ire governmental program. The inability of Berlusconi to get the two separate poles, the Northern League and AN, to work together led to the collapse of the government after only 8 months (Fabbrini 2000). 26 Parker (1997: 129) identifies 88 pledges, but Ne vague or general that it is impossible to tell, after the event, whether or not they have been

PAGE 383

383 League l matters (19 pledges). The government acted up 9 of the 15 pledges (60%) made in the area of the economy and taxation. They fulfilled 10 of the 19 commitments (53%) made in the area of the government and legal matters (Newell 2000, 172) 27 The most pressing issues facing the government were the entry into the European Monetary so the government made sorting out these issues a large priority. Here, the government succeeded in steering Italy toward a path to admission by revamping the budget and reining in public debt. Administrative reform, although encapsulating the highest number of promises, was treated a secondary set of goals. The Olive Tree was successful in passing some legislation, but disagreements over the issue eventually led to the collapse of the government. Prior to the 1990 s, Italy had one of the worst records in dealing with EU demands. This is despite the fact that Italy has consistently been one of the most pro EU member states (a fact which is a democracy). Ensuring Italian participation in the EMU had been the Olive Tree primary campaign pledge, and they used their victory in the election, along with its Tesi as a mandate for enact ing quick, decisive economic reforms that would ensure entry. When the center left made the commitment to be among the first wave of countries entering the EMU in 1999, the government was obligated to put economic and financial reforms high on its agenda. The task in front of them was not small. In 1996, when the government took office, payment of public debt stood at 123% of GDP and a current budget deficit of 130 trillion Lira (54.18 billion dollars) or 7% of GDP (P arker 1997, 137). 27 Newell (2000, 174) points out that these percentages compare favorably with other parliamentary democracies.

PAGE 384

384 Acceptance into the EMU requires a budget deficit of no more than 3% of GDP and the cost of paying for the public debt can not cost more than 60% of GDP. Getting the issues so high on the gov the state budget squarely in the spotlight. The fact that the Olive Tree campaigned on the issue of entry into the EMU and ng Italians, and the serious reforms that the country had to make necessitated swift, decisive government action. This was not what would be considered a hallmark of the First Republic, but the Olive Tree was able to enact substantial reforms to the econo mic sector in a short period of time. They did this by passing austere budgets and by obtaining a remarkable number of legislative degrees (decreti de lega te: the power of the government to issue decrees with legislati ve force when the legislature delega te s to them the power through a vote of parliament). In the two years of the 13 th legislature (1996 1998) obtained over 70 such Delegations granting them broad powers to deal with not only the public debt, budget, and taxes but a range of policies from admi nistrative reform to im migration reform (Fabbrini 2000, 124). The ability of the government to obtain this number of Delegations is noteworthy for two reasons: the number was substantially higher than past legislatures and the fragmented nature of their coalition. The 70 Delegations were a high compared to the 12 th Delegations (1994 1996), the 11 th 1994), 26 Delegations in the 10 th legislature (1987 1992), and only 14 Delegations during the 9 th (1983 1987) (Fabbrin i 2000, 124). The success that the Prodi government had in

PAGE 385

385 securing these broad Delegations was a great achievement since his government did not even hold a majority in the Chamber of Deputies 28 When coupled with aggressive budgeting, these Delegations w ere used to enact a series of policies that aided the transition to the EMU. budget cuts aimed at reducing public debt levels as well as the budget deficit. Skeptics question ed whether the government and Italians would have the discipline and patience to deal with su ch belt tightening (Parker 1997, meeting the 3% deficit/GDP ratio was clear during the 1997 budget legislation. The package passed quickly through both houses of parliament and produced a surplus of 6.7% of GDP and the deficit/GDP hit its target rate of 3%. Debt levels also continued to decline, falling from a high of 124.9% of GDP in 1995 to 123.8% in 1996 to 122.6% in 1997 ( Fel sen 2000, 164). The 1998 budget was the most controversial of the three passed under the Premiership of Prodi. The RC, who the Olive Tree needed for legislative majorities in the Chamber, balked at the continued efforts toward fiscal consolidation. T he RC wanted a budget that was heavily weighted toward creating new employment, which would have required additional spending, and creating a 35 hour work week. The RC and its leader Bertinotti decided to hold up the passage of the 1998 budget until its d emands were met. Because the Olive Tree maintain the support of its largest member the PDS, the government did not have to reorganize or cede to the demands of the RC. Instead, the Olive Tree appealed to the Italian public directly and 28 The Communist Refoundation supported coalition, but outside the government. They never agreed to up (Hine and Vassallo 2000).

PAGE 386

386 openly criticized the RC for attempting to damage the efforts to enter into the European currency. The public sided fully with the government and the RC was eventually pressured into giving up on its demands. The budget that passed in late 1998 reduced the defi cit to 2.7% of GDP (Felsen 2000, 164). The consistent deficit reduction policies of the Olive Tree government helped produce consistent surpluses, reduced borrowing below 3% of GDP, and lowered the netary policy also reduced inflation and interest rates as well as stabilizing exchange rates. Through their actions Italy satisfied Maastrich criteria and was selected to be among the first wave of countries set to join the single currency set to come in to existence in early 1999 (Felsen 1999). The acceptance into the EMU was the Olive Tree achievement. But, it was not the only legislative accomplishment. The large number of legislative decrees they were able to obtain from the par liament were used to good measure as well. Legislative decrees were used to reform and improve several areas. Significant overhauls were made to the budget, budgetary processes, and the Ministry of the Budget. These reforms helped remove some of the in fluence of the legislature over the budget, made the process more open, and centralized decision making in the Ministry of Budget, the Treasury Ministry, and Director General of the Treasury (Felsen 2000 ; Hine and Vassallo 2000 ). These reforms granted mor e power to the Ministers responsible for the budget and took it out of the hands of the legislature. Decrees were also used to aid and accentuate the Bassinini Laws designed to improve and streamline

PAGE 387

387 the unwieldy public administration 29 Public employment and the higher civil service were also reformed through decree legislations. This type of legislation was also used to redesign personal income taxes, remove small tax charges on businesses, and gave regional authorities more fiscal a utonomy (Hine and Va ssallo 2000, 36). Considering all of these reforms and the success of EMU entry, the Prodi led Olive Tree had a fairly impressive two year legislative record. But, the successes were not without difficulty. The RC continually objected to the austerity o f the budget process and the lack of government action on their priorities. Their dissatisfaction, along with increasing up of the ster in late 1998. With this collapse, the legislative accomplishments of the center left stalled as well. The first governmental crisis took place as 1998 was coming to a close. The RC had consistently showed dissatisfaction with the budgets passed by t he Olive Tree government. As early as 1997 Fausto Bertinotti, the Party Secretary of the RC, (Fabbrini 2000, 131) Bertinotti was a bit out front of his fellow party members who disavowed his words. But, the underlying point that the RC was unhappy with the budget was right on point, and by the autumn of 1998, their concerns could no longer be d committee voted in favor of withdrawing support for the government in October, and a 29 I discuss this in an earlier section of this chapter on corruption and improvements made in that area.

PAGE 388

388 confidence vote was scheduled soon afterward. On October 9 th the government lost a confide nce vote 313 to 312 (Fabbrini 2000 131). Olive Tree Prior to the collapse, there was increasing strain between the leaders of the expand the c oalition to include the UDR, that included a group of leaders who opposed the Olive Tree during the election of 1996. Prodi was completely opposed to the idea that a party group could be a replacement for an elected member of the coalition, as this would be a complete distortion of what the people voted for. So, In reality, with the onset of the crisis brought about by the RC secretary, the Olive Tree party leaders had the numbers to resolve the crisis, while Prodi had the symbols. And when push came to shove the former counted more than the latter. Thus, although the no confidence vote motion had been carried by the slimmest of margins, the Olive Tree Party leaders were able to implement the principle that they had constantly sought to assert: namely that the legitimacy of parliamentary governments resides in the parliament and not in the electorate. (Fabbrini 1997, 132) Mass as a result of jockeying between the old parties of the Olive Tree (DS, PD, PPI, RI, Verdi) as well as two new groups the UDR and the Partito dei Comunisti Italiani (a splinter from t he RC). This government was clearly a post electoral one, and had consequences for policy making. Where the Prodi led Olive Tree government went to the voters with a coalition agreement and series of policies and received a mandate from the voters to pur sue them. They were able to use that mandate to push the coalition into making tough policy choices on the issues that I mention earlier. But, the

PAGE 389

389 result of horse trading and ba rgaining among parties. This behavior resembled the way governments were formed under the pre reform period, where there was really no connection between voters and the government. As a result of this governments during the First Republic had difficulty passing meaningful policies 30 reforms that he supported 31 It did manage the economy reasonable well with sound macroeconomic policies, kept the public debt under control, and were able to stay within the guidelines for members of the Euro (Gilbert and Pasquino 2001). Hindered by the challenges to his leadership position. He fi nally lost a vote of confidence in 2000 and was replaced by Amato. Prior to the election of 2001 he was replaced as the leader of the center left by Francesco Rutelli, the little known, former Mayor of Rome. Rutelli and the center left were crushed in th e 2001 elections by a rejuvenated center right and their leader Silvio Berlusconi. The elections held in May of 2001, for the first time in the history of the Italian Republic, produced alternation in government as a direct result of an electoral victory b y the opposition. The center right coalition, the House of Freedoms, made up of Forza Italia, Northern League the AN, and a smaller group of center, Catholic parties, defeated the incumbent Olive Tree coalition. Silvio Berlusconi rose once again to the premiership, and with substantial majorities in both houses of Parliament. Because of 30 Governments during this time relied instead on leggine (small, clientelistic laws passed in committee) or re iterated d ecrees. For more information on these see Chapter 4. 31

PAGE 390

390 a series of campaign pledges that spelled out what the government would do o n key and the economy. The new government also declared it would act swiftly and decisively fice and a long term plan that included changes in the labor market, tax system, educational system, and count 155; Cotta and Verzichelli 2003). With their electoral victory, a secured parliamentary majority, and a plan for action the center right seemed to be in a good position to act decisively on these pledges. Despite what seemed like favorable conditions, the center right was unable to deliver on much of its policy agenda. The coalition, despite holding majorities in both houses, remained fragmented and disagreements over key policy issues made the efficient, decisive action promised, virtually impossible. Almost immediately after taking office, the government began backtracking from its promises on taxes, pensions, an d work. They declared that the last budget passed by the center left had a buco di bilancio back some of their promises. A contentious debate over the state of the public treasury followed, and the gover nment passed some minor changes to tax policy and pensions, but far short of what they promised during the election (Cotta 2002, 155 156). This would be a continuing theme for the House of Freedoms. Time and again the government would declare its intenti on to pass significant reform only to back away due to political realities. Some of these included the constraints of the European Growth disagreements over policy within the center right coalition.

PAGE 391

391 external challenge, one that would highlight many of the existing problems the historic position of support for integration with Europe, but had to deal with internal pressures from members within the coalition who were Euro skeptics. Once entry into the single European currency had been achieved, the Pact for Stability and Growth s everely on economic issues. This fact, coupled with the decline in economic growth within the country made pre election promises to drastically reduce taxes and increase government spending a near impossibility. This is part of the reason that Berlusconi and the center right backtracked on these issues almost immediately. Europe was also and issue for the coalition itself, especially among the Northern League The Northern League was a consistent supporter of de centralization and federalism, and criticized the pro Europe stance out of fear between a convergence between European and domestic centralization (Cotta and Verzichelli 2003). The austere measures that the center left passed in the 1990 s in order to guar antee entry into the EMU were not one time commitments. In order to maintain a member of the monetary union in good standing, the government could not simply ignore the commitments the country had made. In terms of pursuing policies that it had promised, Europe provided constraints on the government and also highlighted the fragmented nature of the governing coalition. Another reason the government was unable to pursue its aggressive agenda was the economic situation within the country. Much of what had been promised in 2001 was predicated on rosy economic predictions and the belief that the economy would

PAGE 392

392 continue to grow as it had. In 2000 and 2001 GDP grew 3% and 1.8% respectively, higher than the 10 year previous average of 1.7% (Guarneiri and Newell 2005, 32). In 2002, 2003, and 2004 the economic growth slowed considerably to 0.4%, 0.3%, and 1.3%. The center and increase spending (education, welfare, public order, etc.), economic gr owth was vital. Despite the bad economic climate, the government did pass a limited form of its tax cuts in 2004. The state of the economy was not just damaging for the government he electorate had economy as a whole what he had done for his own companies, the gap between expectations and reality began to feed the perception that the economy and living 32) These perceptions were companies: Fiat and Parmalat (Cotta and Verzichelli 2003; Onado 2005). The economy provided a barrier, but perhaps the largest issue with the center right was the divergent interests within itself. The policy distance between two of the major pieces of the coalition, the Northern League and the AN, were a continual problem for passing a coheren t legislative program. The differences among Forza Italia, the AN, and the Northern League were not new issues on the center right. The first Berlusconi led government in 1994, collapsed after only 8 months because of the irreconcilable differences between the parties. The coalition that took over after the 2001 election was in a different position. Forza Italia was clearly the largest party, although short of an absolute majority, and had more

PAGE 393

393 flexibility since it did not need both the AN and Northern Le ague to continue governing. clear subordination of the entire coalition to his leadership which were clearly evident during the electoral phase seemed to be replaced by a more uncertain state of affairs 46) Conflict within the government became common place, with scholars identifying 36 major public clashes over policy in 2002 (Cotto and Verzichelli 2003 51) and 32 major issues dividing the majority in 2003 (Donovan 2004, 86). The government repeatedly was forced to tie its policy goals to votes of confidence, and the 14 th Legislature called more than had previously been the norm. The vote of conf idence gave priority to government bills and forced a roll call vote, which made defection more difficult (Vassallo 2007). In and of itself, the use of no confidence votes is a useful tool, especially under heterogeneous coalitions like the center right ( and center left) in Italy. What was hose easing the penalties for illegally hidden accounts overseas to be returne d to the country and reforms of the judicial system making it more difficult to prosecute corruption crimes (Vassallo 2007). The yearly budget, passed at the end of the year, was also subject to a no confidence vote (a practice that began under Prodi and the center left). The Finance Law, passed favored provisions. Since the budget was virtually guaranteed to pass, these provisions

PAGE 394

394 would go through as well. Taken together, i t seemed as though the center right was returning to a time where spoils and corruption were permitted 32 The point of this section on policy making is not to say that the center left was effective and the center right was ineffective. What I intend is tha t the center left, while fractured and heterogeneous, was able to achieve its main policy objective: entrance into the EMU. The center right, although somewhat less fractured and still heterogeneous, had a much more difficult time with its own policy agen da. What I argue is that the policy successes of the center left, coupled with the perceptions that corruption was going down led to higher levels of support among Italian citizens. After the center right was elected in 2001, the momentum that had been a chieved in the previous seven years began to wane. Silvio Berlusconi and the center right, made a concerted effort to block corruption prosecutions, ignored the issue of conflicts of interest, and attacked the judic iary. This led to much more pessimistic feelings about corruption, and a feeling that the old way of doing things had returned, was pervasive among the population. When the government failed to enact significant policy promises, I believe that the support had been built was lost. The followin g section looks at the rise and fall of public support in Italy. Discussion: Support in Italy under MMM 1994 2005 In order to judge the level of support among Italians, I use the measure satisfaction with democracy. This is a survey item found in the Euro barometer and asks 32 See earlier section on corruption and the center right for more detail.

PAGE 395

395 by scholars, but its value is trumped by others who use it in mul tiple studies 33 Satisfaction with democracy in Italy follows the historical narrative that I lay out in previous sections. For complete results see Figure 6 3 Figure 6 3 display s democracy. T sa tisfaction and dissatisfaction. In 1994, the first year of the new electoral system 25% reported satisfaction with, 73% dissatisfaction. Italians reached their lowest point in 1995, with only 19% if Italians satisfied, with 79% dissatisfied. 1996 began an upward tread, reaching a high in the year 2000 where an unprecedented 69% reported began a downward trend finishing in 2004. Why did satisfaction rise, peaking in 2000, while falling in the years to follow? I believe the historical narrative I describe in the earlier sections answers this question. What these satisfaction with democracy resul ts show is a similar upside down U to the indices of corruption. I d o not believe that this is coincidental. General satisfaction rises as the perceived levels of corruption fall. I theorize at the end of Chapter 4 and the b eginning of Chapter 6 that Italian support would only rise if they felt that the persistent corru ption unearthed by the Clean Hands investigations was significantly improved. The electoral system could aid this process by creating bipolar competition, with one coalition taking over the government and governing according to effective, 33 For a complete discussion of the use of this var iable see Chapter 2.

PAGE 396

396 coherent policy. To a certain extent, this happened from 1996 2000. The Prodi led government tried to project an image of competent, honest government. They made several efforts at dealing with the problem of corruption legislatively, as did the ensuing to led center left governments. Ultimately, these efforts were unsuccessful, but public attitudes toward corruption improved, I believe because the have been from the succe ssful administrative reforms undertaken by the Olive Tree which were designed to streamline and improve the inefficient public bureaucracy. By doing so, a major reason for clientelism was removed. Factors having little to do with the electoral system ce rtainly played a role as well. The Clean Hands investigations were still ongoing and continued to bring charges against prominent political elites. Also, in 1994 and 1996, there were high numbers of parliamentary turnover. I believe this created a real eelings about corruption nation wide. The two indices, the Perceptions of Corruption Index and Control of Corruption Index, are evi dence that this occurred. Both the POC and the COC, show feelings about corruption improving until 2001, and as a result feelings of satisfaction with democracy improved as well. Corruption was not the only element responsible for the improved attitudes o f support in Italy. Despite the fractured nature of the center lef t, they were able to pass the key policy initiative, entry in to the EMU. This was not a small task. Italian public debt and the deficit to GDP ratio were well outside of the Maastrich re quirements. The Olive Tree passed austere budgets in 1996, 1997, and 1998 that improved those

PAGE 397

397 indicators significantly. While the Italians never reached the Maastrich targets, they had shown enough improvement to be entered into the first wave of countri es entering the Euro. This was a significant achievement, and since the economy was growing at 1.7%, with a high of 3% in 2000, the Olive Tree got credit for being competent managers of the economy. This also contributed to the higher feelings of satisfa ction among Italians. Public attitudes fell following 2000 as Italians felt that there had been significant backslides on the issue of corruption and management of policy. The Berlusconi government was simply a disaster in the area of corruption managemen t. Not only did they not make any efforts at curbing the problem, they went out of their way to block efforts to prosecute corruption crimes. This was especially the case for the Prime Minister himself, who was the target of several investigations concer ning his massive business empire and their financial activities. The government passed legislation to make it more difficult to prosecute those who hid money overseas, protected businesses from accounting practices designed to hide bribes, as well as avoi ded the issue of conflicts of interest. In keeping with that, the government passed purchase additional assets, and blocking the public sale of a rival network the RAI. Berl usconi also continued his broad side attacks at the judiciary calling them at one feelings about the level of corruption in their country. The POC declined as well as the improved in the area of corruption were rampant. This is a large part of the reason that support declined in the ensuing years as well.

PAGE 398

398 I believe that the actions of the center right government on corruption were important reasons for the falling levels of s upport among Italians, but its policy making abili ty certainly did not help. The government promised Italians major tax reforms and spending increases. But due to the constraints of the EMU, the economic conditions in the country, and the fractured nature of the House of Freedoms, they were unable to deliver on these promises. This also contributed, in my estimation, to the loss of support. Conclusion: Thinking about Electoral Reform in Italy The rise and fall of support in Italy following electoral system reform suggests a few things about the efficacy of the majoritarian model. Italy may not be the purest tes of this model because of the continued fragmentati on of the party system. But, the fact bipolar competition between two alternative governing coalitions emerged suggests that the model can help inform us about what happens when majoritarianism is increased. This section details some of the conclusions th at I draw from the performance of Italian democracy after electoral system reform. These include how the majoritarian model informs support in Italy after electoral reform, the role of the party system, and other general conclusions. The goal of this proj ect is to determine how and if electoral systems influence support. Increased m ajoritarian ism through a change in electoral system might lead to higher levels of support in the following manner. A decrease in proportionality leads to increased mechanical /psychological pressures to cast a vote for larger parties at the expense of third parties. This reduces the number of parties by eliminating smaller parties or forcing parties to enter into pre election coalitions. By reducing the number of parties, the goal of the system is to elect not only parties to office, but clear

PAGE 399

399 governments. When choices are reduced in this manner, parties (or coalitions of parties) run for office with clear policy agendas that they can present to the voters. Voters vote for o r against parties based on the programs they offer or the effectiveness in office. This is the key process of accountability. When parties are given firm control of the government, they are free to act according to the needs of the country. The thought is that this control leads to more efficient and effective government (better policy). Support comes from the ability of voters to vote in or out governments that are passing or not passing effective policy. How does this model do when applied to Italy a fter 1994? majoritarian elements of the elections did encourage parties to form pre election coalitions. The regional strength of smaller parties, however, guaranteed that the part y syste m remain fragmented. Within the coalitions there was a tremendous amount of difference. These differences had, at times, significant consequences for the ability of governments to pass their policy agenda or even hold themselves together. For ins tance, the first gov ernment elected under MMM last eight months because the Northern League and the AN initially proved to be incompatible coalition partners. The center left, while able to hold together for close to three years, also fell victim to inter nal squabbling and in fighting eventually formin g several governments in their five year term. The center right government elected in 2001 was also plagued by internal disagreement, especially between the ideologically opposed Northern League and AN. Wha t is clear is that the promise of electoral reform producing stable, cohesive governments formed by coalitions was only partially met.

PAGE 400

400 The majoritarian electoral system did provide a measure of accountability. Coalitions that were unable to present to the voters a united front were punished by the electoral system and subsequently lost the election. In 1994, the center left was unable to come to pre election agreements, but were able to take advantage of the divided center right in 1996. The center right was able to correct the problems of 1994 and 1996 in 2001 and win an election over a fragmented center left. Feelings toward both 1994 and 1996. In the past, it was the party that controlled the selection of list of the reasons I believe attitudes to ward corruption im proved. But, after an initial \ of the party system, parliamentary turnover slowed in the ensuing years. Once the party system reached a sort of equilibrium, the old behaviors of political elites began to re emerge. Many of th ese old behaviors manifested themselves in the policy making activities of the both the center left and center right governments. The center left had some policy successes, namely the efforts toward reining in public spending, steering the country into th e EMU, and passing significant administration reform. The coalition was able to achieve those accomplishments when they were able to hold themselves together. By the end of 1998, parties resorted to the old bickering and horse trading that they had in th e past by reformulating the coalition on the fly despite what the election results dictated. It is not a coincidence that the major policy initiatives stopped when the old bargaining reemerged. This is another example of how accountability and

PAGE 401

401 good polic y go hand in hand. The center models into question. The Belusconi led government committed itself to policies that made it more difficult to prosecute corruption and shielding its supports from corruption charges. The center Northern League and the AN within the center right itself. The rise and fall in support follow ing electoral system change has more to do with the behaviors of the party system than the electoral system. The electoral system was at least moderately successful in pushing parties into two coalitions. What the partie s did afterward is what affected t he level of support in the country. When the center left was able to hold itself together, they passed effective policies, made efforts toward curbing corruption, and kept many of the promises they made before the election. But, inter party bargaining, h orsetrading, and logrolling took over following 1998. This is a point that I make in Chapter 2. The models only work when parties behave responsibly within the given rules. When the models do work there is a positive result, when they do not the result is variable. Further evidence of this is the behavior of the center right parties. What theory of electoral systems can explain Silvio Berlusconi? How can a particular electoral system inform why the center right decided to pursue a course of policies t hat undermined efforts to curb corruption? There really is no institutional explanation. The center right pursued its own interests because it could, the coalition won a decisive election and was free to pass policies as they saw fit, so they did. They were only hindered by the ideological difference that existed between parties within the coalition.

PAGE 402

402 What this demonstrates is a point that I have made before, that it is the behavior of the party system that matters in the vari ability of support. What C hapter 6 demonstrates is that a neat theoretical link can only be drawn when things fall into place from electoral systems and support when everything falls into place. For e xample, after the 1996 election, the more majoritarian electoral rules punished a divided center right and allowed the center left to get elected based on a concrete, detailed coalition agreement. The ensuing Olive Tree government passed significant polices in the areas of administrative reform and debt reduction. This led to higher levels of support in the country. When the government could no longer hold itself together, the electoral system punished them and the country elected a center right government. This government did very little in the areas of corruption and continually f ailed to pass significant portions of his policy agenda. Support for democracy fell as a result. The electoral system was the same through these highs and lows in support, what changed were the behavior of the parties. We should question the efficacy of theories that only work contingent on a completely separate set of variables. Elections produce winner and losers, what those winners and losers do following the election is up for grabs. This is what the review of electoral system change in Italy demon strates and is a point that I make in the conclusion to follow.

PAGE 403

403 Table 6 1 Electoral Coalitons under MMM in the Chamber of Deputies Italy 1994 2001 Coalitions and Parties 1994 1996 2001 FPP Proportional Total FPP Proportional Total FPP Proportional Total % Seats % Seats % Seats % Seats % Seats % Seats House of Freedoms/ Pole of Good Government 46.1 302 46.4 64 366 43.2 169 42.1 77 246 45.4 282 49.5 86 368 Alliance of Progressives 33.0 164 34.5 49 213 Pact for Italy 15.6 4 15.8 42 46 The Olive Tree 45.7 247 34.7 38 285 43.5 189 35 58 247 Northern League 10.8 39 10.1 20 59 Communist Refoundation 2.7 15 8.6 20 35 5.0 11 11 Others 1.3 5 1.3 0 5 4.6 5 4.4 0 5 10.9 4 10.7 0 4

PAGE 404

404 Table 6 2. Electoral Coalitions under MMM in the Senate, Italy 1994 2005 Parties and Coalitions 1994 1996 2001 FPP PR Total FPP PR Total FPP PR Total % Seats Seats Seats % Seats Seats Seats % Seats Seats Seats House/Pole o f Freedoms 42.5 128 28 156 37.3 67 49 116 42.5 152 24 176 Alliance of Progressive/Olive Tree 33.4 97 26 123 41.2 134 23 157 39.2 77 51 128 Pact for Italy 16.7 3 28 31 Northern League 10.4 18 9 27 Communist Refoundation 2.9 10 0 10 5.0 0 4 4 South Tyrolean .7 3 0 3 .5 2 0 2 .4 0 2 2 European Democracy 3.2 0 2 2 Valee Union .1 1 0 1 .1 1 0 1 .1 1 0 1 Italy of Values/Di Pietro List 3.4 0 1 1 O thers 5.0 0 0 0 7.6 0 2 2 6.1 0 2 2

PAGE 405

405 Table 6 3. Italian Parliamentary Seniority by Group, Chamber of Deputies 1994 and 1996 Party Groups Newly Elected (%) 3 or More Legislatures (%) Average Seniority (In Legislatures) 1994 1996 1994 1996 1994 1996 Com munist Refoundation 61.5 40.0 5.9 2.9 1.6 1.9 Democratic Party of the Left 58.7 39.9 5.6 9.4 1.7 2.0 PPI 51.5 70.1 15.2 10.4 2.2 1.6 Italian Movement 73.1 3.8 1.4 Northern League 69.5 50.8 1.7 1.3 1.7 CCD CDU (CCD) 77.3 46.7 9.1 16.7 1.6 2.0 Forza Italia 90.5 47.2 3.4 4.1 1.5 1.6 National Alliance 74.5 32.3 11.8 15.1 1.7 2.1 Total 70.8 44.8 6.0 8.9 1.6 1.9 *Data acquired from Verzichelli (1997, 149)

PAGE 406

406 Table 6 4. Italian Parliamentary Seniority by Group, Senate 1994 and 1996 Party Groups Ne wly Elected (%) 3 or More Legislatures (%) Average Seniority (In Legislatures) 1994 1996 1994 1996 1994 1996 Communist Refoundation 38.9 9.1 27.8 9.1 2.4 2.5 Greens La Rete 69.2 57.1 7.7 7.1 1.4 1.8 Democratic Party of the Left 45.5 27.6 5.2 10.2 1.8 2.2 PPI 45.2 40.7 6.5 18.5 1.9 2.2 Italian Movement 54.5 18.2 1.9 Northern League 61.7 48.1 1.3 1.9 CCD CDU (CCD) 66.7 48.1 16.7 12.0 1.8 2.0 Forza Italia 94.4 58.3 2.8 2.1 1.1 1.5 National Alliance 64.6 32.6 12.5 16.3 1.8 2.2 Total 58.4 4 0.3 7.9 9.5 1.7 2.0 Data acquired from Verzichelli (1997, 149)

PAGE 407

407 Figure 6 1. Control of corruption i ndex, Italy 1994 2005

PAGE 408

408 F igure 6 2. Corruption perception index Italy 1994 2005

PAGE 409

409 Table 6 5. Italian public perceptions of corruption in 2004 c ompar ed to 1992 Answer % Gone up a lot 13.7 Gone up a little 11.0 Remain Unchanged 46.5 Gone down a little 3.3 Gone down a lot 20.7 4.7 Total 100.0 Data obta ined from Bull and Newell (2005, 12)

PAGE 410

41 0 Table 6 6. Italian perceptions on the cle an hands i nvestigations After 12 Years Extent of Agreement (%) Hypothetical Judgements Very Much in Agreement Generally in Agreement Not Much in Agreement Definitely not in Agreement Know/No Response Total Tangentopoli was never stopped, and the sy stem of corruption continues to operate as it used to do 32.5 37.9 17.0 6.6 6.0 100.0 Corruption has always existed and always will exist, and there is nothing that can be done about it 31.8 25.1 21.5 16.9 4.7 100.0 The Clean Hands judges did not have th e chance to complete their work because the politicians blocked them 29.3 31.9 16.1 12.5 10.2 100.0 Data obtained from Bull and Newell (2005, 113)

PAGE 411

411 Figure 6 3. Satisfaction with d emocracy, Italy 1994 2005

PAGE 412

412 CHAPTER 7 WHAT DO ELECTORAL SYSTEMS HAVE T O DO WITH SUPPORT? Introduction systems and s upport. I hold in Chapter 1 and Chapter 2 that there is a natural fit between these two literatures that is both fruitful for study and under examined. In order to fill the gap that exists, I derive from the literature on electoral systems and support, competing models linking how diff erent electoral systems can provide higher levels of support among the population. I forward New Zealand and Italy as cases for study for two reasons. New Zealand and Italy represented excellent examples of majoritarian and proportional systems respectiv ely. Both systems, after a long run of success, turned to mixed member systems when support among their citizens dropped to precipitous lows. Changes in electoral systems, designed to correct some of the perceived ills, in both countries proved to be les s than successful in improving the level of support. The question remains, what do the models tell us about the relationship between electoral systems and support and how do the experiences of New Zealand and Italy inform us about that relationship? A p reponderance of the evidence leads to the conclusion that electoral systems, whether majoritarian or proportional, cannot be linked to more supportive attitudes. Efforts to increase proportionality or majoritarianism also do not appear to be successful in creating higher levels of support among the population. This conclusion highlights evidence from both New Zealand and Italy before and after their electoral reforms. The reason I reject the connection between electoral systems and support has to do with the

PAGE 413

413 variability of behavior demonstrated by individuals, parties, and governments in both Italy and New Zealand. Electoral systems largely produced the expected immediate effects and the first steps of the theoretical models were generally met. The more distal effects of electoral systems are more problematic. This conclusion describes how the models generally fail, and why no clear link between electoral systems and support can be made. My overall conclusion is to reject any implicit connection between electoral systems and support, but this does not mean that particular electoral regimes cannot foster support. New Zealand and Italy, prior to their reforms, are often critiqued by eir inherent flaws. I believe that the evidence in this work demonstrates that these conclusions are not accurate. Majoritarian and proportional systems can produce outcomes that lead to higher levels of support. Efforts to increase proportionality and majoritarianism in both New Zealand and Italy, while ultimately unsuccessful in terms of raising support, had some successes as well. I believe that it is instructive to look at how these systems worked to inform both academics and policy makers. In th is Conclusion I highlight a few key findings from this work. The main argument is that electoral systems cannot be credited for the variability in support in a given country. Next, I look at how majoritarian and proportional systems, or increasing propor tionality or majoritarianism, can operate in a manner that might lead to higher levels of support. There are also confounding and contributing factors outside of the electoral system that impact support including: political context, institutions (includin g contamination effects of mixed systems), and agency. Finally, I explain how the

PAGE 414

414 conclusions of this study can inform policy makers when designing rules and academics studying the political consequences of electoral systems. The Connection between Electo ral Systems and Support The main conclusion of this work is that there is no real connection between electoral systems and support. That is not to say that particular electoral systems cannot be successful and lead to higher levels of support. This part of the conclusion discusses two main findings. The first is the ultimate failure of the models to link electoral systems and support. The second is that increased majoritarianism and increased proportionality can lead to higher levels of support. Unders tanding what those conditions can broaden understanding as well as aid both scholars and policy makers especially in the design and reform of electoral institutions. Before discussing the findings of this work it may be helpful to pause and review why sch olars have held that different electoral systems might lead to more supportive attitudes. The theoret ical links I draw in Chapter 2 are derived from the larger body of the literature on the political consequences of electoral systems. At the most basic level, the expectation is that the type of electoral system will have a mechanical consequence that leads to a psychological reaction from the voters. These, more immediate effects of the electoral system, help to determine the number of parties able to w in seats in the legislature. The number of parties impact the type of government formed, the way that government operates, the kinds of policies that are adopted, and ultimately the level of support the population expresses. This logic is expressed by sc holars in the following manner, Logically, the establishment of electoral procedures and institutions precedes the measurement of support for the system, this assumes a chain of causality running from institutions, party performance, and party system

PAGE 415

415 perfo rmance, on the one hand and satisfaction with democracy on the other result in particular kinds of party and party system performance, which in turn should affect citizen satisfaction wi th democra tic governance. (Anderson 1998, 575) This argument seems to make intuitive sense, but it is my belief that if there is a connection between electoral systems and support, then that connection should be made explicitly. The above quote does not go far enough. In order to fill this gap in the literature, I derive two distinct models that theorize the effects that increased proportionality or increased majoritarianism can have on the subsequent variables ending with differing levels of support I ncreased proportionality might lead to higher levels of support in the following (Lijphart 1984). This means that they have mechanical effects that make it easier for smaller parties to transfer votes to seats (Duverger 1954; Rae 1971; Riker 1982). These more accommodating mechanical effects lead to different psychological effects. Voters and parties recognize the increased opportunity to gain seats, so voters are freer to v ote for a party that represents their ideological interest and parties are encouraged to run since their opportunities to win seats are increased (Duverger 1954; Cox 1997). When more parties are represented in the party system, the ideological diversity i ncreases as well (Cox 1997). This leads to coalition governments that are more representative of the population as a whole (Powell 200 0 making process, the policies that are passed are bett er 1 The key process of representation and better policy lead to higher levels of 1 There are a number of works that show that countries that use proportional representation have higher gross national products, lower unemployment, lower inflation, policies are more egalitarian, and public

PAGE 416

416 support in the population. This is the more popular belief among scholars (Amy 1993; Anderson 1998; Banducci, Donovan, and Karp 1999). There is less of a consensus in the scholarly community around the potential benefits of more majoritarian systems. There is a theoretical case that a link between more majoritarian systems and support however and the systems do have their proponents. More majoritiarian systems have mecha nical effects that are penal to smaller parties and tend to produce two party systems as a result (Duverger 1954; Rae 1971; Riker 1982). This creates a psychological pressure on voters to vote strategically or to vote for against their least favored alter native out of fear of wasting their vote (D uverger 1954; Rae 1971; Cox 1997 ). The goal of more majoritarian systems is disproportionality and to elect a governing majority rather than guarantee everyone representation. By promoting a two party system the hope is that those parties will present clear alternatives to voters. Voters then base their votes on these alternatives election promises. More majoritarian systems help produce vertical ac countability (Schattschneder 1950). This accountability is the trade off for disproportionality and the hope is that parties, when given governing majorities, will produce policies that meet the needs of the country efficiently and decisively (Duverger 19 84; Powell 2000). The combination of accountability and effective policy might lead to higher levels of support. These are the two competing models that I test in New Zealand and Italy, both before and after electoral system change. New Zealand was an excellent example of a majoritarian system and sought to add proportionality by switching to a MMP system in goods are distributed more efficiently (Baylis 1989; Snyder 1989; Powell 2000; Lijphart 1999; Lizzeria and Perisco 2001).

PAGE 417

417 1996. Italy, on the other hand, was an excellent example of a proportional system that wanted to increase majoritarianism by changing their electo ral system to MMM in 1994. I draw conclusions from both countries about the nature of majoritarian and proportional systems and their ability to influence support. I also look at whether electoral system changes (both in the majoritarian and proportional directions) could help foster supportive attitudes as well. Ultimately what I find is that electoral systems have little to do with support. Electoral systems did influence the numbers of parties that were in the system, but the behavior of those partie s was variable. Sometimes parties and governments were accountable or representative. At times parties and governments were able to pass policies that addressed the pressing issues of their respective countries. But, as we move down the theoretical chai n, away from the immediate impacts of electoral systems (the number of political parties in the system) into the more distal effects (behavior of governments or policies passed), the more unpredictable the results. What I find is that support is contingen t on the more distal effects. Can Electoral Systems Foster Supportive Attitudes? The evidence from New Zealand and Italy, both before and after electoral system change, leads me to conclude that there is no real connection between electoral systems and sup port. In terms of the immediate effects of electoral systems (represented by the initial steps in the theoretical chain), the models do quite well. It is when we move down the theoretical chain where the more distal effects of the electoral systems are t here is more opportunity for the parties to break down. This is especially true for both New Zealand and Italy prior to electoral system change. Despite their success was coupled with a change in party system behavior in both countries, absent

PAGE 418

418 a change in the immediate effects of electoral systems, which led to a decline in support. After electoral system change, there were successes and failures in both counties, and I al so examine why that was the case. New Zealand operated under majoritarian electoral rules until adopting MMP in 1996 (after a 1993 referendum on the electoral system). These electoral rules continually functioned as they were designed. They overrepresent ed the two largest parties, Labour and National, and provided them with a significant electoral advantage. Between 1935 and 1993, Labour and National combined to win 1494 of the 1516 seats, roughly 98.5% of all the seats contested (Bale and Roberts 20 0 2; 11). Also, during the same time period one of the two main parties won a majority of seats in the parliament and formed a single party government. This was the primary function and most immediate effect of the majoritarian electoral rules, which consiste ntly functioned as intended. What changed was the level of accountability of governments and the types of policies that they passed. From 1935 to 1984 accountability was something that both Labour and National took very serious. Governments went throug h great pains to state their intentions clearly in a pre election manifesto, consult with their back bench, and hold regular meetings with the interest groups that would be affected by potential policy decisions. There is also ample evidence that prime mi the manifesto was represented at all times. These governments generally promised and delivered generous social welfare and protectionist economic policies. During this time New Zealanders offered high levels of support for their system which was evident

PAGE 419

419 by the high levels of public trust and satisfaction, robust party membership, and impressive voter turnout rates. Here, it looks like the majoritarian model is a good predictor of support, since all of the links in the theoretical chain have been satisfied. Majoritarian elections promoted a two party system and one party, majority governments. These governments were accounta ble and passed policies that the people wanted and received support as a result. A major change in party and government accountability as well as a radical new policy direction began in 1984. The Labour manifesto, a vague, bland document printed publicly prior to the election in 1984, offered no real change in policy direction and included language that promised an active role for government (Mulgan 1995). When they stood for reelection they did not even bother to offer a manifesto, instead they printed o nly several pages of unpopularity, the National manifesto offered before the election of 1990, promised a the Labour policies of the previous six years. Rather than follow the mandate, both parties reversed decades of standing policy. From 1984 to 1990, Labour eliminated farm subsides, flattened the tax structure, sold off state a ssets, and significantly liberalized trade barriers. National gutted much of the social benefits that New Zealanders had enjoyed. The government drastically cut spending on health, welfare, education, housing, and social security. This change in policy direction, by both parties, was in direct contrast with what the parties promised before they were elected. This loss of accountability and policy had a devastating effect on support. Turnout, party membership, trust and satisfaction all declined during this period. New Zealanders also

PAGE 420

420 rejected the first past the post electoral system in 1993 in favor of MMP. Statist ical models run in Chapter 3 provide evidence that the change in electoral system can be attributed to the change in policies and the lack of respect the parties showed for their electoral mandates. Italy prior to electoral system change demonstrated similar variability in party system behavior. This inconsistency took place despite the proportional electoral system performing as expected. party system that is expected under the proportional model. Proportional elections in Italy offered virtually no barriers to representation and smaller, third parties were able to take advantage, winning seats. From 1948 to 1979, between 10 and 12 parties were represented in the chamber and the senate. Disproportionality was normally less than 1%, but on the rare occasions it did occur it was usually minimal favoring the largest party, the DC, by only 2% to 4%. The party system itself was ideologically diverse and included: centrist parties (DC), left wing parties (PSI and PCI), right wing parties (MSI and Monarchists), and a host of lay parties on the center right and ce nter left (PLI, PRI, PSDI). The initial steps of the proportional model were met, the electoral system produced proportional results and the party system contained ideological diversity. From 1948 to 1979, that ideological diverse party system was also a ble to foster representative processes and pass effective policy, which led to supportive attitudes. Generally speaking, coalition governments might be the place to look to determine whether decision making processes include more representative interests. In Italy that is not the case. Parties fostered representation in two different ways: by creating deep societal ties with their constituents and by practicing broad based, cooperative decision

PAGE 421

421 making within the legislature. The DC and PCI as the two lar gest parties in the system created a number of ancillary organizations to increase the connection with their constituents. These organizations touched almost all aspects of Italian life including: sports clubs, coffee bars, newspapers, movie houses, and t elevision networks. Within parliament the DC, as the largest party, was a part of every post war government and controlled the premiership until 1980. They formed minority governments or included one of the smaller lay parties or the PSI. Governments in Italy were not necessarily the place where representation was evident. This is due to the nature of Cold War politics and the exclusion of the PCI, the second largest party, from government. Decisions, prior to 1979 were made only when there was a large consensus. In order to generate that consensus, parties turned to a practice that was used during the liberal regime of the late 19 th and early 20 th centuries called transformismo. This is when elites set aside ideologies in favor of broad based coaliti ons in order to deliver for their respective constituencies. This was especially evident in the political behavior of the PCI. As the second largest party in parliament, the PCI could have effectively blocked much of the Instead, the PCI supported almost 75% and at times ma 1976; Spotts and Wieser 1986, 113). Transformismo was effective in calibrating government action to the changing political mood of the time. Evi dence of transform ismo took place during the 1960s and the 1970 s in two diffe rent respects. During the 1960 s massive amounts of legislation, supported by up to 80% of parliamentary forces, was passed through ften were clientelis tic in nature. During the 1970 s there was broader consensus and cooperation between parties, especially

PAGE 422

422 during the historic compromise between the DC and PCI. From 1976 to 1979, the DC governed with the support of the PCI as shadow me mber of the government. Representation, as a process, was generally met from 1948 to 1979. In order to get anything done in parliament, there had to be wide consensus between almost all political parties in the system. While not a perfect example of par liamentary government or the proportional model, parties in the Italian legislature were able to represent the interests of their constituents and pass effective policy as well. From the end of WWII to 1979, two types of policies were passed. The first, l eggine, were small laws that serviced a particular clientelistic base. Some hold that that there were a tremendous number of leggine passed, especially in the 1960 s, but the Italian legislature was able to pass a number of significant policy achievem ents as well. During the 1950s and 1960 s, the economy grew at an impressive rate and the co untry went through a period of rapid modernization. The Italian legislature, during this time, was able to build an impressive system of public transportation including highways and an extensive network of rail lines. The country also improved its system of education, illiteracy stood at ab out 90% after WWII, by the 1970 s it was all but eliminated (LaPalombara 1987, 163). The legislature also created a social welfare state through a number of legislative measures including: putting a floor on poverty, cr eating wage scales that narrowed the gap between rich and poor, as well as creating a program for univ ersal health care. By the 1970 s, many Italians were in an improved economic situation with higher wages, more social services, and a quality of life that

PAGE 423

423 saw many owning second homes and more vacation time. Because of this representation and policy success, there were a number of variables that suggest that people were supportive of the system. One of the indicators was the remarkable consistency of elec tion results, Italians really identified with the parties that they voted for. This was evident in the high levels of party membership and activity in the ancillary groups associated with the political parties. Voter turnout was also among the highest in the world, even though Italians had to travel to their hometowns to cast a ballot. Like New Zealand, changes in policy and process in Italy led to a massive decline in support, a collapse of the party system, and a change in electoral system. The proport ional electoral system continued to produce proportional results and allowed a cross section of ideologies to win seats in parliament. The other representative processes be gan to unravel in the late 1970 s, however. Citizens began to disengage in the anci llary organizations that were the basis of representation during the 1950s and 1960 s. Part of the reason was the growing secularism that hurt Catholic organizations and the replacement of communist organization by the private market. This detachment saw both large parties lose electoral support after the Historic Compromise fell apart in 1979. The loss of electoral support left the door open for the PSI to gain th e premiership in the early 1980 s. While the balance of power of the party system was changi ng, the parties themselves and the bureaucracy went through a generational change. This new generation viewed government as a business opportunity to improve their own economic standing. Wide spread corr uption exploded during the 1980 s, as the new party elites used the existing clientelistic ties to profit off the distribution of public goods. Another change in the process in government was the

PAGE 424

424 collapse of cooperative behavior characterized by transformismo. The PSI led governments in the 1980 s increasi ngly relied on the emergency decree to pass the small, clientelistic policies they could no longer garner the consensus required to pass through committee. Policy successes were few and far between during the 1980 s. Governments were able, through the us e of the decree, to satisfy their clientelistic bases but they could not garner the cooperation in the legislature necessary to achieve wider policy goals. Continued clientelism led to a massive increase in the public debt, but governments were unable to generate support for any austerity measures. This became more and more problematic as the decade came to a close. Italy was nearing joining the European Union and came under increasing pressure from other European states to get their public spending unde r control. By the time the Cold War ended Italy was mired in a series of corruption investigations, saddled large amounts of public debt, and declining levels of support. This loss of support is evident in a few areas. The public began to move away from political parties, both party membership and attachment to political parties declined. Satisfaction with democracy also fell into the low teens, a historically low level. Turnout also dropped ten points, and those who did vote supported fringe parties a t a higher level as they moved away from the DC and the PCI. The combination of these factors led Italians to vote close to 90% in favor of a change of electoral system in 1992. The evidence from New Zealand and Italy offers some evidence about the lack of connection between electoral systems and support. The first is that both proportional and majoritarian elections in both countries produced the results that they were

PAGE 425

425 supposed to. New Zealand had a two party system and Italy had a multi party system. These are the first steps of the model and the most immediate effects of the electoral system. As the effects become more distal, the more variable the results became. In New Zealand, parties and governments were both accountable while passing effectiv e policy and unaccountable while changing policy direction. In Italy, coalitions of parties were both able to cooperate and pass effective policies and be unrepresentative, corrupt, while impotent to address the pressing issues facing the country. The on ly constant came from the electoral systems, so it is only logical that the variability in party system behavior reflects the changing levels of support in both countries. This work also examines whether changes in electoral system could help correct some of the problems of the previous system. New Zealand sought to increase proportionality to increase representation and Italy looked to increased majoritarianism to improve accountability. Both reforms produced the desired consequences in terms of immedia te effects (number of parties), but saw similar levels variability in party system behavior that led to changing levels of support. New Zealand switched to a system of MMP to increase the levels of proportionality and provide wider representation. For the most part, this was largely achieved. This is evidenced by the drop in disproportionality from a high of 18.19 in 1993 to 3.43 in 1996, 2.97 in 1999, 2.37 in 2002, and 1.13 in 2008 2 In terms of the number of parties winning seats, they increased as wel l from only 2 in 1993 to 6 in 1996, 7 in both 1999 and 2002, and 8 in 2005. Ideological representation was also increased as well. Parties included those on the right (National, ACT, and United Future), the center (NZ First and United 2 Disproportionality here is measured by the least squares index where higher numbers represent higher disproporrtionality. Data acquired from Michael Gallagh er and Mitchell (2005).

PAGE 426

426 NZ), and left (Labo ur, Greens, Alliance, and Progressives). Considering the direct consequences of the increased proportionality of the new electoral system, New Zealand clearly has a more representative party system. But, as we move away from the more direct impacts of th e electoral system, the more variable the party system behavior became in terms of process and policy. Increased proportionality, according to the proportional model, should lead to increased representation in the decision making processes of government. This was not the case in New Zealand. Forming governing coalitions continually proved problematic. The first coalition formed between National and NZ First was formed amid great public controversy. Almost no one in the country wanted or expected those t wo parties to form a coalition. The coalitions formed after 1996 did not inspire as much debate, but parties remained loath to name clearly before the election which parties they would prefer to work with when the election was over. Another strange proce ss that undermined representation was the continued practice of party switching. independent parties (presumably for the increased salary and party funds entitled to under la w). The difficulty forming representative bonds with citizens also translated into the policy making arena as well. Policy never returned to the generous benefits characterized by pre 1984 levels. This is due to the fact that while parliament included mo re parties, governments (usually characterized by one large party and one or two smaller parties) still controlled the entire policy making apparatus. Governments, whether led by National (1996 1999) or Labour (1999 2005) simply did not want to return to the policies of the past. Some

PAGE 427

427 social welfare benefits were returned, including more money for education and neighborhood development, but unpopular welfare and social security policies remained in place. Only at the beginning of the new Labour and Allia nce government in 1999 did it seem as though activist government would return, but that policy momentum slowed. Government policy direction can best be described as piece meal after the adopting of MMP, characterized by half measures and centrist solution s, rather than the proactive, interventionist government policies that the public supported. Support for MMP was impacted by both process and policy. The preference for an alternate system to MMP rose almost immediately after the election of 1996, and onl y increased in the ensuing years. 1998 marked a low for supporting MMP. Only briefly in 1999 did New Zealanders favor MMP over alternative systems, this was the one year when government coalitions were formed smoothly and there were significant policy ac hievements. After another controversy over party switching and evidence that the policy direction was not going to lead to more generous social welfare polices, support for MMP fell in the subsequent years. Preference for alternative systems continued to be higher than support for MMP during the remaining years examined in this study. electoral system seemed to function as intended but accountability and policy achie vements of the party system were variable. I hold in Chapter 6 that MMM in Italy helped to produce bipolar competition. This was achieved due to the pressures the system put on both parties and voters. Parties, from 1994 to 2001, increasingly engaged in strategic behavior. This is evidenced by the number of parties joining one of the two center left (the Olive Tree) or center right

PAGE 428

428 (House of Freedoms) coalitions. In 1994, the Pact for Italy tried to form a center pole but was only able to win 4 seats an d was eventually subsumed by one of the two larger coalitions. This is evidence that learning took place among parties, and eventually the parties within the two large coalitions developed elaborate stand down agreements so as to not split their vote with in districts. This learning is evidenced by district level data, showing that the effective number of electoral parties, in this coalitions, was reduced from 3.07 in 1994 to 2.43 in 1996 to 2.4 1 in 2001 (Cox and Schoppa 2002, 1046). The two main coalitio ns were also able to significantly increase their vote share as well. In 2001 coalitions collected 89% of the votes and 98% of seats in the chamber and 82% and 97% respectively in the Senate. This was up from 80% of the vote and 92% of seats in the chamb er and 67% and 85% in the Senate 3 The change in electoral system had mixed results in terms of producing accountable and effective government however. Italians adopted MMM to increase the level of accountability in the country. They wanted to create bipolar competition, but they also wanted alternation in office. The ability to vote out an underperforming government was something that was lacking under the proportional electoral regime used in the First Republic. In this regard, the electoral syste m change was a success, a center right government was elected in 1994 (which quickly fell apart after 8 months and replaced by a technical government) and lost the ensuing election to the center left in 1996. The Olive Tree was then defeated by the House of Freedoms in 2001. For the first time, Italians had a real choice 3 Despite the success of the electoral system in encouraging bipolar competition, the coalitions themselves remained fragmented. It is difficult to say whether, over time, coalitions would remain electoral cartels or become catch a ll parties because MMM was only used for three election cycles. learning that I describe here suggests that it was at least a possibility.

PAGE 429

429 between alternative governments and prime ministers. Another positive was the turnover that took place in parliament both in the senate and the chamber. This could be partially attribut turnover can also be attributed to the collapse of the old party system). Another key process, perhaps attributed to this increased accountability, was the general feeling among Italians from 1996 to 2001 that corruption was going down. I mention that since corruption was such a major factor in the loss of support in the electoral regime of the First Republic. These improvements in process were coupled with policy achievements as well. T he Olive Tree was elected in 1996 and was able to keep many of the pre election promises they made. The major accomplishment of this government was successfully meeting the fiscal criteria for entry into the European Monetary Union. They passed a series of austere budgets in 1996, 1997, and 1998 that curbed the amount of public debt and the budget deficit. When Italy was accepted in the EMU, it was a great policy achievement for the Olive Tree government. Another achievement was a series of bills overha uling the excessively bureaucratized public administration. The Olive Tree eliminated a number of unnecessary rules and regulations as well as conflicts that made dealing with the bureaucracy confusing for the public. Much responsibility was devolved to the regional and local level, giving communities a degree of flexibility that did not exist previously. Perhaps the most important part of the policy successes the Olive Tree had was the fact that prior to the election they stated clearly what they inten ded to do. When they were elected to office, they made an honest effort to enact those policies. This is

PAGE 430

430 the kind of accountability that was lacking during the First Republic. Satisfaction with democracy increased from around 30% to over 60% from 1994 t o 2001. I believe the increased accountability, policy successes, and feelings that corruption was going down can be attributed for this improvement in support. Following 2001, satisfaction with democracy fell down below 1994 levels. The reason for this is the decline in accountability, feelings that efforts toward eliminating corruption had failed, and the inability of the center right government to pass much of its policy agenda. Silvio Berlusconi returned to the premiership in 2001 after his House of Freedoms coalition won a large victory over the center promised to deal effectively with corruption and his own personal conflicts of interest. Rather than tackle these problems, his government actually blocked ef forts to prosecute corruption, passed legislation that made it easier for businessmen to avoid prosecution, and did not really deal with conflicts of interest at all. This was coupled with the slowing of parliamentary turnover and the professionalization of a new political class. Citizens reacted unfavorably to these developments and people began believing that old, corrupt behavior had returned. Despite large parliamentary majorities, the center right had a difficult time passing its policy agenda, effo rts to reform pensions and stimulate the economy ultimately failed. As a result of the loss of accountability, feelings that corruption had not been dealt with, and a lack of policy achievement satisfaction with democracy feel in Italy. The evidence from New Zealand and Italy before and after electoral system change informs us in several ways. The first is that electoral systems can shape the type of party system that a country has. Majoritarian systems produce pressures

PAGE 431

431 toward a two party system. Incre asing majoritarianism in a country has a similar effect. Proportional systems and efforts to increase proportionality ease the path for smaller parties to gain seats and increase the level of ideological representation in a party system. This is what bot h the majoritarian and proportional models expect. Electoral systems have less of an impact on the type of behavior party systems engage in however, and this is where the models fail. Institutions can shape or encourage certain behaviors, but they cannot guarantee those behaviors. It may make sense for governments to be accountable in an increasingly majoritarian system, but that does not mean that they will. The same is true for increasingly proportionality. Just because more parties are represented i n a system does not mean that representation will be put into action once those parties get to the legislature. The Electoral systems featured in this study produced fairly consistent results, but the party system behavior varied rather widely. This vari ation in the party system behavior is why support varied in those countries as well. Just because no credible link between electoral systems and support is evident does not mean that our inquiries should stop. There were several instances, some for long periods of time, where a credible link could be drawn from the election to the party system, to party system behavior, to the types of policies that were passed, then finally support. I believe we can learn a great deal about the nature of elections and support in those instances. Specifically, there are two linked conclusion. One conclusion that I draw is that the processes of accountability and representation should be thought of as expectations rather than consequences of particular elections. Elect that go with them and the public expects that those ethics will be followed. Ethics may

PAGE 432

432 be expected and encouraged, but that does not mean they will take place. The second conclusion I make is that it is very difficult to differentiate policy and process when examining the consequences of elections. Meeting the ethics of representation and accountability necessarily includes following through with some type of policy. The following passage looks at these two conclusions in more detail Lessons from New Zealand and Italy I believe the evidence presented in this project demonstrates that electoral systems have little to do with support. A critic might suggest that I have built elaborate theories drawing suspect links from different elec toral systems and support only to refute them. It is not my intention to merely tear down theories linking electoral systems and support. In fact, I count myself among those who believe that the critiques of pure proportional and majoritarian systems to be overstated. This belief is best represented Member Systems: The Best of Both Worlds? contain[s] within themselves the preexisting and the road to electoral reform is open (Shugart and Wattenberg 2001: 27). The evidence presented in this work sugges ts otherwise. Elections have particular ethics, or presumed outcomes. In proportional elections that ethic is representation and in majoritarian elections the ethic is accountability. When Italy and New Zealand were able to meet the expected ethics of t heir elections, they had a great deal of success in passing policy and generating support. This goes the same for efforts to increase majoritarianism and proportionality. Examining how New Zealand and Italy were successful under different electoral rules provides evidence for this conclusion.

PAGE 433

433 Accountability as an ethic of majoritarian elections has its roots in the literature on electoral systems. The presumption is that majoritarian elections offer the most direct electoral connection between voters and governments, which provides a few opportunities. Voters have the chance to choose between clear alternatives and judge the winner by the extent to which those promises are met. Parties have the ability to promote a particular agenda, have it judged, and a chance to enact it (Duverger 1954, 1984; Powell 2000). Here the connection between accountability and policy is clear. In New Zealand prior to electoral system change, this relationship was in place. Parties presented manifestos prior to the election and then followed through. Accountability was lost after 1984 when both Labour and National decided to pass policies that were in direct conflict with their manifesto or not mentioned at all. The loss of the electoral mandate through the removal of the manifesto and a change in policy are statistically connected with support for changing the electoral system. Something similar happened in Italy, the electoral system change had a number of goals, but the underneath all of them was a desire to increase a ccountability. The system had its most success when accountability was high. It was high when there was alternation in office between center right and center left and when voters had a chance to vote out of office a generation of elites that were corrupt The center left elected in 1996, entered off with a clear policy agenda spelled out in its coalition agreement. The government was successful in passing their economic policy goals, several important administrative reforms, and had convinced the countr y that the level of corruption was going down. Satisfaction with democracy was at an all time high of 61% when governments were accountable by passing the policies that they promised.

PAGE 434

434 Accountability suffered under the second Berlusconi administration tha t started in 2001. The center right presented to the voters a clear policy agenda that the presumptive through on the promise however, and satisfaction with democracy fell down below 30%. The evidence presented here demonstrates that systems that are wholly majoritarian or seek to increase majoritarianism are only successful when they have accountability. This means that there are clear sets of alternatives for the voter s to choose from, one winner, and that winner pursues the policy agenda they laid out before the election. When one of those objectives fails support lags as a result. Something similar can be said of representation under proportional or systems that see k to improve proportionality. Representation is meant to incorporate more ideas into the policy making arena and produce better policies as a result. When either of those qualifications is not met, support suffers. Representation is the key ethic in prop ortional systems. The idea developed by scholar, and represented in the proportional model, is that proportional elections will try to represent the population as accurately as possible. This means that there is less of a connection between the election and how decisions get made. Parties decide after the election how to organize, make decisions, and pass policies. The most popular way of doing this is the formation of a government, but the expectation from the public is that their interests will be rep resented (Lijphart 1984; Powell 2000). In Italy prior to 1979, parties were generally able to agree to pass policies to service their respective constituencies. They did this through broad based cooperation as evidenced by the extensive use of the commit tee sys tem in the 1950s and 1960 s. They were also able to

PAGE 435

435 pass significant amounts of legislation through conventional legislative procedures as well. When parties satisfied the ethic of representation, they passed policies that produced public support. In t he 1980 s, representation to a large degree was lost. The broad based cooperation that characterized the previous decades was replaced by a more centralized government. This government was unable to pass significant legislation and relied on a proced ure, the emergency decree, which was not cooper ative at all. By the late 1980 s, satisfaction with democracy fell down historic lows. The connection between representation and policy seems to be an important factor in the level of support. The same can al for adopting this system was to increase the amount of representation in parliament, since it had been dominated by one of the two main political parties for the previous 100 years. There was on ly one brief period where MMP was favored over alternative systems (evidence of support). This was in 1999, when Labour and the Alliance campaigned together, clearly stating that they would form a government following the election. When they were elected at the outset they passed polices in the areas of social welfare, which was supported by the public. New Zealand political parties generally struggled with representation under MMP. There were frequent governments formed that did not consider the publi in 1996 and 2005. When the primary decision makers are unrepresentative the policies they pass are not going to be fully supported by the public. This is why MMP continually lagged behind alternative syst ems in public polls. What I take from this is that increased ideological representation is not enough. Just because more parties are represented,

PAGE 436

436 does not mean that they are going to be involved in decision making, or that the policies passed are going t o be representative of the electorate. The main lessons I believe we can take from the successes and failures of the majoritarian and proportional models is the importance of the ethics of accountability and representation and the connection to those ethi cs and policy. Majoritairan and proportional elections (or increased majoritarianism or proportionality) do not guarantee accountability and representation, but there is an expectation that they will. Implicit in the satisfaction with the ethics of accou ntability and representation is the delivery of policy. Without policy, accountability and representation lose their meaning. Accountability requires that parties or candidates promise something then be judged on whether they deliver. Absent the promise and delivery, there is no accountability. Representation is similar. It requires that interests and ideology be represented, not just in theory or policy positions, but policy action. The ethics of elections signal how elected officials, parties, and g overnments should behave. Majoritarian, or increased majoritarian, elections encourage accountability by picking a winner based on policy promises then keep or remove that winner based on how they keep those promises. Proportional, or increased proportio nal, elections encourage the representation of interests in decision making processes and the production of policies keeping all of those interests in mind. The variability of party system behavior in New Zealand and Italy before and after electoral syste m change demonstrates that these electoral ethics are not guaranteed. The research presented in this work offers a few reasons why the ethics of accountability or representation may be met at times and why they might be ignored.

PAGE 437

437 Contributing and Confoundi ng Factors Elections are not enough to guarantee that individuals, parties, and governments will behave accountably or in a representative manner. Evidence from New Zealand and Italy can help explain the variability in this behavior. I identify in this p assage the factors that can contribute or confound the ethics of elections. Elections can help shape political actions, but they do so within a larger political context. Economic conditions, scandals, or the emergence of a new party system can certainly have an impact on the ability of political actors to be accountable or representative. Another factor is institutions outside of the electoral system. Institutional rules like the use of committees to pass laws, caucus voting rules, and bicameral or unic ameral arrangements can also matter a great deal. Finally, there is the role of agency. Ultimately, the decisions that are reached are made by people and their actions can have either contribute to achieving the ethics of elections or prevent those ethic s from being met. This passage discusses contributing and confounding factors which include: political context, institutional factors, and agency. Political Context as a Contributing and Confounding Factor Much of this work has spent time discussing h ow electoral system shape political outcomes like party systems, governments, and public support. I should note that electoral systems themselves are imbedded in a political context that may aid or hinder the ability of that system to produce expected pol itical outcomes. Robert Moser in several works describes how electoral systems in post communist democracies did not always conform to Duvergerian expectations. For example, in many cases the number of political parties was higher in the SMD tier of mixe d systems. He attributes these

PAGE 438

438 party system institutionalization (Moser 1999, 2001). I describe in earlier passages how prior to electoral system change, both New Zealand and Italy had measures of success t and Wattenberg 2001, 27). I have described their success in terms of the models I draw in Chapter 2, but the success of the systems coincides with a fa vorable political environment. When that political climate changed, the systems were not able to produce that same kind of positive outcomes. Political context is also important for understanding why the changes to mixed systems were ultimately less than successful. Cha pter 3 and Chapter 4 describe how electoral systems in New Zealand and Italy helped cultivate support among the population, what changed, and how support declined as a result of those changes. There are important political contexts to note that also help to understand why the changes took place when they did. The high point for support in New Zealand, as described in Cha pter 3 was during the 1950s, 1960s, and early 1970 s. During this time, the New Zealand economy was in a very strong pos ition though their favorable trade relationship with Great Britain. The relationship accentuated the natural agricultural advantages enjoyed by New Zealand, and the economy went through a steady period of growth. With this financial security, the two mai n political parties, Labour and National, competed over how best to deliver the rich spoils this prosperity brought. This is probably evidenced best by the types of policies both parties supported for decades including: universal health coverage, free pub lic education, protectionist trade policy, and generous farm subsidies among other policies. It is no coincidence that during this period New Zealanders felt that they had one of the highest qualities of life in the world. In a sense, it is easy to meet the criteria laid out in

PAGE 439

439 the majoritarian model, especially accountability and good policy, when the country is in a favored economic situation. Parties did not have to stake out differing policy prescriptions or offer the public tough choices. Both part ies favored generous social and economic policies and then delivered them after the election. It is when the anged in the middle of the 1970 s that accountability and when Great Britain ended their favored trade relationship when they joined the European Community. This event also coincided with a series of oil shocks that raised the price of petroleum dramatically. Together, these had a dramatic negative impact on t he New Zealand economy. Political elites were slow to adjust to the new economic reality and turned to the policies that characterized the previous decades. Massive public works projects and federal spending led to an increasing public debt and inflation By the time Labour was elected in 1984, they faced a dire economic situation. I am not excusing the behavior of Labour politicians. According to the ethic of majoritarian systems, in order to maintain accountability they should have made their case du increased public spending in a time of economic downturn was not the right path. This might be easy to say in hindsight, but the point that I would like to make that it easy to be accountable when what your promise contai ns only good news. Politicians in New Zealand failed at this when the new they had to deliver was bad. Italy prior to the 1970 s also faced a particular political context that at times helped foster support and at times did not. Like New Zealand, the heig ht of political support in Italy was during the 1950s and 1960 s. This also coincides with a rapid period of

PAGE 440

440 time that the Italian parties were able to represent their con stituents and make government work. Italian parties turned to committees to pass clientelistic measures known as leggine to deliver spoils to their particular group. In this respect, Italian democracy was similar to New Zealand democracy. When there wer e plenty of resources to divvy out, parties and governments were able to cooperate in order to do so. It was when the economy slowed in the 1970s and 1980 s that cooperation become more difficult. Italy was also impacted by the oil shocks of the periods, but had additional pressure from the European Community to get its financial house in order. During a brief period the DC and PCI were able to govern together to overcome those difficulties, but this behavior did not continue. Cooperation was basical ly n on existent during the 1980 s, as governments turned to the emergency decree to satisfy their political clients. The lesson here, like New Zealand, is that it is easier to be representative and deliver policy that the public supports when the economy is gr owing. Outside of the economy, there are other important political contexts to consider in Italy. There are also two important things to consider in Italy as relates to political contexts: the Cold War and the ideological fragmentation of Italians. In ma ny ways, the story of Italian history can not be told without reference to the Cold War. In fact, the entire political order of the Italian First Republic was organized based on the conflict as shut out of every post war government. The party was still able to influence policy by coming to agreements with DC led governments. As long as the Cold War continued Italians were seemingly stuck with the DC as the largest party and a central part of every government. Take a counter factual where there was no Cold war. Would there have

PAGE 441

441 been a grand coalition between the two largest parties the PCI and DC? It is of course impossible to say, but considering that both parties did exhibit a large amoun t of 1970 s, it is not inconceivable. It is only when the Cold War ended in 1989, the PCI disbanded and reformulated itself into the PDS. It is not coincidental that this c oincided with the dramatic rise in corruption investigations and prosecutions and a rapid decline in support among the population for the existing political order. When the threat of the PCI and communism was removed, the logic of the status quo forwarded by the DC and PSI was gone. The second consideration that impacted how proportional elections were able to meet the ideals of the proportional model was the level of ideological polarization among the Italian electorate. Italian politics includes commu nists, fascists, monarchists, Christian Democrats, socialists, and secularists among others. The proportional electoral system did not create this political reality it only allowed it to be represented. In fact, the adoption of this proportional electora l system in 1946 was advanced by these existing ideological interests to guarantee that they would receive representation in the forthcoming matching the expressions of the p artisan and ideological preferences of the public as close as possible. What the elections do not do is choose what the subject of that photograph is. In Italy, proportional elections allowed the ideological polarization that existed in the country to be represented in the legislature. The distance between the parties certainly made the attributes in the proportional model like cooperation more difficult at the outset. At times, the parties were able to overcome their ideological

PAGE 442

442 differences, but more o ften than not they resorted to the lowest common denominator like small, clientelistic policies targeted at a narrow interest. Truthfully, the fact that Italian democracy had the success that it did is remarkable considering the real ideologically divisio ns in the country. The previous paragraphs describe political contexts that both aided and hindered the ability of the electoral system to be linked to supportive attitudes before electoral system change. Political context was in important factor in both countries following the change as well. Adopting a new electoral system is not like hitting a reset button. Both Italy and New Zealand took significant baggage with them into the new electoral regime. In New Zealand, electoral system change took place a gainst the backdrop of tremendous anti party sentiment. One of the main reasons for the change as I describe in Chapter 3 was the belief that elites were in the habit of making promises and then not following through. The initial optimism that greeted MMP was immediately lost after a contentious, secretive coalition negotiation resulted in a New Zealand First and National coalition government. There was widespread belief among the electorate that Winston Peters had signaled clearly his antipathy for Na tional and that his party would not join National in a coalition. The fact that there was a coalition negotiation where a party leader took a good deal for himself and his party should not have been unexpected. There was plenty of commentary that this co uld be a possibility when the electorate voted for MMP after a lengthy campaign before the referendum. The reason voters reacted so vehemently was the remnants of the distrust that precipitated the electoral system change in the first place. The issue of party switching also caught the

PAGE 443

443 parties to form independent parties with themselves as the leader. Again, like the first coalition formation process, party switchin g reinforced the belief that the word of a politicians was not worth anything. Both the coalition formation and party switching are linked to the decline in support for MMP. It is worth wondering how much these events might have mattered in the absence o f the deep anti party sentiment that was left over from what occurred before the electoral system change. I hold in Chapter 5 that coalition formation and party switching helped undermine representation, a key process if New Zealand was to be more proport ional. I believe that is true, but political context also mattered as well. Italy also presents evidence that new electoral regimes have to deal with leftover political baggage that precipitated the change. Some of the political context that might have i mpacted the level of political support in Italy include: the wave of corruption scandals, the history of party fragmentation, and the pressures of joining the EMU. Italy adopted a new electoral system amidst a wave of scandal. The existing party system w as completely dismantled as political elites were embroiled in the Clean Hands investigations into the widespread corruption that permeated all features of the political class. I mention in Chapter 6 that I believe support was going to be positively or n egatively impacted based on whether citizens felt that the issue of corruption was improved. While the first center left government made some efforts toward improvement in that area, it is likely that the scandals themselves had more of an impac t. As I d iscuss in Chapter 5 there was tremendous turnover in elected officials in the first two elections in 1994 and 1996. This turnover slowed in 2001. It is possible that the improved attitudes toward corruption and higher levels of support are not due to an y affirmative action by new governments, but by some other

PAGE 444

444 mechanism like judicial investigations. Another issue related to political context, is the history of party fragmentation in Italy. Moser (2001) names the strong, historical tendency of Italian s to support smaller parties as one of the reasons that the party system remained fragmented despite the majoritarian pressures of the new electoral majoritarian system wi active and persistent force in Italian politics. Both authors attribute this pull to a long political history in Italy and entrenched tendencies to support smaller parties 4 A final note on political context in Italy has to do with the influence of the EMU on some of the successes of the first center left governme nt. I attribute, in Chapter 6 the connection between the new, more majoritarian electoral system and these policy successes, espe cially the center admittance. The electoral system had a role to play by encouraging pre elections coalitions and providing a majority in both Chambers. Part of the center left cess in steering the country into the EMU was at least in part attributable to political context. Entrance to the EMU was widely supported by the public and political elite. Italy was one of the founding members of the European Community and failure to m eet the standards set would have been a great 4 This is a contested point. Parties did engage in a tremendous amount of strategic behavior under the newer, more majoritarian rules. This strategic behavior was most often represented by stand down agreements between parties that were a part of a larger coaliti on. With respect to Moser (2001) and that in the long term voters would not abandoned smaller parties as Duverger (1954) suggests. Large parties were not willing to take on smaller parties in the short term by risking having an ideologically similar, small party steal some of their vote. This does not mean that learning would not take place over time as Duverger (1954) suggests it will. Unfortunately since there were only three elections it is impossible to determine if the number of small parties would have shrunk, any suggestions otherwise is just conjecture.

PAGE 445

445 embarrassment. So, there was tremendous internal and external pressure of the Italian legislature to get things done quickly. This is not to say that the new electoral system had nothing to do with the govern attained, policy success was much more difficult to find. This applies to both the center left and the newly elected center right in 2001. Considering that the PCI DC compromise gove rnments of the 1970 s had s imilar success passing austerity measures under similar international pressure under a proportional system, political context is something that should not be ignored. The success or failure of electoral systems to produce the conditions that allow for high er levels of support is greatly influenced by political context. In a sense it is easy to be accountable (majoritarian) or representative (proportional) and deliver policies that the public supports when the economy is growing. Both, New Zealand and Ital y had their greatest successes when this was the case. It is increasingly difficult to meet those standards during tougher times. Also, Italy and New Zealand demonstrate that electoral system changes do not happen in a vacuum, the new electoral regime ha s to deal with the remnants of the old one. Actions by the new governments under new electoral rules are judged with the memories of the old regime. In New Zealand, the difficult first coalition process and party switching may not have had the detrimenta l effects they had if it were not for the deep anti party sentiment that preceded the moves. In Italy, the initial success of the reforms (higher levels of support) might be attributable to the wave of political elites swept form office after the Clean Ha nds judicial investigations. Political context has an impact on the ability of elites to be accountable

PAGE 446

446 or representative, the policy achievements, and how the public frames their actions and gives support accordingly. Political context is not the only fa ctor that helps or hurts the connection between electoral systems and support. Electoral rules help pick winners and losers on Election Day, but those winners and losers have to operate in an institutional context. There are a number of institutions and rules that mattered in New Zealand and Italy and these institutions had a contributing influence on the level of support. Institutional Rules as a Contributing and Confounding Factor Chapter 1 of this project outlines the variables that can impact support among these are the types of policies that are passed and processes by which the polit ical elites behave. Chapter 2 is a discussion of how particular electoral systems can help to foster certain processes and policies that lead to higher levels of suppo rt among the population. What became increasingly clear throughout my research of New Zealand and Italy is that there are a number of other institutions outside of the electoral system that can influence the processes and policies of a given country. The re is the thorny presence of mixed electoral rules in New Zealand and Italy. The theoretical component of this work is drawn from the literature on proportional and majoritarian systems and it is framed in terms of what happens when either proportionality or majoritarianism is increased. Mixed systems fit into this framework only if they adhere to the logic of the political impact of electoral systems of Duverger and other electoral system scholars (Duverger 1954; Rae 1971; Riker 1982 ). As I describe in Chapter 2 scholars are divided on this point with some suggesting that there are contamination effects from having mixed electoral rules which can impact the politica l outcomes (Nishikawa and Herron 2001; Ferrara, Herron, and Nishikawa 2005). The evidenc e from New Zealand

PAGE 447

447 and Italy following the electoral system changes will not provide a definitive answer to the contamination question, but the some of the outcomes of mixed rules in New Zealand and Italy are worth noting. In this segment I detail some of the institutional rules that aided and hindered policy and process and, by extension, support in both countries. New Zealand, prior to 1984, seemingly had all of the institutions necessary to promote majoritarian outcomes. That is, the electoral system w as plurality, contested in single member districts. The party that won a majority of seats in the legislature formed a government with no outside institutions to contend with. There was no second chamber, Supreme Court, or constitutional rules (that coul d not be changed by a majority of the parliament) to prevent that majority from governing as it s aw fit. I note in Chapter 3 that governments increased accountability by relying on the back bench, the caucus, and interest groups to help curb the tendency of one party control of the entire ). In 1984, led by the Minister of the Treasury Roger Douglass, the Labour government instituted several small rule changes that consolidated power into the hands of a small number of influential members of the government. One of these changes was to require the government under the threat of being expelled from the party. Another rule change instituted the practice of majority voting within a smaller number of government ministers to determine the policy direction of the government. These two small rule changes, adopted by the National government in 1990 as well, helped a small,

PAGE 448

448 test fo r these changes in Chapter 3 feelings about the importance of the caucus system as it related to their support for an alternative electoral system. It maybe that small rule changes like the ones described are just not on the radar of average citizens. This does not mean that we should ignore rule changes however. These new rules allowed motivated ministe rs to consolidate power in the hands of fewer and fewer people and allowed the major reforms of policy to pass following 1984. As I describe in Chapter 3 these policies certainly had a role in the declining levels of support nation wide and the change in electoral system. Italy also had institutions and rules, outside of the electoral system, that both aided and hindered representative processes and good policy formation. One of the more important institutional features of the Italian Legislature (provid ed for in the Constitution) was the extraordinary powers granted to the committees. Committees, after granted the purview by the parties inside and outside of government, were able to bypass the normal legislative procedure and pass laws without going bac k to both chambers for full approval. This constitutional feature of the Italian system was vital to the ability of the Italian legislature to govern. Scholars repeatedly mention this feature of the Italian system as important, especially in aiding the p represent the interests of their constituents and pass policies that helped keep the system moving forward. If committees are a good example of institutional rules that helped aid representation and policy, the emergen cy decree is an example of a rule that undermined the process of representation. The emergence decree allowed the government to replace the role of the committee as the means of passing micro legislation, but without the consensus and cooperation that typ ified the use of the

PAGE 449

449 committee. Another institutional feature of the Italian legislature was the use of the secret ballot for roll call votes. This practice was increasingly used by D C back benchers during the 1980 s to sabotage the PSI led government tha t their party supported. Again, this institutional feature helped make policy making more difficult and helped contribute to the less representative use of emergency decrees during the period. These are examples of pre reform institutional impacts. Ther e is also the question of the potential confounding impacts of mixed systems adopted by both New Zealand and Italy. Also, several other institutional features of the systems should be noted, particularly the lack of change in any other institution of gove rnment. Before discussing the role of a change to mixed rules in Italy and New Zealand (rather than full changes to a majoritarian or proportional system), it is important to take a minute to discuss the institutional changes that were not made in either c ountry. Both New Zealand and Italy made electoral system changes, but did not buttress those designed to create a system of bipolar competition, reduce the number of parties, and help create a government that could act decisively to pass effective policy. In essence, the goal was to create a more majoritarian system. But, they did nothing to the other institutions of government. For instance, Italy maintained its bi cameral legislature. This type of legislative arrangement slows down the policy making process, promotes consensus, and makes passing effective policy more difficult than a one house legislature (Lijphart 19 84). I point out in Chapter 6 that policy whic h would have addressed conflict of interest and the prosecution of corrupt behavior passed one chamber but failed to pass the other. Broader efforts to reform the constitution also

PAGE 450

450 failed in the same manner. Even with large majorities in both Chambers, t he Berlusconi led government elected in 2001 had a difficult time passing its legislative program. Frequently, the government was able to negotiate one chamber only to have its momentum stopped in the other. The reason I mention this is that there are m ultiple ways to improve efficiency and accountability in government. Changing the electoral system is certainly one of those means, but it is not the only one. New Zealand also changed its electoral system, but like Italy made no other changes to its con stitutional order. New Zealand has an interesting constitutional order. As Lijphart describes, it is an example of a majoritarian system with a single house, no written constitution, and prior to the electoral system change single member district pluralit y e lections. During the late 1980 s when there was an ongoing debate over the electoral system, other constitutional changes were discussed as well. One of the suggestions was to add a second house to the legislature. This proposed Senate would have simi lar powers to the old House of Lords in the British system, where Senators would have the ability to affirm legislation, provide advice, and slow the legislative process down. This reform never came to fruition, but it would have been one mechanism to red uce the majoritarian nature of the system. The electoral system change increased proportionality, but no other institutional changes were made to assist achievement of the goals of that change. New Zealanders were looking to include more voices in the sy stem. They did change the electoral system, but other institutional adjustments could have been made as well. A final note I would like to make about institutions in New Zealand and Italy are the

PAGE 451

451 potentially confounding political outcomes of the mixed na ture of the electoral systems adopted in both countries. I hold in the body of this work that the mixed systems provided New Zealand and Italy both increased proportionality and increased majoritarianism. I will not reiterate those arguments here, but I w ill offer some contrary evidence to consider. There are some signs that there were contamination effects present in both countries. In New Zealand, the risk was that the tendency toward a two party system in the SMD tier would contaminate the PR tier of the electoral system favoring Labour and National. party dominated MMP system (Duverger 1986). Scholars generally reject the notion of contamination effects in New Zealand (Barker et al. 2001; Karp et al. 2002; Vowles 2005; Karp 2009). There should be some pause, however, considering that Labour and National remain the two largest members of the party system. Both parties made up 67.5%, 73.3%, 64%, and 81.6% of the seats in the elections of 1996, 199 9, 2002, and 2005. Labour and National dominated the electoral tier of the mixed system as well. They claimed 86.1%, 94%, 95.6%, and 89.8% of the electoral seats in the same election. Despite the evidence that New Zealanders split their vote between ele ctoral and list tiers, voting strategically on the electoral ballot and genuinely on the list ballot, the continued domination of the two largest parties should be noted 5 Another feature of MMP in New Zealand that should clause for the list threshold. Parties had to attain at least 5% of the vote to be awarded seats, unless they won at least one electoral district. In 1999, Winston Peters, the leader of the NZ First Party and of Maori descent, was 5 For evidence of vote splitting in New Zealand, see Karp et al. (2002) and Karp (2009).

PAGE 452

452 able to hold on to a Ma ori district and as a result his party was awarded 4 list seats. This allowed NZ First to stay in parliament despite is tremendously unpopular performance in government over the previous 3 years (their percentage of the list share fell from 13.35% to 4.26 %). Without this feature of the system, Peters and his party would not have been around to influence the coalition formation in 2005. The presence of single member districts in this case gave a nationally unpopular politician an anchor. There is perhaps more evidence of a contamination effect in Italy. The goal of the electoral system change was to reduce the number of political parties and create a system of bipolar competition. While the later was largely achieved, the former proved more problematic. The two coalitions on the center right and the center left increased their total percentage of seats in the Chamber from 80% of the vote and 92% of seats in 1994 to 89% of the vote and 98% of seats in 2001. In the Senate these numbers increased from 67% of the vote and 85% of seats in the Senate in 1994 to 82% and 97% of seats. These numbers do not accurately reflect the number of parties or groups winning seats in the Chamber and the Senate respectively. The number of parties winning seats in the Chamb er was 14, 11, and 14 in the elections of 1994, 1996, and 2001 (Bartolini et al. 2004: 10). Parliamentary groups in the Senate were also high in the same elections numberin 2004: 9). There was a sign reasons. The largest parties generally found it easier to cede safe districts to a member of their larger coalition in order to prevent losing the district to the other coalition. This wa s a fear because smaller parties retained loyal, regional support and could count on enough voters to throw the election to the other side. Therefore, due to their electoral

PAGE 453

453 blackmail power, smaller parties were able to negotiate seats for themselves prio r to the election (Bardi 2007). This is not the type of contamination effects that scholars expected however (Cox and Schoppa 2002). The expectation is that small parties in and have more success in the list tier as a result. This is not what occurred in Italy, so what are we to make of these conflicting results? The answer is not readily apparent. There is evidence that some measure of Duvergerian learning was taking plac e with increasing numbers of parties organizing into one of the two largest coalitions, but the number of those parties remained high. The hope was that over time, the coalitions would consolidate into two large catch all parties. Some scholars were skep tical about the ability of the Italian party system to the electoral system changed again in 2005, this argument will not be settled. It is true that Italy has a his tory of party fragmentation, and the coalitions were fragmented after electoral system change, but does that preclude consolidation into a two party system? The nature of catch all parties is to include various factions that broadly fit into an ideologica l characterization. For instance, the Republican Party in the United States has elements of religious conservatives, libertarians, secular business interests, and foreign policy hawks. Within that party there is a blend of policy preferences that blend s eemingly irreconcilable views that advocate government action and centralized power as well as decentralization and a hands off approach. The only point I am trying to make here is that ideological purity is not a prerequisite for a political party. It i s

PAGE 454

454 certainly conceivable that over time the party system in Italy could have consolidated into two catch all political parties. Institutions outside of the electoral system and the mixed nature of electoral rules have important impacts to consider in any di scussion of the connection between electoral systems and support. Institutions outside of the electoral system can help policy making or promote good processes. For instance, the unique power of the committees aided both policy making and cooperation in Italy, while the secret ballot became a mechanism to undermine cooperation. It New Zealand, a small change in caucus rules in the Labour Party helped pass a dramatic change in policy starting in 1984. Also, I argue that mixed rules are one mechanism to i ncrease proportionality or majoritarianism. It may also be true that there are potential contamination effects that should be considered. The role of institutions is important, but individuals have to operate within those rules. In both Italy and New Ze aland, individuals have played a decisive role in the execution of policy and process, and as an extension support among their respective populations. Agency as a Contributing and Confounding Factor One of the persistent themes throughout this work is the consequential role that the actions of individuals had on support in both New Zealand and Italy. The models that I draw linking electoral systems to support, as well as many institutional theories in general, count on rules to shape behavior. Even the m ost well designed institutions and rules count on individuals or groups of individuals to navigate and operate within those contexts. Individuals are often faced with a degree of latitude when it comes to their actions and choices. While researching both countries, I thought about a work on a

PAGE 455

455 new democracies. He notes that the actions of people in political office and other He finds examples of individuals who perform deliberately, openly, and admirably even amid widespread institutional failings and corrupt behavior by peers. These honest leaders can inspire others to follow thei r example and create favorable public opinion as they reach the top, they too often fai make a decision. There is ample evidence of his point, however. New Zealand and Italy are full of examples of leaders who, when faced with choices, either behaved with the responsibility of their office in mind or acted in their own immediate, self interest. What is clear is that when leaders behaved with the ethic of their particular system in mind, they more often than not, achieved results that helped increase support among the population. The inverse of this statement is also true, when leaders betrayed that particular ethic, the public responded unfavorably. For instance, the principle process behind majoritarian systems, as I discuss throughout this work, is accountability. Whenever leaders, through their actions, fostered this result the public responded positively. When they were not accoun table, support suffered. In proportional systems, the main ethic is representation. Citizens responded to the extent that leaders encapsulated that ethic of representation. There are plenty of examples of both to draw from.

PAGE 456

456 There are examples of leaders behaving accountably and less so in New Zealand before electoral system change and in Italy after elector al system change. Chapter 3 describes how important the concept of the mandate was in New Zealand. One of the features of accountability was politic ians stating clearly and decisively what they were going to do when they were elected in a pre election manifesto. This was more of a tradition that candidates for prime minister kept for a long period of time, than an institutional rule required under el ection law. Labour Prime Minister James Kirk kept a made. National Prime Minister Robe parties shared this ethic (Mulgan 1995). Starting in 1984, successive Labour and 1984 and National in 1990, or eschewed the tr adition altogether as Labour did in 1987. Abandoning the tradition of the mandate came down to a political choice made by leaders and both parties and ran contrary to the expectations of accountability that the public had grown to expect. As I discuss ea rlier, political context is a contributing factor about why this might happen, but it does not explain fully why elites choose the strategies they do. What kept the Labour party from leveling with the people of New compl etely sure. Chapter 3 simply demonstrates how jettisoning the mandate was detrimental to support in New Zealand and precipitated the electoral system change. Italy did not have the same electoral institutions or governing institutions as New Zealand, th ey chose a mixed member alternative as their new electoral system.

PAGE 457

457 However, Italy was looking for more accountability through a number of processes. The goal was to achieve alternation in government, clearly defined, competing pre election policy agendas between those alternatives, and a chance to remove corrupt officials from elected office. The actions of individuals at times aided accountability and at other times their actions did not. For instance, in 1996 the center left government was comprised o f multiple parties on the left and the center. They were led by Romano internal coup. Prodi was vehemently opposed to adding parties that were outside of the pre election coalition to the government, because he felt that the coalition had a disagreed, and invited part ies from the center right to join the coalition regardless of the electoral mandate (Fabbrini 2000). Perhaps no one exemplifies the lack of accountability than Silvio Berlusconi. The leader of the center right government was re elected in 2001 based on a promise to Italians including major economic reforms and a tough stance on corruption. Soon after his election, Berlusconi committed his government to slowing the ability of prosecutors to pursue corruption related crimes, made it easier for businesses t o return assets to the country without penalty, and did nothing to deal with his own conflicts of interests due to his vast media empire (Hine 2002, Della Porta and Vannucci 2007). Proportional systems (or systems that increase proportionality, li ke New Zealand after 1996) have an ethic of representation. Again, this work demonstrates that individual action plays a significant role in promoting or ignoring that ethic. In Italy, prior

PAGE 458

458 to their elector al system change in the 1990 s, we find examples of both type s of behavior. During the 1960s and 1970 s political elites sought mechanisms to make difficult political situations work. They used committees to pass legislation at times, and came to a historic DC/PCI compr omise in the middle of the 1970s. The 1980 s saw an explosion of corrupt behavior by a new class of politicians dubbed business politicians. These elites used the existing political structures developed by the previous leaders to enrich themselves. This behavior was typified by the lead er of the PSI Bettino Craxi who developed an intricate network to funnel bribes from political clients to the main party coffers (and his own pocket). Corruption reigned due to the perception among a large group of different individuals that payment for p ublic services was the norm, and a e. No one, until the late 1980 s, stood up and declared that it was wrong. New Zealand, after electoral system change, sought increased representation as well. There were sev eral individuals who engaged in behavior that undermined that process. One person who receives a lot of the blame is the leader of the New Zealand First Party, Winston Peters. Peters was central in two different post election coalition deals that were ex tremely unpopular with the public. The first occurred after the initial government post for himself after an acrimonious and secretive negotiation between his party and Lab our and National. His actions contributed to the loss in support for the new electoral regime. He repeated this performance in 2005, negotiating a last minute deal to become foreign minister after flatly stating before the election that his party would h ave no role in government Chapter 5 demonstrates that New Zealanders who

PAGE 459

459 favored parties stating clearly the parties they would form post election coalitions with prior to the election favored a system other than MMP. The behavior of Winston Peters, I b elieve, was a major contributing factor in the failure of MMP to improve levels of support in New Zealand. Another example of agency and its role in the loss of support was the odd practice of party switching that became a major issue after MMP was adopte d. One example of the practice was Alamein Kopu, an Alliance list MP, who left the party to form her own independent party. As the leader of the party (with herself as the lone member), Kopu was entitled to higher salary and funds for party organization. She was not the only MP to switch parties, but her switch set off a debate about the representative nature of the new system (Geddis 2002, 2006). Chapt er 5 details how party switching proved to be a key variable in support for alternative systems among New Zealand citizens. What is interesting to note is that all of the negative behaviors I describe, save the corrupt behavior of business politicians, were within the rules of the game. For instance, Winston Peters was well within his rights as a party le ader to demand and negotiate the best deal for his party and as an extension, his constituents. There certainly were no rules to keep him from doing so, and the type of coalition behavior characterized by negotiation and bargaining between parties should probably have been expected. The coalition process can be a messy one and that fact is certainly documented in the literature and was raised prior to the electoral system change. But, Peters was also criticized for the way that he went about the coalitio n negotiation. He spent a good deal of public money, was accused of drunkenness at a restaurant, fought with the media (almost literally), and generally made a spectacle of himself and the

PAGE 460

460 process. Silvio Berlusconi is another example of the impact that sheer force of personality can have on a political system. Berlusconi, his media empire, and his vast fortune remained (until very recently) a conundrum for Italian democracy. Individuals like Berlusconi and Peters can have a significant impact, but that does not mean there are not institutional controls that can help to organize the behavior of individuals. There a few suggestions I make in this regard in the following subsection of the conclusion. Moving Forward : Thinking about Electoral Systems and S upport I believe this work provides a few insights into the relationship between electoral systems and support. The primary finding of this work is that there is no clear connection between different types of electoral systems and support. Party system behavior varies, under different electoral rules, and that variation is what determines the level of support among a population. This conclusion does not mean that proportional or majoritarian electoral system cannot work. I hold that as long as these s ystems maintain their respective ethic, either representation or accountability, they will be able to foster supportive attitudes. Electoral system changes present another challenge. Increasing proportionality or majoritarianism was less than successful in New Zealand and Italy. There are a few lessons from the electoral system changes in both countries that can be instructive for policy makers and scholars moving forward. There are also several suggestions for how to increase the amount of representati on in countries looking to increased proportionality and accountability in those counties seeking increased accountability. A following subsection concludes with a few suggestions for future scholarship in the area of electoral systems and support.

PAGE 461

461 How t o Increase Accountability and Representation The connection between the ethics of electoral systems, good policy, and support is something that I highlight in this conclusion. In New Zealand, a majoritarian system, support was at its highest when govern ments went to the people with a manifesto, were elected, and followed through on that promised action. In Italy, with increased majoritarianism, the improvement in accountability came when coalitions of parties presented a united front to the voters based on a series of policy promises. The limited time that Italy had success under new electoral rules was when the Olive Tree coalition followed through on the promise they made during the election. In New Zealand, accountability was lost when successive La bour and National governments either did not publish a manifesto or reversed the promises they made in them. In Italy, accountability was lost when the center right government led by Berlusconi was not able to deliver on pre election promises to improve t he economy and deal with the issue of corruption. Representation is the important feature of proportional systems or countries that seek to increase proportionality. In Italy, representation worked when parties fostered context specific relationships w ith their constituents. In the North, this meant ancillary organizations supporting the political parties and parties developing clienteltic relationships with citizens in the South where the organizational traditions were not as strong. Parties were als o representative by cooperating within the legislature in order to make policy achievements. Repr esentation was lost in the 1980 s as associational ties weakened and the clientest ties were replaced with outright corrupt behavior. In New Zealand, there wa s only a short period of time when parties and governments were

PAGE 462

462 representative. This was in 1999 when Labour and the Alliance clearly stated before the elections that they would work together following the election, formed a government, and passed some so cial welfare policies that more closely resembled pre electoral reform New Zealand. Generally speaking however, this type of representation was not evident. Parties seemed unwilling to name, prior to the election, which parties they would be willing to f orm a coalition with and even if they did make those intentions known, following the election those promises were readily dismissed. Representation was further undermined by the practice of party switching, where list changed their party alle giance while remaining in the legislature. The question is what can be done to increase accountability or representation? The experiences of New Zealand and Italy offer several answers. For countries looking to add proportionality or majoritarianism t here are broader institutional changes that can and should be made in addition to changing the electoral system. Broader institutional changes are not always necessary, as countries could also make smaller rule changes that would help to increase accounta bility or representation. Finally, voters and elected officials also bare responsibility for their own actions since rules changes can only go so far. One recommendation that I would make is that electoral system should not be the only institution to cha nge if a country is looking to improve support among its population. The type of institutional changes that are made should aid and assist the ethic of the accompanying electoral system change. In Italy, there was a change to a MMM electoral system, but no other institutional alterations. For instance, Italians kept their bicameral legislature with equal powers in the Senate and Chamber of deputies. If

PAGE 463

463 the goal was to increase accountability and effective policy making, the maintenance of bicameralism i s not the most effective way to do so. A number of institutional changes could be made to increase majoritarianism. One means would be to eliminate, or substantially reduce the powers of, one of the chambers of the legislature. Bicameral legislatures ar e institutional features of systems that seek to improve proportionality (Lijphart 1984). So, while the electoral systems were promoting one ethic (accountability), the remaining institutional arrangements promoted another. Similarly, the institutional a rrangements New Zealand did not change outside of the electoral system. The one house legislature remained the sole institution responsible for passing legislation. So, while the electoral system allowed more representation the institutional make up of t he legislature promoted consolidated power. There was discussion of adding a Senate to the legislature in New Zealand with the powers of amendment and delay. The key point here is that electoral system changes are only going to provide a certain measure of accountability or representation, but how countries are going to make decisions and pass policy are going to be affect by other institutions. My recommendation is that countries adopt additional institutional changes that buttress the electoral system changes. Lijphart (1984) offers a package of institutional arrangements for proportional and majoritarian governments, some of which I mention, that countries should consider if they are looking to add increased representation or accountability. Outside of wholesale institutional changes, there are a number of smaller rule changes that could help countries increase accountability or representation. In New Zealand, prior to electoral system change, one of the informal features of the system

PAGE 464

464 that helped in crease accountability was the manifesto. Also, in Italy the first center left government presented to the public a detailed coalition agreement that stated prior to the election what the center left government would do following the elections. One sugges tion I make is to require parties to craft a specific, public policy manifesto and present that document to the electorate prior to the election. One of the important aspects of accountability is for parties to state what they would do if they were electe d. Voters then cast votes based on those policy positions. By requiring parties to state their policy intentions prior to the election, it will at the very least give voters the opportunity to judge the parties both before and after the election. There a re a number of rule changes that could help increase representation. One of the problems in New Zealand after electoral system change was the formation of coalition governments after the election. The voting public was never really given the opportunity to judge potential government before the election because parties were unwilling to name which parties they would work with. I would suggest a rule change that required parties to name prior to the election, which parties they would enter into a coalition with after the election. These are the rules that I believe would help increase accountability and representation in systems seeking those particular ethics. In New Zealand and Italy, there might have been particular rule changes that would have helped as well. Party switching became a major issue in New Zealand after electoral system change. I would recommend that two rules be in place. The first is a rule that would require any MP who switched parties have to resign their seat. If that seat was a l ist seat, then the party that sponsored that candidate would then nominate the next person on the list to that seat. If it is a SMD seat, then the MP should have to stand for a special election in that

PAGE 465

465 district. To prevent situations like those that occu rred in New Zealand in 2001 where the Alliance party split without technically any MP leaving the party, I would grant either leaving their parties. Another rule t hat I would not suggest was the use of the secret ballot for legislative roll call votes in Italy. This does not strike me as a useful tool to improve representation. Open and honest legislative action would be preferable to secret votes. There is a fin al point I would like to offer in terms of suggestions for rule and institutions. I make the above suggestions knowing fully that ultimately people have to make the existing rules and institutions work. There is an old saying in Rhode Island a serious point in that saying. All of the rules I suggest only work if you have earnest political elites and voters who want to make the rules work, honestly and effective ly. If political elites are going to look for every opportunity to circumvent the rules, then institutional fixes are more like band aids. If voters are going to apathetically sit at home and decry the behavior of political elites, rather than be activel y engaged in their own political system, then rule changes will be cosmetic. An example of this type of behavior is Italy in the late 1970 s and s. Citizens and political elites became entrenched in a culture that became more and more corrupt. It beca me routine to break petty administrative laws and bribes became commonplace. It is difficult to see how any institutional arrangement could work within this type of environment. Institutional failure is another word for human failure, because it is peopl e that run institutions. My only

PAGE 466

466 suggestion in terms of agency is that political elites need to do a better job and citizens need to see that they do by being active and engaged. Thoughts for Further Research There are two suggestions that I would like to make for further research on the topic of the relationship between electoral systems and support. The first is a point that I raise in Chapter 2 about the potential for there to be alternative explanations for why proportional systems might have highe r levels of support. One of the potential reasons has to do with proportional elections and the ability of smaller parties to win seats. Theoretically, citizens might be more satisfied because of just the increased representation of parties regardless of whether they have any influence on policy or not. If this is the case, then proportionality itself might lead to higher levels of support. I have held in this work that in order for representation to lead to higher levels of support, more parties must b e involved in decision making processes, which must lead to widely accepted policies. The alternative hypothesis should be tested. I would suggest a larger N study that looks at new democracies using proportionality as an independent variable and a measu re of support as a dependent variable. If proportionality has an significant effect on supportt, after controlling for some kind of policy and process variables, then there would be evidence that proportionality has an independent effect on support. I wo uld also suggest that the theories that I draw in Chapter 2 be tested in alternative cases. As I have mentioned in this work, I feel that newer democracies (other than countries that change their electoral system) present the best opportunity to study th is question. Like the study I offer above, I would suggest a large N study with

PAGE 467

467 survey data from new democracies. Here, proportionality and support would remain the independent and dependent variables respectively. Instead of using process and policy va riables for control variables, I would suggest running some sort of two stage logit model. This type of model would run from proportionality to policy or process then to support rather than using policy and process as controls. This type of study could p rovide confirmation for some of the conclusions that I draw in this study. These two studies could be used in conjunction with this study to form a more complete picture of the relationship between electoral systems and support. Chapter 1 describes why I think support is an important topic of study. Electoral systems are an institution that can be changed, so understanding the relationship they could potentially have with support is of great consequence for scholars.

PAGE 468

468 LIST OF REFERE NCES Abramson, Paul, and Ronald Inglehart. 1970. The Development of Systemic Support in Four Western Democracies. Comparative Political Studies 2 (4): 419 443. Aimer, Peter. 1998. "Old and New Party Choices." In Voters' Victory? New Zealand's First Electio n Under Proportional Representation eds. Jack Vowles, Peter Aimer, Susan Banducci and Jeffrey Karp. Auckland: Auckland University Press. 48 64. Aimer, Peter, and Raymond Miller. 2002. "New Zealand Politics in the 1990 s." In Proportional Representation on Trial: The 1999 New Zealand Election and the Fate of MMP eds. Jack Vowles, Peter Aimer, Jeffrey Karp, Susan Banducci, Raymond Miller and Ann Sullivan. Auckland: Auckland University Press. 1 15. Aimer, Peter, and Jack Vowles. 2004. "What Happened at the 20 02 Election?" In Voters' Veto: The 2002 Election in New Zealand and the Consolidation of Minority Government eds. Jack Vowles, Peter Aimer, Susan Banducci, Jeffrey Karp and Raymond Miller. Auckland: Auckland University Press. 16 32. Aldrich, John. 1995. W hy Parties? The Origin and Transformation of Political Parties in America Chicago: University of Chicago Press. Almond, Gabriel, and Sidney Verba. 1963. The Civic Culture: Political Attitudes and Democracy in Five Nations Princeton: Princeton University Press. Amy, Douglas. 1993. Real Choices/New Voices: How Proportional Representation Elections Could Revitalize American Democracy New York: Columbia University Press. Anderson, Chris. 1998. Parties, Party Systems, and Satisfaction with Democratic Perform ance in the New Europe. Political Studies 46 (3): 572 589. Anderson, Christopher, Andre Blais, Shaun Bowler, Todd Donovan, and Ola Listhaug. 2005. Losers' Consent: Elections and Democratic Legitimacy Oxford: Oxford University Press. Anderson, Christopher, and Christine Guillory. 1997. Political Instituions and Satisfaction with Democracy. American Political Science Review 91 (1): 66 82. Ardizzoni, Michela. 2007. North/South, East/West: Mapping Italianness on Television Landham: Lexington Books. Bale, Tim and Nigel Roberts. 2002. Plus ca Change...? Anti Party Sentiment and Electoral System Change: A New Zealand Case Study. Commonwealth and Comparative Politics 40 (2): 1 20.

PAGE 469

469 Banducci, Susan, Todd Donovan, and Jeffrey Karp. 1999. Proportional Representation and Attitudes about Politics: Results from New Zealand. Electoral Studies 18 (4): 533 555. Banfield, Edward. 1958. The Moral Basis of a Backward Society New York: Free Press. Bardi, Luciano. 2002. "Italian Parties: Change and Functionality." In Political Parties in Advanced Industrial Democracies eds. Paul Webb, David Farrell and Ian Holliday. New York: Oxford University Press. 46 76. 2007. Electoral Change and its Impact on the Party System in Italy. West European Politics 30 (4): 711 732. Barker, Fiona, Jonathan Boston, Stephen Levine, Elizabeth McLeay, and Nigel Roberts. 2001. "An Initial Assessment of the Consequences of MMP in New Zealand." In Mixed Member Electoral Systems: The Best of Both Worlds? eds. Matthew Shugart and Martin Wattenberg. O xford: Oxford University Press. 297 323. Bartolini, Stefano, Alessandro Chiaramonte, and Roberto D'Alimonte. 2004. The Italian Party System between Parties and Coalitions. West European Politics 27 (1): 1 19. Baylis, Thomas. 1989. Governing by Committee: C ollegial Leadership in Advanced Societies Albany, NY: State University of New York Press. Beetham, David. 1994. Defining and Measuring Democracy New York: Sage Publications. Bensel, Richard, and Elizabeth Sanders. 1979. The Effect of Electoral Rules on V oting Behavior: The Electoral College and Shift Voting. Public Choice 34 (1): 69 85. Blais, Adre, and Agnieszka Dobrzynska. 1998. Turnout in Electoral Democracies. European Journal of Political Research 32 (2): 239 261. Booth, John A. 1991. Socioeconomic and Political Roots of National Revolts in Central America. Latin American Research Review 26 (1): 33 73. Booth, John A, and Mitchell Seligson. 2009. The Legitimacy Puzzle in Latin America: Political Support and Democracy in Eight Nations New York: Cambri dge University Press. Bowler, Shaun, and Todd Donovan. 2007. Reasoning about Institutional Change: Winners, Losers, and Support for Electoral Reforms. British Journal of Political Science 37 (3): 455 476. Bowler, Shaun, and David Farrell. 2006. We Know Whi ch One We Prefer but we Don't Really Know Why: The Curious Case of Mixed Member Electoral Systems. British Journal of Political Science 8 (2): 445 460.

PAGE 470

470 Bowler, Shaun, Jeffrey Karp, and Todd Donovan. 2010. Strategic Coalition Voting: Evidence from New Zeal and. Electoral Studies 30 (1): 1 8. Brittan, Samuel. 1975. The Economic Contradiction of Democracy. British Journal of Political Science 5 (2): 129 159. Bull, Martin, and James Newell. 2005. Italian Politics: Adjustment Under Duress Malden: Polity Press. Cain, Bruce. 1978. Strategic Voting in Britain. American Journal of Political Science 22 (3): 639 655. Calise, Mauro. 1993. Remaking the Italian Party System: How Duverger got it Wrong by Saying it Right. West European Politics 16 (4): 545 560. Campbell, Angus, Philip Converse, Warren Miller, and Donald Stokes. 1960. The American Voter Chicago: University of Chicago Press. Canache, Damarys. 2002. Public Opinion and Protest in a Fragile Democracy Coral Gables: North South Center Press at the University of Miami. Canache, Damarys, Jeffrey Mondak, and Mitchell Seligson. 2001. Meaning and Measurement in Cross National Research on Satisfaction with Democracy. Public Opinion Quarterly 65 (4): 506 28. Cartocci, Roberto. 1998. Omens of an Early Winter: The Propo rtional Vote and the Changing Italian Party System. European Journal of Political Research 34 (1): 35 61. Challies, Edward, and Warwick Murray. 2008. Towards Post neoliberalism? The Comparative Politico economic transition of New Zealand and Chile. Asia Pa cific Viewpoint 49 (2): 228 253. Chang, Eric, and Yun han Chu. 2006. Corruption and Trust: Exceptionalism in Asian Democracies. Journal of Politics 68 (2): 259 271. Chang, Eric, and Miriam Golden. 2006. Electoral Systems, District Magnitude, and Corruption British Journal of Political Science 37 (1): 115 138. Citrin, Jack. 1974. Comment: The Political Relevance of Trust in Government. American Political Science Review 68 (3): 973 989. Clarke, Harold, Nitish Dutt, and Allan Kornberg. 1993. The Political Eco nomy of Attitudes Toward Politiy and Society in Western European Democracies. Journal of Politics 55 (4): 988 1012. Coleman, Kenneth M. 1976. Diffuse Support in Mexico: The Potential for Crisis Beverly Hills: Sage Publications.

PAGE 471

471 Cotta, Maurizio. 2002. "Ber lusconi's Second Governmental Test." In Italian Politics: The Return of Berlusconi eds. Paolo Bellucci and Martin Bull. New York, NY: Bergham Books. 146 166. Cotta, Maurizio, and Luca Verzichelli. 2003. "The Second Berlusconi Government put to the Test: A Year of Complications." In Italian Politics: The Second Berlusconi Government eds. Jean Blondel and Paolo Segatti. New York: Berghahn Books. 37 58. Cox, Gary. 1997. Making Votes Count Cambridge: Cambridge University Press. 1999. Electoral Rules and Electoral Coordination. Annual Reviews of Political Science 2 (1): 145 162. Cox, K.E., and L.J. Schoppa. 2002. Interaction Effect in Mixed Member Electoral Systems. Comparative Political Studies 35 (9): 1027 1053. Craig, Stephen. 1980. Political Disconten t and Political Action. The Journal of Politics 43 (2): 514 522. 1993. The Malevolent Leaders: Popular Discontent in America Boulder: Westview Press. ed. 1996. The Angry Voter? Politics and Popular Discontent in the 1990s Edited by Stephen Cra ig, Broken Contract? Changing Relationships between Americans and Their Government Boulder: Westview. Craig, Stephen, Richard Niemi, and Franco Mattei. 1991. Measuring Internal Political Efficacy in the 1988 National Election Study. American Political Sc ience Review 85 (4): 1407 1413. Craig, Stephen, Richard Niemi, and Glenn Silver. 1990. Political Efficacy and Trust: A Report on the NES Pilot Study Items. Political Behavior 12 (3): 289 314. Crozier, Michel, Joji Watanuki, and Samuel Huntington. 1975. The Crisis of Democracy New York: New York University Press. D'Alimonte, Robert o 2001. "Mixed Electoral Rules, Partisan Realignment, and Party System Change in Italy." In Mixed Member Electoral Systems: The Best of Both Worlds? eds. Matthew Shugart and Mar tin Wattenberg. Oxford: Oxford University Press. 323 351. D'Alimonte, Roberto. 2005. "Italy: A Case of Fragmented Bipolarism." In The Politics of Electoral Systems eds. Michael Gallagher and Paul Mitchell. Oxford: Oxford University Press. 253 276. D'Alimo nte, Roberto, and David Nelken, eds. 1997. Italian Politics: The Center Left in Power Boulder, CO: Westview Press.

PAGE 472

472 Dahl, Robert. 1971. Polyarchy: Participation and Opposition New Haven: Yale University Press. Dalton, Russell. 1985. Political Parties and Political Representation. Comparative Political Studies 17 (2): 267 299. 2004. Democratic Challenges, Democratic Choices: The Erosion of Political Support in Advanced Industrial Democracies New York: Oxford University Press. 2006. Social Modern ization and the End of the Ideology Debate: Patters of Ideological Polarization. Japanese Journal of Political Science 7 (1): 1 22. Darwall, Rupert. 2003. Market Reform: Lessons from New Zealand. Policy Review 18 (2): 61 71. Del Monte, Alfredo, and Erasmo Papagni. 2007. The Determinants of Corruption in Italy: Regional Panel Data Analysis. European Journal of Political Economy 23 (2): 379 396. Della Porta, Donatella. 1996. Actors in Corruption: Business Politicians in Italy. International Social Science Jou rnal 48 (149): 349 364. 2001. A Judges' Revolution? Political Corruption and the Judiciary in Italy. European Journal of Political Research 39 (1): 1 21. Della Porta, Donatella, and Alberto Vannucci. 1999. Corrupt Exchanges: Actors, Resources, and Mec hanisms of Political Corruption New York: Aldine de Gruyter. 2007. Corruption and Anti Corruption: The Political Defeat of 'Clean Hands' in Italy. West European Politics 30 (4): 830 853. Della Sala, Vincent. 1993. "The Scalfaro Election: In the Shad ow of Presidentialism?" In Italian Politics: A Review eds. Gianfranco Pasquino and S Hellman. London: Pinter. 34 49. DelliCarpini, M, and S Keeter. 1996. What Americans know about Politics and Why it Matters New Haven: Yale University Press. Denemark, D avid, ed. 2001. Choosing MMP in New Zealand: Explaining the 1993 Electoral Reform Edited by Matthew Shugart and Martin Wattenberg, Mixed Member Electoral Systems: The Best of Both Worlds? Oxford: Oxford University Press. Dennis, Jack. 1966. Support for th e Party System by the Mass Public. The American Political Science Review 60 (3): 600 615.

PAGE 473

473 Di Palma, Giuseppe. 1976. Institutional Rules and Legislative Outcomes in the Italian Parliament. Legislative Studies Quarterly 1 (2): 147 179. 1977. Surviving W ithout Governing Berkley: University of California Press. Donovan, Mark. 2004. "The Governance of the Center Right Coalition." In Italian Politics: Italy Between Europeanization and Domestic Politics eds. Sergio Fabbrini and Vincent Della Salla. New York : Berghahn Books. 80 98. Downs, Anthony. 1957. An Economic Theory of Democracy New York, NY: Addison Wesley. Duncan, Ian. 1996. "Natural Resource Management." In A Study of Economic Reform: The Case of New Zealand eds. Brian Silverstone, Allan Bollard an d Ralph Lattimore. New York: Elsevier. 388 410. Duverger, Maurice. 1954. Political Parties New York: Wiley and Sons. 1972. Party Politics and Pressure Groups: A Comparative Introduction New York: Cromwell. 1984. "Which is the Best Electoral Sys tem?" In Choosing an Electoral System ed. Arend Lijphart and Bernard Grofman. Westport: Praeger. 31 41. 1986. "Duverger's Law: Forty Years Later." In Electoral Laws and Their Political Consequences ed. Arend Lijphart and Bernard Grofman. New York: A gathon. 69 85. Easton, David. 1965. A Systems Analysis of Political Life Chicago: University of Chicago Press. 1975. A Re assessment of the Concept of Political Support. British Journal of Political Science 5 (1): 435 457. Eckstein, Harry. 1988. A Cu lturalist Theory of Political Change. American Political Science Review 82 (4): 789 804. Fabbrini, Sergio. 2000. "From the Prodi Government to the D'Alema Government: Continuity or Discontinuity." In Italian Politics: The Return of Politics eds. David Hin e and Salvatore Vassallo. New York: Berghahn Books. 121 138. 2000. Political Change without Institutional Transformation: What Can We Learn from the Italian Crisis of the 1990s? International Political Science Review 21 (2): 173 196. Felsen, David. 20 00. "Changes to the Italian Budgetary Regime: The Reforms of Law, no. 94/1997." In Italian Politics: The Return of Politics eds. David Hine and Salvatore Vassallo. New York: Berghahn Books. 157 175.

PAGE 474

474 Ferrara, F, Erik Herron, and Misa Nishikawa. 2005. Mixed Electoral Systems: Contamination and its Consequences New York: Palgrave Macmillan. Fiorina, Morris. 1981. Retrospective Voting in American National Elections New Haven: Yale University Press. Fiorina, Morris, and Roger Noll. 1978. Voters, Bureaucrats, and Legislators: A Rational Choice Perspectives on the Growth of Bureaucracy. Journal of Public Economics 9 (2): 239 254. Fisher, Stephen. 1973. The Wasted Vote Thesis: West German Evidence. Comparative Politics 5 (2): 293 299. Foley, Michael. 1996. Layin g the Groundwork: The Struggle for Civil Society in El Salvador. Journal of Interamerican Politics and World Affairs 38 (1): 67 104. Fuchs, Dieter. 1993. "Trends in Political Support." In Political Culture in Germany eds. Dirk Berg Scholosser and Ralf Ryt lewski. New York: St. Martin's Press. 232 270. Furlong, Paul. 2002. "The Italian Political System in 2001: Radical Change and Work in Progress." In The Italian General Election of 2001 ed. James Newell. New York: Manchester University Press. 11 28. Gallag her, Michael, and Paul Mitchell. 2005. "Introduction to Electoral Systems." In The Politics of Electoral Systems eds. Michael Gallagher and Paul Mitchell. Oxford: Oxford University Press. 3 25. Gamson, William A. 1968. Power and Discontent Homewood: Dors ey Press. Geddes, Barbara, and John Zaller. 1989. Source of Popular Support for Authoritarian Regimes. American Journal of Political Science 33 (2): 319 347. Geddis, Andrew. 2002. Gang Aft A Gley: New Zealand's Attempt to Combat 'Party Hopping' by Elected Representatives. Election Law Journal 1 (4): 557 571. 2006. Proportional Representation 'Party Hopping' and the Limits of Electoral Regulation: A Cautionary Tale from New Zealand. Common Law World Review 35 (1): 24 50. Gilbert, Mark. 1995. The Italian Revolution: The End of Politics, Italian Style? Boulder: Westview Press. 2000. "The Bassinini Laws: A Half Way House in Local Government Reform." In Italian Politics: The Return of Politics eds. David Hine and Salvatore Vassallo. New York: Berghahn Books. 139 156.

PAGE 475

475 Gilbert, Mark, and Gianfranco Pasquino. 2000. "Introduction: The Faltering Transition." In Italian Politics: The Faltering Transition eds. Mark Gilbert and Gianfranco Pasquino. New York: Berghahn Books. 21 32. eds. 2000. Italian Polit ics: The Faltering Transition New York: Bergham Books. Ginsborg, Paul. 1990. A History of Contemporary Italy: Society and Politics 1943 1988 New York: Penguin Books. Golden, Miriam. 2003. Electoral Connections: The Effects of the Personal Vote on Politi cal Patronage, Bureaucracy, and Legislation in Postwar Italy. British Journal of Political Science 33 (2): 189 212. Gschwend, Thomas, and Henk van der Kolk. 2006. Split Ticket Voting in Mixed Member Proportional Systems: The Hypothetical Case of the Nether lands. Acta Politics 41 (1): 163 179. Guarnieri, Carlo, and James Newell. 2005. "Introduction: 2004 a Year 'On Hold'." In Italian Politics: Quo Vadis? eds. Carlo Guarnieri and James Newell. New York: Bergham Books. 29 46. Hainmueller, Jens, and Holger Ke rn. 2008. Incumbency as a source of spillover effects in mixed systems: Evidence from a regression discontinuity design. Electoral Studies 27 (2): 213 227. Harmel, Robert, and John Robertson. 1986. Government Stability and Regime Support: A Cross Nationa l Analysis. The Journal of Politics 48 (4): 1029 1040. Hermens, Ferdinand. 1984. "Representation and Proportional Representation." In Choosing an Electoral System eds. Arend Lijphart and Bernard Grofman. Westport: Praeger. 15 30. Hetherington, Mark. 1998. The Political Relevance of Political Trust. American Political Science Review 92 (4): 791 808. 2005. Why Trust Matters: Declining Political Trust and the Demise of American Liberalism Princeton: Princeton University Press. Hibberd, Matthew. 2004. "R AI Under the Center Right: Wither 50 Years of Public Service Television." In Italian Politics: Italy Between Europeanization and Domestic Politics eds. Sergio Fabbrini and Vincent Della Salla. New York: Berghahn Books. 150 165. 2007. Conflicts of Int erest and Media Pluaralism in Italian Broadcasting. West European Politics 30 (4): 881 902.

PAGE 476

476 Hibbing, John, and Elizabeth Theiss Morse. 2002. Stealth Democracy: Americans' Beliefs About how Governmnet Should Work New York: Cambridge University Press. Hine, David. 1993. Governing Italy: The Politics of Bargained Pluralism New York: Oxford University Press. 2002. "Silvio Berlusconi, the Media, and the Conflict of Interest Issue." In Italian Politics: The Return of Berlusconi eds. Paolo Bellucci and Mar tin Bull. New York: Bergham Books. 261 276. Hine, David, and Salvatore Vassallo. 2000. "Introduction: One Step Toward Europe; Two Steps Back from Institutional Reform." In Italian Politics: The Return of Politics eds. David Hine and Salvatore Vassallo. Ne w York: Berghahn Books. 33 46. Holmberg, Soren. 1999. "Down and Down We Go: Political Trust in Sweden." In Critical Citizens: Global Support for Democratic Governance ed. Pippa Norris. Oxford: Oxford University Press. 103 122. 2003. Are Political P arties Necessary? Electoral Studies 22 (2): 287 299. Ignazi, Piero. 2001. Italy. European Journal of Political Research 40 (3): 340 347. Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society Princeton: Princeton University Press. Jack Vowl es, Susan Banducci, and Jeffrey Karp. 1999. "New Zealand Election Study." In, ed. University of Waikato. Waikato: Foundation for Research, Science, and Technology. Jackson, John. 1975. Issues, Party Choices, and Presidential Votes. American Journal of Poli tical Science 19 (2): 161 185. Jackson, Keith, and McRobie. 1998. New Zealand Adopts Proportional Representation Brookfield Ashgate. Jacoby, William. 1988. The Impact of Party Identification on Issue Attitudes. American Journal of Political Science 32 (3) : 643 661. Jesson, Bruce. 1997. "Winston in Otahuhu." New Zealand Political Review Johnson, R.J. 1984. "Seats, Votes, Redistricting, and the Allocation of Power in Electoral Systems." In Choosing and Electoral System eds. Arend Lijphart and Bernard Groma n. Westport: Praeger. 59 72. Johnston, R, and C Pattie. 2008. Money and Votes: A New Zealand Example. 27 1 (113 133).

PAGE 477

477 Joseph, Phillip. 2008. "MMP and the Constitution: Future Constitutional Challenges." Presented at the MMP and the Constitution: 15 years p ast; 15 years forward, New Zealand Centre for Public Law, Victoria University of Wellington. Kaase, Max. 1988. "Political Alienation and Protest." In Comparing Pluralist Democracies ed. Mattie Dogan. Boulder: Westview Press. 114 142. Karp, Jeffrey. 2009. Candidate Effects and Spill over in Mixed Systems: Evidence from New Zealand. Electoral Studies 28 (1): 41 50. Karp, Jeffrey, and Susan Banducci. 1998. "Voter Satisfaction After Electoral System Change." In Voters' Victory? New Zealand's First Election Und er Proportional Representation eds. Jack Vowles, Peter Aimer, Susan Banducci and Jeffrey Karp. Wellington: Auckland University Press. 135 153. Karp, Jeffrey, Jack Vowles, Susan Banducci, and Todd Donovan. 2002. Strategic Voting, Party Activity, and Candid ate Effects: Testing Explanations for Split Voting in New Zealand's New Mixed System. Electoral Studies 21 (1): 1 22. Katz, Richard. 1996. Electoral Reform and the Transformation of Party Politics in Italy. Party Politics 2 (1): 31 53. 1997. Democracy and Elections New York: Oxford University Press. 2001. The Problem of Candidate Selection and Models of Party Democracy. Party Politics 7 (3): 277 296. ed. 2001. Reforming the Italian Electoral Law, 1993 Edited by Matthew Shugart and Martin Wa ttenberg, Mixed Member Electoral Systems: The Best of Both Worlds? Oxford: Oxford University Press. King, Anthony. 1975. Overload: Problems of Governing in the 1970s. Political Studies 23 (2): 284 296. Klingeman, H, R Hofferbert, and I Budge. 1994. Parties Policy, and Democracy Boulder: Westview Press. Klingeman, Hans Dieter. 1999. "Mappin g Political Support in the 1990 s: A Global Analysis." In Critical Citizens: Global Support for Democratic Government ed. Pippa Norris. Oxford: Oxford University Press. 31 56. Klingeman Hans Dieter, and Dieter Fuchs. 1995. Citizens and the State Oxford: Oxford University Press. Koff, Sandra, and Stephen Koff. 2000. Italy: From the First to the Second Republic New York: Routledge.

PAGE 478

478 Kornberg, Allan, and Harold Clarke. 199 2. Citizens and Community: Political Support in a Representative Democracy New York: Cambridge University Press. 1994. Beliefs about Democracy and Satisfaction with Democratic Government: The Canadian Case. Political Research Quarterly 47 (3): 537 5 63. Kreppel, Amie. 1997. The Impact of Parties in Government on Legislative Output in Italy. European Journal of Political Research 31 (2): 327 350. 2009. Executive Legislative Relations and Legislative Agenda Setting in Italy: From Leggine to Decreti and Deleghe. Bulletin of Italian Politics 1 (2): 183 209. Kuechler, Manfred. 1986. Maximizing Utility at the Polls? A Replication of Himmelweit's 'Consumer Model of Voting' with German Election Data from 1983. European Journal of Political Research 14 (1 ): 81 95. 1991. "The Dynamics of Mass Political Support in Western Europe: Methodological Problems and Preliminary Findings." In Eurobarometer: The Dynamics of European Public Opinion, Essays in Honor of Jacques Rene Rabier eds. Karlheinz Reif and Ro nald Inglehart. New York: St. Martin's Press. 275 293. Laakso, M, and R Taagepera. 1979. Effective Number of Parties: A Measure with Application to West Europe. Comparative Political Studies 12 (1): 3 28. Ladner, Andreas, and Henry Milner. 1999. Do Voters Turn Out More Under Proportional than Majoritarian Systems? The Evidence from Swiss Communal Elections. Electoral Studies 18 (2): 235 251. Lakeman, Enid. 1970. How Democracies Vote: A Study of Majority and Proportional Electoral Systems London: Faber. 1984. "The Case for Proportional Representation." In Choosing and Electoral System eds. Arend Lijphart and Bernard Grofman. Westport: Praeger. 41 52. Lamare, James. 1991. The Growth of Antinuclearism in New Zealand. Australian Journal of Political Scien ce 26 (3): 472 487. Lamare, James, and Jack Vowles. 1996. Party Interests, Public Opinion, and Institutional Preferences: Electoral System Change in New Zealand. Australian Journal of Political Science 31 (3): 321 346. Lambsdorff, Johann. 2007. "The Method ology of the Corruption Perceptions Index." ed Transparency International. Passau: University of Passau. Lange, Peter. 1980. "Crisis and Consent, Change and Compromise: Dilemmas of Italian Communism in the 1970s." In Italy in Transition: Conflict and Conse nsus eds. Peter Lange and Sidney Tarrow. London Frank Cass and Co. 110 132.

PAGE 479

479 LaPalombara, Joseph. 1987. Democracy Italian Style New Haven: Yale University Press. Lawrence, Robert. 1997. "Is it Really the Economy Stupid?" In Why People Don't Trust Governme nt eds. J Nye, PD Zelikow and DC King. Cambridge: Harvard University Press. 111 132. Leonardi, Robert, Raffaella Nanetti, and Gianfranco Pasquino. 1978. Institutionalization of Parliament and Parliamentarization of Parties in Italy. Legislative Studies Quarterly 3 (1): 161 186. Levine, Stephen. 1979. The New Zealand Political System: Politics in a Small Society Boston: Littlehampton Book Services LTD. Levine, Stephen, and Alan Robinson. 1976. The New Zealand Voter: A Survey of Public Opinion and Elector al Behavior Auckland: Price Milburn for New Zealand University Press. Lijphart, Arendt. 1984. Democracies Cambridge: Harvard University Press. 1994. Electoral Systems and Party Systems: A Study of Twenty Seven Democracies, 1945 1990 New York: Oxfor d University Press. 1999. Patterns of Democracy New Haven: Yale University Press. Lijphart, Arendt, and Bernard Grofman, eds. 1986. Electoral Laws and their Political Consequences New York: Algora Publishing. Linz, Juan. 1990. The Perils of Presiden tialism. Journal of Democracy 1 (1): 51 69. Linz, Juan, and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post Communist Europe Baltimore, MD: John Hopkins University Press. Lipsett, Seymour Martin, and William Schneider. 1983. The Confidence Gap: Business, Labor, and Government in the Public Mind New York: Free Press. Lizzeri, Alessandro, and Nicola Perisco. 2001. The Provision of Public Goods Under Alternative Electoral Incentives. The Amer ican Economic Review 91 (1): 225 240. Lockerbie, Brad. 1993. Economic Dissatisfaction and Political Alienation in Western Europe. European Journal of Political Research 23 (3): 281 293. Loewenberg, Gerhard. 1971. The Influence of Parliamentary Behavior on Regime Stability. Comparative Politics 3 (2): 177 201.

PAGE 480

480 Mainwaring, Scott. 1991. "Clientelism, Patromonialism, and Economic Crisis: Brazil Since 1979." Presented at the International Conference for the Latin American Studies Association, Washington, DC. Man nheimer, Renato. 1989. "Italy: Consulting the Oracle." In Choosing Europe? eds. C van der Eijk and Mark Franklin. Ann Arbor: University of Michigan Press. 186 200. March, James, and Johan Olsen. 1984. The New Institutionalism: Organizational Factors in P olitical Life. The American Political Science Review 78 (3): 734 750. Markoff, John. 1995. The Great Wave of Democracy in Historical Perspective. Cornell Studies in International Affairs 34: 1 87. McAllister, Ian, and Stephen White. 2007. Political Parties and Democratic Consolidation in Post Communist Societies. Party Politics 13 (2): 197 216. McCarthy, Patrick. 1995. The Crisis of the Italian State New York, NY: St. Martin's Press. 1997. "Forza Italia: Old Problems Linger On." In Italian Politics: T he Center Left in Power eds. Roberto D'Alimonte and David Nelken. Boulder, Co: Westview Press. 51 65. McClelland, Alison, and Susan St John. 2006. Social Policy Responses to Globalisation in Australia and New Zealand, 1980 2005. Australian Journal of Poli tical Science 41 (2): 177 191. McClintock, Brent. 1998. Whatever Happened to New Zealand? The Great Capitalist Restoration Reconsidered. Journal of Economic Issues 32 (2): 497 503. McTaggart, Stephen. 2005. Monitoring the Impact of Social Policy, 1980 200 1 Auckland: Spear. Miller, Arthur. 1974. Political Issues and Trust in Government: 1964 1970. The American Political Science Review 68 (3): 951 973. 1974. Rejoinder to 'Comment' by Jack Citrin: Political Discontent or Ritualism? American Political Sc ience Review 68 (3): 989 1001. Miller, Arthur, and Ola Listhaug. 1999. "Political Performance and Institutional Trust." In Critical Citizens: Globabl Support for Democratic Government ed. Pippa Norris. Oxford: Oxford University Press. 204 216. Miller, Ray mond. 1998. "Coaltion Government: The People's Choice?" In Voters' Victory? New Zealand's First Election Under Proportional Representation eds. Jack Vowles, Peter Aimer, Susan Banducci and Jeffrey Karp. Wellington: Auckland University Press. 120 135.

PAGE 481

481 2002. "Coalition Government: The Labour Alliance Pact." In Proportional Representation on Trial: The 1999 New Zealand Election and the Fate of MMP eds. Jack Vowles, Peter Aimer, Jeffrey Karp, Susan Banducci, Raymond Miller and Ann Sullivan. Auckland: Auc kland University Press. 114 129. 2006. Political Leadership in New Zealand Auckland: Auckland University Press. Miller, Raymond, and Jennifer Curtin. 2008. "Counting the Costs of Coalition: The Case of New Zealand's Minor Parties." Presented at the C anadian Political Science Association Annual Conference, University of British Columbia. Mishler, William, and Richard Rose. 2002. Learning and Re learning Regime Support: The Dynamics of Post Communist Regimes. European Journal of Political Research 41 (1 ): 5 37. Mitchell, Austin. 1969. Politics and People in New Zealand Wellington: Whitcombe and Tombs. Morlino, Leonardo. 1996. Crisis of Parties and Change of Party System in Italy. Party Politics 2 (1): 5 30. Morlino, Leonardo, and Marco Tarchi. 1996. The Dissatisfied Society: The Roots of Political Chance in Italy. European Journal of Political Research 30 (1): 41 63. Moser, Robert. 1999. Electoral Systems and the Number of Parties in Post Communist States. World Politics 51 (3): 359 385. 2001. Unexp ected Outcomes: Electoral Systems, Political Parties, and Representation in Russia Pittsburgh: University of Pittsburgh Press. Moser, Robert, and Ethan Scheiner. 2005. Strategic Ticket Splitting and the Personal Vote in Mixed Member Electoral Systems. Jou rnal of Legislative Studies 30 (2): 259 276. Moss, Donald. 1995. Patronage Revisited: The Dynamics of Information and Reputation. Journal of Modern Italian Studies 1 (1): 58 93. Mulgan, Richard. 1990. "The Changing Electoral Mandate." In The Fourth Labout Government: Politics and Policy in New Zealand eds. Martin Holland and Jonothan Boston. New York: Oxford University Press. 11 21. 1992. "The Elective Dictatorship in New Zealand." In New Zealand Politics in Perspective ed. Hyam Gold. Auckland: Longm an Paul. 446 456. 1995. The Democratic Failure of Single Party Government: The New Zealand Experience. Australian Journal of Political Science 30 (5): 82 96.

PAGE 482

482 Nagel, Jack. 1994. What Political Scientists Can Learn from the 1993 Electoral Reform in New Zealand. PS: Political Science and Politics 27 (3): 525 529. 1998. Social Choice in a Plutarian Democracy: The Politics of Market Liberalization in New Zealand. British Journal of Political Science 28 (2): 223 267. Newell, James. 2000. Parties and Dem ocracy in Italy Burlington: Ashgate Publishing. 2010. The Politics of Italy Cambridge: Cambridge University Press. Newell, James, and Martin Bull. 1996. The April 1996 Italian General Election: The Left on Top or on Tap. Parliamentary Affairs 49 (4) : 616 647. 1997. Between Crisis and Transition: Italian Politics in the 1990s. West European Politics 20 (1): 1 20. Newton, Kenneth. 1999. "Social and Political Trust." In Critical Citizens: Global Support for Democratic Government ed. Pippa Norris. Oxford: Oxford Univesity Press. 169 187. Nishikawa, Misa, and Erik Herron. 2001. Contamination Effects and the Number of Political Parties in Superposition Electoral Systems. Electoral Studies 20 (1): 63 86. 2004. Mixed Electoral Rules' Impact on Part y Systems. Electoral Studies 23 (4): 753 768. Norris, Pippa. 1985. Women in European Legislative Elites. West European Politics 8 (4): 90 102. 1987. Politics and Sexual Equality Boulder, CO: Rienner. 1997. Chosing Electoral Systems: Proportional Majoritarian, and Mixed Systems. International Political Science Review 18 (3): 297 313. 1999. "Conclusions: The Growth of Critical Citizens and its Consequences." In Critical Citizens: Global Support for Democratic Government ed. Pippa Norris. Oxf ord: Oxford University Press. 257 272. 1999. "Introduction: The Growth of Critical Citizens?" In Critical Citizens: Global Support for Democratic Government ed. Pippa Norris. Oxford: Oxford University Press. 1 30. 2002. Democratic Phoenix: Polit ical Activism Cambridge: Cambridge University Press.

PAGE 483

483 ed. 1999. Critical Citizens: Global Support for Democratic Government Oxford: Oxford University Press. North, Douglas. 1981. Structure and Change in Economic History New York: Norton O'Donnell, Guillermo. 1996. Illusions about Consolidation. Journal of Democracy 7 (2): 34 51. 1998. Horizontal Accountability in New Democracies. Journal of Democracy 9 (3): 112 126. Onado, Marco. 2005. "The Collapse of Parmalat." In Italian Politics: Quo Vadis? ed. James Newell. New York: Bergham Books. 190 203. Palmer, Geoffrey. 1979. Unbridled Power? An Interpretation of New Zealand's Constitution and Government Wellington: Oxford University Press. Parisi, Arturo, and Gianfranco Pasquino. 1980. "Changes in I talian Electoral Behaviour: The Relationships Between Parties and Voters." In Italy in Transition: Conflict and Consensus eds. Sidney Tarrow and Peter Lange. London: Frank Cass and Co. 6 30. Parker, Simon. 1997. "The Government of the Ulivo." In Italian P olitics: The Center Left in Power eds. Roberto D'Alimonte and David Nelken. Boulder: Westview Press. 125 142. Pasquino, Gianfranco. 1980. "Italian Christian Democracy: A Party for all Seasons?" In Italy in Transition: Conflict and Consensus eds. Sidney T arrow and Peter Lange. London: Frank Cass Publishing Co. 88 109. 1997. No Longer a Tarty State? Institutions, Power and the Problems of Italian Reform. West European Politics 20 (1): 34 53. 2001. The Italian National Elections of 13 May 2001. Jo urnal of Modern Italian Studies 6 (3): 371 387. 2002. "The Political Context 1996 2001." In The Italian General Election of 2001 ed. James Newell. New York: Manchester University Press. 29 48. 2004. "The Restructuring of the Italian Party System ." Presented at the Annual Meeting of the Political Studies Association, Lincoln, NE. 2007. The Five Faces of Silvio Berlusconi: The Knight of Anti Politics. Modern Italy 12 (1): 39 54. Peters, Guy. 1998. Managing Horizontal Governmnet Public Administ ration 76 (2): 295 311.

PAGE 484

484 Pharr, S. 1997. "Public Trust and Democracy in Japan." In Why People Don't Trust Government eds. J Nye, PD Zelikow and DC King. Cambridge: Harvard University Press. 237 252. Piattoni, Simona. 1998. "Virtous Clientelism: The Southe rn Question Resolved?" In Italy's 'Southern Question': Orientalism in One Country ed. J Schneider. Oxford: Berg. 225 243. Popkin, Samuel L. 1991. The Reasoning Voter: Communication and Persuasion in Presidential Campaigns Chicago: University of Chicago Press. Powell, G Bingham. 1980. "Voting Participation in Thirty Democracies: Effects of Socioeconomic, Legal, and Partisan Environments." In Party and Electoral Systems ed. G Bingham Powell. London: Sage. 18 35. 2000. Elections as Instruments of Demo cracy New Haven: Yale University Press. Przeworksi, Adam, Susan Stokes, and Bernard Manin, eds. 1999. Democracy, Accountability, and Representation New York: Cambridge University Press. Przeworski, Adam. 1986. Capitalism and Social Democracy New York: C ambridge University Press. Putnam, Robert. 1973. The Political Attitudes of Senior Civil Servants in Western Europe: A Preliminary Report. British Journal of Political Science 3 (3): 257 290. 1993. Making Democracy Work: Civic Traditions in Modern I taly Princeton: Princeton University Press. Rae, Douglas. 1971. The Political Consequences of Electoral Laws, Revised Edition New Haven: Yale University Press. (RCES), Royal Commission on the Electoral System. 1986. Report of the Royal Commission on the Electoral System: Toward a Better Democracy Wellington: V.R. Ward, Government Printer. Reilly, Ben. 2002. Electoral Systems for Divided Societies. Journal of Democracy 13 (2): 156 170. Renwick, Alan. 2010. The Politics of Electoral Reform: Changing the Ru les of Democracy New York: Cambridge University Press. Riker, William. 1976. The Number of Political Parties: A Re examination of Duverger's Law. Comparative Politics 9 (1): 93 107. 1982. Liberalism Against Populism: A Confrontation Between the Theor y of Democracy and the Theory of Social Choice San Francisco: W.H. Freeman.

PAGE 485

485 1986. "Duverger's Law Revisited." In Electoral Laws and Their Political Consequences eds. Arend Lijphart and Bernard Grofman. New York: Agathon Press. 19 42. Robinson, Willi am. 1950. Ecological Correlations and the Behavior of Individuals. American Sociological Review 15 (2): 351 357. Rogowski, Ronald, and Mark Kayser. 2002. Majoritarian Systems and Consumer Power: Price Level Evidence from the OECD Countries. American Journa l of Political Science 46 (3): 526 539. Rose, Richard. 1969. Dynamic Tendencies in the Authority of Regimes. World Politics 21 (4): 602 626. 1984. Do Parties Make a Difference? London: MacMillan. 2007. Learning to Support New Regimes in Europe. J ournal of Democracy 18 (3): 111 126. 2008. Understanding Post Communist Transformation: A Bottom Up Approach New York: Routledge Press. Rose, Richard, and Ian McAllister. 1990. The Loyalties of Voters: A Lifetime Learning Model Newbury Park: Sage. Rose, Richard, and William Mishler. 1996. Testing the Churchill Hypothesis: Popular Support for Democracy and its Alternatives. Journal of Public Policy 16 (1): 29 59. Rose, Richard, William Mishler, and Christian Haerpfer. 1998. Democracy and Its Alternat ives: Understanding Post Communist Societies Baltimore: John Hopkins University Press. Rose, Richard, Doh C Shin, and Neil Munro. 1999. "Tensions Between the Democratic Ideal and Reality: South Korea." In Critical Citizens: Global Support for Democratic Government ed. Pippa Norris. Oxford: Oxford University Press. 146 165 Sakamoto, Takayuki. 1999. Explaning Electoral Reform: Japan versus Italy and New Zealand. Party Politics 5 (4): 419 538. Sanchez, Omar. 2002. An Open Ended Transition: The Effects of Electoral Reform in Italy. Journal of European Area Studies 10 (2): 243 258. Sartori, Giovanni. 1976. Parties and Party Systems Cambridge: Cambridge University Press.

PAGE 486

486 1986. "The Influence of Electoral Systems: Faulty Laws or Faulty Method?" In Electo ral Laws and their Political Consequences eds. Bernard Grofman and Arend Lijphart. New York: Agathon Press. 43 68. 1994. Comparative Constitutional Engineering New York: New York University Press. Sassoon, Donald. 1995. Tangentopoli and the Democrat ization of Corruption. Journal of Modern Italian Studies 1 (1): 124 143. Sautet, Frederic. 2006. Why Have Kiwis not become Tigers? Reforms, Entrepreneurship, and Economic Performance in New Zealand. The Independent Review 10 (4): 573 597. Schattschneider, EE. 1950. Toward a More Responsible Two Party System. American Political Science Review 44 (3): 1 99. Scheiner, Ethan. 2008. Does Electoral System Reform Work? Electoral System Lessons from Reforms of the 1990s. Annual Reviews of Political Science 11 (1): 161 181. Schmitt, Hermann, and Soren Holmberg. 1995. "Political Parties in Decline?" In Citizens and the State eds. Hans Dieter Klingemann and Dieter Fuchs. Oxford: Oxford University Press. 352 371. Seligson, Mitchell. 1987. "Development, Democratization, and Decay: Central America at a Crossroads." In Authoritarians and Democrats ed. James Malloy and Mitchell Seligson. Pittsburgh: University of Pittsburgh Press. 167 192. 2005. Improving the Quality of Survey Research in Democratizing Countries. PS: Political Science and Politics 38 (1): 51 56. Shugart, Matthew, ed. 2001. "Extreme" Electoral Systems and the Appeal of the Mixed Member Alternative Edited by Matthew Shugart and Martin Wattenberg, Mixed Member Electoral Systems: The Best of Both Worlds? Oxford: Oxford University Press. Shugart, Matthew, and Martin Wattenberg. 2001. "Mixed Member Electoral Systems: A Definition and Typology." In Mixed Member Electoral Systems: The Best of Both Worlds? eds. Matthew Shugart and Martin Wattenberg. Oxford: Ox ford University Press. 9 24. Siaroff, Alan. 2003. Spurious Majorities, Electoral Systems, and Electoral System Change. Commonwealth and Comparative Politics 41 (2): 143 161. Silverstone, Brian, Allan Bollard, and Ralph Lattimore, eds. 1996. A Study of Econ omic Reform: the Case of New Zealand Amsterdam: Elsevier.

PAGE 487

487 Sinclair, Keith. 2003. A History of New Zealand Auckland: Auckland University Press. Snyder, James. 1989. Election Goals and the Allocation of Campaign Resources. Econometrica 57 (3): 637 661. Spa fford, Duff. 1972. Electoral Systems and Voter's Behavior. Comparative Politics 5 (1): 129 134. Spotts, Frederic, and Theodor Wieser. 1986. Italy: A Difficult Democracy New York: Cambridge University Press. Taagepera, Rein, and Matthew Shugart. 1993. Seat s and Votes New Haven: Yale University Press. Thompson, Dennis. 1970. The Democratic Citizen Cambridge: Cambridge University Press. Tsebelis, George. 1990. Nested Games: Rational Choice in Comparative Perspective Berkley, CA: University of California Pr ess. 1999. Veto Players and Law Production in Parliamentary Democracies: An Empirical Analysis. American Political Science Review 93 (3): 591 608. Vassallo, Salvatore. 2005. "The Constitutional Reforms of the Center Right." In Italian Politics: Quo V adis? eds. Carlo Guarnieri and James Newell. New York: Bergham Books. 117 135. 2007. Government Under Berlusconi: The Functioning of the Core Institutions in Italy. West European Politics 30 (4): 692 711. Verzichelli, Luca. 1997. "The Majoritarian Sy stem, Act II: Parliament and Parliamentarians in 1996." In Italian Politics: The Center Left in Power eds. Roberto D'Alimonte and David Nelken. Boulder: Westview Press. 143 168. Verzichelli, Luca, and Francesco Zucchini. 2002. "The New Parliament and the Start of a Decisive Legislature." In The Italian General Election of 2001 ed. James Newell. New York: Manchester University Press. 217 237. Vowles, Jack. 1992. Party Strategies and Class Composition: The New Zealand Labour and National Parties in 1988 and Beyond. New Zealand Sociology 7 (1): 36 61. 1995. The Politics of Electoral Reform in New Zealand. International Political Science Review 16 (1): 95 115. 1998. "Countdown to MMP." In Voters' Victory? New Zealand's First Election Under Proportion al Representation eds. Jack Vowles, Peter Aimer, Susan Banducci and Jeffrey Karp. Wellington: Auckland University Press. 12 28.

PAGE 488

488 2000. Introducing Proportional Representation: the New Zealand Experience. Parliamentary Affairs 53 (4): 680 696. 200 2. Proportional representation on trial Auckland, N.Z.: Auckland University Press. 2005. New Zealand. European Journal of Political Economy 44 (4): 1135 1140. 2005. "New Zealand: The Consolidation of Reform?" In The Politics of Electoral Systems eds. Michael Gallagher and Paul Mitchell. Oxford: Oxford University Press. 295 313. 2006. New Zealand. European Journal of Political Research 45 (4): 1207 1220. Vowles, Jack, Peter Aimer, Susan Banducci, and Jeffrey Karp. 1998. "Expectations of Chan ge." In Voters' Victory? New Zealand's First Election Under Proportional Representation eds. Jack Vowles, Peter Aimer, Susan Banducci and Jeffrey Karp. Wellington: Auckland University Press. 1 11. Vowles, Jack, Peter Aimer, Helena Catt, Raymond Miller, an d James Lamare. 1993. "New Zealand Election Study." In, ed. University of Auckland Department of Political Studies. Auckland: Foundation for Research, Science, and Technology. Vowles, Jack, Susan Banducci, Jeffrey Karp, Peter Aimer, Helena Catt, and Raymon d Miller. 1996. "New Zealand Election Study." In, ed. Waiko School of Social Science Research Committee. Auckland: University of Auckland Research Committees and Lottery Sciences. Vowles, Jack, Susan Banducci, Jeffrey Karp, Peter Aimer, and Raymond Miller. 2002. "New Zealand Election Study." In. Auckland: Foundation for Research, Science, and Technology. Vowles, Jack, Susan Banducci, Jeffrey Karp, Raymond Miller, and Ann Sullivan. 2005. "New Zealand Election Study." In, ed. University of Auckland. Auckland: New Zealand Electoral Commission. Vowles, Jack, Susan Banducci, Jeffrey Karp, Raymond Miller, Ann Sullivan, and Jennifer Curtin. 2008. "New Zealand Election Study." In, ed. University of Auckland. Auckland. Vowles, Jack, Jeffrey Karp, Susan Banducci, and Peter Aimer. 2002. "Public Opinion, Public Knowledge, and the Electoral System." In Proportional Representation on Trial: The 1999 New Zealand General Election and the Fate of MMP eds. Jack Vowles, Peter Aimer, Jeffrey Karp, Susan Banducci, Raymond Miller and Ann Sullivan. Auckland: Auckland University Press. 160 174. Weatherford, Stephen. 1987. How Does Government Performance Influence Political Support. Political Behavior 9 (1): 5 29.

PAGE 489

489 1991. Mapping the Ties that Bind: Legitimacy, Representation, and Alienation. Western Political Quarterly 44 (2): 251 277. 1992. Measuring Political Legitimacy. The American Political Science Review 86 (1): 149 167. Zuckerman, A. 1997. Transforming a Peripheral Region: The Consolidation and Collapse of Christian D emocratic Domination in Basilicata. Regional and Federal Studies 2 (1): 1 24.

PAGE 490

490 BIOGRAPHICAL SKETCH Jeffrey S. Hamill is from Narragansett, Rhode Island and moved to Gainesville to pursue a Ph D in political science at the University of Florida. He ea rned a Ph.D with a specialty in comparative politics in the spring 2012. He receiv ed a Masters of Arts in Political Science from the University of Rhode Island in 2002 after earning a Bachelor of Arts in history and p oli tical s cience the same university i n 2000. His interest in politics and elections dates back to the term he served on the Na rragansett School Board from 1996 to 1998. His career goal is to study the relationship between elections and whether they can increase the attachments between citiz ens and their governing i nstitutions