This item is only available as the following downloads:
1 WHAT DETERMINES ECONOMIC SATISFACTION IN ETHNICALLY POLARIZED SOCIETIES? THE IMPORTANCE OF TRUST IN POLITICAL INSTITUTIONS By AMANDA RAGNAUTH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTI AL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2012
2 2012 Amanda Ragnauth
3 To the people of Guyana
4 ACKNOWLEDGMENTS I thank my parents for their continuing love and guidance. I am here only b ecause of their sacrifices. I am indebted to my chair, Dr. Robert White, for all of his help throughout the conducting of this study and for sharing his expertise in empirical research. I thank my committee member, Dr. Alin Ceobanu for encouraging me to further my studies in sociology and for his advice throughout the course of this research. Finally, I am very grateful to my brother and my friends for all of their love, support and advice.
5 TABLE OF CONTENTS page A CKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 LIST OF ABBR EVIATIONS ................................ ................................ ............................. 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 2 ECONOMIC WELF ARE AND TRUST IN POLITICAL INSTITUTIONS ................... 16 Economic Satisfaction ................................ ................................ ............................. 16 Political Trust ................................ ................................ ................................ .......... 21 3 POLITICAL FRAGMENTATION AND ETHNIC POLARIZATION IN GUYANA ....... 31 4 ANALYTIC STRATEGY ................................ ................................ .......................... 36 Economic Satisfacti on ................................ ................................ ............................. 36 Political Trust Index ................................ ................................ ................................ 38 Socio Demographic Measures ................................ ................................ ................ 40 5 R ESULTS ................................ ................................ ................................ ............... 50 6 DISCUSSION ................................ ................................ ................................ ......... 58 7 CONCLUSION ................................ ................................ ................................ ........ 62 LIST OF REFERENCES ................................ ................................ ............................... 66 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 70
6 LIST OF TABLES Table page 4 1 Descriptive statistics (standar d deviations appear in parentheses below the mean values) ................................ ................................ ................................ ...... 46 5 1 Predicting economic satisfaction with SES and political trust (N = 2,329). ......... 55 5 2 Comparing results between multiply imputed sample and observed sample ..... 56 5 3 Effects of political trust and SES on economic satisfaction, stratified by ethnic group ................................ ................................ ................................ .................. 57
7 LIST OF FIGURES Figure page 2 1 The theoretical relationship between political trust and economic satisfaction. .. 30 4 1 Economic satisfaction by ethnicity ................................ ................................ ...... 47 4 2 Mean political trust by economic satisfaction and ethnicity ................................ 48 4 3 Income distribution by e thnicity ................................ ................................ ........... 49
8 LIST OF ABBREVIATION S LAPOP L atin American Public Opinion Project PNC PPP/C UF United Force
9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts WHAT DETERMINES ECONOMIC SATISFACTION IN ETHNICALLY POLARIZED SOCIETIES? THE IMPORTANCE OF TRUST IN P OLITICAL INSTITUTIONS By Amanda Ragnauth August 2012 Chair: Robert White Major: Sociology This study examines the importance of political trust in ethnic disparities in economic satisfaction. Previous research has shown that economic satisfaction is an important factor in voter preference and is highly correlated with socio economic factors, such as education, income, and recent change in individual economic circumstance. In polarized societies, ethnic differences in economic welfare have often been th ought to exert an additional influence on differences in economic satisfaction. This study examines the political origins of ethnic differences in economic satisfaction. Trust in political institutions is an important element in civic engagement and vote r preference. I examine the importance of political trust in the paradoxical setting of Guyana: a country in which the ethnic groups with the highest income and education levels display the lowest levels of economic satisfaction. A history of racial politics has led to a situation in which the party in power predominantly represents a particular ethnic group and is perceived to enact discriminatory policies against other ethnic groups. In this setting, political trust is not only positively associated with economic satisfaction, but it also explains a large share of ethnic differences in economic satisfaction. These results
10 point to the important role of political trust in voter choice and its lasting effects for racial politics.
11 CHAPTER 1 INTRODUCTION As individuals cast their votes in upcoming elections, they will do so upon evaluation of certain criteria and the importance ascribed to each criterion. Research has consistently pointed to the importance of economic conditions in explaining voter choice. The degree to which voters select political candidates and incumbents based on the performance of the economy has been long debated by political scientists. This tr in economic conditions or in their perceptions (supposedly) lead to changes in individual voting preferences and via that in 1). When selectin g a candidate, individuals evaluate their personal economic circumstances and attribute some degree of responsibility to political institutions and agents. Economic voting involves both a national and a personal component; the state of the national econo my is considered (such as in terms of economic growth, job creation, economic status, recent changes in economic situation, comparisons made with other individuals, and expectations for his or her future economic situation. Some theories of economic satisfaction have focused on the importance of the economic status (SES) and recent changes in eco nomic circumstances. These theories focus on the comparisons which individuals make between their past and present economic circumstances. While other theories emphasize the importance of expectations for future economic performance, they fail to suggest an adequate explanation of how expectations for the future are formed. Some
12 state that those with higher SES will have higher expectations than will those with lower SES. This assumes that high SES individuals expect their past economic prosperity to co ntinue in the future and that low SES individuals also expect their past economic hardships to continue. However, we know that individuals frequently face improvements or worsening of their economic situations. In light of these temporal changes in econo mic situations, it is inadequate to think that individuals base their economic satisfaction on expectations for the continuation of their previous economic circumstances. Rather, the height of their expectations depends on their evaluations of the nationa l economy and on the distribution of resources within the society. Both national economic situation and the distribution of resources are partially dependent on the functioning of political institutions and the actions of political agents. To the extent that individuals trust these political actors to distribute resources in a way that benefits the individual, his or her economic expectations may be raised or lowered. An individual with a high degree of trust in political institutions may believe that th ese institutions will perform in a manner that political institutions of the society. An in dividual with high expectations will believe that his or her economic situation will improve under the incumbent regime and will be likely to rate his or her economic situation positively. Political trust (i.e. the trust that individuals place in politic al institutions) is likely to political cleavages. In these societies, the identification of one or more political party
13 with a particular ethnic group leads to the build ing of trust among individuals of one group and distrust among individuals of other groups. Individuals often come to perceive political actions in terms of their effects on the ethnic group as a whole, and not simply in terms of their effects on the indi vidual. When a political party has a history of discrimination (whether perceived or real) against one or more ethnic group, individuals within this group build distrust for that party. This distrust is predicated on the perception of past discrimination and the expectation of the continuation of discrimination if this party is elected to power. Conversely, individuals belonging to the ethnic group that is aligned with or favored by this party, are unlikely to perceive the inatory and are likely to trust this party to continue to act are partially determined by the degree of trust they have in political institutions. In ethnically pol arized societies, therefore, actual economic circumstances (i.e. SES) may be a less important factor in determining economic satisfaction than in other societies, as SES may be independent of expectations for future economic situation. SES may be more im portant in determining economic satisfaction in societies where ethnic polarization is less extreme. In such societies, individuals may attribute less responsibility to government and/or may believe that government is relatively ineffective in shaping eco nomic circumstances. A history of discrimination by any political party is less salient or absent, and so, individuals feel less certain about the way any party will affect their personal economic situations. Therefore, political trust is expected to var y more greatly among individuals within any particular ethnic group in a non ethnically polarized society than in an ethnically polarized society. In the former,
14 compared to the latter, SES is likely to play a larger role in determining economic satisfact ion because the role of political trust in mediating the effect of perceived discrimination on economic expectations will likely be less significant or entirely insignificant. To assess the relationship between political trust and economic satisfaction, Guyana is used as a case study. Guyana presents an extreme example of ethnic polarization and is especially suited for this analysis because its ethnic cleavages align with political cleavages. It also presents a paradox, as the ethnic groups with the highest income and education levels display the lowest levels of economic satisfaction. A history of racial politics has led to a situation in which the party in power predominantly represents a particular ethnic group and is perceived to enact discrimin atory policies against other ethnic groups. The presence of perceived discrimination by the incumbent party creates a case in which political trust is expected to exert a direct effect on economic satisfaction by altering economic expectations. To estim ate the effect of political trust on economic satisfaction in Guyana, I have analyzed 2009 LAPOP data. Economic satisfaction is measured by an ordinal variable, ranging from one to five. The variable derives from a question that asks respondents to evalu ate their current economic situations. Political trust is measured by an index that is composed of nine indicators, each asking respondents to evaluate their trust in a particular political institution. I compare nested regression models to determine the effects of SES, change in economic situation, and political trust on economic satisfaction.
15 In Chapter 2 I discuss the theoretical motivation for this paper, including discussions of the literature on economic voting, economic satisfaction, political trust, and perceived discrimination. Chapter 3 explains the political and ethnic context of Guyana. In Chapter 4 I discuss the data and methods for measuring the variables, as well as important bivariate relationships. That is followed by a report on my findings and checks for their robustness in Chapter 5 Finally, the paper is concluded with a discussion of the main findings in Chapter 6 and the significance and limitations of this study in Chapter 7
16 CHAPTER 2 ECONOMIC WELFARE AND TRUST IN POLI TICAL INSTITUTIONS economic status has been shown to be an important determinant or her material needs, such as food security and adequate hous ing, as well as his or her position in society. Recent changes in SES, such as an increase in income, have also been shown to be associated with economic satisfaction. However, studies on economic satisfaction, while focusing on SES and changes in SES, h ave omitted the importance of political trust in explaining economic satisfaction. This section reviews the scholarship on economic satisfaction and political trust and points to the need for the inclusion of political trust in theories of economic satisf action. Economic Satisfaction Economic satisfaction is the degree to which an individual is satisfied with his or her economic situation. Research has shown a clear positive correlation between objective socio economic conditions and economic satisfaction Level of income, ownership of goods, access to resources, standard of living, and education all comprise economic subsistence. The theory of absolute utility predicts that satisfaction results from the ability of an individual to satisfy his or her mate rial needs (Veenhoven 2005; Veenhoven and Hagerty 2006). Rising levels of income and ownership of assets should lead to increases in satisfaction because people will have the means to meet their needs and wants. This hypothesis finds considerable support in the literature (Cheung and Leung 2004; Clarke 2000; Alwin 1987). People with higher levels of income and education are more likely to have high satisfaction (Bjrnskov, Dreher, and Fischer 2008). Income is shown to have a positive correlation with s atisfaction at all levels of
17 income, though the strength of the relationshi p decreases as income rises. Furthermore, SES is found to affect satisfaction similarly across race groups, income category, educational level, and counties of drasti cally differen t average income (Diener et al. 1993). These results point to the importance of SES, and unimportance of relative factors (such as comparisons with local others and differences in expectations) in determining satisfaction. Moreover, families with compar able incomes but more children tend to be less satisfied than families with fewer children, a trend that indicates that the sharing of resources among more family members may lead to financial strain and lower levels of satisfaction (Alwin 1987). Economi c satisfaction may also be derived from relative factors. Research indicates that individuals make comparisons with others and are more satisfied when their economic situations are better than the situations of others (Luttmer 2005). Firebaugh and Schroe der (2009) assess the neighborhood effects of income on satisfaction. Interestingly, they find that individuals residing in richer neighborhoods are happier than individuals of comparable income who live in poorer neighborhoods. Yet, individuals who live in poorer counties are happier than individuals of comparable income who live in richer counties. The authors state that these trends indicate that the high quality amenities found in wealthy neighborhoods lead to increased levels of happiness among resi dents, but that individuals are happier when they live in poorer counties because in such counties, the individual has more relative wealth. Differences in economic satisfaction across groups may persist even if the group with lower levels of satisfac tion earns higher income than the group with higher levels of satisfaction. This may be due to expectations of earnings and awareness of historical
18 trends. In a study of US doctors in 1987, white doctors are found to have lower average earnings than mino rity doctors. However, minority doctors report lower satisfaction with their earnings than did white doctors. The authors expect that this is due to awareness of earnings trends. That is, black doctors are aware that blacks face an earnings disadvantage in American society and perceive this disadvantage even though they earn more than whites (Hampton and Heywood 1999). In this case, the authors may the lower satisfac tion of black relative to white doctors may be due to the awareness of historical and contemporary discrimination against blacks. In the absence of knowledge about the earnings of white doctors, black doctors are likely to believe that they are still the victims of discrimination, which results in lower earnings relative to white doctors. This perceived disadvantage may also be due to comparisons which blacks are making between their current earnings and their expected earnings. Even if black doctors ar e aware that they are earning more than whites, they may think that their earnings are still lower than what is deserved by their skill because they deem it unlikely that they are being compensated equally with whites of similar skill level. When they ass ess their satisfaction, they may be comparing their current earnings to the earnings they would expect to receive under a non discriminatory political regime. Thus, they may be comparing their current earnings to their expected earnings under favorable po litical conditions. In addition to making comparisons with others, research indicates that individuals also make comparisons with their previous economic situation. Many scholars emphasize the role of recent changes in economic situation. Easterlin ass erts that
19 socio economic factors only lead to increases in satisfaction in the short term, but as income increases, so too do aspirations. Thus, the more an individual has, the more he or she wants; this is known as the adaptation hypothesis (Easterlin 20 01). Although claim that increases in per capita GDP are not associated with in creases in satisfaction. Tian and Yang (1993) argue that increases in income lead to higher levels of satisfaction, but only to a point; they call this point the critical level. After the critical level is reached, increases in income are no longer corre lated with satisfaction. Though the effects may be short lived, research indicates that individuals who have experienced recent improvements in their economic situations are more likely to approve of their current economic situations than are those who ha ve experienced no improvement (Fiorina 1978; Key and Cummings 1966; Remmer 1991; E. Diener et al. 1993). The importance of recent economic changes has been emphasized in the literature on economic satisfaction; in addition to being satisfied with recent i mprovements in economic situation, individuals may also derive satisfaction from the belief that they will continue to experience improvements in the future. Social scientists debate whether comparisons with the past or expectations for the future are mo re influential in determining voter choice. Proponents of the retrospective hypothesis argue that individuals judge their current economic situations in relation to their past economic situations. On the other hand, proponents of the prospective hypothes is contend that individuals place greater emphasis on expectations for future economic performance than on past economic performance (H. W. Chappell and Keech
20 1985; MacKuen, Erikson, and Stimson 1992). Expectations are partially shaped by perceptions of j ustice. Individuals who believe that their current income levels are lower than the income they deserve are less satisfied with their material circumstances than are individuals who perceive their income levels to be fair or above what they deserve, given any level of income (Younts and Mueller 2001; Alwin 1987). The difference between actual income and deserved income is found to be much more important for determining material satisfaction than is actual income. Furthermore, controlling for income, thos e with higher education are less satisfied than are those with lower education (Alwin 1987 ; Lee and Choi 2005 ). This suggests that education may be a strong determinant of expected income, as highly educated people are likely to expect high levels of inco me and may be more susceptible to disappointment. Individuals with high expectations for future economic conditions may face more frustration with their economic circumstances than individuals with low expectations (Rattinger 1991; Powell and Whitten 199 3; Carlsen 2000; Dorussen 2002). Although income and wealth correlate directly with economic satisfaction, wealthier individuals generally have higher economic expectations than do poorer individuals (Graham and Pettinato 2006; Birdsall, Graham, and Pettin ato 2000). If the expectations of the wealthy are very high, this may lead to lower economic satisfaction among the wealthy than among the middle and lower classes. In addition to individual level factors, the political and economic contexts of a state are important factors in economic voting. In societies where political partisanship is low, economic voting is more likely to be important (Kayser and Wlezien 2011; Lockerbie 1993). The degree of centralization of political power is also likely to affect economic
21 voting; voters are more likely to assign responsibility for economic conditions to political actors in states with a high degree of political centralization (Chappell and Gonalves Veiga 2000; Anderson 2006; Remmer 1991; Powell and Whitt en 1993). The degree of responsibility that voters assign to political institutions for their economic situations depends on their perceptions of the effectiveness of these institutions and the fairness by which they operate. People who have different vi ews on the way that income is distributed in society are likely to have different levels of satisfaction. At similar levels of income, those who believe that income is distributed fairly are more likely to be satisfied than are those who believe that inco me is distributed unfairly (Bjrnskov et al. 2008). This can be explained by varying levels of expectation s as those who are unsatisfied may believe that they are not being rewarded fairly and that they deserve more than their current income. Individual when evaluating their satisfaction, and their perceptions of fairness are reflected in the degree of trust that they have in these institutions. Political Trust In this study, political trust will be defined as the probability that the political system (or some part of it) will produce preferred outcomes even if left untended (Gamson 1968). Trust implies a relationship between two parties one that is giving trust and one that is receiving it. T rust is based on past and present experiences between two parties, with expectations for future experiences. An individual who gives trust to the these institutions If the institutions fail to act in accordance with the expectations of the
22 The trust of the population has important consequences for political institutions. A democratic government must maintain an adequate level of political trust from the population in the hopes of maintaining stability and gaining re election. In a given society, trust in government may be lower for some groups than for others. Political cynicism typically arises from a prolonge d period of political alienation and discontent seated social conflict which, for some segment of the population, has been translated into a negative orientation toward the political system because their sen se of insufficient political influence implies a futility in bringing about desired social change or control through po 951). The previous policies of a government, along with the characteristics of the majority of officehold ers (e.g. race and sex), may lead to low trust among a particular segment of the population. Discrepancies in political trust have received much attention since the 1970s after noting a dramatic decline in trust among Americans between 1964 and 1970 (Mille r 1974). In the U.S., survey research has relied upon the four to five questions intended to measure political trust in the National Election Survey. These questions measure is right; whether the government is run by a few big interests or for all of the people; the efficiency with which tax money is used by government; and the crookedness of government officials. These questions, though poorly worded, are intended to uncove r Most studi es of political trust in the United States have used a four item index composed of these indicators; some have used three of these items (Bennett 2001) and
23 o thers have used two (Citrin and Green 1986). Justification for the use of some items over others has been inadequate, as political scientists typically neglect to discuss the underlying concept of trust which these items are intended to measure. Some, how evaluated. While the standard political trust index measures trust in the national government, it fails to account for trust in the state government, which may be greater among those who fav or small government (Hibbing and Smith 2004). There is also debate about whether individuals value political outcomes more than political processes. Those who argue for the importance of political outcomes claim that the smaller the discrepancy between i by government, the greater trust individuals have in government (Citrin and Green 1986; Hetherington 1998). Others contend that the means by which these policies are enacted are more important. Thus more than policy outcomes, individuals may value the honesty of government officials (Lipset and Schneider 1987) and the degree to which officials are influenced by private interest groups (Blendon et al. 1997). In many underdeveloped nations, po litical trust has been measured differently than whole, political trust items typically measure trust in particular institutions of the state. Some studies differentiate b etween purely state institutions (e.g. the judiciary and the Parliament) and state social institutions (e.g. mass media and the education system) (Finkel, Sabatini, and Bevis 2000). In Latin America and the Caribbean, military institutions usually garner the most trust, while the judicial system typically receives the least (Seligson 2007). In China, studies find that rural Chinese citizens are likely to trust
24 the central government, but perceive the local government as corrupt (Bernstein and L 2000; Shi 2001; Li 2004). The trust that individuals afford to political institutions is based on past, current, and expectations for future relationships between individuals and institutions. Individuals trust governments which may be thought to have performed well in the past. Good performance may include the maintenance of a strong economy (Chanley, Rudolph, and Rahn 2000; Wong, Wan, and Hsiao 2011), low levels of corruption (Anderson and Tverdova 2003; Seligson 2006), and high institutional capacity (Hutchis on and Johnson 2011), while fulfilling pre election promises (Roberts and Arce 1998). performance of political institutions in the future, and the effect that this perfor mance will trust. Individuals are likely to have high levels of political trust if they perceive political representatives as honest and uncorrupt (Gronlund and Setala 2 010) and if they have positive expectations for the future of the economy (Lockerbie 1993; Hetherington and economic situation; those with greater trust in governm ent may have higher expectations than those with less trust because the former believe that the government will work to improv e their economic situations (Figure 2 1). Government policies will affect individuals disproportionately based on their social po sitions (e.g. a tax break for high income earners will benefit high SES individuals, but to the detriment of low SES individuals, who depend on the public services that are funded by tax revenues). So,
25 for low SES individuals, the election of a political candidate who campaigned on a willingness to improve their economic situations. Economic expectations may mediate the relationship between political trust and econo mic satisfaction in two different ways ( Figure 2 1). First, trust in government may lead to higher expectations, as discussed above. Individuals with high expectations, however, will require greater improvement in their economic situations to be satisfie d, compared with individuals with low expectations. Even if trustful individuals experience some improvement in their economic situations, they still may be dissatisfied if their improvement did not meet their expectations. On the other hand, distrustful individuals may have low expectations for their future economic situations; in this case, only a slight improvement in their economic circumstances may lead to their satisfaction. While in this model, distrustful individuals may seem more easily satisfi ed than government to be altered. Individuals who are cynical toward government may believe that either the government should not be involved in economic affairs or that the g overnment is ineffective in carrying out its policies. This negative orientation toward government may cause one to attribute any improvements in economic circumstances to either oneself or to non political forces and any deterioration to the government. situation has recently improved. This may occur if the individual believes that the improvem ent. The individual might believe that his or her economic situation would be
26 better had the government not intervened in the economy or had it chosen to enact a e economy and in the effectiveness of government may affect his or her economic satisfaction. evaluations of their past and current economic situations. Evans and Pickup (2010), using National Election Studies data from 2000, 2002, and 2004, find no evidence for a significant effect of economic satisfaction on presidential approval; rather, they find strong evidence for the effect of presidential approval on economic satisfaction Respondents were more likely to evaluate their economic situations favorably if they approved of the incumbent government than if they disapproved of it. These results r economic situations. In the American context, where partisanship is fairly strong, beliefs about the government are an important determinant of economic satisfaction. ec onomic situations. Though some individuals may have high expectations, which remain unmet, they may be economically satisfied if they trust that the current government is doing all it can to improve their economic situations. On the other hand, individua ls with low trust in government may believe that the government is not doing enough for them; rather, they may believe that their economic situations would be better under a different government. Political trust, therefore, may affect economic satisfactio n either directly or indirectly through altering economic expectations or views on government ( Figure 2 1).
27 The effects of trust on economic satisfaction may be especially pronounced in societies characterized by a high degree of ethnic polarization. In such societies, political partisanship may align with ethnic cleavages. In this case, political parties may tend to enact policies that disproportionately benefit the ethnic group from which it draws the majority of its support. Some regimes develop a history of racist policies, and the lingering of this history may lead to trustfulness among the supporting ethnic group and cynicism among the excluded ethnic group. The Civil Rights Movement provided the opportunity for the study of differences in pol itical trust between different racial groups in the United States. Aberbach and Walker (1970) analyzed the relationship between attitudes toward civil rights and political trust for blacks and whites living in Detroit in 1967 during the time that the city was affected riots in Detroit, and support for militant black leaders) had lower le vels of political trust than blacks who did not adhere to this ideology. White integrationists had higher levels of political trust than did white segregationists. Interestingly, they found that among lower status blacks, their perception of deprivation is negatively associated with political trust; however, socio economic indicators such as income, education, and occupation are not associated with political trust (Aberbach and Walker 1970). The perception of deprivation may be rooted in the historical legacy of state discrimination against blacks. Thus, blacks may be perceiving deprivation even if they have higher levels of income and education than comparable whites. This is because they know that discrimination has existed in the past and they expect that it continues to
28 exist. So, when they assess the level of trust that they have for political institutions, they may be doing so by making comparisons between their current economic situations and the economic situations they would have if they were n ot affected by discriminatory practices. The direction of the relationship between perceived deprivation and political trust is not clear, however. It may be that blacks who have greater trust in the political institutions are less likely to perceive dis crimination and deprivation than are blacks with low political trust. previous performance of political regimes, the historical legacy of these regimes may be enduring and may that feel particularly marginalized by the incumbent regime can be expected to report lower levels of trust in government. Trusting individuals are more likely to believe the incumb ent government is working for their benefit and will continue to do so. In this individuals believe that the incumbent government is not benefiting them and if another regime was in power, then their economic situations would improve. Thus, political trust reach their economic expectations. The effect that political trust has on eco nomic satisfaction also depends on the extent that individuals believe government is responsible for their economic situations and that government is effective in influencing economic events. Contexts in which discriminatory legacies are attached to one or more regime may lead to the interpretation of political acts in terms of their discriminatory effects on one
29 group of people. A number of studies have found a negative correlation between perceived discrimination and political trust (Schildkraut 2005; Sanchez and Brock 1996; Avery 2009; Uslaner 2008). These findings indicate that perceived discrimination may act through political trust to affect economic satisfaction. In the United States, blacks were found to have higher levels of political trust in a city where a black mayor and black administration had held office for eight years than in include evaluations of political incumbents (Howell and Fagan 1988). That blacks are considering the race of the holders of political office indicates that they have greater trust in black than they do in white office holders. This may be because they have greater interpersonal trust in blacks than in whites, but it may also b e because they expect more fair treatment from black officeholders than from white officeholders. In a context in which black officeholders have historically been few and blacks have suffered from continued discrimination from white officeholders, blacks may come to place great value on the race of the officeholder. In ethnically polarized societies, political cleavages often arise which lead to exclusionary policies by the incumbent regime. The legacy of these policies tends to persist even if policie s become neutral over time. When ethnic groups are politically divided, individuals within these groups may perceive the actions of the rival political regime as discriminatory. This perception of discrimination generally leads to political distrust. Po litical distrust may lower the economic satisfaction of individuals who affiliate with the political regime that is out of power. This idea will be applied to the context of Guyana, as its history of ethnic and political cleavages make it an ideal case s tudy.
30 Figure 2 1. The theoretical relationship between political trust and economic satisfaction.
31 CHAPTER 3 POLITICAL FRAGMENTAT ION AND ETHNIC POLARIZATION IN GUYANA Guyana is studied here because it exemplifies extreme ethnic polarization. The two largest ethnic groups, East Indians and Blacks, comprise 44% and 30% of the population, respectively. These two groups have developed separately since the 1800s, and they remain informally segregated by residence, occupation, and cultural institutions With events beginning in the 1950s, they also became divided in the patterns. It presents a paradox in that the ethnic group with the most political power (Indi ans) is not the one with the most economic power. While Blacks report higher levels of income, consumption, and education than Indians, Blacks also report lower satisfaction with their economic situations. I hypothesize that these relatively low levels o f economic satisfaction among Blacks, compared with groups of lower SES, are partially explained by different levels of political trust. With relatively high SES, but under representation in government, Blacks may have high expectations for their future e conomic situation, but little trust that the government will facilitate the meeting of these expectations. Indians, on the other hand, may have lower expectations due to their lower SES, but great trust in government due to their political over representa tion. Ethnic politics in Guyana developed through a series of events that caused political divisions to align with ethnic ones. Blacks and Indians have voted for two different political parties for the past fifty years. These two parties were once unit ed under one name individuals, one Indian (Cheddi Jagan) and one Black (Forbes Burnham). However, due to British and US intervention in the 1950s (Waters and Daniels 2005; Rabe 2005),
32 these two leaders became rivals and Burnham and his faction broke away from the PPP the populace, with the overwhelming majority of Blacks voting ritually for the PNC and 1982). Both parties were determined to lead the country to independence, which occurred in 1966. While the PPP won the elections in 1961, it lost to the PNC in 1964, after the PNC formed a coalition wi th the United Force party (UF). The PNC, which split from the UF in 1968, rigged every election until 1992, when international observers were present to ensure fair elections (Gafar 2004; Rabe 2005). Since independence in 1966, there has been only one regime change, which occurred in 1992 when the PPP 1992, older Indians recall the discriminatory practices by the PNC regime in the 1970s and 1980s. The youth are being socialized into this context of ethnic animosity and racialized politics continue to divide the population The political divisions among ethnic groups are compounded by economic differences. As previously mentioned, I ndians and Blacks form the majority of the population. These two groups have developed separately since their introduction to the country. Many Blacks, brought to Guyana as slaves by the British, migrated to the urban centers after emancipation in the mi d 1800s. Indians were brought to the country as indentured servants, but most elected to stay in the rural areas even after their contracts expired (Waters and Daniels 2005; Premdas 1995). Blacks were preferred by the British for public sector jobs, such as teaching or the police force, due to their
33 greater extent of creolization 1 (Hintzen and Premdas 1982). Today, they continue to form the majority of public sector workers, though this sector has become much more heterogeneous than in the past (Brown 19 99) With neoliberalism and its prescriptions for smaller government, the public sector was drastically reduced in size in the 1990s; this reduction disproportionately hurt the Black population (Gafar 2004). Indians, on the other hand, form the majority of the farming population and are more likely to be self employed. Their economic welfare is closely tied to international prices of rice and sugar and to domestic policies that affect irrigation, infrastructure, crop yields and preferred markets. Diffe rences in occupation between Blacks and Indians have caused the economic policies of the PPP and PNC to be interpreted in ethnic terms. Ethnic animosities erupted into ethnic conflict in the aftermath of elections and policies that were perceived as discr iminatory. After the PNC UF coalition won the national elections in 1964, the PPP and its supporters felt cheated out of a victory and immediately staged demonstrations and boycotts in protest of the PNC UF regime. Ethnic animosities were further exacerb ated when the PNC UF coalition arrested several PPP activists, withdrew subsidies that had been granted to rice farmers (who were predominantly Indian) under the PPP administration, and eliminated the Cuban market for rice (which had previously yielded a h igh price to rice farmers). The PNC continued to alienate that Indians should be rapidly recruited into the African dominated Police Force and the Defence Force to achiev e ethnic balance (Rabe 2005; Premdas 1995). The PNC also 1 Creolization is a term used to describe the extent to which a group of people, brought to the Caribbean, have acculturated to the new climate.
34 staffed public agencies such as banks and bauxite companies with its communal supporters, drastically expanded the police force (which was still mostly composed of blacks), and attempted to locate l and less blacks on lands previously leased to Indians. The Indians in turn, boycotted rice production under the direction of Jagan, which created grave food shortages in Guyana (Hintzen and Premdas 1982; Premdas 1995). Since 1992, the PPP has held power Blacks, who view the PPP policies to be pro Indian (Misir 2002; Gafar 2004). Many of poverty policies are targeted to rural and interior areas, which are predominantly inhabited by Indian s and Amerindians, respectively. While little evidence exists to support a socio economic disadvantage of Blacks (since, on average, they have higher levels of education and income than do Indians), the disadvantage perceived by Blacks may be due to feeli ngs of political exclusion derived from their insufficient representation in the government. Since voting continues along ethnic lines, elections have been reduced to a numbers game; whether the PPP or PNC is the party in power depends simply on the numbe r of Indians and Blacks. This study seeks to determine whether differences in political trust explain some of the difference in economic satisfaction among ethnic groups in Guyana. First, I expect to find that political trust is a significant predictor of economic satisfaction for all ethnic groups. Political trust may mediate the effect of socio economic status indicators on economic satisfaction. For instance, an individual with high income but low political trust may be less satisfied with his or he r economic situation than an individual with high income and high political trust if the former individual believes that the political institutions do not represent his or her interests; in fact, he or she may believe that if the
35 political institutions wer e responsive to his or her needs, he or she would have even more affluence. Thus, individuals may be comparing their current incomes to their institutions in place. In dividuals with high SES but low political trust may also be uncertain about the status of their wealth. They may be wary of political institutions that may adopt redistributive policies and increase taxes on the wealthy So, this uncertainty may cause th ese individuals to have lower economic satisfaction than would be predicted by their SES. Second, I expect Blacks and Mixed people with low levels of political trust to be the least economically satisfied. On average, these groups have higher SES than Ind ians and Amerindians. This high SES may cause them to have higher expectations for their economic conditions, and with low levels of political trust, they are expected to be very dissatisfied with their economic situations. When comparing their current e conomic situations to their potential economic situations, they are doing so with greater economic resources than other groups have, but Blacks and Mixed people may also have disproportionately high expectations.
36 CHAPTER 4 ANALYTIC STRATEGY This study ana lyzes Latin American Public Opinion Project data collected in 2009 combination of stratified, cluster, and weighted sampling, the survey included 2,514 voting aged, non insti tutionalized adults in Guyana. The sampling distribution was based on 2002 Census data. The sample was first stratified into seven geographical areas. A multi stage cluster sample selected municipalities, villages, enumeration districts, and finally hou seholds. Data was collected using in person interviews from one adult per household. In my analyses, I restrict the sample to Blacks, Indians, Amerindians, and Mixed peoples due to the very low numbers from each of the other ethnic groups. 1 The analyse s are based on 2,329 cases; 35.5% of respondents are Indian, 29.3% are Black, 21.2% are Mixed, and 13.2% are Amerindian (Table 4 1) Economic Satisfaction of their overall econ would you describe your overall good nor bad (fair) 4) Bad 5) Very Bad. These answer choices were recoded so that 1 indicates Very Bad and 5 indicates Very Good overall economic situation. This question 1 There were thr ee additional ethnic categories: Chinese, Portuguese, and Other. These were excluded because only 18 respondents identified themselves as belonging to one of these categories. In Guyana, ethnic categories are more clearly demarcated (with the exception o f the Mixed group) than in most Caribbean and Latin American nations.
37 ref f 0.90. Differences in economic satisfaction by ethnic group are depicted in Figure 4 1 Blacks were most likely to have very low and low levels of economic satisfaction and least likely to have high economic satisfaction On the other hand, Indians were most likely to have high and very high levels of economic satisfaction and least likely to have very low and low economic satisfaction. Interestingly, Amerindians, by far the most economically disadvantaged group, appear to be more economically satisfied than are Blacks. The relationship betwee n economic s atisfaction and ethnicity is statistically significant, with a chi square value of 90.3 (p=.000). The bivariate relationship between economic satisfaction and ethnicity is as predicted Indians have the highest economic satisfaction and Blacks have the lowest. This coincides with their degree of political representation in the national government; the party in power is predominantly composed of Indians, while the main opposition party is predominant ly composed of Blacks. Figure 4 1 reveal s that the group with greater political representation is also the group with greatest economic satisfaction. My analysis will be based on ordered logit regression models. The dependent variable is economic satis faction and the independent variables will be: an index for political trust, ethnicity, and controls for household income, education, an index for asset ownership, non liquid assets, change in economic situation, region, gender, and age. The descriptive s tatistics for all variables are reported in Table 4 1. I expect that political trust is able to explain the differences in economic situation among different
38 ethnic groups. Though economic expectations are not observed, I think that political trust may m ediate the effect of ethnicity on economic satisfaction by altering economic expectations. Political Trust Index Respondents were asked to rate their degree of trust in fourteen different political institutions on a scale of one to seven, where one is no t rust and seven is a lot of trust. Exploratory factor analysis was conducted to determine which items reflect a common underlying concept of political trust. Fourteen items were included in the analysis, which measured trust in political, legal, and burea ucratic institutions. I used the Maximum Likelihood extraction method, because it allows for the determination of statistical significance of the factor loadings and correlations between factors, as long as the data are normally distributed (i.e. the abso lute value of skewness is below 2 and the kurtosis is below 7) (Fabrigar et al. 19 99 ). Since the data are relatively normally distributed, Maximum Likelihood extraction was utilized. The variables were rotated obliquely using direct oblimin rotation. Ob lique rotation was deemed preferable to orthogonal rotation, because oblique rotation allows for correlation between factors and it is expected that there will be some correlation between different factors of political trust. Since the different types of oblique rotation have been shown to yield similar results, I selected direct oblimin rotation, as it is the most commonly used (Costello and Osborne 2005). The exploratory factor analysis yielded a single factor (p=.000), which explains 52.2% of variance in the underlying concept. Factor loadings range from .587 to .806 and communalities range from .345 to .649; trust in the Guyana Defence Force and
39 analyses, because less than 40% of the variation of these items were shared by the other items. I also excluded trust in Regional Democratic Councils, the Attorney General, and the Integrity Commission due to high numbers of missing cases. A political trust index was computed by summing the remaining nine items, each with equal weight. These nine items measured trust in: the justice system, the Guyana Elections Commission (GECOM) the Parliament, National Government, Guyana Police Force, the political parties, the President, Supreme Court, and elections. Each item was measured on a seven point scale, where 1 indicates very little trust and 7 indicates very high trust. Descriptive statistics for the individual political trust indicators are found in Table 4 1. Each indicato r was recoded from 1 7 to 0 6 and subsequently summed to form a political trust index. The index has a high internal reliability, with a and standard deviation of 13. Missing values for the nine political trust indicators ranged from 43 to 163. To maximize the sample size and correct for non random missingness these missing values were replaced with imputed values generated from multiple imputation. The missing va lues of each political trust indicator were predicted using chained ordinal logit equations, which contained covariates for the other eight political trust indicators, as well as income, education, age, sex, region, and ethnicity. 2 For each political trus t indicator, ten predicted values were generated and summed in a political trust index. The mean of the multiply imputed index is slightly lower than the mean of the index containing only observed values. Since the sample size of the latter is substantia lly 2 While missing values for education and income were also imputed in this procedure, their imputed values were excluded from the analytic sample.
40 smaller (it omits 15% of cases), the multiply imputed index is used for analysis (N=2,329). The political trust index yields a pattern of ethnic differences in political trust. Table 4 1 reports t he average levels of political trust by ethnic grou p, which are: 22.9 for Blacks, 26.3 for Mixed people, 34.5 for Indians, and 34.4 for Amerindians (p=.000). Differences in political trust across ethnic groups are also examined for different levels of economic satisfaction ( Figure 4 2 ). As economic satis faction improves, the mean values of political trust increase (represented by the red dots). Across all levels of economic satisfaction, political trust is consistently higher for Amerindians and Indians than for Blacks and the Mixed group Additionally, for all groups except Amerindians, political trust appears to increase linearly as economic satisfaction improves. Socio Demographic Measures I expect that differences in economic satisfaction correlate with differences in socio e conomic indicators. I in clude the following variables as controls in subsequent analyses: direct measures of economic resources (income, education, and assets) and indirect measures of poverty (potable water and indoor bathroom). 3 Income is measured as an ordinal level variable which of the following income ranges does the total monthly income of this household fit, including remittances from abroad and the income of all the working adults and ranging from No Income 4 to a bove $250,001. Categories 6 ($90,001 $120,000) and 7 ($120,001 $150,000) were 3 Occupation is not included as a soci o economic indicator due to the high proportion of missing cases (47% of all cases had missing values). 4 category.
41 collapsed into one category and categories 8 ($150,001 $200,000), 9 ($200,001 $250,000), and 10 (Above $250,000) were also collapsed into a nother category, because there were very few cases in each of these initial categories. Altogether, t here were 2,466 valid cases. The average total monthly household income of respondents fell between the range of $20,000 to $40,000 and $40,001 to $60,00 0. Differences in income by ethnic group are shown in Figure 4 3 25.7% of Mixed people had incomes of $20,000 or below, compared to 29.3% of Blacks, 31.7% of Indians, and 43.7% of Amerindians. Conversely, 30.1% of Mixed people had incomes of $60,00 1 or above, compared with 26.2% of Blacks, 20.6% of Indians, and 12% of Amerindians. The Chi Square value of 92.5 indicates that the relationship between income and ethnicity is statistically significant (p=.000). Thus, the data suggest that with respect to total monthly household income, Mixed people fare the best, followed by Blacks, Indians, and Amerindians, respectively. To determine the living standard of respondents, the following question was your house: television, refrigerator, conventional or landline telephone, cellular phone, vehicle (one, two, or three or more), washing machine, microwave oven, motorcycle, computer, flat panel TV, hicle was recoded as a dichotomous variable (0 = no vehicle, 1 = 1 or more vehicles), because only 19% of respondents owned one or more vehicles. The inter item correlations revealed low to moderate positive correlations between all items. Motorcycle, hi gh speed internet, and Flat Panel TV were excluded from future analyses, because they were present in very few households: only 12% of respondents possessed a motorcycle, 6% of respondents
42 possessed high speed internet, and 4% of respondents possessed a fl at panel TV. Factor analysis was conducted with the remaining items to determine which to include in a simple count asset index. This yielded two factors with eigenvalues of above 1.0, which explained 45.3% of variance. The first factor included luxury items: computer, vehicle, washing machine, and microwave. The second factor included basic use items: television, refrigerator, and cellular telephone. Factor analysis was then conducted for each of the two factors to ensure that they both reflected inde pendent underlying variables; this was confirmed. A simple count asset index was constructed, which included the seven basic and luxury assets. With an alpha score of .79, the index has sufficient internal reliability. The asset i ndex ranges from 0 to 7, with a mean of 3.27 and a standard deviation of 1.93. Blacks, on average, own 3.68 assets, compared with an average of 3.54 for Indians, 3.58 for Mixed people, and 1.08 for Amerindians. Clearly, Amerindians have the lowest standards of living, as indi cated by the ownership of durable goods. The ratio of between group variation to within group variation is statistically significant, with an F score of 205.8 (p=.000). Other methods of measuring the ownership of durable goods were attempted. I include d the assets (TV, refrigerator, cellular phone, computer, vehicle, washer, and mi crowave) as dummy variables in the full model I also computed a basic asset index (which included ownership of a television, refrigerator, and cell phone) and a luxury asset index (which included ownership of a vehicle, computer, washer, and microwave) to account for the differences in monetary values for these two sets of assets and included these two indices in the model. Additionally, I regressed the above seven
43 assets (a nd all control variables) on political trust. The coefficients for the assets were then used as weights in the computation of a weighted asset index. After including each of these different measures separately the final results of the models displayed n o difference from the model that included the simple count index. The simple count index was included in the final model because it adequately represents variations in wealth and is more interpretable. As Guyana strives to provide improved access to cle an water and sanitation to all of its citizens, the access to potable water indoors and an indoor bathroom are if you have the following in your house: potable water in Potable Water and Indoor Bathroom were included as dummy variables. 63% of respondents reported having potable water indoors and 55% reported having an indoor bathroom. Indians fare the best with regards to safe drinking wa ter, with 73% of them having access to potable water indoors, compared to 69% of Blacks, 62% of Mixed people, and only 18% of Amerindians. There is a statistically si gnificant relationship between p otable w ater and ethnicity (Chi Square = 335; p = .000). 65% of Blacks have an indoor bathroom, compared with 60% of Indians, 60% of Mixed people, and 11% of Amerindians. Again, Amerindians are clearly the most disadvantaged group. level variable that rang ed from zero years of education to eighteen or more. Primary school includes years one school includes years twelve to thirteen, and university or tertiary schooling incl udes years twelve and above. The average number of years of schooling is 8.5, the standard
44 deviation is 3.7 years, and the median is 10 years. 68% of respondents have between 4.8 and 12.2 years of education. On average, Blacks have higher levels of scho oling with 9.5 years. Mixed people average 9.4 years, Indians average 8 2 years, and Amerindians average 6.7 years of schooling. The ratio of between group variation to within group variation has an F value of 74.8, which is statistically significant (p= .000). Previous studies have depicted the effect of recent changes in economic indicators of well being (e.g. income, ownership of goods) on current perception of well being (Easterlin 2002). To control for this effect, a variable for change in responden current economic situation is compared to 12 months ago? Is it better, same, or del as two dummy variables for better economic situation and worse economic s ituation. 29% of respondents indicated that their economic situations have improved over the previous twelve months, while 24% indicated that economic situations have worsened over the previous twelve months. Guyana is one o f few countries in the Americas with a predominantly rural population. Since there are marked differences between standards of living and lifestyles of rural and urban populations, a dummy variable for urban residence is included in the model. 31% of res pondents reside in an urban area. The inclusion of the urban variable also attempts to control for relative deprivation. That is, it allows for the comparison of differences in economic satisfaction among only urban residents, and again among only rural residents. This is admittedly a crude control for the effects of relative deprivation, but in the absence of Census data, it is the best proxy that is
45 available. Control variables are also included for gender and age to exclude the possibility of confoun ding effects.
46 Table 4 1. Descriptive s tatistics ( s tandard deviations a ppear in parentheses below the mean v alues) N Mean SD Black Indian Amerindian Mixed Mean 0.29 0.35 0.13 0.21 Political Trust 2,514 38.1 13.0 22.9 (12.2) 34.5 (11.2) 34.4 ( 10.7) 26.3 (13.4) Justice System 2,410 4.22 1.76 3.84 4.62 4.76 3.98 Elections Commission 2,361 4.20 1.96 3.32 5.00 4.85 3.95 Parliament 2,375 4.28 1.79 3.62 4.88 4.93 3.95 National Government 2,413 4.24 1.84 3.49 4.88 4.97 3.98 Guyana Police Force 2, 471 3.98 1.90 3.47 4.43 4.51 3.69 Political Parties 2,409 3.93 1.73 3.48 4.34 4.37 3.76 President 2,441 4.57 2.12 3.50 5.54 5.37 4.10 Supreme Court 2,351 4.48 1.65 4.07 4.86 4.88 4.33 Elections 2,421 4.18 1.97 3.29 5.00 4.78 3.93 Economic Satisfaction 2,490 3.04 0.90 Very Low 124 .06 .08 7 .0 41 .0 52 .0 66 Low 334 .17 .22 3 119 213 168 Neutral 942 .47 475 453 447 505 High 541 .27 195 338 267 228 Very High 73 .04 .0 21 .0 49 .0 21 .0 34 Income: No Income 158 .07 .0 55 .0 71 .0 95 .0 57 < $10,000 195 .08 .0 75 .0 92 111 .0 55 $10,001 $20,000 388 .16 163 154 231 145 $20,001 $40,000 590 .25 241 294 268 176 $40,001 $60,000 484 .21 205 182 175 264 $60,001 $90,000 270 .11 141 100 .0 58 135 $90,001 $150,000 1 84 .08 .0 83 .0 67 .0 37 109 > $150,000 95 .04 .0 37 .0 39 .0 24 .0 5 7 Education 2,509 8.49 3.72 9.5 (3.4) 8.2 (3.7) 6.7 (3.7) 9.4 (3.5) Assets 2,514 3.27 1.93 3.72 3.59 1.08 3.59 Potable Water 2,514 0.63 0.48 .71 .73 .18 .60 Indoor Bathroom 2,51 4 0.55 0.50 .66 .60 .12 .59 Worse Economic Situation 2,488 0.29 0.45 .37 .25 .25 .29 Better Economic Situation 2,488 0.24 0.43 .20 .27 .24 .27 Urban 2,514 0.31 0.46 .46 .18 .07 .39 Female 2,514 0.50 0.50 .45 .47 .55 .52 Age 2,514 38.4 14.6 38.6 (15.3) 38.5 (14.4) 38.5 (13.7) 35.5 (14.2) N 737 892 332 532
47 Figure 4 1. Economic satisfaction by ethnicity
48 Figure 4 2. Mean political trust by economic satisfaction and e thnicity
49 Figure 4 3. Income d istribution by e thnicity
50 CHAPTER 5 RESUL TS Table 5 1 reports the ordered logit regression results, with logged odds listed beside the standard errors. Model 1 shows the baseline model estimates. This model is limited to race, region, sex and age to demonstrate the magnitude of race differences in economic satisfaction. Consistent with the differences illustrated in Figure 4 1 the estimates display large race differences in economic satisfaction. The logged odds of Blacks having high levels of satisfaction are .434 that of Indians (p=.000), co ntrolling for all other variables. The logged odds of having high satisfaction for Amerindians are .583 times that for Indians (p=.000). Similarly, the logged odds of high levels of satisfaction for Mixed people are .568 times that for Indians (p=.000), controlling for all other variables. Clearly, the greatest discrepancy in economic satisfaction occurs between Indians and Blacks. Region and sex have no statistically significant effect on economic satisfaction. Older individuals are less likely to be economically satisfied than are younger individuals, controlling for all other variables (p = .000). Ethnicity, region, sex, and age explain 2.2 % of the variation in economic satisfaction. Model 2 adds education and a set of income dummy variables as controls. As education increases by one year, the expected logged odds of economic satisfaction increases by .032, controlling for the other variables (p<.01). Compared with respondents who reported no income, respondents who reported monthly household i ncomes of $40,001 $60,000, $60,001 $90,000 (p<.01), and above $150,000 displayed higher economic satisfaction (p < .05). These two SES indicators appear to exert a suppression effect on economic satisfaction for Blacks. After controlling for SES and demographic variables, the logged odds of having high economic satisfaction for
51 Blacks are .399 that of Indians (p=.000); that is, accounting for SES increases the gap in economic satisfaction between Indians and Blacks. This finding runs counter to theor ies that argue for the importance of SES in determining economic satisfaction, but lower economic satisfaction than Indians. Controlling for SES slightly reduces the differ ences in economic satisfaction between Amerindians and Indians and between the Mixed group and Indians. Adding education and income to the mo del improves the model fit by .5 % (pseudo R 2 = .027 ). Model 3 includes an asset index and the dichotomous vari ables for access to potable water indoors and ownership of an indoor bathroom. The asset index has no significant effect on economic satisfaction. Controlling for all other variables, those who have access to potable water are more likely than those who lack access to be economically satisfied (p<.01). Having an indoor bathroom has no statistically significant effect on economic satisfaction. Including these three variables in the model explains much of the difference in economic satisfaction between Am erindians and Indians (p<.05). For Amerindians, living standards appear to be an important cause of economic satisfaction. Controlling for these measures significantly reduces the effects of schooling and income on economic satisfaction. These three pro xy measures of living standa rds improve the model fit by 0.2 % (pseudo R 2 = .029 ). economic situations over the previous twelve months. These variables both have large effects on economic satisfaction. Controlling for all other variables, those whose economic situations have recently improved are almost twice as likely to have high
52 economic satisfaction as are those whose economic situations have remained stable (p=.000). Net of all other variables, people whose economic situations have recently worsened are much less likely to have high economic satisfaction than are people whose economic situations have remained unchanged (p = .000). These results confirm t he predictions of mo st theories, which have found that recent changes in economic situation are important for economic satisfaction. Controlling for economic change also reduces the effect of income and modestly reduces the difference in economic satisfaction between Blacks and Indians. Model 4 indicates that an important their current and past economic situations. The inclusion of change in economic situation improves the mo del fit by 5.6 % (pseudo R 2 = .085 ). Although this remains a modest measure of explained variance, the increase in pseudo R 2 underscores the relative improvement in explanatory power from including the measures of economic change. In Model 5, the political trust index is added. The findings indicate a positive relationship between political trust and economic satisfaction. Controlling for all other variables, a one unit increase in political trust is associated with a .022 increase in the logged odds of being economically satisfied; one stan dard deviation increase in trust (13 points in the political trust index) improves the likelihood of reporting high economic satisfaction by 31% (p=.000). The inclusion of political trust in the model significantly reduces the differences in economic sati sfaction between Blacks and Indians and between the Mixed group and Indians. P olitical trust explains about 24 % of the difference in economic satisfaction between Blacks and Indians and 22% of the difference between the Mixed group and Indians (p=.000). Thus, trust in political
53 institutions is an important cause of economic satisfaction and is able to explain a significant proportion of the differences across ethnic groups, even after controlling for demographic factors, SES indicators, and recent changes in economic situation. The addition of political trust to the mod el improves the model fit by 0.6 % (pseudo R 2 = .091 ). Model 5 is also estimated using a smaller analytic sample (N=2,014) which includes only observed values for political trust (columns 3a and 3b of Table 3). Unlike the model with the multiply imputed sample, this model reports no significant effects of any of the income categories or education on economic satisfaction. It deceptively shows that controlling for political trust eliminate s the effect of SES on economic satisfaction. Second, analysis of only observed values overestimates the effect of political trust on reducing differences in economic satisfaction between Blacks and Indians and between the Mixed group and Indians; analysi s of the multiply imputed sample indicate that this reduction is slightly more modest. The observed sample also shows that controlling for political trust eliminates the difference in economic satisfaction between Amerindians and Indians, while the multip ly imputed sample indicates that a difference between these groups persists. The differences in estimates between these two models may be attributable to different rates of missingness across ethnic and SES groups. This non random missingness is controll ed for in the multiply imputed sample. Since the descriptive statistics indicate that SES and political trust may have different effects on economic satisfaction for different ethnic groups, the full model (using the multiply imputed analytic sample) is stratified by ethnic group. The results show that political trust affects economic satisfaction for all groups, except Amerindians
54 (Table 5 3 ). The effects for Indians and Blacks are not significantly different from each other. For the Mixed group, poli tical trust is slightly less important than for Indians and Blacks. The unimportance of political trust for Amerindians may be due to their relative isolation compared with the rest of the population and the low perceived effectiveness of government polic ies in reaching the hinterland and its Amerindian population. The Mixed group, Indians, and Blacks, on the other hand, reside in the urban and rural regions, where government policies are more capable of reaching. The results also show that recent econo mic improvement is much more important for Amerindians and Mixed people than for Blacks and Indians. Economic change may be less important for Blacks and Indians because these two groups are more entrenched in ethnic and political animosity than are the A merindian and Mixed groups. The support of the majority of Indians for one political party and of the majority of Blacks for the other major party is largely attributable to the historical (and contemporary) political and ethnic polarization between these groups. Political trust is more likely to matter for groups that display a high degree of political partisanship, especially when this partisanship is rooted in ethnic conflict. As previously mentioned, Amerindians are relatively isolated from the other ethnic groups and are not part of this ethnic conflict (though, this is not to say that they are unaffected by it). The Mixed group, with many of its members being both Indian and Black, display lower levels of partisanship as a whole. The partisanship that members of this group may show is likely to be largely attributable to their identification with one ethnic group over the others.
55 Table 5 1 Predicting economic s atisfaction with SES and political t rust (N = 2,329). (1) (2) (3) (4) (5) Econom ic s atisfaction Black (Indian) 0.434*** 0.399*** 0.411*** 0.461 *** 0.553*** (0.042) (0.041) (0.042) (0.048) (0.060) Amerindian 0.583*** 0.624*** 0.738* 0.723 0.724* (0.071) (0.077) (0.103) (0.102) (0.103) Mixed 0.568*** 0.560*** 0.597*** 0 .599 *** 0.669*** (0.059) (0.061) (0.067) (0.068) (0.077) Urban 0.955 0.872 0.820* 0.818 0.880 (0.082) (0.080) (0.076) (0.077) (0.084) Female 0.983 1.005 0.999 0.991 0.993 (0.073) (0.078) (0.078) (0.078) (0.078) Age 0.982*** 0.984*** 0.984*** 0.98 7 *** 0.988*** (0.003) (0.003) (0.003) (0.003) (0.003) Years s chooling 1.032** 1.023+ 1.022 + 1.027* (0.012) (0.013) (0.013) (0.013) Income 1 10,000 (no income) 1.023 1.056 1.021 1.011 (0.209) (0.217) (0.212) (0.210) 10 20 (000s) 1.254 1.283 1.120 1.101 (0.231) (0.237) (0.210) (0.206) 20 40 1.293 1.297 1.261 1.224 (0.225) (0.226) (0.223) (0.217) 40 60 1.611** 1.582* 1.395 + 1.365+ (0.288) (0.284) (0.255) (0.249) 60 90 1.743** 1.661* 1.513 1.502* (0.342) (0.329) (0.304) (0.301) 90 150 1.325 1.243 1.117 1.115 (0.279) (0.265) (0.244) (0.243) >150 1.805* 1.737* 1.429 1.508 (0.462) (0.453) (0.382) (0.406) Asset i ndex 0.994 1.003 0.998 (0.031) (0.031) (0.031) Potable water indoors 1.299** 1.222 1.199+ (0.130) (0.124) (0.121) Indoor bathroom 1.163 1.100 1.114 (0.130) (0.124) (0.126) Econ. change w orse (Same) 0.259 *** 0.271*** (0.025) (0.027) Better 1.954 *** 1.902*** (0.197) (0.193) Political trust 1.020*** (0.004 ) Log Likelihood 2,959 2,944 2,936 2,768 2,750 R 2 pseudo 0.022 0.027 0.029 0.085 0.091 Exponentiated coefficients; Standard errors in parentheses ; Reference categories in parentheses. + p < 0.10, p < 0.05, ** p < 0.01, *** p < 0.001
56 Table 5 2 Compari ng results between multiply imputed sample and observed s ample (1a) (1b) (2a) (2b) (3a) (3b) Years s chooling 1.018 1.028* 1.015 1.022+ 1.021 1.027* (0.012) (0.011) (0.014) (0.013) (0.014) (0.013) Income 1 10,000 (none) 0.879 0.955 0.96 8 1.021 0.946 1.011 (0.201) (0.194) (0.225) (0.212) (0.220) (0.210) 10 20 (000s) 1.125 1.127 1.208 1.120 1.169 1.101 (0.231) (0.205) (0.254) (0.210) (0.245) (0.206) 20 40 1.184 1.259 1.272 1.261 1.220 1.224 (0.230) (0.217) (0.252) (0.223) (0.24 1) (0.217) 40 60 1.370 1.470* 1.392 1.395+ 1.347 1.365+ (0.271) (0.259) (0.283) (0.255) (0.273) (0.249) 60 90 1.388 1.531* 1.486+ 1.513* 1.447+ 1.502* (0.296) (0.295) (0.327) (0.304) (0.318) (0.301) 90 150 1.091 1.194 1.061 1.117 1.050 1.115 (0.249) (0.247) (0.253) (0.244) (0.251) (0.243) >150 1.665+ 1.731* 1.446 1.429 1.542 1.508 (0.449) (0.432) (0.413) (0.382) (0.443) (0.406) Black (Indian) 0.489*** 0.461*** 0.596*** 0.553*** (0.054) (0.048) (0.069) (0.060) Amerindian 0.818 0.723* 0.801 0.724* (0.126) (0.102) (0.124) (0.103) Mixed 0.652*** 0.599*** 0.739* 0.669*** (0.079) (0.068) (0.091) (0.077) Urban 0.766** 0.818* 0.826+ 0.880 (0.078) (0.077) (0.085) (0.084) Female 0.991 0.991 0.992 0.993 (0.084) (0.078) (0.084) (0.078) Age 0.988*** 0.987*** 0.988*** 0.988*** (0.003) (0.003) (0.003) (0.003) Asset i ndex 1.006 1.003 0.997 0.998 (0.034) (0.031) (0.033) (0.031) Potable water indoors 1.155 1.222* 1.126 1.199+ (0.126) (0.124) (0.124 ) (0.121) Indoor bathroom 1.167 1.100 1.186 1.114 (0.142) (0.124) (0.145) (0.126) Econ. c hange w orse (Same) 0.267*** 0.259*** 0.280*** 0.271*** (0.028) (0.025) (0.030) (0.027) Better 1.866*** 1.954*** 1.807*** 1.902*** (0.201) (0.197 ) (0.195) (0.193) Political t rust 1.022*** 1.020*** (0.004) (0.004) Log Lik elihood 2,606 3,011 2,406 2,768 2,388 2,750 R 2 pseudo 0.003 0.005 0.080 0.085 0.087 0.091 Exponentiated coefficients; Standard errors in parentheses ; Reference c ategories in parentheses. Models 1a, 2a, and 3a are based on an analytic sample of only observed values (N=2,014). Models 1b, 2b, and 3b are based on an analytic sample including multiply imputed values (N=2,329). + p < 0.10, p < 0.05, ** p < 0.01, *** p < 0.001
57 Table 5 3 Effects of political trust and SES on economic satisfaction, stratified by ethnic g roup Full Sample Amerindians Indians Blacks Mixed Economic s atisfaction Urban 0.880 1.746 0.905 0.769+ 0.907 (0.084) (0.871) (0.159) (0.119) ( 0.185) Female 0.993 0.686+ 0.991 1.099 0.943 (0.078) (0.151) (0.133) (0.163) (0.165) Age 0.988*** 0.988 0.977*** 0.994 0.992 (0.003) (0.008) (0.005) (0.005) (0.006) Years s chooling 1.027* 1.038 0.996 1.056* 1.014 (0.013) (0.034) (0.021) (0.027) (0 .030) Income 1 10,000 (no income) 1.011 0.619 1.102 0.786 1.989 (0.210) (0.316) (0.370) (0.321) (1.041) 10 20 (000s) 1.101 0.904 1.105 1.062 1.094 (0.206) (0.427) (0.338) (0.390) (0.496) 20 40 1.224 1.199 1.279 0.870 1.716 (0.217) (0.560) (0.3 57) (0.306) (0.750) 40 60 1.365+ 0.871 1.771+ 1.006 1.552 (0.249) (0.445) (0.528) (0.363) (0.653) 60 90 1.502* 1.136 2.131* 0.966 2.099+ (0.301) (0.704) (0.710) (0.372) (0.937) 90 150 1.115 0.526 1.295 0.712 1.704 (0.243) (0.357) (0.467) (0. 301) (0.809) >150 1.508 0.325 1.305 3.255* 1.993 (0.406) (0.257) (0.603) (1.686) (1.095) Asset i ndex 0.998 0.994 0.914+ 1.092 1.012 (0.031) (0.107) (0.046) (0.065) (0.071) Potable water indoors 1.199+ 0.584 1.406* 1.351+ 1.173 (0.121) (0.195) (0 .239) (0.246) (0.266) Indoor bathroom 1.114 2.662* 1.085 0.831 1.270 (0.126) (1.220) (0.189) (0.172) (0.346) Econ. c hange w orse (Same) 0.271*** 0.224*** 0.263*** 0.277*** 0.264*** (0.027) (0.061) (0.046) (0.048) (0.058) Better 1.902*** 2.312** 1.7 78*** 1.618* 2.364*** (0.193) (0.660) (0.291) (0.330) (0.525) Political trust 1.020*** 0.988 1.030*** 1.026*** 1.023** (0.004) (0.010) (0.006) (0.006) (0.008) R 2 pseudo 0.091 0.097 0.088 0.093 0.090 Observations 2329 318 832 675 486 Exponentiated c oefficients; Standard errors in parentheses Reference categories in parentheses. + p < 0.10, p < 0.05, ** p < 0.01, *** p < 0.001
58 CHAPTER 6 DISCUSSION The findings show that political trust is a significant predictor of economic satisfaction. Important ly, controlling for p olitical trust explains about 24 % of the difference in economic satisfaction between Blacks and Indians and about 22 % of the difference between Mixed people and Indians. The results indicate that Guyanese citizens take into considera tion evaluations of the performance of political institutions when they evaluate their own economic situations. Those who have more trust in the political institutions are more likely to have high levels of economic satisfaction. A possible mechanism whi ch explains this relationship may be that individuals are making comparisons between their current household income and their expected household income. In the context of occupational and residential segregation, government policies may disproportionately benefit a particular ethnic group over the others, even in the absence of the intent to discriminate. For instance, setting up solar panels in the hinterland to provide energy to Amerindian communities, which almost exclusively comprise the population of this region, may lead to higher levels of political trust for Amerindians than for other groups (United Nations Development Programme 2011). Alternatively, improving the irrigation system may benefit Indians disproportionately, as they are more likely to be farmers; in turn, Indians may have higher levels of political trust than other groups. On the other hand, failure to make improvements to areas that predominantly consist of Black residents, such as renovating roads, may be seen by these residents as indication of discrimination (Stabroek News 2011).
59 poverty policies have led to improvements These groups may perceive that their current economic situations are close to their expected economic situations, given their lower levels of education compared with wealthier groups. So, their economic satisfaction is relatively high because they perceive the conditions put in place by the political institutions to be conducive to good economic circumstances. Since, on average, Indians and Amerindians have lower SES than do Blacks and the Mixed group, there may be a smaller discrepancy between current and expected economic situation for In dians and Amerindians than for Blacks and Mixed people. The latter groups may perceive the conditions produced by the government to be unfavorable or even discriminatory. For instance, groups with higher poverty policies to be a hindrance to economic growth. They may believe that investments in the poor are to the detriment of the middle and upper classes. In this case, public policies geared toward the economic interests of the middle and upper class w ould lead to substantial improvements in the economic situation of these classes. In the absence of these policies, however, the higher SES groups may believe there to be a great difference between their current and their expected incomes. So, people wit h higher SES report lower levels of economic satisfaction than people with lower SES because the former perceive a greater discrepancy between their current and expected economic situations. As the data show, the groups with lowest levels of income are t he Indians and they seem to be placing more trust in the political institutions than are Blacks and Mixed
60 people. Of course, since the PPP is widely perceived to be a pro Indian party, the higher levels of political trust reported by Indians may be simply due to spill over effects of Indians having higher levels of trust in members of their own group. However, the spillover effect does not explain the high levels of trust reported by Amerindians, who are by far the most economically disadvantaged group. If individuals are evaluating their economic situations based on the performance of national and local political institutions, then we expect to find a positive relations hip between improvements in economic situation and economic satisfaction. Put another way, government policies that benefit a group and lead to real improvements in this am ong the members of this group. In this sense, individuals are comparing their current economic situations to their past economic situations, and may attribute the change partially to government policies. The degree to which they assign responsibility for these changes to government depends on their view of the role of government (i.e. whether government has a regulatory role to play in the economy and whether it should be responsible for redistributing resources) and on the effectiveness of government (e. g. whether government allocates tax revenues efficiently and is able to carry out its intentions through legislation and policies). With the direct relationship between political trust and economic satisfaction, it appears that Guyanese citizens are attri buting some responsibility to government for their economic situations. The theory of temporal relative deprivation finds support in the case of Guyana. There is a positive relationship between improvement in economic situation and economic satisfaction and a negative relationship between worsening of economic situation and economic satisfaction.
61 Nevertheless, political trust remains a significant predictor of economic satisfaction, even after controlling for change in economic situation. The inclusion of political trust in studies of economic satisfaction provides a partial explanation for dramatic differences in economic satisfaction among different ethnic groups in an ethnically polarized society. Ethic group differences in political trust indicate p erceptions of government discrimination and help to shape economic expectations for the future. The study of economic satisfaction proves important Assuming the validity of social contract theory (which holds that individuals enter a contract with the state, in which individuals sacrifice some liberties for protection and security from the state), the state is responsible for providing at least the necessities for all of its citizens. In contrast to life satisfaction, which may be affected heavily by strongly on macro level forces. Even with increased deregulation in recent decades, governments still play an important role in the economic process, whether through direct ownership of resources or through the regulation of private sector corporations and labor standards. Furthermore, economic growth (or perceptions thereof) has been cit ed as significant predictors of popular support for government (Avery 2009; Chanley, Rudolph, and Rahn 2000). It is, therefore, in the interests of the public and policy makers to improve economic satisfaction among the population.
62 CHAPTER 7 CONCLUSION This study depicts the importance of political trust for theories of economic satisfaction. Individuals consider a number of factors when evaluating their economic satisfaction, including SES, recent change in economic situation, and economic situation c ompared to others. In ethnically polarized societies, this study finds that political trust also matters for economic satisfaction. Given that individuals believe that government has the ability to shape their economic situations, individuals consider th e trustworthiness of the government when forming their economic expectations for the future. Their trust in government is likely to alter their expectations and, therefore, alter their comparisons between their current and expected future economic situati ons. In ethnically polarized societies where ethnic and political divisions align, one political party often draws the majority of its support from a particular ethnic group. Some of the actions that this party took in the past come to be identified as favoritism toward this ethnic group and as discrimination against other groups. In this case, voters are likely to consider the history of discrimination and patronage of political parties when making their selections for public office. The remembrance o f this history prolongs political partisanship and reinforces ethnic divisions. Guyana is used as a case study because it is an ethnically polarized society with a history of racial politics. Since the two main political parties are perceived to favor a particular ethnic group over others, political trust is expected to be an important factor in determining economic satisfaction. LAPOP data from 2009 are analyzed using 2,329 cases. Nested regression models are estimated to determine the effects of poli tical trust, SES, and change in economic situation on economic satisfaction.
63 Consistent with previous research, this study finds that socio economic status is an important determinant of economic satisfaction. Individuals with high income and high level o f schooling are likely to have higher economic satisfaction compared to those with low income and education. Standard of living is also found to be very important for Amerindians, the most economically disadvantaged group; ownership of an indoor bathroom explains much of the difference in economic satisfaction between Amerindians and Indians. In addition to SES, recent changes in economic situation are shown to be very important factors for economic satisfaction. Compared to those who have experienced no recent change in their economic situations, those who have experienced recent improvements are much more likely to be economically satisfied; on the other hand, those who have experienced recent worsening in their economic situations are much less likely to be economically satisfied. SES and economic change do little to explain the differences in economic satisfaction across ethnic groups, however. A significant proportion of these ethnic differences are explained by variation in political trust; when ac counting for political trust, the differences in economic satisfaction between Indians and Blacks and between Indians and the Mixed group are significantly reduced. After controlling for socio demographic factors and change in economic situation, politic al trust is found to have a significant direct effect on economic satisfaction. This finding is as expected. In Guyana, individuals consider the trustworthiness of political institutions as they evaluate their economic situations. Since Indians are show n to have the highest levels of political trust, this may explain why they also have higher economic satisfaction than do Blacks and Mixed people, though the latter have higher SES. The
64 results point to the insufficiency of a model which omits political t rust. In ethnically polarized societies, individuals not only consider their SES and recent changes in economic situation, but they also consider the political environment when assessing their economic situations. The belief in the ability and willingnes s of the government to for the future. This study, however, is limited to the case of Guyana a society characterized by extreme ethnic polarization. Though it is ex pected that political trust will have similar importance for economic satisfaction in other ethnically polarized societies, the results of this study should not be extrapolated to other societies. The importance of political trust is even less known for n on ethnically polarized societies. Furthermore, this study omitted a key factor in socio economic status and economic satisfaction occupation. Due to the large number of missing cases for this variable, it could not be included in the analysis. Occupa tional status (i.e. whether an individual is employed, underemployed, unemployed, etc.) is also omitted from the analysis because the response categories for this question were not mutually exclusive. There is little reason to believe, however, that the i nclusion of occupation and occupational status in the analysis would have reduced the effect of political trust on economic satisfaction. SES with local others. The mos t specific spatial units included in the LAPOP questionnaire are the ten regions by which Guyana is divided. Most of these regions include both urban and rural areas and are not homogeneous with respect to occupations or socio economic status of residents A binary variable indicating
65 residence in an urban or rural area is included in the analysis in an attempt to control for differences among individuals living in totally dissimilar residential areas, but this is admittedly a very crude control. While I propose that political trust affects economic satisfaction by altering economic expectations for the future, this mechanism cannot be verified in this study. Economic expectations are not observed and their dependence on political trust cannot be stated with certainty. Nevertheless, the findings of this study indicate a direct relationship between political trust and economic satisfaction. Individuals appear to be assessing the political environment when forming evaluations of their economic situation s. Whether or not they believe that the government is working to their benefit affects their economic outlook. Previous studies on political trust focus on the effect of economic satisfaction on political trust, but this study indicates that the relation ship is not unidirectional. By modifying expectations for the future, political trust affects economic satisfaction. While it is important for political candidates in most, if not all, societies to focus on economic growth and job creation, this study po ints to the importance of building trust with voters. Trust has both a direct and indirect effect on voter choice. Voters who have high levels of trust will be more likely to vote for a candidate based on his or her personal characteristics (e.g. honest and fair), and will also be more likely to vote for this candidate based on his or her economic policies (because trust improves economic satisfaction). It is in the interest, therefore, of governments to build trust among its citizens through policies th at benefit all groups and to expand its support to include voters who were previously alienated.
66 LIST OF REFERENCES Aberbach, Joel and Jack Walker. 1970 tical Trust and Racial Ideology. The American Political Science Review 64 ( 4 ) : 1199 1219. Alwi n, Duane. 1987 action with Material Well Being. American Sociological Review 52(1): 83 95. Anderson Cameron D Compar ative Individual Level Analysis, Economic Voting and Multilevel Governance: A Comparative Individual America n Journal of Political Science 50(2): 449 463 Political Sys Political Research Quarterly 62(1): 132 145. Bennett, Stephen. 2001. "Were the Halcyon Days Really Golden? An Analysis of Americans' Attitudes about the Political System, 1945 1965." From What is it About Government that Americans Dislike? John H ibbing and Elizabeth Theiss Morse, eds. Cambridge: Cambridge University Press. Bernstein, Thomas The China Quarterly 163: 742 763. Bird SSRN eLibrary Retrieved April 18, 2012 (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=277162). Bjrnskov, Christian, Axel Dreher, and Jus tina country Determinants of L ife S atisfaction: E xploring D ifferent D eterminants across G roups in S Social Choice and Welfare 30(1): 119 173. Blendon, Robert, John Benson, Richard Morin, Drew Altman, Mollyann Brodie, Mario Br ossard, and Matt James. 1997. "Changing Attitudes in America." From Why People Don't Trust Government ed. Joseph Nye Jr., Philip Zelikow, and David King. Cambridge, MA: Harvard University Press. Brown, Deryck. 1999. "Ethnic Politics and Public Sector Man agement in Trinidad and Guyana." Public Administration and Development 19: 367 379. Inflation and Government P opularity are there P artisan E Electoral Studies 19(2 3): 141 150.
67 Chappell, Henry W. and Willia The American Political Science Review 79(1): 10 27. Chappell lections in Western Europe: 1960 Electoral Studi es 19(2 3): 183 197. Citrin, Jack British Journal of Political Science 16(04): 431 453. Diener, Ed, Ed Sandvik, Larry Seidlitz, and Marissa Diener. 1 The relationship between Income and Subjective Well B Social Indicators Research 28(3): 195 223. Dorussen, Han and Michaell Taylor. 2002. Economic Voting London: Psychology Press. Evans, Geoffrey and Mark Pickup. 2010. "Reversi ng the Causal Arrow: The Political Conditioning of Economic Perceptions in the 2000 2004 U.S. Presidential Election Cycle." The Journal of Politics 72(4): 1236 1251. Finkel, Steve, Christopher Sabatini, and Gwendolyn S ociety, and Political Mistrust in a Developing Democracy: The Case of the World Development 28(11): 1851 1874. Elections: A Micro American Journal of Political Science 22(2): 426 443. Gafar, John. 2004. "Income Distribution, Inequality, and Poverty during Economic Reforms in Guyana." The Journal of Developing Areas 38(1): 55 77. Gamson, William. 1968. Power and Discontent Homewood: Dorsey Press. and Subjective Well Annals of the American Academy of Political and Social Science 606: 128 153. The American Political Science Review 92(4): 791 808. Hibbing, John R. American Politics Research 32(6): 652 678.
68 Hintzen, Percy and Ralph Premdas. 1982. "Guyana: Coercion and Control in Political Change." Journal of Interamerican Studies and World Affairs 24(3): 337 354. K Patte rns of Partisanship and the Economic V European Journal of Political Research 50(3): 365 394. Key, Valdimer Orlando, and Milton C. Cummings. 1966. The Responsible Electorate: Rationality in P residential V oting, 1936 1960 Cambridge: Belknap Press of H arvard University Press. Modern China 30(2): 228 258. Reagan Years, 1981 Political Science Quarterly 102(1): 1 23. issatisfa ction and P olitical Alienation in Western Europe Europea n Journal of Political Research 23(3): 281 293. Luttmer, Erzo The Quarterly Journ al of Economics 120(3): 963 1002. The American Political Science Review 86(3): 597 611. tical Issues and Trust in Government: 1964 The American Political Science Review 68(3): 951 972. Misir, Prem. 2002. "The Social Construction of Race Ethnic Conflict in Guyana." http://www.uog.edu.gy/files/documents/prochancellor/The_Social_Construc tion_o f_Race Ethnic_Confilct.pdf Powell, G. Bingham National Analysis of Economic American Journal of Political Science 37(2): 391 414. Premdas, Ralph. 1995. Ethnic Confl ict & Development: The Case of Guyana. Aldershot, Brookfield, USA, Hong Kong, Singapore and Sydney: Avebury. Rabe, Stephen. 2005. U.S. Intervention in British Guiana: a Cold War Story Chapel Hill: University of North Carolina Press.
69 Rattinger, Hans. 19 From Helmut Norpoth, Michael Lewis Beck, and Jean Dominque Lafay. Economics and Politics: The Calculus of Support University of Michigan Press. 49 63. ic Crisis in Latin America in The American Political Science Review 85(3): 777 800. Sanchez, Juan I. The A cademy of Management Journal 39(3): 704 719. Political Behavior 27(3): 285 312. Comparative Politics 33(4): 401 419. Stabroek News. 2011. Georgetown Road Already Deteriorating and the Rains ccessed on November 21 2011. http://www.stabroeknews.com/2011/opinion/letters/05/13/lethem georgetown road already deteriorating and the rains have only just begun/ United Nations Development Programme. 2011. Energy Access at Community Level for MDG Achievement in Hinterland Area. The Public Opinion Quarterly 72(4): 725 740. Waters, Robert The AFL Diplomatic History 29(2): 279 307. American Sociological Review 66(1): 1 25 145.
70 BIOGRAPHICAL SKETCH Amanda Ragnauth received her B.A. in s ociology from the University of Florida in 2006. She is currently worki ng toward her M.A. in s ociology at the University of Florida. Her areas of interest are race and ethnic studies p olitical economy, international development, and the Caribbean and Latin American regions.