The Mark of Cain

Material Information

The Mark of Cain Explaining the Adoption of Severe Felon-Collateral Consequence Policies in the United States
Christie, Natasha
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (147 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Political Science
Committee Chair:
Hedge, David M.
Committee Members:
Austin, Sharon D.
Scher, Richard K.
Martinez, Michael D.
Akers, Ronald L.
Graduation Date:


Subjects / Keywords:
African Americans ( jstor )
Criminal justice ( jstor )
Criminal punishment ( jstor )
Criminals ( jstor )
Disenfranchisement ( jstor )
Felons ( jstor )
Legislators ( jstor )
Population growth ( jstor )
Slavery ( jstor )
Voting ( jstor )
Political Science -- Dissertations, Academic -- UF
civil, class, collateral, consequences, disenfranchisement, felon, race, rights
Electronic Thesis or Dissertation
born-digital ( sobekcm )
Political Science thesis, Ph.D.


Collateral consequences of a felony conviction refer to a set of civil punishments put in place post-conviction that restrict the civil rights of felons and ex-felons. Each state varies on which collateral consequences it adopts and the severity of the consequence adopted in terms of their effect on the targeted population. More generally, this study focuses on the variation in the type and number of collateral consequences adopted by each state. However, we specifically focus on two of the most widespread felon collateral consequences: felon disenfranchisement and the felon welfare ban. I argue that the main motivation for the adoption of harsh felon collateral consequences are racially motivated and provide testable hypotheses. Further, I suggest the need for testing alternative hypotheses that include cultural, class-based and political explanations. As such, this study aimed to address possible explanations for the severity and magnitude of these policies adopted by testing race-based, cultural, political, and class-based explanations. Our findings revealed that, overwhelmingly, race-based explanations accounted for state adoption of a higher number of collateral consequences and the most severe levels of felon collateral consequences adopted with class-based explanations following close behind. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis (Ph.D.)--University of Florida, 2008.
Adviser: Hedge, David M.
Electronic Access:
Statement of Responsibility:
by Natasha Christie.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Christie, Natasha. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
LD1780 2008 ( lcc )


This item has the following downloads:

Full Text




2008 Natasha V. Christie 2


To equal justice 3


ACKNOWLEDGMENTS I thank Dr. Mary Katzenstein for planting the se ed for this project dur ing my last semester as an undergraduate at Cornell University. Her course on prisons in the United States guided my passion for this subject. I thank my current and former committee members for all of the time and effort expended to help me achieve this goal Special thanks go to my committee chair, Dr. David Hedge, for his patience and commitment to this project. I thank Dr. Ron Akers for signing on at the last minute to ensure that I had a fu ll committee to complete the requirements for this degree. I thank the late Dr. James Button for convi ncing me to pursue an advanced degree at the University of Florida and for believing in this project from the very beginning. I thank Dr. Renee Johnson for the wealth of wisdom and guida nce. I also thank her for sharing her caring nature and genuine belief in me as a scholar and as an individual. I extend a special thanks to the Eighth Judicial Circuit Court of Florida for thei r whole-hearted support during the final stages of this project. I thank the Jones family for all of their love and support throughout th e years. Thanks to my parents for never allowing me to settle for le ss than what they knew I could achieve. To my brothers, Jelani and Saeed, thanks for conti nuously keeping me motivated. To my husband Dwight, thanks for putting up with my obsession with this project and my love for the subject. Lastly, but certainly not least, Tyler thanks for going to bed on time (on most nights) and allowing Mom to work through the late hours of the night. 4


TABLE OF CONTENTS Page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 ABSTRACT.....................................................................................................................................9 CHAPTER 1 INTRODUCTION................................................................................................................. .10 Collateral Consequences: A C ontroversial Topic of Study...................................................12 Importance of This Study....................................................................................................... 15 2 BACKGROUND................................................................................................................... .23 Criminal and Political Methods of R acial SubordinationThen and Now...........................23 Then: Slavery to the Civil Rights Era..............................................................................24 Now: Post Civil Rights Era to the Present.......................................................................38 New Methods: Crime Meets Welfare.....................................................................................45 Conclusions.............................................................................................................................47 3 THEORY AND HYPOTHESES............................................................................................53 Racial Explanations................................................................................................................54 Culture........................................................................................................................ .....55 Group-Threat Theories....................................................................................................59 Alternative Explanations....................................................................................................... .63 Political Explanations......................................................................................................63 Class-Based Explanations...............................................................................................65 4 DATA AND METHODS.......................................................................................................68 Data.........................................................................................................................................68 Measure of Dependent Variables....................................................................................68 Measurement of Independent Variables..........................................................................70 Methods..................................................................................................................................77 Level of Felon Disenfranchisement.................................................................................77 Number of Collateral Consequences...............................................................................78 Felon-Welfare Ban..........................................................................................................78 5 FINDINGS AND DISCUSSION...........................................................................................85 Felon Disenfranchisement Laws.............................................................................................86 Race Models....................................................................................................................86 Culture Models................................................................................................................9 1 5


Political Model................................................................................................................ 94 Class Model.....................................................................................................................97 Full Model.......................................................................................................................97 Number of Collateral Consequences......................................................................................98 Race Models....................................................................................................................98 Culture Model................................................................................................................100 Political Model..............................................................................................................10 1 Class Model...................................................................................................................103 Full Model.....................................................................................................................104 Severity of the Felon Welfare Ban.......................................................................................105 Race Models..................................................................................................................105 Culture Model................................................................................................................109 Political Model..............................................................................................................11 3 Class Model...................................................................................................................114 Full Model.....................................................................................................................115 Change in Level of Felon Welfare Ban................................................................................115 Race Model....................................................................................................................116 Culture Model................................................................................................................117 Political Model..............................................................................................................11 7 Class-based Model.........................................................................................................118 6 CONCLUSION: CAN WE EXPLAIN STATE FELON-COLATERAL CONSEQUENCE POLICY ADOPTION?..........................................................................127 Why Should We Care?.........................................................................................................127 What Do We Know?.............................................................................................................12 9 Implications: How Do Our Findings Improve the Study of Race, Class, and Criminal Justice Policies?................................................................................................................131 Future Research: Ways to Improve This Study....................................................................132 APPENDIX A BIVARIATE RELATIONSHIPS.........................................................................................134 LIST OF REFERENCES.............................................................................................................136 BIOGRAPHICAL SKETCH.......................................................................................................147 6


LIST OF TABLES Tables Page 1-1 Collateral consequences as reported by Burton, Cullen, and Travis (1986 data)..............17 1-1 Continued.................................................................................................................. .........18 1-2 Collateral consequences as reported by Olivares and Burton (1996 data)........................19 1-2 Continued.................................................................................................................. .........20 1-3 Collateral consequences as reported by Buckler and Travis (2001 data)..........................21 1-3 Continued.................................................................................................................. .........22 2-1 Level of felon disenfra nchisement by state (2006)............................................................50 2-1 Continued.................................................................................................................. .........51 2-2 Level of felon welfare ban by year 1996-2006..................................................................52 4-1 Level of severity of felon disenf ranchisement laws (2006) by State.................................80 4-2 Number of collateral conseque nces (1986, 1996, and 2006) by state...............................81 4-2 Continued.................................................................................................................. .........82 4-3 Level of felon welfare ban 1997, 2006, and the difference in levels from 1997-2006......83 4-3 Continued.................................................................................................................. .........84 5-1 Severity of felon disenfranchiseme nt laws (2006) ordered logit models........................119 5-1 Continued.................................................................................................................. .......120 5-2 Poisson regression model for the number of collateral cons equences 1986, 1996, and 2006..................................................................................................................................121 5-2 Continued.................................................................................................................. .......122 5-3 Level of felon welfare ban 1997: Ordered logit models for race -based and cultural variables...........................................................................................................................123 5-4 Level of felon welfare ban 1997: Ordered logit model for political and class-based variables...........................................................................................................................124 5-5 Change in the level of felon we lfare ban (2006) ordered logit models...........................125 7


5-5 Continued.................................................................................................................. .......126 A-1 Sample of bivariate relationships: Seve rity of Felon Disenfranchisement model...........134 8


Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE MARK OF CAIN: EXPLAINING THE ADOPTION OF SEVERE FELONCOLLATERAL CONSEQUENCE POLICI ES IN THE UNITED STATES By Natasha V. Christie August 2008 Chair: David Hedge Major: Political Science Collateral consequences of a felony conviction refer to a set of civil punishments put in place post-conviction that restrict the civil rights of felons and ex-felons. Each state varies on which collateral consequences it adopts and the severity of the consequence adopted in terms of their effect on the targeted population. More generally, this study focuses on the variation in the type and number of collateral consequences adopted by each state. However, we specifically focus on two of the most widespread felon collateral consequences: felon disenfranchisement and the felon welfare ban. I argue that the main motivation for the adopti on of harsh felon collateral consequences are racially motivated and provide testable hypothese s. Further, I suggest the need for testing alternative hypotheses that include cultural, class-based and political explanations. As such, this study aimed to address possible explanations for the severity and magnitude of these policies adopted by testing race-based, cultural, political, and classbased explanations. Our findings revealed that, overwhelmingly, race-based explanations accounted for state adoption of a higher number of collateral consequences and the most severe levels of felon collateral consequences adopted with class-based explanations following close behind. 9


CHAPTER 1 INTRODUCTION After God sent Adam and Eve out of the Garden of Eden, they joined together and Eve gave birth to Cain, who became a worker of the soil, then to Abel, who grew up to be a shepherd of the flocks. In time, Cain made an offering to the Lord of fruit from the ground. And Abel brought the firstborn of his flock with its fullness and fat. The Lord respected Abel's gift, but had no regard for Cain's. And Cain was angry and his face fell. And the Lord said, "Why are you so upset? If you do well you will be a ccepted, but if you don't do right, sin is waiting for you by your door. And sin will want you. But you can conquer it." One day, when the brothers were in the field, Cain rose up against Abel and killed him. The Lord said to Cain, "Where is your brother, Abel?" And Cain replied, "I do not know. Am I my brother's keeper?" And God said, "What have you done? The voice of your brother's blood cries out to me from the ground. Now you are cursed! You are cursed from the earth which opened her mouth to receive your brother's blood from your hand! When you work the soil, you'll get next to nothing. You will be a fugitive and a wanderer on the earth." Cain said to God, "My punishment is greater than I can bear! I am banished from the face of the earth and from the face of God. I will wander without purpose and whoever finds me will try to kill me." But God said to Cain, "I promise! Anyone who kills Cain will suffer vengeance seven times over." And the Lord set a mark on Cain so that whoever came upon him would not strike him down. And Cain went away from the presence of the Lord, dwelling, in the land of Nod, wandering east of Eden (Narrative written by Elizabeth Swados in conjunction with Bill Moyers Genesis on PBS, 10/16/1996,, Access date: 9/30/06. Adapted from the Bible: The book of Genesis 4:1-16) Felons in todays society bear a similar mark to that of Cain, which follows them through every facet of their lives, from their everyday ac tivities (such as obtaini ng employment) to their more civic duties (such as voting). Although they have completed their se ntences, they are still treated as property of the state, in some instances forever bearing the mark of Cain. According to Webb Hubbell, former Associate Attorney General of the United States and himself an ex-felon, I soon learned, as have millions of other Americans, that I carry a mark that keeps me behind bars, even on the outside. In the prison reform movement, it's called the "mark of Cain," but contrary to the biblical injunction, God's mercy isn't attached. Rather, it shackles former offenders like me with restrictions barring us -often permanently -from the means to live a normal life. (Hubbel 2001) 10


Collateral consequence describe s the restriction of the civil rights of felons and ex-felons. More specifically, they refer to rights or privileges lost upon convi ction that serve as punishments in addition to conviction and sent ence imposed by the court (Burton, Cullen, and Travis 1987, 52). Moreover, thes e restrictions last during the incarceration peri od but are more often prolonged well past the peri od of incarceration. Some states have gone so far as to keep these felons civil rights restrictions permanent. According to Travis (2002), since ancient times, governments have denied felons certain rights and privileges of citizenship. In Ancient Rome, the government imposed the penalty of outlawry on offenders. Ancient Athens placed the penalty of infamy on the offender. The nomenclature may differ but the idea was identical from society to society: to deny certain rights and benefits of citizenship to those who had broken the law. Medieval Europeans usually imposed civil de ath on offenders serving a life term in prison, whereby they lost the right to inherit or bequeath property, enter into contracts, and vote (Travis 2002, 17). This tradition found its way to the United States, which used civil death to prohibit felons (usually those sentenced to life imprisonment) from en joying all the rights and privileges granted to free citizens. Civil death evolved into what we now refer to as collateral consequences. Gradually, the scope of civil death expanded. The mutation began when states applied civil death procedures to felons who were not serving life sentences, th ereby extending the reach of this type of punishment. Todays collateral consequences reach a far grea ter number of individuals under various levels of criminal-jus tice supervision, even extending to those who are no longer under the authority of the criminal-justice system, than we have seen in the history of our American society. 11


Collateral Consequences: A Controversial Topic of Study Various researchers have examined a range of collateral sanctions associated with a criminal conviction, in addition to how both the state and federal governments apply these laws. To date, the most notable types of research have examined how governments apply these collateral consequences of a cr iminal conviction over time, and more so, specifically focus on whether these governments have used these conseque nces in a moreor le ss-restrictive manner. Further, they attempt to provide possible explanations for why st ates have become more or less restrictive in their use of collateral consequences. Wheelock (2005) attempts to classify collateral consequences into different categories that reflect the specific injuries they impose on felons and ex-felons. He classifies collateral consequences into four distinct categories: civic, service and aid, employment/occupational and other restrictions. Civic restric tions include the loss of voting right s, the right to serve on a jury, and the right to run for public office. Service and aid restrictions p rohibit ex-felons from receiving scholarships and grants, welfare, public housing, military benefits, and any other form of public assistance (Wheelock 2005, 85). Employment and occupational restrictions consist of any restrictions that ban or disqualify ex-felons from holding certain governmental positions and occupational licenses. Other restrictions can consis t of the loss of parental rights and the use of a felony conviction as grounds for divorce. De pending on how collateral consequences are classified a state can have a nywhere between one and 140 collat eral consequences of a felony conviction 1 thus documenting the vast amount of va riation that exists between states. Burton, Cullen, and Travis (1987) focus on the variation between what types of collateral consequences governments choose to adoptless or more severe. They study states use of nine 1 The Ohio Collateral Consequences project documents 140 collateral consequences of a criminal conviction in the state of Ohio. Access date: January 5, 2006. 12


particular collateral c onsequences: those on voting, parenting, divorce, public employment, jury duty, holding public office, firearm ownership, crim inal registration, and ci vil death, to assess the current legal restrictions placed on the civil ri ghts of persons convicted of a felony for all 50 states and the District of Columbia Their findings suggest that the rights to serve as a juror, own a firearm, and remain married after a conviction were the most restricted of the nine rights studied, meaning more states adopted these collate ral consequences. On the other hand, the least restrictive rights, accordi ng to the 1986 legal codes, were th e rights to obtain public employment, register as a criminal, and the imposition of civil death. Overall, Burton, Cullen, and Travis (1987) conclude that, based on their analysis of the 1986 legal codes, that a smaller number of states were adopting severe collateral consequences. However, in 1996 Olivares and Burton decided to update the resear ch of Burton, Cullen, and Travis (1987) to determine changes in the level of severity/restr ictiveness of collateralconsequence policy adoption between 1986 and 1996. Their findings revealed that states had become more restrictive betw een 1986 and 1996, with six of the nine rightsvoting, holding public office, parenting, divorce, firearm ownership, and crimin al registration,experiencing increases in restrictiveness. Seven years later, Buckler and Tr avis (2003) updated Olivares and Burtons (1996) research using 2001 legal codes 2 and found another increase in the restrictiveness of parental rights, firearm ownership, and criminal registration (See Table 1-3). Overall, these studies confirmed the variation in the number of colla teral consequences a state adopts as well as the type of collateral consequences adop tedmore severe vs. less severe. Tables 1-1, 1-2, and 1-3 consolid ate the findings of Burton, Culle n, and Travis (1987); Olivares and Burton (1996); and Buckler and Travis (2003) to further explore th e variation in state2 Buckler and Travis (2003) use eight of the original nine civil death instruments (with the exception of public employment), and an additional right, the right to receive welfare benefits. 13


collateral consequence adoption over time. The si gnificant change is the number of collateral consequences states have adopted ove r this nearly two-decade span. Between 1987 and 2001, the number of states that have adopted criminal registration as a collateral consequence has drastically increase d. In 1987, only eight states (Alabama, Arizona, California, Florida, Mississippi, Nebraska, Tennessee, and Utah) re quired criminal registration. By 2001, all states and the District of Columbia required criminal registration. However, it is worth noting that Buckler and Travis (2003) indi cate that the dramatic increase in criminal registration is a response to fede ral requirements of sex-offender registration. In the same way, the number of states that utilized the collateral consequence which re stricted parental rights also increased significantly. In 1986, Bu rton, Cullen, and Travis (1987) fi nd that in 16 states a court could terminate an offenders parental rights upo n conviction or imprisonment. That number increases slightly to 19 states in 1996, but up to an astounding 48 states in 2001. However, no valid explanation has yet emerged for the increas e in the number of states that terminate an offenders parental rights upon conviction. Additionally, Table 1-1 (1986 data ) shows the number of states that impose civil death to be relatively low (four out of 50 states). It re mains at this number in 1996, then decreases by half (to two) when the 2001 le gal codes are analyzed. Al so, the remaining collateral consequences remain relatively stable over this fifteen-year period. Not only is the analysis of the changes in the specific collateral consequences important, but it is also important to look at some trends at the state level. At the state level, in 1986, North Carolin a had the lowest number of collateral consequenceszero, followed by Vermont with one Mississippi had th e highest, with eight collateral consequences, with Alabama and Rhod e Island close behind with seven. Table 1-2 14


indicates that the analys is of the 1996 legal codes reveal Mi ssissippi and Alabama maintaining a higher number of collateral consequences; whereas, Rhode Island actually decreases its number. North Carolina, on the other hand, increases its nu mber of collateral consequences to one while Vermont continues to maintain its sole collateral consequence. Trends only tell us part of th e story, however. To get a more in depth understanding of the phenomena we need to research explanations for these trends at the st ate level. This study proposes to give us greater in sight into state implementation of collateral consequences by testing explanations that can account for the adoption of severe collateral consequences. To take a deeper look at this question, I ha ve chosen to focus on two specific collateral consequences: the felon-welfare ban and felon-di senfranchisement laws. I have chosen these areas particularly because they are two of four collate ral consequences that produce a substantial effect on felons, according to Wheelock (2005). Furthermore, each corresponds to a different classification of collateral conseq uence: voting affects the political process, whereas welfare falls under the service and aid category. Importance of This Study Why is this study important? The larger goal of this study is to create an intersection of political science and criminology that will im prove the discourse between these fields. Furthermore, it will serve as example of why academics should pursue interdisciplinary studies. Overall, this work improves past research by updating the important studies of collateralconsequence policies done by Olivares and Burton (1987), Burton, Cullen, and Travis (1996) and Travis (2003). It does so by ex tending the reach of their work to include the year 2006, thereby showing the change in the adoption of these collateral consequences over the past twenty years. Lastly, it involves an in-depth analysis of speci fic collateral consequences and states reasons for adoption in both the area of fel on disenfranchisement and the fel on-welfare ban, an area that has 15


not been well developed in prior studies. However, most importantly, th is study allows us to investigate the role of state demographics in policy decision making and whether the effect of these policies potentially violate the 14th Amendment of the Constitution. 16


Table 1-1. Collateral consequences as reported by Burton, Cullen, and Travis (1986 data) State Voting Parental Divorce Public employment Jury Public office Firearm Registration Civil death Total AL x x x x x x x 7 AK x x x 3 AZ x x x 3 AR x x x x x 5 CA x x x x x 5 CO x 1 CT x x 2 DE x x x x 4 DC x x x x 4 FL x x x x x 5 GA x x x x 4 HI x x 2 ID x x x 3 IL x x 2 IN x x x x 4 IA x x x x x 5 KS x x 2 KY x x x x 4 LA x 1 ME x x 2 MD x x 2 MA x x x 3 MI x x 2 MN x 1 MS x x x x x x x x 8 MO x 1 MT x x 2 NE x x 2 NV x x x x x x 6 NH x x 2 NJ x x x 3 NM x x x x 4 NY x x x x x 5 NC 0 ND x 1 OH x x x 3 OK x x x 3 OR x x 2 PA x x x 3 RI x x x x x x x 7 SC x x x 3 SD x x 2 TN x x x x x x 6 TX x x x x 4 UT x x x 3 VT x 1 17


Table 1-1. Continued State Voting Parental Divorce Public employment Jury Public office Firearm Registration Civil death Total VA x x x x 4 WA 0 WV x x 2 WI x x x 3 WY x x 2 Total 11 16 28 6 31 23 31 8 4 Source: Burton, Velmer S. Jr., Francis T. Cullen, and Lawrence F. Travis III. 1987. The Collateral Consequences of a Felony Conviction: A National St udy of State Statutes. Federal Probation 51:5260. 18


Table 1-2. Collateral consequences as reported by Olivares and Burton (1996 data) State Voting Parental Divorce Public employment Jury Public office Firearm Registration Civil death Total AL x x x x x x x 7 AK x x x 3 AZ x x x 3 AR x x x x x x 6 CA x x x x x 5 CO x x 2 CT x x x 3 DE x x x x x x 6 DC x x x x 4 FL x x x x x 5 GA x x x x x 5 HI x x x 3 ID x x x x x 5 IL x x x 3 IN x x x x x 5 IA x x x x x x 6 KS x x x 3 KY x x x x x 5 LA x x x 3 ME x x x 3 MD x x x x 4 MA x x x 3 MI x x 2 MN x x 2 MS x x x x x x x x 8 MO x x 2 MT x x x 3 NE x x x 3 NV x x x x x x 6 NH x x x 3 NJ x x x x 4 NM x x x x x 5 NY x x x x x x 6 NC x 1 ND x x 2 OH x x x x 4 OK x x x 3 OR x x x 3 PA x x x 3 RI x x x x x x 6 SC x x x x 4 SD x x x x 4 TN x x x x x x 6 TX x x x x x 5 19


Table 1-2. Continued State Voting Parental Divorce Public employment Jury Public office Firearm Registration Civil death Total Source: Olivares, Kathleen M. and Velmer S. Burton. 1996. The Collateral Consequences of a Felony Conviction: A National Study of State Legal Codes 10 Years Later. Federal Probation 60:10-17. UT x x x x 4 VT x 1 VA x x x x x x 6 WA x x x 3 WV x x x x x 5 WI x x x x x 5 WY x x x x 4 Total 14 19 28 6 29 25 33 47 4 20


Table 1-3. Collateral consequences as reported by Buckler and Travis (2001 data) State Voting Parental Divorce Public employment Jury Public office Firearm Registration Civil death Total AL x x x x x x x 7 AK x x x 3 AZ x x x x x 5 AR x x x x x x x 7 CA x x x x 4 CO x x x x 4 CT x x 2 DE x x x x x x x 7 DC x x 2 FL x x x x x x 6 GA x x x x x x 6 HI x x x 3 ID x x x x 4 IL x x x x 4 IN x x x x x 5 IA x x x x x 5 KS x x 2 KY x x x x x x 6 LA x x x x x 5 ME x x 2 MD x x x x x 5 MA x x x 3 MI x x x 3 MN x x x 3 MS x x x x x x x 7 MO x x 2 MT x x x 3 NE x x x x x x 6 NV x x x x x x 6 NH x x x x x 5 NJ x x x x x 5 NM x x x x x 5 NY x x x x 4 NC x x x x x x 6 ND x x x 3 OH x x x x x 5 OK x x x x x 5 OR x x x 3 PA x x x x x 5 RI x x x x x 5 SC x x x x x 5 SD x x x 3 TN x x x x x 5 TX x x x x x 5 UT x x x x x 5 VT x x x x 4 21


Table 1-3. Continued State Voting Parental Divorce Public employment Jury Public office Firearm Registration Civil death Total VA x x x x x x 6 WA x x x x x x 6 WV x x x x 4 WI x x x x x 5 WY x x x 3 Total 16 48 29 6 30 21 26 51 2 Source: This table was adapted from Buckler, Kevin G. and Lawrence F. Travis III. 2003. Reanalyzing the Prevalence and Social Context of Collateral Consequence Statutes. Journal of Criminal Justice 31:435-453. However, I have changed Buckler and Travis coding to resemble that of previous research by using the coding scheme of Burton, Cullen, and Travis (1987) and Olivares and Burton (1996). 22


CHAPTER 2 BACKGROUND Criminal and Political Methods of Racial SubordinationThen and Now these punishments have become instruments of social exclusion; they create a permanent diminution in social status of c onvicted offenders, a distancing between us and them (Travis 2002). Social Darwinism and Eugenics theories ar e often used when talking about us vs. them. These theories were the foundations by which us vs. them laws and policies were created. Social Darwinists are most popularly linked to the phrase the survival of the fittest, meaning that biology plays a very important part in how we define us and them. According to Darwin, Though nature grants vast periods of time for the work of natural se lection, she does not grant an indefinite period; for as all organic beings are strivi ng, it may be said, to seize on each place in the economy of nature, if any one species does not become modified and improved in a corresponding degree with its competitors, it will soon be exterminated (Darwin 1859, 84). In this theory, the us survive and the them become extinct. Darwinistic thought led the way for Eugenics theory, which relied on the idea that the fittest do what is necessary to rid themselves of the burdens of the others. Darwin, specifically, promoted the idea of nature being the force behind natural selection, but he also stressed that nature takes its time. According to Darwin, That natural selection will act with extreme slowness, I admit.But the action of natural selection will probably still oftener depend on some of the inhabitants becoming slowly modified; the mutual relations of many of the other inhabitants being thus disturbed. Nothing can be effected unless favourable variations occur, and variation itself, is apparently always a very slow process. (Darwin 1859, 89-90) Eugenicists could not rely on the slow processes of natura l selection. Instead, they believed in taking matters in their own hands by forc ing natural selection. Eugenics strove to 23


maintain the quality of the human race by cont rolling hereditary factors (Garver and Garver 1991). Therefore, race, poverty, and crime were im portant to eugenicists because they saw these as arising from defects in heredit y. Both Social Darwinists and E ugenicists used their theories to justify slavery and further racial subordination during the Jim Crow era in the United States. I argue that racial subordination is the basis of the adoption of felon collateral-consequence policies. This chapter discusses how legislators and pol iticians use criminal-justice policies and exclusionary politics to create institutional raci al subordination to further distance us from them. More specifically, I argu e that the institution of racial subordination that existed in the United States because of slavery and its aftermat h has never died, but merely changed form over time. I begin with a brief overview of the then, that is, the criminal and political methods used during slavery, the Jim Crow Era, and the Civil Rights Era. Then I shift to an overview of these methods as they pertain to the now the 1970s to the present. Th roughout, I continually trace the subtle transformations in institutional racial subordination through time. Then: Slavery to the Civil Rights Era Criminal: Slave law and Jim Crow justice Slave owners used slave laws to restrict the freedoms enjoyed by slaves in the newly established colonies of North America. Jame stown, Virginia has one of the most extensive volumes of Slave Law and makes a good example from which to draw. Ther efore, I rely heavily on its laws to depict the range of laws used to subordinate the African sl ave and keep him in his place. Slave codes were used to restrict the rights of slaves to ensure that they were denied full citizenship. Because Virginia was one of the first colonies to receive slaves in 1619, it had one of the most comprehensive sets of slave laws, rangi ng from distinguishing Native American slaves 24


from African slaves, deciding the monetary value of a particular type of slave (i.e. women and children), and assessing taxes for slaves, to legalizing the punishment and killing of slaves. Other laws suppressed rebellious activities by slaves, establis hed the procedure for bringing a slave to trial for a capital offense, outlined the punishment of slaves for hog stealing, and declared Negro, Mulatto and Indian slaves to be real estate. Jamestown lawmakers used these types of laws to cement the slaves status as property of their resp ective owners. In 1705, Jamestown statutes explicitly declared that enslaved men, women, and children were to be regarded as real property that became part of the esta te of their masters and would be passed on to his heirs and wi dow upon their deaths1. Not only did these laws determine the official status of slaves and their offspring, th ey also dictated how slaves sh ould behave and how their masters and other Whites in authority were to treat them. Between 1629 and 1750, Jamestown law particularly stressed the position of the African relative to any White in the colony. In 1680, Jamestown laws provided specific punishment for any slave who resisted a White individual. Lawmakers placed particular emphasis on three crimes: leaving a plantation without the masters permission, raising a ha nd against a Christian, and resisting capture after running away. Politic al leaders at this tim e were concerned about slaves running away and starting sl ave colonies or even igniting a slave rebellion. To this end, Jamestown passed a series of laws prohibiting slaves from leaving the territory without a license, defined the punishments for slaves caught runnin g away, and detailed the rewards given to anyone who apprehended a runaway. Jamestown was not the only community that wo rried about slaves running away. Other southern states also began to worry about the threat of runaway slaves as the United States 1 The Jamestown, Virgina, Slave Laws were adapted from 25


population blossomed and its territory expanding westward. The populati on increase affected southern states and their status as slaveholding states in two separate ways. First, southern states were engaged with northern states over the num ber of lower house repres entatives in each state (McGeehan and Gall 2007). Southern states, particularly slave holding states, wanted slaves to be counted for the purpose of representation but not for taxation. On the other hand, Northern states, particularly non-slave holdi ng Northern states, preferred to count only the free population. Nevertheless, members of the Constitutional Convention came to what is known as the ThreeFifths Compromise which provide d for three-fifths of the slave population in each state to be counted for the purposes of both representati on and taxation. Naturally, these decisions disappointed the southern states. In order to keep the southern states happy after the passing of the three-fifths compromise and the banning of slavery in the Northwest Territory, the convention decided to allow the importation of slaves until 1808 and to support the return of fugitive slaves as prescribed in the Fugitive Slave Clause of the Three-Fifths Compromise. Other slave laws governed the punishment for crimes committed by slaves as well as how to carry it out. As an illustration, lawmakers enacted specific laws to dictate the punishment of slaves who stole hogs, as Virginia Law did not pe rmit slaves to own horses, cattle, and hogs. If a slave stole a hog, he or she wa s subject to the following2: That for the first offence of hog stealing commited by a negro or slave he shall be carried before a justice of the peace of the county where the fact was commited before whome being convicted of the said offence by one evidence or by his owne confession he shall by order of the said justice receive on his bare back thirty nine lashes well laid on, and for the second offence such negro or slave upon conviction before a court of record shall stand two hours in the pillory and have both his eares nailed thereto and at the expiration of the said two hours have his ears cutt off close by the nailes, any thing in the aforesaid act or in any other law to the contrary in any wise notwithstanding. 2 Information of the Slave Laws of Jamestown, Virginia were obtained from ,Access date 7/12/07. 26


Not only did the law severely prosecute slav es for stealing hogs, but the decision by the Jamestown colony to prohibit the ownership of cattl e, horses, and hogs by slaves helped diminish the status of the slave within the colony. Further, as of the day the colony passed these laws, slaves were required to give th eir masters possession of all live stock that had previously been theirs. Here again we see an attempt by those in power to take away any independence given to slaves. Because slaves could no longer own livestock, they could not generate a profit from their sale nor rely on them for food. Worse, the slave masters automati cally profited by having livestock turned over to them, while no one compensated the slaves for their loss of incomegenerating property. If a slave master went too far in punishing a slave and ultimately killed him, the master could escape prosecution according to Jamestown law: And if any slave resist his master, or owner, or other person, by his or her order, correcting such slave, and shall happen to be killed in such correction, it shall not be accounted felony; but the master, owner, and every such other person so giving correction, shall be free and acquit of all punishment and accusation for the same, as if such incident had never happened ( 1 ). The status of the African in the new world was obvious, but even more obvious was the threat a free African represented to the settlers. The fact that t hose in power felt it necessary to enact these insurrection laws against a population of freed Africans (a population that was small and virtually powerless) underscore s the extent to which White Am ericans viewed the African as a potential threat. That is why the government instituted laws specifically against any Negro plotting against his master or attemp ting to incite an insurrection. Eventually, the status of the African could no longer remain justifiably subordinate and slavery in the United States met its demise. Inst ead of creating an equal society (which no one at the time except the most idealistic or nave belie ved would happen), it merely necessitated that new forms of racial subordination had to be developed to ensure th at the African could not rise to 27


the status of his former masters. Eventually lawmakers transformed Slave Laws into Black Codes. Shortly after the Civil War, former slaves began enjoying their freedom by partaking in activities previously forbidden to them, such as buying property and creating their own churches and schools. Many former slave owners beca me nervous about their own position in this transitioning society and tried to limit the economic and physical freedom of the formerly enslaved by adopting laws known as Black Code s (Davis 2004, 1). Lawmakers passed Black Codes to severely limit the civil rights and civil liberties of newly freed slaves after the Civil War. However, the Black Codes were short-live d (1865-1866). Some argue that many states did not have an opportunity to institute these Bl ack codes because by 1866, the federal government declared Black Codes illegal (McGeehan and Ga ll 2001). However, lawmakers found a way to bring them back; Jim Crow laws arrive d as the reincarnation of Black Codes. Jim Crow laws were instrumental in creating distinctions between races in private as well as public places. During this period, many cities began erecting separate facilities for colored and White use, segregating public transportati on, passing anti-miscegenation laws, and inviting lynching and the Ku Klux Klan to become the African-Americans newest nightmare. Lawmakers initiated all of these e fforts to ensure that the African knew his place and to impart the message that despite the freedoms given to h im by law, he was still inferior to the White American. One of the most common forms of segregati on seen throughout the South concerned the mixing of races on public transportation. As early as 1896, in Plessy v Ferguson, individuals began challenging the constitutionality of Jim Crow Laws when a man of African descent, Mr. Plessy, was asked by a trolley car operator to s it in the colored section of a trolley car in 28


Louisiana. Although Mr. Plessy was only one-eighth Black, he was still required to sit in the colored section of the railcar. When he refu sed, the police arrested and charged him with violating Louisianas segregation laws. His case made its way through the Louisiana legal channels with no success. Eventually, the United St ates Supreme Court decided to hear his case. The Supreme Court ruled in favor of Louisiana s segregation law. They simultaneously upheld the constitutionality of racial segregation in public places, citing that it did not find any differences between the White section and the Black section of the rail car. This landmark decision affirmed the federal governments positi on in support of this method of institutional racial subordination. Most impor tantly, according to the Supreme Court, Louisianas segregation law did not violate the separate but equal doctrine, which required that facilities could be separate as long as accommodations were equal. One area where separate but equa l failed appallingly was in pub lic education. In Florida, the laws specifically stated that educators were required to have separate teaching facilities for White children and Negro children (Jim Crow LawsFlorida). Many other states throughout the South also adopted such laws. However, al though the facilities were separate they were not by any means equal. Negro schools were always inferior to White schools in terms of the quality of education and, in most cases, the conditions of the schools fa cilities. Many Negro schools did not have the proper materials with which to teach nor an adequate structure to provide lessons to children. Th ese schools were often dilapidated and in need of upkeep. Many schools in the northern part of the South went without heat in th e winter while furnaces warmed their White counterparts. Civil rights lawyers successfully challenged these laws in the 1954 Brown vs. Board of Education of Topeka, Kansas cases. The decision in Brown v Board of Education struck down 29


racial segregation in schools, thereby creating another landmark S upreme Court case. Parents in Topeka, Kansas filed a class-acti on lawsuit against the Topeka school district, citing that racially segregated schools were inherently separate and unequal. Similar to the case of Mr. Plessy, these parents were unsuccessful with the Kansas legal system and decided to bring their suit to the Supreme Court. On May 17, 1954, the United Stat es Supreme Court overturned their previous ruling in Plessy v Ferguson by declaring that creating separa te public schools for Black and White students was inherently unequal (McGee han and Hall 2007). However, that did not mean states automatically desegregated their sc hools; some states remained segregated through the 1970s, at which time federal intervention was required to ensure that desegregation in schools took place. Not only did Jim Crow Laws call for separate f acilities: they also outlawed the intermixing of races in any form, especially the forms of cohabitation, sexual relations and marriage. In Florida, some Jim Crow Laws specifically ba rred marriages between a White person and a Negro or anyone of Negro descent. In some st ates, they went so far as to institute fines or imprisonment for anyone who engaged in sexua l relations, became marri ed or lived with a Negro or anyone of Negro descent. Howeve r, the 1967 United States Supreme Court case of Loving v. Virginia struck down Virginias anti-miscegenation laws as unconstitutional (Loving v. Virginia 388 US 1 (1967)). The findings of this court declared that Virginia had no legitimate overriding purpose independent of i nvidious racial discrimi nation which justifies this classification and such laws violated th e Equal Protection Clause of the Constitution. As with most of these types of laws, states did not automatically repeal their state anti30


miscegenation laws. As recent as 2000, the state of Alabama became the last state in the country to repeal its anti-miscegenation laws 3. The cumulative effect of rendering social beha viors illegal was the social cementing of a two-tiered society, in which White people had more rights and a higher status than Blacks. This effect was augmented by discriminatory enforcement practices. One could see evidence of this throughout this period, when the state attorneys offices or district attorneys offices would rarely prosecute Whites who perpetrated violence agains t Blacks, up to and including lynching. In essence, we see from this brief summary that laws have been used since Americas founding to secure the subordinate status of Blacks within its borders. However, one cannot confine the imposition of institutional racial subordination to the legal system but must also examine how this phenomenon affects our political system. Political: Black disenfranchisement and diluting the Black vote Every device of which human ingenuity is capable has been used to deny the Black citizen his right to vote (President Johnson, March 1965 quoted in Zelden 2002, 29.) From slavery through the Civil Rights Era, various methods of institutional racial subordination prohibited Blacks from participating in the American political system. Prior to slaves gaining their freedom, their enslaved st atus excluded them from attaining the rights of citizenship and consequently thei r dependent status (dependent on their masters) precluded their electoral participation. According to Zelden (2002), establis hed wisdom of the time held that dependency on others for the necessities of life ro bbed individuals of the independence needed to exercise independent decisions su ch as voting. However, the outcome of the Civil War and the passage of the 13th Amendment gave slaves their freedom. Nevertheless, that did not mean that they could actually exercise all the rights as a freed person that other Americans took for granted. 3 article posted the day voters in Alabama voted to ban anti-miscegenation laws in their state. S/stories/11/07/alabama.interracial/ 31


Southern lawmakers instituted a system of Black Codes used to prevent newly freed slaves from enjoying their freedom. These Bl ack Codes served as a countermeasure to the passing of the Civil Rights Act of 1866 and the passage of the 13th Amendment. The Civil Rights Act of 1866 gave Blacks the right to test ify in court against Whites, sit on a jury, participate in lawsuits, and to enter into legal contracts. Although the 13th Amendment had abolished slavery in the United States, to ensure that Blacks would not have any voting rights, lawmakers wrote these codes in such a way as to establish the continued dependency of Blacks on their former masters. In many states, one common requirement was for freed Blacks to engage in some form of employment. This empl oyment, however, usually tied them to a virtual master. These codes discouraged self-sufficien cy for newly freed Blacks, thereby enhancing a form of dependency similar to what had existed between slave and ma ster prior to the Civil War. This dependency, which informally had the effect of denying suffrage prior to the Civil War, was now codified into a formal law. Despite their newly gained freedom, Blacks were still unable to participate fully in the political process. It was not until Congress passed the Fourteenth and Fifteenth Amendments were Black males allowed to vote. The Four teenth Amendment gave African Americans who were born in the United States citizenship and with that, all privileges and immunities that come with it. The Fifteenth Amendment further establ ished the rights of citizenship by stating that, the right of citizens of the United States to vote shall not be denied or abridged by the United States or by any state on account of race, color, or previous condition of servitude(United States Constitution, Fifteenth Amendment 1870). The pa ssage of these two amendments wrote the position and privilege of Black citizens into the laws of the United States. However, that did not mean that these laws were not broken, especially in the South. 32


Those in power systematically used voting rules, procedures, or practices to diminish the ability of Blacks to vote for the candidate of their choice. This practice, known as vote dilution, often made it so difficult for the minority to partake in the electoral process that, instead of jumping through the hoops, they found that it easi er not to vote. Some common examples of vote dilution included the all-White primary, literacy tests, poll taxes, racial gerrymandering, and the more recent use of at-large elections. Throughout the South between 1896 and 1915, st ates began instituting the all-White primary. The all-White primary was a tactic used to prevent Blacks from voting for a candidate of their choosing. During this period, African Americans were ex tremely loyal to the Republican Party, because that party had gran ted their freedom. The Democrats in the South (the majority of Whites in the South were Democrats) knew they c ould control the elections as long as they did not allow the Blacks to vote for th e candidate of their choice. C onsequently, they devised a plan to decrease the number of votes available to the Republican Party by excluding Blacks from voting in the primary. Although lawmakers used this tactic to keep Blacks from voting, it also hindered the progress of the Republican Part y. As time progressed, lawmakers and local politicians used more explicit vote-dilution tactics, especially after the Supreme Court declared the all-White primary unconstitu tional in the 1944 case of Smith v. Allwright Literacy tests and poll taxes were additionally used to keep, primarily, the AfricanAmerican population away from th e polls, even if the law allowe d them to vote. Most newly freed African-Americans were poor and illiterate in fact, in some areas it had been illegal to teach a slave to read or write. Therefore, lawmakers and politicians expected that a large segment of that population, although granted th e right to vote by means of the Fifteenth Amendment, could not do so if local laws were in place requiring that voters know how to read 33


and have enough money to pay the poll tax. Howeve r, these tactics did not exclude all Blacks, because some actually knew how to read and othe rs had the means to pay the poll taxes. Thus, this vote dilution tactic was not damaging enough to the African-American vote to suit the White power base. Further, to diminish African-American influence in th e political process, politicians employed other, more inventive, tactics. According to Zelden (2002), in Georgia, Mississippi and Louisiana, polling places were set up in locations that were inconvenient and sometimes dangerous for Blacks. Some were deliberately set up in the middle of White neighborhoods or directly in front of White businesses that opposed African-American voting. These tactics presented such a challenge to AfricanAmericans that it diminished their likelihood of pa rticipating in the voting process. Some states even went as far as purging Blacks from the voter rolls by closing the re gistration rolls to new applicants at times when they knew Blacks were engaged in planting and harvesting, as Blacks primarily worked in agriculture. Another common practice was adding crimes lawmakers thought Blacks were more likely to commit to the list of felonies that disqualifie d an individual from voting for lifetoday known as felon disenfranchisement. Two other popular tactics were gerrymandering and at-large elections. Again, lawmakers did not use these tactics explicitly to keep African-Americans away from the polls but to prevent them from voting for the candidate of their choos ing; in effect making th eir vote ineffective. Gerrymandering is a method of redrawing congressiona l districts in such a way to give one party electoral advantage over anot her. According to the Merriam-Webster dictionary, gerrymandering is defined as to divide (a territo rial unit) into election districts to give one political party an electoral majority in a large number of districts while concentrating the voting 34


strength of the opposition in as few districts as possible.4 The purpose of gerrymandering is to either concentrate opposition votes into a few distri cts to gain more seats for the majority in surrounding districts (called pack ing), or to diffuse minority strength across many districts (called dilution)5. Politicians have also used gerrymandering, as referenced by the above definition, to reduce the influence of racial minorities. However, the Voting Rights Act of 1965 decreed this to be illegal. On the other hand, at-large elections are still legal. In at-large municipal elections, all members of the municipality vote for the candidates running for each districts seat. When an atlarge election takes place within a municipality, it re quires that all residents of the city vote to fill all seats in the city council, whereas, with a district-based el ection, voting rules allow voters to vote for candidates within their re spective municipal ward. In di strict-based elections, AfricanAmericans and other minorities who are concentrated with in a particular dist rict have a better opportunity to elect an official of their choice who will represent them in on the city council, because the diluting effect of an at-large election is eliminated. Criminal and political: Felon disenfranc hisement in the United StatesThen Felon disenfranchisement is a collateral conse quence that restricts a felons right to vote. However, states have not uniformly adopted it. Some states do not disenf ranchise their felons and ex-felons, whereas other states disenfranchi se their felons and ex-f elons for life unless the state grants the individual clemency or a pardon. Table 2-1 provides a summa ry of the levels of disenfranchisement within the 50 states and the District of Columbia for the year 2006. Maine and Vermont do not disenfranchise their felons and ex-felons, but Idaho disenfranchises its felons and ex-felons permanently. Until recen tly, Florida was one of those states that 4 Accessed February 24, 2007 5 35


disenfranchised both felons and ex-felons for life unless the state granted clemency or a pardon. Additionally, from this table one can also see that there is an inherent ordering by degree of severity of felon disenfranchisement law. If a state chooses to disenfra nchise its ex-felons for life, it has also chosen to disenfranchise those in prison, parole, and proba tion. Specifically, the number of states will decrease as the degree of severity of felon disenf ranchisement increases (indicator of a perfect Guttman scale). Florida is a good place to begin th e discussion of the history of felon disenfranchisement in the United States because it was one of the first states to come up with a creative scheme to disenfranchise felons. During the Reconstruction Er a, if a state wished to re-enter the Union, it was required to rewrite its state constitution to incorporate the 13th, 14th, and 15th Amendments. Florida complied; but also simulta neously restricted all felons, many of whom were Black, from voting for life (Itzkowitz and Oldak 1973). How did they manage to get away with that? Although not formally documented, some suggest that Florida found a loophole while rewriting its constitution that allowed it to exclude a la rge number of Blacks from gaining suffrage by denying felons the right to vote fo r life (Thompson 2001; Shapiro 1993). Other states, becoming more innovative, deci ded that instead of disenfranchising all criminals, they would include di senfranchisement laws only for crimes that were known to be disproportionately committed by Blacks. Unti l 1890, Mississippi law disenfranchised those convicted of any crime. However, during Mississippis Constitutional Convention of 1890, drafters narrowed the disenfranchisement law to exclude from voting only those convicted of certain offenses such as bribery, burglary, th eft, arson, obtaining money or goods under false pretenses, perjury, forgery, embezzlement, or bigamycrimes that were said to be 36


disproportionately committed by African-Ame ricans (Demos 2003; Thompson 2001; Shapiro 1993). In the same way, in 1901 the president of the Alabama Constitutional Convention, John Knox, said Why, it is within the limits imposed by the Federal Constitution, to establish White supremacy in this State (Demos 2003, 6). Demos (2003) also suggests John Knox, once a judge in a predominantly African-American district, us ed his experience on the bench to enumerate the crimes he knew African-Americans were more likely to commit. Using that list, he was able to write those crimes into the disenfranchisement la ws of the state. Acco rdingly, 12 years later, a historian found that the revised constitution of 1901 had disenfranchised approximately ten times as many Blacks as Whites (Demos 2003, 5). These new disenfranchisement laws in combination with the other tactics noted above decimated the number of freed Blacks who were elig ible to vote. In 1867, Mississippi had nearly 70% of the eligible Black populat ion registered to vote, but within two years of the 1890 convention, this number dropped to less than 6%. South Carolina, Louisiana, and Alabama soon followed suit, and the number of freed Black regist ered voters dramatically decreased for each of these states (Hench 1998). In some states, disenfranchisement did not star t out as a specific political agenda. Between 1865 and 1870, states in general adopted more rest rictive disenfranchise ment laws; however, they were sometimes the result of expedience rath er than deliberation. Wh en a colony/territory wished to enter statehood, it automatically adopted disenfranchisement laws by writing them into its constitution as other states before it had done (Behrens, Uggen, and Manza 2003), as a sort of blind adoption. However, fr om 1870-1880, states continued a dding more laws that further restricted voting rights. By th e 1880s, we begin to see a defin ite trend develop toward voting37


rights restrictions. However, in 1881, somethi ng innovative took place: Indi ana relaxed its felon disenfranchisement laws so that they affected only the currently incarcerated, making Indiana the first state to take such a step (Behrens, U ggen and Manza 2003). I ndianas innovation did not spread to other states. Between 1900 and 1940, ex isting states continued to adopt restrictive disenfranchisement laws, and new states continued their rote adoption of them. In 1921, Louisiana disenfranchised felons convi cted in federal court; in 1928, Minnesota did likewise. Prior to these date s, states enfranchised this clas s of offenders. In 1934, New York went a bit further by disenfranchi sing individuals convicted of a crime in another state that would have qualified as a felony in New York. Had someone committed a misdemeanor in North Carolina that New York considered a felo ny, that person would have lost his or her enfranchisement upon taking up residency in New Yo rk. Overall, from 1865 to the present, felon disenfranchisement laws in the United States have often ch anged, ultimately leading to a state trend of liberalizing (becoming less severe ) disenfranchisement laws. States began questioning the purpose of disenfranchising all classe s of felons for life and retreated from their former, more severe policy adoptions. Now: Post Civil Rights Era to the Present Criminal: Crack -cocaine disparity Joseph Spillane, author of Cocaine: From Medical Marvel to Modern Menace in the United States 1884 1920, said that, legal prohibitions on opium, alcohol, cocaine, heroine, and marijuana were all advanced partly on the idea that their users were deviant, dangerous others whom no society could safely tolerate. Each substance was once a significant part of the therapeutic arsenal of legitimate medicine, and each was subsequently identified as illegitimate because of concerns that were at least as influenced by fear of prejudice as by objective evaluations of the harms they caused. (Spillane 2000, 158) 38


Opium laws were models of legislation that intended to limit a specific problem but had the effect of assisting the government in control ling a specific racial or economic class, in this case Chinese Americans. Because opium was the predecessor to cocaine, opium laws provided the model by which present-day cocaine laws are ab le to exert the same control over the African American population as it had done pr eviously with Chinese Americans. It is important to make the connection to opium because it does provide a possible guideline for how the government views crack cocaine today. In the case of opium, the government made a distinction between smoki ng opium and other types of opium. Smoking opium was rather cheap, whereas other forms of opium were more expensive. Going through the necessary legal channels to obtai n opium legally in the United Stat es could cost the consumer a considerable amount of money. Smoking opium was not subjec t to the economic constraints placed upon obtaining opium legally. Consequent ly, smoking opium was the drug of choice for many opium users living in low-in come households and neighborhoods. The government made the distinction betw een these two types of opium because it believed smoking opium to have a more adverse effect on the body than other forms of opium. Coincidentally, from the 1980s to the present the same method used by lawmakers to distinguish opium from smoking opium is being used to make distinctions betw een crack cocaine and powder cocaine. The argument made about smoking opium being more harmful than other forms of opium strikes a remarkable resemblance to the one made today that states crack cocaine has more detrimental effects on the user than pow der cocaine. No plausible scientific evidence supports this claim, so why ar e there federal sentencing dispar ities between crack cocaine and powder cocaine offenses? It is meaningful to note that, cocaine in any form-paste, powder, freebase, or crack-produces the same type of physiological and psychotro pic effects, the onset, 39


intensity, and duration of its effect s are related directly to the me thod of use(The United States Sentencing Commission 1995, 14). This is one of the major discus sions behind the disparity of sentencing of powder cocaine and crack cocaine because proponents of the legislation often argue that crack-cocaine has more harmful effects than powder coca ine. The issue, however, as noted above, is not in the effect s but in the immediacy, intensit y, and duration . (Ibid). Snorted cocaine takes between three and five minutes to enter the bloodstream whereas crack cocaine only takes two minutes. Once the user obtains a high, the effects last for about forty minutes with powder cocaine and with crack cocaine the effects lasts for approximately thirty minutes. Thus, those who smoke cocaine will achieve a high sooner than those who snort it; those who snort it will remain high longer than those who smoke it. The United States Sentencing Commission, wh ose purpose was to develop sentencing guidelines for federal criminal offenses, provided a ll of this evidence. In contrast to what the presidential administration had hoped for, th e commission determined that the sentencing disparity between powder co caine and crack cocaine was not warranted Nevertheless, Congress did not agree with their recomm endation and enacted laws in 1986 and 1988 to deal with crackcocaine offenses. These laws initiated the se ntencing disparity between offenders found with either crack or powder cocaine on their person. With these la ws, Congress enacted mandatory minimum sentences for first time crack cocaine offenders. If someone were found with five grams (or more) of crack or 500 grams (or more) of powder cocaine, he or she would be sentenced to a minimum of five years in prison. If the police were to apprehend someone with 50 grams of crack or 5,000 grams of powder cocaine, the judge would have to sentence the defendant to a mandatory minimum sent ence of ten years in prison (Ibid). It may be useful to point out that, 40


Five grams of crack cocaine, the quantity necessary to trigger the five-year mandatory minimum, represents between 10 and 50 doses and costs between $225 and $750 Five hundred grams of powder cocaine, the quantity necessary to rigger the five-year mandatory minimum, represents between 1000 and 5000 doses and costs between $32,500 and $50,000. (United States Sentencing Commission 1995, 14) The Anti-Drug Abuse Act of 1986 created a rati o to determine how to weigh these drugs against one another. From its inception, crack received the higher end of the ratio where powder cocaine held the lower. Congress passed th e 100-to-1 ratio in the 1986 Anti Drug Abuse Act because it believed that crack cocaine was more harmful than powder cocaine. However, scientifically, that is not the case. Moreover, th e repercussions of these laws seem to be falling disproportio nately on African-Americans. As crack is much cheaper than powder cocaine, it is more readily available in low-income neighborhoods. The majority of the inhabitants in low-income neighborhoods are Af rican-Americans who do not have the necessary economic means to support a powde r cocaine habit. Placing crack cocaine at the higher end of the ratio heavily punishes low-income, primar ily African-American, members of inner-city communities. Political: Minority vote dilution Minority vote dilution practices aim to minimize the voting strength in minority communities. These practices usually take th e form of government mechanisms such as redistricting plans that aim to decrease the power of minority voters in situations where these voters choices would lead to the election of a minority candidate (Southern Poverty Law Center 2002). Two major forms of vote dilu tion practices that have gained attention for their impact on the minority population: at-large vs. district based elections and gerrymandering. Lawmakers and politicians used these prac tices to minimize actual minority vote strength. Each of these practices aims to crippl e the success of both minor ity votes and candidates. 41


At-large elections require all vo ters in all districts to vote for candidates running for office in a specific district. In dire ct contrast, district-based elec tions require only those voters who reside in that district to vote their represen tative into office. Why are at-large elections problematic? Foremost, it takes away the voting stre ngth of district voters in that their votes are not the only votes used to elect a representative for their district. This means that someone with no vested interest (meaning they do not live in th e district nor will be affected by the candidate who takes office in this district) in district 12 ha s an equal say as a distri ct 12 resident about who the next representative of district 12 will be. One could see how this practi ce of at-large elections beco mes an issue of minority-vote dilution. If minorities made up the majority of di strict 12 residents in a district-based election, they would have a better chance of getting th eir candidate elected be cause their votes are concentrated within the district. However, in an at-large election that advantage can disappear, and if the preferred district candidate happens to be a minority then minorities suffer a double loss: loss of the ability to have the candidate of their choice and loss of a minority representative on the city council. Therefore, the practices of at-large elections clearl y dilute the votes of minority constituents in district 12. Minority-vote dilution practices are not limited to at-large elections. District-based elections also fall prey to mi nority-vote dilution pract ices. Gerrymandering is a practice that allows the redrawing of district lines to the advantage of a pa rticular party and dilutes the strength of the minority-voting bloc. Three common practices are pack ing, cracking, and stacking. All of these methods actively dimini sh the strength of the minority-voting bloc. Packing refers to packing as ma ny minority voters into as few districts as possible to limit the number of representatives they are able to vote in to office. If a city has 12 districts, of which 42


three are the result of packing, th e minoritybloc vote will not have an effect in the other nine districts in a district-based election because mi nority voters are not in the majority in those districts. Cracking is a similar practice in wh ich lawmakers divide minority neighborhoods into many districts so that no one dist rict can form a large enough minor ity-voting bloc to affect the outcome of an election. Stacking refers to grouping an opponents constituents in areas where they are a minority. Racially, this practice would call for lawmakers to redraw heavily concentrated minority areas to include heavily co ncentrated non-minority areas. Lawmakers use the practice of stacking in such a way that the mi nority-voting bloc could ne ver attain more than 50% of the vote within the district. Even in our present era, we see how lawmaker s and politicians use political practices to weaken the power of minority vo ters. In earlier times, the em phasis had been specifically on diluting the Black vote, as mentioned in the pr evious sections, but now as the demographic landscape of the United States has changed, the targ et has since widened to include all minorities as one large group not worthy of equal repres entation in Americas political processes. Criminal and political: Felon disenfranchisementNow Felon disenfranchisement laws vary by state; forty-eight states and the District of Columbia prohibit inmates from voting while incarcerated (only Maine and Vermont allow inmates to vote). Thirty-three states prohibit felons from voting while they are on parole and twenty-nine of these states excl ude felony probationers as well. Furthermore, seven states deny the right to vote to all ex-offenders who have completed their sentences, whereas five others disenfranchise some ex-felons for a specified pe riod. Felon disenfranchisement laws strip an estimated 3.9 million Americans of their voti ng rights. African-American men make up 1.4 million of the 3.9 million Americans who are inel igible to vote due to a felony conviction. 43


Further, an additional 1.4 million disenfranchise d persons are ex-offenders who have completed their sentences. Hill (1994), Harvey (1994), and Shapiro (1993) found that race is a plausible explanation of felon disenfranchisement laws because of the disproportionate impact these laws have on Blacks and Latinos. Furthermore, empirical evidence shows a strong link between a states public policies and its minority population (Preuhs 2001). Using a variant of the racial-threat theory, researchers (Hero, 1998; Fording, 1997; Radcliff and Saiz, 1995) have found evidence that suggests that larger minor ity populations lead White lawmakers to be more likely to impose policies that undermine th e interests of minority citizens wh en the minority population in their state is relatively large (P reuhs 2001, 738). Others have explained this phenomenon using theories that suggest competition over political social, and economic resources between the White population and the minority group (Blalo ck 1967; Fording 1997; Keech 1968; Preuhs 2001; Radcliff and Saiz 1995). Furthermore, the do minant groups act in ways to maintain or increase their power relative to the minority gr oup (Ibid). According to Preuhs (2001), Given the disproportionate impact of felon disenfranchisement laws on Black and Latino citizens, coupled with the states histor ic use of procedural means to restrict minority political participation (Hill, 1994), race is a plausible explanation of fe lon-disenfranchisement policy (Harvey, 1994; Shapiro, 1993) (737).. According to Manza and Uggen (2006), Felon disenfranchisement laws raise troubling questions about American society..the adoption and expansion of these laws in the Unite d States is closely tied to the divisive politics of race and the history of racial oppression. Con cerns about the role of race are not limited to matters of historical interests. The extraordinarily high proportion of African American men in the criminal justice system today produces the shocking fact that more than one in seven Black men is currently denied the right to vote, and in several states over one in four Black men are disenfranchised. Just as felon disenfranchisement laws in several states can be traced to patterns of racial exclusion, the current effect in diluting the African American vote is no less significant. (27) 44


The historical and current relationship between race and felon disenfranchisement is clear; however, it is not as easy to de termine a states motivation for adopting these policies. Prior to the Civil Rights Era, racist policies were overt However, being a ra cist today is not as acceptable as it was in the past, so racist policies transformed them selves into what we now term race-neutral policies, thereby making it harder to determine if a policy truly has a racist intent. It is my aim with this research to make a st ronger connection between racist motivations, felon disenfranchisement, and of felon welfare-ban policies. New Methods: Crime Meets Welfare Section 115 of the federal Personal Responsib ility and Work Opportunity Reconciliation Act (PRWORA) of 1996 places a lif etime ban on the receipt of public assistance or food stamps for any individual convicted of a state or federal felony offense involving the use or sale of drugs6. Section 115 reads as follows: (a) IN GENERALAn individual convicted (under Federal or State law) of any offense which is classified as a felony by the law of the jurisdiction involved and which has as an element the possession, use, or distribution of a controlled s ubstance (as defined in section 102(6) of the Controlled Substances Act (21 U.S.C. 802(6))) shall not be eligible for-(1) Assistance under any State program funde d under part A of title IV of the Social Security Act, or (2) Benefits under the food stamp program (as defined in section 3(h) of the Food Stamp Act of 1977) or any State program carried out under the Food Stamp Act of 1977. This legislation was a last-minute floor amen dment written into th e 1996 welfare-reform act by Sen. Phil Gramm, R-Texas7. According to Travis (2002), collateral consequence policies are often added as riders to other major pieces of legislation, and theref ore given scant attention 6 Source: ( 7 Source: ( ). 45


in the public debate over the main event (16). The passage of the felon-welfare ban fits Travis description perfectly. During th is time, the nation focused on what it meant to change welfare as we know it. It was easy for legislators to overlook this last-minut e floor amendment since they were consumed by the larger debatewelfa re reform. Therefore, the public gave little attention to the felon-welfare ban because legislators thought other parts of the legislation to be more important than restricting felo ns from receiving welfare benefits. President Clintons agenda was to change welfare so profoundly that it no longer resembled the welfare of the past. Sen. Gramm picked a strategic opport unity to add the felon provision to the PRWROA, because legislators were too busy trying to figure out how the proposed change in welfare would directly a ffect their states budge ts to consider the amendments ramifications. Consequently, Cong ress passed the felon-welfare ban along with the PRWORA in 1996. At first glance, the passage of the felon-we lfare ban seems draconi an; however, although the legislation calls for a lifetime ban on the r eceipt of public assistan ce or food stamps, the federal government does grant discretion to stat es. States have the choice of adopting the legislation as is, opting out of the legislation, or limiting the period of prohibition. State Elections(A) Opt OutA State may, by specific reference in a law enacted after the date of the enactment of this Act, exempt any or all individuals domiciled in the State from the application of subsection (a). (B) Limit the Period of ProhibitionA State may, by law enacted after the date of the enactment of this Act, limit the period for which subsection (a) shall apply to any or all individuals domiciled in the State. Table 2-2 lists the number of states from 1996-2006 that fall under the various policy adoption choices outlined above. Further, the data in this table seem to suggest that states, in general, have become less restrictive in their implementation of the fel on-welfare ban. In 1997, 46


the year following the passage of this ban, 26 states adopted the felon-welfare ban without modifications. Sixteen adopted it with modifications (partial/te rm denial or dependent on drug treatment) and eight totally opted out of the ban. By 2005, we can see a general shift, with 16 states keeping the felon-welfare ban without modifications, 23 with modifications, and 12 totally opting out (See Table 2-2). Clearly, states now prefer to either cr eate modifications to the felonwelfare ban or opt out completely. The felon welfare ban marked the union of cr iminal justice policies and social welfare policies in the United States. Why is this signifi cant? As the analyses above states, historical methods of racial subordination traditionally i nvolved the criminal and political methods. Welfare, in and of itself, is racially neutral; however, it has been increasingly linked with the African-American population (Peffley and Hurwitz 2002). With the advent of the felon-welfare ban, we are seeing a new type of racial subordination emerge. Therefore, focusing on the felonwelfare ban as a substantive issue will allow me to examine a how, yet again, the forms of institutional racial subordinati on are changing and expanding to a ffect more areas of a felons life. Conclusions Why focus on institutional racial subordinatio n? The simple answer is that race has mattered since the foundation of this country and will always matter. According to Feagin and Vera (2002), The racial or ethnic divisions in our society go very deep. Simple database searches of the term does race matter? reveal how much this topic is studied. Studies on this topic vary from politics to breastfeeding, showing how race is deeply bound to almost everything in our society. Obviously, then, one cannot ignore the effect of race on crim inal justice policies, particularly considering the disproportionate im pact the criminal justice system has on African American and Latino communities. 47


According to Skogan (1996), African Americ ans are disproportionate ly represented at every step in the criminal justice process; from arrest to imprisonmentBlacks are also disproportionately likely to be executed (996). He suggests two majors reasons for this: (1) Blacks commit more crimes relative to their prop ortion of the population and (2) Black offenders are more likely to be arrested because they ar e more likely to commit arrestable crimes. Overall, this makes Blacks more likely to enter the criminal justice system and remain there. Skogans (1996) explanation for Black overrepre sentation in the criminal justice system leads to the question of why Blacks tend to commit crimes that are more imprisonable. The traditional arguments made about the tendency of Bl acks to commit more crimes is the fact that Blacks do not just commit more crimes but they commit more imprisonable crimes (Walker et al. 2004). However, I will question one of the el ements of that argument: why are the crimes that Blacks typically commit also the ones that carry greater penaltie s? Depending on how you choose to phrase the question, one can either place the responsibi lity on the individual or on the lawmakers. One version suggests simply that Blacks commit more crimes, whereas the other seems to imply that lawmakers target the punis hment of certain crimes specifically to place larger portions of the Black or minority community in prison. The same argument holds when you examine felon collateral-consequence policies. Collateral-consequence policies affect all felons and ex-felons alike, howev er minorities, as has already been stated, represent a disproportionate number of the incarcerated population. So the question still remains: Were th ese laws intended to have a disc riminatory effect (to have a disproportionate impact on the minority population)? Alternatively, do members of the minority population just commit a disproportionate amount of imprisonable crimes? 48


Since past studies have shown the impact of race on crime in the United States, I am very well aware these facts and make great effort to not let these facts dictate the outcome of this study. I have used an extreme amount of caution in letting the data speak for itself. As such, the findings presented in later chapters are led by the results of the da ta and are not dictated by what I expect the data to say. 49


Table 2-1. Level of felon disenfranchisement by state (2006) State Prison Probation Parole Ex-felons AL x x x x AK x x x AZ x x x x AR x x x CA x x CO x x CT x x DE x x x x DC x FL x x x x GA x x x HI x ID x x x IL x IN x IA x x x KS x x x KY x x x x LA x x x ME MD x x x x MA x MI x MN x x x MS x x x x MO x x x MT x NE x x x NV x x x x NH x NJ x x x NM x x x NY x x NC x x x ND x OH x OK x x x OR x PA x RI x SC x x x SD x x TN x x x x TX x x x UT x 50


Table 2-1. Continued State Prison Probation Parole Ex-felons VT VA x x x x WA x x x x WV x x x WI x x x WY x x x x TOTAL 49 30 35 12 Note: This table was adapted from the Sentencing Project and the Center for Cognitive Liberty and Ethics. 51


Table 2-2. Level of felon welfare ban by year 1996-2006 Number of states that 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Adopt without modifications 17 26 25 24 23 22 22 18 17 16 12 Partial/term denial 12 6 6 6 7 7 10 10 10 11 14 Drug treatment provision 10 10 11 12 12 13 10 12 12 12 13 Opt out 12 9 9 9 9 9 9 11 12 12 12 Note: Data sources for this table include the Sentencing Project ( ), The Legal Action Center, and the Center for Cognitive Liberty and Ethics ( 52


CHAPTER 3 THEORY AND HYPOTHESES A jar retains for a long time the odour of the liquid it first contained. (Erasmus, circa 1529) For purposes of our discussion, the jar is analogous to a body politic, whereas the odour refers to the institutions of government particularly state-government retention of sentiment and policy reflective of the widespread belief in racial and ethnic inferiority. The effect of the odour the racist sentiment has been to cling to legisl ation and policies enacted long after the overt signs of racial prejudice have dissipated. As political act s have been used to institute methods of racial institutional subordinati on, it is evident that race is intertwined with politics. One cannot easily eras e the racial and ethnic sentimen t that pushed these methods of racial subordination into existence. As gene rations continue to uphold these methods, either knowingly or unknowingly, they continue to carry out these policies negative impact upon the originally targeted members of the population. Therefore, it is easy to understand why these methods of racial institutional subordination ha ve persisted through time. Simply stated, long held beliefs of racial and ethni c inferiority do not go away over ni ght or over generations. Some collateral consequences of a felony convi ction mentioned in this work have existed through time, unscathed by any major changes. Howe ver, today we see that states are beginning to release the odour from the pr overbial jar and make major changes to their felon collateralconsequence policies. In April 2007, Florida, which once imposed lifetime disenfranchisement on ex-felons, made it easier fo r ex-felons to have their civil rights restored. The next section of this work will focus on what specific odours a state retains, specifically as related to felon collateral-conseque nces. Specifically, can we identify what keeps these methods of racial institutional subordina tion, specifically felon collateral-consequence policies, present in todays society? Are there explanations we can examine in order to get a 53


better understanding of why these me thods of racial institutional subordination have continued to exist and flourish? Is it po ssible that these methods have become so embedded in our government over time that we are unable to st op and reexamine their va lidity in our modern society? Or are we instead passively maintainin g the status quo? Alte rnatively, is adoption of these methods unconscious and rooted in long-standing political traditions? Racial Explanations Since Africans first arrived on the shores of the United States, lawmakers created laws and policies to ensure the Africans continued subor dination. Their initial subordination was overt: they were forcibly deprived of their freedom, shackled toge ther in the most horrendous conditions for a voyage through the Middle Passage that many did not survive, arriving on the shores of the United States only to be subsequently sold at an auction a nd regarded as a piece of property. Since then, those in power have alwa ys found ways, through either the criminal law, public policy, or social rules, to continue to promote their subordination. As evidenced by the status of African-Americans in ea rly America, one can clearly see th at laws were used to aid in the blatant institutional subordina tion of African-Americans. Currently, the link is le ss obvious. One must examine the laws and policies more carefully to determine if lawmakers drafted them in such a way as to hide any motivation behind their adoption. Therefore, one must hypothesize about the actual motivation of the lawmakers. I believe their intent is to continue the system of institutional racial subordination that has plagued African-Americans in the United States si nce their arrival in this country. I will advance two major groups of racial theories: group threat and culture, as explanations for the continued use of instituti onal racial subordination in the form of felon collateral-consequence-policy adoption. Culture suggests that although the overt manifestation of racial prejudice may either change or go underground over time, discriminatory racial 54


sentiment carries from one generation to the next, and those negative sentiments in our present era manifest themselves through subtler practices and policies. Group th reat, on the other hand, advances the theory that as minority populati ons get larger, the domin ant group(s) perceives them as a threat and employs methods to reduce the threat. Culture A growing trend in political psychology resear ch suggests that, due to the stereotypes Whites have about Blacks, racial ly neutral policies such as cr iminal justice and welfare are becoming racialized. These stereotypes may be deep-rooted in a concept called symbolic racism. According to Sears and Henry (2003), sy mbolic racism is described as a coherent political belief system whose content embodies four specific themes: the beliefs that (a) Blacks no longer face much prejudice or disc rimination; (b) Blacks failure to progress results from their unwillingness to work hard enough; (c) Blacks are demanding too much too fast; and (d) Blacks have gotten more than they deserve (260). The theory of symbolic racism consists of two major conceptsanti-Black affect and traditional values. Anti-Black affect is referred to as a spontaneous, non-conscious, and negative emotion that manifests itself as fear, avoida nce and a desire for distance, anger, distaste, disgust, contempt, apprehension, unease, or simp le dislike (Sears 1988, 70). Traditional values consist of those values pronounced by the Protes tant ethic such as hard work, individualism, sexual repression and delay of gratification. These two concepts merge to form the foundation of symbolic racism. Furthermore, according to Sears and Henry (2003), the term racism in symbolic racism refers to an underlying prej udice towards Blacks; whereas, the symbolism lies in the fact that the racism is towards Blacks as a whole and not toward a specific individual. Symbolic racism is more concerned with moral values instead of self -interests or personal experience. 55


McConahay and Hough (1976) specify exactly which values are most important for producing symbolic racism: The values that appear to be most important for producing symbolic racism are those associated with what we call American Civil Protestantism. Specific expressions of these values may be multidimensional, but they are, we would hypothesize, derived from the secularized versions of the Protestant ethic: hard work, individualism, sexual repression, and delay of gratification, with a large dose of patriotism and reverence for the past thrown in (41). The belief that Blacks do not honor these valu es fuels typical ster eotypes some people have about Blacks, in general; however, these valu es also help ignite r acial stereotypes about crime and welfare. Furthermore, researchers have discovered that racially neutral policies such as crime and welfare have become racialized. These policies are cognitively connected to negative perceptions of Blacks or certain members of a ta rgeted social group (Federico 2004). According to Peffley and Hurwitz (2002), Unlike affirma tive action or busing, crimelike welfareis not an explicitly racial issue. Yet both issues have become linke d to race in the minds of many Whites who (inaccurately) tend to see the typi cal welfare recipient and criminal as being African-American (59). These studies lead us to ask, Why have these r acially neutral policies become cognitively linked to African Americans? Gilens (1996) found that White opposition to we lfare is shaped more by a belief that Blacks are lazy than that poor people are laz y. Gilens (1996) also found that overall White opposition to welfare is strongly rooted in thei r unfavorable perceptions of Black welfare recipients. According to Peffley and Hurwitz (2 002), if the race-policy c onnection exists in the welfare-policy area then it is reas onable to hypothesize that the re lationship also exists in the crime-policy arena. Further, they agree, crim e, like welfare, has doubtless become racialized in the minds of many Americans, who are exposed to media accounts in which a disproportionate number of stories about crime mention African Americans (Jamieson 1992) (61). 56


It seems the media would be a plausible an swer to the occurrence of negative racial attitudes toward both welfare a nd crime policy. Media portrayal s of Black welfare queens reinforce a negative perception of African-Americans and their use of welfare. These portrayals established an image of the average African-American on welf are as one who relied on the government as a sole source of income while co ntinuing to produce children, thereby, increasing the governments burden. These portrayals often di d not depict the other side of welfare, which suggests that the majority of people on welfar e in the United States are actually White. Furthermore, they did not show the numerous Black families who were struggling to get off welfare. Thus, images of a Black, single moth er, dependent on welfare, helped put negative images of African Americans on welfare in the minds of Americans, in some cases to the exclusion of any other welfare image. More im portantly, these images fit with the symbolic racism theory that suggests this kind of raci sm is strongly fostered by an image of Black disregard for traditional values such as hard work and individualism. Symbolic racists may be prone to use the Black welfare queen as an exam ple of how Blacks cannot abide and are a threat to their traditional values. Similarly, television often portrays crime as a Black problem due to the way local news programs over-represent violent crimes perpet rated by Blacks (Peffley and Hurwitz 2002). Further, Experimental evidence suggests that even a brief visual image of a Black male in a typical nightly news story on crime is powerful and familiar enough to activate viewers negative stereotypes of Blacks, producing ra cially biased evaluations of Bl ack criminal suspects (Peffley, Shields & Williams 1994) (Peffley and Hurwitz 2002, 61). Therefore, if television often portrays criminals as being African American, it is likely that those who e xperience feelings of 57


anti-Black affect (a component of symbolic racism) will be more affected by these images and thus associate crime with Blacks. The concept of symbolic racism is quite difficult to measure because it, over time, disguises itself with other issues. Although it speaks quite directly to the issues I address, it is usually measured at the individual level with survey data. As such, it is not an appropriate concept to measure with state-level data. As su ch, I employ a general con cept of state culture to measure how a states culture influences these colla teral consequence policies. To do so, I have attempted to use the following independent vari ables as indicators fo r culture: government ideology, political culture, and religious fundamentalism. I have chosen these three indicator variables due to their ability to sustain trad ition across time. In government, politics, and religion, ideology and beliefs run deep and ofte n located at the foundatio n of structures. One last area of culture includes a states status as former confederate state. According to Preuhs (2001), The early history of felon disenfranchisement la ws suggests that these laws were a component of the systematic attempts by southern states to deny Blacks the right of political participation following the enactment of the Civil War Amendments (Fellner and Mauer, 1998; Harvey, 1994; Shapiro, 1993). (736) However, I would argue here that it was not the South that we see today; rather, he was referring more to the confederate states and their culture. Part of the reason for the formation of the Confederate government concerned the issue of sl avery with other issues such as states rights following as an afterthought. As such, I argue that the culture of confeder ate states should be different from non-confederate states such that much of the culture that formed this new government can be found in various policies 58


Furthermore, I would hypothesize the following: States with higher government ideology scores, meaning they have more liberal governments, tend to adopt less severe felon disenfranchisement policies, a lower number of collateral consequences, and less restrictive felon-welfare ban policies. States that are ranked as traditionalistic by Elazars political culture measure are more likely to adopt more severe felon disenfranchisement policies, a greater number of felon collateral consequence policies, and harsher felon-welfare ban policies. States with a higher proportion of citizen membership in fundamentalist churches to adopt harsher felon disenfranchisement laws, a greater number of collateral consequences, and harsher felon-welfare ban policies. I would hypothesize that former Confederate states would continue to maintain remnants of this racist motivation by adopting more severe felon collateral consequence policies, a large number of felon collateral consequences and harsher felon-welfare ban policies. Group-Threat Theories The early development of racial threat theori es began with V.O. Key Jr.s (1949) work on Southern politics. Key (1949) noted that in areas of the S outh where Blacks were more concentrated a real problem of politics, broadl y considered, is the maintenance of control by a white minority. Group-threat extends this theo ry in the following way: in situations where a subordinate group gains power at the expense of the dominant group, which results in the dominant group feeling threatened, the dominant group takes actions to reduce the power and therefore the threat of the subordinate gr oup (Behrens, Uggen, and Manza 2003, 15; Blalock 1967; Bobo and Hutchings 1996). According to J acobs and Carmichael (2004), V.O. Key Jr. suggest the following about the consequences of racial threat, Studies of criminal justice outcomes suggest that the threat posed by a larger racial or ethn ic underclass leads to enhanced pressure on political authorities to make greater efforts to control street crime (255). The racialthreat hypothesis has been used in research regarding: 59


The death penalty (Jacobs et al. 2005); partisan identification (Giles 1994); Incarceration rates (Myers 1990; Jacobs & Carmichael 2001); Police brutality (Holmes 2000); spending on law enforcement (Barkan and Cohn 2005; DAlessio et al. 2005); Spending on prisons (Jacobs & Helms 1999); The number of officers and strength of the police force (Liska, Lawrence and Benson 1981; Kent and Jacobs 2005); Arrest rates (Liska, Chamblin & Reed 1984). Support for racist candidates (Giles and Buckner 1993) Furthermore, according to Behrens and Manza (2003), the racial-threat hypothesis emphasizes, to varying degrees: economic competit ion, relative group size and political power. The political-threat hypothesis states that, as the percenta ge of Blacks in the population increases, Blacks may be able to leverage democr atic political institutions to their advantage (Behrens and Manza 2003, 575). According to th e political-threat hypothesis, larger minority populations tend to be associated with a trend in White lawmaker s adopting policies that are not in the best interest of the minority population. In other words, they perceive this growing segment of the population as a threat. To combat the threat, they enact policies that will undermine the interests of minority citizens, in terms of either partic ipatory rights or the distribution of policy benefits (Preuhs 2001, 738). This behavior on the part of lawmakers underscores the fact that they don t see themselves as representing one population of citizens, but rather, two populations whose interest must n ecessarily conflict and that this two population situation harkens back to the deep-ro oted us vs. them way of thinking Blalock (1967) suggested that racial-threat exists, where di fferent racial groups compete for political, social and economic benefits, resulting in the dominant group feeling threatened by these other groups and taking actions to preserve or increase its own power. Blalocks racial60


threat hypothesis suggest that the dominant group uses tools that are to its advantage, such as the criminal law, to control subordinate groups that threaten its interests (B lalock 1967; Eitle et al. 2002; Jacobs & Carmichael 2005). The racial-t hreat hypothesis states that when the Black population gets progressively larg er in size and Black-on-White crime increases, social controls intensify. According to Preuhs ( 2001), If the variation in felon disenfranchisement laws is due to racial politics, one would also expect these laws to be more severe when the targeted population is composed of a higher number of minorities than nonminorities (739). This logic also holds for th e percent minority on state welf are rolls. According to the research of Soss et al. (2001), one of the best predictors of state adoption of tough welfare policies is the percentage of non-White adults on state Aid to Families with Dependent Children and Temporary Assistance to Needy Family rolls. Accordingly, this data was collected for the welfare ban portion of this analysis. The economic-threat hypothesis asserts that competition between Whites and Blacks for jobs and other resources results in an increase in soci al control. Jacobs and Carmichael (2004), however, suggest that the result of economic threat is that imprisonment and other punishments are used to control the excess supply of labor in capitalist eco nomies (256). However, the research in this area, according to Jacobs and Carm ichael (2004) is elusive or contradictory. In their study, they tested the hypothesis that the number of death sentences will be more substantial in jurisdic tions with higher unemployment rate s. However, they found that the unemployment rate is completel y ineffective has no effect on the number of death sentences and conclude that using another meas ure may achieve different results. According to the group-threat theories, when such conditions exis t you can expect intergroup conflict to arise, resulting in the majo rity group becoming reluctant to extend societal 61


privileges to minority groups during these periods (Soule and Earl 2001). An initial thought would be to use only the unemploym ent rate to capture the economi c aspect of the group-threat theories. However, the group-threat theory encompasses more than just economics. It must also take into account the racial competition for econom ic resources. Therefore, I use a measure called idle White males to measure the interaction of the ra cial and economic components of the group-threat theory. This measures the percenta ge of the White male population that is neither unemployed, in a correctional institution, nor enrolle d in an institution of higher education. This measure tells us that if there were more id le White males in the population, then you would expect unemployment to be on the rise and therefore expect to see an increase in the competition for economic resources. Furthermore, I woul d expect the majority group to perceive unemployed minorities as a thre at because of their race a nd their unemployed status. The last measure of group threat I intend to employ is the Black incarceration rate, which measures the rate at which Blacks are incarcerat ed relative to their populat ion within the state. As Blacks are incarcerated at a higher rate than any other grou p in the United States, a states Black-incarceration rate is a good indicator of how states criminal jus tice systems treat their minority residents. Further, the Black incarceration rate measurement is used as a proxy for minority incarceration rate as these rates are not calculated consisten tly for other minority groups. To summarize I would hypothesize the follo wing about the group-threat variables: States with a large percentage of non-Whites in its population are more likely to adopt more severe felon disenfranchisement laws and felon-welfare ban policies, and a large number of collateral consequences. States with a high percentage of idle White males can be expected to institute a greater number of collateral consequences. Further, states with a large proportion of idle White males will also have more severe felon-disenfranchisement and felon welfare-ban policies 62


States with high Black incarceration rates are more likely to adopt more severe levels of felon disenfranchisement laws and felon-welfare ban polices. Additionally, I would hypothesize that these states also adopt a large number of felon collateral consequence policies. State with a high percentage of minorities on its state welfare rolls would be more likely to adopt the more severe levels of felon welfare ban. Alternative Explanations Political Explanations The second set of theories looks more at the political explanations fo r the continued use of institutional racial subordination. Here, I briefly present the polit ical explanations that suggest these methods of institutional racial s ubordination continue to exist because: Those in power compete for the mainstream voter (who in this case is not a member of the minority) and subsequently succumb to the requests of the majority and continue to do so when in elected office. Methods of institutional racial subordination can continue because low-income voters do not turnout in high numbers on election day to challenge policies that may adversely affect them. Since V.O. Keys seminal work on southern state politics (1949), sc holars have recognized the importance that political competition can have on the nature of politics in the states. First, it is generally recognized that competition influences the types of policies a state produces. States with highly competitive environments tend to sp end more on social-wel fare public policies, while noncompetitive states generally spend less on programs for what Key (1949) referred to as the have-nots. Second, it is widely recogni zed that states with competitive environments generally have higher levels of vot er turnout in their elections. According to Cnudde and McCrone (1969), Party competition, by producing some semblance of an organized politics, lessens the difficulty of lower status groups in sorting out political actors and issues, there by enabling them to promote their own interest more effectively (858). Holbrook and Van Dunk (1993) provide further clarification by stating that elected officials in more competitive areas are more re sponsive to their constituent needs and therefore 63


provide greater benefits to those constituents with lower socioeconomic interests (955). They also suggest that elected officials in competitive states are more likely to support policies that benefit the have nots in society than. Barrilleaux, Holbrook, and Langer (2002), also support this reasoning but go further by suggesting the unde r conditions of close party competition, parties have to adjust their policy positions to support the voting class that makes up the majority of the population. As such, if th e majority of the population consis ts of lower income individuals then the parties will adjust their policy positions to favor that group (416). Furthermore, according to V.O. Key (1949), h ave nots loose in di sorganized politics (315). Therefore, I would hypothesize: More competitive states will be less likely to adopt more severe criminal-justice policies, particularly large numbers of collateral consequences and harsher felon-disenfranchisement and welfare-ban policies. Hills (1994) work demonstrates that states have different levels of democracy and that these differences can have policy consequences. He determines a states democracy level by evaluating the extension of voting rights, competitiv eness of political parties, and the level of voter turnout in the state (Hill 1994; Wright 1996). Hill (1994) assigned each state a democracy score based on how inclusive a states democratic pr ocedures were. States with higher scores are more democratic. Therefore, I hypothesize that, States with high democracy scores to be less lik ely to adopt severe felon disenfranchisement laws and felon-welfare ban polices. Additionally, I would expect these states to adopt a low number of felon collateral consequence policies. Squire (1992) based his legi slative professionalism scale on Polsbys (1975) examination of the professionalization movement in which certa in attributes would ch aracterize a legislature as professional. States are cla ssified as more professional if th ey have qualities that mimic the United States Congress. For Polsby (1975), it meant looking at the establishment of a 64


respectable pay scale, provisions for independent staff services, a nd increase in the time allowed for legislatures to sit (Squire 1992). Squire (1992) developed a s cale using these attributes of each state as compared to the United States C ongress, as most view Congress as the ideal professionalized legislature. Therefore, I would hypothesize the following: States with more professionalized legislatures to adopt less severe felon disenfranchisement laws and felon-welfare ban policies, and a low number of felon collateral consequence policies According to Uggen and Manza (2006), Politic al actors must formally introduce, propose, and vote on the creation of, or amendments to, disenfranchisement laws (63). This too becomes important when we think of the highest offi ce within state governme nt: the governor. The governor is the executive branch of state government and as such is responsible for signing laws. A governors political party affiliation should ha ve an impact on their law-making process. Therefore, based on the traditional positions ta ken by the Republican and Democratic parties on crime issues, I would hypothesize that states headed by Republican governors are more likely to adopt harsher felon disenfranchisement laws and felon-welfare ban policies. These states would also adopt a greater number of overall felon-collateral consequence polices Class-Based Explanations Is it race or is it class? In todays society as race blurs income levels, ones race cannot be tied to their social class. Th rough the efforts of affirmative action and anti-discrimination laws, minorities are moving up the corporate ladder and taking lead roles in some of this nations most prestigious organizations. Due to these cha nges and other societal factors class-based explanations for specific policy adoptions have taken on a new momentum. Those who support the class-based over race-based explanations suggest that class, rather than race, motivate policy decisions. However, the difference may not be distinct enough to 65


exclude either factor. In a study of perceptions of injust ice in the United States, Hagan and Albonetti (1982) found that both race and class influence perceptions of injustice. Furthermore, Black Americans were more likely than White Am ericans to perceive that the criminal-justice system discriminated against marginalized group s (Hagan and Albonetti, 1982). Just the same, differences amongst people of different soci o-economic groups were also found. These differences suggest that unemployed people were more likely to perceive injustice than employed people. Hagan and Albonetti (1982) also concluded that class, somehow, conditioned the relationship between race and pe rception of injustice. They came to this conclusion because differences in perceptions were most apparent among members of the highest social classes whereas among unemployed individuals, socio-econom ic status was more influential than race (Hagan and Albonetti 1982). We can see at the individual level that both race and class intertwine when talking about the criminal justice system. Theref ore, it is likely that these two concepts also play an integral role in criminal-justice policy making. We must al so remember that ultimately it is individuals who come together to make these policy decisions, ther efore, it is possible for these issues to become dominant at the state-policy making decision level. Additiona lly, according to LewisBeck (1977) one must consider the influence of socio-economic variables on state policies, hence, even more reason to not dismiss class as a possible determinant in state-criminal justice policy making. For the purpose of the influence of class-based explanations on disenfranchisement laws, it is also important to note that Blacks were not the only target of the earlier practices of disenfranchisement. The poor were also targeted (poll taxes). Therefore, I hypothesize the following: States with high lower income turnout would be more likely to adopt policies which are less severe and lower in magnitude than states with low levels of lower income voter turnout. 66


States with high levels of income inequality would be more likely adopt more severe felondisenfranchisement laws, a larger number of collateral consequences, and the most severe level felon-welfare ban policies. States with a high unemployment rate would be more likely adopt the most severe of all of these policies to maintain social order within the state. 67


CHAPTER 4 DATA AND METHODS This section introduces methodol ogical procedures for investig ating the correlates of the following: The severity of felon-disenfranchisement laws (2006); The number of collateral consequences in the fifty states and the District of Columbia (1986, 1996, and 2006); The severity of felon-welfare bans (19972006). I divided this chapter according to the dependent variables listed above. For each dependent variable, I outlined the exact met hodological procedures used to investigate the relationship between the dependent and the independent variables. A sample of the bivariate relationships between these variab les are provided for the felon disenfranchisement law model in Table A-1 of the Appendix. Data Measure of Dependent Variables Severity of disenfra nchisement (2006): I compiled data for the level of disenfranchisement laws using various pub lications produced by the Sentencing Project. Subsequently, I coded the severity of felon disenfranchisement to show the increasing level of severity from no disenfranchisement to the disenfra nchisement of ex-felons. The lowest level of severity (coded as low) includes states that do not disenfranchise felons during their incarceration period or while they are on probation. The next leve l (coded as moderate) includes all states that disenfranchise felons who are on parole. Last, the most severe (coded as high) category of felon disenfranchisement includes onl y states that disenfra nchise felons postincarceration period (or probation period) and those on parole; that is, they permanently deny felons the right to vote. 68


Consequently, one could classify some states as having more than one level of felondisenfranchisement law because they disenfranc hise both probationers and parolees. In other words, such a state meets the cr iteria for both the low and modera te-severity cate gories of felondisenfranchisement. However, to avoid any ove rlap, I placed states into categories based on their highest level of disenfranchisement. Table 41 depicts this categoriza tion of the severity of felon disenfranchisement laws and a listing of each state that would fall into the associated category. Number of collateral consequences 1986, 1996, and 2006 : I collected data for the number of collateral consequences by state fo r the years 1986 and 1996 from the work of Burton, Cullen, and Travis (1987), and Olivares and Bu rton (1996). These researchers systematically reviewed the legal codes and statutes attributed to nine collateral consequences for all fifty states: voting, parenting, divorce, pub lic employment, jury duty, hol ding public office, firearm ownership, criminal registration, and civil death. Additionally, in cases where statutes were unclear, all authors consulted case law for clarifi cation. In some cases, these researchers went a step further by mailing letters to the Attorney General of each state, which summarized their findings on the collateral-conseque nce statutes, asked them to re view the researchers findings, and to make clarifications where necessary. I chose to focus primarily on specific collatera l consequences to incl ude: loss of parental rights, prohibition from public employment, voting ri ghts restrictions, right to own a firearm, and right to serve on a jury. The motivation for c hoosing these specific colla teral consequences was that each of these collateral c onsequences either poses a barrie r to family life, professional success, or civic participation as defined by the Center for Cognitive Liberty and Ethics1. From 1 69


the data derived from the previous works of Burton, Cullen and Travis (1987), Olivares and Burton (1996) and data I collected for 2006 from the Center for Cognitive Liberty and Ethics online library. Further, states ranged from havi ng zero to all five collateral consequences as defined above (See Table 4-2). The severity of felon welfare ban laws (1997 and 2006): I compiled data for the felonwelfare bans from various sour ces, including the Sentencing Proj ect, The Legal Action Center, the Lexis Nexus database, the Drug Policy Alliance, the National Governors Association, the United States Department of Agriculture, and the Center for Cognitive Liberty and Ethics (solely for the 2006 data). I coded the level of felon-we lfare bans to reflect the increasing level of severity. The lowest level of the felon-welfare ban is the opt out cate gory where states opted out of the requirements of the felon-welfare ban as proposed by the federal government. For that reason, I coded this category as level one. Next, category two in cludes states that adopted the felon-welfare ban but added a drug-treatment pr ovision to the legisl ation. Category three includes states that adopted some parts of the fe lon welfare ban or have instituted a time limit on the federally imposed welfare ban. Lastly, the most severe level of felonwelfare ban, level four, includes states that adopted the felon-welfare ba n without modifications. These states adopted the federally proposed felon-welf are ban as is and thus instit uted the lifetime ban on welfare benefits for anyone convicted of a drug felony (See Table 4-3). Measurement of Independent Variables Group threat variables: racial and economic Percent non-White (population): This variable consists of the number of individuals in a state considered non-White for Census purposes. I collected this data from the United States Census publication titled Residen t Population by Race and State in the Statistical Abstract of the United States for the years 1980, 1996, and 2005. I created this variable by, first, subtracting 70


a states total White population from its total population to obtain the number of non-White residents within a particular state. To obtain the percentage of the st ates population who are non-White, I divided the states non-White popula tion by its total population and converted the quotient to a percentage, thereby creating the percent non-White population variable. Percent non-White on welfare rolls: I included this variable in the felon-welfare ban portion of this analysis as a pred ictor for state adoption of the mo st severe levels of the felon welfare ban. I compiled data for this variable from the United St ates Department of Health and Human Services: Office of Family Assistance for the years 1997 and 2006. Black incarceration rate: This measurement tells us th e rate at which Blacks are incarcerated within a particular state. These me asurements are also available for other minority groups; however, the data are sparse and could not be used in this study. I complied data for this variable from the Bureau of Justice Statistics bulletin titled Prisoners and Jail Inmates at Midyear for the various years under study. Idle White males: This measures the per centage of White males, ages 15-39, unemployed or neither in the labor force nor in school. Manza and Ugge n (2006) report the following on their measurement of idle White males, Using historical data from the United States Census Bureau, we calculate...[the rate of White male idleness and unemployment in each state]by dividing the number of White males aged 15 to 39 who were either unemployed or idle (defined as neither attending school nor participating in the labor force) by the total White male population aged 15 to 39 (62). Therefore, this measure indicates that if the number of idle White ma les in the population is higher than expected, we should expect a rise in the competition for economic resources. Sources for this measure include the Bureau of Labor Statistics and the Integrated Public Use Microdata Series produced by the Minnesota Population Center ( ). 71


Culture variables Government ideology: Preuhs (2001) believes that one can use the dominant ideology of a state to determine a states inclination toward criminal justice policies. Further, Due to the path-breaking work of Wright, Erikson and McIver (1993) and their subsequent research, there is consensus that interstate differences in public ideology are important in accounting for notable differences among the states in the policies they adopt (Brace et al. 2002, 1). Therefore, to capture this concept I us ed the state government-ideology measurement created by Berry et al. (1998), which measures the dominant ideology of the states political elite. According to Berry et al. (1998), state-government ideo logy is the mean position on the same continuum of the elected public officials in a state, weighted according to the power they have over public policy decisions (328). Many researchers over time attempted to improve upon the measurement of government ideology including the widely us ed Wright, Erikson, and McIver measures of 1985 and 1993. However, Berry et al. (1998) argue that this measure also has flaw s, due to its insensitivity to annual changes in government ideology and the role of the governor in state policy-making; thus, they assume that government ideo logy is a static concept. Berry et al. (1998) believe that one can improve upon this concept by obtaining ideology scores for the governor and the major party delegations in each house of the state legislature. One can then aggregate these scores based on certain assumptions about the distribu tion of power among policy-makers. Therefore, their measure of government ideology takes into account the ideology of the governor and the two major parties within the state legislature. With Be rry et al. (1998) government-ideology scale, higher scores represent states with mo re liberal governments, whereas, lower scores correspond to governments that are more conserva tive. Richard Fording (2007) updated Berry et 72


al. (1998) government ideology variable to in clude 1994-2006, since Berry et al. (1998) data ended with the year 1993. Political culture: One can describe a states culture as moralistic, traditionalistic or individualistic, according to Elazar (1966). Political culture refers to what people believe and feel about government, and furthermore, how they think people should act toward the government (Elazar 1966). Elazar provided three political culture types in the United States: Moralist in this type of political culture, society is more important than the individual is; community comes first; issues are important; government is seen as being positive; good government is measured by the degree in which it serves the public good rather than the individual. Individualist private concerns are more important than public concerns; the government is instituted for utilitarian reasons; government action should be restricted to areas that encourage private initiative and widespread access to the marketplace (Elazar, 87). Traditionalist social and family ties are very important; accepts hierarchy as a natural ordering of things which means that those at the top have a dominant role in the government; government is seen as being of use only when it is to secure the maintenance of the existing social order. Based on the stark differences between these classifications, the cultures of the states clearly emerge when they are required to make certain decisions. I obtained the measures for these variables from Elazars 1966 work and as a result do not vary for each year under study. Religious fundamentalism: Religious fundamentalism meas ures the percentage of a states population who are members of fundamenta list churches. Jacobs and Carmichael (2004) measure religious fundamentalism based on the log of a scale created by Morgan and Watson (1991) that relied on church membership data enum erated by Quinn et al. (1982). They further indicate the following in reference to the use of this scale, This scale is only available for 1980, but Newport (1979) suggests that church affiliatio ns in large aggregates like states are stable (Jacobs and Carmichael 2004, 260). Jacobs and Ca rmichael (2004) provided the data for this variable. 73


Confederate state: I classified states as Confederat e based on whether they we once a Confederate state (dummy variable). The form er confederate states were South Carolina, Mississippi, Florida, Alabama, Ge orgia, Louisiana, Texas, Virg inia, Arkansas, North Carolina, and Tennessee. These variables were coded with a one to indicate their status as former confederate states and all othe r states were coded as zero. Political variables Level of legislative professionalism: Squire (1992) measures legislative professionalism using a scale where he awards mo re professionalized legislatures higher scores. Squire (2007) updates his original work to include the ye ars 1979, 1988, 1996, and 2003. As such, I used his 1988 data in the 1986 model and the 1996 data in both the 1996 and 1997 models (since they were separate models) and the 2003 data as a proxy for 2006 data. Democracy score: Preuhs (2001) used state democracy score to capture a states disposition toward democratic inclusion (741). Preuhs (2001) adopted this measure from Hill (1994) for the years 1980 and provided the data for this variable. In line with the method adopted by Preuhs (2001), I also used Hills (1994) democracy score measure for all years in this study. Inter-party competition: This variable is based on the di fference of proportions for seats controlled by each major party (Democrat a nd Republican) in each states lower and upper house (Soss et al. 2001, 392). It ranges from zero to one where values closer to one indicate greater inter-party competiti on. I have recreated all data for 1986, 1996, and 2006 from the Statistical Abstract of the United States by creating a measure that takes the difference of proportions of total seats held by the Democrats and the Republi cans in each house. As an illustration, if Republicans had 34 of the 62 total seats (both houses) and Democrats had 28 of the 62 seats. The difference in proportion of the number of seats held between these two parties 74


is 0.0967 indicating that they are almost equal in the number of seats that they share. Then I subtracted this number from one to create a scale in which higher values equal greater competition. Republican governor: For all years under study, I constructed a dummy variable to indicate the presence of a Republican governor. Accordingly, a indicates that the state has a Republican governor and a indicates the presence of a non-Republican governor. The National Governors Association publishes this data and it was used for all years under study. Class-based variables Lower-class voter turnout: I have based my measurements of class-based voter turnout on the work of Hill and Leighley (1992). However, I obtained the data for this variable from the Dataverse website of William D. Berry. Unfortuna tely, this data is only available for the year 1986. For that reason, I have had to recreate this variable for the years 1996, 1997 and 2006 by following the methods of Hill and Leighley ( 1992) that use the U.S. Census Bureaus Current Population Survey: Voter Supplement Files. These files are available through the University of California at Los Angeles Institute for Social Sc ience Research Data Arch ives for all years under study 2 The method for constructing this variable consists of dividing the respondents by state then dividing the percentage of a states voter s who are wealthy (or poor) by the percentage of the states adult population whic h is wealthy (or poor) (Hi ll and Leighley 1992, 354). The measure is multiplied by 100 to create an index of class bias. A value of 100 represents an equal representation of high and low-in come voters at the polls. Wher eas, values below 100 indicate higher levels of low-income voter turnout and values above 100 indicate higher levels of highincome voter turnout. 2 75


Income Inequality: This variable measures the distribu tion of income inequality within a state, also known as the Gini index. The Gini index is calculat ed uses various income related Census items to compute the measure of income in equality. The scores range from zero to one with a zero indicating almost perfect income equality and one indicating greater income inequality. The United States Ce nsus Bureau houses this data on its website for the years 19992006. As such I have used the measuremen ts for the corresponding years under study. Unemployment rate: I used data for state unemployment rate from the United States Bureau of Labor Statistics report titled, Geographic Profile of Employment and Unemployment. The Bureau of Labor Statistics collects this data every month. I used the date reported in January for each of the years under study: 1986, 1996, 1997, and 2006. Control variables Incarceration rate: Uggen and Manza (2006) suggest using incarceration rate as a variable in these types of studies because one can use it to isolate the effects of racial threat from the effects over overall state punitiveness (63). I obtained the measure of incarceration rate by state from the Sourcebook of Criminal Ju stice Statistics Table 6.29.2004. As reported in the Sourcebook, this rate represents sentenced prisoners per 100,000 residents under jurisdiction of state and federal correctional authorit ies on December 31, by region and state, 1980, 19842000. According to this report, the Sourcebook of Cr iminal Justice Statistics compiles their data using a year-end census of prisoners in state and federal custody. The table presented by the Sourcebook of Criminal Justice Statistics used various governmental sources in their data collection procedure, which are included in the reference section of this study. 76


Total state population: According to the United States Census Bureau, Estimates are made for the United States, states, counties, places, and metropolitan areas within the United States. The timing of estimates and availability of demographic detail vary by geographic level. The schedule of releases is available at I derived this data from the United States Census Bureaus webs ite for all years under study. Population density: Daley and Garand (2005) suggest using population density as a measure because urban populations are often more supportive of governmental solutions to policy problems than less urban populations (6 27). The usual procedure for determining population density includes dividing the total po pulation in the state by the land area of each state measured in square miles. I relied on the calculations provided by the United States Census Bureau Land Area, Population, and Density for States and Counties for 1990 and 2000 Census years for this variable. Methods Level of Felon Disenfranchisement The level of severity of felon disenfranchisement reflects an ordered outcome. Thus, I ranked the category of ex-felons higher than the other categories in terms of level of severity of felon disenfranchisement. For th at reason, it is appropriate to es timate the relationships between the level of disenfranchisement and the explan atory variables using ordered logit regression model. Although the response categories are not inherently ordere d, I created an unobserved ordering by having each category refer to a states propensity for adopting severe disenfranchisement laws. Therefore, a state lo cated in the high severity category of felon disenfranchisement adopted more severe felon-disenfranchisement laws than states not in that category. 77


Similar to the work of Preuhs (2001), I esti mated four models to test the relationship between a states current level of severity of disenfranchisement laws and race-based, classbased, cultural, and political variables. The firs t model is the race-based model, which test the relationship between race-related va riables and a states level of felon disenfranchisement. The remaining models separately test the impact of culture, class-based, and political variables to identify the individual effects of these particular groupings of va riables. Lastly, I ran a full model to identify the simultaneous effects of th e race-based, class-based, and political variables on a states current level of sever ity of felon disenfranchisement. This full model only contained variables found to have a substantia l impact in the individual models. Number of Collateral Consequences Generalized Estimating Equations are an exte nsion of the general linear model for data collected on the same units across different poin ts in time. Because GEE methods take these correlations into account these models providi ng valid estimates and standard errors. GEE allows one to compare observati ons over time without actually employing time-series analyses. As such, I use a GEE model with poisson specified as the family and the log link specified as the distribution function to estimate the relationship between the numb ers of collateral consequences at the various time points. Felon-Welfare Ban Unlike the methods used for the level of dise nfranchisement in which the specific ordering of the dependent variable were considered by wa y of an ordered logit mo del, since so many of the models violated the parallel regression assumption alternate models had to be used. For the race model I used a generalized ordered logit model (Williams 2005). However, for the culture model, I used a multinomial logit model to estima te the relationship between a states level of felon-welfare ban and the cultural variables. Las tly, the political and class models were able to 78


be estimated using a simple ordered logit methods With the ordered m odels, increases in the category of felon welfare ban from one to four in dicate the increasing severity of this law. In other words, as the categories of felon welfare ban increase from one to four the level of severity of felon welfare ban also increases. Additionally, I have measured th is relationship at two separate time points: a year after the initial adoption (1997) and the cha nge in level of felon-welfare ba n ten years after their initial adoption, in 2006. I chose to use these time points to, first, determine which of the independent variables influenced a states ini tial decision to adopt a specific level of fe lon welfare ban (less severe versus more severe); second, to determine what accounted for a states change in level of felon-welfare ban policy from 1997 to 2006. 79


Table 4-1. Level of severity of felon disenfranchisement laws (2006) by State Low Moderate High D.C., Hawaii, Indiana, Iowa, Maine, Massachusetts, Michigan, Montana, New Hampshire, North Dakota, Ohio, Oregon, Pennsylvania South Dakota, Utah, Vermont Alaska, Arkansas, California, Colorado, Connecticut, Georgia, Idaho, Illinois, Kansas, Louisiana, Minnesota, Missouri, New Jersey, New Mexico, New York, North Carolina, Oklahoma, Rhode Island, South Carolina, West Virginia, Wisconsin Alabama, Arizona, Delaware, Florida, Kentucky, Maryland, Mississippi, Nebraska, Nevada, Tennessee, Virginia, Washington, Wyoming 80


Table 4-2. Number of collateral consequences (1986, 1996, and 2006) by state State 1986 1996 2006 Alabama 4 4 5 Alaska 2 1 3 Arizona 2 2 1 Arkansas 3 3 3 California 3 3 3 Colorado 1 1 1 Connecticut 1 1 3 Delaware 3 3 3 D.C. 2 1 2 Florida 3 3 3 Georgia 2 2 3 Hawaii 2 2 2 Idaho 1 2 2 Illinois 1 1 2 Indiana 3 2 2 Iowa 4 4 2 Kansas 2 2 2 Kentucky 3 3 3 Louisiana 0 1 2 Maine 1 1 2 Maryland 1 2 3 Massachusetts 2 2 3 Michigan 2 1 3 Minnesota 0 0 2 Mississippi 4 4 2 Missouri 1 1 2 Montana 2 2 3 Nebraska 2 2 4 Nevada 4 4 2 New Hampshire 1 1 2 New Jersey 1 1 1 New Mexico 3 3 1 New York 2 2 3 North Carolina 0 0 2 North Dakota 0 0 2 Ohio 1 1 2 Oklahoma 2 2 3 Oregon 2 2 1 Pennsylvania 2 1 4 Rhode Island 4 4 1 South Carolina 2 2 2 South Dakota 1 2 1 Tennessee 3 3 3 Texas 2 2 3 Utah 1 2 3 81


Table 4-2. Continued State 1986 1996 2006 Vermont 0 0 1 Virginia 2 3 3 Washington 0 2 1 West Virginia 1 2 2 Wisconsin 2 3 2 Wyoming 2 3 2 82


Table 4-3. Level of felon welfare ban 1997, 2006, and the difference in levels from 1997-2006 State 1997 2006 1997-2006 Alabama 4 4 0 Alaska 4 4 0 Arizona 4 3 1 Arkansas 3 3 0 California 4 4 0 Colorado 2 3 1 Connecticut 1 3 2 Delaware 4 2 2 D.C. 1 1 0 Florida 3 3 0 Georgia 4 4 0 Hawaii 2 2 0 Idaho 4 3 1 Illinois 2 3 1 Indiana 4 4 0 Iowa 2 2 0 Kansas 4 2 2 Kentucky 4 2 2 Louisiana 3 3 0 Maine 4 1 3 Maryland 2 2 0 Massachusetts 4 3 1 Michigan 1 3 2 Minnesota 3 2 1 Mississippi 4 4 0 Missouri 4 4 0 Montana 4 2 2 Nebraska 4 3 1 Nevada 2 2 0 New Hampshire 1 1 0 New Jersey 4 2 2 New Mexico 4 1 3 New York 1 4 3 North Carolina 2 2 0 North Dakota 4 4 0 Ohio 1 1 0 Oklahoma 1 1 0 Oregon 1 3 2 Pennsylvania 4 1 3 Rhode Island 3 1 2 South Carolina 2 2 0 South Dakota 4 4 0 Tennessee 4 1 3 Texas 4 4 0 Utah 3 3 0 Vermont 1 1 0 Virginia 4 2 2 83


Table 4-3. Continued State 1997 2006 1997-2006 Washington 2 1 1 West Virginia 4 4 0 Wisconsin 2 2 0 Wyoming 4 1 3 Note: For 1997 and 2006: 1= Opt out; 2=Adopt with term/partial Modification; 3=Adopt with a drug treatment requirement; 4= Adopt ban as is from federal legislation. 84


CHAPTER 5 FINDINGS AND DISCUSSION What does the data reveal a bout the relationship between r ace-based, political, cultural, and class-based variables and state felon-colla teral consequence policy choices? Were the specified models adequate to explain what type s of factors states are most reliant on when making decisions? This chapter hi ghlights the findings of the mode ls specified in Chapter 4 and discusses the implications of t hose findings in relation to the theories and hypotheses advanced in Chapter 3. Before I begin reporting the findings, I must firs t digress to discuss an important issue with the interpretation of the models presented in this chapter. The fifty states and the District of Columbia make up the population of states with in the United States. As such, the models presented are not using inferential statistics based on sample data as is done in the majority of studies using inferential statistics. Researchers rarely analyze popul ation data because for some populations, like adults living in the United States, it is virtua lly impossible to obtain data fo r all members in the population. As a result, they rely on inferential statistics for sample data, which is a re presentative subset of the population, and apply those findings to the p opulation. These findings are used to measure the likelihood of such results actually occurr ing due to a true rela tionship between these variables or due to random error (chance). In other words, what is the probability of the observed result being due to chance? By conventi on, social scientists are not willing to accept a probability above 0.05 or 5%. The rationale above is the basis of hypothesis testing. In all of the models presented in this chapter, the null hypothesi s states that no relationship (no association) exists between each of the inde pendent variables and the dependent variable. Accordingly, if the true state of affairs is no association, what is the probability of observing 85


the reported coefficient or one more extreme (American College of Physicians 2001)? This is the question answered with hypothesis testing. Ho wever, if you recall the data used in these models are population data and the coefficients produced reveal th e true state of affairs. Therefore, traditional interpretation of a hypothesis test has no value here. Further, three specific factors influence statistical significance: (1) the size of the coefficient; (2) the number of observations; and (3) the spread of the data. It is noted that small N studies (studies with a small sample size) often report non-significance even when there are important effects (American College of Physicians 2001). Further, in this an alysis the spread of the data is limited to the number of categories created. Due to all of the reasons mentioned above, it is my point to focus on the magnitude of the relationshi p instead of the results of a hypot hesis test. If the magnitude is closer to zero then we can say that our independent variab le has no effect on the dependent variable. Felon Disenfranchisement Laws Race Models This model contained three race specific va riables: Black incarceration rate, minority population, and the percent of idle White males. There were two va riants of this model to test specific hypotheses about a states non-White popul ation and its level of severity of felon disenfranchisement laws. Model 1 included th e percent minority in the population variable, whereas the second model included an additional variant of the percen t non-White variable, squared percent non-White in the population. Model two tests the alternate hypothesis that minority threat increases then decreases as the minority population increases exponentially. The coefficient for black incarceration rate was -0.730 (See Table 5-1 Race model 1). However, for ease of interpretation I will transl ate that coefficient into an odds ratio by exponentiating the coefficient usi ng base e as follows (odds= e-0.730). This indicates that for a one 86


percent increase in a states Black incarceration ra te, the odds of a state fa lling into the high level of severity of felon disenfranchisement polic y versus the combined moderate and low felon disenfranchisement categories is 0.482 times lower, given that percent idle White male and percent non-White population are held constant. Furt her, the probability of a state such as South Dakota, with very high Black in carceration rate, falling into th e high felon disenfranchisement severity category is 0.16. Overa ll, we can conclude that states with higher Black incarceration rates are less likely to adopt the high severity level of fel on disenfranchisement category. For a one percent increase in a states idle White male population, the odds of a state being observed in a high level of felon disenfranchisement category is 0.665 times lower than being observed in the combined moderate and low cate gories, holding Black incarceration rate and percent non-White in the population constant. Cons equently, such a state would most likely fall into the moderate and low categories of felon disenfranchisement. Additionally, Michigan, the state with the highest idle-White male population has a 12% chance of adopting a high severity felon disenfranchisement law. Lastly, the odds of a state th at increases its minority populatio n by one percent falling into the combined moderate and high categories vers us the low category is 40.731 times greater. Further, a state like Alabam a with a high minority population would have a 62% chance of adopting a high severity felon di senfranchisement law and a 30% chance of adopting a moderate severity felon disenfranchisement law. On th e other hand, the results fo r the alternative race hypothesis (percent minority square d) reveal the opposite. Furtherm ore, since we have evidence that there is a relationship between a states minority population squared and the level of felon disenfranchisement we must ackno wledge the true relationship is in fact curvilinear and not linear. In fact, those results suggest that as a states minority population increases exponentially 87


the chances of a state with a high minority popu lation falling into the high severity felon disenfranchisement category decreases drama tically. Therefore, we have overwhelming evidence that the relationship be tween the size of the minority population and the severity of felon disenfranchisement laws is curvilinear. We originally hypothesized that states with high Black incarc eration rates would be more likely to adopt more severe felon disenfranchise ment laws. Our results, however, did not yield any evidence in support of this hypothesis. In actuality, our results reve aled the opposite: states with high Black incarceration rates had a 45% chance of adop ting a moderate-severity felon disenfranchisement law and a 38% chance of adop ting a low-severity felo n disenfranchisement law. This tells us that as the number of Blacks in prison relative to the number of Blacks in a states population increases, st ates are not as likely to adopt high-severity felon disenfranchisement laws. Further, if the target population of high-severity fel on-disenfranchisement laws is minorities in the population, it would make sense that states with a disproportionate number of minorities in prison relative to minorities in th e population would be more likely to adopt moderate-and low-severity felon-disenfranchisement laws. In e ssence, the target population, the number of minorities in the population, had decreased due to the high rate in which they were incarcerated within such states. Therefore, there no longer existed a need to target minorities in the population with high-severity felon-disenfranchi sement laws. In sum, this explanation seems even more plausible when we examine the results of the percent minorities in the population variable which supported the hypothesis that stat es with large minority populations would be more likely to adopt more severe felon-disenfranchisement laws. Again, if we assumed the 88


target of high-severity felon-disenfranchisement laws to be minorities in the population, the evidence provided thus far would support this theory. Referring back to group-threat th eory, particularly Blalocks racial-threat hypothesis, we could argue that states are usi ng felon-disenfranchisement laws to control minority groups in the population. The premise of the raci al threat hypothesis is that as the number of minorities in the population increases, the interests of the majority group become threatened. When this happens, the majority group uses social-control mechanisms, such as criminal laws, to subordinate the impending minority threat. Additionally, Preuhs (2001) stated, If th e variation in felon disenfranchisement laws is due to racial politics, one would also expect these laws to be more severe when the targeted population is compos ed of a higher number of minorities than non minorities (739). The evidence provided thus far s hould lead one to think th at the racial threat hypothesis is supported by these findings. Likewise, Preuhs (2001) presented an alternate to the racial threat hypothesis that suggested that minority threat: fi rst, increases as the minority popul ation increases then decreases as the minority population increa ses exponentially. Our findings confirmed that as a states minority population increased expon entially, the odds of a state a dopting a high-severity felon disenfranchisement law was unlikely. Since our findings confirmed this hypothesis it gave me an opportunity to examine possible expl anations for such a relationship. Another variant of group-threat theory is the political-threat hypothesi s, which states that growing minority populations constit ute a threat to the majority population. This is partially because this growing minority group would be able to leverage democratic political institutions to their advantage (Behrens and Manza 2003). To combat this threat, the majority group, through its government, enacts policies that aim to undermine the interests of minority citizens, 89


in terms of either participatory rights or the di stribution of policy benef its (Preuhs 2001). We would argue here that felon-dise nfranchisement laws were used to combat the threat that a growing minority population had on the political interest s of the majority population. Further, such states adopted high-severity felon-disenfranchisement laws specifica lly to impact minoritygroup participatory rights. If this group were a political threat to the majority group, it would be in this groups best interest to limit the political participation of the minority group. Therefore, it was perfectly logical for states to adopt highseverity felon-disenfra nchisement laws that restricted access to participation in government most harshly through its most basic civic duty: voting. If you take away a minority groups voice in government, it is no longer a threat to the majority groups political interests. In contrast, the economic aspect of the group-th reat theory, posits th at competition between Whites and Black for jobs and other economic resources results in an increase in social control. I hypothesized that states with a high percentage of idle White males would be likely to adopt more severe felon-disenfranchisement laws Our findings did not support this hypothesis because increases in a states idle White male population were associated with such states preferring to adopt the lowor moderate-severity felon disenfra nchisement laws. If you follow our original logic, this would mean that as the number of idle White males in the population increases, you would expect to see increased competition between Blacks and Whites for economic resources. However, our findings would lead us to believe that this increase in competition would not result in imprisonment a nd other punishments to control the excess supply of labor (Jacobs and Carm ichael 2004). On the other hand, we are not convinced that the percentage of idle White males was the best variable to operationalize the economic threat hypothesis. If we were to repeat this study, we would use White employment vs. Black 90


employment as a ratio. Using the ratio would al low us to clearly identify the differences in employment levels between these groups as th e source of conflict/thre at, as opposed to only examining the employment status of one group. Lastly, to answer our original and overarchi ng question, does race explain state adoption of high-severity felon-disenfranchise ment laws? The answer is yes. The evidence provided above overwhelmingly supports the racial explanat ions for state adopti on of severe felondisenfranchisement laws. Therefore, we cannot dismiss race as a factor in a states decision to adopt a highvs. a moderateor lowimpact fel on disenfranchisement law. In particular, we must pay attention to the effect that an in creasing minority population has on state lawmakers decision to adopt high-severity felon-disenfranc hisement laws. Clearly, evidence supports the idea that such policies have a greate r effect on a states minority population. Culture Models In order to test the effects of the individual categories of th e political culture variable, I created dummy variables for the moralistic and trad itionalistic states with individualistic states acting as the base category. As such, our analys is revealed that when including these individual political culture variables in this model, the tr aditionalistic category was highly collinear with the Confederate variable, p=0.772. With such a high level of collinear ity, I created an interaction variable that included the traditionalistic categor y and the former Confederate state status. To test whether, these two variable s were essentially measuring the same concept I ran this model twice excluding the traditionali stic category and including the C onfederate state status then including the Confederate state status and excluding the traditi onalistic category. What I found was the findings for the model that excluded the traditionalistic category variable were identical to the model with the newly formed interaction te rm therefore confirming that the traditionalistic variable and the Confederate variable were esse ntially measuring the same concept. For this 91


reason, I decided to omit the traditionalistic vari able from the model and use the Confederate status as a proxy measurement for both concepts. Our findings revealed that state political cu lture and religious fundamentalism played an important role in the level of severity of felon di senfranchisement laws adopted within a state. It was found that traditionalistic st ates and states with a high re ligious fundamentalist population were more likely to adopt high-severity felon dise nfranchisement law. We would argue here that both types of variables portray th e patriotism within a state (adhe rence to their states culture) and the impact of a populati on with strong adherence to th eir religious beliefs. More specifically, these variables addressed the traditiona l-values aspect of a st ates culture with the exception of the moralistic states classification. In another way, the findings for moralistic states suggest that such states we re less likely to be observed in the high severity felon disenfranchisement law category which would also support our hypothesis of traditionalistic states being more likely to fall into the high se verity categories of felon disenfranchisement. One could argue that a states classification as a former Confederate state refers to the patriotism aspect of traditional values and the magnitude of the religious fundamentalist population would also point to trad itionalistic beliefs. Likewise a states classification as a traditionalistic state refers to reverence for the past aspect of a states traditional-values. Regardless, these three variables represent a states ability to sustain values over time despite criticism or modernism. As such, we argue th at states with large fundamentalist populations would be more likely to adopt high-severity felon-disenfranc hisement laws. Further, we hypothesized that traditionalis tic states and former Confederate states would also be more likely to adopt high-severity felon-disenfranchisement laws. Overall, our results support these hypotheses and we conclude that states with large religious fundame ntalist populations, or 92


classified as having a traditionalist political cultur e, or which are former Confederate states were more likely to adopt high-severity felon-disenfra nchisement laws vs. the combined moderate and low levels. Let us begin our discussion with potential explanations for former Confederate states adopting high-severity felon-disenf ranchisement laws. First, we cannot ignore that Confederate states were married to the idea of continuing the impor tation of slaves into the United States. Due to the economic aspects of slavery, Confederat e states were very conc erned about the effect banning this practice would have on its economy, but they were also concerned about how ending the slave trade would leave Whites left to compete with slaves. Accordingly, Spratt, (1861) in a letter addressed to Hon. Mr. Perkin of Louisiana, openly criticized the provisional constitution adopted by the Southern Congress at Montgomery, Alabama. Spratt (1861) clearly states it would not be in any sout hern states interest to ban the sl ave trade. Specifically, he said, In that State [Virginia] there are about 500,000 slaves to about 1,000,000 Whites; and as at least as many slaves as masters are necessary to the constitution of slave society, about 500,000 of the White population are in legitimate relation to the slaves, and the rest are in excess. Like an excess of alkali or acid in chemical experiments, they are unfixed in the social compound. Without legitimate connection with the slave, they are in competition with him. (Spratt 1861) Clearly, Spratt (1861) expresses th e fears of many residing in Confederate states during this period. Furthermore, slavery was a traditional value th at Confederate states were willing to go to war to protect. This is why many southern states many of them former Confederate states continued a sort of de facto slavery in the form of indentured se rvitude or debt slavery after slavery was abolished in 1865. As mentioned previously, a jar fo rever retains the odor of that which it first housed, which would mean that these very traditional values based on a system of Negro slavery still to this day remain in thes e states. Therefore, it would make sense for 93


former Confederate states to adopt high-severity disenfranchisement laws. If Confederate states had a legitimate fear of being in competition with the slaves, it is my contention that this underlying fear has transcended th e many generations that have ar isen since the era during which these states were part of the Confederacy. This underlying fear has surv ived as a part of the traditions of former Confederate states. Again, here we are touching on the deep-rooted racial motivations for the adoption of severe criminal -justice policies that disproportionately impact minority segments of the population. On another note, the above argument gives stre ngth to the group-threat theory previously discussed. These findings give us further reason to believe th at we cannot ignore the racial implications imbedded in the adoption of high-seve rity felon-disenfranchisement laws. In terms of religious fundamentalism, we must admit that the racial elemen ts of the religious fundamentalism component were not as clear as those presented above. This relationship was more closely linked to the overall culture of a state. Political Model I used the level of professionalism of a state s legislature, state demo cracy score, political competition, and the presence of a Republican governor as the political variables in this model. I originally ran this model using an ordered logit model; however, the results of the Brant test of parallel regression assumption i ndicated that the ordered logi t model using these variables violated the parallel regressi on assumption. Therefore, I used gologit2 (Williams 2005), which relaxes the constraints set forth by the ordere d logit model such as the parallel regression assumption. Our findings suggest that states with highly pr ofessionalized legislatur es were less likely to adopt high-severity felon disenfranchisement laws vs. moderate and low severity categories. Similarly, states classified as highly democratic would also be less likely to adopt the high94


severity category of felon disenfranchisement laws vs. lo w and moderate categories. Conversely, states with high leve ls of political competition were more likely to adopt the high and moderate categories vs. the low category of felon disenfranchisement laws. In addition, states with a Republican governor were more lik ely to be observed in the low and moderateseverity categories vs. th e high-severity categories of felon di senfranchisement. Overall, we can conclude that states wi th highly professional legi slatures, classified as being highly democratic, or have a Republican governor w ould be less likely to adopt the high severity category of felon disenfranchisement law. Conversely, we conclu de that states with hi gh levels of political competition would be more likely to adopt a high-severity felon disenfranchisement law. I hypothesized that states with highly professionalized legisl atures would adopt less severe felon-disenfranchisement laws because they woul d have more time to research and deliberate issues before making a decision. Our findings did support this hypothesis in that the more professional a states legislature was, the less likely it wa s to adopt a high-severity felon disenfranchisement law. We also assume that a more professionalized le gislature is also highly experienced and can therefore be mo re deliberative in their decision -making process. If that is the case, such states would take the time to research potential impacts of adopting such legislation and have more of an opportunity to ma ke an informed decision. In this case, a state like California whose legislature most resemble s Congress would be less likely to adopt a highseverity disenfranchisement law. According to our results, highly democratic st ates are less likely to adopt high-severity felon disenfranchisement laws. I would like to th ink that this relationship exists because such states have a fundamental belief in democracy and as such preserve democracy through their legislative decision-making process. Regardle ss, we cannot discount the implications these 95


findings have on our study. These findings confirm some of my original suspicions about this legislation, namely that its adop tion is a threat to democracy. Given that, it would make sense that more democratic states, if they adopt this legislation, would defer to adopting those with the least severity as they are less of a threat to democracy. Additionally, the high level of political compe titiveness in a state makes it more likely to adopt high-severity felon-dise nfranchisement laws. I orig inally hypothesized that highly competitive states would be more likely to adopt high-severity felon disenfranchisement laws for various reasons. One, I assumed that in states with more political competition, legislators would adopt a tough on crime stance in order to please their constituents and th erefore adopt the most severe form of felon-disenfranchisement laws. Accordingly, such states were more likely to adopt high-severity felon-disenfranchisement laws. Lastly, to the contrary of the hypothesized re lationship, states with a Republican governor were less likely to adopt high-seve rity felon-disenfranchisement laws. In general, Republicans tend to favor harsher criminaljustice policies and it would make sense that states with a Republican governor are likely to do the same. As the governor has to sign off on all legislation, it is plausible that he/she as a Republican (and do nt forget as a politicia n) will adopt legislation in line with his/her Republican values and legislation that will get him/her re-elected. However, that does not seem to be the case. In fact, th e opposite is true. Alt hough we tend to think of Republicans as being tough on crime these findings suggest that they are not as tough as everyone seems to think (at least when it comes to the voting rights of ex-felons). We can use Florida as an example here, who under a Republi can governor, Charlie Chri st, changed its felon disenfranchisement laws to allow ex-fe lons voting rights to be restored. 96


Class Model For a one percent increase in cl ass based turnout the odds of a state being observed in the high category of felon disenfranchisement law is 0.991 times lower than being observed in the combined moderate and low categories. Additio nally, for a one percent increase in a states unemployment rate the odds for adopting a highseverity felon disenfranchisement law was 1.084 times greater than the odds for the moderate and low categories. With such a value, we cannot clearly express such as states pref erence in severity category of the felon disenfranchisement law. Except for the findings for income inequality, the findings of the class model suggest no meaningful relationship exists between these independent variables and the level of disenfranchisement with in a state. Lastly, for states with high levels of income inequality (values closer to one ) the odds for adopting a high-seve rity felon disenfranchisement law were 4.402 x 1010 greater than the odds the m oderate and low categories Originally, I hypothesized that st ates with high levels of in come equality would adopt the most severe level of felon-dise nfranchisement law out of a fear of crime. Our findings support this hypothesis and lead us to conc lude that in areas where they are great disparities between the wealthy and the poor, states adopt punitive crimin al justice policies such as the high-severity category of felon disenfranchisement law. My or iginal thought was that in such situations the wealthy, in an effort to protect their wealth, support th e adoption of severe crime control policies to further distance the poor from themselves. Full Model The full model includes variables from the i ndividual models presented above that were found to have a more than negligible impact on th e level of severity of felon disenfranchisement adopted by a state. Examining the individual models separately, I chose variables with coefficients that were not close to zero, since thei r impact translates to a negligible effect on the 97


dependent variable. Overall, this model will allow us to test si multaneously for coefficients that produce a moderate to substantial effect on the level of felon di senfranchisement law adopted by a state. As a conclusion, all of the results of th is model confirm the relati onships revealed in the individual models above. This is important in that this model combines the important elements of all the other models plus the use of statistical control variables and the findings remain the same. Number of Collateral Consequences Since each state contributes up to three years wo rth of the same data we can almost assume that there will be correlati ons within state-years. The above is what is referred as panel data. Generalized estimating equations can be used fo r panel data because it is capable of modeling the within state-year co rrelation and still obtain unbiased estimates (Agresti 1996). Next, I had the task of choosing between a negative binomia l or poisson regression model. Making that decision required a check of the data for overdis persion in the dependent variable. The data revealed that the variance of the dependent vari able was not greater than the mean and there was not an excessive amount of zeros in the data (som e of the first signs of overdispersion). I used the STATA command nbvargr, which graphs observed, negative binomial, and poisson probabilities to examine the data visually to determine which model best fit the data. The graphical output revealed that the poisson regression model was the best choice for this data. Below I present the results of these models and their relationship to the number of collateral consequences adopted by a state (S ee Table 5.2 for statistical output). Race Models Our findings revealed that increases in a states Black incarceration rate, idle-White male population, and minority population result in an incr ease in the expected counts of the number of collateral consequences adopted by a state. Sp ecifically, the expected counts of collateral 98


consequences are expected to increase by 20% for a one percent increas e in a states Black incarceration rate, holding all other variables in the model constant. The percentage of idle White males in a states population did not have a substantial effect on the number of collateral consequences adopted in a state: a 0.6% decrease for a one percen t increase in the idle White male population. We also found support for the alte rnative minority threat hypothesis in that as the minority population increases exponentially the num ber of collateral consequences is expected to decrease by 98%. We hypothesized that states with high Black incarceration rates would adopt a greater number of felon collateral consequences. Our findings revealed that as a states Black incarceration rate increase d, the state was more likely to experi ence an increase in its number of collateral consequences. Clearly, this identifies a positive relationship that would warrant further explanation. What possi ble reasons could justify such a relationship? We also must acknowledge a racial element in this relationship. As such, I would theorize that again if the target population is minorities, as these consequences apply to those no longer incarcerated but those re-enter ing the population, it would follow that states adopt a high number of collateral consequences. However, thes e states are incarcerati ng minorities at a rate disproportionate to their size in the population. What could be the reason for adopting a large number of collateral consequences when minorities are incarcerated at a ra te far surpassing their proportion of the population? If we use the group -threat theory to answer this question, we conclude that despite the high ra te at which minorities are inca rcerated, some threat remains when they re-enter society. Hence, the more mi norities you incarcerate, the more opportunity a state has to affect the power yi elded by a growing minority populat ion. Further, adopting a large 99

PAGE 100

number of collateral consequences allows states to cast a wider ne t and increase the possibility of affecting a vast number of minorities in the population. The above explanation also applies to the perc entage of minorities in a states population. Again, we hypothesized that states with a larg e minority population would adopt a large number of collateral consequences to co mbat the threat posed by this groups increase in population. Our findings supported the alternative racial hypothesis and thereby suggest that as a states minority population increases exponentiall y, the number of collateral cons equences adopted at first increases then decreases over time Lastly, these results gave us more evidence in favor of a racial motivation for the adoption of collateral consequences. Culture Model Unlike our previous models, a st ates status as a former Conf ederate state and classification as a traditionalistic state were not correlate d. However, it was highl y collinear with the fundamentalism variable in this model, p= 0.772 and with the moralis tic variable, p=0.512. Therefore, I omitted the traditionalistic political cu lture variable from this model. Overall, the culture model suggests no substantial relationship exists between a states government ideology and its propensity to adopt a greater number of felon collateral consequence policies. However, the coefficients for religious funda mentalism, moralistic state political culture, and Confederate state would suggest that there is a re lationship between these variables and the number of collateral consequen ces adopted by a state. A one percent increase in a states religious fundamentalist population increases the e xpected number of coll ateral consequences by a factor of 1.898. Essentially, this finding would suggest that this variable has a moderate impact on the number of collateral consequences a stat e adopts by indicating that states with a high religious fundamentalist population would be more likely to adopt a grea ter number of collateral consequences. 100

PAGE 101

On the other hand, being a moralistic state decreases the expected number of collateral consequences by 27%. This suggests that moralistic states are less likely to adopt a high number of collateral consequences. Simila rly, the results for former Confederate states reveal identical finding by indicating that being a former Confeder ate state decreases the expected number of collateral consequences by 33%. Lastly, when we examine the time variables associated with this model we see the culture variables had the following impact on time: compared to 1986, the number of collateral conse quences adopted in 1996 and 2006 increased by 4% and 23% respectively. I hypothesized that former Confederate states would be more likely to adopt a higher number of felon collateral consequences, our findings confirmed the opposite: former Confederate states were less li ke to adopt a large number of collateral consequences. Additionally, as hypothesized, our findings rev ealed that states with large religious fundamentalist populations were more likely to adopt a greater number of collateral consequences. Again, these results speak to the traditionalist aspect of religious fundamentalist churches and their unwillingness to accept modernism. Over th is twenty year period, 1986 to 2006, as a states religious funda mentalist population grew so di d their likelihood to adopt a large number of felon collateral consequences. This suggest that although we saw a trend in states adopting fewer collater al consequences over time, st ates with large religious fundamentalist populations, like Alabamas, would continue to adopt felon collateral consequences at an accelerated pace. Political Model Only two of the four political variables had a modest impact on the number of collateral consequences. Specifically, for every one unit in crease in a states level of professionalism on Squires professionalism index, a st ates expected number of collate ral consequences is expect to 101

PAGE 102

increase by 26.7%. In contrast, a one unit incr ease in the level of political competition as measured with this index decreases the expect ed number of collateral consequences by 19%. Lastly, when comparing the effect of these vari ables on the number of co llateral consequences adopted over time, we found that the number of expected collateral consequences adopted in 2006 were expected to increase by 24.6% when compared to the number adopted in 1986. Overall, this suggests that stat es in this model were expected to have a greater number of collateral consequences as compared to 1986. In chapter three, I hypothesized that less pol itically competitive states would have a high number of collateral consequences. I composed this hypothesis based on the reasoning that less competitive states spend less on resources and pr ograms that would benefit the have nots in society. I would categorize the ex-felon population as belonging to the h ave nots and therefore not eligible to benefit from their states adopti on of fewer collateral consequences. Our findings support these theories, which state that competiti ve states, which the literature has shown to spend more on social-welfare programs, actually adopt fewer collateral consequences than their less competitive counterparts. Why would it benefit a highly competitive state to adopt fewer felon collateral consequences? Holbrook and Va n Dunk (1993) provide a logical explanation that suggests that elected officials in highly competitive states provide greater benefits and are more likely to support liberal policies than non-comp etitive states. Further, it is important to note that highly competitive states have greater levels of voter participation, thereby increasing the possibilities for lower socioeconomic interests to be heard. In that sense, it is more plausible for our findings to suggest that less competitive states would adopt a higher number of collateral consequences. If voter participation is lower in less competitive states then lower socioeconomic interests are 102

PAGE 103

not as likely to be heard, as re search shows this the group is less likely to turn out at the polls on election day. Similarly, I hypothesized that st ates with more professionali zed legislatures would adopt fewer collateral consequences for similar reason s mentioned above. Nonetheless, our findings do not support this hypothesis. Instead, our findings indicate that such states actually adopt a greater number of collateral conseq uences. I argued that states with more professionalized legislatures would have more time for the legislature to sit, which might lead to better decisionmaking. However, increases in the level of a states professionalism are associated with increases in the number of collateral consequences adoption. One could conclude that the extra time these legislatures have to sit allows them to deliberate over issues and decide it is in the states best interest to adopt a large number of collateral consequences. Overall, this finding suggests that states whose legisl atures more closely resemble th at of the United States Congress are more likely to adopt a greater nu mber of collateral consequences. Class Model The income inequality variable was the only va riable in this model to have a less than negligible effect on the number of collateral consequences adopted by a state. Our findings revealed that states that increased their level of income inequality by one unit increased the expected number of collateral consequences by a factor of 556. Lastly, we found that the number of expected collateral consequences adopted in 2006 was expected to increase by 17% when compared to the number adopted in 1986. As hypothesized our findings revealed that st ates with high levels of income inequality would be more likely to adopt a larger number of collate ral consequences. I originally thought it plausible for wealthier states to adopt a great number of collateral conseq uences to combat the threat posed by the criminal population making its re-entry into society. We cannot discount the 103

PAGE 104

effect that the level of income inequality has on the number of collateral consequences adopted by a state. More importantly the magnitude of the effect is greater than an y of the other variables tested thus far which would lead me to think that further investigation into the effects of class/economic variables on the number of coll ateral consequences adopted by a state is warranted. Full Model This full model contains all va riables found to have a more th an negligible effect on the number of collateral consequences adopted by a state. Additiona lly, I have omitted the control variables as the impact was negligible in this model as it was in previous models. Thus, this model only includes variables from the individual m odels with a substantial to moderate impact on the number of collateral consequences adopted by a state. For most of the variables in this model, the results were similar to those reported for the individual models above. However, the minority population and the income inequality variables were contrary to what was expected based on th e results of the individua l models. The findings for the full model concluded that the expected number of collatera l consequences would decrease as the minority population increased by one pe rcent but would increase as the minority population increased exponentially. Similarly, the findings for income inequality in the individual model suggested that st ates that increased their level of income inequality by one unit on the Gini index would be expected to increas e their number of collateral consequences by a factor of 556. However, the result s for the full model indicated in fa ct that states that increased their level of income inequality by one unit ba sed on the Gini index would be expected to decrease their expected number of colla teral consequences by a factor of 0.782. 104

PAGE 105

Severity of the Felon Welfare Ban As the categories of the felon welfare ban increase from one to four, the level of severity of felon welfare ban also increases. To recap, the levels of felon welfare ban were coded accordingly: 1= opt out; 2= adopt with treatment provision; 3=adopt with time limits/partial adoption; 4= adopt the legislati on without modifications. For a ll of these models STATA used category 4 as the base outcome. Accordingly, all findings and analysis are compared to the base category outcome of adoption without modifications (See Table 5-3 Race a nd culture models and Table 5-4 Political and class-based models). Further, I originally intended to use an ordered logit model for this analysis however, I found that some models violated the parallel regression assumption wh ich is an essential component of an ordered logit model. Wherever it was found that a model violated the parallel regression assumption I used the generalized ordered logit, gologit2 (Williams 2005), model that relaxes the condition for the parallel regression assumption to be met. Since the dependent variable consists of four cat egories, the gologit2 command out put contains three panels of coefficients with the fourth acting as a base category. Therefore panel 1 of the gologit2 contains coefficients for category 1 vs. 2+3+4; pane l 2 = 1+2 vs. 3+4; and panel 3=1+2+3 vs. 4. Race Models Our findings indicate that states with high Black incarceration rates are less likely to adopt high-severity felon welfare bans. The results fo r idle-White males on the other hand indicate the opposite. For this variable, we found that states, like Michigan, would be more likely to adopt high-severity felon welfare bans. We also found support for the altern ative racial threat hypothesis, which suggests that as a states minority population increases exponentially threat decreases and the majority group is less likely to a dopt severe criminal just ice policies. Lastly, a variable that I hypothesized to be very important in this model, percent minority of welfare rolls, 105

PAGE 106

resulted in coefficients that had a negligible im pact on the level of severi ty of felon welfare ban adopted by a state. Going back to our original hypotheses, we hypothesized that stat es with high Blackincarceration rates would be more likely to adop t more severe levels of felon-welfare ban policies. Our findings, however, cannot lead us to directly support this hypothesis because states overwhelmingly chose to adopt the least severe levels of felon-welfar e ban. Therefore, we cannot conclude that our findings support our hypothesis and instead concede that a states Black incarceration rate does not influence states to ad opt severe levels of the felon-welfare ban. Specifically, states that increased the number of idle White males in their population would be expected to adopt the high-seve rity felon-welfare ban policies. This would suggest that a rise in White male unemployment has an effect on the le vel of felon-welfare ban adopted by a state. If we revisit the economic component of the raci al-threat hypothesis, it suggests that competition between Blacks and Whites for jobs and other ec onomic resources results in increases in social control. Jacobs and Carmichael (2004) sugge st specifically that im prisonment and other punishments are used to control the supply of labor. Original ly, we hypothesized that if an increase occurs in the number of idle White males in the pop ulation, unemployment would rise and the competition for jobs and other economic resources would intensify. We further hypothesized the majority group would be threat ened because the minority unemployment rate was also on the rise (assuming White unemployment has a positive effect on minority unemployment). To combat this threat, we expect the majority group to use punishments to reduce the excessive labor supply. If the felon-welfare ban were used as a puni shment, we would expect this punishment to affect minorities seeking employme nt. Indirectly, the felon-welfare ban does affect the minority 106

PAGE 107

unemployed population. Specifically, in states where the idle White male population is increasing, there exists competition for jobs a nd economic resources. Ex-felons would be among the competitors. However, they have a subs tantially diminished chance of surviving the competition since most employers prefer not to hire someone with a criminal record. In other words, if the competition is already intense, em ployers have a larger pool of potential employees to draw from, of which the ex-f elons are the least desirable candi dates. The ex-felon, most of whom are minorities due to th e disproportionate imprisonment rates between minorities and nonminorities, are left with few opportunities for gain ful employment. Ex-felons in such situations who desire to become law-abiding citizens turn to welfare for assistance, only to be turned away because the state they live in chose to adopt the felon-welfare ban without modifications. The unmodified felon-welfare ba n requires that ex-felons conv icted of drug felonies be barred for life from receiving welfare benefits. Th is legislation makes it virtually impossible for an ex-felon drug offender to resi st the temptation of going back to the street corner to resume selling drugs. Private employers fear hiring them and the government denies them aid for subsistence, making their alternatives somewhat scarce. I must concede, however, that the implications of these findings are racial but al so indicative of the treat ment of ex-felons in general. States could adopt such legislation to target two specific se gments of the population: minorities and ex-felons. Additionally, similar to some of the previous findings for the percen t minority squared, the alternative racial hypothesis was su pported in this model. These findings conclude that as a states minority population increases exponentially, it is not likely to adopt high-severity felon welfare ban policies. Going back to the alternative minority threat hypothesis we can conclude 107

PAGE 108

that as a states population increa se exponentially the threat to the majority population decreases and results in less severe criminal justice policies, such as the felon welfare ban. What does this imply about the relationship be tween high-severity adoptions of the felonwelfare ban and race? The raci al-threat hypothesis states that the majority group often uses criminal-justice policies to mitigate the threat of a growing minority population. The above findings suggest the continually growing minority population may be a consta nt threat but states are not using the most severe categories of felon welfare ban to reduce the threat. You can also see evidence of this relations hip when you examine the percent minority in the population variable. If given the choice between opting ou t of the felon-welfare ban, adopting it with a treatment provi sion, or adopting it without modifi cations, states that increased their minority population by one percen t preferred to either opt-out of the ban entirely or adopt the ban with treatment provisions. However, wh en given the choice of a partial adoption or adopting without modifications, states preferre d the adopt without modifications category. This would suggest that under these circumstances, the follo wing occurs: when faced with incremental increases in a states minority popu lation, states would prefer the adopt without modifications category relative to the partial adoption option; but, when faced with exponential growth under the same circumstances, states w ould overwhelmingly prefer the partial adoption category relative to adopting the felon-welfare ba n without modifications. This tells us that when states experience growth in their minority populations, regardless of the pace of growth; states generally prefer the less severe categories of the felon-welfare ban. However, when given the option between a partial adoption and adoptin g the ban without modifications, a state that experiences consistent increases (one percent increases) in its minority population would prefer 108

PAGE 109

to adopt the felon welfare ban without modifica tions; but, states that experience exponential growth would prefer the partial adoption category of the felon welfare ban. First, the findings revealed the percent mi nority on state welfare ro lls had virtually no effect on the level of felon-welfar e ban adopted in a state. This finding is in dire ct contrast to what was originally hypothesized. Arguing that severe felon-welfare ban policies were a product of race, we expected this variable to be very powerful in this model and were quite surprised to see that it had no effect on the level of felon-we lfare ban adopted by a state. We can conclude here that the percent minority on state welfare rolls is not an important fact or in a states decision to adopt harsh felon-welfare ban pol icies. Intuitively, that makes se nse, if this legislation is put in place to bar ex-felons from receiving welfare benefits the number of minorities a state has on its welfare rolls should have no effect since th is population would not have been eligible for welfare anyway. Culture Model To start, the political culture variables reveal ed there is truly a difference in felon welfare ban adoption between moralistic, traditionalist, and individualistic states. As such, the findings revealed that when comparing the opt-out, treatment, and partial categories with the adopt without modifications categ ories moralistic states tended to fall into any category other than the adopt without modificati ons category. On the other side traditionalistic states, when comparing the opt-out category with the adop t without modifications categories were more likely to choose the adopt without modifications category. But when comparing the treatment and partial adoption categories with the adopt w ithout modifications categories such states were more likely to adopt the less se vere categories of treatm ent or partial adoption. Additionally, since individualistic st ates were the referent group for the political culture variables we can say that the findings above compare both moralistic states to individualistic states and 109

PAGE 110

traditionalistic states with individu al states. As such, we can safe ly conclude that the differences between these categorizations of political culture do translate to variation in the adoption of the felon welfare ban legislation. The results for government ideology, religious fundamentalism, and former Confederate state are not as uniform. Wh en comparing the opt-out cat egory with the adopt without modifications category for former Confederate stat es we found that such st ates were more likely to choose the adopt without modifications category. However, when you compare the treatment or partial categorie s with the adopt without modifi cations categories we find that former Confederate states are more likely to choose any other category over the adopt without modifications category. Conversely, the inverse of the relationship pr esented above for the former Confederate states exists for the religious fundamentalist va riable. When comparing the opt-out category with the adopt without modifications categor y, we found that states that increased their religious fundamentalist populati ons by one percent were more likely to choose the opt-out category. On the other hand, when comparing the treatment and partial categories with the adopt without modifications cat egory, states that increased their religious fundamentalist populations were more likely to choose the adopt without modifications category. For both of these analyses we found that increases in a states religious fundamentalist population were associated with a 1.518 and 6.056 decrease in the multinomial log odds of the treatment category vs. the adoption without modifications category and the partial vs. the adopt without modifications category, respectively. Last, if a state were to increase its govern ment ideology by one unit, the multinomial logodds for treatment vs. adopt without modifica tion and the partial category vs. the adopt 110

PAGE 111

without modifications categories would be expected to increase by 0.030 and 0.033, respectively, holding all else equal. On the other hand, when you compare the opt-out with the adopt without modifications category we would expect the multinomial log-odds to decrease by 0.000 units (no effect) for a state that increases its government ideology. Overall, the relative risk for each of these categories vs. the adopt without modification categories are essentially equal to one that indicates that states are no more likely to fall into one category vs. the other. When examining the political culture and Confederate findings, the motivation for adoption of severe levels of the fe lon-welfare ban is in direct contrast. Moralistic states prefer the opt-out, treatment, and part ial categories to the adopt wit hout modifications category. On the other hand, traditionalistic and Confederate states prefer the adopt without modifications category relative to the opt-out cat egory; however, they prefer th e treatment and partial adoption choices over the adopt without modifications category. This would suggest that moralistic states when given the less severe alternatives to the more severe alternative were more likely to choose the less severe option; bu t traditionalistic states when given the same options would prefer the treatment and partial adoption category options only. These preferences are quite inte resting to examine in relation to the hypothesis set forth. Originally, I hypothesized that traditionalist states would be more lik ely to adopt the most severe levels of felon-welfare ban. The reason be hind this hypothesis was that Elazars (1966) characterization of traditionalistic states suggested that they pr efer to maintain the status quo without exerting much initiative (i nnovation). This applies to th e adoption of the felon-welfare ban in that these states would much prefer a b lind adoption of a piece of legislation coming from the federal government because they respect hierarchy and woul d prefer to exert as little initiative as is necessary. The federal govern ment gave states the opportunity to take the 111

PAGE 112

initiative and modify this legisl ation but traditionalistic states w ould have preferred to adopt the felon-welfare ban as is from the federal gover nment without any modifications. However, our findings do not lead to those conclusions. Instead, it would seem that traditionalistic stat es and former Confederate states prefer the option of some sort of sanction for these offenders but not the most severe. This is evidenced by the fact that when given the option to choose between opting out and adopting the ban without modifications such states chose to adopt it wi thout modifications because that was the only option with a sanction associated with it. On th e other hand when given options with less severe sanctions, such states preferred that option over the most severe sanction, thereby indicating that traditionalistic and Confederate states prefer to sanction drug offenders but do not necessarily believe in adopting the most severe sanction. Over all, the findings for th ese culture variables do not lead to any direct r acial explanations as much as they te nd to reflect a culture effect, absent any racial implications. When comparing the treatment and partial adoption cate gories with the adopt without modifications category, states the increased their religious fundamentalist populations by one percent were more likely to choose to adopt the felon-welfare ban w ithout modifications. However, in comparing the opt-out category wi th the adopt without modifications category for these variables, we found that states that in creased their religious f undamentalist populations by one percent preferred the opt-out category ov er the adopt without modifications category. This would suggest that when such a state is faced with choosing between the least severe and the most severe levels of the felon welfare ban, they choose to opt-out of the felon welfare ban. However, when faced with what we can term as th e moderate severity levels of the felon-welfare ban relative to the most severe level, they choose the most severe level. 112

PAGE 113

Originally, I hypothesized that st ates with large religious f undamentalist populations would be more likely to adopt high-severity felon-we lfare ban policies. Our findings would support these hypotheses in that when faced with the moderate severity levels of the felon welfare ban such states prefer to adopt the felon-welfare ban without modifications. Nevertheless, states faced with choosing between m oderate-severity felon-welfar e ban policies such as the treatment or partial adoption categories and a harsher categor y prefer the harsher category. However, when faced with categories that are opposite in their levels of severity, such states tend to adopt the harshest category available. Additionally, it was ea rlier argued that one concept of fundamentalist religions is their reliance on trad ition and unwillingness to accept modernism. As such, we contend the that adopti on of severe felon-welfare ban policies by states with large fundamentalism populations reflect their adherence to traditional beliefs in law and order and less belief in modern techniques such as modifying punishments. Overall, race cannot be explicitly linked as an explanation for these relationships. We could always argue that because the intended targets of these policies are overwhelmingly minorities, states with large fundamentalist populati ons tend to adopt the most severe levels of this policy as a way of maintaining traditionalist sentiment that existed in the past. However, I contend that the findings are not strong enough to deduce a clear-cut racial relationship between these variables and their adoption of severe felon-welfare ban policies. Political Model The results for a few of the variables in this model indicate that in creases in specific political variables were associated with a state being more likely to adop t a high-severity felon welfare ban policy. Particularly, if a state were to increase it s level of professionalism by one unit on the professionalism scale, the odds of that state adopting a high seve rity policy relative to a low severity policy would increase by 42%. A dditionally, if a state were to increase its 113

PAGE 114

democracy score by one unit on Hills (1994) state democracy scale, the odds of that state adopting a high vs. a low severity felon welfare ban would be expected to decrease by 15%. Lastly, for states that increase th eir political competition by one pe rcent, the odds of such states adopting a high severity ban vs. a low severity ban would be expected to decrease by 75%. Overall, this tells us that a state such as Califor nia, who is almost always ranked in the top five for states whose level of profe ssionalism most resembles that of the U.S. Congress, would be more likely to adopt the high severi ty felon welfare ban policy. Wh ereas, states such as Montana and Nevada which rank highest in terms of their democracy scores would be less likely to adopt the high-severity felon welfare ban policy. Likewise states such as Illino is and Michigan would also be less likely to adopt a high severity felon welfare ban policy due to the high levels of political competition with their states. We hypothesized that states with more libera l governments (states with higher democracy scores) would be less likel y to adopt severe felon-welfare ban policies. Our findings support this hypothesis and thereby suggest th at highly democratic states, through their tendency to have more inclusive democratic procedures, are less likely to support the adoption of high-severity felon welfare ban policies. If we think of a hi ghly democratic state as having more democratic inclusive procedures it is plausible that those states would not support a policy aimed at diminishing an individuals voting rights. Class Model The results of the class-based model also prov ide contrasting findings for state preference for adopting severe felon welfare ban laws. This model had three variables: class-based turnout, income inequality and unemployment rate. To st art, only the income inequality variable had a less than negligible impact on the level of felon welfare ba n adopted. As hypothesized, the 114

PAGE 115

findings suggested that a state that increased is level of income inequality by one unit on the Gini index would be more likely to adopt the highe r severity felon welf are ban policies. Full Model Similar to the previous full models, the felon we lfare ban full model, which included all of the variables of the individual models reporte d above, was unsuccessful. The major suspect behind this was multicollinearity. A few variables in the individual models were highly correlated with other variables in the other individual models. Per cent of idle White males in the population was highly correlated w ith a states unemployment ra te. The correlation between these two variables was expected since the idle White male m easure does involve some aspect the unemployment rate within a st ate. Therefore, when we merg e these individual models into the full model, we obtain estimates with ex tremely large standard errors, a sign of multicollinearity. Unfortunately, collinearity issu es were present in this model; thus, we could not report the findings. Change in Level of Felon Welfare Ban What specific states experienced changes in their level of felon we lfare ban between 1997 and 2006? In particular, I am looking to examine which states moved to a more severe category of felon welfare ban in 2006. I started with a ye ar after the felon welfar e ban was first adopted so that we could give the states that adopted this le gislation blindly some to figure out how such a decision would affect them. Coincidentally du ring the first year of this legislation, 1996, 17 states chose to deny benefits en tirely to the drug convicted ex-fe lons. However, in 1997 that number increased to 26 states. Further, I wished to examine state changes in the adoption of this legislation at a much later period, specifica lly ten years after th e initial passage. Most states that increased the severity leve l of their felon welfare ban policy chose to remain in the second to highest severity category, partial adoption. Specifically, the only state 115

PAGE 116

that changed to the highest level of severity fr om the lowest level was New York. Three states went from the opt out category in 1997 to the partial adoption category in 2006 (Connecticut, Michigan, and Oregon). Three other states went from the treatment category in 1997 to the partial category in 2006 (Colorado, Illin ois, Maryland). Last, I estimated a binary logit model to examine the relationship between the race-based, cultural, political, and classbased models and their likelihood to experience a ch ange in their level of felon welfare ban in 2006. I coded the change variable to reflect 0=no change or change to a more lenient policy category and 1=more severe policy category. Race Model For a one percent increase in a states Black in carceration rate, the odds of a state changing to a more severe felon welfare ban category vs. no change or a more lenient category would decrease by 45%. On the other hand, a one pe rcent increase in a states idle White male population was associated with a 17 8% increase in the odds of a st ate changing to a high severity felon welfare ban. Similarly, for a one percent increase in a states mi nority population, the odds of a state changing to a more severe felon welfare ban category would increase by 302%. Likewise, a one percent increase in the percentage of minorities on a states welfare rolls is associated with a 1.78% increase in the odds of a state changing to a high severity felon welfare ban category. Essentially, these results would suggest that states with high Black incarceration rates would be less likely to change their level of felon welfare ba n in the more severe direction in 2006. However, the results for the idle White male variable, percen t minority, and percent minority on state welfare rolls would suggest th e opposite. Overall these findings suggest that states with a large idle White male populati on, large minority population or large number of minorities on state welfare rolls would be more likely to change their felon welfare ban in 2006 in a more severe direction. 116

PAGE 117

To put it another way, the results show that a state with an average Black incarceration rate, idle White male population, percent minorit y on state welfare rolls, and minority population has the following probabilities: a 90% probability of keeping their original pol icy choice or adopting a more lenient choice and a 10% probabili ty of adopting a more severe policy category. Even if we compute these probabilities with an idle White male percenta ge of 6.3 (which is the highest level in this dataset) we are still left with a 52% probability of states experiencing no change or more lenient change and a probability of 42% in the more severe direction. Overall these results show us that a state is less likely to experience a more severe change in its level of felon welfare ban policy in 2006 based on the variab les in the race-based variable model. On another note, the one finding worth mentioning is the percent minority sq uared variable, this finding again confirmed the alte rnate minority threat hypothesi s by showing that exponential increase in a states minority population makes a st ate less likely to adopt a more severe felon welfare ban policy. Culture Model The results of the culture model revealed that states are less likely to experience a change in their level of felon welfare ba n vs. adopting a more severe policy. In general we can say that a state with a moralistic culture, an average government ideology sc ore, and an average religious fundamentalist population, has a 90% chance of experiencing no change or choosing a more lenient policy choice and a probability of 10% of adopting a more seve re felon welfare ban policy in 2006. Political Model In general we can say that a state with a more professional legislature, a Republican governor, average level of political competiti on, and average democracy score has a 86% probability of not changing their policy choi ce or adopting a more lenient policy and a 117

PAGE 118

probability of 14% of adopting a more severe le vel of felon welfare ban policy. Again, this confirms that most states are less likely to expe rience a change in their level of felon welfare ban when comparing 1997 with 2006. Further, these fi ndings indicate that incr eases in any of the culture variables are more likely to lead a stat e to experience a change in its level of felon welfare ban in 2006. However, if we compute th e predicted probabilities for a state with a high professionalism score, average democracy sc ore, high level of po litical competition, and a Republican governor the predicted probability of such a state experiencing no change or adopting a more lenient policy is 17% and a probab ility of 83% for a change to a more severe category. From these values, we can see that a states level of professionalism and political competition have the greatest impact on states ch ange in level of felon welfare ban policy to a more severe category in 2006. Class-based Model Only one of the three class-based variables had a substantial impact on the change in the level of felon welfare ba n policy, the income inequality indicato r. The findings for this variable are overwhelming in that the odds of a state that increase its level of income inequality by one level on the Gini index adopting a high severity fe lon welfare ban policy are expected to increase by a factor of 6.837 x 1013. This tells us that this variable has a large effect on a state adopting a more severe felon welfare ban policy in 2006. Sp ecifically, a state with an average class-based turnout, average unemployment rate, and high inco me inequality index has a probability of 61% of adopting a more lenient policy or keeping its original policy choice and a 39% probability of adopting a more severe fel on welfare ban policy in 2006. 118

PAGE 119

Table 5-1. Severity of felon disenfranchisement laws (2006) ordered logit models Independent variables: Race models Cult ure model Political model Class model Full model Model 1 Model 2 Panel 1 Panel 2 Ln Black incarceration rate -0.730 (1.171) -1.282 (1.250) -0.125 (1.940) Idle White male -0.408 (0.309) -0.641 (0.337) Minority population 3.707 (3.331) 25.378 (8.527) 10.000 (17.746) Minority population squared -48.345 (17.706) -17.811 (39.208) Political culture: Moralistic -1.607 (0.714)) -1.195 (1.151) Government ideology 0.002 (0.012)) Religious fundamentalism 1.756 (1.548) -0.658 (1.785) Confederate 0.769 (0.733) Professionalism 1.193 (2.932) -7.783 (6.344) -0.899 119(4.150) Democracy score -0.656 (0.399) 0.436 (0.383) Political competition 0.346 (1.745) 1.597 (2.022) -1.075 (2.001) Republican governor -0.197 (0.718) -0.009 (0.710) Class-based turnout -0.009 (0.008) Income Inequality 24.508 (15.218) Unemployment rate 0.081 (0.320) H.S. Graduation rate -14.414 (8.088)

PAGE 120

Table 5-1. Continued Independent variables: Race models Culture mode l Political model Class model Full model Model 1 Model 2 Panel 1 Panel 2 Total population -0.000 (0.000) Population density -0.001 (0.002) Incarceration rate 0.002 (0.004) Intercept 1 -7.762 -11.452 -0 .953 0.543 -1.144 7.548 -6.560 Intercept 2 -5.671 -9.021 1.499 9.618 -3.680 Overall model Ologit Ologit Ologit Gologit2 Ologit Ologit LR 2 5.73 16.34 15.44 12.89 5.67 21.95 p 0.126 0.003 0.004 0.116 0.129 0.025 N 50 50 50 49 50 48 Cronbachs Alpha 0.572 0.723 0.613 0.327 0.613 0.798 NOTE: Standard errors are reported in parentheses. Hawaii has been omitted from the race models due to its minority population being primarily Asian-Americans and Pacific Islanders whereas minorities in the rest of the United States are primarily African-American and Latino. The culture model and the class-based models exclude the District of Columbia as Elazar(1966) did not compute such a value for this jurisdiction and classbased turnout values are also not available for the District of Columbia. The District of Columbia and Nebraska are excluded from the political model and the full model as the District of Columbia does not have a separate legislature and Nebraska has a unicameral legislature. 120

PAGE 121

Table 5-2. Poisson regression model for the number of collateral consequences 1986, 1996, and 2006. Independent variables: Race models Culture model Political model Class model Full model Model 1 Model 2 Ln Black incarceration rate 0.183 (0.153) 0.220 (0.163) 0.373 (0.191) Idle White male -0.006 (0.039) -0.015 (0.040) Minority population 0.733 (0.594) 2.975 (1.607) -0.434 (0.349) Minority population squared -3.998 (2.747) 4.204 (8.248) Political culture: Moralistic -0.309 (0.182) -0.240 (0.220) Government ideology -0.002 (0.002) Religious fundamentalism 0.641 (0.445) 0.588 121(0.458) Confederate -0.407 (0.246) -0.427 (0.246) Professionalism 0.237 (0.576) 0.133 (0.622) Democracy score 0.080 (0.083) Political competition -0.209 (0.305) -0.288 (0.331) Republican governor 0.061 (0.096)

PAGE 122

Table 5-2. Continued Independent variables: Race models Culture model Political model Class model Full model Model 1 Model 2 Class-based turnout -0.001 (0.001) Income Inequality 6.322 (3.996) -0.246 (3.780) Unemployment rate -0.014 (0.046) 1996 -0.053 (0.146) -0.104 (0.153) 0.041 (0.117) 0.079 (0.115) -0.032 (0.168) -0.092 (0.169) 2006 0.011 (0.181) -0.069 (0.196) 0.209 (0.110) 0.220 (0.110) 0.157 (0.206) -0.146 (0.231) Intercept -0.705 -0.994 0.784 0.681 -1.821 Overall model Wald 2 7.44 9.20 12.24 6.20 9.44 13.97 p 0.190 0.163 0.057 0.402 0.093 0.234 122N Cronbachs alpha 148 0.432 148 0.577 150 0.563 147 0.248 150 0.654 147 0.582 NOTE: Standard errors are reported in parentheses. Hawaii has been omitted from the race models due to its minority population being primarily Asian-Americans and Pacific Islanders whereas minorities in the rest of the United States are primarily African-American and Lati no. The culture model and the class-based models exclude the District of Columbia as Elazar(1966) did not compute such a value for this jurisdiction and classbased turnout values are also not available for the District of Columbia. The District of Columbia and Nebraska are excluded from the political and full models as the District of Columbia does not have a separate legislature and data for Nebraska has a unicameral legislature.

PAGE 123

Table 5-3. Level of felon welfare ban 1997: Ordered logit models for race-based and cultural variables Independent variables: Race model 1 Race model 2 Culture model Panel 1 Panel 2 Panel 3 Panel 1 Panel 2 Panel 3 Opt out Treatment Partial Ln Black incarceration rate -0.303 (1.511) 0.170 (1.540) -1.785 (1.132) -0.088 (1.230) -1.549 (1.946) -1.830 (1.113) Idle White male -0.707 (0.613) 0.099 (0.456) 0.515 (0.306) -0.663 (0.630) 0.376 (0.505) 0.368 (0.332) Minority population -12.587 (6.169) -9.399 (5.940) -0.561 (0.306) -4.147 (17.525) -30.741 (27.656) 13.970 (21.245) Minority population squared -15.455 (30.531) 50.591 (59.387) -27.836 (47.864) % Minority on welfare rolls 0.065 (0.031) 0.060 (0.029) 0.006 (0.027) 0.056 (0.040) 0.062 (0.036) -0.011 (0.031) Political culture Moralistic 0.742 (0.965) 1.158 (0.974) 1.940 (1.555) Government ideology -0.000 (0.017) 0.030 (0.017) 0.033 (0.025) 123Religious fundamentalism 0.757 (2.004) -1.518 (2.172) -6.056 (3.955) Confederate -35.440 (3.42 e07) 0.734 (1.184) 3.452 (1.689) Intercept 4.886 -3.068 11.009 2.923 10.179 11.568 -1.372 -2.515 -3.504 Methodl Gologit2 Gologit2 Mlogit LR 2 15.07 18.29 15.68 p 0.237 0.248 0.206 N 49 49 50 Cronbachs alpha 0.537 0.679 0.536 NOTE: Standard errors are reported in parentheses. The base out come is the adopt without modi fications category. Additional ly, race models 1 and 2 exclude Maine as the Black incarceration rate was not available for 1997. The culture model excludes the Distri ct of Columbia as both political culture and the religious fundamentalist percentages were not available for this jurisdiction. The c ulture model also had to be estimated using a multinomial logit regression as it violated the parallel regression model and resulted in 39 i n-sample cases with a predicted probability of less than zero.

PAGE 124

Table 5-4. Level of felon welfare ban 1997: Ordered logit model for political and class-based variables Independent variables: Political model Class-based model Professionalism 0.350 (2.039) Democracy score -0.158 (0.269) Political competition -1.388 (1.371) Republican governor 0.003 (0.574) Class-based turnout -0.011 (0.005) Income Inequality 1.613 (15.527) Unemployment rate 0.228 (0. 308) 124 Intercept 1 -2.642 -1.998 Intercept 2 -1.518 -0.831 Intercept 3 -0.999 -0.292 Method ologit ologit LR 2 1.39 5.43 p 0.846 0.143 N 49 50 Cronbachs alpha 0.275 0.224 NOTE: Standard errors are reported in parentheses. The base outcome is the adopt without modifications category. The District of Columbia and Nebraska are excluded from the political model as the District of Columbia does not have a separate legislature and data for Nebraska has a unicameral legislature. For the class-based model, the District of Columbia is excluded because the data for class-based turnout in this jurisdiction was unavailable.

PAGE 125

Table 5-5. Change in the level of felon welfare ban (2006) binary logit models Independent variables: Race models Culture model Political model Class model Model 1 Model 2 Ln Black incarceration rate -0.594 (2.001) -0.822 (2.117) Idle White male 1.023 (0.564) 0.872 (0.586) Minority population 1.392 (6.260) 14.282 (22.465) Minority population squared -24.830 (45.047) % Minority on welfare rolls 0.018 (0.029) 0.006 (0.035) Political culture: Moralistic -0.065 (0.883) Government ideology 0.018 (0.019) 125Religious fundamentalism -4.554 (3.903) Confederate Professionalism 5.627 (3.250) Democracy score 0.034 (0.450) Political competition 3.973 (3.445) Republican governor 0.524 (0.988) Class-based turnout 0.015 (0.015) Income Inequality 31.856 (34.601) Unemployment rate 0.057 (0.438)

PAGE 126

Table 5-5. Continued Independent variables: Race models Culture model Political model Class model Model 1 Model 2 Intercept -3.127 -1.261 -2.139 -6.425 -13.147 Overall model LR 2 4.73 5.24 4.65 4.72 4.32 p 0.316 0.387 0.199 0.318 0.229 N 50 50 50 49 50 Cronbachs Alpha 0.591 0.731 0.537 0.262 0.627 NOTE: Standard errors are reported in parentheses. Hawaii has been omitted from the race models due to its minority population being primarily Asian-Americans and Pacific Islanders whereas minorities in the rest of the United States are primarily African-American and Latino. The Confederate variable was dropped from the binary logit analysis as it perfectly predicted failure (which was coded as 0 in this model). Consequently 11 observations were dropped from this model (all of the former Confederate states). Additionally, the culture model and the class-based models exclude the District of Columbia as Elazar (1966) did not compute such a value for this jurisdiction and class-based turnout values are also not available for the District of Columbia. The District of Columbia and Nebraska are excluded from the political model and the full model as the District of Columbia does not have a separate legislature and Nebraska has a unicameral legislature. 126

PAGE 127

CHAPTER 6 CONCLUSION: CAN WE EXPLAIN STA TE FELON-COLATERAL CONSEQUENCE POLICY ADOPTION? Why Should We Care? From the equality of rights springs identity of our highest interests; you cannot subvert your neighbors rights without striking a danger ous blow at your own.Charles Schurz An estimated 5 million citizens in the United States are barred from taking part in one of the rights of citizenship we hold most dear: the right to vote. As the population of the United States is estimated at 301,480,754,1 the 5.3 million-plus Americans affected by the collateral consequences of a felony conviction may seem insignificant. However, the issue goes beyond the numbers. It deals more with the preservation of democracy and civil rights within the United States. Furthermore, the impact of these collateral consequences affects minority populations at a significantly higher rate than the non-minority population. We should all be deeply concerned about this important issue. However, the public both fears and sets itself apart from the felon population. The public tends to picture all felons as violent offenders incapable of redemption, and draw such a bright-line distinction between felons and themselves that they do not equate inroads against a felons civil rights as being an inroad against their own. They tend to forget, a felony conviction may continue to deny these rights of citizenship decades after a sentence has been completed, even for a one-time, nonviolent offense (Mauer 2004). More importantly, they do not question whether felon collateral consequences, which act as additional punishments, may be a violation of every American citizens civil rights. America has engaged in many a war to preserve democracy. It has even gone so far as to engage in warfare to secure democracy for other nations as well. Then why is it so easy for Americans to justify blatant disregard for their own democracy when applying felon collateral 1 The United States Census Bureau, 127

PAGE 128

consequences? Felon collateral-consequence policies clearly threaten the principles of democracy upon which this country was founded. Alexis de Tocqueville, often cited when speaking of democracy, initially intended to study the prison system but instead his research became a study of American democracy. Thus, I am not alone in this endeavor. It seems when one tries to study criminal-justice issues, one is often confronted with democratic theory in American politics. The emphasis of any democratic nation is the right of individuals to elect representatives of their choice. Herein lies the root of the problem with felon collateral-consequence policies, particularly felon disenfranchisement. Felon dise nfranchisement hinders the individuals right to actively participate in government by electing representatives of his or her own choosing. As such, the loss of these rights presents a major obstacle to th e practical applications of democratic principles in our nation. Some may argue it is just to restrict felons rights while they are serving their sentences because this practice has been commonplace in our nation since before its founding. A common conception of collateral consequences is that they are a way of branding felons to make it known that they are not to enjoy the full rights of citizenship. Initially instituted for pragmatic reasons by ancient governments, over time felon disenfranchisement has become punitive, and in many states, lasts far beyond the terms of felons sentences and in some cases indefinitely. In effect, once someone becomes a felon, never again a citizen. How is this possible in a democratic nation? How is it possible for a large segment of the population to have lost their voting rights and be consequently barred from enjoying the full participation in the one democratic process we hold dear: the right to choose our representatives? By disenfranchising ex-felons, are we creating a subclass of United States citizenry? Why should we be concerned about doing this? History teaches us that ultimately, oppressive imbalances within democratic societies are corrected. In the United States, the American Revolution ended tax 128

PAGE 129

oppression, the Civil War ended slavery, the Womens Suffrage movement ended gender-based voter-disenfranchisement, and the Civil Rights movement ended Jim Crow. Democracy and civil rights go hand in hand; a conversation dealing with one will inevitably lead to the other. The Eighth Amendment to the Constitution specifically looks at the intersection of civil rights and punishment of criminal offenders. It specifica lly states, Excessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted (The United States Constitution). A strong case can be made that the adoption of felon collateral consequences, in general, are a violation of civil rights under the restrictions of the Eighth Amendment. Because collateral consequences are punishments put in place post-conviction, one can consider them cruel and unusual because the offender has already been punished, and in most cases has already paid restitution for his or her crime; however, they continue to be plagued by the collateral sanctions placed on them by the state in which they live. In essence, the ex-offender is being continually punished for the same crime. In some states, these extra punishments are short term; in others, they can last indefinitely. Further, it is clear that the implementation of these collateral consequences is retributive because in many instances they c ontinue post-conviction, often even after all other punishment or community-supervision requirements have been met. What Do We Know? Overall, our findings reveal overwhelming support for the racial explanations of felon collateral-consequence policy adoption. The implications of such findings is controversial in that we have evidence that justice is not applied equally, even at the tail end of ones punishment. There are numerous studies documenting the racial disparities of sentencing but few that take these studies even further by documenting post-conviction racial motivations for adopting the most severe collateral consequence policies. Various factors influence a states decision to adopt a specific felon collateral consequence or the level of severity of that consequence. Our main concern with this study was to show a strong 129

PAGE 130

racial motivation behind the adoption of this legislation. To that effect, of all the groups of variables tested, the racial variables had the strongest effect on the levels of severity of felon disenfranchisement, felon-welfare ban, and the number of collateral consequences adopted. Overall, we found that the percentage of minorities in a state had the effect of impacting the level of severity of felon disenfranchisement law adopted by a state. Further, it was found specifically that states with a high percentage of minorities in their population were more likely to adopt the highest severity felon disenfranchisement laws and a high-severity categories of the felon welfare ban. To do so would mean that those states would agree to impose a lifetime ban on the receipt of welfare and cash assistance for anyone who was convicted of a drug felony. Lastly, in terms of the number of collateral consequences adopted by a state, states with a high percentage of minorities in their population were more likely to adopt a greater number of collateral consequences. Based on the results for the other racial variables, we cannot conclude, as confidently as with did above, that those variables influence states to adopt the highest severity of all tested categories of collateral consequences. However, we cannot ignore their effects in each of these models individually either, but it is important to note the rate at which the magnitude of the minority population had an effect on every aspect of felon collateral-consequence policy studied in this work. To sum, of all the racial variables tested, we can conclude that states that experience increases in their minority populations are more likely to adopt the most severe felon disenfranchisement, felon-welfare ban, and a larger number of felon collateral-consequence policies. Further, this relationship is characterized by a curvilinear relationship as a states minority population increases exponentially. Specifically, as a states population begins to increase exponentially the likelihood of that state adopting severe policies decreases. This is not to say that the other race findings were not influential in these models but that the percentage of minorities in a states population had an effect in every model. This would mean that the number of minorities in a states population is a good indicator for the severity of felon collateral130

PAGE 131

consequence policies adopted. This finding in and of itself is telling of the importance of race in state criminal justice policy making. One other such variable would be the income inequality variable. This variable had a large and positive effect on all of the dependent variables. Such an effect cannot be ignored and we must therefore acknowledge that both race and class had a strong effect on these dependent variables. In both of these major race and class findings indicate that increases in a states minority population or increases in a states income inequality make a state more likely to adopt a large number of collateral consequences and high severity felon disenfranchisement and felon welfare ban policies. Implications: How Do Our Findings Improve the Study of Race, Class, and Criminal Justice Policies? We acknowledge that in terms of felon collateral-consequences policies, there is an obvious racial connection between state adoptions of these policies and the racial composition of their states. Further, we cannot ignore the possibility of racist motivations underlying both the creation and application of these policies throughout the fifty states and the District of Columbia. These findings overwhelmingly prove that race plays a significant role in state adoption of these types of criminaljustice policies. Therefore, we cannot ignore the idea that racist motivations do exist for the adoption of these policies. We can clearly see that the adop tion of these policies cannot be completely clear of some kind of bias. This bias, some may argue, could be against the felon population in general, but if that were the case the effect of the minority population would not have been so substantial. We must therefore assume, given our findings, that we cannot rule out racial bias. More importantly, these biases have infiltrated to the government, where politicians react by adopting policies to control their growing minority population. We must concede, although, that yet another underlying bias exists: that of the relationship between minorities and crime. To accept these findings, we have to be willing to assume that policy makers believe minorities commit more crime and their increase in the population could mean a rise in the states crime rate. Therefore, the impact 131

PAGE 132

of the findings of our study are two-fold, in that policy makers decision to adopt severe collateral consequences may be both a result of fear of crime produced by the minority population and fear of the minority population itself. These findings should urge researchers to pay closer attention to the relationship between race and these types of criminal justice policies. Further, it should urge the greater population to question the equal application of justice in our society as well as our democratic principles. Based on the fundamental rights that we, as citizens of the United States have granted to us by our nations most sacred document, the United States Constitution, we should be concerned that our rights are being diminished before our eyes. Felons, ex-felons, and minorities alike are still citizens of th e United States and therefore should be afforded the same rights as any free citizen once they have served their incarceration/ punishment period. It is our hope that our study provided the evidence nece ssary to garner support for the protection of felons, ex-felons, and minority civil rights because a threat to one citizens rights is a threat to all of us. But, let us not forget class. There has been a longstanding argument regarding criminal justice policies and their racial or class based effects. This research showed evidence in favor of both and we cannot simply ask the question is it race or is it class but must instead acknowledge that it is both. Differences in the makeup of a states minority population and the disparity between its rich and poor residents influence states to react more severely to felon collateral consequence policies. These findings would suggest an even greater problem than I had originally suspected with these policies: they are adopted with both a racial and a class-based bias. Future Research: Ways to Improve This Study First, one of the most obvious ways to improve our study is to include more alternative explanations for state adoption of severe felon collateral-consequence policies. By doing so, we may be better able to rule out alternative explanations in favor of our leading explanation: race. In 132

PAGE 133

addition, within the categories of alternative explanations we may get better results if we focused on variables that more closely operationalized what we were trying to measure. Second, I would be interested in knowing how the number of collateral consequences of a felony conviction varied by state. For the purpose of our study, we limited the number of collateral consequences. However, it would be interesting to test these explanations against the total number of collateral consequences adopted by a state. Last, the next step in this research would be to link the findings to possible 14th Amendment violations. We show clearly here that states are violating the rights of its minority and lower-income inhabitants by adopting these severe felon collateral consequence policies. Could our lawmakers be driven to investigate the findings presen t in an attempt to rectify possible 14th Amendment violations being carried out by states? One would hope that th is research and its future improvements continue along the path of ensuring equal justice in the United States. 133

PAGE 134

APPENDIX A BIVARIATE RELATIONSHIPS A-1. Sample of bivariate relationships: Severity of Felon Disenfranchisement model Black inc. % nonWhite % nonWhite2 Idle White males Confederate Moral Individual Traditional Religious Fund. Black inc. % non-White -0.6575 % non-White2 -0.6429 0.9185 Idle White males 0.0559 -0.3235 -0.3614 Confederate -0.2503 0.4017 0.2116 -0.2877 Moral 0.5336 -0.4957 -0.3217 0.0562 -0.3922 Individual -0.3780 0.2304 0.2115 0.0784 -0.3746 -0.5075 Traditional -0.1636 0.2727 0.1150 -0.1355 0.7727 -0.5075 -0.4848 Reg. fund. 0.0310 0.2330 0.0998 0.0636 0.3825 -0.3333 -0.1567 0.4950 Gov. ideology -0.1852 -0.0456 0.0055 0.2142 -0.2342 0.0536 0.0917 -0.1461 -0.4180 Competition 0.2641 -0.0980 -0.2179 0. 2350 0.0757 0.1253 -0.2667 0.1395 0.2276 Dem. score 0.1069 -0.2535 -0.1650 -0. 1110 -0.2641 0.3104 -0.0713 -0.2437 -0.6662 Profess -0.0487 0.1663 0.0773 0.2633 -0.2122 -0.0149 0.2482 -0.2331 -0.0186 Rep. Governor -0.1762 0.1881 0.1837 -0 .1263 0.0923 -0.1179 0.1911 -0.0714 0.1523 134Unemployment -0.1085 0.1377 -0.0567 0.6056 0.2358 -0.2103 -0.0506 0.2640 0.4075 Gini coefficient -0.3013 0.3907 0.1583 0.0578 0.5177 -0.5379 -0.0088 0.5548 0.2834 Turnout. 0.1388 -0.3205 -0.1016 -0.2585 -0.1054 0.4029 -0.2246 -0.1843 -0.2522 Total population -0.0064 0.1490 -0.0091 0.0769 0.1862 -0.1001 0.0197 0.0820 0.0775 Pop. density -0.3176 0.1490 0.0639 0.0248 -0.1107 -0.3519 0.5681 -0.2110 -0.2567 Incarceration rate 0.0785 0.4216 0.2154 -0.1398 0.5703 -0.4069 -0.1665 0.5796 0.5388 H.S. grad. rate 0.3271 -0.4610 -0.2538 -0.0789 -0.4092 0.4919 -0.1566 -0.3428 -0.3776 Gov. ideology Comp. Dem. score Profess Re p. Gov. Unemploy Gini Turnout Total Pop Gov. ideology Competition 0.1363 Dem. score 0.3589 0.0074 Profess 0.1335 0.0271 -0.0831 Rep. Governor -0.5706 -0.4019 -0.2887 0.0514 Unemployment -0.0867 0.3435 -0.2624 0.1780 0.0394 Gini coefficient 0.0181 0.1210 -0.2195 0.3334 0.1092 0.3609 Turnout. 0.1433 -0.1478 0.1538 -0.3091 -0.0197 -0.5114 -0.3525 Total population -0.1201 0.1502 -0.2071 0.7321 0.1374 0.1464 0.5261 -0.2464 Pop. density 0.3581 -0.2909 -0.0178 0.3089 0.0978 -0.0006 0.2796 -0.1954 0.1710 Incarceration Rate -0.4309 0.1300 -0.3182 0.0130 0.0974 0.3280 0.4876 -0.3867 0.2332 H.S. grad. rate -0.0303 -0.2052 0.1770 -0.2440 -0.0960 -0.3447 -0.5163 0.4318 -0.3440

PAGE 135

A-1. Continued Pop. density Inc. rate H.S. grad. Pop. density Inc. rate -0.2315 H.S. grad. -0.2572 -0.3747 135

PAGE 136

LIST OF REFERENCES American Civil Liberties Union. 2001. Ex-Of fenders and Voting Rights: Black State Legislators et al. v Moore. Miami, FL: American Civil Liberties Union. ature_courts/legal_departm ent/briefs_complaints/exfelons__voting_rights.cfm June 3, 2008. Allard, Patricia and Marc Mauer. 2000. Reg aining the Vote: An Assessment of Activity Relating to Felon Disenfranc hisement Laws. Washington, DC: The Sentencing Project. October 11, 2005. Allison, Paul D. 1984. Event History Analysis: Regression for Longitudinal Event Data. Sage University Paper series on Quantitative Applications in the Social Sciences 07-046. Beverly Hills, CA: Sage. American College of Physicians. 2001. Prime r on Statistical Significance and P Values. Effective Clinical Practice 4:183-185. Barber, Kathleen L. 2001. A Right to Representa tion :Proportional Systems for the Twenty-First Century. In Urban life and Urban Landscape series. Columbus, OH: Ohio State University Press. Barkan, Steven and Steven Cohn. May 2005. W hy Whites Favor Spending More Money to Fight Crime: The role of racial prejudice. Social Problems 52:300-314. Barrilleaux, Charles, Thomas Holbrook, and Laura Langer. 2002. Electoral Competition, Legislative Balance, and American State Welfare Policy. American Journal of Political Science 46:415-427. Beck, Allen and Christopher J. Mumola. Prisoners in 1998. Bureau of Justice Statistics Bulletin Washington, D.C.: U.S. Department of Justice, June 4, 2008. Beck, Allen J et al. 2000. Co rrectional Populations in the United States, 1997. Washington, D.C.: U.S. Department of Justice. Behrens, Angela, Christopher Uggen and Je ff Manza. 2003. Ballot Manipulation and the Menace of Negro Domination: Racial Threat and Felon Disenfranchisement in the United States, 1850-2002. American Journal of Sociology 109:559605. Berry, Frances S. 1994. Sizing Up State Policy Innovation Research. Policy Studies Journal 22:442-456. 136

PAGE 137

Berry, Frances S. and William D. Berry. 1990. Sta te Lottery Adoptions as Policy Innovations: An Event History Analysis. American Political Science Review 84:395-415. ---1992. Tax Innovation in the States: Ca pitalizing on Political Opportunity. American Journal of Political Science 36:715-742. ---1999. Innovation and Diffusion Mode ls in Policy Research. In Theories of the Policy Process, ed. Paul A. Sabatier. Boul der, CO: Westview Press. Berry, William D. Evan Ringquist, Richar d Fording, and Russel Hanson. 1998. Measuring Citizen and Government Ideology in the American States, 1960-93. American Journal of Political Science 42:327-348 -2007. "Replication data for: Measuring Ci tizen and Government Ideology in the American States, 1960-1993." hdl:1902.1/10383 UNF:3:5/2OriaMj8+VAbqLArvUQA== William D. Berry [Distributor], June 3, 2008. --2007.Updated & Revised Citizen and Government Ideology Measures through 2006. June 3, 2008. Bibby, John F. and Thomas M. Holbrook. 1999. Parties and Elections. Politics in the American States: A Comparative Analysis eds. Virginia Gray, Russell L. Hanson, and Herbert Jacob. Washington, D.C.: CQ Press. Blalock, Hubert M. Jr. 1967. Toward a Theory of Minority-Group Relations New York: Capricorn Books. Bobo, Lawrence and Vincent Hutchings. 1996. P erceptions of Racial Group Competition: Extending Blumer's Theory of Group Positi on to a Multiracial Social Context. American Sociological Review 61:951-972. Boehmke, Frederick J. and Richard Witmer. 2004. Disentangling Diffusion: The Effects of Social Learning and Economic Competiti on on State Policy Innova tion and Expansion. Political Research Quarterly 57:39-51. Boes, Stefan. 2006. goprobit A Stata Module to Estimate Generalized Ordered Probit Models software/bocbocode/s456603.htm June 3, 2008. Boire, Richard Glen. 2007. Life Sentences: Coll ateral Sanctions Associ ated with Marijuana Offenses. Center for Cognitive Liberty and Ethics. June 3, 2008. Borooah, Vani K. 2002. Logit and Probit: Ordered and Multinomial Models. Sage University Papers Series on Quantitative Appl ications in the Social Sciences 07-138. Thousand Oaks, CA: Sage. 137

PAGE 138

Box-Steffensmeier, Janet M. and Christopher J.W. Zorn. 2001. Duration Models and Proportional Hazards in Political Science. American Journal of Political Science 45:972988. Box-Steffensmeier, Janet M. and Bradford S. Jones. 1997. Time is of the Essence: Event History Models in Political Science. American Journal of Political Science 41:1414-1461. Brace, Paul. 1991. The Changing Cont ext of State Political Economy. Journal of Politics 53:297-317. Brace, Paul, Kevin Arceneaux, Martin Johnson, a nd Stacy G. Ulbig. 2002. Does State Political Ideology Change Over Time? Political Research Quarterly 57:529-540 Brown v Board of Education of Topeka, 347 U.S. 483 (1954). Buckler, Kevin G. and Lawrence F. Travis III. 2003. Reanalyzing the Pr evalence and Social Context of Collateral Consequence Statutes. Journal of Criminal Justice 31:435-453. Burton, Velmer S. Jr., Francis T. Cullen, and Lawrence F. Travis III. 1987. The Collateral Consequences of a Felony Conviction: A National Study of State Statutes. Federal Probation 51:52-60. Chiricos, Theodore and Gordon Waldo. 1975. Socioeconomic Status and Criminal Sentencing: An Empirical Assessment of a Conflict Proposition. American Sociological Review 40:753-772. Cnudde, Charles, and David McCrone. 1969. Party Competition and Welfare in the American States. American Political Science Review 63:858-866. Cummins, Eric. 1994. The Rise and Fall of the California Prison Movement Stanford, CA: Stanford University Press. D'Alessio Stewart J., David Eitle, and Lisa St olzenberg. 2005. The Impact of Serious Crime, Racial Threat, and Economic Inequality on Private Police Size. Social Science Research 34:267-282. Daley, Dorothy M. and James C. Garand. 2005. H orizontal Diffusion, Ve rtical Diffusion, and Internal Pressure in State Environmental Policymaking, 1989-1998. American Politics Research 33:615-644. Darwin, Charles. 1859. The Origin of the Species. ed.Gillian Beer, 1996. Oxford: Oxford University Press. Davis, Ronald. 2004. The History of Jim Crow. Northridge, CA: California State University. 9/14/07. 138

PAGE 139

Demos. Democracy Denied: The Racial History a nd Impact of Disenfranchisement Laws in the United States. New York, NY: Demos. June 4, 2008 Eitle, David, Stewart J. DAlessio, and Lisa Stolzenberg. 2002. Racial Threat and Social Control: A Test of the Political, Economic and Threat of Black Crime Hypothesis. Social Forces 81:557-576. Elazar, Daniel. 1966. American Federalism: A View From States New York: Thomas Y.Cromwell Company. Ewald, Alec. 2002. Civil Death: The Ideological Paradox of Criminal Disenfranchisement Law in the United States. University of Wisconsin Law Review 2002:1045-1137. Fairchild, Erika S. and Vincent J. Webb. 1985. The Politics of Crime and Criminal Justice. In Perspectives in Criminal Justice, ed. John A. Conley. Thousand Oaks, CA: Sage Publications. Feagin, Joe, and Hernan M. Vera. 2002. Old Wine in New Bottles: The Reality of Modern Racism. In The New Politics of Race: From DuBois to the 21st Century ed. Marlese Durr. Westport, CT: Praeger Publishers. Federico, Christopher. 2004. When do Welfare A ttitudes Become Racialized?: The Paradoxical Effect of Education. American Journal of Political Science 48:374-391. Fellner, Jamie. 1998. Losing the Vote: The Impact of Felony Disenfranchisement Laws in the United States. Washington, D.C.: The Sentencing Project. October 11, 2005. Fording, Richard C. 1997. The Conditional Effect of Violence as a Political Tactic: Mass Insurgency, Welfare Generosity, and Elector al Context in the American States. American Journal of Political Science 41:1. Frant, H. 1991. Specifying a Mode l of State Policy Innovation. American Political Science Review 85:571-79. Friedman, Anne Kiehl. 2005. Vot er Disenfranchisement and Policy Toward Election Reforms. Review of Policy Research 22:787-810. Garver, Kenneth and Bettylee Garver. 1991. E ugenics: Past, Present, and Future. American Journal of Human Genetics 49:1109-1118. Giles, Michael and Melanie Buckner. 1993. David Duke and Black Threat: An Old Hypothesis Revisited. The Journal of Politics 53:702-713. 139

PAGE 140

Gilens, Martin. 1996. 'Race Coding' and White Opposition to Welfare. American Political Science Review 90:593-604. --1999. Why Americans Hate Welfare: Race, Medi a, and the Politics of Antipoverty Policy Chicago: University of Chicago. Gill, Jeff. 2001. Generalized Linear Models: A Unified Approach In Sage University Papers Series on Quantitative Applications in th e Social Sciences, 01-134. Thousand Oaks, CA: Sage. Glick, Henry R. and Scott P. Hays. 1991. Innova tion and Reinvention in State Policymaking: Theory and the Evolution of Living Will Laws. Journal of Politics 53:835-850. Gray, Virginia. 1973. Innovation in the States: A Diffusion Study. American Political Science Review 67:1174-1185. Green, Eva, Christian Staerkle, and David Sears. 2006. Symbolic Racism and Whites Attitudes Towards Punitive and Preventive Crime Policies. Law and Human Behavior 30: 435-454. Grossback, Lawrence J, Sean Nicholson-Cro tty, and David M. Peterson. 2004. Ideology and Learning in Policy Diffusion. American Politics Research 32: 521-545. Grupp, Fred. and Alan R. Richards. 1975. Variations in Elite Perceptions of American States as Referents for Public Policy Making. American Political Science Review 69:850-858. Hagan, John and Celesta Albonetti. 1982. Race, Cl ass, Perception of Criminal Justice in America. American Journal of Sociology 88:329-355. Harrison, Paige and Allen Beck. 2005. Prisoners in 2004. Bureau of Justice Statistics Bulletin Washington, D.C.: U.S. Department of Justice. June 4, 2008. Harvey, Alice E. 1994. Ex-Felon Disenfranchisement and Its Influence on the Black Vote: The Need for a Second Look. University of Pennsylvania Law Review 142:1145. Hays, Scott. P and Henry R. Glick 1997. The Role of Agenda Setting in Policy Innovation-an Event History Analysis of Living-Will Laws. American Politics Quarterly 25:497-516. Hench, Virginia E. 1998. The Death of Voti ng Rights: The Legal Disenfranchisement of Minority Voters. Case Western Law Review 48:727-798. Hero, Rodney. 1998. Faces of Inequality New York: Oxford University Press. Hill, Kim Q: 1994. Democracy in the Fifty States Lincoln, NE: University of Nebraska Press. Hill, Kim Q. and Jan E. Leighley. 1992. The Po licy Consequences of Class Bias in State Electorates. American Journal of Political Science 36:351-365. 140

PAGE 141

Holbrook, Thomas M. and Stephen L. Percy. 1992. Exploring Variations in State Laws Providing Protections for Persons with Disabilities. The Western Political Quarterly 45:201-220. Holbrook, Thomas M. and Emily Van Dunk. 1993. Electoral Competition in American States. American Political Science Review 87:955-962. Holmes, Malcolm. 2000. Minority Threat and Poli ce Brutality: Determinants of Civil Rights Criminal Complaints in U.S. Municipalities. Criminology 38:343-368. Hubbel, Webb. 2001. The Mark of Cain. The San Francisco Chronicle June 10, 2001. San Franchisco, CA: The San Francisco Chronicle June 11, 2008. Huberman, Michael and Ma tthew B. Miles. 2002. The Qualitative Researcher's Companion Thousand Oaks, CA: Sage Publications. Hurwitz, John, and Mark Peffley. 1997. Public Pe rceptions of Race and Crime: The Role of Racial Stereotypes. American Journal of Political Science 41:375-401. Itzkowitz, Howard and Lauren Oldak. 1973. Restoring the Ex-Offenders Right to Vote: Background and Developments. American Criminal Law Review 11:721-770. Jacobs, Herbert and Virginia Gray. 1966. "Parties and Elections" In Politics in the American States 6th ed. Washington D.C.: CQ Press. Jacobs, David and Jason T. Carmichael. 2001. T he Politics of Punishment Across Time and Space: A Pooled Time-Series Analysis of Imprisonment Rates. Social Forces 80:61-89. ---Jacobs, David and Jason T. Carmichael. 2004. Ideology, Social Threat, and the Death Sentence: Capital Sentences across Time and Space. Social Forces 83:249-278. Jacobs, David, Jason T. Carmichael, and Stepha nie Kent. 2005. Vigilant ism, Current Racial Threat, and Death Sentences. American Sociological Review 70(4): 656-677. Jacobs D., Helms R. June 1999. Collective Outbur sts, Politics, and Punitive resources: Toward a Political Sociology of Spending on Social Control. Social Forces 77:1497-1523. Jamieson, Kathleen. H. 1992. Dirty Politics: Deception, Distraction, and Democracy New York: Oxford University Press. Jim Crow Laws: Florida, crow/insidesouth.cgi?state=Florida August 17, 2007. Keech, William R. 1968. The Impact of Negro Voting: The Role of the Vote in the Quest for Equality Chicago: Rand McNally. 141

PAGE 142

Kent, Stephanie L. and David Jacobs. Aug 2005. Minority Threat and Police Strength from 1980 to 2000: A Fixed-Effects Analysis of Non linear and Interactive Effects in Large US Cities. Criminology 43:731-760. Key, Valdimer O. 1949. Southern Politics in State and Nation New York: Vintage. LaFollette, Hugh. 2005. Collateral Consequences of Punishment: Civil Penalties Accompanying Formal Punishment. Journal of Applied Philosophy 22:241-261. Lee, Matthew. 2000. Concentrated Poverty, Race, and Homicide. The Sociological Quarterly 41:189-206. Legal Action Center. 2001. Whats the La w. New York, NY: Legal Action Center. setting_the_rec ord_straight.pdf June 4, 2008. Levi, Robin and Judith Appel. 2003. Collateral C onsequences: Denial of Basic Social Services Based Upon Drug Use. New York, NY: Drug Policy Alliance. stincarceration_abuses_memo.pdf June 4, 2008. Lewis-Beck, Michael. 1977. The Relative Importa nce of Socioeconomic and Political Variables for Public Policy. American Political Science Association 71:559-566. Liao, Tim F. 1994. Interpreting Probability Models: Logit, Probit, and Other Generalized Linear Models. Sage University Paper series on Quantitative Applications in the Social Sciences 07-101. Thousand Oaks, CA: Sage. Liska, Allen, Joseph Lawrence, and Michael Benson. 1981. Perspectives on the Legal Order: The Capacity for Social Control. American Journal of Sociology 87:413-426. Liska, Allen Mitchell Chamblin, and Mark Reed. 1984. Testing the Economic Production and Conflict Models of Crime Control. Social Forces 64:119-38. Manza, Jeff, Clem Brooks and Christopher Ugge n. 2004. Civil Death or Civil Rights? Public Attitudes Towards Felon Disenfranchisement in the United States. Public Opinion Quarterly 68:275-286. Manza, Jeff and Christopher Uggen. 2006. Locked Out: Felon Disenfranchisement and American Democracy New York: Oxford University Press. Mauer, Marc. 2003. Invisible Punishment: Bloc k Housing, Education, Voting Policies Called Irrational, Counterproductive. Forces May/June:3-4. --Winter 2004. Felon Disenfranchisement: A Policy Whose Time Has Passed? Human Rights Magazine 31:16-17. --1999. Race to Incarcerate. New York: New Press. 142

PAGE 143

McConahay, John and Joseph Hough. 1976. Symbolic Racism. Journal of Social Issues 32:23-46. McElrath, Jessica. The Black Codes of 1865. New York, NY: About, Inc. /od/Blackcodes/a/Blackcodes1865.htm September 12, 2007. McGeehan, John and Morris Gall. 2007. Lets Review: U.S. History and Government 4th ed. Barrons Educational Series, Inc. Menzel, Donald C. and Irwin Feller. 1977. Leade rship and Interaction Pa tterns in Diffusion of Innovations Among American States. Western Political Quarterly 30:528-536. Mintrom, Michael. 1997. Policy Entrep reneurs and the Diffusion of Innovation. American Journal of Political Science 41:738-770. Mooney, Christopher Z. 2001. Modeling Region al Effects on State Policy Diffusion. Political Research Quarterly 54:103-24. Mooney, Christopher Z. and Mei-Hsien Lee. 1995. L egislating Morality in the American States: The Case of Pre-Roe Abortion Regulation Reform. American Journal of Political Science 39:599-627. ---2001. The Temporal Diffusion of Morality Policy: The Case of Death Penalty Legislation in the American States. In The Public Clash of Private Values New York: Chatham House Publishers. Morgan, David, and Sheilah Watson. 1991. Political Culture, Political System Characteristics, and Public Policies Among the American States. The Journal of Federalism 21:31-48. Myers, Martha A. 1990. Black Threat a nd Incarceration in Postbellum Georgia. Social Forces 69:373-393. Olivares, Kathleen M. and Velmer S. Burton. 1996. The Collateral Consequences of a Felony Conviction: A National Study of St ate Legal Codes 10 Years Later. Federal Probation 60:10-17. Peffley, Mark and Jon Hurwitz. 2002. The Raci al Components of RaceNeutral Crime Policy Attitudes. Political Psychology 23:59-75. Plessy v. Ferguson, 163 U.S. 537 (1896). Polsby, Nelson W. 1975. Legislatures. In Governmental Institutions and Processes Ed. Fred L. Greenstein and Nelson W. Po lsby. Reading, MA:Addison-Wesley. Preuhs, Robert R. 2001. State Fe lon Disenfranchisement Policy. Social Science Quarterly 82:733-748. 143

PAGE 144

Quinn, Bernard, Herman Anderson, Martin Br adly, Paul Goetting, and Peggy Shriver. 1982. Churches and Church Membership in the United States 1980 Atlanta:Glenmary Research Center. Radcliff, Benjamin and Martin Saiz. 1995. Race, Turnout, and Public Policy in the American States. Political Research Quarterly. 48:775-794. Rogers, Everett. 1995. Diffusion of Innovation New York: The Free Press. Rosenson, Beth. 2005. Shadowlands of Conduct: Et hics and State Politics Washington, D.C.: Georgetown University Press. Savage, Robert L. 1978. Policy Innovativen ess as a Trait of American States. The Journal of Politics 40:212-224. Schram, Sanford R., Joe Soss, and Richard C. Fording. 2003. Race and the Politics of Welfare Reform Ann Arbor, MI: Univer sity of Michigan. Sears, David. 1988. Symbolic Racism. In Eliminating Racism: Profiles in Controversy ed. P. Katz and D. Taylor. New York: Plenum Press. Sears, David and P.J. Henry. 2003. The Origins of Symbolic Racism. Journal of Personality and Social Psychology 85:259-275. Shapiro, Andrew L. 1993. Note: Challenging Cr iminal Disenfranchisement under the Voting Rights Act: A New Strategy. Yale Law Journal 103:537. Simon, Elizabeth. 2002. Justice Denied: How Felony Disenfranchisement Laws Undermine American Democracy. Washington, D.C.: Am ericans for Democratic Action Education Fund. October 11, 2005. Singer, Judith D. and John B. Willett. 1993. It s About Time: Using Discrete-Time Survival Analysis to Study Duration and the Timing of Events. Journal of Educ ation Statistics 18:155-195. Skogan, Wesley G. 1996. "Crime and Punishment." In Politics in the American States: A Comparative Analysis ed. Virginia Gray and Herbert Jacob. 6th ed. Washington, D.C: CQ Press. Smith vs. Allwright, 321 US 649 (1994) Soss, Joe, Sanford F. Schram, Thomas P. Vart anian, and Erin OBrien. 2001. Setting the Terms of Relief: Explanaing States Policy Choices in the Devolution Revolution. American Journal of Political Science 45:378-395. Soule, Sarah A. and Jennifer Earl. 2001. The En actment of State-Level Hate Crime Law in the United States: Intrastate and Interstate Factors. Sociological Perspectives 44:281-305. 144

PAGE 145

Southern Poverty Law Center. 2002. Drawing the Line: Protect Your Voting Rights During Redistricting Montgomery, AL: Southern Poverty Law Center. /static/drawing_the_line.pdf June 11, 2008. Spillane, Joseph. 2000. Cocaine: From Medical Marvel to Modern Menace in the United States 1884-1920. Baltimore: John Hopkins University Press. Spratt, Leonida W. 1861. The Philosophy of Secession: A Southern View. Documenting the American South Chapel Hill: The University of North Carolina Chapel Hill. June 11, 2008. Squire, Peverill. 1992. Legislative Professionali zation and Membership Diversity in State Legislature. Legislative Studies Quarterly 17:69-79. --.2007. Measuring State Legislative Professi onalism: The Squire Index Revisited. State Politics and Policy Quarterly 7:211-227. Stucky, Thomas D, Karen Heimer, and Joseph B. Lang. 2005. Partisan Politics, Electoral Competition and Imprisonment: An An alysis of States Over Time. Criminology 43:211247. Swados, Elizabeth and Bill Moye rs. 1996. The First Murder. Genesis. PBS, June 11, 2008. Talrico, Susette M. 1985. Courts and Crim inal Justice: Emerging Issues. In Perspectives in Criminal Justice Beverly Hills, CA: Sage Publications. The United States Bureau of Labor.1980. Geographic Profile of Employment and Unemployment. Washington, D.C: U.S. Bureau of Labor. January 30,2008. The United States Census Bureau. 1980. Statistical Abstract of the United States Washington, D.C,: United States Census Bureau. June 11, 2008. --Current Population Survey: Voter Supplement Files Los Angeles, CA: University of California Los Angeles Institute for Social Science Research Data Archives ( sr/da/index/techinfo/M1551.HTM, January 1-30-08) The United States Constitution. 1787. The United States Department of Justi ce, Bureau of Jus tice Statistics. 1982. Prisoners 1925-81. Washington, D.C.: U.S. Department of Justice. The United States Sentencing Commission. 1995. R eport on Cocaine and Federal Sentencing Policy. Washington, D.C.: United States Sentencing Commission. June 4, 2008. 145

PAGE 146

Thompson, Nicholas. 2001. Locking up the Vote: Disenfranchisement of Former Felons was the Real Crime in Florida. Washington Monthly January/February. Tolbert, Caroline and Rodney Hero. 1996. Race/Et hnicity and Direct Democracy: An Analysis of Californias Illegal I mmigration Initiative. Journal of Politics 58:806-18. Tompson, Randall and Matthew Zingraff. 1981. Detecting Se ntencing Disparity: Some Problems and Evidence. The American Journal of Sociology 86:869-880. Travis, Jeremy. 2002. Invisible Punishment an Instrument of Social Exclusion. In Invisible Punishment: The Collateral Cons equences of Mass Imprisonment ed. Marc Mauer and Meda Chesney-Lind. New York: The New Press. Tyran, Jean-Robert and Rupert Susgruber. 2005. The Diffusion of Policy Innovations-an Experimental Investigation. Journal of Evolutionary Economics 15:423-442. Uggen, Christopher and Jeff Manza. 2002. Democrati c Contraction? Political Consequences of Felon Disenfranchisement in the United States. American Sociological Review 67:777803. Uggen, Christopher, Jeff Manza, and Melissa Thompson. 2006. Citizenship, Democracy, and Civic Reintegration. The Annals of the American Academy of Political and Social Science 605:281310. Walker, Jack L. 1969. The Diffusion of Innovations among the American States. American Political Science Review 63:880-99. Walker, Samuel, Cassia Spohn, and Miriam DeLeone. 2004. The Color of Justice: Race, Ethnicity, and Crime in America 3rd ed. Belmont, CA: Wadsworth. Wheelock, Darren. 2005. Collateral Conse quences and Racial Inequality. Journal of Contemporary Criminal Justice 21:82-90. Williams, Richard, 2005. Gologit2: Generalized Logistic Regression Models for Ordinal Dependent Variables.North American STATA Users Group Meeting 2005 21. June 11, 2008. Wright, Gerald. C. 1996. Review: De mocracy in the Fifty States. The American Political Science Review 90:191-192. Wright, Gerald, Robert Erikson, and John Mc Iver. 1985. Measuring State Partisanship and Ideology with Survey Data. Journal of Politics 47:469-489. Yates, Jeff and Richard Fording. 2005. Politics and State Punitiveness in Black and White. Journal of Politics 67:1099-1121. Zelden, Charles L. 2002. Voting Rights on Trial Santa Barbara, CA: ABC-CLIO, Inc. 146

PAGE 147

BIOGRAPHICAL SKETCH Natasha Christie was born in Bronx, NY, where she resided until attending Cornell University in Ithaca, NY. In 2001, she graduated from Cornell University with a Bachelor of Arts degree in history with a law and society concentration. In May 2004, she received a Master of Arts degree in political science from the University of Florida where she also received her doctorate in 2008. Dr. Christie currently resides with her husband and children in Florida where she is employed as a criminal justice analyst.