INTRODUCTION
Political, social, and civil rights are fundamental to democratic citizenship (Marshall Reference Marshall1950). Political rights are associated with participation in governance, social rights involve benefits that reduce economic vulnerability and promote social security, and civil rights cover the exercise of particular freedoms like speech and privacy. Arguably, the historical trend in the United States has been expansion of these sets of rights, often in response to contentious movements and in order to maintain vigilance against retrograde politics (Keyssar Reference Keyssar2009). However, rather than an expansion of rights, persons with felony convictions tend to face a contraction of their civil liberties and freedoms (Holloway Reference Holloway2014; Keyssar Reference Keyssar2009). These restrictions are known as the collateral consequences of a criminal conviction, which include the legal sanctions and restrictions imposed upon people because of their criminal record (Dawson-Edwards Reference Dawson-Edwards2008; Heumann et al., Reference Heumann, Pinaire and Clark2005; Uggen and Manza, Reference Uggen and Manza2002).
Majoritarian politics—when status-based majorities exert influence over society—tend to produce these collateral consequences (Holloway Reference Holloway2014; Katzenstein et al., Reference Katzenstein, Mohsen Ibrahim and Rubin2010; Manza and Uggen, Reference Manza and Uggen2006; Mauer and Chesney-Lind, Reference Mauer and Chesney-Lind2002; Mele and Miller, Reference Mele and Miller2005; Owens and Smith, Reference Owens and Smith2012; Travis Reference Travis, Mauer and Chesny-Lind2002). Since most people in legislative and executive office want to appear tough on crime and criminals, they tend to produce policies that seek to regulate the behavior of felons as citizens, to deny them entitlements during and after completion of penal sentences and restitution, and to exclude felons from political communities and economies (Mauer and Chesny-Lind, Reference Mauer and Chesney-Lind2002; Mele and Miller, Reference Mele and Miller2005; Pager Reference Pager2007; The Pew Charitable Trusts 2010).
Ostensibly, these collateral consequences of felony convictions are race-neutral policies in that they do not explicitly target any particular racial-ethnic group. States commonly adopt and enforce criminal justice policies, such as capital punishment, mandatory minimum sentences, “Three Strikes” laws, and social welfare policies like bans on cash and food assistance for poor families that primarily affect individual felons. Unlike those policies, politically-oriented collateral consequences, inclusive of voting disenfranchisement, the inability to hold public office, and the denial of service on juries, influence broader society in terms of civic participation and substantive representation in governance. In other words, the denial of political rights to felons can affect everyone. Moreover, by reducing political engagement, empowerment, and efficacy, the denial of political rights disproportionately affects the citizenship of racial minorities, especially African Americans, who as a group are disproportionately represented in the criminal justice system (Bowers and Preuhs, Reference Bowers and Preuhs2009; King and Mauer, Reference King and Mauer2004; Mazna and Uggens, Reference Manza and Uggen2006). Thus, while the restrictions on political rights for felons are ostensibly nonracial, the effects certainly have disparate consequences, especially if support for the restrictions is based on racial sentiments.
Racialized opinions promote political bias and discrimination and can ultimately lead to the suppression of freedoms and therefore citizenship. If public sentiment toward the restoration of political rights for felons is tainted by racial attitudes that exist over and above principled reasons on the issue, then the standards of democracy are themselves threatened. Yet, public opinion on the issue of felon political rights is quite limited (Dawson-Edwards Reference Dawson-Edwards2008). While the existing literature suggests the American public is quite supportive of felon political rights, especially related to voting (Manza et al., Reference Manza, Brooks and Uggen2004; Pinaire et al., Reference Pinaire, Heumann and Bilotta2003; Uggen and Manza, Reference Uggen and Manza2002), the research citing explanations of support for or opposition to these rights has yet to offer an empirical account of how racial attitudes affect opinion.Footnote 1
Going beyond the suggestion that politically-oriented collateral consequences are racially codified, this article examines the extent to which racial attitudes, specifically racial resentment, affect attitudes about the extent to which society should restore political rights to felons, inclusive of restitutions such as voting in federal elections and holding public office. Further, we investigate how political ideologies, which serve as the principled basis for stances on political criminal justice issues, influence and interact with racial attitudes to bias collateral consequences for felons in the mind of the body politic (Feldman and Huddy, Reference Feldman and Huddy2005; Gilens et al., Reference Gilens, Sniderman and Kuklinski1998; Sniderman and Carmines, Reference Sniderman and Carmines1997).
Our work builds on findings from survey-based research on support for the political rights of felons and variations in support by race (Dawson-Edwards Reference Dawson-Edwards2008; Dawson-Edwards and Higgins, Reference Dawson-Edwards and Higgins2013; Manza et al., Reference Manza, Brooks and Uggen2004; Pinaire et al., Reference Pinaire, Heumann and Bilotta2003) as well as a felon’s status (e.g., on probation or parole) and criminal background (Manza et al., Reference Manza, Brooks and Uggen2004). It also leverages scholarship on the social and political psychology of racial attitudes and their punitive effects on criminal justice issues (Bobo and Johnson, Reference Bobo and Johnson2004; Hurwitz and Peffley, Reference Hurwitz and Peffley2005; Johnson Reference Johnson2008, Reference Johnson2009; Peffley and Hurwitz Reference Peffley and Hurwitz2010). We use survey data from a nationally representative sample of adults in the United States, thereby addressing the need for more generalizable studies of public opinion about felon rights (Dawson-Edwards Reference Dawson-Edwards2008; Pinaire et al., Reference Pinaire, Heumann and Bilotta2003).
ATTITUDES TOWARD FELON RIGHTS AND VOTING
Despite an abundance of scholarship on the design, implementation and institutionalization, and effects of restrictions on the political rights of felons, there are few studies of public attitudes about their political rights in the United States (Dawson-Edwards Reference Dawson-Edwards2008; Dawson-Edwards and Higgins, Reference Dawson-Edwards and Higgins2013; Manza et al., Reference Manza, Brooks and Uggen2004; Manza and Uggen, Reference Manza and Uggen2006; Pinaire et al., Reference Pinaire, Heumann and Bilotta2003). The extant studies with nationally representative samples focus exclusively on voting rights for felons and primarily report descriptive statistics. They conclude that majorities (60% to 80%) of Americans agree that felons should have full voting rights at some point following their convictions (Manza et al., Reference Manza, Brooks and Uggen2004; Manza and Uggen, Reference Manza and Uggen2006; Pinaire et al., Reference Pinaire, Heumann and Bilotta2003). Additionally, experimental evidence suggests that public support for restoration of felons’ voting rights depends on factors related to the status of felons (i.e., stage of correctional control) and the nature of their crimes (e.g., violent vs. nonviolent) (Manza and Uggen, Reference Manza and Uggen2006; Manza et al., Reference Manza, Brooks and Uggen2004). Finally, while majorities of both Whites and Blacks support voting rights restoration at some point for felons, racial differences in support exist. In particular, Whites are more likely than Blacks to favor disenfranchisement for the imprisoned. Moreover, while no Blacks favor permanent denial of voting rights to felons, 16% of Whites favor it (Pinaire et al., Reference Pinaire, Heumann and Bilotta2003).
The aforementioned results suggest that support for the political rights of felons, at least as measured by voting rights, are conditional and depend on readily accessible information. Notably, however, measures of racial attitudes, which underlie some of the most important cleavages shaping American society, especially as they pertain to crime and criminal justice issues, are absent from public opinion studies.Footnote 2 Such an omission is of particular concern, given the systematic evidence of the effects of racial attitudes on nonracial crime and welfare policies (Gilens Reference Gilens1996, Reference Gilens1999; Kinder and Sanders, Reference Kinder and Sanders1996; Peffley et al., Reference Peffley, Hurwitz and Sniderman1997) and the many anecdotes and politicized statements about the racialized nature of rights and crimes in America (Behrens et al., Reference Behrens, Uggen and Manza2003). Thus, one gap in the literature is an understanding of the extent to which policies affected by racial considerations produce political collateral consequences for felons.
HOW RACE-NEUTRAL POLICIES AND ISSUES BECOME RACIALIZED
Laws that limit the political rights of felons and ex-felons are positioned as race neutral. However, given that individuals tend to associate images of criminals and felons with Blacks (Entman and Rojecki, Reference Entman and Rojecki2001; Gilliam and Iyengar, Reference Gilliam and Iyengar2000; Hurwitz and Peffley, Reference Hurwitz and Peffley1997, Reference Hurwitz and Peffley2005; Peffley et al., Reference Peffley, Hurwitz and Sniderman1997), and the outcomes of such laws disproportionately affect Blacks, laws offering political equality for felons and race become essentially non-separable in the minds of Whites and other non-Black minorities. That is, when many think of crime and punitive crime policies or collateral consequences, they are more likely to think of Black than White offenders (Eberhardt et al., Reference Eberhardt, Atiba Goff, Purdie and Davies2004; Welch Reference Welch2007). Similar to how other race-neutral policies—such as “Three Strikes laws,” illicit drug sentencing practices, welfare and food stamps benefits, the death penalty, urban renewal, and Head-Start programs—are shaped by perceptions of Blacks and the poor (Gilens Reference Gilens1996, Reference Gilens1999; Kinder and Sanders, Reference Kinder and Sanders1996; Manville Reference Manville2012; Peffley and Hurwitz, Reference Peffley, Hurwitz and Sniderman1997), tacit approval of felon political rights may also become racially codified, and, as a result, may become another platform for conveying racial resentment toward Blacks.
Numerous studies demonstrate the relationship between perceptions of Blacks and attitudes towards policies in the domains of criminal justice, social welfare, and even urban planning (Bobo and Johnson, Reference Bobo and Johnson2004; Gilens Reference Gilens1996, Reference Gilens1999; Hurwitz and Peffley, Reference Hurwitz and Peffley2005; Johnson Reference Johnson2008, Reference Johnson2009; King and Wheelock, Reference King and Wheelock2007; Manville Reference Manville2012). Generally, the studies find that negative perceptions of Blacks tend to reduce support for ameliorative public policies, particularly those deemed to advantage Blacks. Regarding criminal justice policies, Lawrence D. Bobo and Devon Johnson (Reference Bobo and Johnson2004) conclude that racial animus is key to understanding opinion on criminal justice policies. They find that the punitive attitudes of Whites are less tied to instrumental concerns about crimes, but instead are rooted in racial biases (Johnson Reference Johnson2008). Social welfare studies tend to draw similar conclusions. Martin Gilens (Reference Gilens1996) demonstrated that Whites’ opposition to welfare is rooted in their stereotypes of Blacks; specifically, that they are lazy. For example, in Gilens’ welfare mother experiment, participants were asked their impressions of a welfare recipient described as either Black or White. They also were asked how likely it was that the woman would try hard to find a job and how likely it was that she would have more children as a means of receiving larger welfare checks. Based on this research, Gilens (Reference Gilens1996) concluded, “White Americans’ welfare views are clearly not ‘race-neutral’ expressions of their economic self-interest, commitment to individualism, or evaluations of poor people in general. Instead, those views are strongly rooted in their beliefs about Blacks” (p. 600). Thus, the tendency to associate circumstances such as criminality or poverty with negative Black stereotypes converts seemingly race-neutral issues or policies into race-based issues and policies (Peffley and Hurwitz, Reference Peffley and Hurwitz2010). Therefore, when individuals racialize criminal justice and social welfare issues, they also racialize policies related to those domains thereby affecting Blacks, inclusive of collateral consequences for felony convictions that cover political and social rights.Footnote 3
In short, perceptions of race-neutral policies, such as those that pertain to criminal justice and poverty, are inextricably connected to perceptions about Blacks. This connection is forged in the minds of Whites by racial stereotypes and perceptions that Blacks are disproportionately involved in the criminal justice and welfare systems. Despite being race neutral, criminal justice statutes, unintentionally or not, become a code for communicating racial information without appearing overtly racist. Attitudes toward policies restoring the political rights of felons are expected to be similar to attitudes toward other race-neutral criminal justice issues and, more generally, to those of race-conscious policies.
Additionally, the racial attitude and race-neutral policy linkage helps to better comprehend collateral consequences for felony convictions that serve as political practices historically tied to racial domination and social control in the United States. One should recall that the origins of politically-oriented collateral consequences in many states, especially in the South, were race based, designed to diminish the political participation and influence of African Americans at all levels of government (Behrens et al., Reference Behrens, Uggen and Manza2003; Dinan Reference Dinan2007; Holloway Reference Holloway2014). Moreover, the adoption of contemporary collateral consequences among the states supports enduring levels of racial domination and social control. To cite an example, states with majoritarian racial systems—those governing populations with higher percentages of non-White residents—tend to be more punitive when denying political rights to felons, especially regarding felony disenfranchisement (Behrens et al., Reference Behrens, Uggen and Manza2003; Manza and Uggen, Reference Manza and Uggen2006).
RACIALIZED RESENTMENT
In response to the changing nature of racism, from overt to more covert expressions, the concept of racial resentment captures Whites’ disdain for Blacks’ seemingly illegitimate demands on society and the opportunities offered by public and private institutions to those who are undeserving. Following Donald Kinder and Lynn Sanders’ (Reference Kinder and Sanders1996) definition, racial resentment is the contention that “Blacks do not try hard enough to overcome difficulties they face and they take what they have not earned” (pp. 105–106).
Consistent with its generalized notion, resentment is a form of anger that results from perceptions of undeservingness. Norman Feather (Reference Feather2006, Reference Feather2008) suggests that resentment is likely to occur when a person is perceived to enjoy a positive outcome (e.g., the right to vote, welfare benefits) when that person’s actions are perceived by others as negative (e.g., the person was convicted of a crime or sexual promiscuity) or when that person is perceived to possess negative qualities (e.g., lazy, violent, or of low moral character) that would not be acceptable in a just world. Thus, ideas about deservingness as well as perceptions or stereotyping of out-groups (including attribution errors) and justice interact to produce feelings of resentment. Further, evidence suggests a connection between resentment and schadenfreude (i.e., the pleasure in another’s misfortune) (Feather Reference Feather2008; Feather and Sherman, Reference Feather and Sherman2002). This form of animosity is likely to produce reactions that go beyond an emotional response and lead to a desire for punitive treatment.
Drawing from social psychology literature on general resentment, David Wilson and Darren Davis (Reference Wilson and Davis2011) defined racial resentment as an explicit feeling of animosity or irritation toward Blacks who are perceived to seek or receive unfair and unjust advantages or opportunities that come at the expense of others. Wilson and Davis elaborate that racial resentment stems from beliefs that racial groups, specifically Blacks, are trying to skirt the sociopolitical rules that govern fair play and that charges of racism and discrimination are mere excuses for their the group’s own shortcomings. Accordingly, alleged advantages and opportunities that are perceived to be due to race are viewed as unfair, unearned, and undeserved, which ultimately challenges one’s sense of justice (Darley and Pittman, Reference Darley and Pittman2003; Feather Reference Feather2008).
Following this reasoning and based on the inseparability of race and beliefs about unfairness resulting from “special” considerations of one’s situation, racial resentment may affect reactions to the restoration of felons’ political rights. We contend that individuals who support the denial of political rights for felons believe that criminals have violated the social contract and, thus, deserve to lose their benefits of citizenship, particularly many political rights, even after serving their sentences. Any generalized request to have voting rights restored will be seen as a special consideration for unearned and undeserved rights. Moreover, we suspect that individuals are likely to believe that, because Blacks are ultimately to blame for their own criminality and have benefitted in the past from illegitimate and unjust benefits—special considerations—because of race, they too are undeserving of further civil benefits. Thus, racial resentment should have a strong negative effect on support for felons’ rights.
CONSERVATISM AND RACIAL RESENTMENT
Ideology and racial resentment overlap because attempts—typically via government policy—to equalize society violate norms of individualism; they excuse a lack of self-discipline or immoral behavior and ignore moral justice (e.g., people get what they deserve). These violations of values strike a nerve among mostly self-described conservatives. As Paul Sniderman and Edward Carmines (Reference Sniderman and Carmines1997) suggest, how Whites feel about Blacks should be a secondary consideration for conservatives since they already have reasons for opposing racial policies, regardless of their feelings for Blacks; that is, conservatives are perhaps “principled” rather than prejudiced in their beliefs about ameliorative racial policies (Gilens et al., Reference Gilens, Sniderman and Kuklinski1998; Sniderman and Carmines, Reference Sniderman and Carmines1997). Yet, conservatives also have more difficulty supporting ameliorative policies when targeted at Blacks than when those policies are targeted at women or the poor (Sniderman and Piazza, Reference Sniderman and Piazza1993; Wilson et al., Reference Wilson, Moore, McKay and Avery2008). For conservatives, race is an especially salient source of anxiety as Blacks, more than any other group in society, are perceived to violate conservative ideals of individualism, self-sacrifice, and discipline. These perceptions may lead conservatives to conflate race, criminal behavior, and policies designed to alleviate collateral consequences.
Liberals as well are not free from the influence of racial resentment, and, if the evidence is correct, racial resentment should more powerfully shape their reactions to racial policies. While liberals and conservatives may have similar beliefs about Blacks, albeit at different levels (Feldman and Huddy, Reference Feldman and Huddy2005; Tessler and Sears, Reference Tessler and Sears2010), self-reported liberals lack a principled reason for rejecting equal rights or opportunity policies other than their perceptions of Blacks (Gilens et al., Reference Gilens, Sniderman and Kuklinski1998; Sniderman and Carmines, Reference Sniderman and Carmines1997). Opposing ameliorative racial policies is not an easy decision for liberals, as it goes against their beliefs about equality and opportunities for all, especially Blacks. In a somewhat counterintuitive fashion, the influence of racial resentment should be greater for them than for conservatives. Sniderman and Carmines (Reference Sniderman and Carmines1997) identify the irony in the notion that “while prejudice is weakest, that is, on the political left, its political impact may be strongest, that is, most capable of bolstering opposition to government efforts to assist blacks” (p. 78).
These liberal and conservative inclinations are supported by systematic evidence. For a variety of different racial policies (e.g., government spending for programs to help Blacks get jobs, fighting discrimination against Blacks in jobs, welfare spending, job quotas for Blacks, and preferential admissions to universities for Blacks), Sniderman and Carmines show that conservatives are less supportive of racial policies and that there are small differences between highly-prejudiced conservatives and lesser-prejudiced conservatives. There are, however, large differences between lesser-prejudiced liberals and highly-prejudiced liberals in that highly-prejudiced liberals’ reactions to racial policies are very close to those of highly-prejudiced conservatives. For example, Stanley Feldman and Leonie Huddy (Reference Feldman and Huddy2005), using a measure of racial resentment, show a similar unwillingness among liberals and conservatives to support a college scholarship program targeted at Blacks.
Drawing from these studies, we tested four key hypotheses. First, racial resentment more powerfully shapes beliefs about felon political rights than do other explanations. Second, liberal-conservative self-identification is an important consideration, as it reflects a more principled approach to law and order, and, thus, conservatives should be more likely than liberals to oppose efforts to restore political rights to felons.Footnote 4 Third, based on the overlap of racial resentment and ideology, the level of support for felons’ rights is expected to be lower for conservatives than for liberals and relatively consistent across levels of racial resentment; that is, regardless of conservatives’ beliefs about Blacks, their support for felons’ rights should not vary much. Fourth, the impact of racial resentment should be greater among liberals than conservatives, as they do not have a plausible reason for opposing felon political rights other than their views about Blacks.
DATA AND MEASURES
The data for this research come from the 2010 Cooperative Congressional Election Study (CCES). Conducted by YouGov/Polimetrix, the CCES is a stratified (i.e., by states and congressional districts), multi-stage random national sample of the U.S. population. YouGov/Polimetrix surveys are conducted online using an opt-in panel of respondents. Economies of scale are achieved by allowing teams of scholars to collaborate on the survey with CCES principle investigators who provide the “common content” for the questionnaire. The data used in our analyses are from both the common and team content collected during the national pre-election (September and October 2010) and post-election (November 2010) periods.Footnote 5 The original data included 1,000 respondents; however, only non-Black respondents received questions measuring racial resentment. In the final sample size of 882 (which excluded the 118 African Americans), the demographics were as follows: Whites = 757, Hispanics = 60, Asians = 11, Native Americans = 8, Mixed (non-Black) = 2, and Other = 26. All analyses were weighted to reflect the demographics of the general population.
Measures of Attitudes toward Felon Political Rights
We use three different measures to reflect individual positions on the restoration of political rights for felons (cf. Manza et al., Reference Manza, Brooks and Uggen2004; Pinaire et al., Reference Pinaire, Heumann and Bilotta2003). We gauge stances by: (1) whether an individual would support Congressional legislation to restore the right to vote in federal elections to ex-prisoners; (2) whether society would be better if felons were reintegrated into communities by restoring their voting rights or would be worse by giving felons a chance to influence elections; and (3) whether felons should be allowed to hold public office after completion of their sentences. Each focus captures a different aspect of support for the restoration of political rights to felons. The appendix provides wording of the questions and response options.Footnote 6 A benefit of using three different measures is that we are thereby able to examine reactions to the restoration of political rights for felons under different conditions and, perhaps, identify nuances in opinions.
The relationships among the items—congressional action and the effects on society (χ 2 (4) = 610.8, p < .01, Cramer’s V = .59), congressional action and office holding (χ 2 (8) = 260.3, p < 0.01, Cramer’s V = 0.60), and effects on society and office holding (χ 2 (8) = 208.6, p < 0.01, Cramer’s V = 0.54)—are all significant and large in magnitude, which indicates strong internal consistency among the items.
Racial Resentment Measures
We employ a racial resentment scale comprised of five items that measure the feeling of animosity and antipathy toward Blacks as undeserving, focusing on their seemingly illegitimate demands and opportunities rooted in their racial history and status (Wilson and Davis, Reference Wilson and Davis2011).Footnote 7 The items contain statements related to: (a) understanding why race is a special problem; (b) the unfairness of special considerations based on race; (c) the use of racism and slavery as excuses; (d) how special considerations for Blacks are unfair; and (e) concerns about how Blacks use race to gain an advantage (see the appendix for wording). We computed the final racial resentment scale by summing responses to the five items and then scaling the values to range from 0 to 1, higher values indicate higher scores on racial resentment. Results from principal components (PCA) and confirmatory factor (CFA) analysis of our CCESS data reveal good measurement properties: the PCA produced a single factor, with eigenvalue = 3.7, 74.1% variance explained; Cronbach’s alpha (α) = 0.91; and good CFA fit (CFI = 99.1, NFI = 99.0, RFI = 96.9, and RMSEA = 0.07).Footnote 8
Liberal-Conservative Dimension
Political ideology was measured by 5–point Likert scale created from self-reports in the CCES. The scale is comprised of very liberal (coded 1), liberal, moderate, conservative, and very conservative (coded 5) responses. Given this classification—where the most conservative category is coded with the highest value—we label this variable “conservatism.”
Control Variables
Because we did not expect racial resentment to be the only determinant of reactions to policies supporting felon political rights, several demographic and political variables were included in our analyses. We measure age in years. Gender is a dummy coded variable that contrasts males (coded 1) with females (coded 0). Education is a 6–point ordinal measure (1 = less than high school, 2 = high school graduate, 3 = some college, 4 = two-year college graduate, 5 = four-year college graduate, and 6 = post-graduate education). Family income is a 14–point ordinal measure, which ranges from less than $15,000 (coded 1) to $100,000 or more (coded 14).Footnote 9 Race/ethnicity is a dummy variable that contrasts Whites (coded 1) against all other racial-ethnic groups (coded 0), excluding Blacks (i.e., racial-ethnic minorities). Additionally, we measured the region in which the respondent lived as a dummy variable that contrasted the South (coded 1) versus all other regions (coded 0). We also measured whether respondents had social exposure (see Rose and Clear, Reference Rose and Clear2004) to persons who had been incarcerated in prison and/or jail, derived from responses to the following question: “Of all the people that you are acquainted with—those you know by name and would stop and talk to if you ran into them on the street or in a shopping mall—how many of them have served time in prison or jail? Please provide a percentage (between 0% and 100%).” We converted the continuous measure into a dichotomous measure of respondents having ex-prisoners as acquaintances (1 = yes, 0 = no) due to the skewed nature of the distribution (mean = 2.9, median = 1, SD = 12.5, range = 0–300).
ANALYSIS AND RESULTS
Racial Resentment and Support for Felon Political Rights
We first examined the distribution of responses to the three different questions about felons’ political rights, as seen in Table 1. All of the items are constructed such that a high number indicates more support for felon rights, a middle number reflects ambivalence on the issue, and a low number indicates lower support for felon political subjugation. Overall, reactions to the restrictions to felon political opportunities are somewhat nuanced.
Table 1. Distribution of Responses to Felon Disenfranchisement Measures
Note: Non-African American respondents only; data are weighted; totals may exceed 100% due to rounding.
Source: 2010 Cooperative Congressional Election Study.
For the first item, which pertains to congressional action to restore voting rights to ex-prisoners, 38% oppose this effort. While a sizeable percentage (27%) are unsure or ambivalent, 35% support their Congressman’s restoring the right to vote to ex-prisoners. This would seem to be an easy issue for individuals, but the high number of uncertain responses indicates that congressional action in restoring felons’ voting rights may, in fact, be a difficult issue to evaluate for a substantial number of respondents. People may not have heard much about felon disenfranchisement laws, they may have mixed or conflicting beliefs about race and criminal justice issues or may want to simply play it safe, the response options may not sufficiently represent their attitudes, or they simply might not care about the issue. Regardless, congressional action to restore felon voting rights is not an easy issue for the public.
A consideration of the second item, which concerns how restoring felons’ voting rights affects society, evokes a somewhat similar pattern. The results show that 35% indicate that restoring felons’ voting rights would be better for society, 27% indicate it would worsen society, and 39% are unsure. Like the measure asking about congressional action, this item also produces a significant number of people who are uncertain. Nevertheless, the selection of these “middle positions” may be caused by a number of different factors (Schuman and Presser, Reference Schuman and Presser1981).
The last question, which pertains to ex-felons being able to hold an elected office, results in a more nuanced set of responses than do the previous questions. Several changes are apparent. First, a larger percentage of individuals (45%) are opposed to restoring rights to hold political office, compared to the percentages seen in previous questions. Second, the percentage of individuals who took an uncertain position is reduced to 19%. Third, approximately 36% of individuals would support the restoration of felons’ rights to hold political office under some conditions. This percentage is essentially equivalent to the percentage who favor restoring felons’ voting rights in the previous questions. Based on the responses to the last question, it is tempting to suggest that the number of individuals who select the middle position in the first two items are really individuals who oppose felons’ voting rights, but this is only speculation— individuals may genuinely be unsure of their stances on felon voting rights. Yet, it does raise a question about the extent to which individuals are consistent in their positions across all three of the items.
We pursued this question of belief consistency across the three measures. Despite the differences in how the felon political rights items were constructed, there were individuals who took consistently punitive, lenient, or indifferent positions across the three questions, which revealed the extent of principled reactions toward felons’ political rights. We assume that individuals who take a consistent position across the three items are most principled in their stance on the restoration of political rights to felons. The extent to which individuals give an indifferent response to all three items shows a general level of ambivalence.
Using the three items, we created an ad-hoc categorical distribution of responses. Respondents who took pro-felons’ rights positions on all three items comprised one end of the distribution; next were those with two pro-felons’ rights responses; those who were unsure of their position on all three items; those with two anti-felons’ rights responses; and finally, those who took anti-felons’ rights positions on all three items.Footnote 10
Table 2 shows the distribution of responses to this new variable: 36% of individuals responded in an ambivalent manner to the felon political rights measures. Regardless of how the issues are presented, a sizeable percentage of individuals will consistently adopt a vague position. Equally important, roughly equal percentages of individuals consistently opposed restoring (20%) and supported restoring rights (18%). Additionally, while 14% opposed restoring rights on two of the three items, 12% similarly supported restoring rights. With this ad-hoc distribution in place we can take a preliminary gauge of the relationship between opinions on the restoration of felon political rights and racial resentment. The far right column of Table 2 reveals that attitudes toward these political rights for felons are not arbitrary. In particular, and supporting our hypothesis, people who have higher levels of racial resentment show greater opposition to restoring political rights to felons, and vice versa. The analysis also shows that opponents of rights restoration tend to be more homogeneous in their resentments toward Blacks than are supporters, as reflected by the standard deviations. Thus, while the restoration of political rights to felons might be an issue that produces quite a bit of indifference, confusion, or contradiction, we interpret the results to indicate that by and large such attitudes are coherent, since individuals can rely on predispositions about Blacks and criminal justice to form an opinion.
Table 2. Response Distribution across All Three Political Rights Restoration Items and Average Levels of Racial Resentment
Note: Non-African American respondents only; data are weighted.
a Descriptive statistics for index: M = 2.94, SD = 1.33;bANOVA results: F(4,680) = 40.5, p < .01, η 2 = .192.
Source: 2010 Cooperative Congressional Election Study.
With initial evidence for the negative relationship between racial resentment and support for political rights restoration, we turn our attention to whether ideology and socio-demographic factors change the relationship. To accomplish this, we used multinomial logistic regression (MLR) to examine the durability of racial resentment. Using each of the restoration of political rights items as dependent variables, we estimated the effects of the modeled factors and present the coefficients and related standard errors in Table 3. The MLR results produce two columns for each dependent variable, one predicting a supportive stance on the restoration of rights versus opposition, and another predicting uncertainty about restoring rights versus opposition. Positive coefficients indicate each factor is associated with more support (or uncertainty) for the restoration of felons’ political rights and negative coefficients indicate each factor is associated with more opposition. We focus primarily on the results predicting favorable versus unfavorable opinions on felon political rights.
Table 3. Multinomial Logistic Regression Estimates of Model Predicting Political Rights Restoration Items
Note: Non-African American respondents only.
* p < .05, **p < .01
Source: 2010 Cooperative Congressional Election Study.
The MLR results indicate that racial resentment is the strongest and most consistent predictor of stances on all three restoration of felon political rights measures. The coefficients for racial resentment are all positive, high in magnitude, and statistically significant, which indicates that the higher one’s racial resentment score, the less likely they are to support the restoration of political rights to felons. This confirms the previous finding in Table 2. Self-report ideology (conservatism) also has consistent effects, although they appear strongest for the political rights questions addressing voting. As conservatism increases, individuals are more likely to have negative views on the restoration of felon political rights. Several demographic factors are significant as well. Whites are more likely than non-African American racial minorities to believe that restoring voting rights will make society better. Those with higher family incomes are also less likely to support congressional action to restore federal voting rights to felons, but those with lower incomes tend to offer more unsure responses to the same question, as well as the question about states restoring office holding rights. Other effects reflected among the demographic factors include males being more likely than females to believe the restoration of felon voting rights will worsen society. Males are also less likely than females to offer unsure responses to both questions about felon voting rights. We also find that as age increases individuals are less likely to offer an unsure response to the question about congressional action to restore federal voting rights and the question about whether states should restore rights for felons to hold office.
Our most prominent findings thus far lead us to conclude that a main psychological obstacle to supporting the restoration of felon political rights are resentments toward African Americans over what respondents perceive to be unjust demands and opportunities that serve to unfairly advantage the group. Perhaps untethered political rights for felons are viewed among some Whites as another undeserved benefit or opportunity for persons who are likely to be Black (Gilens Reference Gilens1996, Reference Gilens1999; Hurwitz and Peffley, Reference Hurwitz and Peffley1997, Reference Hurwitz and Peffley2005; Peffley and Hurwitz, Reference Peffley and Hurwitz2002; Peffley et al., Reference Peffley, Hurwitz and Sniderman1997; Welch, Reference Welch2007) and who should suffer great consequences for their criminal activities and other violations of societal values.
Ideology: Conditional Effects
The effects of racial resentment on the reactions to restoring felon political rights policies are quite strong, but important questions remain about the extent to which the effects of racial resentment are constant across the ideological spectrum of liberals, moderates, and conservatives.
To disentangle the effects of ideology and racial resentment on support for felons’ political rights we ran additional MLR models for the three political rights variables. The models consisted of the same variables presented in our previous MLR models (Table 3), but each included an additional racial resentment by conservatism interaction term. We focus our attention primarily on the estimated effects and predicted probabilities for the interactions between racial resentment and liberal-conservative categories.
The analyses (not presented in tables) reveal statistically significant racial resentment times ideology interaction terms for all three felon rights measures. Examining only the effects for favorable versus unfavorable stances—and not unfavorable versus unsure results—we find patterns showing racial resentment has its strongest effects among liberals. These patterns are presented in Figures 1, 2, and 3.
Fig. 1. Predicted Probability of Support for Congressional Action to Restore Voting Rights for Felons: Racial Resentment and Political Ideology
Note: The magnitude of the slopes for all three levels of ideology are statistically significant. Liberals: B=-.75, SE=.05, p<.01, 95% C.I.=-.84, -.66; Moderates: B=-.47, SE=.04, p<.01, 95% C.I.=-.54, -.39; Conservatives: B=-.26, SE=.03, p<.01, 95% C.I.=-.31, -.20
Source: 2010 Cooperative Congressional Election Study.
Fig. 2. Predicted Probability for Belief Restoring Voting Rights Betters Society: Racial Resentment and Ideology
Note: The magnitudes of the slopes for all three levels of ideology are statistically significant from zero. Liberals: B=-.74, SE=.04, p<.01, 95% C.I.=-.82, -.66; Moderates: B=-.36, SE=.04, p<.01, 95% C.I.=-.43, -.29; Conservatives: B=-.14, SE=.03, p<.01, 95% C.I.=-.19, -.08
Source: 2010 Cooperative Congressional Election Study.
Fig. 3. Predicted Probability of Support for Restoring Right of Felons to Hold Public Office: Racial Resentment and Ideology
Note: Only the magnitudes of the slopes for liberals and moderates are statistically significant from zero. Liberals: B=-.93, SE=.03, p<.01, 95% C.I.=-.99, -.86; Moderates: B=-.48, SE=.04, p<.01, 95% C.I.=-.55, -.40; Conservatives: B=.01, SE=.03, n.s., 95% C.I.=-.06, .08
Source: 2010 Cooperative Congressional Election Study.
Figure 1 shows the estimated average levels of support for congressional action to restore federal voting rights for felons for the racial resentment times ideology interaction effect (B = 1.61, S.E. = .55, p < 0.01). Higher scores on the racial resentment scale for all three ideological groups—conservatives, moderates, and liberals—are associated with more punitive positions on felon political rights policies. While the same overall pattern exists—a negative relationship between support for felons’ rights and racial resentment—for all ideological groups, liberals exhibit a stronger racial resentment effect than moderates and conservatives. As can be seen in Figure 1, although they do not start off in such a fashion, liberals come to resemble moderates and conservatives on this issue when racial resentment is high. The movement of conservatives is relatively meager compared to that of moderates and liberals.
Figure 2 shows the racial resentment by ideology interaction (B = 2.10, S.E. = 0.69, p < 0.01) for the item, “Restoring voting rights will make society better.” While conservatives are again at lower levels of support for felony marginalization at all levels of racial resentment, the slope of the line is steeper for liberals. Racial resentment has a greater influence (larger slope) for liberals, but their level of support for felons’ voting rights remains distinct from that of conservatives. Similarly, the racial resentment by ideology interaction effect (B = 2.10, S.E. = 0.69, p < 0.01) on the item pertaining to “the rights of felons to hold office” shows (in Figure 4) once again that racial resentment has the greatest effect among liberals. There is no significant racial resentment effect for conservatives, and the effects for moderates are significant but smaller in magnitudes than the effect for liberals. The steep slope of the line for liberals adds to the evidence that racial resentment is a stronger driver of their estimated probability of supporting rights. In fact, liberals, along with moderates, with the higher levels of racial resentment, show the lowest support for restoring felons’ rights to hold public office.
The findings for the three individual measures of political rights restoration for felons are consistent with the research that suggests that racial resentment operates more strongly among liberals than among conservatives (Feldman and Huddy, Reference Feldman and Huddy2005). Our results, however, reveal a statistically significant racial resentment effect for conservatives on the two voting rights items, countering the claim that racial resentment is only a bias held among liberals. In part, that finding by others may be due to previous studies having stronger social desirability effects because they take place in telephone interview contexts which might reduce a willingness to express more resentful sentiments (Krysan Reference Krysan1998). Our data, however, were collected in a self-administered fashion and, thus, individuals, particularly conservatives, may have been more willing to express their forthright racial views. Without more evidence this remains only speculation on our part.
DISCUSSION AND CONCLUSION
Public opinion is the chief means for citizens to voice their ideas to elected officials who decide if and how to translate public preferences into public policy. Public attitudes set the mood for political action and help to inform political debates about important issues. While there are studies examining public attitudes toward the restoration of felon rights, few employ national samples and none provide strong evidence of an empirical link to racial attitudes.
The goal of this research was to determine the factors that influence attitudes towards the restoration of political rights for felons. We assert in this paper that, despite the race-neutral ways in which policymakers discuss politically-oriented collateral consequences for felony convictions, the laws and policies producing the consequences and race become essentially non-separable in the minds of non-African Americans, who tend to conflate criminality with Blacks. Our analyses show a consistently strong influence of racial resentment and a liberal-conservative identity on attitudes toward the restoration of felons’ political rights; however, racial resentment has larger effects between the two, and the effects of both racial resentment and liberal-conservative identity are not strictly linear.
Review of Findings
First, racial resentment is the primary driver of attitudes regarding the restoration of political rights for felons. As levels of racial resentment increase, individuals are less likely to support congressional action to restore felons' voting rights, to believe that restoring voting rights will make society better, and to believe that felons should be allowed to hold public office, even after completing their sentences. Second, conservatives, at all levels of racial resentment, tend to be more opposed than liberals to restoring voting rights for felons. Nevertheless, racial resentment does shape how conservatives think about such restorations, as conservatives with low racial resentment are less opposed to restoring felons’ voting rights than are conservatives with high racial resentment. Conservatives, despite having a principled reason to oppose restoring the political rights of felons, are, indeed, affected by negative beliefs about African Americans. Third, when it comes to felons, liberals are not, as expected, the defenders of political rights or free from the effects of racial resentment. Liberals with low racial resentment are most supportive of restoring political rights for felons. However, as feelings of racial resentment increase among liberals, their attitudes toward restoring rights to felons begin to mirror those of conservatives, especially conservatives with high racial resentment. Strong feelings of racial resentment produce almost identical reactions to restoring political rights to felons, regardless of ideological views.
What do these results mean for the restoration of political rights, as well as social and civil rights, for felons in the United States? Our findings may contribute to the debates about the design, adoption, and implementation of collateral consequences for felony convictions and campaigns to reduce or abolish them, as well as punitive policymaking and criminal justice issues more generally. The way that people think about race does not bode well for the full restoration of political rights for felons. That is, racial thinking is not simply about color, it contains emotionally laden beliefs about an assortment of traits related to stereotypes about African Americans—overgeneralizations about their behavior. Unfortunately, the informational use of these stereotypes is not confined to any one political ideological group.
If political liberals have been assumed to defend individual rights, this is partly true because liberals are not as likely as conservatives to have racial resentment toward Blacks. However, if liberals are affected by hostility and resentment toward Blacks in the same way that conservatives are affected, traditional policies, both racially conscious and race-neutral, may not enjoy support. As a result, felon rights, regardless of how one thinks about fairness and equality, may be dismissed.
As images of crime and criminal behavior continue to be conflated with stereotypes about Blacks, the stances of liberals and conservatives potentially may converge. Moreover, images in political ads, news stories, and other portrayals of felons seeking voting rights that include African Americans may only exacerbate the belief that felons are undeserving (Entman and Rojecki, Reference Entman and Rojecki2001). The saving grace for these racialized—nonracial—issues may be the ability to use framing to appeal to the principled rights and liberties aspects of these laws rather than focusing on the victimized group (Lakoff Reference Lakoff2002). Making explicit appeals to constitutional principles may override racialized ideas about narrowing or broadening the political rights of felons, as well as prime considerations of democratic fairness and justice, participation, and governance. However, the empirical results of such practices are unknown at this point. We are left with the glum conclusion that, absent any counterstrategy or change among a critical mass of citizenry or elected officials, as a result of race-crime-ideology convergence, it is probable that felons will continue to possess far fewer political rights than they did before their criminal convictions, and that the public will continue its ambivalence toward the diminished citizenship of felons in the United States for some time to come.
APPENDIX
Question Wording
Positions on the Restoration of Political Rights for Felons
[NCB26–Congressional Action] Congress is considering legislation that would restore the right to vote in federal elections to ex-prisoners who are U.S. citizens. Should your Congressman support or oppose restoring the right to vote in federal elections to ex-prisoners?
1. Support
2. Unsure
3. Oppose
[NCB29–Effects on Society] Some people argue that permitting felons to vote after they have served their time would better society by improving their ability to integrate back into society. Other people argue that permitting felons to vote would worsen society by giving criminals a chance to influence elections. Which of the following statements is closer to your own opinion?
1. Permitting felons to vote after serving their time would better society
2. I’m not sure
3. Permitting felons to vote after serving their time would worsen society
[NCB418–Holding Office] States are debating whether convicted felons should be allowed to hold public office after they have completed their sentences. Which of the following proposals is closest to your view?
1. Automatically restore the right to hold public office after the completion of their sentence
2. Restore the right to hold public office after a certain period of time beyond the completion of their sentence
3. Restore the right to hold public office after receiving a pardon and/or return of other civil rights
4. Never restore the right to hold public office even after the completion of their sentence
5. I'm not sure
Racial Resentment
Please indicate how much you agree or disagree with the following statements:
[NCB19] I don’t understand why race is any different from what other people have to deal with.
[NCB20] I resent any special considerations that Africans Americans receive because it’s unfair to other Americans.
[NCB21] For African Americans to succeed, they need to stop using racism and slavery as excuses.
[NCB22] Special considerations for African Americans place me at an unfair disadvantage because I have done nothing to harm them.
[NCB23] African Americans bring up race only when they need to make an excuse for their failure.
1. Strongly disagree
2. Disagree
3. Neither agree nor disagree
4. Agree
5. Strongly agree
