INTRODUCTION
For decades, American scholars have studied the intersection of race, crime, justice, and presidential politics, focusing particularly on the Goldwater, Nixon, and Reagan eras, which were characterized by the emergence and success of the Republican “southern strategy” and broader themes of “law and order” and “war on crime.” Recently, a series of highly-publicized deaths of African American citizens during encounters with police—and the Black Lives Matter movement that has emerged in their wake—have increased the visibility of this racially-driven intersection, drawing political attention to issues involving contact with and attitudes toward the police.
The present study explores the impact of these events on the shape and directions of national political behavior in the contemporary era. We posit that citizen contact with the criminal justice system and perceptions of the police, in conjunction with racial feelings and resentments, are likely to be highly influential in predicting candidate selection and the likelihood of citizen participation in the 2016 Presidential election. We test this perspective using data from the 2016 Pilot Study of the American National Election Studies, a representative survey of voting-age citizens in the United States. In new and revealing ways, the results suggest that political behavior in the 2016 presidential election will be influenced by citizens’ experiences with and attitudes about race, justice, and policing, adding a new chapter to the long American history of race, crime, and presidential politics.
CONTEMPORARY POLITICAL BEHAVIOR IN U.S. RACIAL HISTORICAL CONTEXT
To understand the current political climate surrounding race, crime, and the 2016 presidential election, it is helpful to consider how these issues became intertwined in the midst of the Civil Rights era. In 1961 and 1962, as organizations began to register Black voters across southern states where post-Reconstruction policies had disenfranchised them (McAdam Reference McAdam1988), Southern politicians who were threatened by the massive change that would accompany an enfranchised African American electorate began redefining political action in support of these policies as disorderly and unlawful (Beckett Reference Beckett1997; Beckett and Sasson, Reference Beckett and Sasson2004; Tonry Reference Tonry2011; Wacquant Reference Wacquant, Pratt, Brown, Brown, Hallsworth and Morrison2005; Weaver Reference Weaver2007). Efforts to register voters or otherwise advocate for political civil rights were also met with violent repression, often organized by the state through the mobilization of the police (Branch Reference Branch1998; Hampton and Fayer, Reference Hampton and Fayer1990).
In the summer of 1964, mass racial protests and violence erupted in Rochester, Harlem, and Philadelphia. Similar protests and violence occurred during the Watts Rebellion in Los Angeles in the summer of 1965, and in Detroit, Newark and other cities across the country throughout the summer of 1967. Each of these uprisings was sparked by a specific incident of perceived police mistreatment of Black citizens, although in each case tensions were already high between the police and the Black community.
As a result of these political and social actions, segregation, Black disenfranchisement, urban unrest, and law and order were major themes in several presidential elections, beginning with the transformative and consequential 1964 election. From the end of Reconstruction—when White southerners, including many former Confederate soldiers and politicians, re-established social and political dominance over Blacks in their states—until 1964, voters in “deep south” states had reliably voted against Republican presidential candidates. Footnote 1 The 1964 presidential election brought a reversal of this trend; for the first time, deep south states voted for the Republican candidate, Barry Goldwater, while the majority of the country voted for Democrat Lyndon B. Johnson. This pattern has persisted over time, with the deep south consistently voting against the Democratic presidential candidate ever since. Footnote 2
The 1964 election was consequential for a variety of reasons. It was the first stage of the “southern strategy,” where Republicans attempted to peel large numbers of southern White voters away from the Democratic Party by appealing to their discomfort with the racial changes posed by the Civil Rights Movement (Carmines and Stimson, Reference Carmines and Stimson1989; Edsall and Edsall, Reference Edsall and Edsall1991; Tonry Reference Tonry2011; Wacquant Reference Wacquant, Pratt, Brown, Brown, Hallsworth and Morrison2005). The 1964 election also saw the emergence of a racially-coded “tough on crime” rhetoric that initially sought to criminalize those involved in political protest, and which later evolved into support for “wars” on crime and drugs in the 1980s and 1990s (Beckett Reference Beckett1997; Weaver Reference Weaver2007). The policies and practices of these “wars” ultimately resulted in a significant uptick in aggressive policing tactics and massive increases in state and federal incarceration, all of which disproportionately affected Black communities (Alexander Reference Alexander2010; Beckett and Sasson, Reference Beckett and Sasson2004; National Research Council 2014; Tonry Reference Tonry2011). In addition, the 1964 presidential contest, and the election of Lyndon Johnson, also marked a major moment for the civil rights movement. Lyndon Johnson was able to mobilize congressional support and passage of the Civil Rights Act of 1964, the Voting Rights Act of 1965, and eventually the Civil Rights Act of 1968.
Lessons from this era suggest a basic model for understanding the role of race, crime, and the police in politics: issues of crime and policing are invoked not only as part of strategies to maintain status-quo racial group positions in the face of perceived threats, but also in efforts to challenge this status quo (Beckett and Sasson, Reference Beckett and Sasson2004; Blumer Reference Blumer1958; Feagin Reference Feagin2009; Tonry Reference Tonry2011). When protestors in the 1960s sought to draw attention to the mistreatment of Black citizens by the police as well as other issues of racial equality, some political actors responded by reframing these protests as promoting lawlessness and disorder and by otherwise highlighting images of Black crime to delegitimize the protesters’ grievances and call to action (Beckett Reference Beckett1997; Beckett and Sasson, Reference Beckett and Sasson2004; Tonry Reference Tonry2011; Wacquant Reference Wacquant, Pratt, Brown, Brown, Hallsworth and Morrison2005; Weaver Reference Weaver2007). This “law and order” and “tough on crime” rhetoric served as a racial “dog whistle” (López Reference López2014), rallying support from those opposed to the expansion of African Americans’ civil rights, a usage that continues in subsequent elections, especially during the “wars” on crime and drugs (Beckett Reference Beckett1997; Beckett and Sasson, Reference Beckett and Sasson2004; Edsall and Edsall, Reference Edsall and Edsall1991; Tonry Reference Tonry2011). In addition, the criminal justice system was used as part of the effort to reclaim White dominance in the political sphere through the criminalization of political protest and voter registration efforts, intimidation of prospective voters, and eventual state-level revocations of voting rights through felon disenfranchisement laws (Beckett and Sasson, Reference Beckett and Sasson2004; Behrens et al., Reference Behrens, Uggen and Manza2003; Manza and Uggen, Reference Manza and Uggen2006: Tonry Reference Tonry2011; Uggen and Manza, Reference Uggen and Manza2002). On the other hand, widespread perceptions of injustice—rooted in real bias and mistreatment on the part of the criminal justice system, and in particular the police—served as a flashpoint to motivate and rally support for civil rights efforts (Branch Reference Branch1998).
The 2016 Presidential Election
The politics of race, crime, and criminal justice that emerged in the 1960s left an indelible mark on the American electorate. Despite progress over the last sixty years, race has retained critical importance in American politics, and issues involving crime and the police remain highly racialized.
In a striking parallel to the events of the early 1960s, public and political attention to the relationship between the police and the Black community increased significantly following the deaths of many Black citizens during encounters with police in the years preceding the 2016 election. Attention to police treatment of African American suspects began with the deaths of Eric Garner in New York City and Michael Brown in Ferguson, Missouri, and continued through the deaths of Tamir Rice, Freddie Gray, Laquan McDonald, Alton Sterling, Philandro Castile, and Terence Crutcher, among others. These events have inspired mass public protests in cities such as Ferguson, St. Louis, New York, Baltimore, Cleveland, Baton Rouge, and St. Paul. In an echo of the past, some politicians have framed the protests as disorderly, violent, and unlawful, while expressing strong support for the police (Campbell Reference Campbell2015; Diamond Reference Diamond2016; Stranahan Reference Stranahan2016).
These events have intensified the social and political action of the Black Lives Matter movement, and have revived a national dialogue about police biases and brutality against African American citizens and discriminatory criminal justice practices more generally. This is all occurring at the same time that a bipartisan coalition of politicians is pushing for criminal justice reform (Grawert Reference Grawert2016; Whack Reference Whack2016). Black Lives Matter activists have protested at campaign events of both Democratic and Republican presidential candidates as political rhetoric about race, crime, and criminal justice in U.S. presidential politics has been re-energized. The result is a revival of framing and counterframing efforts utilizing the symbolic value of these issues to address questions about the relative status of racial groups in the United States.
The major party presidential candidates have distinct positions on crime and policing issues. Republican candidate Donald Trump has explicitly identified himself as pro-police and pro–“law and order” (Alcindor Reference Alcindor2016); he has made claims about “inner-city crime” reaching record levels (Fields Reference Fields2016), and infamously tweeted false and misleading statistics about African Americans and violent crime (Greenberg Reference Greenberg2015). Trump has resisted holding police accountable for unprovoked violence against citizens (Greenberg Reference Greenberg2015) and has openly criticized the Black Lives Matter Movement, saying in September of 2015 that “I think they’re trouble. I think they’re looking for trouble” (Campbell Reference Campbell2015). In November 2015, he commented on an incident where a Black Lives Matter demonstrator had allegedly been assaulted by Trump supporters, noting that “maybe he should have been roughed up” (Johnson and Jordan, Reference Johnson and Jordan2015). Although Trump is not alone in his views, his rhetoric is widely seen as inflammatory. His responses may yield support from White voters who feel they have lost status relative to minorities, and—at the extreme—from White nationalists (Berger Reference Berger2016; Confessore Reference Confessore2016; Klinker Reference Klinker2016). In short, then, it appears that crime and justice rhetoric is once again being used to signal to those concerned about threats to the status quo racial hierarchy.
Democratic candidate Hillary Clinton has been criticized for her association with the 1994 Clinton Crime Bill and mass incarceration, and for her use of racially-coded language when referring to juvenile offenders as “superpredators” during the 1990s (Capehart Reference Capehart2016). Clinton has since repudiated a “tough on crime” policy approach (Beinart Reference Beinart2015), openly called for reform in the aftermath of the 2015 Baltimore protests following the death of Freddie Gray while in police custody (Bouie Reference Bouie2015; Grawert Reference Grawert2016). Her positions on criminal justice reform set her apart from Democratic candidates in previous elections. Her reform efforts align with some of the concerns raised by the Black Lives Matter movement, and she has used injustice frames to rally support among those concerned about the differential treatment of Black citizens by the police and the relative status of Black citizens more generally. In sum, the anecdotal evidence suggests that issues of crime, justice, and policing are highly salient—and fiercely contested—in the 2016 presidential election.
Our goal is to assess the general impact of these recent developments on the 2016 presidential election. Specifically, we are interested in the influence of two distinct factors on prospective political participation and candidate choice in the 2016 election: 1) contemporary citizen contact with the criminal justice system; and 2) perceptions of police injustice. In the following section, we review research and further develop a theoretical model for this purpose.
RESEARCH AND THEORY ABOUT JUSTICE SYSTEM CONTACT, PERCEPTIONS OF POLICE INJUSTICE, AND POLITICAL BEHAVIOR
Criminal Justice Contact
Police stops and questioning have increased in recent decades and are often justified by police authorities as aggressive or zero-tolerance strategies for misdemeanor, morality, and “broken window” law enforcement. These police actions are often referred to as “Terry stops,” in reference to Terry v. Ohio 392 U.S. 1 (1968), which effectively lowered the evidentiary requirement from “probable cause” to “reasonable suspicion” of an infraction. Such stops rarely result in arrests or convictions and disproportionately target economically disadvantaged, minority citizens. At the height of this policy in New York City in 2011, for instance, 685,724 stops were made, with no charges or arrests in 88% of the cases. Footnote 3 More than half of the citizens stopped were Black (53%), with Latinos making up more than half of the remaining stops (34%).
Given the number and the racial and ethnic disproportionality of the stops, it is unsurprising that many perceive them to be unjust (Fagan and Davies, Reference Fagan and Davies2000; Gau and Brunson, Reference Gau and Brunson2010; Silverman and Della-Giustina, Reference Silverman and Della-Giustina2001; Tyler and Waslak, Reference Tyler and Waslak2004; Weitzer and Tuch, Reference Weitzer and Tuch2004a, Reference Weitzer and Tuchb). Among those who feel a stop was unwarranted and/or the product of police bias, the experience can be humiliating and infuriating. For many, police stops may be their primary point of contact with the criminal justice system and their main point of contact with the government.
Work on procedural justice suggests that when the police are perceived to be disrespectful or biased, compliance and cooperation with law enforcement suffers (Brunson and Weitzer, Reference Brunson and Weitzer2009; Tyler and Waslak, Reference Tyler and Waslak2004; Weitzer and Tuch, Reference Weitzer and Tuch2005). One consequence of this is a sense of legal cynicism—a cultural frame in which the police are seen not only seen as unjust and therefore illegitimate, but also as failing at crime prevention, protection, and in providing public safety (Anderson Reference Anderson1999; Carr et al., Reference Carr, Napolitano and Keating2007; Hagan et al., Reference Hagan, Kaiser and Hanson2016; Kirk and Papachristos, Reference Kirk and Papachristos2011; Kirk and Matsuda, Reference Kirk and Matsuda2011; Sampson and Bartusch, Reference Sampson and Bartusch1998).
In addition to aggressive police stops, incarceration has increased dramatically since the 1970s, and has disproportionately affected minority citizens (Rosich Reference Rosich2007; Western Reference Western2006). Recent research documents the devastating personal, familial, social, and economic consequences of incarceration (Clear Reference Clear2007; Clear and Frost, Reference Clear and Frost2014; Mauer and Chesney-Lind, Reference Mauer and Chesney-Lind2002; Western Reference Western2006), resulting in long term exclusion from mainstream civic and public life (Lerman and Weaver, Reference Lerman and Weaver2014; Levy-Pounds Reference Levy-Pounds2013; Pettit Reference Pettit2012; Pettit and Western, Reference Pettit and Western2004; Travis Reference Travis, Chesney-Lind and Mauer2002). Crucially in terms of political enfranchisement, many currently or formerly incarcerated persons are legally prohibited from voting (Manza and Uggen, Reference Manza and Uggen2006).
Perceptions of Police Injustice
One consequence of negative encounters with the police is perceived police injustice (see Gau and Brunson, Reference Gau and Brunson2010; Hagan and Albonetti, Reference Hagan and Albonetti1982; Hagan et al. Reference Hagan, Shedd and Payne2005; Wortley et al., Reference Wortley, Macmillan and Hagan1997). These perceptions can proliferate as a result of both direct and indirect experience. Thus exposure to news coverage of police abuses provokes more critical views of the police, even among Whites who generally have more positive baseline views of the police (Weitzer and Tuch, Reference Weitzer and Tuch2004a, Reference Weitzer and Tuchb). In contrast, racial animus toward Blacks can diminish perceptions of racial injustice by the police and the justice system more broadly (Drakulich Reference Drakulich2015a, Reference Drakulichb; Matsueda and Drakulich, Reference Matsueda and Drakulich2009).
Police practices and (mis)behavior vary both across the race of individuals who encounter the police (e.g., Fryer Reference Fryer2016) and the racial composition of neighborhoods (Brunson and Weitzer, Reference Brunson and Weitzer2009; Fagan and Davies, Reference Fagan and Davies2000; Hagan and Albonetti, Reference Hagan and Albonetti1982; Kane Reference Kane2002; Mastrofski et al., Reference Mastrofski, Reisig and McClusky2002; Smith Reference Smith, Reiss and Tonry1986; Terrill and Reisig, Reference Terrill and Reisig2003; Waddington and Braddock, Reference Waddington and Braddock1991; Weitzer and Tuch, Reference Weitzer and Tuch1999, Reference Weitzer and Tuch2002), with interracial differences in perceptions of the police explained largely by differential experiences with the police (Brunson and Weitzer, Reference Brunson and Weitzer2009; Hagan et al., Reference Hagan, Shedd and Payne2005; Hurst et al., Reference Hurst, Frank and Browning2000; Skogan Reference Skogan2005; Weitzer and Tuch, Reference Weitzer and Tuch2004b). One consequence is a “crisis of legitimacy,” in which perceptions of police bias influence citizen views of police responsiveness as well as their likely behavior serving on juries in cases where the police may have acted in biased ways (Bobo and Thompson, Reference Bobo and Thompson2006, p. 467). High-profile incidents of police misconduct even appear to suppress the public’s willingness to call 911 (Desmond et al., Reference Desmond, Papachristos and Kirk2016).
Perceptions of police actions are distinguishable from general affective evaluations of the police. In the recent context of the Black Lives Matter movement and proposed police reform, some politicians and voters have been accused of being motivated by a dislike or hatred of the police (Greer Reference Greer2015), while other have expressed strong affective endorsements of the police (Nuzzi Reference Nuzzi2016). One interpretation is that such views represent an expressive concern: that “hating” the police, or, alternatively, “loving” the police, and, indeed, voting on the basis of it, is an expression of one’s identity (Brennan and Hamlin, Reference Brennan and Hamlin1998).
Political Behavior
Our collective responses to rule violations are fundamental to the operation of society, and our justice system represents one of the largest and most impactful public institutions. Thus, there are good reasons to suspect that contact with the criminal justice system as well as broader beliefs about the fairness of the system are likely to impact political engagement, participation, and candidate choice.
There are several key mechanisms by which criminal justice contact and perceptions of police injustice may be related to political behavior. A line of political thought with a long history suggests that government institutions matter to the development of its citizens’ political behavior. Alexis de Tocqueville (Reference de Tocqueville and Goldhammer2004[1840]), for example, wrote of the importance of jury service in educating citizens about the purposes of the justice system. More recent work similarly suggests that people’s interactions with government representatives and institutions are important mechanisms by which citizens learn about the political process and civic matters more broadly (Landy Reference Landy1993; Lipsky Reference Lipsky1980; Soss Reference Soss, Schneider and Ingram2005). Scholars like Joe Soss (Reference Soss, Schneider and Ingram2005) have found that direct contact with government agents plays a particularly important role in shaping how poorer and more marginalized people perceive government, writ large.
More recently, Amy Lerman and Vesla Weaver (Reference Lerman and Weaver2014) tested the relationship between justice system contact and political behavior using the National Longitudinal Study of Adolescent Health and the Fragile Families and Child Wellbeing Study. Lerman and Weaver argue that contact with agents of law enforcement is unlike contact with other government bureaucracies and suggest that this contact can have profound effects on views of the government, the civic body, and the place of the citizen within it. In fact, they found that citizens often form their opinions about the entire government based almost exclusively upon their personal experience with the criminal justice system. As a result, these citizens often see the focus of government to be the exercise of control, not the provision of services or the protection of citizens. This can result in the kind of legal cynicism noted above (see especially Kirk and Papachristos, Reference Kirk and Papachristos2011; Hagan et al., Reference Hagan, Kaiser and Hanson2016).
Lerman and Weaver’s research uncovered weak, negative relationships between arrest, conviction, imprisonment, and “low-impact” political activities like contacting a representative; but the negative effects were much stronger on the likelihood of registering to vote or voting in the previous presidential election. The magnitude of the effect on political behavior increased with the punitiveness of the justice system encounter, and was robust to controls for differences in socioeconomic status, demographic factors, and even respondent drug use. These results suggest that the experience of being arrested and/or punished by the justice system may lead citizens to believe that they are part of a stigmatized class that has been symbolically severed from the polity. The implication is that such contact is marginalizing and exclusionary, driving citizens away from political engagement and participation.
However, Lerman and Weaver (Reference Lerman and Weaver2014) also discovered an unexpected (but small in magnitude) positive relationship between simply being questioned by police and political activities like engagement in civic organizations or contacting a representative. Consistent with this “unexpected” finding, Victor Rios (Reference Rios2011) finds that disadvantaged youth who have experienced police harassment become more politically engaged in order to protest perceived injustice. In sum, it appears that very punitive encounters with the criminal justice system such as incarceration may drive individuals away from political participation, while less impactful actions such as police stops may actually increase engagement and participation.
Perceptions of police injustice may be relevant to political behavior for at least two reasons. The first draws on an idea from the collective action framing perspective, which describes how actors seek to frame social problems for political purposes (Benford and Snow, Reference Benford and Snow2000; Gitlin Reference Gitlin1980; Goffman Reference Goffman1974). The goals of such framing efforts are to build consensus around the causes of and solutions to a problem, but also to motivate action around an issue. To spur such action, frames will emphasize the severity and urgency of the negative consequences of the issue, but they may also emphasize a moral imperative—the propriety or rightness of action (Benford Reference Benford1993). As such, viewing a problem as an injustice rather than a misfortune can be a powerful motivating frame for an issue (Snow and Benford, Reference Snow, Benford, Morris and Mueller1992; Turner Reference Turner1969). Thus, perceptions of police behavior not just as unfortunate but as unjust may motivate both political participation and support for political candidates whose statements conform to such a justice frame.
In line with this perspective, anecdotal evidence suggests that the death of Michael Brown, viewed by many as the result of an unjust police action, prompted increased turnout for local elections in Ferguson, Missouri (Eligon Reference Eligon2015). In addition, research shows that perceptions of police injustice may be related to candidate choice. Using a 2006 political survey, Ross Matsueda and colleagues (2011) found that perceptions of police injustice were associated with a reduced likelihood of the respondent reporting that they would vote for George W. Bush over Bill Clinton in a hypothetical election, even after controlling for political party and ideological identification.
Another possibility deserves further consideration: that pro-police rhetoric or perceptions of the police as fair may also motivate political participation and increase support for candidates espousing similar rhetorical and perceptual frames. As noted above, some politicians have used references to crime to signal and attract voters who are uncomfortable with perceived changes in the relative standing of racial groups (Beckett and Sasson, Reference Beckett and Sasson2004; Edsall and Edsall, Reference Edsall and Edsall1991; Hagan Reference Hagan2010; Mendelberg Reference Mendelberg1997; Tonry Reference Tonry2011). Recent work suggests this connection between racialized feelings and perceptions of crime and (in)justice persists (Drakulich Reference Drakulich2015a, Reference Drakulichb) and polarizes voters into two camps: those who perceive the police to be fair, impartial protectors of law-abiding citizens; and those who perceive the police (and the broader justice system) to be prejudicially biased and punitively discriminatory against minority citizens (Peffley and Hurwitz, Reference Peffley and Hurwitz2010).
This raises a final important consideration. If crime and justice rhetoric is frequently employed as a signal to voters concerned about threats to the existing racial order, this suggests that these views may act, in part, as a proxy for racial feelings and attitudes. Indeed, abundant evidence demonstrates a connection between racial feelings, attitudes, and perceptions of crime and justice, including perceptions of crime as a problem, explanations for racial disparities in contact with the criminal justice system, and support for punitive responses to crime (Bobo and Johnson, Reference Bobo and Johnson2004; Drakulich Reference Drakulich2015a, Reference Drakulichb; Johnson Reference Johnson2001, Reference Johnson2008, Reference Johnson2009; Matsueda and Drakulich, Reference Matsueda and Drakulich2009; Soss et al., Reference Soss, Langbein and Metelko2003; Unnever and Cullen, Reference Unnever and Cullen2007, Reference Unnever and Cullen2010; Wozniak Reference Wozniak2016). In addition, prior work has also identified a direct role for racial feelings and attitudes in voter choice (McElwee and McDaniel, Reference McElwee and McDaniel2016).
Key to this discussion is an understanding of the often hidden or implicit role of racial anxieties and racism in contemporary politics. Beginning in the Civil Rights era, open expressions of racial antipathy towards Blacks declined and were replaced by a norm of equality (Mendelberg Reference Mendelberg2001). Contemporaneously, new racial logic emerged that deemphasized racial group distinctions in favor of individualism, while ignoring that not all individuals have the same opportunities given historical group disparities (Feagin Reference Feagin2009; Jackman and Muha, Reference Jackman and Muha1984; Schuman et al., Reference Schuman, Steeh, Bobo and Krysan1997). Scholars have described this new racial logic—which serves to maintain inequalities by rejecting group demands—as symbolic, laissez-faire, or colorblind racism (Bobo Reference Bobo, Krysan and Lewis2004; Bobo and Kluegel, Reference Bobo, Kluegel, Tuch and Martin1997; Bobo and Smith, Reference Bobo, Smith, Katkin, Landsman and Tyree1998; Bobo et al., Reference Bobo, Kluegel, Smith, Tuch and Martin1997; Bonilla-Silva Reference Bonilla-Silva2010; Kinder Reference Kinder1986; Kinder and Sears, Reference Kinder and Sears1981; Sears Reference Sears, Katz and Taylor1988). As a result, politicians have relied on implicit racial messaging and racial code words (such as “law and order”) to appeal to White voters concerned about relative racial group positions without overtly referencing race (López Reference López2014; Mendelberg Reference Mendelberg2001). Thus, it is possible that views of the police serve as proxies for racial feelings and attitudes. Ross Matsueda and Kevin Drakulich (Reference Matsueda and Drakulich2009), for example, find that perceptions of police injustice are associated with a variety of political policy positions through an association with racial resentment. This measure represents one attempt to capture this new racial logic (Henry and Sears, Reference Henry and Sears2002).
Research Questions
In light of the theory and research evidence reviewed above, we explore several key questions about the role of criminal justice contact and perceptions of police injustice in the 2016 U. S. presidential election.
The first question involves the impact of criminal justice contact on political participation. Negative police contact, especially if it is perceived to be unjust, may motivate political participation with the goal of changing the system. However, criminal justice contact that results in stigmatic marginalization or exclusion may deter individuals from participating. Our hypothesis is therefore that police stops motivate political action, while incarceration decreases political participation. A second related question focuses on the impact of police contact on candidate choice. Our hypothesis is that experience with police stops motivates support for Democratic presidential nominee Hillary Clinton. In this election, there were clear distinctions between the criminal justice policy positions of the major party candidates. Both major Democratic candidates (Hillary Clinton and Bernie Sanders) advocated for criminal justice and police reform. The major Republican candidates generally, and the eventual nominee Donald Trump in particular, tended to express support for the police rather than propose police reform.
The third and fourth questions involve the impact of perceived police injustice on political participation and candidate choice. Once again, several hypotheses present themselves. Perceptions of injustice may act as a powerful motivational frame inspiring political participation and support for candidates who promote justice-relevant reforms. However, views of the police as fundamentally fair may serve as a rallying cry for those on the other side of the political spectrum, especially in an election year in which the police are perceived to be unfairly criticized. In other words, strong views either way may drive both groups to the voting booth, but to vote for different candidates. This suggests a potentially strong effect on candidate choice, but also the possibility of a confounding effect on participation.
A fifth question asks whether the role of perceived police injustice can be distinguished from a more general affective evaluation of the police, or in simplistic terms: “hatred of the police.” Given our theoretical perspective on why perceptions of injustice matter, we expect the role of perceived injustice to be both independent of, and more important than, affective evaluations of the police.
A final question, based on the long history of law and order rhetoric being used as a racial code word or dog whistle, asks whether the effect of these feelings about attitudes or the police are independent of measures of feelings and attitudes about race.
METHODOLOGY
Data
To test the impact of recent experiences with and attitudes toward police and the criminal justice system, the American National Election Studies agreed to include questions about criminal justice contact and perceptions in their 2016 Pilot Survey (ANES 2016). The ANES collected surveys from 1200 respondents and included weights designed to make the sample representative of the larger population of U.S. citizens age eighteen or older on the basis of age, gender, race-ethnicity, education, region, and party identification. Footnote 4 The survey was conducted over the internet drawing on respondents from an existing panel. Footnote 5 The survey was conducted in late January 2016, just before primary voting began in early February.
Measures
Political Behavior
The analysis seeks to explain two forms of political behavior: participation and choice. Participation is captured by two indicators. The first is a measure of engagement, asking respondents to rate themselves on a scale reflecting how frequently they “follow what’s going on in government and public affairs” (ranging from “hardly at all” to “most of the time”). The second measure captures intended behavior, the self-assessed percent chance, on a scale from zero to 100 that the respondent will vote in 2016. Table 1 describes the sample on key variables. Footnote 6 The majority of respondents—78%—reported that they followed politics some or all of the time, and respondents on average reported a 79% chance that they would vote in the 2016 Presidential election. Table 1 also presents descriptive means and significance tests for two different sub-group comparisons: Black versus White respondents, and Trump versus Clinton supporters. There were no racial differences in political engagement or participation, nor were there any differences between Trump and Clinton supporters in these measures.
Table 1. Descriptive Information for the Sample on Key Measures
*** p<.001, **p<.01, *p<.05.
Political choice is also captured with two different questions. The first taps a preference for the Republican nominee, Donald Trump. Respondents were asked “regardless of whether you will vote in the Republican primary this year, which Republican candidate do you prefer?” Respondents were provided a list of the top nine candidates at that point plus the options “another Republican candidate” or “none” (about 28% chose this option). The resulting variable is coded as 1 for those who indicated a preference for Trump and 0 for those who indicated a preference for a different Republican nominee. Of those who expressed support for some Republican candidate, about a third chose Trump, though this was a significantly more popular choice among White versus Black respondents.
The second question moves to the general election. Footnote 7 Respondents were asked “If the 2016 presidential election were between Hillary Clinton for the Democrats and Donald Trump for the Republicans, would you vote for Hillary Clinton, Donald Trump, someone else, or probably not vote?” The resulting variable was coded 1 for those indicating they would vote for Trump in this case, and zero if they planned to vote for Clinton, someone else, or not vote.
The results suggest polling was close at this point in time (January 2016): around 39% expressed support for Clinton versus 37% for Trump, with 11% voting for a third party candidate, and 14% not voting. Trump was again more popular among White relative to Black respondents. Because of the large number of potential general election matchups at this point in the primary season, respondents were randomly selected to answer only two. As a result, about 40% of the sample (487 respondents) were asked this question. With the exception of those who were randomly chosen not to answer the question about a Trump-Clinton matchup, there is no missing data for any of these questions.
Criminal Justice Contact
The survey included questions about a variety of different kinds of contact with the criminal justice system. Many, such as whether the respondent had been arrested for, charged with, or convicted of a crime were highly interrelated. We focus on two different points of contact. The first measure asked respondents whether they have been stopped and questioned by a police officer. The second asked if the respondent had spent one or more nights in jail or prison. Respondents were randomly selected to be asked these questions either in reference to the last twelve months or ever. For each model and outcome, exploratory interactions were added between a dummy variable identifying those respondents who were asked about contact in the last twelve months versus ever and each of the two measures of contact. In no case was the interaction statistically significant, suggesting that despite the different time periods, criminal justice contact was related to the outcomes in similar ways. Missing data was rare for these questions: one respondent skipped the question about police stops and two skipped the question about time in jail or prison. Black respondents were more than twice as likely to report being stopped and questioned by the police in the last year.
Perceptions of the Police
Two measures were included to reflect respondent perceptions of the police. We distinguish a basic perception of police fairness from an affective evaluation of the police, which we developed from the literature on perceived injustice and legal cynicism. The affective measure employs a “thermometer scale” question, a format common in ANES and other political surveys. Respondents were asked to rate their feelings toward a list of individual persons and groups on a scale ranging from 0 to 100, with 0 representing very cold and unfavorable feelings and 100 representing very warm or favorable feelings. Respondents were asked specifically how warmly or coldly they felt toward “the police.” On average, respondents reported feeling modestly warm toward the police—67 on the 100-point scale. White respondents and Trump supporters reported feeling significantly more warmly toward the police than Black respondents and Clinton supporters.
The second measure taps into a very different evaluation of the police, one focused specifically on whether the police act in discriminatory ways. Specifically, the variable is captured as factor scores from a confirmatory factor analysis with three indicators. Footnote 8 The first indicator is a question asking respondent, on a seven-item scale, whether they believe that “in general, the police treat Whites better than Blacks, treat Blacks better than Whites, or treat them both the same.” Just over 60% of respondents reported believing that Whites were treated better than Blacks by the police. Clinton supporters and Black respondents were significantly more likely than Trump supporters and White respondents to report this—nearly 90% of Clinton supporters compared to less than 40% of Trump supporters indicated that Whites were treated better than Blacks by the police.
The second and third indicators are based on a series of questions about specific police behavior. Respondents were asked (on five-item scales) how often they think police officers stop people on the street without good reasons and how often they use more force than is necessary. They were asked each of these questions twice: once concerning how the police act when dealing with White people, and once concerning how the police act towards Black people. About half the sample, respectively, felt that Black people were stopped by the police and subject to unnecessary police force, compared to less than 15% of the sample reporting the same for the treatment of White people. Interestingly, there was widespread agreement that Whites are not often stopped without reason or subjected to unnecessary force, but Clinton supporters and Black respondents were significantly more likely to feel that Blacks experienced these kinds of police mistreatment. Two simple measures were created—one for unnecessary stops and one for unnecessary force—as the difference between perceptions of police behavior towards Blacks and Whites, with positive values representing those respondents who believed Whites were treated better and negative values reflecting those who perceived Blacks as treated better than Whites. Missing data were again rare: 1 respondent did not answer the questions about unjustified police stops, 1 skipped the general question about police discrimination, and none skipped the question about police force.
Racial Attitudes
Two measures were added to capture racial attitudes and feelings. The first is a simple measure of racial animus against Blacks relative to Whites based on the same affective thermometer scale used to capture feelings toward the police. The measure is the difference between how coldly the respondent feels toward Blacks versus Whites, with higher values indicating respondents who felt more coldly towards Blacks than Whites. The average respondent reported feeling about 5 degrees more warmly toward Whites than they did toward Blacks, with both racial groups showing in-group preferences. Clinton and Trump supporters did not differ in their feelings toward Whites, but Trump supporters felt significantly more coldly toward Blacks than did Clinton supporters.
The second measure captures racial resentment, a dimension of “symbolic racism” widely used in prior work (Henry and Sears, Reference Henry and Sears2002; Kinder Reference Kinder1986; Kinder and Sears, Reference Kinder and Sears1981; Sears Reference Sears, Katz and Taylor1988). The perspective suggests that because Whites view African Americans as inferior on key “Protestant Ethic” dimensions like being hard-working, they resent perceived harms resulting from these deficiencies, including African Americans getting “unfair” special access to jobs or educational opportunities. The measure is captured as factor scores from a confirmatory factor model with four indicators. A majority of the sample, nearly two-thirds, believed that Blacks should overcome prejudice and work their way up without “special favors.” Around half the sample disagreed that slavery and discrimination created conditions that remained significant barriers for lower-class Blacks, disagreed that Blacks had gotten less than they deserved, and agreed that inequalities would be solved if Blacks tried harder. However, White respondents were more likely to agree with these statements than were Black respondents, and Trump supporters were especially likely to agree with each of the racial resentment statements. Missing data were again rare: 3 respondents total skipped any of the racial resentment questions, and 4 skipped the thermometer scales against either Blacks or Whites.
Political Identity and Beliefs
We control for three measures of political sentiment when examining candidate choice: identification as conservative, identification as Republican, and belief in a limited government. One case is missing for Republican Party identification as are five for conservative ideology. Regarding preference for a limited government, 103 respondents chose the option “I haven’t thought much about this.” To avoid reducing the N, these cases were recoded into the middle-category—those who did not strongly agree with either statement. However, models were also run dropping these cases and did not differ substantively from the reported models.
Demographic and Biographical Controls
We control for a variety of standard demographic and biographical characteristics: gender, age, marital status, education, income, employment status, and race-ethnicity. There were no missing data for any of these questions other than income, for which 147 respondents chose not to answer. In the interest of not dropping cases, all of the missing cases were mean-replaced. However, models were also run dropping these cases and did not differ substantively from the reported models. Table 2 presents basic descriptive information on the sample for all of these control variables.
Table 2. Descriptive Information for the Sample on Control Measures
***p<.001, **p<.01, *p<.05.
Methods
Reported effects are estimated with survey-weighted generalized linear models run using the “survey” package in R (Lumley Reference Lumley2014; R Core Team 2016). The four category measure indicating political engagement is treated as ordinal and modeled using an ordinal logit. The percent likelihood of voting in 2016 is treated as continuous. The measures of vote choice are dichotomous and are fit with a quasibinomial family logit link. Cases missing data on any of the outcomes or key substantive variables (contact, police, and racism) were dropped (as discussed above, missing data on these variables were infrequent), as were those for control variables with small numbers of cases missing. In total, only fourteen of the 1200 cases were dropped before the analyses.
The survey captures political preferences and inclinations from a point in time prior to the Presidential election itself, and as with all voting polls the respondent preferences may change and they may act differently on November 8. The survey was conducted between January 22 and 28, 2016, prior to the major political parties’ nomination of candidates. According to the poll aggregation website realclearpolitics.com, on the first day of the survey Donald Trump led the Republican field with about 35% of voters expressing a preference for him, versus about 19% for Ted Cruz, and 11% for Marco Rubio, the second and third most popular candidates at that time (Real Clear Politics 2016). Among Democrats, 51% supported Hillary Clinton while 38% favored Bernie Sanders (Real Clear Politics 2016). Although asked in a slightly different fashion, the question about candidate preference from the ANES survey reflects this same pattern. Among those who indicated a preference for one of the Republican candidates, the top choice with 34% of support was Donald Trump followed by Ted Cruz at 14%, and then Marco Rubio at 7%. Among those who expressed a preference for Democratic candidates, 45% favored Clinton versus 36% for Sanders. In other words, although the survey occurred early in the election process, the survey data appear representative of voting preferences at that point in time.
The data are cross-sectional, and as such we cannot definitively determine the causal direction of reported relationships. It is certainly possible, for instance, that feelings of coldness toward the police or perceptions of police injustice are formed as a consequence of following political reporting or by adopting the positions of a candidate chosen for other reasons (although, in fact, we explore the possibility that racial animus rather than perceptions of the police are a motivating factor). However, given our theoretical reasons for expecting an effect on political outcomes and given the dearth of direct research on these questions, we feel there is substantial value in examining the results.
Results
The results are presented first for models predicting political engagement and participation, and then for candidate choice.
Political Engagement and Participation
Table 3 presents two models predicting political participation and engagement. The first three columns present coefficients, standard errors, and odds ratios for an ordered logit model predicting the measure of political engagement—how closely the respondent follows politics. Among the control variables, men, older persons, single persons, and those who are employed, earn more, and have more years of education are all more likely to report that they follow politics most of the time. Footnote 9 Race and ethnicity are not associated with the likelihood of following politics.
Table 3. Coefficients from Models Predicting Political Engagement and Likelihood of Voting in 2016
* p<.05, **p<.01, ***p<.001 (2-tailed). N=1186.
Follow politics employs survey-weighted ordinal logit, Likely to vote uses survey-weighted generalized linear model.
Both criminal justice contact and perceptions of the police are relevant to political engagement, although the results suggest very different substantive reasons for each of the effects. First, those who reported having been stopped by the police were substantially more likely to report that they follow politics most of the time—the odds of following politics closely are 1.8 times higher for those who have been stopped. Footnote 10 On the other hand, those who have spent a night in jail or prison were substantially less likely to report following politics most of the time.
Perceptions of and affective feelings toward the police also had divergent effects on the likelihood of following politics. Those who felt more coldly toward the police were less likely to report following politics most of the time, while those who perceived the police as biased towards Blacks relative to Whites were more likely to report following politics most of the time. Thus in general, a simple affective dislike of the police seems to drive people away from politics, while a perception of the police as acting in unjust ways drives greater interest. Notably, both of these effects are relatively modest.
However, as we suggested in the discussion of research questions, it is likely that people with different political orientations will be differentially motivated by these perceptions. Those who perceive the police as treating Blacks and Whites the same, or even those who believe—in the spirit of the view that anti-White bias is now more prevalent than anti-Black bias (Norton and Sommers, Reference Norton and Sommers2011)—that the police treat Whites worse than Blacks, may also be motivated to be more politically engaged. In this way, the same sense of injustice driving political engagement and participation may occur at both ends of the scale but for different groups. Conservative respondents may be more likely to be driven by a feeling that the police are biased against Whites while liberal respondents may be more motivated by a sense that the police are biased against Blacks. Similarly, while conservative groups have accused liberals of being motivated by anti-police feelings, pro-police messages have been common at Trump speeches and rallies (Alcindor Reference Alcindor2016).
In fact, both measures of perceptions of the police have significant interactions with conservative identification. The top two panels of Figure 1 presents predicted values from this interaction. Footnote 11 The top panel suggests that feeling more warmly toward the police is associated with an increased attention to politics, but only among those who identify as more conservative. Footnote 12 Feeling warmly (or, conversely, coldly) toward the police does not seem to be associated with following politics among those who identify as more liberal. The middle panel suggest the opposite pattern for perceptions of police injustice. For those who identify as more conservative, perceptions of the police as unjust toward Whites versus Blacks do not appear to be associated with following politics. For those who identify as more liberal, however, bigger perceptions of police bias against Blacks are associated with a larger likelihood of following politics closely.
Fig. 1. Predicted levels of political engagement and participation from interaction of perceptions of police and political ideology.
The final two columns in Table 3 present coefficients and standard errors from a weighted generalized linear model predicting the self-assessed percent likelihood that the respondent will vote in the 2016 election. Older persons, the employed, and those with more years of education and higher family income all reported a greater likelihood that they would vote in the election. Hispanic respondents tended to report a lower likelihood of voting.
Both forms of criminal justice contact appear important to the likelihood of voting, again in opposite directions. Just as being stopped by the police is associated with more political engagement, it is also associated with a higher estimated likelihood of voting—those who experienced police contact were 8% more likely to vote. The experience of having spent at least a night in jail or prison, on the other hand, appears to be more marginalizing or exclusionary, and is associated with about a 14% lower chance of voting.
Feeling coldly toward the police or perceiving the police to be racially biased, on the other, hand, are not directly significantly related to the likelihood of voting. As with political engagement, however, it may be that the effect depends on ideological identification. Interactions suggest that the feeling coldly toward the police is not strongly associated with the likelihood of voting for conservatives or liberals. However, perceptions of police bias are associated with both, and in opposite directions (b = -2.3, p<.001). As the bottom panel of Figure 1 illustrates, for liberals, those who believe the police act in biased ways toward Blacks were more likely to say they would vote in 2016. For conservatives, it was those who believed that the police were more biased against Whites who reported a higher likelihood of voting. Thus, perceptions of police injustice do appear related to people’s political participation, but the direction of the effect depends on one’s political ideological identification.
Candidate Choice
Table 4 presents two models of candidate choice: the choice of Trump among the Republican candidates in the primary, and the choice of Trump relative to Clinton in the general election. The first three columns present coefficients, standard errors, and odds ratios from a model of a preference for Donald Trump among the Republican presidential candidates. Footnote 13 Education is among the strongest predictors of support for Trump, as those with fewer years of education were substantially more likely to express support for Trump among the Republican candidates. Relatedly, those who are unemployed were more likely to support Trump. In terms of race, Black and Hispanic voters were less likely than White voters to support Trump. All of this is consistent with the general narrative that Trump is pulling substantial support from Whites who have less education and experience greater unemployment (Thompson Reference Thompson2016). However, even after accounting for these associations, those who perceive greater bias against Blacks were much less likely to state support for Trump in the primary—conversely, of course, this means those who believe the police may be biased against Whites were more likely to support Trump. Notably, neither contact with the criminal justice system nor more affective negative feelings toward the police appear to be relevant to support for Trump.
Table 4. Coefficients from Models Predicting Candidate Preference (survey-weighted, quasibinomial-family logit-link)
*** p<.001, **p<.01, *p<.05
The next three columns in Table 4 present a second model for Trump support in the primary, this time adding two measures of racial attitudes and feelings toward Blacks. Explicit animus toward Blacks—those who openly admit to feeling more warmly toward Whites than Blacks—is strongly and positively associated with support for Trump. Similarly, racial resentment of Blacks is also strongly and positively related to a preference for Trump. Interestingly, including these measures of racial attitudes have two important impacts on other estimates in the model. First, the size of the estimate for the difference between Black and White respondents is cut in half and drops below conventional significance. In other words, racial animus and racial resentment appear important to explaining the different preferences among White and Black voters in the Republican primary. Second, the estimate for the effect of perceived police injustice is also cut nearly in half and drops below conventional significance. In other words, if racial attitudes were ignored, it would appear that people’s perceptions of police injustice was important in distinguishing Trump supporters from those who support other candidates. However, the full results suggest such perceptions may be acting as a proxy for racial attitudes among these voters. In other words, when choosing among the Republican candidates, perceptions of the police were less important than attitudes toward Blacks.
The third set of columns switches the focus to the general election, contrasting those who indicate they would likely vote for Donald Trump in a general election against Hillary Clinton versus those who would vote for Clinton, a third-party candidate, or no one. Footnote 14 Not surprisingly, partisanship is an important factor in the general election, with those who identify most with the Republican Party expressing the strongest support for Donald Trump in a match-up with Hillary Clinton. Relatedly, “small government” advocates are more likely to support Donald Trump—an interesting finding given that Donald Trump, in contrast to previous Republican candidates, has not made “small government” a campaign issue and has in fact proposed significant government expansions (Appelbaum Reference Appelbaum2016; French Reference French2016). However, even after political party identification is controlled for, perceptions of the police as biased against Blacks is strongly and negatively associated with opposition to Donald Trump. Once again, neither contact with the criminal justice system nor more affective negative feelings toward the police appear to be relevant to support for Trump.
The final set of models adds the two measures of racial attitudes. Although explicit animus is not significantly associated with candidate choice in the general election, those respondents who scored high on the racial resentment measure were substantially more likely to express support for Donald Trump. Including this measure results in a substantial drop in the direct effect of a preference for limited government, complementing prior work suggesting that individualistic or anti-big government ideological positions often serve as a mask for racial resentments (Bobo et al., Reference Bobo, Kluegel, Smith, Tuch and Martin1997; Bonilla-Silva Reference Bonilla-Silva2010; Kinder and Sears,Reference Kinder and Sears1981; Jackman and Muha, Reference Jackman and Muha1984). There is also a modest decrease in the direct effect of perceptions of police injustice, though, unlike the primary model, perceptions of police injustice remain an important predictor of opposition to Donald Trump in the general election.
DISCUSSION AND CONCLUSION
During the Civil Rights era, which was a particularly turbulent moment in American political history, intersecting issues of race, crime, policing, and justice had great salience for political behavior. Specifically, issues of crime, justice, and the police became important sites for conflict about the relative status of racial groups. Outrage over the unfair treatment of Black citizens by the police set off massive protests across the country, ultimately contributing to the passage of several landmark pieces of civil rights legislation. At the same time, White politicians reframed these protests as criminal to delegitimize them, used law and order rhetoric as a signal to rally White voters uncomfortable with the changes posed by the Civil Rights Movement, and simultaneously used law enforcement officers to try to suppress political activity by civil rights protestors.
Thus, claims of unjust treatment at the hands of the police were used to motivate action to attempt to address problems of racial inequality. On the other side, “law and order” rhetoric as well as the actual institutions of the criminal justice system served as tools in efforts to maintain White superiority. These tools play a critical role in the new racial logic justifying White privilege, one that de-emphasizes group distinctions and instead favors individualistic explanations for inequalities (Bobo et al., Reference Bobo, Kluegel, Smith, Tuch and Martin1997; Bonilla-Silva Reference Bonilla-Silva2010; Jackman and Muha, Reference Jackman and Muha1984). In this model, the “law and order” rhetoric serves as an implicit racially-coded appeal to rally those concerned about challenges to the racial order without mentioning race or legitimizing the importance of the racial structure (López Reference López2014; Mendelberg Reference Mendelberg2001). Contemporary criminal justice institutions also serve an important role, mirroring some of the functions of older systems of racial confinement and subjugation like slavery and the Jim Crow system, but without the overtly-stated racial motivations and goals of those institutions (Alexander Reference Alexander2010; Wacquant Reference Wacquant, Pratt, Brown, Brown, Hallsworth and Morrison2005).
The findings presented in this paper reveal the continued relevance of the intersection of race, crime, justice, and policing to American political behavior. Just as in that earlier era, though, the relevance is far from simple or one-dimensional. The results suggest, for instance, that the impact of contact with the criminal justice system depends on the form and punitiveness of contact. While incarceration is a severe, marginalizing, and exclusionary form of contact that tends to drive citizens away from political engagement and participation, the experience of police stops and questioning appears to do the opposite, by motivating citizens to become politically engaged and to vote.
We further find perceptions of police injustice to be an important factor both in political participation and candidate choice. The impact of these perceptions, however, differs along lines of political ideology. We find that perceptions of police injustice against Blacks motivate liberals to vote, and to vote for the Democratic presidential candidate, Hillary Clinton. Perceptions of the police as fair or even biased against Whites, on the other hand, motivates turnout among conservatives and greater support for the Republican candidate, Donald Trump.
Notably, we find that perceptions of the police are not simply expressions of affective evaluations of the police. Moreover, we find that affect is not significantly associated with candidate choice, or, for the most part, with turnout. For example, feeling warmly toward the police only motivated political engagement among conservatives. Alternatively, we found no effect of feeling warmly or coldly toward the police among liberals.
The simplest and most general conclusion from our analysis is that both criminal justice contact and perceptions of police injustice matter for political behavior in 2016. This indicates that researchers in political science, political sociology, and political criminology should take these race, crime, justice, and policing issues seriously when trying to understand contemporary political behavior.
One finding is particularly troubling. Consistent with Lerman and Weaver (Reference Lerman and Weaver2014), we find that certain forms of contact with the criminal justice system marginalize and exclude citizens from political engagement. Those who had spent a night in jail or prison were much less likely to report following politics and rated their likelihood of voting in 2016 as much lower. This complements and extends work on the most direct effect of incarceration—felon disenfranchisement—in its critique of the corrosively antidemocratic impact of the carceral state (Manza and Uggen, Reference Manza and Uggen2006; Weaver and Lerman, Reference Weaver and Lerman2010).
Finally, the influence of perceptions of police injustice has both encouraging and worrisome implications. On the one hand, consistent with the broader narrative from the Civil Rights era, perceptions of the police as disproportionately mistreating Black citizens motivates political engagement, participation, and support for the presidential candidate explicitly proposing police reforms. On the other hand, perceptions of the police as unbiased or as biased against Whites motivates engagement, participation and support for the candidate who has been unconditionally supportive of police and antagonistic towards the accountability concerns of the Black Lives Matter movement. We have found evidence that perceptions of police injustice in the latter group may be a proxy for anti-Black sentiment.
For fifty years, politicians have used racialized allusions to crime and disorder in order to prey on voters’ fears, secure electoral gains, and enact punitive criminal justice policies. The same types of conflicts between police and communities of color that sparked the most controversial and polarizing events during the Civil Rights Movement are occurring once again across the country. Despite recent progress toward bipartisan criminal justice reform in congress and numerous state legislatures, this study provides further evidence that the same racialized tensions that polarized the electorate in the 1960s are still shaping national politics today.
