The Racial Implications of Voter Identification Laws in America

Abstract

Over 40 states have considered voter identification laws in recent years, with several adopting laws requiring voters to show a valid ID before they cast a ballot. We argue that such laws have a disenfranchising affect on racial and ethnic minorities, who are less likely than Whites to possess a valid ID. Leveraging a unique national dataset, we offer a comprehensive portrait of who does and does not have access to a valid piece of voter identification. We find clear evidence that people of color are less likely to have an ID. Moreover, these disparities persist after controlling for a host of relevant covariates.

Keywords

voter ID laws racial and ethnic politics

Introduction

Early challenges to voter identification laws equated them with poll taxes, given it costs money to obtain identification through the department of motor vehicles (Shanton, 2010). Presented with disproportionately negative implications for Blacks and Latinos, federal judges stayed laws in Georgia, Texas, and South Carolina via Section 5 of the Voting Rights Act (VRA). However, the U.S. Supreme Court ruled that the risk of fraud outweighed burdens to voters. The decision in Crawford v. Marion County (2008), together with the invalidation of Section 5 preclearance under Shelby County v. Holder (2013), rendered Section 2 of the VRA the principal means to federally challenge ID laws. State constitutional challenges have been inconsistent, leading to a patchwork outcome. Activists have mounted federal challenges in Texas, North Carolina, Kansas, and Wisconsin, among other states, a controversy that heated up in the approach to and aftermath of the 2016 election.

Although challenges to the validity of voter ID laws often draw on data collected and evaluated by expert witnesses, case law has developed largely in the absence of comprehensive research on differential access to ID among population subgroups. Instead, analyses for these cases are limited to the data in that state, and to bivariate relationships of haves and have nots, raising questions about the role of other variables, like education, income, and age, in impacting access to an ID. Finally, analyses connecting ID laws to turnout suffer from data limitations, and findings are contested. Importantly, we argue that research overly focused on turnout misses the point of de facto disenfranchisement, on which we elaborate below. We draw on a unique, comprehensive dataset to describe the nature and scope of differential access to ID among racial subgroups. We situate our analysis within the legal and social science framework at play around voter ID laws to centralize questions of access to the franchise. We do this to reinforce the argument that racial differences in access to ID required to vote result from historical institutional racial exclusion. From this vantage point, questions of impact on aggregate turnout are secondary. Turnout rates are never universal, and rise and fall every cycle with competitiveness and quality of candidates. Put simply, the contest over voter ID laws is one of power, access to democracy, and the value of civic voice. As such, who has access to documents which allow you to vote is of primary importance.

Relevant Literature

Concerns over voter fraud propelled the popularity of voter identification laws after the 2000 election and the 2002 passage of the Help America Vote Act (HAVA). Despite the fact that there are few documented instances of the kind of in-person fraud that voter ID laws would prevent, they have withstood constitutional scrutiny when confronted with the demand for electoral integrity (Minnite, 2007; Stoughton, 2013). Despite claims to electoral integrity, scholars demonstrate that partisan motivations enhance the popularity of these laws (Bentele & O’brien, 2013; Biggers & Hanmer, 2017; Hicks, McKee, Sellers, & Smith, 2015). Investigating the conditions favorable to their adoption, Hicks et al. (2015) find that competitive legislatures where Republicans have a slight edge are most likely to pass ID requirements. Republicans strategically leverage such laws to support turnout among their base while undercutting the turnout of Democratic voters (Grossmann & Hopkins, 2015).

Eroding turnout among Democratics is sometimes crafted directly into ID laws. In Texas, hunting and gun permits, which Whites are statistically more likely to possess, are legitimate forms of ID but social service cards, more often held by Blacks and Latinos, are not (Bachu & O’Connell, 1995; Shanton, 2010). The passage of Alabama’s ID law was accompanied by the closure of nearly half of the state’s Department of Motor Vehicles (DMV) locations, with most closures in disproportionately poor and Black counties (Watson, 2015). Even when states offer remedial assistance, like Wisconsin’s provision of a free ID, underlying documentation required for identification is an onerous burden. One study found nearly 15% of eligible voters, and 20% of Latinos, in Milwaukee County without the documents to get a free ID (Barreto & Sanchez, 2012a).

Despite the underlying motivation to curtail democratic turnout, the impact of ID laws on voting is contested. A handful of studies find voter ID laws have little impact (Erikson & Minnite, 2009; Muhlhausen & Sikich, 2007; Mycoff, Wagner, & Wilson, 2009). Two studies find that the strictest laws diminish turnout among Blacks and Latinos (Hajnal, Lajevardi, & Nielson, 2017; Vercellotti & Anderson, 2006). Still other research from Georgia found that Black turnout in the state increased in 2008 following a strict ID law. However, this line of inquiry is fettered by data limitations (Grimmer, Hersh, Meredith, Mummolo, & Nall, 2018).

Mixed findings around turnout obscure the importance of the legal framework within which ID laws operate. Republican lawmakers design ID laws with marginalized voters in mind. Why and how marginalized citizens overcome barriers intended to keep them from voting is a point of inquiry important to the study of power in American politics. In Georgia in 2008, for example, Barack Obama’s historic campaign and activists’ mobilization efforts energized citizens who had a valid piece of ID. Higher turnout among co-ethnic community members with valid ID does not equate to the negligible impact of voter ID laws. Singular focus on turnout without centralizing the real impact of such burdens on access to the franchise is one-dimensional, operating within the subtext of racial power to reproduce the inequalities that demand the attention of political scientists in the first place (Murakawa & Beckett, 2010).

We therefore turn attention to assessing the extent to which ID laws amount to a racially disparate barrier to the franchise, should one wish to cast a ballot. Expert reports in several cases challenging ID laws demonstrate by a variety of methods that Blacks and Latinos are less likely than Whites to possess an appropriate ID (Barreto & Sanchez, 2012a, 2012b, 2014; Marker, 2014). In Texas, for example, using database matching Stephen Ansolabehere demonstrated a two-to-five percentage point difference between Hispanic and White voters possessing a valid ID, which grew to four-to-eight percentage points for Blacks (“Plaintiffs and plaintiff-intervenors’ proposed,” 2014). Barreto, Nuno, and Sanchez (2009) demonstrate disparate rates of access to an ID in Indiana, and work by Stewart (2013) suggests that this trend holds more generally.¹ Yet, very little has been published in academic venues corroborating disparate rates of ID possession, nor have these differences been subjected to more rigorous analysis.

Citizens across the nation face barriers to voting in the form of ID laws. To address the shortcomings of existing research on this topic, we offer evidence in three parts. First, we show that ID laws present a greater barrier to voting for minorities than for Whites, and that these disparities are national in scope. Second, we demonstrate that racial differences persist after accounting for relevant covariates, including socioeconomic status. Finally, we assess the underlying factors that uniquely impact access to an ID among racial subgroups. Faced with inconclusive evaluations of voter ID on turnout, a comprehensive portrait of “who does or does not have the kinds of identity documents mandated in recent voter identification legislation” should be “enough to raise concerns about a disparate impact of voter ID laws” (Erikson & Minnite, 2009, p. 98).

Data and Method

We leverage six datasets, collected between 2008 and 2014. Surveys were conducted in Wisconsin, Indiana, Pennsylvania, and Texas, were designed to measure access to an ID given the laws in each state, and to coincide with the time period when the laws were in place but legally contested. The survey in Indiana was fielded in October, 2007; in Wisconsin from December, 2011, to January, 2012; in Pennsylvania in June of 2012; and in Texas from March to April, 2014. We pair these state datasets with two national surveys: the 2008 Collaborative Multiracial Post-Election Survey (CMPS), fielded from November, 2008, to January, 2009, and the 2012 American National Election Study (ANES), fielded from September, 2012, to January, 2013. Across surveys, respondents were asked whether they had access to an ID, with multiple follow-up questions to ensure their ID would meet state guidelines. Sampling techniques employed in Wisconsin, Indiana, Pennsylvania, Texas, and in the CMPS ensure a robust sample of non-White and low socioeconomic status eligible voters, who are most likely to lack an ID and are hard to reach by traditional sampling methods. Combined, these amount to 18,186 completed interviews, includuing 4,528 Latinos, 4,289 Blacks, 1,064 Asians, 7,763 White non-Hispanics, and 542 of “other race.”

Key issues faced when combing multiple datasets include differences in sample design, population, question wording, and survey administration (Tourangeau, 2003). Modes of data collection included telephone, face-to-face (ANES), and web administration (ANES). The target population varied from registered voters (in the 2008 CMPS) to all citizen eligible adults (in Wisconsin, Indiana, Pennsylvania, Texas, and the ANES). Whether one has a valid ID is coded according to state law in each of the state datasets. In the two national datasets, valid ID is assessed by asking about a non-expired, state issued ID. Question wording is thus an issue insofar as we have combined precise measures in the state datasets with less precise measures in the national datasets. We do this to enhance the generalizability of the findings, and to gain analytical leverage among racial subgroups.² Each individual dataset is weighted to bring its demographic profile in line with Census estimates for the eligible or registered population (depending on the sample), and then an overall weight is constructed such that the final data matche the national citizen, adult population, as per Osborne (2011). We include fixed affects for dataset in analyses of the pooled data. Finally, pooling across multiple sources is not without precedent, and the statistical leverage gained from increasing the sample size legitimizes the methodological decision to do so (Kohnen & Reiter, 2009).³

Findings

White respondents were statistically more likely to possess a valid form of ID than other racial groups in a model only accounting for race across every dataset included in the analysis. Table 1 displays the percent of each racial subgroup possessing a valid ID, among both the individual and pooled datasets. In both nationally representative datasets, Whites were significantly more likely to possess a valid ID than were all other racial groups. In all datasets but Texas, Blacks were statistically less likely to possess an ID than were Whites. The same was true for Latinos in all datasets but Pennsylvania. In the combined dataset, about 81% of Blacks possessed a valid ID, compared with 91% of Whites, 82% of Latinos, 85% of Asians, and 86% of those who identify some other way. In the pooled dataset, Asians and those who identify some other way are statistically less likely to possess a valid ID than are Whites, although these relationships are not consistent.⁴

Table 1.

Percent Possessing a Valid Piece of Voter ID, by Race and Dataset.

	Combined	ANES	CMPS	Indiana	Pennsylvania	Texas	Wisconsin
White	90.5	93.1	88.1	85.8	86.7	90.8	92.7
Black	81.2***	82.7***	78.0***	73.6***	84.4**	86.9	82.7***
Latino	82.0***	81.5***	80.8***	72.7	78.4*	83.2***	83.2***
Asian	84.9***	86.2***	80.9**	91.7	100.0	100.0	87.5
Other race	85.5***	84.0***	−	79.6	69.6**	92.9	89.4

Note. ANES = American National Election Study; CMPS = Collaborative Multiracial Post-Election Survey.

p < .1. **p < .05. ***p < .01; Significance levels reflect logistic regression analysis, and racial categories are statistically different from White; estimates are weighted.

The disparate impact of voter ID laws on Blacks, Latinos, and those of some other race persists after controlling for a variety of relevant covariates. Figure 1 displays the logistic regression coefficients and confidence intervals resulting from an abbreviated and fully specified multivariate analysis of access to a valid ID among eligible voters in the pooled sample.⁵ After including appropriate controls, the relationship between possession of a valid ID and identifying as Asian continues to be negative but is no longer statistically significant. Figure 1 is most useful for comparing the relative effect size for each of the variables in the model, and reveals that the magnitude of the negative impact of race on the likelihood of having a valid ID is substantial, outstripping other relevant variables like age, gender, and having been born outside the United States. Figure 2 displays the predicted probability of possessing an ID among each racial subgroup. Whites have a probability of having a valid ID of about 90% both in the abbreviated and multivariate models. In contrast, Blacks in the abbreviated model have a predicted probability of ID possession of .81, which improves to .85 in the multivariate model. Likewise, Latinos in the abbreviated model have a predicted probability of having an appropriate ID of about .82, which improves to .85. Among Asian respondents, the likelihood of possessing a valid ID improves from .85 to .87, and in the fully specified model is no longer statistically distinguishable from Whites. Among those who identify with some other racial group, controlling for relevant covariates does not diminish the spastically negative relationship between race and ID possession.

Figure 1.

The marginal effects of relevant variables on likelihood of possessing a valid piece of voter ID, in an abbreviated and fully specified model.

Figure 2.

The predicted likelihood of possessing a valid piece of voter identification by race and model.

In addition to age, gender, and having been born outside the United States, education, income, and homeownership also impact ID possession (Figure 1). Age may negatively impact the likelihood of having an ID by way of expiration, where the elderly are less likely to drive and thus less likely to have an updated license. Younger individuals may rely on a student ID issued by a university, which often do not conform to ID regulations in a given state. Women may be less likely to have an ID as result of changing one’s name after marriage. Education, income, and homeownership are associated with an increased likelihood of ID possession insofar as individuals with greater civic knowledge and material resources are likely to have the skills needed to navigate public agencies, and the resources to secure appropriate ID.

The above analysis highlights that, while courts rely on bivariate evidence of racially disparate impact, when subjected to more rigorous analysis the independent, negative effects of race persist. This raises questions around underlying factors that might account for the enduring race gap. We explore this further by examining differences in the various factors that are associated with ID possession among racial subsamples, displayed in Figure 3. Some factors consistently influence ID possession across all groups, like homeownership and income. However, there are differences. Among Blacks, education is positively and statistically associated with the likelihood of having an ID. Education is also positively associated with ID possession among Latinos, although it is not statistically significant. Among Whites, other factors are important predictors of lacking an ID, including being over the age of 65 years, a Democrat, and female.

Figure 3.

The marginal effects of relevant variables on likelihood of possessing a valid piece of voter identification, among racial subgroups.

While these factors likely compound barriers to accessing an ID among Blacks and Latinos, socioeconomic factors like education and income are of primary importance for these groups. An examination of the predicted probabilities of having a valid ID given less than a high school education compared with having a postgraduate degree reveals that Whites with lesser education are no more likely to have a valid ID than are their educated counterparts (.94 among those without a high school degree compared with .937 among those with a postgraduate degree). In contrast, high levels of education increase the likelihood of having an ID by about eight percentage points among Blacks (.79 compared with .87) and three percentage points among Latinos (.81 compared with .84).

Conclusion

This analysis was undertaken to offer a comprehensive evaluation of the scope of the racially disparate impact of voter ID laws. Existing research demonstrates that voter ID laws are partisan tools, designed with the marginalized fringe of the Democratic party in mind, to shape the electorate primarily in favor of state Republican legislatures facing competitive elections. Voting rights activists levy challenges to such laws, focusing on disparate access to appropriate ID among people of color. Legal precedent has developed largely in the absence of evidence that the disparate impact of ID laws extends beyond a few key states, endures beyond class, and diminishes turnout. The best evidence available suggests that voter ID laws have a negative, racially disparate impact on turnout across the states (Hajnal, Kuk, & Lajevardi, 2018; Hajnal et al., 2017). Our analysis joins this research to demonstrate that racial disparities in access to identification appropriate for voting persist even after accounting for important covariates like education and income, underscoring the privileges accrued to Whites through a history of institutional racial exclusion.

Supplemental Material

Appendix_(1) – Supplemental material for The Racial Implications of Voter Identification Laws in America

Supplemental material, Appendix_(1) for The Racial Implications of Voter Identification Laws in America by Matt A. Barreto, Stephen Nuño, Gabriel R. Sanchez and Hannah L. Walker in American Politics Research

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biographies

Matt A. Barreto is a professor of Political Science and Chicana/o Studies at the University of California, Los Angeles, and co-cofounder of the research firm Latino Decisions.

Stephen Nuno is an associate professor and Chair of Politics and International Affairs at Northern Arizona University.

Gabriel R. Sanchez is a professor of Political Science at the University of New Mexico, where he is also the Executive Director of the Center for Social Policy.

Hannah L. Walker is an assistant professor of Political Science and Criminal Justice at Rutgers University.

References

Bachu

O’Connell

(1995). Mothers who receive AFDC payments (Statistical brief). U.S. Census Bureau. Retrieved from http://www.census.gov/population/socdemo/statbriefs/sb2-95.html

Barreto

M. A.

Nuno

S. A.

Sanchez

G. R.

(2009). The disproportionate impact of voter-ID requirements on the electorate—New evidence from Indiana. Political Science & Politics, 42, 111-116.

Barreto

M. A.

Sanchez

(2012a). Rates of possession of accepted photo identification, among different subgroups in the eligible voter population, Milwaukee County, Wisconsin (Expert report). American Civil Liberties Union. Retrieved from https://www.aclu.org/files/assets/062-10-exhibitjexpertreport.pdf

Barreto

M. A.

Sanchez

(2012b). Rates of possession of valid photo identification, and public knowledge of the voter ID law in Pennsylvania (Expert report). American Civil Liberties Union. Retrieved from https://www.aclupa.org/files/2513/7960/9091/Barreto.pdf

Barreto

M. A

Sanchez

(2014). Accepted photo identification and different subgroups in the eligible voter population, State of Texas, 2014 (Expert report). Latino Decisions. Retrieved from http://www.latinodecisions.com/blog/wp-content/uploads/2014/10/Texas-Voter-ID-Expert-Report_Barreto_Sanchez.pdf

Bentele

K. G.

O’brien

E. E.

(2013). Jim Crow 2.0? Why states consider and adopt restrictive voter access policies. Perspectives on Politics, 11, 1088-1116.

Biggers

D. R.

Hanmer

M. J.

(2017). Understanding the adoption of voter identification laws in the American states. American Politics Research, 45, 560-588.

Erikson

R. S.

Minnite

L. C.

(2009). Modeling problems in the voter identification—Voter turnout debate. Election Law Journal, 8, 85-101.

Grimmer

Hersh

Meredith

Mummolo

Nall

(2018). Obstacles to estimating voter ID laws? Effect on turnout. The Journal of Politics, 80(3), 1045-1051.

10.

Grossmann

Hopkins

D. A.

(2015). Ideological republicans and group interest democrats: The asymmetry of American party politics. Perspectives on Politics, 13, 119-139.

11.

Hajnal

Kuk

Lajevardi

(2018). We all agree: Strict voter ID laws disproportionately burden minorities. The Journal of Politics, 80(3), 1052-1059.

12.

Hajnal

Lajevardi

Nielson

(2017). Voter identification laws and the suppression of minority votes. The Journal of Politics, 79, 363-379.

13.

Hicks

W. D.

McKee

S. C.

Sellers

M. D.

Smith

D. A.

(2015). A principle or a strategy? Voter identification laws and partisan competition in the American States. Political Research Quarterly, 68(1), 18-33.

14.

Kohnen

C. N.

Reiter

J. P.

(2009). Multiple imputation for combining confidential data owned by two agencies. Journal of the Royal Statistical Society: Series A, 172, 511-528.

15.

Marker

D. A.

(2014). The statistical role in voter identification (ID) laws. Statistics and Public Policy, 1, 46-50.

16.

Minnite

(2007). The politics of voter fraud (Report). Project Vote. Retrieved from http://www.projectvote.org/wp-content/uploads/2007/03/Politics_of_Voter_Fraud_Final.pdf

17.

Muhlhausen

D. B.

Sikich

K. W.

(2007). New analysis shows voter identification laws do not reduce turnout (A report of the Heritage Center for Data Analysis). Retrieved from https://www.heritage.org/election-integrity/report/new-analysis-shows-voter-identification-laws-do-not-reduce-turnout

18.

Murakawa

Beckett

(2010). The penology of racial innocence: The erasure of racism in the study and practice of punishment. Law & Society Review, 44, 695-730.

19.

Mycoff

J. D.

Wagner

M. W.

Wilson

D. C.

(2009). The empirical effects of voter-ID laws: Present or absent? Political Science & Politics, 42, 121-126.

20.

Osborne

J. W.

(2011). Best practices in using large, complex samples: The importance of using appropriate weights and design effect compensation. Practical Assessment, Research and Evaluation, 16(12), 1-7.

21.

Plaintiffs and plaintiff-intervenors’ proposed findings of fact and conclusions of law. (2014). Expert report of Brennan Center for Justice. Retrieved from https://www.brennancenter.org/sites/default/files/legal-work/Findings.pdf

22.

Shanton

(2010). A portrait of hunters and hunting license trends: National report (national Shooting Sports Foundation). Retrieved from http://www.michigandnr.com/FTP/wildlife/aversb/MWL/Hunting_Heritage/Southwick%20NSSF%20Hunter%20Churn%20Natl%20Report.pdf

23.

Stewart

(2013). Voter ID: Who has them; who shows them. Oklahoma Law Review, 66, 21-52.

24.

Stoughton

K. M.

(2013). A new approach to voter ID challenges: Section 2 of the Voting Rights Act. The George Washington Law Review, 81, 292-327.

25.

Tourangeau

(2003). Recurring surveys: Issues and opportunities. Washington, DC: National Science Foundation.

26.

Vercellotti

Anderson

(2006). Protecting the franchise, or restricting it? The effects of voter identification requirements on turnout. Manuscript, Rutgers University, New Brunswick, NJ.

27.

Watson

(2015). Alabama’s DMV shutdown has everything to do with race (Report). American Civil Liberties Union. Retrieved from https://www.aclu.org/blog/speak-freely/alabamas-dmv-shutdown-has-everything-do-race

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.21 MB