Voter Suppression In A Post-Shelby World

In the first general election since Shelby County v. Holder (2013), restrictive voting rules in place in jurisdictions previously covered for preclearance by the Department of Justice might have played a role in the election results. Such tactics limiting access to the ballot box include shortened early voting hours, reduced polling locations and hours, new requirements for photo ID, and increased voting restrictions for felons. These tactics have been contested in the courts for their negative impact on minority voters. For example, the three judge panel that struck down an omnibus bill passed in North Carolina after Shelby County described it as “target(ing) African Americans with almost surgical precision.”

Selective voter identification requirements were among the provisions in the North Carolina bill, and were one of the most hotly contested voting regulations heading into the 2016 election. Advocates of voter identification laws argue that they are important to maintaining the integrity of elections. Research suggests that they are most often adopted in states where Republicans control the legislature, but elections are highly competitive. Opponents worry about their disparate impact on racial minorities and lower-income voters, and battles over their validity are waged in the courts.

In addition to North Carolina, during the summer of 2016 courts struck down, stayed or amended laws in North Dakota, Texas and Wisconsin. Yet, strict requirements remained in nine states, and cases in key states were pending appeal at the time of the election. For example, although the 5th Circuit Court found that Texas’ law was in violation of Section 2 of the Voting Rights Act, a version of the law remained in effect, problems surfaced with its administration in the weeks leading up to November 8, and the State applied for a writ of certiori in the end of September. Questions abound over the exact impact of voter identification laws on voter turnout, and whether or not they give an edge to their Republican proponents. The 2016 presidential election brings these questions into sharp relief, where for only the fourth time in American history the winner took the Electoral College while losing the popular vote, which Trump lost by nearly a million ballots.

Questions over the impact of voter suppression are particularly acute in highly competitive swing states like Wisconsin, home to one of the strictest and most contested voter identification laws in the nation. While the exact impact on voter turnout is unclear, evidence suggests that Wisconsin’s law had the potential to impact election outcomes.

Voter Identification in Wisconsin

2016 marked the first general election in which Wisconsin’s voter identification law was in effect. The law requires that ID have a photo and a name conforming to that listed on the voter rolls, and enumerates only eight variations that are acceptable. Moreover, if the ID is not proof of address individuals must provide additional documentation to this effect. Research indicates that voter identification laws are most onerous for low income and minority voters, and on Election Day officials confirmed that turnout was lower than in previous years in districts most likely to be impacted by voter ID.

In the city of Milwaukee, a Democratic stronghold, 41,000 fewer votes were cast in 2016 than in 2012. This decrease cannot be attributed to a shrinking citizen voting age population, which according to the U.S. Census increased by over 8,000 between 2011 and 2015. Turnout was also down by over 22,000 votes in Douglass County, also reliably Democratic, and in certain counties flipped by Trump in 2016, such as Kenosha and Racine county where over 20,000 fewer votes were cast. Research examining the impact of voter identification laws on voter turnout estimate that the strictest laws, of which Wisconsin’s is one, depress minority turnout as much as 10 percentage points. Hillary Clinton lost the state of Wisconsin to Donald Trump by only 22,871 votes, raising the possibility that voter ID played a part in giving Republicans the edge.

Survey data collected in Milwaukee County in 2012 measuring access to a piece of identification that met the requirements of the law indicates that nearly 10 percent of Milwaukee County residents did not have access to an appropriate ID at the time of the survey. Possession of a valid piece of ID differed among racial subgroups, where 13.2 percent of Black respondents and 14.9 percent of Latino respondents did not have appropriate ID, compared to only 7.3 percent of white respondents. Drawing on census estimates of the citizen voting age population, this translates to an estimated 63,085 voters in Milwaukee County who lacked an appropriate ID for voting, including 7,453 Latino, 20,939 Black and 30,894 white eligible voters. While estimates of vote choice by race are not yet available for Wisconsin, national exit polls suggest 37 percent of whites, 88 percent of Blacks and 65 percent of Latinos voted for Clinton, generating 34,700 eligible voters likely to support Clinton whose vote was threatened by voter ID laws. If we correct the Latino estimate to reflect Latino Decisions’ Election Eve Poll, which found 77 percent support for Clinton in Wisconsin, the approximate number of threatened votes grows to 35,595.

Voter Suppression around the Country

Despite the fact that North Carolina’s restrictive voting law was struck down in July election officials in the state sought to alter the electorate through alternative means. For example, to get around the Court’s ruling against eliminating a full week of early voting county election boards instead severely reduced the number of polling stations available during the early voting period. In the first week of early voting 158 fewer polling stations were available across 40 counties with high concentrations of Black voters than in 2012. Reports suggest that in the first seven days Black turnout was down by 16 percent in the first week of early voting compared to the same week in 2012. Drawing on the estimated number of Black voters who voted in the first week of early voting in 2012, that amounts to about 40,409 potential votes for Clinton, without accounting for other suppression tactics like last minute purges to voter registration records. Clinton only lost North Carolina by 177,529 total votes. Likewise, in Florida where Clinton lost by only 119,770 votes, felon disenfranchisement barred a quarter of Florida’s Black voting age population from the polls in 2016. Florida is one of a handful of states, including swing states like Arizona and Iowa that permanently bar felons from voting.

These numbers highlight the increasing importance of voter suppression to election outcomes. The impact of voter ID and other voter suppression tactics on the presidential race likely will require more thorough rigorous analysis. But, reports from volunteers for the League of Women Voters deployed to observe the election across Wisconsin indicate that individuals were indeed negatively impacted by voter ID, having been turned away without being offered a provisional ballot. This, together with evidence that voter identification laws have a disproportionately negative effect on minority turnout in the aggregate, suggests that Wisconsin’s law was not inconsequential to the outcome of its highly competitive election.

Hannah Walker, a guest blogger for Latino Decisions, is a post doctoral fellow with the Prisons and Justice Initiative at Georgetown University and joins Rutgers University as an assistant professor of Political Science and Criminal Justice next fall.