My colleagues Nazita Lajevardi and Lindsay Nielson and I analyzed validated voting data from the Cooperative Congressional Election Study in order to follow voter turnout from 2006 through 2014 among members of different groups — almost a quarter-million Americans in all — in states with and without strict ID laws. The patterns are stark. Where strict identification laws are instituted, racial and ethnic minority turnout significantly declines. One way we analyzed the data was to compare the gap in turnout among races and ethnic groups. It is well established that minorities turn out less than whites in most elections in the United States. Our research shows that the racial turnout gap doubles or triples in states that enact strict ID laws. Latinos are the biggest losers. Their turnout is 7.1 percentage points lower in general elections and 5.3 percentage points lower in primaries in strict ID states than it is in other states. Strict ID laws lower African American, Asian American and multi-racial American turnout as well. In fact, where these laws are implemented, white turnout goes up marginally, compared with non-voter ID states.
The racial and ethnic patterns persist even after we control for factors other than voter ID laws. We ran the data to check the influence of other state-level electoral laws that encourage or discourage participation, of particular issues in each state and congressional district, of the overall partisanship of each state, and of an array of individual demographic characteristics. Regardless of how we looked at the data, we found that strict voter ID laws suppress minority votes.
It is unlikely that the falloff in turnout is due to a reduction in actual voter fraud. Voter ID laws can only prevent voter impersonation, where someone votes in another person’s place. Despite widespread efforts to find such fraud, documented instances are almost nonexistent. Justin Levitt, law professor and now a deputy assistant attorney general in the Justice Department, tracked voter-impersonation allegations from 2000 through 2014 in all kinds of U.S. elections — general, primary, special and municipal. As of August 2014, he found 31 credible instances out of more than 1 billion votes cast in general and primary elections alone.
The suppression patterns in voter ID states have real political consequences. In states where the voices of Latinos, blacks and Asian Americans become more muted and the relative influence of white America grows, the influence of Democrats and liberals wanes and the power of Republicans grows. It should thus not be surprising that strict voter ID laws have been passed almost exclusively by Republican legislatures.