Voter ID laws don't seem to suppress minority votes – despite what many claim

Strict voter ID laws require residents to possess a valid, state-approved identification in order to vote.

Support and opposition to these laws primarily fall along party lines. Proponents – mainly Republicans – argue they are needed to protect the integrity of the electoral process. Opponents, who tend to be Democrats, say they’re not necessary to reduce voter fraud.

Democrats have a point: In-person voting fraud is almost nonexistent^[1]. President Donald Trump’s now-defunct Voter Fraud Commission, which was supposed to investigate voter fraud during the 2016 election, was unable to unearth any significant evidence^[2].

Critics claim^[3] Republicans don’t really care about electoral integrity – that voter ID laws are about suppressing the turnout of minority voters, since these voters are less likely^[4] to possess legal forms of identification. Democratic candidates and activists routinely^[5] evoke^[6] these laws as tools of voter suppression.

But a growing body of evidence – which includes a new study^[7] we just published – finds that strict voter ID laws do not appear to disproportionately suppress voter turnout among African Americans, Asian Americans or people of mixed races.

A partisan ploy

In 2005, Georgia and Indiana became the first two states to pass strict voter ID laws, although the statutes weren’t implemented until the Supreme Court’s 2008 decision in Crawford v. Marion County Election Board^[8]. Since then, eight other states^[9] have adopted strict ID laws.

These laws appear to be partisan in nature. From 2006 to 2011, every law^[10] requiring a photo ID or proof of citizenship was passed by a Republican-controlled legislature.

Legislators who support voter ID laws claim they want to protect the country’s elections^[11] from in-person voter fraud. However, this type of fraud is extremely rare^[12]: Voter ID expert Justin Levitt estimates that from 2000 to 2012, there were just 31 credible instances^[13] of in-person voter fraud, out of more than 1 billion votes cast.

Most voting fraud claims^[14] turn out to be inadvertent errors by voters or polling officials. Even the conservative Heritage Foundation finds only 1,177 proven instances^[15] of overall election fraud since 1948.

Although it’s difficult to gauge the true motives of the legislation, research^[17] on the enactment of the laws suggests they are designed to restrict minority turnout for a partisan advantage. As the electorate continues to become more racially and ethnically diverse^[18], Republicans stand to benefit politically from such laws since minority voters dependably support^[19] Democratic candidates.

It’s true that the readiness to produce identification varies by race^[20]. Studies have found that minorities are less likely to have the records needed^[21] to verify their identity. A 2013 national study^[22] found that 63 percent of African Americans and 73 percent of Hispanic Americans had valid driver’s licenses, the most common form of photo ID, compared with 84 percent of whites. Moreover, a 2017 study^[23] found that African Americans and Hispanic Americans in Texas were significantly less likely than white Americans to have proper ID to vote.

For those who lack an acceptable form of ID, the costs are clear^[24] – they need to spend time, energy and money to obtain the necessary ID or documents.

Analyzing the law’s effects

But the research on voter ID laws and minority voter turnout has found mixed^[25] results.

In 2014, the United States Government Accountability Office reviewed 10 early studies^[26] to determine the effects of voter ID laws. Five of the studies found that the laws had no effect on turnout, four noted decreases and one discovered an increase. Of the four studies that found decreases, estimates ranged from 1.5 to 3.9 percentage points.

In 2017, a study appeared to have resolved the issue. Political scientists Zoltan Hajnal, Nazita Lajevardi and Lindsay Nielson published a voter ID article^[27] demonstrating strict ID laws negatively affect minority voters but not white voters. The article, which received national attention^[28], reasoned that the effects of ID laws were now easier to establish because more strict laws had been enacted, giving researchers more data to analyze.

They found, for instance, that in general elections, Hispanic Americans were 10 percent less likely to vote in states with strict ID laws. They determined that strict ID laws could potentially bias United States democracy by favoring whites and the political right.

Their findings seemed like a breakthrough – proof that strict ID laws disenfranchise minority voters. But a subsequent article^[29] pointed out some possible errors in Hajnal, Lajevardi and Nielson’s work. Once these errors were accounted for, it appeared their original findings were not firmly supported. Hajnal and his coauthors, however, disagreed^[30] with this assessment.

New data set, different results

Using new data, we decided to replicate Hajnal’s study to see whether we could confirm the results.

In our study^[31], we analyzed the turnout of more than 285,000 registered voters in midterm and general elections from 2008 to 2014. Whereas Hajnal, Lajevardi and Nielson only used data from the Cooperative Congressional Election Study, we tested data from both the Cooperative Congressional Election Study and the United States Census, which provides the most accurate representation of turnout by state.

After we analyzed the Cooperative Congressional Election Survey data, we found that among Hispanic American voters, there does seem to be a relationship between strict voter ID laws and lower turnout.

However, when we examined the census data, we found strict ID laws did not disproportionately disenfranchise minority voters, whether it was Hispanic Americans, African Americans or anyone else. There’s also a working paper^[32] by economists Enrico Cantoni and Vincent Pons that’s generating some buzz. It found results that are similar to ours – namely, strict ID laws do not appear to negatively affect minority voters.

Are voter ID laws backfiring?

Given that minorities are less likely to possess IDs, why are studies finding that these laws don’t suppress turnout?

Political scientists Jason Mycoff, Michael Wagner and David Wilson write^[33] that these laws likely don’t influence turnout because a voter’s political interest is strong enough to overcome the costs of having to obtain an ID. There’s also some evidence^[34] to suggest the laws actually act as a catalyst, inspiring and mobilizing minority voters. After a controversial North Dakota voter ID law was passed that made it more difficult for Native Americans to vote, Native Americans turned out in record numbers during the 2018 midterm elections^[35].

Of course, accurately capturing the effect of voter ID laws across different states and years can be challenging. Voters make countless mental calculations when deciding whether to vote. The candidates running matter. And statistical models can’t account for every factor on Election Day – even something as arbitrary as weather conditions.

At this point, conclusions – including ours – should be interpreted with some caution. The research is still in its infancy. After all, these laws have only been in place for 11 years. If states continue to enact more stringent ID laws, more reliable results should emerge.

For now, the effect of voter ID laws appears to be minimal. However this doesn’t mean policymakers should discount the laws’ intentions to suppress voting. Nor should they ignore other strategies to warp elections, whether it’s gerrymandering, shuttering polling locations or limiting the ability of former felons to vote.

References

1. ^{^} is almost nonexistent (www.washingtonpost.com)
2. ^{^} was unable to unearth any significant evidence (www.nytimes.com)
3. ^{^} Critics claim (www.npr.org)
4. ^{^} since these voters are less likely (journals.sagepub.com)
5. ^{^} routinely (thehill.com)
6. ^{^} evoke (www.texasmonthly.com)
7. ^{^} a new study (onlinelibrary.wiley.com)
8. ^{^} Crawford v. Marion County Election Board (www.law.cornell.edu)
9. ^{^} eight other states (www.ncsl.org)
10. ^{^} every law (www.cambridge.org)
11. ^{^} claim they want to protect the country’s elections (www.nytimes.com)
12. ^{^} extremely rare (www.brennancenter.org)
13. ^{^} 31 credible instances (www.washingtonpost.com)
14. ^{^} voting fraud claims (www.projectvote.org)
15. ^{^} 1,177 proven instances (www.heritage.org)
16. ^{^} Reuters/Chris Keane (pictures.reuters.com)
17. ^{^} research (doi.org)
18. ^{^} racially and ethnically diverse (www.pewsocialtrends.org)
19. ^{^} dependably support (www.people-press.org)
20. ^{^} varies by race (doi.org)
21. ^{^} the records needed (citeseerx.ist.psu.edu)
22. ^{^} study (digitalcommons.law.ou.edu)
23. ^{^} study (amstat.tandfonline.com)
24. ^{^} the costs are clear (books.google.com)
25. ^{^} mixed (www.gao.gov)
26. ^{^} 10 early studies (www.gao.gov)
27. ^{^} a voter ID article (doi.org)
28. ^{^} national attention (www.theatlantic.com)
29. ^{^} a subsequent article (www.journals.uchicago.edu)
30. ^{^} disagreed (www.journals.uchicago.edu)
31. ^{^} In our study (onlinelibrary.wiley.com)
32. ^{^} working paper (www.nytimes.com)
33. ^{^} write (www.brennancenter.org)
34. ^{^} some evidence (files.www.chismstrategies.com)
35. ^{^} Native Americans turned out in record numbers during the 2018 midterm elections (time.com)