"This was a racist attack on our sacred right to vote," North Carolina NAACP President Rev. William Barber declared earlier this week. Barber was denouncing some recent changes to North Carolina's election laws, which are being challenged in the federal court case NAACP vs. McCrory. The changes, Barber said, are "the worst voter suppression law we've seen since the 1960s."
Among other things, the new North Carolina law eliminated same-day voter registration, cut back early voting, and ended pre-registration of eligible high school students. But the main objection to the changes involves the requirement that voters show photo identification at polling places. In its brief, the NAACP argues that African Americans are "disproportionately" affected by this requirement. The brief also claims that the new laws were "enacted with the intention of suppressing the number of votes cast by African-Americans."
My general view is that voting should be made easier, although as we'll see, there is precious little evidence that it makes any significant difference in electoral results. I also don't have much regard for the idea that we need tighter voter ID laws to prevent voter fraud, since most research shows that voter-impersonation fraud is extremely rare.
But if the intent of the new laws really is to suppress minority group voting, it's not likely to work. There has been a lot of academic research recently on the effects of stricter voter ID requirements, and—contrary to Barber's apocalyptic statements—they don't seem to have much of an impact on minority turnout at all.
Take "The Politics of Race and Voter ID Laws in the States: The Return of Jim Crow?," a 2013 study published in the Political Research Quarterly. In it, political scientists Rene Rocha and Tetsuya Matsubayashi find that states in which Republicans hold a majority in the legislature and the governorship are more likely to adopt strict voter ID laws. (And so it was in North Carolina when the new laws were adopted.) Then they look at how changes in electoral rules may have affected voter turnout by comparing election results before and after voter ID law changes in 49 states between 1980 and 2010.
"Our primary explanatory variables, photo ID and nonphoto ID laws, have no statistically discernible relationship with the probability that whites, blacks, and Latinos voted in the general elections between 1980 and 2010 except that the nonphoto ID law has a positive and significant relationship with Latino turnout," they find. "In short, more stringent ID requirements for voting have no deterring effect on individual turnout across different racial and ethnic groups."
Rocha and Matsubayashi speculate that any suppression effects the new laws may induce are being more than counteracted by get-out-the-vote efforts by partisan organizations that aim to mobilize minorities. The two also find that "universal mail voting, no-excuse absentee voting, and early in-person voting, have no systematic effect on turnout when racial and ethnic groups are analyzed separately."
In another recent paper, three political scientists from Berkeley and Columbia report the results of a field experiment involving voters in Appalachian Tennessee and Virginia matched for income, age, and minority-status in the 2012 general election. They also targeted voters in predominately black neighborhoods in Knoxville, Tennessee, and Roanoke, Virginia. Tennessee requires a photo ID to vote whereas Virginia requires a nonphoto ID, e.g., utility bills and bank statements showing the voter's address. The researchers mailed three different postcards to selected voters. One reminded them to vote; the second warned that voter ID requirements had changed ("warning"); the third warned that the ID requirements had changed but also explained how to get the appropriate IDs ("help"). A control group received no postcards.
The researchers reported that both white and black voters sent the reminder postcard voted at essentially no greater rate than the controls who did not receive postcards. On the other hand, both the warning and help postcards apparently boosted turnout by both white and black voters by around one percent. The researchers conclude. "We find no evidence that [stricter voter ID requirements] have a net demobilizing effect."
In a 2015 study that is currently under review, Lindsay Nielson—a political scientist at the University of San Diego—parses the effects of stricter voter ID laws on the voting patterns of the young, the elderly, the poor, and racial minorities. Using data on 100,000 respondents in Cooperative Congressional Election Survey, Nielson examines how voter ID laws affected turnout in both primary and general elections in 2010 and 2012.
Nielson finds that stricter voter ID laws do change the probability that someone will vote in primary elections, but not in general elections. In primaries, she reports, whites and minorities vote at approximately similar rates; turnout declines for people of all races from 43 to 31 percent, as ID requirements become stricter. Turnout among voters over age 65 declines from 57 to 48 percent in primary elections; among those ages 35 to 64, it drops from 42 to 34 percent; the young vote decreases from 30 to 22 percent. Income makes no difference; turnout declines about 10 percent both for people who make more than $40,000 per year and those who make less. She found similar results when the income cutoff was set at $20,000 per year.
General elections are another matter. Nielson finds that "there is little evidence that racial minorities are less likely than whites to vote when states institute voter identification requirements." The elderly vote drops a few percentage points when IDs are required, but the turnout of middle-aged and young voters does not change. There is also "no statistically significant gap in estimated turnout [between high-income and low-income voters] when the identification law becomes stricter."
Ultimately Lindsey reckons that her "findings demonstrate that the evidence that voter identification laws demobilize potential voters is not as strong as opponents of the laws might wish and that the controversy over these laws may be exaggerated." Whatever lawmakers' intentions may be when imposing more stringent voter ID requirements—to honorably prevent fraud or to dishonorably suppress votes—the data suggest that the requirements are a big waste of time and money.