Turnout, Voter Fraud, and Voter IDs

With the leak yesterday regarding the Department of Justice investigation into ACORN's registration efforts in various battleground states, there is a continued attempt to keep the voter fraud canard alive. Those following carefully know that there has been a deliberate effort to conflate voter registration fraud with fraud at the polling place. The two are quite distinct and only one of the two directly threatens the integrity of the voting process. When you have a system where you pay people to gather registrations, their incentives are going to be to register as many people as possible, regardless of whether they are accurate or false. There is, however, no evidence that these falsely registered people actually show up to vote. Careful examinations of instances of alleged voter fraud at the polling place, including the Brennan Center's report last year, show that none can be substantiated and most are a function of clerical error.

Nonetheless, the claims of voter fraud have already had an impact on the electoral landscape. There are now 24 states that require ID from voters at the polls, including Indiana, Florida, Ohio, Missouri, Virginia, and Colorado. These requirements were adopted in response to concerns about voter fraud at the polling place, even in the absence of such fraud occuring. Of course, it is no mistake that those least likely to have government-issued ID are the poor, racial minorities, and the very elderly. Increasing the costs of voting for those populations means lower turnout. In April of this year, the Supreme Court upheld these laws even in the admitted absence of evidence of voter fraud, because such laws help to reassure voters about the integrity of the electoral system (although note Stephen Ansolabehere's work establishing that voters show no change in level of confidence in states with voter ID laws). The burdens imposed were incidental, according to the majority opinion. As political scientists know, though, even "incidental" burdens on voting can reduce turnout.

What does this mean for this election? As a barrier to victory for Obama, probably very little. Obama's leads in the states that he needs to win are sizeable enough that the influence of these kinds of laws are unlikely to be decisive. However, they can influence the kind of victory that Obama has. I don't buy into the notion that a "mandate" is necessary or even means anything substantive, but I do think that redrawing the map of American politics could matter. Obama wins in the south and the Mountain West can change the perception of the acceptability of Democratic policies. And it is on the margin, in close red states, that voter ID laws may have an impact. It is unfortunately too late to change them in this election cycle, but hopefully the issue will be revisited in coming years.

Oh, and thanks to Larry for working to deny me tenure by getting me to post on his blog rather than doing my research.