Three scholars at the University of California San Diego caused a stir when they released a recent paper that estimated the impact of voter ID laws, finding a significantly larger drop in minority turnout, especially among Hispanics, than other studies had found. “Voter ID laws skew democracy in favor of whites and those on the political right,” they concluded.
But now a group of experts at Stanford, in a paper of their own, say they found “large data inaccuracies” in the UCSD study. The new study concludes that the available data don’t allow us to make definitive conclusions about the impact of voter ID on turnout.
You may or may not be surprised to know that I don’t have anywhere near the statistical chops to figure out who’s right here. But if nothing else, the controversy underlines the reality that, as I’ve written before, measuring the exact impact of voter ID laws and other election rules on turnout is an extremely complicated business, in part because numerous factors can affect turnout, so its hard to compare one election to another. In short, the picture remains murky, and claims of certainty from either side should be met with skepticism.
But this kind of measurement, though important for political scientists to do, is also beside the point in terms of the merits of the policy. There’s no dispute that large numbers of registered voters lack photo ID (around 800,000 in Texas, around 300,000 in Wisconsin); that they’re much more likely to be black or Hispanic than white; and that several states with ID laws have failed to make IDs easily accessible. Exactly how many people these laws keep from the polls might matter if they prevented a comparable amount of in-person fraud. Since they don’t, it’s essentially an argument over whether we’re disenfranchising 10,000 or 100,000 people.