One particularly unfortunate aspect of the report: like others, it apparently gives credence to the argument that you can assess the impact of an ID law on minorities in places like Indiana and Georgia by looking at turnout results from an election or two. Rep. Todd Rokita made that same argument again earlier this week.
As I’ve said before, trying to assess the impact of any given law by looking at aggregate turnout in an election or two is, simply, incredibly bad statistics. There are hundreds of reasons why turnout goes up or down in a particular year: particular candidates, particular hot-button issues, particularly meaningful races, particular local mobilization, particular weather.
Trying to assess the impact of ID on minority citizens in Indiana and Georgia in 2008 is particularly foolish: both states were, for the first time ever, contested battleground states in a presidential election with a minority candidate at the top of the ticket. Knowing that minority turnout jumped from 2004 to 2008 doesn’t tell you anything about the role of a single law in the mix. Maybe the turnout would have been twice as high if not for the law.
Update: In today’s Washington Times, Hans von Spakovsky makes the same claim Rokita did above — and adds 2010 as additional evidence. 2010 is less facially ridiculous than 2008 … but four data points, two of which don’t pass the laugh test, still fail Statistics 101 as any sort of causal assertion.