A good reminder to self to read, rather than scan the article if I am planning on writing about it. The Federalist had an interesting article about a National Bureau of Economic Research study on Voter ID. I had somehow taken it into my head that the study showed that requiring voter ID did not actually reduce the number of illegals voting. That's what happens when you skim, and arrogantly think you know everything.
I thought it was a surprising but plausible result, and immediately compared it to gun control. Gun regulators are just sure that legislation A or B is going to result in less violent crime, because it just has to. I thought this was an interesting mirror issue, of conservatives believing that requiring voter ID would just have to reduce the number of illegals voting. It's just one of those assumptions they make, but without data it's not any better than the gun controller's argument that increasing registration or waiting period is just obviously going to work. "Just obviously" is not an argument. It is in fact dangerous territory. I was composing this post in my head, working from that assumption.
The study is actually something different. It may still be true that voter ID doesn't reduce illegal voting, but there is nothing here that gives evidence for it. That wasn't the intent. The NBER was looking for something else. The study shows that requiring voter ID does not depress minority turnout, as has been widely believed. If this evidence holds up, it removes the only serious obstacle to states requiring the ID.
Now that it has come up, though, I'd be curious if requiring voter ID actually does reduce illegal votes. I'm not in favor of regulations that only look like they are going to work, just because.