You need an ID to get on a plane, to open a checking account, to buy beer, you need an ID for all sorts of things in order to prove your identity. So at first blush, it seems only logical that you ought to be required to present an ID in order to vote, too, right? You don't want somebody stealing your identity to open a checking account, and you don't want somebody stealing your identity to cast a vote for a different candidate. And besides, getting an ID is pretty easy, all things considered.
This is all very appealing to plain common sense. Unfortunately, it's also wrong. The reasons why it is wrong are counter-intuitive, but they are also relatively simple to explain. I don't recall ever seeing the reasons summarized succinctly, logically, and dispassionately (as we will soon see, this debate tends to inflame passions due to its relationship to race and class issues). So without further ado, I offer here my attempt to explain as simply as possible why ID requirements for voting are a bad thing:
In a nutshell, it's because of two reasons. One, there are significant negative side-effects; and, two, there is virtually no benefit in terms of preventing meaningful voter fraud.
Most Americans drive. So you've already got an ID. For most of us, producing an ID seems like the simplest thing in the world. Within this context, our "common sense" seems to tell us that only the most irredeemable fuckups are not going to have an ID.
But as always, we are better served by looking at the data than we are by consulting our "common sense", and the data is clear: Laws which require voters to present ID disproportionately affect minorities, the poor, and Democratic voters. We don't have to speculate about this or that, because the data is already there.
The first two groups are historically disenfranchised already, so it seems to me like a demonstrably Bad Thing to disenfranchise them even further. If you think that preventing minorities and the poor from voting is a positive thing, well, I probably cannot convince you of anything (but I would like to tell you to bite me, please). But for those of us who, at the very least, feel that it's downright shameful to disproportionately exclude black voters from the political process, especially in a nation that so recently was regularly lynching it's not-white-enough citizens, the data speaks for itself.
Whether or not the disproportionate disenfranchisement of Democratic voters is a good thing I imagine constitutes a rather straightforward political Rohrshach. But if nothing else, it ought to call into question the motives of the exclusively-Republican legislators who are the ones pushing for such laws: It directly affects their odds of future employment. If that's not a conflict of interest, I don't know what is.
But all of this might still be excusable if such laws really did reduce voter fraud in a meaningful way. Do they?
More than once in this post, I've drawn an analogy to the act of presenting an ID to open a bank account, so let's continue in that vein. Let's say there was a bank that didn't require any sort of ID. Somebody could come in with just my name and address and open a checking account in my name, write all sorts of bad checks, and I'd be on the hook for it. That's pretty shabby, I agree.
But wait a minute now, there are problems with this analogy. What if there were about a 57% chance that I would be coming into that very same bank, on the very same day in order to open a legitimate checking account of my own? What if I already got there first? Now it will be mighty embarrassing for the would-be fraudster! Even if the fraudster beat me to it, someone might remember what she looked like, and the odds of getting caught are non-trivial.
And yet, the analogy gets even worse. Turns out that the fraudster can only get fifty bucks or so for each fraudulent bank account she opens. In order to actually make it worthwhile, she will have to visit dozens of no-ID banks, she'll have to do it all on the same day, and for each identity she steals there is a better than 50/50 chance that the person in question will be visiting the same bank on the same day to open an account.
This is starting to look like a pretty lousy way to make a buck. And so it is with this method of voter fraud: You have to steal a lot of votes in order to make it worthwhile, and each vote you steal comes with a non-trivial risk of getting caught.
This analogy was a bit of an over-simplification -- for instance, in reality, you are better off stealing the identity of the recently deceased rather than a living person, since there is a 0% chance they will be making it to the polling place that day, but this still presents problems in that a) you have to obtain a non-trivial number of names of the recently deceased in order to make it worthwhile; and b) given that polling places tend to service a relatively small community, there's still a damn good chance of getting caught. The point remains: This is a really inefficient way to influence an election. Especially when there are much more effective means of tipping the scales which also carry a relatively low risk of getting caught.
So yeah, I admit it's counter-intuitive. In this era of identity theft paranoia -- I just had some fraudulent charges on my debit card less than a week ago! -- it almost seems absurd not to require an ID to vote. But when you crunch the numbers in regards to who is affected, when you look at the feasibility of actually pulling off an election-changing fraud that could be thwarted by ID requirements, it just doesn't add up.