In this recent post (drawing from a just-released preview of my forthcoming book, The Voting Wars), I argued that many Democratic claims against voter identification laws and other Republican-backed laws (part of what Democrats term the “GOP Var on Voting”) either are exaggerated or based upon a total lack of information about the likely effects of such laws.

Does that mean that states should feel free to pass voter identification laws, especially given that identification requirements tend to be supported by majorities in public opinion polls?  No, for the following reasons.

1. The tough new voter id laws serve no purpose.  Of course, some identification requirement is necessary to prevent voter fraud.  But many states use signature matches or other tools (including the ability to have witnesses verify identity) besides a state photographic identification requirement.  Further, HAVA requires first time voters anywhere in the country who register first by mail to provide identifying information.  As the chapter and book explain, there is no good evidence that impersonation voter fraud, the only kind of voter fraud an ID law can prevent, is a real problem.  I go through the claims made by those supporting ID on this basis, as well as the evidence from federal and Texas investigations—voter impersonation fraud is not a problem.  Nor is the use of a voter id law necessary for public confidence in the integrity of the process, as an important article by Ansolabehere and Persily demonstrated.

2. Many Republican legislators support them for the wrong reason.  If Republican leaders really believed that voter fraud was rampant, perhaps support for such laws would be partially justified.  But the best evidence is that those who have studied the issue do not believe such fraud is rampant.  Instead they support the laws to depress the vote of  those likely to vote Democratic, and, as I explain in the last post and the book, they see it as a great “wedge issue” to fire up the base.

3. Even if the effects are small, they can still matter in razor-thin elections.  Perhaps voter i.d. laws deter about 1% of voters from voting (and more who forget their ids and don’t return to vote again—a real concern for Democrats about low motivation voters).  One percent may not seem like much, but we have had enough razor-thin elections in the last few years that some elections could well turn on such a small effect.  Further, we don’t know if some of the newer voter id laws—such as Texas’s laws which won’t accept student i.d.s (many college students won’t have drivers licenses)–will have a greater effect.  In an environment where we know the laws are not necessary to deter fraud, but where we know they will have some (perhaps very small) effect on turnout, the smart choice is not to adopt the i.d. law.

4. Even when effects are too small to affect the outcome of elections, we should not make it harder for people to vote for no good reason.  Voting should not be a difficult activity; everyone entitled to vote should be able to vote.  If you accept those viewpoints (and I know that not everyone does), then we should not put stumbling blocks in front of voters to solve a non-existent problem.  The dignity of the voter is just too high.

5. What we really need is universal voter registration supported by the government, and national voter identification cards with optional thumbprints.  All the costs of such a system would be paid for by the government, including the costs to get the documents necessary to get a free id.  I make the case for this in the book.  And so far, the proposal has united Democrats and Republicans—against the plan!  But until we get such a national system, most state voter id requirements are misguided at best and nefarious at worst.