Over-Registered?

Perhaps someone can enlighten me on this. You often hear someone or other bemoaning how poor voter turnout is in the United States with numbers as low as 35% sometimes being cited. Is it true?

Since turnout is presumably calculated as the number of vote relative to the number of registered voters there are two ways that a low turnout could be observed: either, as is commonly believed, lots of Americans don’t vote or Americans are actually over-registered to vote in substantial numbers.

I doubt that information from the Census Bureau can be relied on for this purpose since I gather that they rely largely on self-identification which is completely unreliable. One methodology that would be better would be to start with the birth records, add the number of naturalizations, and apply the actuarial tables. That’s a pretty arduous task so I doubt that those who are complaining about Americans’ voting behavior are doing it.

It’s enormously easier to register to vote than it is to be removed from the voter rolls. Based on my own experience as an election judge I’ve seen people who’ve died or moved away being left on the rolls for five or even ten years.

Here’s a typical scenario. Kids register to vote in their parents’ home precinct. Over time kids move away, parents die or retire out of state. Those who’ve moved away may register to vote in their new homes and, presto!, the number of registered voters exceeds the number of people eligible to vote.

Here in Chicago the procedure for verifying the voter rolls goes something like this. Every so often the Board of Election Commissioners does what’s called “telephone canvassing”. That means that they call the telephone number they have on file for all of the registered voters in the city and ask if the voters still live there. People may not be home when they’re called, they may not answer their phones, or they may lie about who’s still living there and eligible to vote. Additionally, the number they’ve given may be a cell phone and not be a reliable gauge of residence.

The records of people who fail to answer or who have moved away are marked and the election judges are required to ask them for identification if they come in and vote. The identification they give may be obsolete—we have no way of telling. Indeed, it would be possible for the same person to be voting in person in our precinct and absentee out of state. There’s no systematic way of determining that.

Old-fashioned canvassing in which you go door to door and verify whether a voter still lives there is better than phone canvassing although a lot more expensive. To the best of my knowledge there’s been a real canvassing in Chicago three times over the period of the last 25 years. It has some of the same problems as phone canvassing does: people lie or are mistaken.

Our current system of voter registration is hopelessly antiquated, based on assumptions that haven’t been true for decades if they were ever true. As a people we are quite mobile. Relatively few of us live in the same houses or even the same towns our grandparents did and we may move to different places in the country several times in the courses of our lives.

Although the system clearly doesn’t insure that everyone who’s eligible to vote is registered, it does a lot better job at that than at ensuring that people are registered to vote legitimately. From a technical standpoint it would be possible to have a voter registration system under which everyone who’s eligible is enrolled automatically and their registration remains intact regardless of where they live. There are those who oppose such a thing on privacy grounds. I’m not sure how you can reconcile the privacy considerations with a registration system in which we can be confident and anything less than such a system undermines confidence in our system, generally.