What More Can We Learn from SC?

Doug Chapin’s post today on his blog digs down into the Department of Justice’s data request from South Carolina, seeking more detailed data concerning who does, and who doesn’t, have the identification required to vote in that state, as a consequence of their new voter ID law. I agree entirely with Doug’s top-line reaction — At last! Some real data.

At the same time, the request seems to miss an opportunity to find out more about whether voter identification laws will have a disenfranchising effect, and in particular, a disproportional effect on minority voters. The reason is that the disproportional effect may not be so much on whether whites and blacks have drivers licenses, but whether they have drivers licenses with the voter’s current address.

African Americans are more mobile than whites. When there is mobility, there is the chance that the voter registry and the drivers license registry will be out of sync, since in most states the two lists are segregated in real-time.

In the most recent Voting and Registration Supplement released by the Census Bureau as part of its Current Population Survey, 38% of white respondents in 2010 reported they had lived in their current address for four years or less. This compares with 50% of African American respondents.

Turn now to drivers license statistics. In 2008, several colleagues and I conducted the Survey of the Performance of American Elections. One question we asked was whether the respondent had a drivers license. Of those who answered “yes,” we followed up, to ascertain whether the DL had the respondent’s legal name, current address, and whether the license had expired. Nationwide, 3% of whites, but 19% of blacks said they did not possess a drivers license. Furthermore, 9% of whites and 16% of blacks who had a DL said the address on the license was not their current address. Put these two statistics together, we have 87% of whites and 68% of blacks with a drivers license that has the voter’s current address on it.

We only had 200 responses from SC, so it is difficult to know whether this pattern extends to the Palmetto State. Among the 141 white respondents, 98% had a drivers license, 90% had a drivers license with the correct address. Among the 56 black respondents, the corresponding percentages were 87% and 78%.

The racial disparities in South Carolina on both scores — having a DL and having a DL with the correct address — are smaller than the national differences, but they still exist. At least that’s what the survey research evidence stays. The real question is whether the voter and drivers license lists say that, too.

I first encountered the voting/DL registry mismatch when I was observing the polls in Massachusetts in 2004.  There, the issue was not mobility, but the common practice of keeping your car registered (and your drivers license address) in a small town far from Boston, even if you live in the city, in order to keep insurance premiums down.  I doubt the same games are played in South Carolina, but I do wonder what other folk practices have emerged to drive a wedge there between the accuracy of the DL list and the voting list.