The Fix: One key factor explaining voter registration differences by state: Ethnicity

By Philip Bump February 26 at 4:11 PM

The map below, spotted by The Fix's very own Chris Cillizza, shows a remarkable bit of data: the percentage of people in each state who are not registered to vote.

Notice that it isn't interactive; the arrows don't do anything. It's a still from a very cool tool created by the Public Religion Research Institute. (Which also lets you explore a bunch of other data, so go look at it.) (After you finish reading this.)

The question that could/might/should arise from that map is: What accounts for the discrepancy? Why are so many people in Missouri registered to vote and so few in, say, Texas?

The answer? States with higher Hispanic populations have lower rates of registration.

We can show that fairly easily by comparing the unregistered population (from PRRI) with the density of the Hispanic population in each state (from the Census Bureau). We've done the same with black population density, for comparison. As Hispanic population density increases, so do non-registered rates. A black population density rises -- it has no effect. (Each dot is data from one state.)

There are two reasons for this. First, Hispanic residents have lower citizenship rates. And, second, even among citizens, registration rates are lower. Using voter registration data from the Census Bureau for the 2012 election, we can compare those rates.

(Since the population of Asian residents is significantly smaller, there is a much smaller effect on statewide registration rates.)

Another way of looking at it: In most states for which there are enough Hispanic residents to be statistically significant, registration in that community in 2012 trailed overall registration significantly.

This is another reason why the threat posed to Republicans by America's shifting demographics is a lot further away than a year and 10 months.

Philip Bump writes about politics for The Fix. He is based in New York City.