Monday, October 13, 2014

Lakisha and Jamal Go to the Polls

Also, now they're Latino.

A very interesting study out of USC tests responsiveness of legislators to Latino versus Caucasian citizens with concerns about how to vote. State legislators received the following message:
Hello (Representative/Senator NAME),

My name is (voter NAME) and I have heard a lot in the news lately about identification being required at the polls. I do not have a driver’s license. Can I still vote in November? Thank you for your help.

(voter NAME)
For half, the voter's name was "Jacob Smith," and for the other half it was "Santiago Rodriguez". In all of the states tested, the actual answer to this question was "yes" (a driver's license is not required to vote).

The study found that legislators were considerably more likely to respond to Jacob Smith than Santiago Rodriguez. And that gap exploded when one compared voter ID supporters to voter ID opponents. Opponents of voter ID responded the Caucasian-sounding constituent 50% of the time compared to 43% of the time for his Latino-sounding peer (a seven point gap). For proponents of Voter ID, by contrast, that split was 45/27.5 (a gap of 17.5 points). In other words, proponents of voter ID are far less likely to respond to Latino constituents who have simple questions about the voting process.

This gap still exists for opponents of voter ID, but it is almost purely a result of partisan differences. Republican proponents of voter ID evinced a nearly 40 point gap in response rates between White and Latino constituents, while Republican opponents of these laws showed only a 16 point gap. Democratic opponents of voter ID laws, by contrast, responded to White and Latino voters at equal rates (within the margin of error). There were too few Democratic supporters of voter ID laws to measure.

The draft paper is available for download here.


Ben Faber said...

Interesting, disturbing, relevant. Not entirely surprising, but data is always better than inference. Is this sort of finding something that courts can meaningfully look to when they interpret / review voter ID laws?

I also have a tangentially related question I am hoping you can answer. In reading the recent ruling by 5th Circuit Appeals staying the District Court's injunction of Texas SB 14 on voter ID (5CA ruling here, I was confused by the court's analysis of 'success on the merits'. The court basically says "The state's claim has merit that a stay should be issued now" because the election is too soon (pg 9). But isn't that a question of procedures, and success on the merits would mean the appropriateness of SB 14 in general / in future elections?

David Schraub said...

Depends what you mean by "can". Courts have been very skeptical of even rigorous evidence of generalized race discrimination (e.g., in criminal convictions) because it cannot show that any particular actor was motivated by race (see McKleskey v. Kemp).

On the latter question, normally you'd be right, but the court seems to indicate (not unreasonably) that Purcell stands for the proposition that part of the "merits" prong in the election context is whether the lower court order would substantially disrupt an eminent election. From my vantage point, this would be better folded in to the irreparable harm prong; but I'm sympathetic to the 5th Circuit because the Court wasn't entirely clear in Purcell how this special election-related-rule impact/supplemented/supplanted the general rules for issuing a stay. That's my off the cuff take, anyway.

Ben Faber said...

Re: SB 14, thank you, that makes sense.

Re: statistical evidence, thanks for the reference, having looked it up I think I understand positions on both sides about the distinction between discriminatory intent and discriminatory act. It seems problematic that statistical data collected under methodologically valid circumstances is just about the best empirical assessment of truth we can make, except methodologically valid scientific experiments, and yet the courts are reluctant to employ these data. (Especially given psychology findings that challenge the model of humans as unified reasoned decision makers with free will, the distinction between effects and acts is tenuous.) Are you aware of legal philosophy / theory projects ongoing that attempt to incorporate this newer understandings? This might best be done legislatively rather than by judicially, but I think it is more best done than left undone.