Saturday, October 18, 2014

Errors of omission to discourage voting

There was an interesting study done by Christian Grose at the University of Southern California on human behavior using voter ID laws and political party.  They sent a simple question to 1,871 legislators in 14 states with large Latino populations.   “Do I need an ID to vote?”  And the emails were signed Jacob or Santiago, to see if there would be a difference in response rate to the different ethnic names.

The answer in all of these states was a definite “No.”  No Voter ID laws had been passed in any of the states.  The researchers were not studying whether the lawmakers would lie about it.  They were simply measuring whether the lawmaker’s office sent a reply.  If not, the assumption was that by not responding, the lawmaker could discourage poorer citizens, who often trend Democratic and are less likely to have ID, from voting because if they didn’t have ID they wouldn’t know they didn’t need one. Of course, lots of politicians are too busy to reply, but that shouldn’t be different between political parties.  So any significant difference could be due to this bias.

They did find a difference.  Republican legislators were less likely to respond than Democratic legislators.  But this difference was small.  The big finding was between Republicans who supported Voter ID laws and those who didn’t.  Republican legislators who supported these laws were 40% less likely to respond to the question.  This was “one of the largest gaps” the researchers had seen in any of their studies.  It doesn’t prove cause and effect.  Perhaps Republicans who support Voter ID laws coincidentally were more busy during this period that others for some other reason.  But . . .