[EL] re “Kansas’ state voter ID law tested in August primary”

[Robbin Stewart]

http://electionlawblog.org/?p=40653
The 2012 Kansas primary gives us some new data on the effects of voter ID.
We learn that of 405 provisional ballots for not having ID, 251 did not
return with more ID and did not have their ballots counted. So, in the
population of people casting provisional ballots, voter ID requirements
deterred 60% of these people from finalizing their votes.

This is voter disenfranchisement in a weak sense; they tried to vote but
their votes were not counted. In most cases probably, it is not voter
disenfranchisement in the stronger sense of, most of these people could
have voted if they put more time effort and money into it, but they didn't
bother. It is not entirely reasonable to extrapolate this more generally
and say that voter ID keeps 60% of voters away from the primary; the
populations are not exactly comparable. However, now we have some numbers
where before we had guesses.
If our concern is with the integrity of the elections, it should matter
that voter ID requirements deter 60%, at least of this small sample.
Primary turnout is already low enough that some people think the result
promote extremists (like Kobach), and that the focus for election integrity
should be on getting more voters, by reducing the costs of voting.
As far as I know, none of the litigation so far has really focused on this
point, deterrence rather than disenfranchisement in the strong sense. The
state interest typically cited by states is election integrity. Instead of
focusing  argument on how individual voting rights outweigh the state
interest, one could argue that the means are a poor fit, and do not
actually promote election integrity after all, so if one is doing an
Anderson balancing test, there's nothing on that side of the balance. 251
votes out of 1/3 of a million is not a lot, so in one sense the program is
a "success" in that it does less measurable harm than some predicted. My
sense is that voting is like an iceberg, 9/10th invisible. If there were
405 provisional ID-related ballots, there were probably 4050 people
deterred from voting (and 4 or 5 actually fraudulent votes deterred.) But I
have no way to prove or measure this. In my neighborhood we often have
elections decided by a handful of votes. I did not get around to attending
the Libertarian convention this year in Las Vegas as a delegate, as I have
sometimes done before, and a guy know pretty well lost the party
chairmanship election by one vote on the 5th ballot. In the Indiana primary
this year, I was turned away from the polling place because I'm not willing
to show a voter ID. This time, I didn't bother to cast a provisional
ballot. They don't, in Indiana, keep any records of how many people show
up, want to vote, but are sent away for no ID. Having such records might be
useful. Eventually Indiana's voter ID will be back in court on an
as-applied, data-based challenge,and the more data there is the more the
court can fairly engage in balancing competing interests.

I just wanted to point out that the kansas numbers can be read one way, as
a great success for voter ID, but looked at another way, show a great
failure.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20120925/79a6b215/attachment.html>