Some More Thoughts on Election Reform

Here are some edited versions of some postings I've recently made to the election law listserv:

Steve Ansolabehere and Mike Alvarez made the following observations in a 2002 Demos report: "“Voter fraud is a persistent concern of those who run elections. Fraud is hard to detect and to measure. It may involve only enough ballots to win an election, and there may be no paper trail that would allow election officials to determine what exactly occurred.” R. Michael Alvarez & Stephen Ansolabehere, California Votes: The Promise of Election Day Registration 14 (2002), available at: (last visited Feb. 16, 2005).

But the difficulty of proving actual fraud doesn't mean it doesn't
exist. Beyond anecdotal studies, there have been some recent newpaper
studies of double voting by voters living in two states. (For
citations to those newspaper studies, and more discussion about the
fraud issues, seee the text accompanying footnotes 113-116 of my Beyond
the Margin of Litigation draft posted at:
http://ssrn.com/abstract=698201).

The bottom line: we know some fraud happens but we don't know how much
of it there is. But we also know that a large number of voters believe
that there is a great deal of fraud in elections. (see pages 7-10 of my
article for some rather stark statistics in this regard).

Voter i.d. can help to alleviate concerns about fraudulent voting and
help restore some voter confidence in the process. The problem with
voter i.d. alone is that while it helps some voters with confidence, it
can suppress the votes of others, particularly the poor and mobile, who
will have greater difficulty providing i.d. It doesn't help to
alleviate concerns about vote fraud by raising fears about vote
suppression.

That's why a part of my proposed solution for falling voter confidence
is government issued voter i.d. with biometric information (like
fingerprints). If you show up with your finger, you don't need an i.d.
card. And I'd couple it with universal voter registration accomplished
by the government. These two items alone could reduce as much as 60%
of the litigation surrounding our elections and do much to restore
confidence in our elections.

One other argument that is made against voter i.d. is that it does
nothing to stop fraud by election officials. That's right. We need
protection from that as well, as well as the appearance of bias by
elected officials. For that reason, at the least all state chief
elections officers should abide by the IDEA Code of Conduct, and
refrain from participating in any political activities (like serving as
a chair of a presidential candidate's state committee). And ideally,
we should move to nonpartisan administration of elections, with checks
to make sure that no fraud happens at the level of election officials.
***
I agree with much of what Lori has to say on the need for more data. But I think even Lori would agree there is some strong evidence of, at the least, registration fraud, and in 2004 we have very strong evidence of some double voting across states. There was perhaps enough of it to have affected the outcome of a very close presidential election in Florida (if one had occurred again in 2004). I think there's little doubt that such reports undermine confidence in the electoral process.

I also believe there is real evidence of attempts to suppress minority voter turnout in the name of protecting the integrity of the voting process. How much of suppression goes on is also very difficult to measure (like fraud), but I have no doubt that some of it occurs, in some places. Because we can't measure it precisely, shall we decide not to address it?

We need a solution that gets beyond a data-poor dispute between those concerned about "access" and "integrity," toward a system that everyone can have confidence in. Most democracies require i.d. to vote, and many democracies have more centralized databases of voters so as to protect fraud. My twin proposal of government registration voters, with government issued voter i.d., should go a long way toward alleviating concerns over both access and integrity. On top of that, we need faith in election administrators, which is why we should join most democracies in using nonpartisan election administrators.

As for absentee ballots by mail, how about a fingerprint to go along with that ballot? The only objection I can see to this process is a privacy concern, but (as I argue in the paper), we need to balance privacy concerns over concerns about voter confidence in the electoral process.

Posted by Rick Hasen at April 22, 2005 08:55 AM