Readers Debate the Merits of Post-election Audits

by Wendy Underhill

The September issue of NCSL's elections newsletter, The Canvass, addressed what I thought was a sleepy topic: post-election audits. (As a way to double-check that the procedures, voting equipment and vote-counting software yielded the correct result, election officials run a post-election audit by hand-counting the ballots from a random set of precincts or machines.)

So I was surprised that this issue received more responses than politically-charged and publicly debated issues, such as those on Voter ID or Voter Registration.

A professor of state and local government affairs, wrote, "I had never heard of "post-election audits" or the idea that such things might be mandated by state law.  I would guess a lot of legislators have never heard of such things either. "

Kim Alexander, president and founder of The California Voter Foundation, wrote, "I really appreciated the list of reports and studies featured in the newsletter relating to PEA's." And she added a report that I had missed: Post-Election Audit Standards Working Group Report, from California's Secretary of State.

And then there was one full-on critique. I quote here from Frank Sampsell, a former election official in Berrien County, Michigan:

I don't necessarily condemn post-election audits, at least not in principle.  I see the validity of audits as spot checks for performance or procedural issues.  However, I know there has been a call placed by our colleagues in the election community to begin drumming up data on all things elections.  I would ask legislators to demand this data before enacting laws which may eventually produce virtually no return. 

In the case of post-election audits, I think this data would be relatively easy to develop, especially as some states already have these audits on the books.  How long does the audit of a single precinct take?  How much does that practice cost?  What percentage of these audits yield a significant issue?  What types of issues are found?  How often do these issues endanger the legitimacy of the election outcome? 

At some point, the field of elections administration really does need to consider the least popular of all questions:  Which is a greater danger to our republic?  Fractions of 1 percent of our population being disenfranchised due to error, or hundreds of millions of dollars being spent across the country to clean up those errors?  Patriotically and ideologically speaking, there is something un-American about accepted the first option, but economically and administratively, there is legitimate credibility to the second point--if we already get it right 99.5% of the time, how much is that last .5% worth?  Moreover, if that last drop in the bucket is caused by human error (as would likely be caught by an audit), will it ever be eradicated by the addition of any new procedure?  

Besides, if there is any one lesson the past decade has taught us, it is that virtually any significant election that finishes within 1 percent in the tally is eventually going to be decided by the Judicial branch.  

His final comment is an excellent segue. The October issue of The Canvass will address election law and the courts. If you have thoughts on either post-election audits or election law challenges—or if you would like to receive a copy of Frank's full comments, please let us know.

Comments

[Paul Stokes]

We have an effective post-election audit in New Mexico.

The cost of a good audit will cost a few million dollars across the country, assuming a close race - say, a 1% margin.

In addition to spotting errors, a benefit of the audit is to deter fraudulent manipulation of the vote which has been shown to be feasible with voting machines operated by software. It is hard to know what that deterrent is worth.