A student of mine just drew my attention to the Pennsylvania Voter ID Handout, which appears to be aimed at educating voters about the new voter ID law.

Two things occur to me.

First, the second page is a visual mess.

Second, the language describing the requirements of the bill is not exactly aimed at voters who may have low levels of education — precisely those voters who are likely to lack the correct ID in the first place.  When I run the text through a Flesch-Kincaid readability calculator, it comes back with a 16th grade reading level (i.e., college graduate).  The readability score (20) means its “very difficult to read” under this metric.

Are there better designed and/or more readable notifications out there, for the new Pennsylvania law?  I would be happy to post examples.

This article notes an unfortunate ruling from a federal district judge in Nevada, striking down that state’s generation-long practice of including the choice “none of these candidates” (i.e., “none of the above”) on its ballots.

Although the suit was brought by the Republican Party, which was worried that this choice would siphon votes away from Mitt Romney, my beef isn’t partisan. It’s that the presence of this choice on the ballot helps to us gauge the lower bound of the residual vote rate, which is the most direct measure we have of voter confusion and machine malfunctions. This choice gives voters who want to abstain a way of indicating their preferences explicitly, making it clear that voters who cast a blank ballot (or who over-vote) have made an error of some sort.

The election geek in me hopes the ruling is overturned.

Professor Elizabeth Bergman at Cal State University East Bay has designed a voting app for iPad, iPhones and Android devices. The Voter Guide Now app was developed to provide an easy method for voters to access information about candidates and legislation on federal, state, county and city ballots. All the user has to do is enter an email and street address. You can read more about the app here. This is interesting and innovative work and I commend Professor Bergman for her efforts in helping to build greater voter literacy.

There is a recently published paper on election fraud by Andrew T. Little in the Quarterly Journal of Political Science, titled “Elections, Fraud, and Election Monitoring in the Shadow of Revolution.”

Here’s the abstract:

Elections are modeled as a public signal in an incomplete information game of revolution. By changing beliefs about the general level of anti-regime sentiment, elections can make citizens more or less apt to rebel and hence make a successful revolution more or less likely. This effect makes elections valuable to incumbents that are not secure in office as they have more to gain by good results than they have to lose from bad results. Electoral fraud is modeled as a distortion of the public signal, and election monitoring is incorporated as changing the cost of this distortion. In equilibrium, citizens discount the distortion, so the average protest size and probability of revolution are the same as when the incumbent cannot commit fraud. This makes election monitoring valuable to incumbents as it ties their hands and lowers the equilibrium amount of fraud. So, elections may be held that would not occur in the absence of monitoring.

Today’s primary election day here in California, and once again I’ll be out in polling places watching and observing the process. Updates on that as the day progresses.

But I had an interesting experience this morning when I went to vote at my local polling place with my daughter.

I obtained my ballot and inserted it into the Inkavote Plus reader. After working my way through the candidate races I noticed that the vote recorder didn’t seem well aligned with the positions on the ballot underneath — so I wiggled the ballot a bit so that it seemed better aligned with the vote recorder for the measures at the end of the ballot. I pulled the ballot out, noticed the alignment of markings wasn’t quite right, and showed it to the poll worker; her response was to tell me to put it into the Inkavote Plus scanning device and see what it said.

Well, the ballot was rejected — and I got a little paper receipt saying that I had overvoted in one of the candidate races. That was a surprise, I’ve never overvoted before!

So they voided that ballot and gave me another. I repeated the procedure but on a second vote recorder, and still noticed some alignment problems with the vote recorder when I got to the measures at the end of the ballot. I voted, making sure to mark my ballot well. This time, when I pulled the ballot out, I looked at it closely — and indeed, there was a stray ink mark right below one of my candidate choices in that same election that I had overvoted before. I showed it to the pollworker, and I compared it to the sample ballot. No question that the stray mark on this ballot was the problem. The pollworker advised checking it in the scanner, and again — Ballot Alert! Overvote in the same election!

Now I was quite perplexed. But my guess was that I was somehow shifting the ballot in the vote recorder such that it was smearing some ink into some other ballot positions in that one candidate race, or that I was marking the ballot too hard. So they voided that ballot.

The pollworker got a sample ballot, and we both positioned it in the voter recorder, and shifted it around until it looked well aligned. I pulled it out, and got my third ballot (this was my final attempt). I positioned the ballot in the vote recorder, and flipped from front to back to make sure it was well aligned before I started. I voted a third time, making sure not to move the ballot at all as I marked it.

I looked it over carefully, and it looked okay. We put it into the scanner. The good news — no overvote! So the third time was the charm. All in all, not a big deal as the pollworker was pleasant and helpful, and the technology seemed to be catching whatever error was caused by my interaction with the vote recorder. But an odd experience nonetheless, as it’s never happened to me before!

On Friday I gave a talk to the nice people at the annual Social Science Librarians Boot Camp, held at Tufts University, about using data to judge whether elections have gotten better in America over the past decade.  That talk has given rise to me corresponding with social science librarians around the region, who get requests for election data, and don’t always know how to refer people.  One of the resources I’ve been referring them to is Pew’s Election Administration by the Numbers, which should be better known as a primer to the uses of election data to tell a story about the function (and functioning) of elections in America.  Check it out if you haven’t seen it.  (The shameless plug aspect of this is that I had a hand in producing the document.)

Tomorrow’s election in Wisconsin will be one to watch carefully, as only a few times in American history has the recall process been used to throw state Governors out of office (I believe it has only been used successfully twice before, here in California in 2003 and in 1921 in North Dakota). We’ll be closely following the Wisconsin elections tomorrow, in particular for news of problems the polls and voter confusion.

I thought I’d repost this, originally posted on March 13, 2011, “Wisconsin Recall Election — Lessons From the 2003 California Recall.”

I’ve been hearing a bit of buzz in academic circles recently about potential recall elections in Wisconsin.

For a lot of reasons, recall elections haven’t been widely studied by political scientists: they are not terribly common in the types of elections we tend to study, and when they do occur they are often over and done before anyone has had a chance to mount a large-scale research study.

Of course, the 2003 recall in California was an exception, and there have been a number of important studies of that election. That body of research on the 2003 recall in California should be of great interest in the contemporary discussion of potential recall elections in Wisconsin.

First, there was a symposium in PS: Political Science & Politics published in 2004 on the California recall election. There were a number of papers in that symposium that are of interest to those interested in learning more about recall elections; readers of this blog might especially like the paper that Melanie Goodrich, Thad Hall, Rod Kiewiet, Sarah Sled and I published in that issue, “The Complexity of the California Recall Election.” Our paper discusses the administrative complexities of recall elections, and the effects of that complexity on voters.

Second, there was an volume published in 2006, edited by Shaun Bowler and Bruce Cain, “Clicker Politics: Essays on the California Recall”. This volume has a number of important studies of the 2003 recall election, especially papers on some of the legal issues of the election (Rick Hasen) and a variety of papers on behavioral issues in the recall election (for example, Becki Scola and Lisa Garcia Bedolla on race and gender; Matt Barreto and Ricardo Ramirez on race and the recall; Betsy Sinclair, Rod Kiewiet and I on rationalistic voting behavior).

Finally, there are a number of independent studies that consider important political and behavioral issues regarding the 2003 recall election:

Elisabeth Garrett, “Democracy in the Wake of the California Recall,” University of Pennsylvania Law Review.

R. Michael Alvarez and D. Roderick Kiewiet, “Rationality and Rationalistic Choice in the California Recall”,British Journal of Political Science.

Gary M. Segura and Luis R. Fraga, “Race and the Recall: Racial and Ethnic Polarization in the California Recall Election.”American Journal of Political Science.

This is by no means an exhaustive list of studies, but these are just the ones that I have drawn upon in my recent research.

This story from The Thicket, an online blog of the National Conference of State Legislatures, is a cause for concern. A likely close election, fiscal and pension crises (compare the fiscal health of states at the Pew Center), and a burst of inexperienced legislators is a recipe for potential legislative follies (a link to David Canon’s classic study on amateurs in Congress).

There will surely be a burst of activity on the elections front.

Here’s a toast to Tim Storey and the NCSL staff, who will have to redouble their efforts to help provide information and guidance to these new legislators in 2013.

Back in 2000, when the Caltech/MIT Voting Technology Project began, we had a lot of trouble tracking down reliable data on the costs of election administration throughout the US. That difficulty led us to recommend to some of our graduate students that they look into this important question, and luckily for those of us who study election administration, Sarah A. Hill (now a professor at CSU Fullerton) took the challenge.

Sarah’s study is in the American Review of Public Administration, and here is the abstract for her paper, “Election Administration Finance in California Counties”:

Over the past decade, the federal and state governments have made large financial investments to improve election administration, but there is little to no understanding of the real workings and implications of election administration finance. This article takes a first look at election administration finance by examining election expenditures in California counties for fiscal years 1992 through 2008 using a public sector cost model. Regression analysis shows that economies of scale and voting technology are significant determinants of election expenditures, as are other factors affecting the cost of the production of election administration. Factors that are expected to affect the demand for election administration are generally shown not to be significant. These results will hopefully be beneficial for policy makers as they face important decisions about changes in voting technology and election administration.

I hope others will do similar studies in other states!

Charles Stewart, Paul Gronke, and I are attending the HAVA@10 Conference at the Moritz College of Law at The Ohio State University today.  The morning panels covered What HAVA Did and Did Not Do and about Federal and State aspects of HAVA.

Charles and I spoke separately but both covered the history of HAVA and the issues the led to its enactment and why it looks as it does.  I discussed the voter registration component of HAVA and how the language of HAVA – and the goals of the commissions and research groups that helped to shape HAVA – recommended the creation of statewide voter registration databases and how the implementation of these systems varied across states.  Some adopted the top-down statewide system discussed in HAVA and some states developed “bottom up” systems that were more of a stitching together of local systems.  I presented some data showing that top-down systems may outperform bottom up systems when it comes to voters reporting problems with voter registration.

Charles noted how the problems with elections were first identified in Florida, how these changed to some extent when the Congress started to address election reform (adding, for example, disability access to the pool of issues being considered), and how the HAVA law and implementation again are different.  He noted that much of election administration is governed by state laws that are based on years of tradition and culture that is hard to overcome.  For example, Charles noted that there may be differences between how states that had provisional balloting before HAVA treat provisional ballots compared to states that had provisional voting pushed on them by HAVA.

Dan Tokaji discussed the legal aspects of HAVA in the courts and found, in his study, that there are relatively few lawsuits use HAVA as the basis for election-related lawsuits.  Part of this might have to do with the lack of any private right of action in HAVA.

David Kimball did a nice presentation noting that counties have very different capacities and different demands.  He discussed three types of counties:

1. SMALL (Less than 1,000 voters), where the local election official (LEO) is like a school principal,
2. MEDIUM (Between 1,001 and 50,000 voters), where the LEO is like a restaurant chain owner, and
3. LARGE (More than 50,000 ballots), where the LEO is like the CEO of a large corporation.

What is interesting is this:  there are 5,149 small jurisdictions but they only serve 1.8 million voters.  The 457 large jurisdictions serve 88 million voters.  The 4,893 medium jurisdictions serve 43 million voters.  Most voters live in large jurisdictions but most jurisdictions are small.  David argues that size matters.  Large jurisdictions have lower residual vote rates compared to small jurisdictions but they also report more problems in the voting process, largely because they have more challenging populations to serve – more mobile, more diverse.