More on turnover in state houses

Karl Kurtz of The Thicket provides some additional information on legislative turnover rates in state houses.

Do Oregon Voters Really Find Election Day So Special?

In a recent editorial, the Oregonian asserted that more citizens are choosing to wait to turn in their ballots until Election Day. The piece claims that this behavior reveals a sort of synthesis of the pro- and anti-mail ballot arguments rolled into one: Election Day traditions are able to survive even while no one is forced to follow them.

This is a nice idea, and I have no doubt there are still quite a few citizens who vote on Election Day because that’s how their parents did it. But the data just does not support the claim that more citizens are suddenly beginning to realize the value (whether it be intrinsic—as the article asserts—or perhaps even utilitarian) of last-day voting.

Here is a graph showing the number of ballots casted on Election Day in Oregon elections from 2000-2010, as a percentage of total ballots submitted:

Percentage of Ballots Returned on Final Day of Voting

(Data found at: Oregon Secretary of State.)

Since 2000, the level of last-day voting has decreased a few times, but has regularly hovered around 25%. It is not that I think the Oregonian is plain wrong—I have no reason to doubt that a portion of the individuals voting on Election Day do so because their parents did the same—but these numbers do not reveal any sort of aggregate chance in behavior.

Notice much change? It is not that I think the Oregonian is utterly false—the 20-odd % that still chooses to vote on election day can make those decisions for whatever reason they want—but there is no trend in the last 10 years that seems to show Oregon citizens as changing their behavior in any aggregate way.

Now, the Oregonian may have a hunch about this new trend, but it will take additional research and evidence to convince me otherwise.

BTW – this post is not written by your regularly EVIC blogger, Paul. My name is Jacob Canter, and I’m the new RA for Paul. I’ll be adding to the blog every so often, hopefully providing something interesting to look at and think about. Feel free to ask questions and requests posts in the future.

Oregonians finally notice that voting by mail doesn’t necessarily mean by mail

Candidate Support by Result Reports, Portland Mayors Race

(Graphic courtesy of Kari Chisholm,

I have noted in the past that a substantial percentage of Oregon ballots are hand delivered by voters to county offices or satellite drop boxes on election day.

The state’s major paper, the Oregonian, seems to have finally taken note of this trend, likely because in the most recent mayoral contest, the first returns were based primarily on “by mail” ballots while the final returns included ballots dropped off on election day.  The results were quite different.  Early returns indicated that Charlie Hales was ahead by as many as 10 points over Jefferson Smith, while the final returns showed them apart by only 4.4% (37.2% to 32.9%).

Final returns are here and the hour by hour reports are here (scroll down to view reports 1-10).

There has been some local speculation about what this indicates about who supports Smith (young voters?  late deciders?) and how his GOTV operation worked.

What’s interesting to me as a scholar of early voting, however, is what this shows about the voting by mail system.  Observers who are less knowledgable about VBM describe the system as if every ballot came through the postal service, but election officials in Oregon, Washington, California, and other states know that a significant number of “by mail” voters still hold their ballot until the end and deliver it “in person” on election day.

Size matters. And so do a lot of other things.

I appreciated Doug Chapin’s posting about David Kimball (FULL professor now, folks) and Brady Baybeck’s paper titled “Size Matters in Election Administration“, presented at OSU Moritz School of Law’s “HAVA at 10” conference.

I’ll leave you to Doug’s posting for the nitty gritty, but I wanted to add an important thought for anyone who does comparative election study in the United States: because “size matters” so much in the U.S., a lot of other things matter as well, and it’s vital to take them all into account. It may be the case that large jurisdictions face different problems than small jurisdictions.

But it’s not enough to just show that large jurisdictions process, for example, 89% of the provisional ballots cast in the U.S., because large jurisdictions also 63% of the voters.  It’s the difference between the two–89%-63%–that is the quantity of interest. Furthermore, it may not be “size” that matters, but other things that covary with size: number of lower income voters, number of Latino voters, or the number of mobile voters.

My first takeaway from Kimball and Baybeck was: excellent first take at the disparate situation faced by jurisdictions in the U.S.

And my second takeaway was:  someone out there needs to connect the characteristics of LEO’s, jurisdictions, states, and citizens to really help disentangle these effects.  This is a great next project for some enterprising graduate student at CalTech, MIT, University of Maryland, Ohio State, University of Minnesota, University of Utah, or University of Missouri-St Louis (to name a few of the usual suspects!).

Minnesota discovers the dangers of complexity Minnesota discovers the dangers of complexity

I’ve been reading a lot more about public policy in the past few years, undoubtedly the result of the insidious influence of Doug Chapin and Thad Hall. The field has made great strides since I was in graduate school two decades ago. I cut my teeth on John Kingdon’s Alternative, Agendas, and Public Policies (now a Longman “classic”–what does that say about me!). Now the field is replete with “punctuated equilibria,” “policy narratives,” and “advocacy coalition frameworks” (all describe in Paul Sabatier’s classic text).

While these complex models of unpredictable systems may frustrate a quantitative generalist like myself, they are obviously necessary. And Minnesota’s recently passed election law demonstrates this fact as well as anything.

As Mark Fischenich writes so effectively in the Mankato Post, the law seems straightforward to legislators, but election officials realize that the “devil is in the details.” And a lot of devilishness there is!

Among the potential legal and administrative conflicts that were apparently not considered by legislators:

  • UOCAVA problems: if the law is read strictly, overseas (and other absentee) voters will be unable to cast a ballot because they cannot show an ID.
  • Same day registration problems: identity has to be verified “prior to casting a ballot”, but what does that imply for someone who shows up to register and vote at the same time?
  • Provisional avalanche: If all these same-day registration voters and voters without sufficiently validated IDs are given provisional ballots, will this result in an unanticipated avalanche?

I have written about the interdependencies of various aspects of election laws in the past, and I’m trying to finish a project on early voting that lays many of these out. But the examples and anecdotes keep changing. Voter ID adds a new layer of complexity!

Maybe Doug and Thad can sort this one out.

* Image courtesy of

Early and Absentee Votes in Cuyahoga County, Ohio

Rob Richie and I have been arguing for the use of ranked choice ballots for overseas voters, and potentially all absentee voters, in the presidential primary process. Our concern is that candidates who have withdrawn from the race remain on those absentee ballots, and overseas voters in particular have to mail their ballots back without realizing that some candidates have withdrawn.

The recent Ohio primary provides only mixed support. While we don’t have information on when the absentee ballot were returned, it does appear that Perry and Huntsman received a larger percentage of their votes on absentee ballots. Still, there are obviously a lot of election day voters who cast a ballot for one of these two candidates.

It might be interesting to observe “non running” candidates as a measure of voter dissatisfaction with the current crop of candidates. Nice dissertation topic? I’ve never seen anyone do this analysis before.

Cuyahoga Co., OH, 2012 PPP

Early voting updates

Apologies to our regular readers for my absence for a few weeks. I’m back to update you on all things early voting.

Some of you may have seen an editorial in Roll Call that Rob Richie and I authored, arguing for ranked-choice voting in presidential primaries for overseas absentee ballots. If anyone has reactions, I’d love to hear them. I think this is a great idea, and I’m pondering whether I should work with Rich to push this more systematically in a few states for 2016.

Voting reform is pushing ahead in Connecticut. It looks like online voting registration–an initiative of the Pew Center on the States--will be put in place. The legislature may also relax no-excuse absentee voting requirements. This means my standing comment in powerpoint presentations about the Northeast may need to be amended!

Early voting rates in Ohio seem to be lagging behind 2008. Some local officials speculate that it may because of a fear that candidates will drop out, but I don’t find that particularly convincing. None of the leading four candidates is showing any signs of withdrawing at this point. As Mike Alvarez has argued in his book, it is probably because voters remain uncertain and candidate support is fluid in the Buckeye State.

Another Ohio controversy is brewing over changes to the period for early voting.

Miami-Dade by mode of balloting

Another county’s results are plotted below, and while it’s too soon to assume a lot from just two counties, there is still a lot of election day votes cast for Cain, as a proportion of his total. Perry, Bachmann, and Huntsman look more like I’d have expected.

Miami-Dade, by mode of balloting

Romney and Gingrich did equally well in the early votes in Duval County, FL

Interesting breakdowns of the vote totals by mode of balloting in Duval County, FL. Gingrich and Romney performed about the same in early, absentee, and election voting. Perry got a lot more absentee votes, no surprise there. As to Herman Cain? Most of his votes in Duval came on election day!

Choice Polling Absentee Early Voting Provisional Unscanned Total Percent
Michele Bachmann 43 155 8 0 0 206 0.24%
Herman Cain 119 74 31 0 0 224 0.26%
Newt Gingrich 23,005 5,433 5,307 0 0 33,745 38.72%
Jon Huntsman 54 133 12 0 0 199 0.23%
Gary Johnson 14 24 7 0 0 45 0.05%
Ron Paul 3,821 1,206 662 0 0 5,689 6.53%
Rick Perry 79 234 17 0 0 330 0.38%
Mitt Romney 21,641 7,558 5,926 0 0 35,125 40.31%
Rick Santorum 8,388 1,683 1,509 0 0 11,580 13.29%