A standard line in my opening lecture to new students of political science is some variant of this:
Studying politics can be exciting and can be frustrating because political actors are also strategic actors. They make the rules, the break the rules, and then they rewrite the rules. While you may be able to generalize about political actors, it’s very hard to generalize about political outcomes.
This lesson applies to this year’s coverage of early voting. Both campaigns have learned lessons from past elections. Both campaigns have been monitoring legal changes in the states. And both campaigns are spending millions of dollars trying to mobilize the early vote by whatever means necessary.
Image courtesy of Q2learning.com
Early voting is a moving target, and shooting at the bullseye in 2008 is almost surely going to miss the target in 2012.
Coverage a week ago was trumpeting a Democratic absentee advantage in Iowa. Today’s story in Politico? “GOP Gains Ground in Iowa Early Voting.”
If you line these and other stories up, it’s clear that the Obama campaign focused a lot of effort on recruiting more Democrats to apply for and cast absentee ballots in Iowa. Result: an early Democratic surge in absentee votes. Reaction: Romney campaign has redoubled their mobilization efforts in response.
As I posted a few days ago, the same thing appears to be happening in Florida. Given the uncertainty over the early in-person voting period, the Obama campaign redirected resources to encourage Democratic-leaning voters to request and cast absentee ballots. Result? An “advantage” for the Democrats in absentee ballots!
This is why I’ve been resisting making broad conclusions about what these early early voting numbers mean. Not every commentator has been so circumspect. This might make for a nice story that will be forgotten a week later (see: Iowa) but it doesn’t make for informed political commentary.
Image courtesy of the Pew Center on the States
I think Doug Chapin is seriously underselling the benefits online voter registration in his recent posting about Pew’s OVR initiative.
It may seem counter-intuitive, but online voter registration is not just about adding more citizens to the voter rolls. In fact, that might be it’s least important contribution.
OVR should result in registration rolls that are more accurate, more efficient, and cost a lot less money to maintain.
These are the takeaways from Pew’s report on the current voter registration systems: Inaccurate, Costly, and Inefficient: Evidence That America’s Voter Registration System Needs an Upgrade. The report authors write:
At a time when government budgets are significantly strained, our antiquated paperbased system remains costly and inefficient…
The paper-based processes of most registration systems present several opportunities for error…
Election officials administer a system that is fundamentally inefficient in a number of ways
It may be the case that these new systems have resulted in a higher increase in the number of new registrants than we would have otherwise seen in a presidential election year, but until we can conduct some comparative analyses after the election, we won’t know the answer.
We do know for certain that the registration records added in the thirteen states are far less likely to be invalidated due to errors, can be quickly and easily cross-checked with other statewide data systems if a state chooses to do so, and have cost the states less money than the old paper forms.
Cheaper, less prone to error, less vulnerable to fraud. And maybe more registrations than before.
Just like voting by mail was initially misbranded only as a way to increase turnout, OVR is not only about more registrants. It’s about taking advantage of technology to modernize our elections system.
What’s not to like?
A Boise Public Radio story today describes Oregon and Washington, the only fully vote by mail states in the nation, as late to the party:
But the push for “early voting” across the country is making vote-by-mail states look like late arrivals to the party. In Idaho, voters in some counties have been going to the polls since late September.
Here’s an alternative interpretation: Oregon and Washington realize that it does not take two months to deliver a vote by mail ballot a few miles verssus the few thousand miles that it takes to deliver a UOCAVA ballot.
Perhaps Oregon and Washington are latecomers to the party. All that is left on the table are a few meager morsels. The bar closed long ago.
Perhaps they are not late to the party after all. Perhaps the thirty states that mailed their absentee ballots in September (led by North Carolina, a superbly administered state, yet mailed ballots way back on September 6th) are like those early arriving guests, knocking on your door when you don’t even have the hors d’oeuvres ready. Give them some cold cheese slices!
Somehow, Oregon and Washington manage to mail their ballots just over two weeks before Election Day yet still rank near the top in terms of voter participation. It seems to me that the two states time things just right, and it’s those states that encourage voters to cast a ballot two months before Election Day that may need to rethink things.
Image courtesy of the NY Times
There is a really cool graphic in today’s 538 blog at the NY TImes that is making its way around the internet. The graphic is creative and awfully pretty, but the focus on individual states as independent entities, “swinging” in response to individual presidential candidacies, is almost certainly wrong.
As Jim Stimson showed a decade ago in Tides of Consent, and Ben Page and Bob Shapiro (find me in the acknowledgements!) showed two decades ago in The Rational Public, the nation as a whole has swung in fairly consistent patterns between liberal and conservative policy positions.
Image courtesy of the Policy Mood project at UNC
What’s revealing to me in the Times figure is which states maintain a position consistently in the middle of the distribution, the consistent swing states, but also those states that move most in response to individual candidacies.
The ability to mouse over and view a state’s trajectory is very instructive. I can imagine every state politics instructor today is showing this to his students.
But let’s not overstate the independent movement of individual states, as the authors do at the start:
The latest FiveThirtyEight forecast shows many states shifting to the right. Florida, North Carolina and Indiana are more likely than not to shift back to Republicans.
The nation as a whole is shifting slightly back to Romney. We only pay attention to Florida, North Carolina, and Indiana because they are on the cusp.
I don’t want to suggest that there are not unique, idiosyncratic policy issues and ideological responses tied to a states history or political culture, but what I see primarily in the graphic is a reflection of policy mood, not states swinging back and forth on individual trajectories.
Absentee ballot returns are piling up in Florida, and both campaigns are trying to spin the numbers.
Bloomberg reports that 275,000 ballots have come back, 44% from registered Republicans and 40% from registered Democrats. As long time readers of this blog know, Republican voters have historically used the absentee / by mail option more than Democrats, so the fact that these numbers are so close leaves Dave Weldon, a former member of Congress working on the Romney campaign, “certainly concerned.”
You haven’t seen any analysis of these data here, however, because of an anomaly in Florida law. Absentee ballot requests and returns are
…exempt from state public records law. Political parties and candidates are able to obtain and release the information.
I describe this as anomalous because Florida early in-person voting returns are easily accessible on the Internet. There is no reason to provide one piece of information and not the other, since they are in essence the same thing–early ballot returns–and only vary by mode.
This makes it very difficult to assess the real information from the absentee returns. Both campaigns have surely compared the return rates to voter history files, which provide information on past voter turnout, including the mode of balloting. This same information is contained in the vote mobilization files assembled on the Dem side by Catalist and the GOP side by VoterVault. The rest of us are forced to rely on the information released by the campaigns and by overall totals released by the state.
And is there good reason to think there has been change since 2008?
Yes there is–as Bloomberg reports:
Democrats have closed the gap by targeting voters who might not have otherwise cast a ballot, said Eric Jotkoff, communications director for Obama in Florida. The Democrats’ focus on absentee voting followed a state law last year that cut early voting at the polls from 15 to eight days.
This makes it very difficult to draw any firm conclusions from the absentee voter data at this point.
Somehow I missed the publication of this article by Melanie Springer, “State Electoral Institutions and Voter Turnout In Presidential Elections, 1920–2000” in State Politics and Policy Quarterly (gated).
The abstract:
Expansive and restrictive state electoral institutions have been instrumental in structuring the vote throughout American history. Studies focused on a small number of reforms, years, or states lack the scope necessary to comprehensively evaluate the effects of institutional change over time. This work, however, places recent reforms in historical context and offers a long-term perspective. Using an original data set, it identifies the institutions that have generated the most substantial effects on state turnout rates during presidential elections from 1920 to 2000. Findings demonstrate that restrictive laws (those aiming to limit the vote or make voting more costly) produced large and consistently negative effects in the Southern and non-Southern states alike, but the effects associated with expansive reforms (those making participation more convenient or less costly) vary. Although a few expansive laws have increased turnout in the non-Southern states, they have had no effect in the Southern states where turnout rates are lowest.
My day in Nashville has been wonderful–thanks again to John Geer and the Center for the Study of Democratic Institutions for hosting me.
My lecture had one lowlight and one highlight. I wonder if the highlight is shared by my good friends in the elections community such as Doug Chapin, Charles Stewart, Dan Smith, and others.
The lowlight is easy: I never realized that this was a course on elections, not introductory American politics! When I answered in response to a question that the demographic profile of the early voter mirrored many of the biases in American politics that they have “read about in the interest groups chapter”–even using the “what accent the heavenly chorus” quote from Schattschneider–I wonder if the students knew what class they were in.
To Geer’s credit, he told me afterward that the students will probably be panicked, and maybe that’s a good thing! Continue reading
This is a really well-implemented tool, and I encourage everyone to look at it.
I’m not quite sure what to conclude from the early voting numbers, plotted here and available by clicking this link (bloggers who want to grab a permalink from this site–click on the “share” button and you can grab the URL).
On the one hand, they show a pretty consistent 15 point advantage in the Obama vote among those respondents who say they have cast an early vote.
I looked more closely at the data, and they show 36.9% of the respondent pool thus far say they are Democrats versus 31.7% Republican. That is 3% above (Dem) / below (GOP) Gallup’s current estimate of party affiliation among likely voters.
What makes this hard to evaluate is that the states which currently have sent out absentee ballots are not a random subset of the nation as a whole. Nonetheless, in another week or so, these figures might start to give us a real sense of how the early vote is shaping up.