As we move towards election day, we will begin to see predictions about turnout. Of course, predicting voter turnout is even more of a guessing game than reading economic indicators, and Secretary of State Matt Dunlap has been even less competent at predictions than the state’s chief economist, Charles Colgan.
We saw this just last year, when Dunlap warned clerks to prepare for high turnout, predicted it himself, and was, well, wrong. Of course that’s not quite as embarrassing as some bloggers who reported that there was high turnout based on that prediction. That’s like trying to write a biennial budget based on Colgan’s projections. Oh wait, that’s what the Legislature always does.
No doubt in the next few days we’ll see a series of statements from Dunlap predicting huge turnout. He says these sorts of things every election, and he’s even occasionally right. In 2008 he was not – turnout was actually slightly lower than in 2004. The numbers below come from the election with the highest vote total. This is actually slightly less than the total ballots cast, but Maine only recently began keeping track of that number.
As you can see, Maine turnout has generally trended upwards, but is unpredictable. For referendum-only elections – 2001, 2003, 2005, and 2007 – the highest turnout was 2003. Like 2009, 2003 was a busy referendum election. The 55% state education mandate was on the ballot, as was the racino initiative, the tribal casino vote, and three separate bonds. Turnout rocketed up to just above 51% – which remarkably was actually higher than the turnout in the 2002 Governor’s race.
2005 and 2007 were much lower vote totals. In 2005 the gay rights referendum shared the ballot only with bonds, while in 2007 the only non-bond issues on the ballot were the Washington County tribal racino initiative and a term limits extension referendum. With 2009 being the most crowded referendum election since 2003, it’s reasonable to expect higher turnout. Maine media will likely report Dunlap’s predictions without much questioning or analysis. In fact, they too tend to just predict higher turnout each election cycle without much analysis.
No doubt next week we’ll see a wave of these articles. We have already seen some reports on the number of absentee ballot requests. The real question, of course, is whether the high number of absentee ballot requests actually indicates high turnout. It could just be that a higher percentage of people are voting absentee as more Mainers get used to the idea of easy absentee voting.
Recent articles reporting the increase in absentee ballot requests have compared the number to 2007, the last off-year referendum election. This is, of course, the easiest comparison to make, but it’s not really accurate. As we noted above, there was very low turnout in 2007 compared to 2005 and 2003. Going back to earlier, more controversial referendum elections would be a more accurate comparison than 2007. Keep this in mind when you hear tales of “much higher turnout than the last referendum election”. Take any one of the controversial issues off the ballot and turnout would probably still be higher than 2007.
As campaigns master absentee ballot strategy, and people become more accustomed to excuse-free absentee voting, it will likely increase every election – but will it drive turnout? It certainly did not in 2008, when there was a record number of absentee ballots but lower turnout. People will vote absentee a lot in this election, to be sure, but it is not a reasonable predictor of turnout.
Turnout will be important for all of the ballot questions in this election, though, and the crowded ballot will produce some interesting interactions. For example, a higher turnout in southern Maine (especially around Portland) will likely be beneficial to No On 1. Will it hurt the effort to repeal school consolidation? After all, most of the communities in the Portland area have been exempted from consolidation. How those voters might feel on the issue is unpredictable. Or the vote on repealing school consolidation may drive higher turnout in more conservative rural areas that have been most affected by the issue, and help the Yes On One campaign.
The fiscal issues are more difficult to predict. Polling reveals that the partisan divide on the excise tax and the Taxpayers Bill of Rights is not as pronounced as one might expect. Moreover, there is not quite the geographical division on the fiscal issues as there are on school consolidation and the people’s veto.
Turnout is likely to be high in this election, but it is not likely to be driven by absentee voters, and will not necessarily set records. It’s reasonable to expect turnout in the 40% – 50% range based on past off-year referendum elections. For the marriage question, turnout efforts by each campaign will likely determine the final result – and may trickle down to affect other elections.
