Maine — Question 1 — “Do you want to reject the new law that lets same-sex couples marry and allows individuals and religious groups to refuse to perform these marriages?” .
The Positions: A ‘Yes’ vote overturns same-sex marriage, which the state legislature approved in May. A ‘No’ vote preserves same-sex marriage.
The Polls: As in the other elections this cycle, recent polling shows somewhat contradictory results. A Research 2000 / Daily Kos poll shows Question 1 losing by one point. A Pan Atlantic SMS poll shows the initiative losing by 11 points. But a fresh PPP poll shows it passing by 4 points. The Research 2000 and Pan Atlantic polls showed slight movement toward the ‘No’ side from their previous surveys, while the PPP poll showed movement toward the ‘Yes’ side. The only other pollster to have surveyed the race is Democracy Corps, which had shown Question 1 losing by 9 points among both registered and likely voters, although that poll is now five weeks old.
Analysis: All the polls show a very low number of undecideds, so like most close elections, it’s a question of turnout. And the pollsters have different opinions about what turnout is liable to be. PPP has people under 45 representing about 38 percent of the electorate, whereas Research 2000 has them at 51 percent of the electorate. PPP’s figures are a closer match for Maine’s 2006 electorate, when 36 percent of voters were 45 or under.
On the other hand, PPP shows conservatives outnumbering liberals 36-23, whereas those numbers have been about evenly split in exit polling of Maine’s elections in 2004, 2006 and 2008. Were the liberal-conservative split to match 2006, for example, when Maine’s electorate was 26 percent liberal, 26 percent conservative, and 48 percent moderate, then Question 1 would fail 46-53, according to PPP’s internals.
While an electorate that favorable to liberals might be somewhat unlikely in an off-year election, there is also not a lot of evidence that conservatives have the edge in terms of organization or enthusiasm. On the contrary, the No-on-1 campaign has received contributions from 9 times as many Mainers as the Yes-on-1 side, and Yes-on-1’s messaging has been haphazard, to put it generously. With that said, the gay marriage question is one on which conservatives have typically had an enthusiasm advantage, although that may be changing, with conservatives devoting more of their energies to abortion and fiscal policy.
One last methodological issue worth mentioning may be cellphone-only households, which continue to make up a higher and higher percentage of the survey base and which none of these pollsters, to my knowledge, are including in their surveys. Some previous studies have found a particularly strong split on the gay marriage question based on cellphone usage, with the younger and perhaps more sociable cellphone-only crowd tending to be more supportive of gay marriage.
The Odds: A statistical analysis I conducted last month, which was based on the results from previous gay marriage referenda in other states, gave the Yes on 1 side just an 11 percent chance of prevailing, although the fraction rises to 32 percent after an ad-hoc adjustment for the fact that this is an off-year election. In spite of the PPP poll, I’m not especially persuaded to deviate substantially from those numbers: the polling average still favors the ‘No’ side, albeit narrowly; the ‘No’ side seems to have run the superior campaign, and the cellphone issue may be worth a point or two. The tight polling, certainly, should keep everybody on their toes, and gay marriage could quite easily be overturned. But I’d still put the Yes on 1 side as about a 5-to-2 underdog.