Analyzing the same-sex marriage ban vote

Republicans at the Minnesota Capitol today announced attempts to put before voters a constitutional amendment banning same-sex marriage. It comes as no surprise since that’s exactly what they said they’d do not long after last November’s election, which wiped out the last vestige of DFL influence in the Legislature: the Senate.

In early interviews in January, the leaders, however, said “there’d be time” for the issue at some point in the session, but the focus would be on passing a state budget.

We’re a month from the end of the session, and there’s no sign of a budget being passed. Time is running out to get the same-sex marriage amendment on the 2012 ballot. Today, five Republicans filed SF1308.

Opponents of the amendment have annually suggested that it’s a way to get Republican conservatives to turn out for a re-election bid. In Indiana last month, state Sen. Dennis Kruse said here are 30 states that now have a marriage amendment to their constitution, “and they all passed in all 30 states by an average of 68 percent of the electorate,” Kruse said.

Does an amendment increase voter turnout?

In Maine last November, 53-percent of the voters voted to repeal the state’s same-sex marriage law. Turnout was heavy, but both sides said a heavy turnout favored those in favor of allowing same sex marriage. “It means we succeeded in reaching younger people and others who don’t always vote,” said Mark Sullivan, spokesman for No on 1/Protect Maine Equality, the coalition seeking to uphold the law.

In a paper released last month (open in Word), researchers in Wisconsin pointed out that regardless of what pre-election polls said about opposition to same-sex marriage, the percentage of people actually voting for a ban on same-sex marriage was markedly higher.

In North Dakota in 2004, for example, polls showed 52% of those surveyed were against same-sex marriage. On Election Day, 68% voted against it. In Wisconsin in 2006, 59% voted against same-sex marriage, compared to 53% who voiced opposition in a late pre-election poll.

Possibly more significant, however, is this finding: The question could make Republican allies of a traditional DFL voter: The African American.

“The higher the proportion of African Americans in the county, the higher the vote for banning same sex marriage. While this fails to confirm our hypothesis, this result seems easily explained as an indication that, on the issue of same-sex marriage, the LGBT community’s argument that marriage is a civil right, and not a moral question, has failed to win favor among black citizens.”

But, the researchers said, there are three “striking” conclusions from their study (emphasis mine):

First, the noticeable impact that time has. There is clearly an increase in support for gay marriage over time and it will be particularly interesting to see if this issue follows a similar trajectory to gays serving in the military. That is a highly controversial issue that is met with a compromise that provides only partial equality (Don’t Ask Don’t Tell in the military case, civil unions in the case of gay marriage) followed a generation later by full recognition of equality. Second, the very big impact that turnout has and the possible implications for results. To take one concrete example, Colorado in 2006 voted 55-45 to ban same sex marriage. Turnout among registered voters was 19 percent higher in the 2008 presidential election. If we assume a 19 percent increase in turnout, our regression equation predicts almost a 4 point drop in the yes vote. Such a drop would transform this race from a very comfortable ten point win for the referendum to a nail biter at 51-49. The third impact we believe is worthy of more extensive consideration is the differences in referendum language. This is because this issue has broader implications beyond just this question, but because of the significance for democratic theory and discussions of referendum as a democratic tool. That is a discussion for another paper.