A slew of polls have surveyed voters’ beliefs about marriage equality since President Obama’s endorsement last week, but the data collection is quickly becoming lazy and the interpretation sloppy.Monday’s CBS/New York Times poll has been roundly criticized for its incredibly small sample size (615) and the odd framing of its questions. Fox News unsurprisingly conducted a poll of its own in the same fashion and eagerly spun the results to accommodate its anti-equality agenda:
A majority of voters don’t support allowing gays and lesbians to marry legally, yet at the same time a majority also opposes a constitutional amendment banning same-sex marriage.
According to a Fox News poll released Wednesday, 37 percent of voters believe gays and lesbians should be allowed to get married legally. While that’s unchanged from 2010, when the question was most recently asked, it’s nearly double the 20 percent who felt that way in March 2004, the first time it was asked.
There is actually nothing in the data that supports this conclusion. What the Fox News article doesn’t mention until its fourth paragraph is that it asked its question the same way the CBS/NYT poll did: forcing a choice between same-sex marriage, legal unions not called marriage, or no legal recognition. The true result of this poll is that 70 percent believe there should be legal recognition for same-sex couples, which was actually 8 points higher than what Monday’s CBS/NYT poll found.
But the problem with both polls is that they never force respondents to choose between same-sex marriage and nothing, creating an incomplete picture of where voters stand. Consider therecent polling from Colorado, which found that 62 percent support civil unions, but that 53 support full marriage equality as well. Forcing respondents to make an either/or choice about marriage and civil unions instead of allowing consideration for both separately creates a distorted view of where voters actually stand.
The Times’ Ross Douthat attempts to spin the interpretation the other way, suggesting that because so many “prefer” civil unions, their support for marriage equality when not provided with an alternative is “reluctant.” And it’s because of that reluctance, he believes, that the results of ballot measures don’t match the polling. This, of course, is a conclusion that can only be drawn from the strange construction of the question in these polls, and it also ignores the reality that many complex factors impact these plebiscites. In North Carolina, the most current example, polling showed that voters were largely uninformed (or misinformed) about the actual impact of Amendment One, and thus did not realize they were voting to ban civil unions and domestic partnerships in addition to marriage — against their wishes. Plus, as Nathaniel Frank points out, polls on social issues are simply “notoriously bad at predicting [voter] behavior.”
Fox News wants to be able to claim it has data opposing the conclusion that a majority of Americans support the freedom to marry, despite consistent national polling over the past two years that shows otherwise. Any poll can be structured and framed to deliver a certain bias to the results, but the true momentum for marriage equality cannot be disregarded so easily.