Poll poled

The New York Times for December 21 reported on the results of a Times/CBS News poll taken the previous week regarding same-sex marriages. The report said it revealed "strong support" for a constitutional amendment that would define marriage as being between one man and one woman.

There is, it seems to me, a little bit of wag the dog going on here. (The results are available in .pdf format here.) In fact, it's likely that either the results are seriously skewed by what respondents thought was a good answer or the sample was seriously unrepresentative.

Right off, the "strong support" was 55% in favor, which is certainly a majority, but the adjective "strong" is subjective. Whether or not it would be enough to pass such an amendment would depend on just how those people are distributed among the states. A constitutional amendment requires, once it passes both houses of Congress by a two-thirds majority, endorsement by the legislatures of at least three-fourths of the states. Which means, for example, if support for the proposed amendment is very high in the South but moderate or weak elsewhere (averaging out to 55% nationwide), it would likely not be enough to get it through.

It's also true that in a number of questions about same-sex marriage or "civil unions," support was around 40% and even somewhat more. That appears a significant base of support. Banning such marriages altogether is by no means a slam dunk.

At the same time, other questions in the survey raise doubts as to whether people were saying what they actually believed or were influenced by what they thought - sometimes wrongly - that others believed. For example, only 8% of respondents thought other people would accept same-sex marriages in their "church, synagogue, or place of worship." But when asked if they themselves would accept it, 32% said yes. How much did the perception of opposition by others discourage people from acknowledging their own willingness? Perhaps not at all - but how do we know?

Then consider that when asked if "homosexual relations between consenting adults" should be legal, in July the answer was "yes" by 54%-39%. When the same question was asked in December, the answer was "no" by 41%-49%. Such shifts have been attributed to the Massachusetts Supreme Judicial Court decision allowing for same-sex marriages having brought the issue forward in people's minds. But even with that, a 13 percentage point shift is one hell of a jump which should have raised a few eyebrows of doubt.

But there is one area that really makes me wonder how much respondents are influenced by what they think others think. When asked if they voted in the 2000 presidential election, just about two-thirds said yes. Actual voter participation was 51% of registered voters. So even if all the respondents were registered voters, either a significant number aren't telling the truth (but knowing they should vote, say they did) or the sample is seriously skewed.

In addition, when those who said they voted were asked who they voted for, Bush came out on top of Gore by roughly 49.5%-42.5% (the rest going to Buchanan, Nader, and "won't say). Since that is so clearly divergent from the actual results, it again indicates the tendency of people to try to be where they imagine the majority is.

I don't really fault the pollsters here; they tried to ask questions different ways to ferret out actual results and it was their own data that brought the results they reported into question. I do fault the Times for failing to recognize or mention those doubts, preferring to run with the "gay rights is a loser for the Democrats" political meme.

Footnote: On the other hand, the most depressing single answer in the entire survey was to the question "Do you think being homosexual is something people CHOOSE to be, or do you think it is something they cannot change?" People split down the middle, 44%-44% (the rest being "don't know" or "did not answer"), a figure unchanged from a decade earlier.

Despite all that's been learned, despite all that's been discovered (including ample evidence of homosexuality among animals in the wild), roughly half of us still believe that people would freely choose to be the targets of opprobrium, ridicule, hatred, violence, superstition, suspicion, and loss of civil rights. Amazing. And saddening.

Previous posts on this can be found here, here, here, and here