Revisiting Gay Marriage vs. Civil Unions

In March, I reported on a Gallup survey that found a surge in public support for gay civil unions, but suggested part of the change was due to the interviewing process itself, not necessarily a real change in opinion. Our question in March found 54% of Americans supporting civil unions, compared with only 40% the previous July. But in the March survey, Gallup asked respondents first about their support for gay marriage and then about civil unions, while the previous July, Gallup had asked only about civil unions. Did this change in question context affect the results? I referred to an experiment testing just this question in a Pew Research Center poll, conducted in October 2003. The Pew results: Question order makes a difference.

While this information helped inform our interpretation of the new data, we could not say for sure how much of the increase in Gallup's new survey was due to question order, and how much was due to a real change in public opinion. In a recent Gallup survey, we re-asked the gay marriage and gay civil union questions, rotating the order -- so that half the sample heard the questions in one order, and half in reverse order.

The results confirm the following:

• There is a "real" increase in public support for gay civil unions over last year. When the civil unions question is asked first, Gallup now finds the public evenly divided in its view -- with 49% in favor and 48% opposed. Last July, Americans opposed gay civil unions by 57% to 40%. Current results show an eight-point increase in support and an eight-point decline in opposition. Support now is about the same as it was in May 2003.

Would you favor or oppose a law that would allow homosexual couples to legally form civil unions, giving them some of the legal rights of married couples?

Trend: Polls in which civil union question asked before gay marriage question or in which no gay marriage questions were asked

• There is also a question-order effect -- with people more likely to express support for gay civil unions after being asked about gay marriage than when they are asked the question about civil unions first. These results are similar to what the Pew Center found.

Support for gay marriage has also increased. When that question is asked first, 42% say "marriages between homosexuals should be recognized by the law as valid, with the same rights as traditional marriages," and 55% say they should not be valid. In the March survey, when the gay marriage question was also asked first, 61% said such marriages should not be valid, and just 33% said they should. Current results show a nine-point increase in support, and a six-point decline in opposition, the highest level of support measured by Gallup.

Do you think marriages between homosexuals should or should not be recognized by the law as valid, with the same rights as traditional marriages?

Trends for Comparison: Polls in which gay marriage question asked before civil union question or in which no civil union questions were asked

• Finally, unlike the findings of the Pew Center, Gallup found that support for gay marriage is lower after being asked a question about civil unions than when asked about gay marriage first. Pew found no difference in results when the gay marriage question was asked first or second.

Interpreting the Experimental Results: The experiments show many people support gay civil unions as a substitute for gay marriage, though support for both has increased in the past several months. The hypothesized reasons for the difference in support, depending on which question is asked first, are:

• When some people are asked about gay marriage first (not knowing they will be asked about civil unions), they express support for the proposal -- because they generally support equality for gays. But when people with these same views are asked about gay marriage after civil unions, they are less likely to support gay marriage than when that question was asked first, because they have already indicated their support for civil unions.
• When people are asked about gay civil unions first, many indicate their opposition even though it's really gay marriage that bothers them -- and they just don't want the interviewer to think that their support for gay civil unions means support for gay marriage. But when respondents are asked about gay marriage first, and are able to indicate their opposition to that idea, then some are more likely to say they support civil unions -- implicitly as an alternative to gay marriage.

Probably the best way to ask the question would be to notify the respondents ahead of time that we are going to ask them about both gay marriage and civil unions. Knowing that they will be given an opportunity to express their opinions about both proposals, the question-order effect would probably disappear. That is something we will have to try the next time.