- - Wednesday, June 18, 2014

It’s become an article of faith in media reporting that there has been a massive shift in public opinion in favor of same-sex ‘marriage.’ Reporters point to a series of surveys that purport to show growing support for this idea. And it’s certainly what those who wish to redefine marriage want people to believe. However, it’s wrong to suggest that the American people have changed their minds, and there is very strong evidence to show that they haven’t. The definition of marriage remains one of the most hotly contested issues in the country.

Let me point to four polls since Election Day 2012 that look at these issues.

On Election Day 2012, NOM commissioned a survey of randomly selected people who had actually cast ballots in the 2012 elections. We asked them this basic question: “Do you agree or disagree that marriage is between one man and one woman.” Sixty percent (60%) of respondents said they agreed, while only thirty-four percent (34%) disagreed.

The liberal website Politico sponsored a survey in May of this year of likely voters in competitive US House and Senate races. They asked if these voters “support or oppose allowing same-sex couples to marry” and found that 52% of those voters opposed allowing same-sex couples to marry. Among those who felt strongly about the issue, opponents of same sex-marriage enjoyed an 11 point lead, 33% to 22%.

A survey last month in Michigan found a four point drop in support for same-sex marriage in that state over the past year, along with a five point increase in opposition over that same period. This equates to a shift of nearly 10% in voter support for same-sex marriage. The issue of same-sex marriage in that key state stood at 47% in favor and 46% opposed.

Finally, a very interesting study on attitudes regarding same-sex marriage was published by Rice University in June 2013. Rice looked at the responses of 1,294 Americans surveyed in 2006 on same-sex marriage, and then re-surveyed those same people in 2012. This is the only survey I am aware of that has actually tracked people over time. And what the Rice survey found was that overall attitudes about same-sex marriage were essentially unchanged between 2006 and 2012. Moreover, of those who did change their views, more of them moved to oppose same-sex marriage than moved to support it. The researchers wrote, “What is surprising in light of other polls and the dominant media reports that Americans are moving in droves from defining marriage as one man and one woman to an expanded definition is the movement of people in the other direction as well, a fact missed by surveys that do not follow the same people over time.”

It may not surprise you that the Rice University study received very little coverage in the mainstream media. Neither does, it seems, any survey showing that traditional marriage enjoys strong support ever get much media exposure. Why might that be?

One answer to the question might be found in a different type of study, one conducted by researchers at the Pew Center’s Project for Excellence in Journalism. Pew looked at nearly 500 news stories on same-sex marriage published over a two-month period in 2013 and found that those stories with more statements supporting same-sex ‘marriage’ outweighed those with more statements opposing it by a margin of roughly 5-to-1.

Aside from media bias, what other factors might account for seemingly widely differing polling results?

I’ve been managing issue campaigns for the better part of thirty years, and have been involved in hundreds of surveys designed to understand how voters feel about an issue. I have seen the good, the bad and the ugly of polling instruments. There is a lot of the “bad” and “ugly” when it comes to surveys on same-sex marriage.

One factor that can influence polling results is something researchers call “priming.” Priming occurs when a preceding question, or a series of preceding questions, essentially “tees up” an answer to a subsequent question. For example, one of the best-known polling organizations, Gallup, “primes” its question on same-sex marriage by first asking whether respondents “think gay or lesbian relations between consenting adults should or should not be legal.” It’s such an odd question, because gay relationships have been legal in every state for decades. Its purpose, one can assume, is to predispose the answer to the next question asking, “do you think marriages between same-sex couples should or should not be recognized by the law as valid, with the same rights as traditional marriages?” Having just told Gallup that gays relationships should be legal, some poll respondents will now be loathe to tell them that same-sex couples should not be able to marry and obtain “rights.” This may explain why Gallup consistently shows high support for same-sex marriage in its surveys.

In a 2013 article at National Review Online, sociologist Mark Regnerus examined Gallup’s practice of priming its questions on same-sex marriage, and found that the company did not always do this. Interestingly, when it varied its practice — priming on some surveys and not others — support for same-sex marriage varied. When Gallup did not prime, support for gay marriage totaled, on average, 6 to 7 percentage points less than when it did. Regnerus goes on to report that Gallup now primes on every survey related to same-sex marriage.

A major factor that impacts results on an issue like gay marriage is called “social desirability bias.” Many survey respondents are willing to alter their answers to give what they perceive to be the socially desirable answer. That’s why, for example, when asked if they voted in a particular election, a significantly higher percentage of survey respondents will say “yes” than those who actually voted. In the case of same-sex marriage, the politically correct answer has become to say you favor it. This may help explain why polls in campaigns where traditional marriage measures were on the ballot have consistently underestimated support for traditional marriage by 6-7 points compared to the actual results of the elections.

Another factor that can impact results is question wording. For example, many surveys including those for media outlets like CBS and the New York Times ask whether same-sex marriages should be “legal or not legal.” That wording could easily bias results since the concept of making something illegal conjures up notions of punishment and consequences.

An additional factor in polling is the psychology of response preferences. People have a natural tendency to prefer to be “for” something rather than “against” something. In virtually every media poll I have seen, the positive response is to be supportive of same-sex marriage, while the negative response is associated with traditional marriage. This is compounded by including words such as “allow,” “rights” and “recognize” in the question language. The Rice University survey is the only one that associated a positive response with traditional marriage. NOM’s survey also associated a positive “agree” with support for traditional marriage.

Story Continues →