1) It is much safer to say that a majority of the public approves of same-sex marriage
Below are the 2013-2014 polls on same-sex marriage (compiled from PollingReport.com). There are some polls reporting an estimate below 50 percent, indicating that there is still some uncertainty depending on the source of the poll. However, for every one poll that reports less than majority support, there are five polls that report majority support. There is not much evidence that the Supreme Court’s June 2013 decisions produced a major shift in approval. Before the decision, an average of 52 percent supported same-sex marriage. After the decision, 54 percent did.
The national trend using all surveys asking opinions on same-sex marriage also suggests that the majority of Americans now supports same-sex marriage. In fact, in the past year, the change in support has increased more rapidly than ever. (The graph excludes questions that separate out marriage equality, civil unions, or no recognition — but these questions also show that support for same-sex marriage is increasing.)
2) Support for same-sex marriage is accelerating.
Here is some evidence that support for same-sex marriage is truly accelerating. Below I compare how well two different kinds of trends fit the pattern in the polls: a simple linear trend, which assumes that the rate of change is stable, and a polynomial trend, which assumes that the rate of change is increasing.
The accelerated regression line better fits both the data prior to 2004 and the upward trend after 2012. Additional statistical tests confirm that the accelerated trend does indeed out-perform the linear trend.
This has important implications for where public opinion is headed. If the stable linear trend were the right one, then by 2016 just over 56 percent of the public would be expected to support same-sex marriage. However, the accelerated trend predicts that support for same-sex marriage will be about 5 points higher by 2016. It is appropriate to infer that opinions are trending positively and changing exponentially as time goes on.
3) A majority of the shift in attitudes is coming from people’s changing their minds, not more younger, more liberal generations replacing older, more conservative generations.
In a post about a year ago, Nate Silver concluded that about half of the change per year was due to changing minds and half was due to generational replacement. Silver estimated that 1 point of change per year was due to generational replacement. If that is correct, then anything beyond that 1 percent change a year would be attributable to changing opinions. Taking the 2013-2014 results and forecasting opinion in 2016 using the accelerated trend, I estimate that opinion change is greater than 2 percent a year. Thus, a majority of the increase in support appears to derive from changing minds. In short, people who once opposed same-sex marriage are beginning to support it, and this will be the driving force behind any future changes.
Right after the Supreme Court’s decisions in June, Patrick Egan postulated on this blog that the changing views of same sex marriage were more similar to changing views of interracial marriage than of abortion. In the latter case, the Supreme Court’s decision in Roe v. Wade likely hardened preferences, leading to little ebb-and-flow in opinion. In the former case, the Supreme Court’s decision in Loving v. Virginia likely facilitated the upward trend in support for interracial marriage. Given the little evidence of backlash against the recent same-sex marriage cases and the data provided here, trends should continue in the Loving direction.