National longitudinal U.S. public opinion polls
on same-sex marriage (SSM) & civil unions
Webmaster's comments & conclusions
about polls' accuracy, reliability, etc.
Other 2015 polls on SSM are found in the previous essay
How can polls be designed to give biased results?:
Polls on such an emotionally laden topic as same-sex marriage
and other legal recognitions of the relationships of loving, committed same-sex
couples are sensitive to many factors. The results will vary according to:
- The question(s) raised before the question on
same-sex marriage: If they result in a negative response from the
subject, they may carry this mood over to the SSM question(s).
- The precise question asked: The term "homosexual"
is sometimes used as a snarl word by religious and social conservatives. "Gay"
and "lesbian" often invoke a less negative response. "Same-sex marriage"
may have the least emotional response at all. If a pollster is interested in
swaying the results, they may decide to select the questions accordingly.
- Time of day: One method of getting an
accurate survey is to phone in early evening, when persons who work out of the home are more
likely to be at home. All of the polls that we have seen on these topics indicate that support for the LGBT community drops with the respondent's age. Thus, if a pollster wants to sway the results in a negative
direction, all they have to do is to phone subjects during the daytime, when
more retired folks are probably home and employed individuals probably cannot be reached.
- Use of land lines vs. cell phones: Older respondents can be over-sampled by simply phoning people at random using land lines rather than including a blend of cell phones and land lines. This will sway data towards opposition to SSM.
How can a poll of 1,000 people in a country of over 300 million give any meaningful results?:
Webmaster's comments: bias alert:
I personally feel that the three most serious deficiencies in North American public schools are:
- Under funding of inner-city schools: They are the ones that need the best facilities, but often they receive the least funding.
- The lack of courses that teach skeptical analysis: As this essay is being updated, the U.S. Presidential elections are approaching. Massive levels of deceitful misinformation and disinformation are being broadcast by all parties via every media to manipulate the public. Americans are in desperate need of training to separate the bull sPit from the small amount of reliable information provided.
- The lack of courses that teach basic statistics: Every high school graduate needs to have taken a course in elementary statistics so that they can understand what polling results actually mean. A quickie course follows:
The only way to learn what the precise opinion of the entire population is would be to ask every person. A precise national average would be clearly beyond reach financially. So, polling companies always ask a much smaller number of people. The result is an approximation of the true value.
If 1,000 people are chosen at random from the total population of, say, 200,000,000 individuals, then the "95% confidence margin of error" (MOE), if the opinion is more or less evenly split, will be ±3.1 percentage points. If 550 of 1,000 individuals asked say that they approve of same-sex marriage, then one can be 95% confident that the true value if one had asked the entire population would be within ±3.1 percentage points of 55%. 95% of the time, the true value would be between 51.9% and 58.1%. [The confidence level is normally 95%, but 90% or 99% are less often used.]
A survey should always report its MOE; without that value, it is meaningless. Alternately, the MOE can be calculated if the sample size -- the number of people asked -- is known.
A given survey's margin of error will change very slightly for different total populations. For example, the MOE for a population size of 200,000,000 persons is ±3.1 percentage points; for 100,000 it is ±3.08 percentage points; for 10,000 it is ±2.94 percentage points.
The American Research Group, Inc. has a handy margin of error calculator that you can access for free. 1
Current status of public opinion towards same-sex couples:
As of late 2012-AUG, numerous polls show that a majority of American adults now favor equal marriage rights for all loving, committed couples, whether they be of the same sex or opposite sex. This majority is also increasing over time. Support experienced a major surge during 2011 and 2012 which we attribute partly to:
- The extensive discussions during 2012/2011 concerning the repeal of the military's "Don't Ask, Don't Tell" policy that had prevented lesbian, gay and bisexual service members from openly revealing their sexual orientation.
- Many federal courts finding that the Defense of Marriage Act is unconstitutional on multiple grounds.
- The executive order by President Obama that gave same-sex partners visitation rights to each other in essentially all hospitals.
- The successful, high profile battle to legalize SSM in New York state.
- The public declarations of support for SSM by President Obama, vice-president Biden, and other administration leaders during 2012-MAY.
- Decisions by millions within the LGBT community to openly reveal their sexual orientation and gender identity to friends, family and employers. Studies have shown that the most effective way to change a persons beliefs about sexual orientation and gender identity is for someone they know as a friend or relative to come out of the closet.
Meanwhile, by a large margin, older youths and young adults now favor enlarging the definition of
marriage to include same-sex as well as opposite-sex couples. We suspect that
these trends are based on the belief by a growing number of persons under 30 years-of-age that same-gender sexual behavior by gays and lesbians is morally equivalent to opposite-gender sexual behavior by persons with a heterosexual
orientation. They typically view sexual orientation as
beyond the control of the individual, as unchosen, and unchangeable.
Opposition to homosexual behavior generally -- and same-sex marriages and civil
unions in particular -- is still very high among older members of the Republican Party, fundamentalists of all religions, and other evangelical Christians, as well as by their denominations, traditions, and parachurch groups.
One might expect that Roman Catholics would reject SSM
and civil unions in large numbers because of their church's unalterable
opposition to both. However, support of SSM by the Catholics laity is generally higher than the among general
A possible future pushback:
present trends continue, most Americans will no longer view sexual minorities
as a problem group. Instead, they will increasingly view homophobia 2
-- opposition to equal rights for gays, lesbians, bisexuals and SSM -- as the
real social problem. In future decades, the majority may well consider opposition to marriage
equality for all loving, committed couples as immoral, as one form of homophobia. Religious and social conservatives would then be faced with a difficult choice:
- They may choose to continue political action to deny the right of gays and
lesbian to marry or enter civil unions. For fundamentalist Christians and other evangelicals, this would detract from their great commission: to evangelize the world. It would also accelerate the current flight of large numbers of older teens and young adults from conservative faith groups, or
- To quietly reduce their activity against same-sex partnerships
and allow gays and lesbians to attain marriage equality.
The previous instance when marriage was redefined in the U,S, occurred in 1967 when the U.S. Supreme Court legalized interracial marriage across the country. Some 72% of American adults at that time were opposed to marriage by persons of different races. Over the next generation, opposition dropped by about 1 percentage point a year as more liberal older teens and young adults entered the voting pool and many of the less inclusive elderly left it. By 1991, those adults who were still opposed to interracial marriage became a minority. Currently about 13% of marriages are interracial. Many people today would negatively regard a person as as racist if they express opposition to interracial marriage.
We expect that public opinion support for same-sex marriage will continue to increase in the future, much as support for interracial marriage did during the late 20th century. As a result of this shift, we expect that a growing percentage of the population will consider opposition to SSM as homophobic; it will be regard as despicable and immoral as is racism today.
If this happens, then we expect that public opinion polls will become less accurate over time. They will start to over-report support for SSM while under-reporting opposition. This is because many people who are surveyed will be reluctant to admit to the pollster their opposition towards SSM. They would not want to be considered as holding an immoral opinion, even to an anonymous stranger over the telephone. This same effect has been observed for decades when people are polled about attendance at religious services. About 40% of Americans typically say that they attend weekly; only about 20% actually do. Attending church, synagogue, mosque, etc. regularly is held in high esteem in the culture and many adults want to be thought of as "good" people. And so they lie to the pollster, making the survey results less accurate.
How reliable are current SSM surveys?:
Survey results taken prior to plebiscites are rarely good predictors of the voting results:
- One reason is that concern over expressing immoral opinions may be causing opponents to SSM to lie, as described in the previous paragraphs.
- A second reason is that state plebiscites are generally proceeded by a blitz of millions of dollars worth of TV advertisements such as "A Gathering Storm" which was launched by the National Organization for Marriage in 2009. [The National Organization for Marriage is the main nation-wide group opposing marriage for same-sex couples.] Fear is a strong motivator against any type of change. Such ads are quite effective in temporarily swaying public opinion against SSM in advance of a vote.
- A third reason is the stealth nature of some citizen initiatives. The 2012 plebiscite in North Carolina is a superb example. It was heavily promoted by religious and social conservatives as simply a method of banning SSM in the state by amending its constitution. However, it was actually written to ban SSM, civil unions, domestic partnerships, and even minor legislation by which the state government could recognize either same-sex and opposite-sex relationships by unmarried couples. Loving, committed, unmarried couples were given the status of "legal strangers" -- and treated as simple roommates, with no protection for the couple and their children.
A survey taken before the vote showed that about 60% of the public would oppose a constitutional change that would prohibit same-sex marriage, and civil unions, and other forms of recognition of same-sex couples. But the actual vote was about 60% in favor of the amendment which accomplished exactly this. By confusing most of the voters throughout the entire state that the amendment affected only same-sex marriage, they implemented a most impressive public relations accomplishment!
The above three reasons mean that a plebiscite to approve SSM within a state will probably succeed only if there is about a 10% margin in support of SSM as the election day approaches. Perhaps a 20% margin would be needed if it is another stealth measure. Four civil initiatives to approve, prevent, or terminate SSM in Maine, Maryland, Minnesota, and Washington State were scheduled for election day on 2012-NOV-06. None of them are stealth measures. In a remarkable event.
- SSMs were legalized in Maine, Maryland and Washington State by a modest margin,
- The prohibition of SSM in Minnesota was defeated.
This was followed up by a remarkable series of legislative and court initiatives that further advanced marriage equality during 2013. By year end, the total number of states where SSMs are legal increased to 18!
The following information sources were used to prepare and update the above
essay. The hyperlinks are not necessarily still active today.
- "margin of error Calculator," American Research Group, Inc., at: http://www.americanresearchgroup.com/moe.html
- The term "homophobia" has a variety of
definitions. We use the above definition.
- Polling Report.com has a large report on polls involving same-sex marriage and equal rights for the LGBT community at: http://www.pollingreport.com/
Copyright © 2011 to 2013 by Ontario Consultants on Religious Tolerance
Originally written: 2011
Latest major expansion and update: 2013-DEC-28
Author: B.A. Robinson