Anti-gay marriage ballot measures don't come cheap. Proposition 8 now holds the record with its whopping $83 million price tag, a figure representing the combined total both sides spent in California where voters in 2008 narrowly approved a constitutional amendment, rolling back same-sex marriage rights.
On a smaller scale in Maine, the warring sides last fall spent a total of $9.5 million on Question 1, with voters approving a ballot initiative that repealed same-sex civil marriage legislation signed into law by the governor.
What then has all this money bought? Apparently nothing, according to a recent study conducted by a New York University ( NYU ) assistant professor. "The study indicates that when all of [ the campaigning ] is over, the result is a wash," NYU's Patrick Egan told reporters at press conference at San Francisco City Hall June 15.
In other words, during the heat of a campaign, hard-hitting television ads had little effect in changing voters' views, one way or the other.
Egan, a political scientist, undertook the study, an analysis of pre-election polling data, spanning the decade from 1988 to 2008. The study includes data from states that held ballot measures on marriage equality or domestic partnershipsa total of 167 surveys on 32 different ballot measures. In 30 states, voters approved anti gay-marriage initiatives, mostly constitutional amendments, while two other ballot measures focused on domestic partnerships laws.
Unlike LGBT Americans in 30 other states, Illinois' gay community has not faced any ballot measure fight over partnership recognition. While the Illinois General Assembly enacted legislation in 1996 restricting marriage to the union of one man and one woman, lawmakers could overturn marriage inequality by a simple majority. The state has no constitutional ban on same-sex marriage. In fact, amending the Illinois Constitution is not all that easy to do, either through the legislature or by citizens' initiated petition. ( See the sidebar for a discussion of lessons applicable from the study for Illinois. )
The San Francisco-based non-profit Evelyn & Walter Haas Jr. Fund commissioned the study. The full report, "Findings from a Decade of Polling on Ballot Measures Regarding the Legal Status of Same-Sex Couples," is available at www.haasjr.org/sites/default/files/Marriage percent20Polling.pdf.
The study revealed, among other key findings, that pre-election polls consistently under estimate voter support for bans the "Yes" voteon the legal recognition of same-sex relationships. At the same time, "No" vote estimates prove to be fairly accurate indicators of how marriage equality backers end up voting, according to the report.
"The share of voters … saying they will vote to ban same sex marriage [ runs ] typically seven percentage points [ below ] the actual vote on Election Day," the study states.
Furthermore, the report found, polling surveys do not become any more accurate as Election Day approaches. In fact, "Support for a ban on same-sex marriage projected from final polling typically falls three percentage points short of actual results," the study states. Accordingly, polls done in the final weeks and days beforehand apparently understate the full share of the vote against legal recognition for same-sex relationships.
During the hour-long press availability, Egan told reporterssome of whom participated through a telephone conference call the impetus for the study resulted from ten years of ballot measures, conventional wisdom concerning the accuracy of polling surveys, and theories about why the polling data are inaccurate.
What accounts for the gap between polling surveys and election outcomes? Two phenomena may come into play. Some voters voice support for same-sex marriage rights and then don't vote that way. Or nearly all the undecided voters support marriage equality bans.
Interestingly, Egan finds no evidence for the "Bradley Effect," a theoretical explanation, which holds that voters conceal their biased intentions, telling pollsters they intend to vote one way, but in fact vote another. The theory's name derives from the 1982 California gubernatorial race where exit polls showed Los Angles Mayor Tom Bradley had won. But Bradley, an African American, lost in a close contest. Some political observers attribute racial bias for his defeat.
In the case of marriage equality, however, "There is very little evidence to support suggestions are lying to pollsters when asked about support for bans," Egan said, referring to the theory of social desirability bias, which may give rise to a politically correct pro-gay response.
Why is that? For one thing, professor Egan's analysis demonstrates remarkable consistency in polling outcomes across an entire decade. The study accounts for surveys from states with a larger LGBT population and, therefore, may well be more pro-gay. Egan's analysis includes polling data from the later part of the decade, when generally speaking, nationwide social attitudes tended to be more gay-tolerant. Finally, the study includes surveys relying on both automated and human interviewers.
Beyond that, Egan said, further investigation is necessary for a fuller explanation of the discrepancies between polling surveys and election outcomes, adding, "preferably research that looks at individuals"and conducted by " [ marriage equality ] advocates."
But could there have been a "massive shift in voter intention [ in California or Maine ] during the very last stage of the campaign?"
"It's possible, but not very likely," Egan said.
In yet another key finding, the study reveals no evidence suggesting that voters are confused aboutor misunderstandthe meaning of a "Yes" or "No" vote.
Still, Egan offered one explanation for the gap between polling number and elections outcomes. "The most likely explanation," he said, "is the problem with the way polls screen likely voters." He added, "Screens that filter out unlikely voters may inaccurately screen out more yes voters than no voters."
In spite of the study's finding that overall money spent during the course of a ballot-measure campaign seem to be a wash, marriage-equality advocates welcomed Egan's findings.
Kate Kendell, executive director of the San Francisco-based National Center for Lesbian Rights ( NCLR ) , offered an assessment. "If we are going to be successful in the future on other bans, the conversation with voters needs to happen now," she said, adding, informing voters, "who are our friends, family members, neighbors, and coworkers," about "our lives, our families, our children, our hopes and dreams of who we are as people and how we fit into the culture."
To that end, Kendell explained, funding will be required for LGBT Californians to "have the courage to engage in those conversations" and "to provide financial resources and organizations and an infrastructure to reach out to voters."
Is there any evidence to suggest that public educations efforts, conducted far in advance of ballot-measure campaigns, really work?
Yes, according to Geoff Kors, executive director of Equality California, who like Kendell, commented on the study's findings during the press conference. Kors pointed to an early 2008 "intensive media campaign, including earned media, paid media, and field work in Santa Barbara" he said, the only county in Southern California where "No" on 8 won. In getting to voters early on, Kors explained, "Before the campaign started, we saw a huge swing there."
The Santa Barbara example, Kors said, shows not only "that voters can be moved," but also that "we have to move them through public education campaigns before" ballot-measure battle lines are sharply drawn and television ads "harden voters." He added, "Once people are for equality, we don't see slippage."
Public education efforts may also prove effective when gay marriage detractors bank their anti-gay message on attack ads playing up fears of children being taught homosexuality in public schools.
"We have to debunk people's fears" early on, said NCLR's Kendell, referring to gay children already attending public schools, and with kids whose parents are gay. When the other side goes viral, misrepresenting what children are taught in the schools, she said, "We need to inoculate" voters so that "folks in our column are already sensitized" to our existence, reassuring them "there is nothing to fear."