A lot of headlines have noted that no ballot initiative to dismantle Proposition 8 will be on the ballot this November in California. Some, citing recent favorable polling, have expressed frustration and dismay over that state of affairs. The hope was that Proposition 8 could have been overturned this year, and maybe it could have.
But there are good reasons to be wary of polling numbers and models when they attempt to predict the outcome of a referendum on equal marriage rights: in particular, polling and modelling have not predicted past results, and there is no obvious reason to believe the current polling would be predictive for this November.
Let's look at the data.
The most recent polling on same-sex marriage in California:
In favor of same-sex marriage:
LA Times/USC | 53 | 40 |
PPIC/USC | 50 | 45 |
In 2008, most polls predicted that Proposition 8 would go down to defeat. According to Ballotpedia:
Month | Survey | Favor | Opposed | Undecided |
May 2008 | Field | 40 | 54 | 6 |
July 2008 | Field | 42 | 51 | 7 |
August 2008 | PPIC | 40 | 54 | 6 |
Sept. 2008 | Field | 38 | 55 | 7 |
Sept. 2008 | SurveyUSA | 44 | 49 | 7 |
Sept. 2008 | PPIC | 41 | 55 | 4 |
Oct. 2008 | Internal polling "No on 8" | 47 | 42 | 11 |
Oct 4-5 2008 | CBS News/SurveyUSA | 47 | 42 | 11 |
Oct 15-16 2008 | SurveyUSA | 48 | 45 | 7 |
October 12-19 | PPIC | 44 | 52 | 6 |
October 18-28 | Field | 44 | 49 | 7 |
Nov. 1-2 | SurveyUSA | 47 | 50 | 3 |
The average of all the polling from October onward was 46.2% in favor, 46.7% against. The last poll had Proposition 8 failing by 3%. Yet Proposition 8 passed by a margin of 52.3% - 47.7% (4.6%) and a vote total differential of almost 600,000 votes.
In Maine, Question 1 was approved, rejecting same-sex marriage, by 52.75% - 47.25%,
a 5.5% differential. While the last poll taken indeed showed it being approved 51% - 47%, the average of all the polling from October onward was 46% - 50%.
State | October polling average | Results |
California |
46.2%-46.7% | 52.3% - 47.7% |
Maine | 46%-50% | 52.75% - 47.25% |
Nate Silver, guru of all things poll related, put together a model and a revised model based on past voting patterns on equal marriage rights, which predicted that Mainers would approve of same-sex marriage in 2009. Even he got it wrong.
In order to have confidence in a possible victory at the polls, equal marriage rights will, I'd estimate, need a solid, consistent majority and a differential of 8% (e.g. 53% - 45%) in the polling in order to withstand the potential attrition that these examples illustrate once people enter the voting booth.
That's not going to be easy. Yes, demographics and the public's attitude is changing, but how fast?
This diary, written immediately after the 2008 election, analyzes the demographics behind the vote. Proposition 8 won by some 600,000 votes. The analysis in that diary shows that, when we take into account young people (who are in favor of same-sex marriage by approximately 2:1) becoming voters, and older people (who are opposed to same-sex marriage by approximately 2:1) dying, and the rate of voting participation for each of those age groups, equal marriage rights nets about 100,000 votes per year.
So even in 2012, if we rely solely on demographic changes, we'll still be behind by 200,000 votes. That may be a manageable margin to overcome. We'll have social changes to go along with demographics, an effective campaign of persuasion and get-out-the-vote by such organizations as the Courage Campaign, and perhaps ads that can effectively counteract the inevitable 'Protect the Children' ads that have been used with such devastating results against equality in the past.
On the contrary, it's really hard to believe that a 400,000 vote deficit could be overcome regardless of the organizational effort and the excellence of any ad campaign. Besides two fewer years of changes in societal attitudes, it's not like its just twice as hard to change 400,000 minds as it is to change 200,000: those second 200,000 are presumably less persuadable and harder to reach than the first 200,000.
Does all this suck? Of course. Should people have to wait around for their rights? In a just world, no. But to quote a keen observer from the past, "That's the way it is."
2010 would have been a real gamble. But why not 2010 and then, it if failed, 2012? That's a good question, but I think one plausible answer is that there might have been enough voters who would not be happy with being asked to weigh in on the question three times a row that there would be a really high chance of a backlash.
In any case now its 2012 or bust. And bust is not an option.