Interview With Dave Fleischer: Prop 8 Report Author

Dave Fleischer is the author of the 509-page Prop 8 Report which was released publicly on Tuesday. I have known Dave for years and he agreed to this on the record interview with MadProfessah.com.

MadProfessah: What are the main results or ideas you want people to gain from this report?

Dave Fleischer:
* The No on 8 campaign made a smart decision to invest in research. As a result, for the first time ever, in any campaign, the LGBT community has daily tracking polling that measures increases and decreases in our support as voters were being exposed to the anti-gay opposition campaign. This is far superior to episodic polls taken when voters are considering the question out of context and have not recently been exposed to the vile but effective opposition campaign. The No on 8 tracking polling by Lake allows us to correlate voter movement to significant changes in the political environment including but not limited to strategic decisions made by the competing campaigns. That's why this report is the first report ever to seriously evaluate why we struggle to compete in these campaigns. It provides evidence that allows us to consider and evaluate the wide range of competing hypotheses offered by a wide variety of observers. Up until now, we just had the hypotheses, not the data. Now, we have significant data. With the data, we are able to reject some hypotheses, confirm others, and notice areas where we need more information. We are on our way to learning more about the electoral reality in which we are operating.

* Based on the data, we know why we lost. We lost because in the final six weeks, when TV ads from both campaigns saturated the airwaves, almost 700,000 votes switched sides and decided to oppose same-sex marriage. It seems fair to me to conclude that Yes on 8 outcampaigned us in the final six weeks. We went from even to losing in the final six weeks. And it's fair to give Yes on 8 credit for what they accomplished, because we lost the most ground by far among the exact group of voters they targeted: parents.

* No on 8 regained some of the ground we lost, but only when we directly rebutted the opposition fear-mongering. "O'Connell" was No on 8's belated rebuttal of the pernicious Yes on 8 ads that exploited and stimulated anti-gay prejudice by making people fear that kids were in danger. The effectiveness of the "O'Connell" ad, while limited, suggests that we can successfully rebut the longstanding lies, at least among some voters. And the fact that it improved the situation compared to how we were doing when we were avoiding the issue makes it clear that avoidance serves us poorly. When the opposition attacks the character of LGBT people, we have to rebut immediately, directly, and clearly.

* As we prepare to go back to the ballot, we have to commit ourselves to gaining insight into how to rebut those attacks effectively. We still have so much to learn; after all, does "O'Connell" represent the best we can do? We simply don't yet know. We need to try out, in real world circumstances that simulate the campaign, what rebuttals help us most with most of the voters who are susceptible to the opposition fear-mongering. Luckily, we have a great immediate opportunity to enlarge our learning, when we canvass. From my point of view, when we are canvassing, this leads to a practical imperative: we should talk about kids to as many voters as we can; test different messages to see what works; and consider the canvass an investment in qualitative research, the largest, longest focus group in social science history. My concern is that if we don't do this -- if we don't talk about kids now, when the stakes are low -- we will be tempted once again to avoid the issue in an campaign, when the stakes are so high. Then history will repeat itself, to our detriment.

MP: How long did the report take to research and write and who paid for it?

DF: 18 months. Paid for by the LGBT Mentoring Project. No money solicited from nor did any come from CA individuals or organizations or the No on 8 campaign on anyone evaluated in the report.

MP: When do you think we should return to the ballot to overturn Proposition 8?

DF: We should return to the ballot when we have a decent chance to win. So we have homework to do. Should only choose the year when we have finished the homework. That could be 2012, or a different year. To me, what's relevant is: have we done our homework?

MP: One of the key findings is in wrong-way voting. Apparently there were 6% of voters who voted No who opposed marriage equality as opposed to 4% of Yes voters who supported marriage equality. Thus there's a net gain of 2 percentage points for the No side. Can you explain how confident you are of this result?

DF: Very confident. The Lake polling data and our methodology to interpret it are all laid in out detail in the report in a special appendix. [Appendix K]

We acknowledge all of the limitations of the data. But here's what we calculate:

* approx 1.525 million people were wrong-way voters

* of those, 875,000 voted No and opposed same-sex marriage

* and 650,000 voted Yes though they favored same-sex marriage

* the difference, 225,000 voters, made the margin closer by approx. 450,000 votes

FYI, David Binder's polling during the No on 8 campaign, commissioned by the No on 8 campaign, corroborated Lake. In the one poll where he asked questions you could use to gauge wrong-way voting, conducted Sept. 2-4, he found that

* 13% of the voters who found same-sex marriage acceptable said they'd vote Yes, and

* 19% of the voters who found same-sex marriage unacceptable said they'd vote No.

This is a less reliable measure of wrong-way voting than because a) it was further in advance of election day, and some voters self-correct; and b) Binder's question gave three options to respondents, so the middle group is sizeable and there's no way to tell if any of them were likely to wrong-way vote. But Binder's finding here is consistent with the idea that our side was the likely net beneficiary of wrong-way voting.

And of course Binder's May 2009 attempt to gauge wrong-way voting is very unlikely to be of any value. Polls are best at detecting wrong-way voting as or before people vote, not afterwards, especially not six months afterwards, when accurate recollection of their confusion is unlikely.

You will also enjoy looking at Lewis & Gossett, cited in the Prop 8 Report. Their excellent paper analyzing the Field and PPIC data uses regression analysis to compare the plausibility of four competing hypotheses to explain the failure of the two polls to gauge what was happening. They make a compelling case for wrong-way voting as the most probable explanation.

MP: What do you say to people that your source of daily tracking poll data by Celinda Lake ends 5 days before the election and the election could have been decided in that time frame?

DF: The Lake data is the best data we have on Prop 8, and the best data set our community has ever had on any of these ballot measures. That said, it is imperfect in a variety of ways, including the one you mention. There could have been significant changes in voter opinion in those final days. The Lake data would of necessity not reveal that. However, to the extent that the Binder tracking polling data in the final days tell us anything, they do not suggest massive voter movement or anything other than trivial voter movement. See the charts in the report with Binder data and Lake data side by side and you'll see what I mean. Binder tracked until election day (but he didn't start until much much later than Lake, when most of the big movement had long occurred).

Thanks, Dave!