Use the time between campaigns—act right now!—to identify messages that help voters (a) support same-sex marriage and (b) resist ads like “Princes” and its phony argument that kids are in danger.
Given that we need to move a substantial number of voters—probably hundreds of thousands—to change their minds and vote with us in support of same-sex marriage, there is every reason to try out persuasion efforts now. Every persuasion project, if it measures its ability to get results, can increase our understanding of what moves voters. Well-designed experiments can give us insight into which forms of communication and which messages persuade some of those who voted against us to reconsider. Just as useful, the experiments can reveal which messages don’t have persuasive power among the voters whose minds we need to change.
The first experiment in persuasion began only two months after Prop 8 was defeated. Starting in January 2009, the LA Gay and Lesbian Center’s Vote For Equality (VFE) project began canvassing voters one-on-one, door-to-door, in LA County precincts where Yes on 8 received 50–65% of the vote. The project continues today. VFE volunteers knock on the doors of all voters, but they keep conversation brief with those who already support same-sex marriage. When volunteers find undecided and unsupportive voters, however, they take much more time to engage in dialogue. The VFE volunteers ask open-ended questions about same-sex marriage to identify and understand voters’ main concerns. They spend much more time with voters than a typical canvass—five, ten, or twenty minutes per voter rather than one, two or three minutes. They use much of this time to ask voters about their perspective on marriage; their actual, lived experience with marriage; and their actual, lived experiences with LGBT people. The volunteers then try one of several persuasion messages on the VFE script, choosing the one that seems most responsive to the concerns of the individual voter with whom they’re speaking.
The results: in their first year, VFE volunteers had over 6,000 conversations with voters. Of the voters who began the conversations on their doorstep undecided or unsupportive, 26% started to move by the end. To more definitively measure its effectiveness, however, VFE has begun calling back subsets of voters who canvassers had spoken with to assess where the voter now stands on marriage and how memorable the first conversation was to them. Telephone follow-up a month or more later that is carefully structured not to elicit a pro-LGBT response has found to date that half of the 26% did in fact move and remain moved. Thus, the bottom line is that of the undecided or unsupportive voters engaged by VFE volunteers at the door, 13% moved in the pro-LGBT direction in an enduring and documentable way.
To see videos of some of the conversations between canvassers and voters go to the Conversations with Voters section. They are posted on the report web site with permission from VFE.
In the interests of full disclosure, I have worked closely with VFE since December 2008, and its activities and approach reflect in part my thinking and recommendations. That said, I can only take very limited credit for the program’s great success. Remarkable leadership of the LA Gay and Lesbian Center’s staff and volunteers has made the substantial and sustained canvassing achievements possible. This report highlights VFE because it is the most promising post-Prop-8 “skunkworks” I know (see below for a fuller discussion of the term). VFE’s work exemplifies how we can add to the traditional ways that ballot measure campaigns go about gathering information about messages. Simultaneously, VFE is one of the most impressive incubators of new talent and new leaders I’ve seen anywhere in the LGBT community in my thirty years of community organizing.
Inspired by VFE, multiple other organizations began to canvass voters in 2009 as well. Equality California began by using the VFE model in locations throughout the state and for a year worked in close cooperation with VFE in Los Angeles. As of this writing, Equality California has adopted parts of the VFE model and has altered others, creating its own model for one-on-one, door-to-door persuasion conversations. The Courage Campaign and Marriage Equality USA also have other types of voter canvassing operations running in various parts of the state, focused not as much on voter persuasion as on identifying supporters and solidifying support. It is beyond the scope of this report to examine in detail each of these projects and others that may have promise, except to say hurrah! California is a big enough place, and the learning we need to do is remarkable enough that having more groups rather than fewer experimenting with messages and message delivery is encouraging.
Although the discussion above focuses on canvassing, there are many types of experiments well worth our attention that do not involve canvassing. All forms of one-on-one dialogue strike me as promising, including having LGBT people and allies talk with people they know in-person, on the phone, at events, and on-line. The strength of dialogue is that it both teaches us more about what others are thinking and also allows us to gauge whether what we say has an impact.
Even one-way communications may have value. The key is to plan in advance how to measure whether we’re getting a result. Specifically, we can’t assume that one-way communications are a) received, b) digested, c) understood, d) believed, and e) effective. Measurement of results is essential if we’re to evaluate whether the effort on our part is worth continuing, expanding, or abandoning in favor of an approach that gets better results.
We benefit by having multiple experiment occur simultaneously.
It is smart for the LGBT and allied communities to initiate experiments that both think outside the box and yield qualitative and quantitative data. The concept of “skunkworks” is worth introducing to LGBT community: widely used in business, engineering, and technical fields, skunkworks is a team unhampered by bureaucracy and past assumptions and engaged in research and development. A skunkworks team aims to accelerate the innovation process. Its loose structure and openness to new and creative thinking can be powerful if combined with a willingness to rigorously test and measure its new ideas, celebrate the ones that work, and discard the ones that don’t. LGBT and allied skunkworks teams have the potential to accelerate acquisition of insight into what it takes to get some of those who voted against us to reconsider their views on same-sex marriage, and to keep us from losing voters to the kids argument and other anti-gay propaganda.
If an experiment isn’t getting results, change it or stop doing it. Then try something else.
This may seem like common sense, but many projects continue even when they don’t produce either qualitative or quantitative results. There’s no shame in trying something, having it fall short, evaluating it, and ending it. That’s far better than institutionalizing it and devoting some of our limited resources to it.
Iterative learning is the process where we try experiments, measure their results, expand upon the ones that are working or show promise, and stop doing the ones that aren’t working. This approach was exemplified in the second rollout of the online presence of the No on 8 campaign. It will serve us well in every part of the campaign, particularly in the ones where we have a great deal to learn.
The single most useful way to evaluate any campaign or experiment is to examine the collective choices made by the team leading it.
This report offers little assessment of the performance of the individuals involved in the Prop 8 campaign. Instead, the report examines the operations of the No on 8 campaign team. The report describes the key campaign decisions as the product of multiple individuals operating as a team.
In my experience, this approach more accurately examines what matters. Campaigns are not individual achievement events. They are collective action. Tony Kushner, in his afterword to Angels in America, argues that when we analyze historical events, we err if the analysis focuses primarily on individuals, even the most exceptional and singular. He calls this tendency a mistaken belief in “the myth of the individual.” Kushner’s point is that all human beings understand reality and instigate change only when they interact with others. The smallest meaningful unit of analysis is therefore at least two people, not one. Assigning credit or blame to an individual is an oversimplification of history that misses the ways in which we depend on each other.
Therefore, throughout these recommendations and particularly in recommending experimentation and skunkworks, I am offering the idea that our progress in public life depends on our building and evaluating the performance of teams, not of individuals. Individuals make a difference, of course. But in my experience, they make a difference by bringing others together. They make a difference in an election when they contribute to a high-functioning team.
The canvass experiments give us the option of exploring the kids argument and learning how to rebut it effectively.
It is particularly important that the VFE team has chosen not only to continue its initial work, but has also begun an experiment to better understand the specific source of voters’ alarm when exposed to the “Princes” ad. As of this writing, in the current VFE canvass, each volunteer brings up the topic of kids to see if we will be able to hold onto both our existing and new supporters once they recall or are exposed to the opposition’s predictable message. By late 2010, when VFE has had hundreds more of these conversations, it will better understand the way that voters see the kids argument, which voters it most affects, how they interpret the “Princes” ad, why the ad affects them, and whether we are able to counter it.
Since polling does a poor job gauging voters’ receptivity to the anti-gay argument about kids, the kind of work being pioneered by VFE is essential. Research methods different from polling may be able to overcome some of its limitations and more fully reveal voters’ thinking to us. Otherwise, there’s a high risk that we could enter a future election still in the dark about why and how some voters are so troubled by an anti-gay campaign based on a falsehood.
Since qualitative data is not typically collected by a door-to-door canvass, VFE is experimenting with several data collection methods. VFE teaches its canvassers to record as many details as possible on conversation tracking forms immediately after the conversation is over. Then it brings all canvassers back together immediately after their canvass shift is completed to debrief all staff and volunteers. Often VFE has multiple debriefs: first with everyone present, then another with the volunteer leadership, then another among the paid staff. In the week after the canvass, key staff and volunteers read and track the information that all canvassers recorded on the conversation tracking forms. Later that week, staff and volunteers watch video of the door-to-door conversations. The last is possible because at every canvass, VFE’s team of videographers accompanies five to ten canvassers to film their conversations with voters.
VFE’s work is singular, but similar work by Equality California on other aspects of persuasion messaging is also encouraging. Both projects are demanding but teachable. Any community organization could try out comparable canvasses if they are willing to recruit a large team of volunteers.
The canvass experiments accelerate our readiness to return to the ballot when they spur team-building and leadership development.
Above and beyond the value of the research, the canvasses train hundreds of people to listen to, understand, and attempt to persuade voters. Doing this work together builds individual competence and also teaches the power, the fun, the effectiveness, and the sustainability of doing this work as a team.
The conversations at the door may also begin to prepare some of the volunteers to talk to people they know in their own lives who may have voted Yes on 8 about why marriage matters to LGBT people. The expanded cadre of skilled and confident volunteers may prove essential as any future campaign seeks to expand the universe of persuadable, partially persuaded, and fully persuaded voters. All of these are valuable, particularly if it turns out that persuasion is a process over time for most voters, rather than an immediate and irrevocable one-contact transformation.
An additional benefit of VFE’s work in particular is that it promotes smart, strategic critical thinking among the next generation of LGBT and allied leaders. Thinking is not reserved for consultants, traditional experts, and established leaders. Valuable as the thinking of those people surely is, VFE’s model helps a much larger group of people act as stakeholders in this fight. VFE goes to great lengths to incorporate the input and ideas of volunteers from all walks of life, whether they are straight allies, LGBT, new to the fight for LGBT equality, or veteran LGBT activists. On an ongoing basis, VFE creates meetings, forums, and trainings that aim to stimulate volunteers’ ability to engage in dialogue and do rigorous critical thinking as a team. VFE will therefore discover whether investing the time in leadership development and community education increases the number of people who are not only episodically involved but also deeply engaged. My hunch is that VFE is on the right track and that this investment will pay huge dividends over time, not only preparing people to lead a strong field effort to repeal Prop 8, but also in creating a farm team of uncommonly capable, tested people who may serve in a wide variety of leadership capacities in any future campaign.
Finally, all of the canvass experiments are an opportunity to change our community expectations about what it means to participate in a campaign. We will be stronger if participants are people who both think and do. We need more of our best thinkers also engaged in doing, not just in meeting; otherwise it is too easy for them to misjudge the challenge before us. Each experiment will therefore maximize its impact if it strives to recruit and mobilize large numbers of volunteers who are participants in the nuts-and-bolts work essential to campaigns. This ability is distinctly different from the ability to turn out attendees at rallies and meetings. Meeting and rally attendees play a valuable role, but they have much less impact on whether we win or lose an election than those who participate in a sustained way in both the thinking and the doing of the non-glamorous hard work of fundraising, voter contact, volunteer recruitment, leadership development, and community leader outreach.
As we learn more about how to persuade, and how to counter anti-gay prejudice, we will want to come up with experiments that allow us to test the potential for voter persuasion in every medium. That’s the only way we’ll impact the very large number of voters we need to win.
Ideally, a variety of experiments will help us learn which communications media are capable of persuasion. Different media include but are not limited to face-to-face communications, phone, mail, radio; TV, texting, instant messaging, Skype, online ads, and online networking.
It’s important that we learn whether persuasion is possible in all of these media, or only in a subset of these media. If the latter, the next campaign should prioritize developing our capacity to utilize the media that facilitate persuasion, because we can’t win the next election without getting a large number of voters to reconsider their view on the issue.