Finding 4: No on 8’s Biggest Mistake

Yes on 8 Knew What it was Doing

“Three weeks after the Yes on 8 campaign had introduced education as a message, the No on 8 campaign responded with what would be their best ad of the campaign. It featured State Superintendent of Public Instruction Jack O’Connell claiming that Prop 8 had nothing to do with education and that our use of children in our ads …

“The response to our ads from the No on 8 campaign was slow and ineffectual.”

—Frank Schubert, Yes on 8 Campaign Consultant

The only clear error in the above statement is when Schubert calls the No on 8 response as “ineffectual” without qualification. As Finding 3 shows, the “O’Connell” ad, which Schubert recognizes as No on 8’s best ad, had an impact, though only fractionally the impact of Yes on 8’s “Princes”.
The most costly mistake by No on 8 was the delay in rebutting “Princes” and the kids argument. No on 8 was only able to regain ground after it rebutted “Princes” in its TV ads.

·         The message of “Princes,” that kids are in danger, was Yes on 8’s main message.

·         No on 8 delayed sixteen days before rebutting “Princes” in its TV ads. Three principal factors contributed to the delay: 1) No on 8 was unprepared for the kids attack; 2) the campaign was under severe financial pressure and at first may have had only enough money to air one ad; and 3) the one ad that appealed most to the new campaign decision makers, and the one they chose, avoided the issue of kids.

·         The long delay in rebuttal and the lack of preparation to rebut reflects a historic pattern. No on 8—like many pro-LGBT campaigns—chose to avoid the issue of kids rather than confront it.

·         Additionally, above and beyond the issue of kids, No on 8, like No on 1 in Maine and other pro-LGBT campaigns, avoided depiction of LGBT people, the centrality of LGBT people in the issue at hand, and even use of the word “gay” or any other word that would communicate the fundamental truth that LGBT people are the ones most directly affected by the ballot measure. Unfortunately, the strategy of avoidance has not yet worked, and there is good reason to doubt it will work.

“Princes” and the Prop 8 Campaign

The Message of “Princes” Was the Principal Yes on 8 Message, and Voters Found It Memorable.

The message of kids in danger was the center of the Yes on 8 campaign. Yes on 8 depended on its TV ads as its principal tactic. All of its six widely broadcasted ads reinforced the same message.

·         “Princes” was Yes on 8’s most widely broadcast ad. Three of its five other ads focused exclusively on the same message: “Massachusetts” showed a mother and father talking about their young son’s exposure to messages about LGBT people that they found inappropriate; and “Field Trip” closed with the face of a forlorn child after she was exposed to a lesbian wedding that was a school field trip. “Hola” featured Latin American soap opera star Eduardo Verástegui appealing to the audience to support Prop 8 to “protect marriage and children.”

·         The remaining two ads (“Newsom” and “Closer”) included the kids message as one of several points.

Evidence of “Princes” Impact Was Its Growing Memorability Among Voters Throughout October.

The Lake Research polling shows that voter memorability of “Princes” increased dramatically every week starting October 8. At first, only 1% of voters recalled and described the ad without prompting when asked what they had heard from Yes on 8. By October 20, 20% of voters recalled the ad with no prompting.

Chart 25: Percentage of Voters Recalling Yes on 8 "Newsom" Ad Over Time

Chart 26: Percentage of Voters Recalling Yes on 8 "Princes" Ad Over Time

Chart 27: Percentage of Voters Recalling Yes on 8 "Gay marriage taught in schools..." Message Over Time.

By contrast, No on 8’s most memorable ad, “Thorons,” was recalled without prompting by less than half the number of voters who recalled “Princes.” See Finding 3 or Appendix D to see the charts showing “Thorons” ad’s memorability.

No on 8 delayed sixteen days before responding to “Princes” on TV

“Princes” first aired in Spanish on October 6 and in English on October 8. But “O’Connell” did not air until October 22. Sixteen days elapsed between the anti-gay attack on kids and schools and No on 8’s direct rebuttal. By the time it aired, only fourteen days remained before election day.

No on 8 Had Warnings that the “Princes” Message Was Coming

No on 8 had a variety of warnings that the “Princes” message was coming. In July, Yes on 8 submitted language that every voter in the state had received in their ballot pamphlet, arguing that schools would be required to teach children about gay marriage. No on 8 filed suit in California Superior Court on July 29 asking to strike the language. On August 8, the court ruled that Yes on 8 could make its argument if it replaced “would” with “could.” The fact that Yes on 8 made this argument and litigated to keep it was a clear signal that this was a key message of the Yes campaign. From this warning alone, No on 8 had two months to prepare to answer the argument.

Also, knowing the history of anti-gay ballot measures provided clear warning that the opposition would raise danger to kids as an issue. Danger to children has been the signature message in many of the most competitive anti-gay ballot measure campaigns. In 1977, Anita Bryant’s Save the Children campaign to repeal Miami-Dade County’s Human Rights Ordinance used the same message. In 1988, so did the highly controversial child-in-danger ad that aired only once in the campaign to repeal then-Governor Goldschmidt’s executive order banning discrimination in employment among Oregon state employees. Direct appeals to anti-gay prejudice, including (but not limited to) the issue of kids, were used throughout the thirty-five state and local anti-gay ballot measures in Oregon over the following sixteen years. In 1998, the principal anti-gay ad used in Hawaii in the very first anti-gay marriage ballot measure rang this theme. It foreshadowed fully the attack in Prop 8: it featured a young boy about to shed a tear after being exposed to a book just like the one featured in “Princes”. To view the Hawaii ad and its strong resemblance to “Princes,” go to the Media section.

Why Did No on 8 Delay Its Response to “Princes?”

No on 8 took so long to respond to “Princes” because the campaign decision makers at the time, installed only a week before “Princes” went on the air, did not choose to directly respond to the attack.

Although several who were integrally involved were kind enough to meet with me and discuss their mind-set at the time, it may be beyond the scope of this report to fully know why. The data and discussion that follow are my best approximation of understanding, the best I can do without having been there.

In the very beginning—before “Princes” went on the air, and in its first twenty-four hours on the air—it is possible that the new decision makers did not appreciate the power of the attack (which of course was easier to know for certain after the fact). Opinion within the campaign was divided at this point about whether the “Princes” ad would work. On October 6, No on 8 pulled off the air the two TV ads up at the time, “Thorons” and “Conversation.” The campaign almost went off TV entirely, airing only a placeholder ad, “Lies.” Almost no one had confidence in this ad, and it was aired in a tiny buy so almost no voters saw it. See Appendix E for the details of the buy backing “Lies” and all other No on 8 and Yes on 8 ads.

Within forty-eight hours, the daily tracking polls showed serious damage to No on 8. At this point, failing to respond represented a judgment call that there was a better alternative than responding directly. The decision makers chose to move ahead with a new ad, “Unfair,” that more generically tried to establish doubt about Prop 8. The new decision makers thought of “Unfair” as a pivot ad meant to turn the campaign away from talking about why gay marriage was a good idea to why Prop 8 was a bad idea—to pivot to a more traditional No campaign. “Unfair” had a lot of black on the screen, with the word “No” always present prominently in red, like a stop sign, and included ominous music as it listed a variety of endorsements of a no vote. Such spots are commonly used to gain no votes on any kind of traditional California ballot measure. “Unfair” ignored the thrust of “Princes” and the opposition’s anti-gay message. The new decision makers paid $3.6 million to air “Unfair,” making it the most heavily funded broadcast ad of the No on 8 campaign. To see all the No on 8 ads, and to see all the information on money spent to broadcast each of the No on 8 and Yes on 8 ads, go to Appendix E.

“Unfair” went on the air on October 14, eight days after “Princes” went up. Part of the delay could easily have been due to uncertainty about which media firm would create the ad. Rather than rely on the media firm that created “Thorons,” “Conversation,” and “Lies,” the new decision makers chose new media consultants to create all ads going forward.

Lack of available money may also have contributed to the delay in getting “Unfair” on the air. Although the campaign bank balance showed substantial money in the bank—$1.7 million as of October 6, rising to $3.4 million on October 9 and to $4.7 million on October 14—an unknown portion may have represented financial commitments already made or encumbered donations, money given to the campaign earmarked for specific purposes. The exact date when enough money was definitely available to put a second ad on the air is not known. The sense that money was scarce and not quickly coming in almost certainly affected the mind-set of the new campaign team, perhaps even past the time that financial weakness was acute.

What is known, however, is that “Unfair” never had an impact. From October 14, its first day of broadcast, “Unfair” never moved the needle in No on 8’s direction in the daily tracking polls. The Lake tracking polling, in both the standard horse race and the “be clear” question, shows the terrible stability during this period: Yes on 8 was above 50% each day and No lagged far behind. By October 16, the decision to persist with “Unfair” is hard to understand. My best guess is that the decision may have reflected the new leadership’s view that the Lake Research polling was too pessimistic, or the fear that confronting the issue of kids and schools could not succeed.

Throughout this time period—from October 6 or a little earlier through October 16—the new decision makers took their position very aware of the alternative point of view. Pressure was consistently exerted from almost every quarter to respond directly to “Princes.” Some previously in a decision-making role felt that ignoring “Princes” was untenable; they predicted serious deterioration until No on 8 rebutted and demonstrated the inaccuracy of “Princes.” They were correct. As Chart XI shows, Yes on 8’s decisive advantage after “Princes” held steady throughout October. At the same time, Chart XXII shows that No on 8 slowly hemorrhaged support and fell further behind. The aftermath of “Princes,” particularly the time before “O’Connell” went on the air, with large numbers of voters casting absentee ballots every day, was almost certainly when No on 8 fell irreparably behind.

In fairness to those who resisted responding to “Princes,” the new campaign manager and his team (still in formation) were facing this crisis after only a week on the job.

Further, it was not obvious how to directly respond. The LGBT community has historically avoided responding directly to the issue of kids, in part out of the belief that no response will defuse the issue, and in part out of a wish not to have to face this unfair, untrue defamation.

Additionally, it was not obvious what the substance of the rebuttal should be. From May through mid-September, No on 8 invested substantial funds in message research. But after finding little in the polling or focus groups to suggest that kids would be an effective attack point, and believing that Yes on 8 might consider other lines of attack, No on 8 did not seriously prepare. No response ad was ready to air when “Princes” hit. Nor did the initial group of No on 8 consultants prepare for other attacks. This was a significant oversight; creating ads responding to attacks takes time, and having a variety of responses in the can is or ought to be standard practice.

The fairest conclusion may be that in August and September, No on 8 did not manage its consultants aggressively enough so that the urgency and concern felt by the campaign committee was translated into practical preparations for anti-gay attack. In October, when the attack came, the new decision makers—on the job for only a week—inherited this inadequate preparation.

In addition, it seems likely that some of the No on 8 decision makers did not want to become reactive to Yes on 8. Some may have felt that responding to the charges would reframe the campaign on terms favorable to Yes on 8, and make the entire campaign about kids and schools. Rather than respond to “Princes,” this line of thinking went, No on 8 should put forcefully forward its own message and frame the issue its own way.

The weakness in this thinking is twofold.

First, Yes on 8 was attacking the trustworthiness of No on 8 and the character and decency of LGBT people. Implicit in the “Princes” attack was the idea that gays are so untrustworthy, so indecent, so depraved, that they would put their interests above those of children. Voters would have no reason to believe a No on 8 message on any other topic if they bought into the idea that the pro-LGBT side was untrustworthy. It’s an axiom in both candidate and ballot measure politics that when one side attacks the honesty or character of the other, the campaign under attack must rebut. It’s all but impossible to prevail in an election if voters don’t trust you. It’s fatal for any campaign to fail to defend its trustworthiness. Voters won’t believe what they hear from a source they don’t trust.

The decision by No on 8 from October 6 through October 19 not to rebut “Princes” (on October 19, the campaign manager gave the green light to the “O’Connell” ad) was therefore a decision not to defend LGBT people as trustworthy. But voters were never going to get this message if not from us. The anti-LGBT character attack works because it reinforces a pernicious, untrue set of assumptions about gay people embedded in the larger culture; voters are exposed to these ideas without realizing it as they grow up. These ideas are not necessarily foregrounded for voters. But when the anti-LGBT side raises them, these ideas resonate because they are consistent with attitudes, fears, and assumptions that voters absorbed long ago. The preexisting prejudice allows the Yes on 8 ads and the shriller Yes on 1 ads in Maine to advance a calibrated dehumanization of LGBT people. Research in the Maine campaign supports the conclusion that receptivity to the anti-gay kids argument is for many Republicans and a slice of Democrats strongly associated with voters’ general feelings about LGBT people and the degree of prejudice they have towards LGBT people.

Granted, it’s difficult for any campaign to make the case that it is trustworthy. Yet allies of the LGBT community have at times done it directly and clearly. Campaigns in Oregon provide some of the best examples. Dawn Laguens’s TV spots “Teacher” and “Dawn” from the 2000 No on 9 campaign represent two different approaches used in the same campaign. Both spots are viewable in Appendix E. One spot, “Dawn,” tells a true story. The other—the campaign’s principal spot, “Teacher”—addresses head-on the ugliest implication of the other side’s “Princes”-style attack. It opens this way: [teacher speaking directly to the camera] “I’ve been teaching in Oregon schools for 20 years, I’ve seen many things, but I’ve never seen anyone promote homosexuality—it doesn’t happen.” Having a teacher vouch for the reasonableness of what happens in schools was also part of the fourth ad in Maine’s No on 1 campaign. This ad was the first to respond to Yes on 1’s equivalent of “Newsom.”

Second, even if the No on 8 decision makers had high hopes for “Unfair,” it is not clear why they chose to bet everything on it. Even if they believed “Unfair” a more promising strategic choice than a direct rebuttal to “Princes,” they could have prepared a backup plan in case “Unfair” tanked. Yet they did not. Part of the delay getting “O’Connell” on the air to rebut “Princes” was because the campaign was waiting for “Unfair” to have an impact on voters—time that could have been put to concurrent use developing a rebuttal to “Princes.” Similarly, part of the delay in creating the “O’Connell” spot or a similar ad with Delaine Easton could have been reduced by putting the pieces in place at the same time that “Unfair” was being produced.

Finally, practical reasons—lack of money, not enough time—explain only partially the delay in responding to “Princes.” It is possible that No on 8 had enough money to rebut “Princes” and also air “Unfair.” As noted above, on October 6, the day “Princes” first aired, No on 8 had $1.7 million in the bank. Cash flow analysis shows that the bank balance climbed to $3.4 million on October 9, and then to $4.7 million on October 14. The balance never fell below $1.6 million until October 22, after No on 8 had put “O’Connell” on the air. For the full day-by-day bank balance and cash flow analysis based on income and expenditures during the campaign, see Appendix G. The reason that this information is only suggestive, however, and not dispositive, is that the raw bank balance numbers tell only part of the story. For example, between October 14 and 22, No on 8 received gifts of $1 million from the California Teachers Association and $2.8 million from Equality California. If those large gifts or others were earmarked for specific purposes, then the No on 8 bank balance could overstate the freedom of action of the campaign decision makers.

At some point prior to October 19, however, No on 8 was no longer broke; it had the funds to go on the air with an additional ad; and it had evidence from a series of daily tracking polls that showed that “Unfair,” the ad No on 8 was running, was having no impact. Any two of these three bits of information could have persuaded the campaign to alter course and replace “Unfair” with an ad rebutting “Princes.”

Time to conceive and create a response ad also appears not to have been an insuperable issue. The concept of the “O’Connell” ad was put forward very quickly, probably on or around October 7, with a similar spokesperson proposed (Delaine Easton, O’Connell’s predecessor as state superintendent) and similar content. If the campaign had prioritized putting a direct rebuttal ad on the air, it could have done so perhaps as early as October 8, more likely by October 9 or 10, in place of airing “Unfair.”

On balance, weighing the limited information available, it seems most likely that the new decision makers made a tactical choice to delay rebutting “Princes”; they chose to go with “Unfair” over rebuttal. That choice was affected by circumstances but not driven only by them; a different choice was possible. The new decision-making team significantly misjudged and underestimated the power of “Princes” and the kids argument, missing the point that what mattered even more than the specific content of the ad was the ways in which the ad and others like it have the capacity to scratch open the anti-gay sentiment and uneasiness that lives just below the surface for a significant number of voters, even in California.

Future campaign teams with no LGBT ballot measure experience are at high risk to repeat this mistake because their polling is likely to understate the persistence of anti-gay prejudice. See Appendix L for a discussion of the larger dynamics, including this one, that lead to history repeating itself in these campaigns.

The No on 8 experience with the power of “Princes” is why this report recommends in the strongest possible terms that the LGBT community begin now to gain the insight necessary to rebut the anti-gay libel about kids and schools. Until the pro-LGBT side reduces voters’ reactivity to ads like “Princes,” we will be hard-pressed to win these elections. It is unwise for the LGBT community to return to the ballot until we are prepared to vitiate this attack. For a fuller discussion of this topic, see Recommendations 2, 3 and 4 in this report.

A Historical Pattern: Avoidance of the “Kids” Issue in Pro-LGBT Campaigns

Avoidance of the kids issue was the first response by the No on 8 campaign

As described in detail above, No on 8 largely avoided the issue of kids for a little over two weeks after the attack ads began. No on 8’s TV advertising relied largely on one ad, “Unfair,” for 87% of its buy for two weeks after “Princes” was on the air. “Unfair” epitomized avoidance of the issue. A full description of the ad as well as an embedded video are in Appendix E. The evidence for the ad’s lack of impact is discussed in detail in Finding 4 (below).

In earned media (not paid media), No on 8 did address the issue of kids quickly, having former Superintendent of Schools Delaine Easton and current Superintendent Jack O’Connell deny the charges. But both No on 8 and Yes on 8 ran campaigns recognizing paid TV as the centerpiece of their communications; to relegate rebuttal to earned media was a futile attempt by No on 8 to keep the issue on the sidelines. Further, it’s an axiom that for campaigns to respond to attack effectively, they have to respond in the same medium in which they were attacked; otherwise, they miss the audience they need to reach.

It’s also worth noting that Yes on 8’s message dominated the earned media.  No on 8 messages about discrimination and equality received less coverage over time than the Yes on 8 message about schools and kids.  For a more detailed analysis of the earned media coverage and the variety of problems encountered by No on 8, see Appendix F, “Non-campaign Media”.

Avoidance of the kids issue occurred again in Maine’s No on 1 campaign

No on 8, however, was not the first or last pro-LGBT same-sex marriage campaign to avoid the issue of kids. In Maine, the No on 1 campaign’s avoidance was just as complete from the point of view of the voters, though perhaps not from the point of view of the campaign leadership. The latter may have felt that their TV ads were in fact responsive to the anti-gay attack, or they may have felt that No on 1 was responding sufficiently in its non-TV campaign communications.

If so, however, their hopes were misplaced. The No on 1 campaign devoted most of its money to TV ads, as did Yes on 1. Voters may have received isolated non-TV communications one or two times from either or both sides, but they were seeing each side’s TV ads ten to twenty times. Non-TV communications could not possibly compensate for or adequately respond to the TV-based attack.

And Yes on 1 did attack. In the closing weeks of the campaign, voters were repeatedly exposed to the explicitly anti-gay “kids are in danger” message in the Yes on 1 TV ads. The only communications voters received equally or more frequently from No on 1 were TV ads that (with only one exception) did not use the word “gay” and across the board did not directly and clearly rebut the anti-gay argument. No on 1’s ads did address an issue involving kids, but not the issue that Yes on 1 was raising so provocatively. No on 1’s ads talked about the need to protect gay kids and kids with gay parents. These were not the kids about which Yes on 1 was sounding the alarm. The two sides were talking about different kids. No on 1 chose to talk about different kids than the ones that moveable voters were being agitated about.

This report is obviously focused on California, and I have not done the same research or had access to the same data about Maine. So although I am able to note the terrible problem with No on 1’s campaign communications—de facto avoidance—I do not know why the campaign made that choice or whether all campaign decision makers realized that that was the choice their campaign was making. But simply looking at the content of the No on 1 ads suggests to me that at least some of those involved in campaign communications were hoping that referencing a different kids issue in the No on 1 ads would cause voter confusion. When Yes on 1 raised its misleading charges that kids were in danger, the first No on 1 response ad (it’s fourth ad of the campaign) played a tiny snippet of the Yes ad and then went on, “But outsiders are trying to harm our kids in schools,” as though this was the harm issue now elevated in voters’ consciousness.

At the point in the campaign when No on 1 aired spot #4, it was well ahead of Yes on 1 in fundraising and was broadcasting ads significantly more frequently. No on 1 may have hoped to outshout Yes on 1 and in that way take control of the kids issue. But Yes on 1 had enough money to buy 500 gross rating points per week, enough for its message to penetrate. Once the provocative Yes on 1 message penetrated, voters were not going to ignore it since it was questioning the safety of their own kids. Voters therefore weren’t confused by No on 1’s pretense that it was addressing the kids argument that made them anxious.

For No on 1 was talking about two disjointed groups of kids that were irrelevant to the targeted voters; the kids most on the voters’ minds were their own. The anti-gay Yes on 1 ads were talking about the (presumed to be heterosexual) kids of heterosexual voters. Yes on 1 argued that homosexuality was being “pushed” on kids (their word, used multiple times in multiple ads). Yes on 1’s message was that heterosexual kids were potential victims, their normality and heterosexuality in danger.

This same point is made by Third Way in its 2010 report Moving the Middle on Marriage: Lessons from Maine and Washington. Their polling found that voters “were less worried about kids in gay families than they were about their own kids.” Third Way also concluded that “the issue is actually broader than schools—it’s about kids more generally.” Its polling found that “40% of [voters in] the middle thought kids would be more likely to experiment with homosexuality if marriage were upheld, and 58% said they were concerned about that issue.”

Third Way made a still broader argument as well: that “[voters in] the middle [see] marriage as an ideal as opposed to a legal construct, and they have yet to be persuaded that gay couples fit into this ideal.” This is consistent with the idea that anti-gay stigma leads some voters to view gay couples as more different from than similar to heterosexual couples. The kids issue may therefore be only one component of the problem. Yet until we answer the kids attack, we will not know if other ideas propping up anti-gay stigma similarly distance voters from accepting same-sex marriage.

Readers can view all the Maine ads from both sides online in Appendix R, where they can also find additional materials on the Question 1 campaign in Maine, including No on 1 campaign literature that, like the TV ads, avoids use of the word “gay” and avoids any clear discussion of the issue of same-sex marriage.

Avoidance of the kids issue is a long-standing tendency, predating both the No on 8 and No on 1 campaigns

In fairness to both the California and Maine decision makers—both of which made strong choices in other parts of their campaigns—avoidance is a longstanding tendency in pro-LGBT campaigns. It was 2002 before the first pro-LGBT ballot measure campaign used the word “gay” in a TV ad (“SAVE Dade,” in its successful effort to stop the repeal of the Miami-Dade County Human Rights Ordinance). It was 2004 before the first pro-LGBT ballot measure campaign put an openly LGBT person on screen in a way that made the person recognizable as LGBT and allowed them to talk (Oregon’s No on 36 ads, “Leanne” and “Heart”). It was 2005 before a pro-LGBT ballot measure campaign showed an openly LGBT couple talking in a TV ad (in Texas, in a series of ads created by the National Gay and Lesbian Task Force); and the buy was very small because research had found the ads to have little impact on African-American voters. With a few exceptions, it was only in the last ten years that our community has begun to try directness in ballot measure campaigns. All of the ads mentioned here can be viewed online on the report website.

It is unknown whether being direct and honest will win marriage campaigns for the LGBT community. We haven’t yet tried this strategy in a full-throated way in a competitive state. The closest was Maine’s No on 1 campaign in 2009. Some No on 1 TV ads depicted LGBT couples and individuals but never used the word “gay” in their ads. Other No on 1 ads had straight family members talk about us: one used the word “gay”; another called us “special.” The Maine campaign deserves credit for taking us one step closer to clarity than any previous marriage campaign. But the sentiments the ads express—including embarrassment, self-pity, and unwillingness to speak up on our own behalf—do not really convey how we feel about ourselves.

Larger dynamics drive pro-LGBT campaigns toward avoidance

Many different LGBT and allied leaders and consultants in different states and in different years have not only chosen avoidance, euphemism, and metaphor over clarity, but have also run campaigns that are very similar in a range of ways. The patterns suggest that the problem is not that one set of leaders or one campaign is particularly dim. Instead, larger dynamics lead each new pro-LGBT campaign—often for reasons that make sense at the time—to repeat past mistakes. Three of these dynamics—misimpressions sometimes created by polling, the gut sense of many consultants that images of gay and lesbian couples will evoke a negative reaction from voters, and campaign structures that give consultants de facto decision-making power—are discussed at greater length in Appendix L.

Is avoidance the best choice?

Of course, all of us who look back after the fact have to concede that there is no guarantee that other untried choices would have fared better. It is possible that avoidance was the best choice in California in 2008 and Maine in 2009; that the pro-LGBT campaigns were not necessarily repeating past mistakes but were instead, each time, using the best arguments available.

Lessons learned in modern candidate campaigns, however, suggest otherwise. Standard practice today is to respond forcefully and immediately to attacks on a candidate’s character or else suffer the fate of the swift-boated John Kerry. Avoidance in the face of attack is widely viewed as ineffective when the opposition is attempting to make the case that a candidate is untrustworthy. The point is just as powerful when the LGBT community is the equivalent of the candidate, its trustworthiness and basic decency impugned by the anti-gay campaign.

Further, the history of anti-LGBT measures on matters other than marriage suggests that avoidance is not the best strategy. Each year, from 1974 through 2000, the LGBT community lost most anti-LGBT ballot measures; most repealed local laws that protected LGBT people from discrimination in employment, housing, and public accommodations even when polling suggested broad voter support for the laws. It is beyond the scope of this report to offer a detailed survey of all of these campaigns. But a limited review of the campaign communications throughout those years shows two overwhelming tendencies:

·         Whenever the anti-gay side had an organized campaign, it made sure that voters knew the issue was about gay people

·         Whenever the pro-gay side had an organized campaign, it ran a campaign avoiding mention of gay people. Pro-LGBT campaigns spoke instead about human rights or protecting the Constitution or saving an unnamed group of people from discrimination—or compared the anti-gay side to well-known bad people and bad behavior (eg Cincinnati 1993, where the pro-LGBT signage used images of Hitler, a Klansman, and Joe McCarthy; Oregon 1988, where the pro-LGBT campaign called the other side a witch hunt).


As recently as 1998–2000, the pro-LGBT side lost the majority of these votes. We won five and lost seven on topics other than marriage (we were five to twelve counting the five losses on marriage during those years).

In 2001, however, the LGBT community had its first winning year, defeating for the first time a majority of the state and local anti-gay measures on the ballot. The pro-LGBT campaigns had changed in a variety of important ways; among them was a willingness to experiment with more direct messaging. The LGBT community’s second winning year was 2002: pro-LGBT victories in Miami-Dade County, Florida; and in Tacoma, Washington, featured the first ever pro-LGBT TV ads that used the word “gay.” In 2003, for the first time since the advent of anti-gay ballot measures, there was both no anti-LGBT ballot measure anywhere around the country, and there was a pro-LGBT ballot measure (in Cleveland Heights, Ohio) that we won. Overall, our community’s record for 2001–2003 was eleven wins and two losses on non-marriage measures. (We also lost one vote on marriage during these years.)

From 2004–2009, our community had seven wins and two losses on non-marriage measures (while going eight to thirty including the marriage measures). Movement toward directness has not been steady; but on issues other than marriage avoidance is no longer the automatic LGBT strategy, and losing these elections is no longer the usual outcome. Many factors contribute to the better election results. It is impossible to separate the decline in social prejudice against LGBT people from the effect of better-run campaigns; or to separate the variety of ways in which we now run better campaigns, to distinguish our improved attention to field or mail from the value of a more direct message. Nevertheless, the movement away from avoidance should make us curious about its value.

In the face of this experience, it is striking that the pro-LGBT marriage campaigns embrace avoidance in whole or in part even though it has only once and temporarily won for us (Arizona, in 2006, where we prevailed on a vote on marriage and domestic partnership; it was reversed by the loss in Arizona in 2008 solely on marriage).

One certain danger of the avoidance strategy

My point about the power of truth is narrower. History teaches us that when our campaigns avoid the most basic, incontestable truth—that the measure on the ballot is about LGBT people—we look like we’re trying to hide the truth; and we are. When we make that mistake, we give credence to the anti-gay argument that we’re untrustworthy. We prove their point for them. We allow the anti-LGBT side to be the one presenting the basic truth that the vote at hand involves LGBT people. The anti-LGBT side presents that truth in an ugly and misleading way, but we are on the defensive and in a poor position to credibly call them out for their gross inaccuracies. We are particularly hard-pressed to contest their depiction of gay people when we are unwilling to depict or even mention them.

This does not mean that accurate depiction of LGBT people and couples by itself will turn voters around. In California, for example, immediately after extensive earned media coverage in June 2008 of gay and lesbian weddings and couples, public support for same-sex marriage temporarily declined, and then rebounded about a month later.

But the portrayal of LGBT people in TV ads of our own creation could be much kinder, gentler, and potentially more effective for us than earned media coverage. Earned media both strives for balance and is drawn to caricature. Sarah Palin is one obvious example of someone who comes across less flatteringly in earned media than she does in a more controlled setting.

In too many campaigns, we have suffered from an inaccurate, grotesque, vicious, distorted, and frequently uncontested, depiction of LGBT people and couples. This is surely an impediment to our winning elections. Yet that is what we guarantee when we leave mention of LGBT people to the anti-gay side.

On balance, considering the available evidence, here’s my current belief and recommendation: As hard as it will be to win these votes on marriage, we don’t need to make it harder by repeating past mistakes. Let’s learn from history and abandon avoidance; and let’s test a variety of alternatives to avoidance so we can learn what works best.  See Appendix L for a detailed discussion of why history repeats itself.