Paid TV ads by No on 8 and Yes on 8 were not the only televised and broadly disseminated communications available to voters as they decided how to vote. At the same time that the two campaigns aired their ads, voters were also exposed to non-campaign media—communications not paid for or controlled by No on 8 or Yes on 8 that reached a large audience on the topic of same-sex marriage or a topic related to it.
To make sure that the time line in this report considers and includes all televised or comparably widely disseminated communications that reached voters at the height of the Prop 8 campaign, this section looks at two distinct types of media: earned media and independent advertisement efforts. Although it is beyond the scope of this report to examine these communications comprehensively, our research team has assembled enough data to evaluate the impact of both: earned media coverage boosted Yes on 8, and independent advertisement efforts had no measureable impact.
A victory for consistent, clear, and scary.
“Earned media” includes publicity through news and editorial coverage—in general, media that is not paid advertising. For this report, we examined the portion of earned media that made its way online, comprising all mainstream coverage reproduced online as well as original online content.
Earned media has the potential to amplify the messaging of one or both sides of the campaign, particularly for those voters who are skeptical of messages presented in political advertising. When news coverage repeats the same points made in paid advertising, it can reinforce and legitimize campaign messaging. On the other hand, news coverage, because it is independent, may present voters with alternative messages that complement or undermine the paid advertising of campaigns.
The data presented below show that during the Prop 8 campaign, Yes on 8 started behind but quickly dominated earned media coverage once it had “Princes” on the air.
We examined online earned media by searching the Google News Archive with search terms designed to isolate the discussion of various messages in conjunction with Prop 8, such as “equality,” “schools,” “children,” and several others. We repeated the searches with various combinations of words to ensure each messaging point was covered, and then aggregated the hits for each message during each week of the campaign. The resulting table provides a chronology of the frequency of each message point.
From early October 2008 through the end of the campaign, earned media coverage reinforced Yes on 8’s message in “Princes” and its redefinition of Prop 8. Yes on 8’s success with earned media began when "Princes" first began airing in English on October 8. Once established, its perspective on Prop 8 remained dominant in earned media coverage through Election Day, even in the closing days when No on 8 dominated paid TV advertising.
The graph below illustrates the devolution of the discussion surrounding Prop 8 in media outlets. Although the search results return a few articles that are not germane, the trends surrounding Prop 8 coverage are striking.
In the first four weeks studied—before the "Princes" TV ad began airing—140 articles about Prop 8 contained references to equality. Only 109 contained the word “school” or “schools.”
In the week of October 6 through 12, however, articles mentioning schools overtook articles mentioning discrimination and equality combined. This remained true in each subsequent week. “Schools” appeared nearly twice as often as both combined in the final week.
As noted above, Yes on 8 began the campaign with its point of view not being the dominant one in earned media. In early September 2008, most public discourse reported in earned media surrounding Prop 8 had to do with whether disallowing same-sex marriage was discrimination, and whether a ban on marriage violated the principle that gay and lesbian people deserved full equality.
News outlets catering primarily to the Christian right contained sporadic mentions of the idea that Prop 8 would affect how children were taught in schools. Quotes from religious conservative leaders occasionally found their way into the mainstream media, but discussion of schools was dwarfed by other matters. For example, in the week following September 21, Steven Spielberg’s donation to the No on 8 campaign generated more media coverage than the effects of Prop 8 on schools. In addition, No on 8 received overwhelming editorial support, including endorsements from almost every major newspaper and the vast majority of smaller newspapers in the state.
Yes on 8 began the campaign with two potential handicaps. First, its message relied on its making tangible and relevant the misleading and tangential relationship between Prop 8, defining marriage as between a man and woman in California’s constitution, and Yes on 8’s central message that the decision facing voters was about the danger to children posed by the curriculum taught in public schools.
Second, Yes on 8’s factual arguments surrounding the effect on California school curriculum were thin. The “Princes” ad is a fictionalized account of the Massachusetts case Parker v. Hurley. The Yes on 8 version of the story is partly true: children did read King and King, a book about same-sex marriage, in school, and parents were not allowed to exempt their children. But the Yes on 8 version is mostly false: the inclusion of King and King in the curriculum had nothing to do with the state’s marriage law; its use predated Massachusetts’ legalization of same-sex marriage. A number of states, including California, already have similar age-appropriate references in their curricula noting the existence of LGBT people and encouraging tolerance toward LGBT families, even without same-sex marriage laws. Prop 8 could not have introduced the topic of LGBT families and other nontraditional families into the California school curricula because the topic was already included. There is always the possibility that students will ask questions about any topic, but same-sex marriage is likely to occasionally come up simply because of the currency and wide discussion of the topic, whether or not marriages are taking place.
4. Yes on 8 succeeded despite the facts, because its message resonated emotionally with some voters.
Yes on 8’s message is emotionally alarming, though not based in fact. Its claim is that children are endangered when they learn about LGBT people. This idea is such a long-standing canard that voters pay attention to it much more than is justified given its factual inaccuracy. The earned media reflected Yes on 8’s ability to dominate and dictate the terms of the debate around Prop 8: voters reacted strongly to the argument, voters talked about it, the No campaign eventually was forced to address it, and the media covered it. Fear is a powerful motivator for voters, and fear of any stigmatized group has an even longer history of being used to win elections and to reinforce and justify social ostracism than fear of LGBT people. Without providing a powerful competing emotional narrative, No on 8 was unable to resist Yes on 8’s redefinition of Prop 8 in the minds of voters.
5. No on 8 lost control of its own message in the earned media. The pro-LGBT message that became most reported was one focused on hate. This message probably had little or no power to persuade undecided voters.
The No on 8 paid advertising campaign was less consistent than Yes on 8 in making one clear point. Messages focused on discrimination, equality, and a factual rebuttal of “Princes.” See Finding 5 for a fuller discussion of message clarity and Findings 2 and 3 respectively for a discussion of Yes and No’s most effective ads. One consequence is that No on 8 lost control of its own message in earned media: the pro-LGBT base adopted another message—hate—that perhaps many in the base found more emotionally compelling. The previous graph shows that the word “hate” began competing with both “discrimination” and “equality” as the campaign intensified and public awareness of the closeness of the contest grew. Then, “hate” became the dominant No on 8 message in earned media, overtaking both “discrimination” and “equality” in earned media coverage in the final week.
Unfortunately for the No on 8 campaign, “hate” is unlikely to be a persuasive message to swing voters. Not only does it poll poorly; but it also makes no logical sense to tell a group of voters that if they are ambivalent, then they are, in part, motivated by their anti-gay hate. Since the majority of those voting against us know LGBT people, they are very unlikely to see themselves as haters of LGBT people; and it is probable that very few voters of any stripe self-identify as haters of anyone.
Independent advertisement efforts—those completely separate from any campaign—also have the potential to play an important role in shaping public opinion. For a portion of the time of the Prop 8 campaign, two (and to the best of our knowledge only two) independent entities created and aired very different TV ads. The data show that both probably had little effect on the choices made by voters.
The first of the two to air was “Garden Wedding.” Its educational mission clear, “Garden Wedding” did not mention Prop 8 and did not recommend to voters whether or how to vote on the ballot measure. The ad was produced prior to Prop 8’s existence and was initially aired prior to Prop 8’s existence by the independent 501(c)(3) organization Let California Ring.
The ad depicts a beautiful female bride encountering several obstacles on the way to meet her groom at the altar. As she is preparing for her wedding, the door handle breaks, she catches her heel in the street, loses her veil, is almost sandbagged by a little girl, is tripped by an old woman with a cane, and collapses yards from her future male groom. The ad analogizes the bride’s difficulties in making it to the altar to the LGBT community’s legal barrier to marrying. It asks “What if you couldn’t marry the person you love?”
This sixty-second spot was the most widely broadcast on the topic of marriage in California in the calendar year 2008. The cost of the TV buy for this one ad was not smaller than $4.6 and possibly as large as $7.1 million. The ad was most widely broadcast in August 2008, but it aired as early as February 18. It therefore ran both before and during the time that Yes on 8 and No on 8 ran ads. The English-language version of the ad was backed by a buy of 600 gross rating points (GRPs) in the Los Angeles, Sacramento, and San Diego media markets. The Spanish language version of the ad ran only in the LA media market in a small buy of 196 GRPs.
However, Lake polling also indicates that there was virtually no movement during this time in voters’ views on same-sex marriage. After asking the horse race question to find out how each voter was planning to vote (yes or no on 8), Lake asked in plain language whether each voter wanted to eliminate marriage for gay and lesbian couples or retain it. This follow-up question began, “Just to be clear”; throughout this report, it’s referred to as the Be Clear question. As the chart above indicates, although there was significant movement in the horse race question, there was movement of only one point in the Be Clear question. This one point gain is within the margin of error and therefore can’t be interpreted as significant movement.
The most likely explanation for the movement in the horse race question is therefore wrong-way voting: more people decided they were going to vote No on 8 even though there were not more people supporting same-sex marriage. Voters moved in No on 8’s direction on the horse race question, but did not move on the issue of same-sex marriage. If “Garden Wedding” affected voters, it would logically show up in the Be Clear question since the ad specifically focuses on marriage. “Garden Wedding” is highly unlikely to have affected voters’ views on Prop 8 without affecting their views on same-sex marriage, since the ad did not mention the ballot measure (it could not, as an educational measure). There is nothing about “Garden Wedding” that would have led more voters to vote no if it did not affect their view of the same-sex marriage issue.
Further, there is no evidence that the impact of “Garden Wedding” survived voters’ exposure to the Yes on 8 messaging in “Newsom” and “Princes.” Polling by the Public Policy Institute of California (PPIC) in 2009 showed overall support for same-sex marriage essentially unchanged from 2003. The PPIC results suggest that pro-marriage (and anti-marriage) messaging of all kinds, including “Garden Wedding,” did not educate voters in a lasting way. Considering the totality of the evidence, “Garden Wedding” cannot be considered effective at changing public opinion.
The most commonly cited argument attempting to make the case for “Garden Wedding,” compares the Election Day results of Santa Barbara and Monterey counties as evidence that the ad had an impact. From January through March 2008, Let California Ring used Santa Barbara as a test market for both “Garden Wedding” and a local field campaign. For the same time period, Let California Ring designated Monterey as a control environment where no educational efforts took place. By the end of March, “Garden Wedding” had aired in Santa Barbara County in Southern California where Prop 8 was defeated. Some concluded that Let California Ring’s efforts boosted the pro–marriage equality No on 8 vote in Santa Barbara.
This argument, however, is not persuasive, for two reasons. First, in October and November Yes on 8 spent significantly more to air their ads in Monterey than they spent in Santa Barbara.
Second, the methodologically best practice for selecting a control group is by chance—randomly—not by having the researcher pick it. It is otherwise too easy for a researcher to inadvertently bias the comparison. To make a valid postelection comparison between Santa Barbara and Monterey counties, Let California Ring would have had to have left to chance, eg a coin toss, which county received the educational treatment and which county served as the control. That selection process was not used here. As a result, Santa Barbara is a single data point with no valid control to which to compare it.
Of course, “Garden Wedding” was an educational ad, not a campaign ad. It was produced before the existence of a No on 8 campaign, and aired independently of it. “Garden Wedding” was therefore not designed to affect the campaign.
“Garden Wedding” may instead have a variety of virtues that this report is ill-equipped to evaluate. “Garden Wedding” may have stimulated discussion among voters who already supported LGBT marriage equality, boosted morale in the LGBT base, and made door-to-door or telephone field work easier by raising the profile of the marriage issue. All of these possibilities are worthy of further investigation apart from this report.
At the same time, it’s fair to note that “Garden Wedding” is a remarkably unclear ad to those who are not already thinking about same-sex marriage and are in favor of it. The difficulties facing the bride are not in any obvious way analogous to the barriers preventing gay and lesbian couples from marrying. It should come as no surprise that the ad did not change public opinion even with a massive TV buy.
In no way, however, does this analysis take anything away from the possibility that educational work, including independent educational TV advertising, has the potential to be a worthwhile investment. Future ads simply need to be clearer and more compelling from the point of view of persuadable voters.
Also, although “Garden Wedding” was the largest expenditure of Let California Ring, its other educational work included early funding to staff nascent LGBT people of color and people of faith organizations. These activities are worthy of serious study in their own right. Let California Ring’s commitment to building a broader set of relationships for the LGBT community is praiseworthy, but evaluation of the effectiveness of its efforts from 2006 to the present is beyond the scope of this report.
The only other independent TV ad on marriage to air in California in 2008 was “Home Invasion.” The sixty-second spot was created and aired by the Courage Campaign Issues Committee, a progressive, 501(c)(4) California-based grassroots and netroots organization. “Home Invasion” was released online and also aired on TV on Election Day; it explicitly asked voters to vote No on Prop 8.
In contrast to the wide release of “Garden Wedding,” “Home Invasion” aired in Los Angeles and the total time buy was a little more than $13,000. According to campaign committee expenditure reports, the Courage Campaign spent a total of $15,163 on media urging voters to vote No on 8, and an estimated $1,800 on production costs. Assuming the remaining budget was used to air “Home Invasion” on cable, the ad would have aired only a few times.
According to the corresponding press release by Courage Campaign, “Home Invasion” was intended to air on cable TV for one day as a final get-out-the-vote reminder. The ad itself follows two men who identify themselves as being from the Church of Latter-day Saints (LDS). They force themselves into the home of two women where they proceed to find and destroy the women’s marriage license. The ad ends with the two men gloating and gleefully considering, “what should we ban next?” “Home Invasion” was intended to go hand-in-hand with a petition the Courage Campaign had delivered to the Church of Latter-day Saints the week prior.
A second Courage Campaign ad, “Gender Auditors,” was released online only. Produced for $1,000, the ad depicts government bureaucrats physically examining the genitalia of a heterosexual couple before they can be granted a marriage license. “We don’t need more government in our lives, or in our pants” was the theme narrated during the exam. The ad was described as satirical on the Web site and prompted visitors to share the ad widely and sign a pledge to vote no.
Both ads seem aimed at the pro-LGBT base. There’s no evidence that either was widely seen beyond it. A potentially problematic aspect of both ads is that their rough tongue-in-cheek messages would offend a wider audience and potentially embarrass the LGBT community in the long run.
Even more troublesome is the hostile tone of “Home Invasion.” Intended to hold the LDS church accountable for its role in Prop 8, it fails to evoke sympathy for No on 8. The discourse in earned media stimulated by the ad can be summarized by the top relevant news hit from the Google News Archive that day: “When marriage debate turns ugly, no one wins.”In any event, there’s no evidence that the Courage Campaign ads had an impact on the election result.