By their own account, many of these citizens might have been drawn to the polls if they had personal connections to the electoral process—for example, if their friends and family members voted, or the candidates were coethnics, or they were personally familiar with the candidates. To motivate them, candidates and parties—and even voter mobilization groups—need to focus on rebuilding trust and repairing the relationship between these citizens and the system from which they feel so disconnected. But who are the trusted messengers who will do this work? We return to this question momentarily.
WHAT MAKES (NON) VOTERS (NOT) VOTE?
To examine these populations more systematically, Latino Decisions built on the findings from the six focus groups by conducting a survey of 1,045 Latinos eligible to vote but not engaged in the electorate. The total sample consisted of 443 respondents who were not registered to vote and another 602 respondents who were registered but not voting.
To evaluate respondents in terms of their propensity to vote, we created an index to account for attitudes about government and politics, civic engagement, information about the election, and prior history of participation. In combination, these factors told us whether people had many different attributes that primed them for more or less political engagement. Based on their responses to these items, we categorized respondents into one of three tiers: low, moderate, or high propensity to vote.
Responses to questions about the following subjects were included in this index:
1.How closely are you following the election?
2.Do you know the names of the candidates running for president?
3.Have you heard about or are you familiar with: Mi Familia Vota, Ya Es Hora, NALEO, and NCLR?
4.What is your partisan identification?
5.Would you register if you received a registration card in the mail?
6.“When watching or listening to news, I feel frustrated or angry.” (agree-disagree)
7.“I will try to register and vote in this election.” (agree-disagree)
8.Reasons for not voting in the 2008 election.
9.Reasons for not registering to vote.
We coded the responses and then used them to create an index that ranged from 0 to 23, with higher values indicating a greater likelihood of participating in politics. Using these scores, we broke all the respondents into three groups.11
FIGURE 4.5Propensity of Latinos to Vote by Age, Sex, Language Ability, and California Residency
Source: Latino Decisions survey, 2013.
The next two figures illustrate these Latinos’ propensity to vote by a series of demographic traits. The moderate-propensity group was the largest for all age, gender, and language ability groups (Figure 4.5). Consistent with all other Americans, the oldest cohort of Latino voters had the largest share of high-probability voters, at 29%. By contrast, young Latinos—those ages eighteen to twenty-four—had the lowest share, only 17%, of high-propensity voters (and the next youngest group didn’t do much better). Figure 4.5 clearly illustrates the age resource disadvantage faced by Latinos. Latinos in California appeared to be somewhat less available for mobilization than Latinos nationwide, while there seem to have been similar distributions between men and women and between English and Spanish speakers. This last finding is something of a surprise, since Spanish speakers have greater difficulty accessing political information and are almost certainly foreign-born.
FIGURE 4.6Propensity of Latinos to Vote, by Registration Status and Demographic Traits
Source: Latino Decisions survey, 2013.
Figure 4.6 shows the nativity, education, income, national origin, and voter registration status of low-, moderate-, and high-propensity voters. Obviously, greater rates of high- and moderate-propensity to vote would be found among registered voters than among those who were no longer registered. Though these nonparticipants had fallen out of the electorate in recent elections, they had paid the costs of registration at some point in their political lives and were more interested in politics than those who had never registered. As we’d expect, more Latinos with higher income and education were in the higher-propensity group. Lower educational attainment is associated with a lower propensity to vote.
Somewhat surprisingly, foreign-born Latinos were more likely to be in the moderate- and high-propensity group relative to their US-born counterparts. This trend echoes the recent political science research, which finds, first, that naturalized voters in recent years have been more active than naturalized voters in the past, and second, that more time in the United States is associated with declines in Latino efficacy.
TABLE 4.1Percentage of Low-, Moderate-, and High-Propensity Latino Voters to Describe Messages as “Really Convincing”
PERSUASIVE MESSAGES TO GET LATINOS ONTO THE ROLLS
In our study, we ran a messaging experiment to see which attempt at persuading respondents to vote would be found to be most persuasive. Recall that we were studying the nonparticipating, so our goal was to identify the messages that had the greatest impact.
Table 4.1 reports the messages we tested and the corresponding share of respondents who described each of them as “really convincing.” What messages appear to work? The results differ considerably across the three propensity groups we created with our index (low, moderate, high). The difference between the high-propensity group and the others is striking. The majority of the high-propensity group found half of the prompts to be very convincing. Appeals to democracy (messages 1 and 3) and fighting anti-Latino policy (messages 2 and 4) were very powerful motivators for this group. The moderate-propensity group showed some similar trends: their top two messages were also about democracy (message 1) and Latino-targeting policies (message 4).
Curiously, the last message did not resonate much with any segment, more or less validating the expressions of alienation and low efficacy from Rafael, Anita, and others in our focus groups. Few Latino respondents who avoid electoral participation really believe that voting can get the attention of elected officials. With alienation high and efficacy low, that message fails to persuade.
MESSENGERS MATTER
We have spent much of the last several years trying to educate clients and leaders alike on the importance of the messenger. As advertising agencies have long understood, what you are trying to sell determines, in large measure, who is going to be good at selling it: the messenger has to be credible and influential and must be someone the target both trusts and sees as worthy of emulation. Civic engagement messages are no different.
FIGURE 4.7The Impact of Latino Messengers on Nonvoters—Endorsers Who Would Make Them More Likely to Register and to Vote
Source: Latino Decisions survey, 2013.
Figure 4.7 illustrates the impact that different endorsers would have on the decision of current nonvoters to register or vote. All of the groups agreed that coethnic firefighters, police, doctors, nurses, and teachers as messengers would have a significant influence on them. These community notables, whose leadership is earned through social position and other avenues of trust, consistently demonstrated the greatest persuasive authority in getting nonparticipant Latinos to consider registering and voting. Well over half of low-, moderate-, and high-propensity voters agreed that these Latino endorsers would have a positive impact on their political participation.
By contrast, entertainment figures (actors, athletes, celebrities) did not have much influence, though Univision anchors Jorge Ramos and Maria Elena Salinas remained trusted figures for high- and moderate-propensity groups. Even Piolin, the syndicated radio personality often credited as one of the significant influences on the massive mobilizations during the 2006 immigration marches, does poorly when compared to nurses, teachers, and firefighters. To be clear, this does not mean that these individuals have a negative effect; we have no evidence that they drive people away from the voting booth. Moreover, different people respond to different messages, and mobilization of this hard-to-reach electorate may require a variety of players and pitches. We d
o, however, clearly find that famous people are less important than trusted community members when mobilizing people to vote. Of course, other factors beyond the boundaries of the Latino community might also have an effect. Figure 4.8 highlights the influence of certain non-Latino individuals on political participation. Family and friends have a consistently strong positive impact for all groups: if asked by a close family member or friend to vote, two-thirds or more said that they would do so, and this finding is consistent with all we know about Latino familial closeness and the importance of family information networks.
Unions are powerful endorsers for high-propensity nonvoters, but not so much for low- and moderate-propensity groups. Organized labor is consistently one of the strongest advocates for Latino civic engagement and immigration reform. This support derives from the fact that the labor movement comprises a growing share of Latinos, a natural consequence of the movement of Latinos into the building trades and the movement of unions into the service sector of the economy. The influence of unions as messengers was greatest among those nonvoters we found to be more readily persuadable.
It is striking, and entirely contrary to popular narrative about Latinos, that priests and ministers rated lowest for moderate- and high-propensity groups. The most informed and engaged nonvoters were less likely to be influenced by the endorsements of religious leaders. This powerful piece of information echoes the earlier observation that Latinos, no matter how religiously observant they are, do not look to their religious leaders for political guidance.
Finally, Figure 4.9 reports the effect of specific family endorsements on voting decisions. The results are essentially what we might expect: mothers, fathers, and spouses or partners had the most positive influence relative to all other family members. It is quite noticeable that the low-propensity voters seemed to be much less moved by family opinions, which were so potent for others in the survey.
FIGURE 4.8The Impact of Non-Latino Messengers on Nonvoters: Endorsers Who Would Make Them More Likely to Register and to Vote
Source: Latino Decisions survey, 2013.
WHAT HAVE WE LEARNED ABOUT THE MISSING LATINO VOTERS?
Eleven million uncast ballots is a huge wasted resource that Latinos can ill afford. Improving voter registration and turnout is the sine qua non of increasing Latino electoral clout.
Some Latino resource disadvantages (age, specifically) will diminish over time, and there are also considerable attitudinal advantages that will keep Latino turnout within striking distance of the goal of parity. The most salient of these advantages is group identity: it is clear that identity concerns drive turnout decisions for as many as one in three Latino voters.
The Latino share of the electorate, though still low, has approximately tripled in the last generation, and it grows with each passing election. Latinos have begun to play a pivotal role in national politics and national elections. We turn next to exactly how, and with what effect, they are taking on this role.
FIGURE 4.9The Impact of Family Messengers on Nonvoters: endorsers Who Would Make Them More Likely to Register and to Vote
Source: Latino Decisions survey, 2013.
Part II
LATINOS AT THE POLLS, 2008–2012
Chapter 5
THE 2008 DEMOCRATIC PRIMARY
With Sylvia Manzano, Gabriel Sanchez, and Ali Valenzuela
Latino voters received unprecedented attention from candidates and media in the 2008 presidential election. During the primaries and general election, campaign staff and political observers repeatedly noted that electoral competitiveness and outcomes hinged on the Latino vote in many states. Frequently referred to as Hillary Clinton’s “Latino firewall” in the primary season, Latinos came through for Clinton: she won the Latino vote by two-to-one margins in nearly every state. It was a defining moment in Latino American politics as Latinos became decisive.
Before 2008, Latinos had had few meaningful opportunities to shape or participate in primary elections. Competitive primaries typically occur early in the primary season, because states that hold their contests early have higher turnout rates and greater influence in determining the party nominees.1 The states that hold the first primaries and caucuses are also among the most homogenous in the country: in 2008, 93% of Iowa’s and New Hampshire’s eligible electorate were non-Hispanic white voters.2 Primary candidates quickly abandon their presidential campaigns after a front-runner has been established.3 Once this happens, voters have little incentive to participate as primaries become noncompetitive and even perfunctory exercises. With America’s Latino electorate geographically concentrated in a handful of states (nearly half of all Latino voters reside in California and Texas alone), the opportunity for Latino voters to participate in primaries of actual consequence or to influence the nomination outcome is especially sensitive to the party-established schedule that dictates when each state holds its primary.
TABLE 5.1States with a Significant Latino Population in 2008
Source: US Census Bureau, Current Population Survey, November 2008; CNN primary exit polls, 2008.
a. CVAP = Voting-Age Population by Citizenship and Race
The Democratic Party changed the 2008 primary schedule, in part to address the lack of Latino representation in the early primary states. That institutional change, coupled with the enduringly competitive contest between Hillary Clinton and Barack Obama, made it possible for the Latino electorate to claim a new and powerful role in choosing the Democratic Party’s presidential nominee. Because so many heavily Latino states held primaries on or before February 5, 2008, Clinton and Obama developed Latino outreach strategies early in the primary season and deployed those resources throughout their campaigns—a first in American politics. Clinton’s Latino effort was more advanced than Obama’s, but both camps signaled their seriousness by including high-profile Latino campaign staffers and consultants, actively seeking endorsements from Latino elected officials and public figures, investing in Spanish and English ads targeted at Latino audiences, and establishing field offices in locales to maximize Latino outreach.4
In the three traditional early primary states, Iowa, New Hampshire, and South Carolina, Latino voters accounted for less than 1% of the statewide electorates. But Latino voters comprised more than 10% of the electorate in New Mexico (35%), California (30%), Arizona (18%), Colorado (16%), Nevada (15%), New Jersey (12%), Florida (11%), Illinois (10%), and New York (10%).5 Conventional wisdom held that Texas, with its March primary date, would be “too late to matter,” but instead, the state played a pivotal role in keeping Hillary Clinton’s campaign alive when she narrowly won. One-third of the electorate was Latino.6
THE POLITICAL CONTEXT OF PRIMARIES
The factors that influence vote choice in presidential primaries and caucuses are fairly well established. Perceived viability, candidate quality, and candidate strength are of the utmost importance. Voters want to support a candidate who has a realistic chance to win the nomination and the general election.7 Candidates with deep pockets tend to win early, creating momentum that funnels more contributions.8 Hillary Clinton entered the race with significant advantages on these points. Many political professionals considered her nomination inevitable, and her capacity to fund-raise was unmatched by any other Democrat at that time.9
Voter preferences are also shaped by issues and ideology. These distinctions can be more difficult for voters to make in primary elections than in general elections.10 In the Clinton and Obama contest, policy and ideological differences were not especially sharp. Both favored immigration reform that included a pathway to citizenship, health care reform, and withdrawal from Iraq. There was no party-defining issue on which the two held opposite views (as we saw, for example, in the 2012 Republican primary when Governor Rick Perry of Texas voiced support for in-state tuition rates for undocumented immigrants). Beyond these traditional factors, targeted mobilization has proven especially effective with Latino voters, who tend to respond positively to coethnic and bilingual outreach
.11 Clinton held a clear advantage in this area.
PRO-CLINTON = ANTI-OBAMA?
During the bruising 2008 Democratic contest, a narrative emerged asserting that racial prejudice among Latino voters accounted for their strong support for Hillary Clinton. Journalists produced many articles echoing the notion that racism kept Latinos from voting for Obama.12 Clinton campaign insiders perpetuated the claim too. Soon after Clinton’s victory in the New Hampshire primary, Sergio Bendixen, who conducted Hispanic polling for the Clinton campaign, told The New Yorker, “The Hispanic voter, and I want to say this carefully, has not shown a lot of willingness or affinity to support black candidates.” Even in late June, when the contest was all but over, Clinton’s campaign chairman, Terry McAuliffe, told reporters that Obama had a problem with the Latino community and could not close out the votes needed with this key constituency.13
Bendixen and McAuliffe were not alone. Following Hillary Clinton’s impressive performance with Latino voters on Super Tuesday, conservative commentator Raoul Contreras wrote in the Los Angeles Times, “When truly given a choice, Latinos will not vote for a black candidate.”14 From this beginning, the story was spun into a broader discussion of black-brown political tension and even became the subject of press conferences by academics.
Latino America: How America's Most Dynamic Population is Poised to Transform the Politics of the Nation Page 7