The Latinx Vote in 2020 (Part I): the Post-Primary State of Play
Some insights from Equis Research’s 2019 data
The Equis Research team is digging into its data to provide context on the Latinx vote in this year’s presidential election. This is the first post in the series.
With less than 200 days until the general election there is a Democratic nominee.
2020 brings a new political environment (the uncharted territory of a global pandemic), a new electorate (Latinos will make up the largest minority group in the electorate at 32 million), and a new battleground map (welcome, Arizona).
In our first post, we’ll look back at the bruising but instructive 2020 primary to pull out insights on Latinx voters that can be applied to the general election.
Equis Research works with a team of polling firms — GBAO, Global Strategy Group, LatinoDecisions, Myers Research, and TargetSmart — to survey Latinx attitudes. We have interviewed 28,227 Latino voters in 11 states since July 2019.
Note: Most of the polling cited here is from 2019 (July-December). Biden likely consolidated some part of the Latino vote after his South Carolina and Super Tuesday wins, as evidenced in his subsequent success in states like Arizona and Florida. Still, lessons remain from these early takes on the Democratic field.
What did we learn from the primary? Where does it leave Joe Biden?
In the 2020 primary, Bernie Sanders seemed to crack the Latino code. He inspired wide and deep support among Latino voters, fueling his big win in Nevada and buttressing his entire campaign, in a way that few outside his most fervent supporters could have anticipated in the run-up to this cycle. (The first article about Sanders’ “barrio cred” caught even many Latino activists by surprise.)
What about his performance with Latino voters do we need to know going into the general election?
We’ll highlight two aspects.
First, the Sanders campaign proved the value of a committed outreach effort to Latino voters.
A simple, if imperfect, comparison between two western states — Nevada and Colorado — suggests, at least, a quantifiable impact from a heavy investment in Latino outreach.
Nevada was a major priority for Sanders, the target of heavy investment that was crescendoing as we polled. Colorado, while important, wasn’t receiving the same attention, and its primary was a couple of weeks away.
If we were going to see the fruits of Bernie’s labor in a Latino electorate, it’d be in Nevada. We wouldn’t expect to see a similar bump in Colorado.
Sure enough, over the last two months of campaigning in Nevada, Sanders’ popularity grew from +12 (December) to +25 (February) among all registered Latinos (not just Dems). Among Latino Democrats, Sanders’s favs rocketed from +37 to a staggering +60.
The same was not true in Colorado. While he started at a similar point in CO (+9), he found himself in more or less the same place in February (+10). Among Colorado Dems, his number stayed fixed at +30.
The disparity in campaigning between the two states also shows up in voters reporting whether they’d been contacted by a campaign or political organization about this year’s election. Only 24% of Latino voters in Colorado said they’d been contacted, compared to 45% in Nevada. The percent of voters saying they’d had a conversation with a campaign was nearly double in Nevada what it was in Colorado.
Two, Sanders resonated deep into the electorate.
Of course, campaigning doesn’t automatically translate into support. Take one Michael Bloomberg. In Texas, where he spent significant dollars targeting Latino voters (behind some good ads), his name ID among Hispanics was comparable to that of the other top contenders, but his popularity was underwater (-8).
While Biden, Sanders and Warren were all well-liked by the Democratic base, Bernie was the only one with positive ratings with the least engaged cohorts of the electorate. We’ve seen this across the states in which we poll.
Naturally, some part of this is name ID. Bernie has been at this since 2016. It is no surprise that he and Biden have the highest name ID among Latinos. Compare this with a candidate like Pete Buttigieg, who in our polling was completely unknown to just over half of Hispanic Democrats in Nevada.
Also true is that, at the time, Bernie had avoided many direct attacks from his primary opponents, whereas Biden was the subject of them, and his numbers reflected the damage.
And yet the disparities in favorability would indicate there was more to it than that. It’s not just that Tío Bernie was a candidate Latino voters could recognize, but that the basic snapshot of him that penetrated to less engaged voters seemed to resonate.
The chart below looks at favorability for some of the leading Democratic candidates broken down by our custom segments. The “Mobilization” cohort represents voters very likely to vote for a Democrat for president but less certain to vote. The “Ambivalent” cohort — as we’ll explain in greater depth below — is voters who aren’t solidly locked into their presidential vote choice.
Sanders was by far the better-liked candidate among Latino mobilization targets, and was the only Democrat to generate positive feelings among the Latino ambivalents.
Meanwhile, much has been said of Sanders’ support among the younger set of voters, across race and ethnicity. Less reported is that among Latinos he was even more popular among younger women than among younger men. (This was true across states.)
Younger women are the largest age/gender cohort in the Latino electorate and still manage to be over-represented in the Dem Base, Mobilization and Ambivalent cohorts. Resonating with Latinas under the age of 50, as Sanders did, is a key to unlocking new potential highs of Latino support and turnout.
Side note: age and gender are the top demographic markers any campaign should be focused on when it comes to Latinx voters. We have so much to say about the topic we’re saving it for a future piece.
Some editorial thoughts beyond the data:
The question for Joe Biden is how much of this is replicable in his campaign. Is there a combination of policy and character that rings the same bells? Or was Sanders, the man and his movement, simply unique? The answer — which Equis will continue to seek in on our upcoming 2020 research — is probably somewhere in the middle. Biden can’t impersonate the anti-establishment populist firebrand, and attempts to do so would cut against the perceived authenticity at the core of Sanders’ appeal. But he has room to maneuver on issues that touch Latino pain points (healthcare and higher education chief among them) and re-introduce himself, connecting authentically on values. It bears remembering that while Sanders had made some inroads with Latino youth in 2016, he was not widely regarded as the Tío Bernie we know today. Putting in the work is half the battle.
The Sanders campaign in Nevada (and California) is useful as a case study because the core elements of their outreach strategy weren’t rocket science. Instead they appeared to be the function of a senior leadership team that asked itself everyday, across departments, what they could be doing to earn the Latino vote. It isn’t one thing. It’s deploying all the assets and tactics available to a campaign, from phones/doors/texts to organic/paid digital to radio/TV/mail to events to press access — and to the candidate’s time.
Will there be holdouts?
Given Sanders’ success with Latino voters, and Biden’s early struggles, will some Latino Bernie supporters join the Democratic Socialists of America and some members of the twitterati in withholding their support for Joe Biden?
In 2016, a significant proportion of voters (25.7%) who supported Bernie Sanders in the primary chose not to vote for Hillary Clinton in the general election. Of the defectors, 12.0% voted for Trump, 10.2% voted for a third party, and 3.5% didn’t vote. In a close race, numbers like these have the potential to affect who wins and who loses.
Our data from December and February don’t allow us to identify Sanders voters, since voting had yet to begin at that point. But we can look at Latinos who had a favorable view of Sanders going into the first primary contests. While favorability ratings do not translate directly into vote choice in the general election, they can still be useful in identifying those who are particularly at risk of defecting to another candidate (or to the couch). We see voters who had a favorable view of Sanders and a negative view of Biden as falling into this category.
Overall, 36.9% of Latinos in our December survey had a favorable view of Sanders. A majority (58.0%) of those in this camp also had a favorable view of Biden, while only 17.3% had an unfavorable view (the remaining 24.8% were neutral/no opinion).
In sum, the Sanders favorable/Biden unfavorable group represents just 6.4% of the Latino electorate in our December survey.
If we break this down by state, Wisconsin emerges as the state with the largest group of potential holdouts (9.5%), while North Carolina has the smallest (4.7%). In our February survey, the Sanders favorable/Biden unfavorable group had increased slightly in Colorado (from 7.9% to 8.7%) and more substantially in Nevada (from 7.8% to 11.7%), perhaps due to the Sanders campaign’s efforts there.
How does this picture of the 2020 contest fit in with what we know about previous elections? Should Sanders holdouts be a major concern? To answer these questions, it may be better to go back to 2008 rather than 2016.
There have been many post-mortems on the reluctance of Sanders supporters to vote for Clinton in the 2016 general election, but these analyses have not focused on the role of Latino voters, perhaps because it was Clinton rather than Sanders who captured the Latino vote in the primary (Clinton won 10 out of 11 states where Latinos made up the largest share of the population).
However, in 2008, Barack Obama faced a challenge similar to the one that Joe Biden faces today. Obama needed substantial support from Latino voters in order to win the presidency, yet they had favored his opponent by nearly a 2:1 margin in the primary.
While a significant percentage of Clinton voters did swing to McCain in the 2008 general election, there is no evidence that Latinos did so at a higher rate than other voters, in spite of their strong support for Clinton in the primary. If anything, survey data suggests that Latino Clinton supporters were more willing to rally to Obama in the general compared to their White counterparts. Indeed, by mid-summer of 2008, Obama’s favorability rating among Latino registered voters (76%) had drawn equal with Clinton’s (73%).
One could fairly speculate that the perceived ideological divide between Sanders and Biden is more significant than the one between Obama and Clinton in 2008, or that the movement from a more moderate candidate to a more liberal one (Clinton->Obama) is smoother than the reverse journey (Sanders->Clinton, Sanders->Biden).
But the data seems to suggest far more opportunity for Biden than risk. There is a path to greater favorability among Latino voters for Biden, even though they favored Sanders in the primary. Indeed, Sanders holdouts could prove to be only a minor problem for Biden if his campaign offsets any small losses by taking advantage of other opportunities for growth.
Where does Biden have room to grow?
There is room for Biden to re-introduce himself with large segments of the Latino electorate.
An obvious starting point for growth is voters who don’t already hold a strong opinion of him.
In the Equis polling there was significant variability in Biden’s net favorability across states. His popularity seemed particularly low, for example, in the Mountain West states of Nevada, Arizona and New Mexico where his average net favorability among all voters was -6 (compared to an average of +9 in the other states).
But if we start by excluding likely Trump voters (those who said they are “definitely” voting for the incumbent), and focus on “potential Dem voters” we get a clearer picture of the voters with whom Biden will need to maximize support.
In most states, there is a large group of voters that are neither favorable or unfavorable toward Biden, falling instead into the neutral or don’t know categories. The Mountain West states top the list.
As with ambivalent voters (discussed in Part II), improving numbers here is likely an imperative for turnout as much as vote choice. The more these voters can be persuaded to choose sides, the greater the chance they will choose to cast a ballot.
Where does Trump start the general election?
As Biden himself is fond of quoting, “Don’t compare me to the Almighty, compare me to the alternative.”
So how is the other choice for president faring?
Pre-COVID, Donald Trump was still widely disliked by Latino voters, and there is no reason in public polling to think opinions have changed during the pandemic. Only in Florida were his numbers less than abysmal, thanks to the Cuban-American vote.
A point on national head-to-heads, for those will be obsessively tracking national surveys: they aren’t the thermometer you’re looking for.
Here are Trump’s re-elect numbers among Latino voters in our polling, combined across battleground states:
December
Dem 64 / Trump 24
November
Dem 64 / Trump 24
September
Dem 66 / Trump 22
July
Dem 64 / Trump 23
Combined
Dem 64 / Trump 23
We recently compared our results to a quality, non-public survey from four years ago this coming month (May 2016): in it, Clinton was leading Trump among Hispanic voters, 63–23.
Not exactly a rollercoaster ride of public opinion. Preferences are mostly locked at the highest level.
To pick up on movement and variance, we have to look at the state level, and at different metrics, with a close eye on the margins and the dynamics of turnout.
To gauge support for Trump, we’ve used a battery of metrics over four waves of polling: job approval (overall, and on economy, national security, immigration, and healthcare); favorability; opposition to impeachment & removal; and a re-elect match-up against a generic Democrat.
As among other voters, Latinos gave Trump his highest marks on his handling of the economy — pre-COVID19. That was when the economy was doing relatively well (for some). None of this polling reflects the current pandemic, or what kind of blame or credit Latinos might give Trump for what comes next.
Already in our earlier polling, reinforced in focus groups, we saw some indication that support for Trump’s economy was passive and didn’t reflect the voters’ personal financial situation.
Trump’s lowest ratings have consistently been on his handling of immigration. As we learned from focus groups, this isn’t merely a matter of policy disagreement — it appears to be an issue of morality and character. Even conservatives who agreed with Trump on some aspects of immigration policy reacted negatively to family separation, children in detention and Trump’s rhetoric.
All together, the Equis polling gives a fairly static picture of attitudes on Trump. Opinions aren’t swinging wildly from one survey to the next. However, there are a wide range of scenarios for Biden support depending on what dominates Latino voters’ minds when they vote — or, more aptly, when they decide whether to vote.
The below chart is an attempt to illustrate the point by choosing low/medium/high scenarios from among our various support metrics to calculate possible Democratic support margins (Biden-Trump).
For poll watchers accustomed to tracking the 50% line, under-performance among Latinos can be hard to spot. In a state like Arizona, a Democrat polling at 62% with Latinos sounds good. But the gulf between a 62–32 (+30) spread and a 65–28 (+38) margin can be terrifyingly wide in a presidential race. That’s where benchmarking support levels becomes critical.
In the next article from Equis Research, we’ll look at turnout…