Insights from the Latino Political Participation Poll
Below are insights from a study of social and psychological factors that impact voting, along with a segmentation of Latino non-voters, both based on a national poll of 2,800 eligible Latino voters.
For actionable takeaways designed specifically for use by organizers, see an accompanying list, “For Organizers: Considerations for Latino Mobilization.”
Table of Contents
- Background
- The Five Types of Potential Latino Voters
- The Clusters
- Should-Be Voters
- Politically Unassimilated
- Uninterested
- Alienated
- Cynics - Lessons
- Lesson 1: The Need for Education About Parties and Ideologies
- Lesson 2: Increasing a Sense of Influence
- Lesson 3: Voting While Uninterested
- Lesson 4: Family and History as a Foothold
- Lesson 5: The Role of Campaign Contact
- Lesson 6: The Engine of Turnout - Additional Reading
Background
Only 50% of eligible Latinos voted in the 2020 election, per Catalist estimates, a rate significantly lower than that of non-Latino AAPI (62%), Black (63%) or White (74%) voters.
Neither demographic nor socioeconomic factors are enough to completely explain this “turnout gap” between Latinos and other racial/ethnic groups, and in any case, there is little that a campaign or organization can do to address those disparities during the course of one, or even several, election cycles. There is some ingredient (or ingredients) at work in this recipe that is unique to Latinos and the Latino experience.
Similarly, the task of closing a 12 or 24 point gap in voting rates seems out of the league of proven get-out-the-vote tactics — which at their best still show very small effects on turnout, and which are not tailored to maximize effectiveness among Latinos.
With these facts in mind, Equis and Sojourn Strategies were invited by the Democracy & Power Innovation Fund to do a deep dive on the factors that explain voting and non-voting among Latinos, with a focus on what is most useful for those who are trying mobilize Latinos in this moment: factors that are unique to the Latino experience, and that are shapeable — values, norms or perceptions that if surfaced can help groups deepen their civic engagement programs.
Working with a set of organizing groups (Florida Rising, Make the Road Nevada, Poder NC, LUCHA, Voces de la Frontera, and PICO California) and a team of researchers (Gabriel Sanchez, Matt Barreto and Sylvia Manzano from BSP Research, and Tyler Reny at Claremont Graduate University), we identified several compelling theories of Latino political participation and designed and executed a study to test them with 2,800 eligible voters nationally who identify as Hispanic or Latino.
We were able to compare self-identified Latinos who said they voted in the 2020 election and those who did not vote, and additionally to compare those who said 2020 was their first-time voting to those who regularly participate.¹
What we found: our research identified several key factors that help to distinguish between voters and non-voters, including political influence, interest and integration. Building on these findings, we further identified five distinct types of non-voters, highlighting the diversity of experiences and attitudes within this group. People don’t all abstain or participate for the same reasons. Nor are the same mobilization tactics likely to be equally effective with all. For some, standard campaign contact might be enough to nudge them over the edge. For others, the opportunity for mobilization may lie in deeper engagement on family tradition and history.
THE FIVE TYPES OF POTENTIAL LATINO VOTERS
Based on a series of multiple regression analyses, we identified several key factors that help to distinguish voters from non-voters in our data.
- Political influence/efficacy — Latinos who believe that “my vote has the power to make change on issues that are important to me” and/or that “Hispanic or Latino voters” as a group have that same power are significantly more likely to have voted in 2020 than those who do not share those beliefs.
- Interest in politics — Latinos who report that they are “very interested” in politics are (unsurprisingly) much more likely to have voted than those who are less interested or not interested at all.
- Partisan integration — This was the biggest predictor in our analysis. Partisan integration is a measure of whether a person feels informed enough to answer questions about the party system and their place in it: their party affiliation, their ideology, or the differences between parties. Those who are highly integrated can easily answer all of these questions (even if they view themselves as moderate or independent), whereas those who are less integrated feel that they don’t know enough to say. These Latinos are much less likely to report having voted compared to their better integrated counterparts.
However, if we stopped here, we might miss some of the important ways in which these factors interact with one another to form unique profiles of non-voters. Therefore, we fed these factors into a k-means cluster analysis, and also threw in measures of perceived discrimination and sense of belonging in the United States. Together they help flesh out the picture of different types of non-voters: who they are and what their experience as Latinos in the U.S. has been like.
Looking at non-voters this way allows us to see the diversity of non-voters: different psychological factors work together in unique combinations to shape the decisions of different segments of the population.
This kind of analysis also demonstrates that there is a way in, so to speak, with every group. For every psychological barrier, there is also a resource or opportunity to invoke. Whether it is the high sense of belonging among the politically unassimilated and the cynics, or the high political interest and group efficacy among the alienated: they all have some attributes that connect them to civic life.
On that note, these clusters are designed to deepen the understanding of potential voters, not for targeting. Most mobilization campaigns will be communicating to multiple segments and should think in terms of strategies that invoke various social/psychological factors.
The Five Clusters
- Should-Be Voters: This cluster looks indistinguishable from a 2020 voter on our key dimensions. Neither influence, interest or partisan integration seem to be a barrier to participation; they have a high sense of all of those. They are older and more educated than our other types of non-voters, and overall are fairly partisan (38% identify as Democrats, 30% as Independent, and 24% as Republican). Interestingly, they are the most foreign-born cluster in this survey.
- Politically Unassimilated: The in-betweeners. They do feel like they belong in this country, but aren’t at all incorporated into the political system: nearly 3-in-4 don’t know enough to say which party they identify with, they say they don’t follow politics, and they have only modest sense that either their vote or the vote of their peers has the power to create change. They don’t have a distinct demographic profile, but notably, they are the most likely to prefer to communicate in Spanish.
- Uninterested: This cluster is checked out of politics. Not because they lack a basic understanding of the party system (partisan integration is high), but because they express a low level of interest in politics. It is the oldest, most female and most Mexican-American cluster. Nearly half identify as Independent; only 2% say they are Republican and some 1-in-5 don’t know enough to say.
- Alienated: A group that is highly knowledgeable of the political system but feels othered. A low sense of personal influence is matched with an increased perception of discrimination against them and a reduced sense of belonging — factors that might interrelate. In spite of it, they believe Latinos as a group are powerful and do report an interest in politics. This is the youngest (83% under 40) and second-most female (57%) cluster, and also the least likely to prefer to communicate in Spanish.
- Cynics: The most male cluster (56% are men). Like the Alienated, they have a low sense of personal influence. Differently, they have only middling levels of political interest, on one end, and a high sense of belonging, on the other. This cluster is also the least likely to have a four-year degree (91% non-college), the most Protestant and the most Republican (despite a plurality still identifying as Democrats).
LESSONS
This analysis suggests some lessons for reaching deeper into the universe of potential Latino voters:
Lesson 1: The Need for Education About Parties and Ideologies
Key clusters: Politically Unassimilated, Uninterested
No other factor was as predictive of voting in our study than partisan integration.
On the surface, the number of people who fit into the low integration category can appear quite small. They also tend to cluster into two segments of the non-voting population. But in this way partisan integration is a cautionary tale about minding the margins. The “don’t know” answers on major partisan questions are often dismissed as an inconvenience. Our analysis confirms that those who “don’t know” are, in reality, a critical part of the story of the Latino electorate. And they aren’t actually that small a group, in aggregate: 1-in-4 Latinos chose the “don’t know enough to say” option in at least one of our partisan questions.
Zoltan Hajnal and Taeku Lee, from whom we adapted this concept, explain why a lack of partisan integration is not a matter of apathy or a failing of the individual but a rational reaction to the political establishment:
When the choices the major parties present to the American public do not match our available stock of political knowledge (information), our deeply held political beliefs (ideology), or how we think of ourselves (identity), nonpartisanship becomes a rationally adaptive strategy.
In other words, those low on partisan integration are the undrafted players of the political system: they remain on the sidelines of politics because they are unclear what team they are supposed to be playing on. The two major parties have not bothered to invite them.
Those less integrated tend to be those who have been without institutions or experiences that would orient them toward matching their ideologies or identities with a political party: those who didn’t go to college, or were born in another country, or don’t attend a church. Most notably, they are women — and, especially, young women.
The example of those born in Puerto Rico is always poignant: voters on the island have some of the highest voting rates in the world. But when they move to the continental United States, they vote at middling rates. Why? Theory and experience would suggest they aren’t integrated: their political preferences don’t fit neatly in the US two-party system (Puerto Rico has its own local political parties), and there aren’t sufficient institutions in place to connect the dots.
We often talk about partisanship in terms of its corrosive, divisive effects. But partisanship and ideology are the grammar of American politics. You cannot understand or be understood in US politics without them. If we believe greater participation is good for our democracy, then we need more partisan and ideological education.
And, yes, this kind of education on the ideological differences and agendas between and among political parties and leaders can be done in a non-partisan manner. For an example, look to the right-wing: the content of PragerU, a non-partisan 501(c)(3) organization often seeks to explain the differences between ideological movements in the United States, tied to the behavior of elected officials.
Lesson 2: Increasing a Sense of Influence
Key cluster(s): Alienated; also Cynics, Uninterested and Politically Unassimilated
A large majority of Latinos agree that, respectively, “my vote has the power to make change on issues that are important to me,” and that “Hispanic or Latinos voters” have that same power. But our analysis shows that those who disagree with those statements are significantly less likely to vote.
The belief that the Latino vote is (or isn’t) influential explains some of the difference between voting and non-voting in our analysis: someone who strongly believes that Latinos, as a group, have the power to make change is about 10 percentage points more likely to vote than someone with similar demographics and political attitudes but who holds a dim view of the power of their group.
Where it gets especially interesting is when we look at the effect by gender.
For women, voting/not-voting is a matter of personal influence: those who believe that their individual vote matters are more likely to vote than those who don’t. It is, in fact, the most meaningful predictor of voting among Latinas.
For men it tends to be a matter of group influence — those who believe that Latinos have power are more likely to participate.
The picture it paints: collective Latino power is almost a given for Latina women, regardless of vote history, whereas personal influence more cleanly sets apart the segment that voted in 2020. The flip occurs with the men: personal influence is assumed regardless of behavior; but those who believe in the power of the group are more likely to vote.
A job for future program and research both at Equis and elsewhere: what moves personal efficacy among women, and what increases group efficacy among men?
Lesson 3: Voting While Uninterested
Key clusters: Uninterested, Unassimilated
The least surprising finding, one that has been such a standard fixture of turnout studies that it is basically conventional wisdom, is that being “very interested in politics” strongly predicts voting.
We include political interest here because (a) it serves as a reminder to activists that voters are not all like them, (b) it helps distinguish between different types of potential voters and, most importantly (c) it is actionable.
Our interest in interest is less about how you get a person to become an avid reader of 538 forecasts or a devoted viewer of cable news, and more about how we create a permission structure that allows someone to feel like they can and should vote even if they are not deeply interested in politics.
Lesson 4: Family and History as a Foothold
Key Cluster: Alienated
Most eligible Latinos — 3-in-4 in our poll — say that honoring the struggles and traditions of their parents, grandparents and ancestors is a “guiding principle in my life” (41% report they agree strongly).
When a social group shares that obligation, and then also shares an understanding of the kind of action needed to meet that obligation, it becomes a social norm. When a group believes this duty calls for taking civic or political action, it is mobilizing.
But different racial and ethnic groups think differently about this duty to ancestors, according to Alison Anoll’s The Obligation Mosaic (from where we take this theory): for some groups, the best way to honor the legacy of family is through economic or academic success or about preserving traditions and culture.
We found that only a third of Latinos see voting in elections as one of the best ways to honor those who came before them. By comparison, a solid majority (54%) choose “working hard to support my family”.
This doesn’t explain the difference between voting and non-voting among Hispanics overall in our study. But in the answers, we see an opportunity. As Anoll says, “participatory social norms are a resource communities can wield to overcome the inherent costs and barriers of engagement.”
Take the “alienated” segment of non-voters: we find they are more likely than other non-voters to believe the best way to honor family is through “learn[ing] my history,” “continu[ing] traditional practices around food, clothing, music, or culture,” or “speak[ing] the language of my family.”
While this has not translated fully into voting, it is not hard to imagine a bridge from history, tradition and language to the importance of civic participation — perhaps by connecting for the very young segment of alienated Latinos the struggle of previous generations of Hispanics to gain their rights in this country with current struggles they face in their communities.
Similarly, Latinos who believe in honoring their families through hard work — the top choice of respondents in our survey — might find in political action a way to elect leaders who value and reward their contributions to the economy.
This opportunity exists, we believe, with anyone who says that honoring family is a guiding principle of their lives.
Lesson 5: The Role of Campaign Contact
Key Cluster: Should-Be Voters
Mobilization as a tactic is often reduced to “campaign contact.” This can in some ways be an over-simplification: as our study shows, the factors that shape a person’s attitudes toward voting are so hefty that it seems reductive to think they can be activated with a “touch,” as if it were a matter of flipping a switch.
However, we did find that a voter who reported being asked by a campaign or organization to participate in the 2020 election was more likely to vote, all else being equal.
Campaign contact is a mechanism for moving all the other factors we describe above. There is power in engagement both as a delivery method for information and in the message the contact itself conveys.
Campaign communication relays information about the parties. It increases a voter’s sense that their vote, or the vote of their group, matters — reporting that you are being contacted by a campaign reveals a sense that you are being courted, that your vote is valued. And, if contact doesn’t outright increase interest, being invited to vote at least creates permission to vote regardless of how closely you follow politics.
Nowhere is the impact of campaign contact more obvious than among the subset of Latino voters who did not vote in 2020 but who in their attitudes are essentially identical to those who did vote. We call these the “should-be” voters. If you compare their answers to those of Latinos who voted for the first time in 2020, one difference jumps out: whereas a majority (56%) of the first-time voters reported being asked to vote, a mere 15% of the “should-be” voters said the same.
Lesson 6: The Engine of Turnout
To close, it might be helpful to see how the theories discussed above fit within a larger model of turnout.
The motor of voter turnout has three interlocking gears: (a) the effort that elites invest in mobilizing voters, (b) each individual voter’s sense of their own empowerment, and © the perceived electoral relevance of the group to which that voter belongs. The calculus of who gets courted, and who feels courted, is related to the perceived political impact of the group (starting with its relative size) within a political coalition.
These gears turn together in a virtuous cycle. When elites (parties, candidates, civic organizations) focus attention on a group of voters by vying for their support — via campaign contact — the members of that group are more likely to become integrated into politics and believe in their own electoral influence. This increased sense of integration and influence leads to increased political participation, which in turn leads both elites and the voters themselves to see their group as more and more pivotal to election outcomes… and invites further mobilization by elites. It goes on and on.
Right now, the turnout motor is broken when it comes to Latino voters. The virtuous cycle has become a vicious cycle. Low relative turnout leads both elites and Latino voters to underestimate their collective power, which leads to underinvestment by elites, which leads to… low turnout.
But it can be repaired.
Increasing Latino participation begins by understanding that turnout isn’t one thing — a simple switch that can be flipped by a single tactic — but rather an inter-connected machine that runs on the complex identities and social dynamics of a set of diverse individuals. More effective mobilization requires a deeper and more multi-faceted approach. At the same time, it isn’t rocket science: people who are invited to a party they want to attend are more likely to show up. When someone feels courted and powerful — when they see that voting is a way to honor the contributions of the people who came before them and as a way to care for their families today — they are going to participate in elections, and the effects of that participation will be passed across people and time.
ADDITIONAL READING
- For more on the “honoring ancestors norm” and other participatory social norms, see Allison P. Anoll’s book, The Obligation Mosaic: Race and Social Norms in US Political Participation
- For a wide-ranging analysis on the disparity in voting between white and non-white voters in the United States, see Bernard Fraga’s book, The Turnout Gap: Race, Ethnicity, and Political Inequality in a Diversifying America
- For more on nonpartisanship, see Zoltan L. Hajnal and Taeku Lee, Why Americans Don’t Join the Party: Race, Immigration, and the Failure (of Political Parties) to Engage the Electorate
- For more on Latinos and belonging, see Angela X. Ocampo’s dissertation (“The Politics of Inclusion: A Sense of Belonging to U.S. Society and Latino Political Participation”) and forthcoming book
- For more on American non-voters more broadly, see the Knight Foundation 2020 report on non-voters, “The 100 Million Project: The Untold Story of American Non-Voters”
- For more on coupling threat and opportunity in messaging to Latinos, see Vanessa Cruz-Nichols, including “How to Sound the Alarms: Untangling Racialized Threat in Latinx Mobilization”
THAT’S ALL FOR NOW
Remember that for actionable takeaways of this article, see “For Organizers: Considerations for Latino Mobilization.”
Follow Equis Research on Medium, or visit our website, to see more on the Latino electorate.
¹ In order to get a broader set of eligible Latinos, we did not use a voter-file-matched sample. As such, voting in this study is self-reported, which is known to overstate turnout. Unsurprisingly, the share of voters in our sample supersedes the Catalist estimates for Latino participation in the 2020 election. While imperfect, we have reason to believe these answers still give us enough for meaningful comparison and analysis. Further research could look to validate the findings with administrative data.