The Latina/o/x umbrella?

Throughout this article, we use the term “Latina/o/x” to refer to immigrants or descendants of immigrants from Latin America. This term purposely reflects the word’s different suffixes, which aim to be inclusive of all gender identities.

The “Latina/o/x vote” is often discussed as a growing electorate, with the power to transform the political landscape of the United States. Resting on the premise of shared goals and interests, the concept of a cohesive group has a strong mobilizing potential within this growing population (Padilla 1985). At the same time, scholars argue that considering Latinas/os/xs only as a singular voting group may homogenize diverse interests within the group (Beltrán 2010). Despite the documented heterogeneity of groups categorized as Latina/o/x, existing research has generally focused on the largest national-origin groups. US Colombians reflect what are termed the “new Latinas/os/xs” (Cepeda 2010), representing growing sources of the Latina/o/x population from more diverse countries of origin, socioeconomic statuses, and contexts of reception upon arrival in the United States. Through a focus on US Colombians, we interrogate the political preferences of this group to understand how “new Latina/o/x” groups may reinforce or challenge notions of “the Latina/o/x vote.”

US ColombiansFootnote 2 encompass the largest national-origin group of Latinas/os/xs in the United States originating from South America, totaling 1.2 million residing in the United States as of 2017 (Noe-Bustamante et al. 2019). Since 2000, the number of Colombian Americans has increased 148%, substantially surpassing the growth rate of Mexican American populations (76%), who constitute the largest national-origin group (Noe-Bustamante et al. 2019). US Colombians are largely foreign-born (61%) and much more likely to be foreign-born compared to the overall Latina/o/x population (33%). In addition to their growing numbers, US Colombians also uniquely reflect the growing diversity within the Latina/o/x community. As the share of immigrants who are from Mexico has been declining, the number of immigrants from other countries in Central and South America has been growing. US Colombians reflect this newer wave of migrants in their mixed motives for migrating, such as fleeing violence and economic concerns (Cohn et al. 2017). US Colombians also have mixed socioeconomic-status backgrounds, a situation that reflects much of the growing diversity in immigrants, including migrants from South America who tend to have higher socioeconomic statuses compared to other Latinas/os/xs. As such, US Colombians represent an interesting group that mirrors much of the growing diversity in the Latina/o/x population in the United States.

Despite the potential for this growing population to contribute to our understanding of how Latinidad plays out in politics, existing studies either do not fully interrogate whether Latinas/os/xs are a cohesive political unit or mainly focus on differences among Mexicans, Cubans and Puerto Ricans. Much of the existing research on US Colombians examines political participation in Colombia (Jones-Correa 1998; Itzigsohn and Giorguli Saucedo 2002) or discusses how organizations mediate civic engagement among US Colombians (Portes et al. 2008). Examining less-studied subgroups is critical for a more comprehensive understanding of whether new Latina/o/x groups are shaping the “Latina/o/x vote” in ways similar to or different from more established groups.

In this article, we examine the political incorporation of US Colombians in comparison to other Latina/o/x subgroups. We define political incorporation as the development of political attitudes and opinions among immigrants and their descendants, including perceptions of political parties and candidates, as well as their involvement in the political system through voting. By examining Colombian Americans, one of the groups that makes up “the new Latinas/os/xs,” we seek to expand the discussion of the “Latina/o/x vote” to consider axes of similarities and differences by country of origin. We present one of the first accounts of the political attitudes and political behaviors of Colombian Americans at the national level, contributing to the emerging field of US Colombian studies. Furthermore, this article speaks to the larger debate around the “Latina/o/x vote” as it is deployed by civic organizations, the media and political elites, as our study highlights the nuances in the political attitudes and behaviors of Latinas/os/xs.

Latinas/os/xs as one group

The concept of Latinidad, and in turn the empirical examination of Latinas/os/xs, has been widely debated within Latina/o/x studies. Considering Latinidad as a political phenomenon can help mobilize the group (Padilla 1985), but can similarly homogenize varying interests, with scholars arguing that it should be permanently debated as a category (Flores 2000; Beltrán 2010). The “Latina/o/x vote” has been used by civic organizations to denote shared interests among a diverse set of groups, yet the term is sometimes used by political elites who do not seriously engage with the interests of this population (Rumbaut 2009; Beltrán 2010). Although quantitative studies of Latinas/os/xs acknowledge the heterogeneity represented by the term, studies of political incorporation generally are unable to interrogate the “Latina/o/x vote” beyond the largest national-origin groups, because of small sample sizes (Beltrán 2010). We build on the depth of qualitative studies of Latina/o/x experiences by examining and comparing the political attitudes and behavior of US Colombians to different Latina/o/x groups at the national level. This analysis will help inform political debates about when the concept of the “Latina/o/x vote” is fruitful and when it is imprecise.

Latina/o/x political incorporation

Studies of Latinas/os/xs in US politics focus on the diverse factors and experiences that shape their political incorporation, attitudes and opinions. Two key issues drive contemporary debates about the Latina/o/x vote: partisan identification and likelihood of voting. The political socialization literature posits that a person’s political attitudes, particularly partisan identification, are adopted early in life and are largely influenced by one’s parents (Campbell et al. 1980; Jennings and Niemi 1968). As such, some scholars argue that Latinas/os/xs are one of the largest “unaligned” groups, given that they are recent entrants into the US political system (Green et al. 2002; Hajnal and Lee 2011). When considering party affiliation, some scholars argue that socialization among Latinas/os/xs can involve a multistage party selection process, where they initially identify as independents or nonpartisan, and eventually decide whether they identify with a specific party (Hajnal and Lee 2011). This trajectory suggests that incorporation into politics occurs with more acculturation, as Hajnal and Lee argue that weaker connections to politics result in lower mobilization rates (Hajnal and Lee 2011). On average, Latinas/os/xs who adopt a partisan identification are more likely to be Democrats than Republicans, which is partly due to the mostly working-class composition of the Latina/o/x population and the documented relationship between socioeconomic status and partisanship (Cain et al. 1991; Garza 2004). Policy preferences are also important, and potentially more so than socioeconomic status, in shaping Latina/o/x partisanship (Alvarez and García Bedolla 2003). Republicans’ anti-Latina/o/x measures and proposals have also pushed Latinas/os/xs closer to the Democratic Party (Barreto et al. 2005; Bowler et al. 2006). Latinas/os/xs are also more likely to identify as Democrats as their length of time in the United States increases (Cain et al. 1991; Barreto and Woods 2003).

In terms of voting, it is debated whether Latinas/os/xs as a whole have lower electoral turnout rates than native-born whites, yet research has shown that foreign-born Latinas/os/xs have higher electoral turnout rates in many cases (Michelson 2003; Barreto 2005). The factors that have been demonstrated to be most influential in determining Latina/o/x political attitudes as a whole are the same characteristics that shape key differences within this group. Scholars argue that resource disparities contribute to lower levels of political knowledge and lower rates of electoral participation (Brown and Bean 2016; García Bedolla 2014; Leighley and Nagler 2016; Lewis-Beck and Stegmaier 2016; Michelson 2005). However, Latinas/os/xs in politics have been studied mostly as a cohesive electorate, with a few notable exceptions (García-Ríos et al. 2019; Fraga et al. 2010; Garza 2004; DeSipio 1998).

Disaggregating the Latina/o/x category has been done mostly in the case of the largest national-origin groups, who at times challenge the notion of the Latina/o/x vote. The studies that have examined different Latina/o/x groups find different perceptions of the political system; for example, South Americans are more likely to perceive that they have political knowledge relative to Mexicans and other Central Americans, but less likely than Cubans, who perceive the greatest political knowledge (Fraga et al. 2010). Cubans are also the most likely to believe that they can influence politics, with Mexicans and South Americans slightly less likely, and Dominicans the least likely (Fraga et al. 2010). On the other hand, South Americans have relatively high levels of electoral turnout in the 2004 election compared to Mexicans, yet somewhat lower than Cubans. Furthermore, Cubans are more likely to be Republicans compared to other groups, which poses the largest threat to a unified “Latina/o/x vote” (Garza 2004). South Americans identify as Democrats at rates similar to Mexicans, yet less than Dominicans (Fraga et al. 2010). However, beyond comparisons of Cubans, Mexicans, Puerto Ricans, and, more recently, Dominicans, the political opinions of other Latina/o/x subgroups have been understudied. We contribute to this line of research through our focus on Colombian Americans’ opinions and how these compare to other Latina/o/x subgroups.

Colombian Americans in comparison to other Latina/o/x subgroups

US Colombians share a number of similarities and differences with other national origin groups, which highlights the importance of studying this emerging subgroup. Colombian migration to the United States dates back to the early twentieth century, with the first significant wave arriving in the late 1960s (Cepeda 2010). Research suggests that Colombian migrants between the 1960s and late 1970s were primarily working class and motivated by economic concerns. Colombia’s drug war and increasing internal conflict throughout the 1980s, coupled with worsening economic conditions, resulted in migrants fleeing violence and poor economic prospects. The number of Colombians migrating to the United States has dramatically accelerated—more than doubling since 1990 (Cepeda 2010)—and is also increasingly composed of more middle- and upper-class migrants (Escobar 2004; Marrow 2005).

Colombian Americans tend to fall in the middle of the socioeconomic spectrum in relation to other Latina/o/x subgroups and are employed in both managerial and professional occupations as well as semiskilled blue-collar occupations, whereas many other Latina/o/x national-origin groups are concentrated on either end of the socioeconomic ladder (Marrow 2005). In addition, Colombian Americans in Los Angeles have higher rates of employment and higher socioeconomic status than those in New York, which highlights how a nationally representative sample may be useful for further study of Colombian Americans (Guarnizo et al. 1999).

At the same time, US Colombians are also likely to be different from larger Latina/o/x groups traditionally examined. First, US Colombians are mostly immigrants, compared with other groups such as Mexican Americans, who are primarily native-born. Colombian Americans also have a higher naturalization rate than other foreign-born groups (DiPietro and Bursik 2012; Migration Policy Institute 2015). These patterns may influence their political attitudes and may result in either higher or lower likelihoods of voting, because of eligibility. Research on US Colombians’ perceptions of discrimination in the workplace argues that they make sense of this marginalization as originating from their accents, origin and ethnic background, whereas research on other Latina/o/x groups suggests they perceive this treatment stemming from race or legal status (Rincón 2017).

Through their varied motives for migration that resemble those of newer migrants from Central America (Cohn et al. 2017), and their middle position on the socioeconomic spectrum (Marrow 2005), US Colombians uniquely reflect much of the diversity within the Latina/o/x umbrella. However, US Colombians also have unique characteristics that warrant the investigation of how they compare relative to other Latinas/os/xs, including their mostly immigrant origins, as well as their political experiences in their home country (Rouse 2017).

Colombian political incorporation

Although little scholarly work has highlighted US Colombians’ involvement in US politics, previous research provides conflicting findings about Colombian involvement in the politics of their home country. Although some scholars suggest that US Colombians have a wide sense of mistrust, due to Colombia’s history of armed conflict, and do not participate in Colombian elections (Guarnizo et al. 1999), other scholars who study the European context find that although Colombian migrants may be pessimistic, they are not uninterested in their country-of-origin politics (Bermudez 2011). Recent statistics from the 2018 Colombian presidential election show moderate electoral turnout among Colombians residing in the United States, as well as conservative political views. Out of the 355,000 Colombians living in the United States who were eligible to vote in the 2018 Colombian presidential election, 31.7% voted, which is lower than the 53.4% of Colombians overall who voted (Registraduría Nacional del Estado Civil 2018). Furthermore, Colombians citizens living in the United States were more likely to vote for Iván Duque, the Colombian conservative candidate, compared to those living in Colombia. Although this may show that US Colombians are less interested in their home-country politics, barriers to participation may also be a factor in why US Colombians do not participate in Colombian elections. Preregistration is required, and polling places are more scarce in geographic locations with fewer US Colombians.

The few existing studies that have examined Colombian American involvement in US politics suggest that their experiences with Colombian politics shape their likelihood of politically engaging in the United States. Some scholars suggest that mistrust of the political system in Colombia, in part fueled by decades of unresolved internal conflict, violence, and the drug war, may influence attitudes toward politics in the United States and may result in lower political involvement (Guarnizo et al. 1999; Escobar 2004). Scholars argue that this hesitancy to participate in politics emerges in the US context among Colombians in terms of electoral and civic participation (Escobar 2010). On the other hand, recent work based in Europe finds that, for Colombian migrants who were engaged in left-wing politics in their home country, engaging in politics in their destination country is important (Bermudez 2010).

Even though their experiences with Colombian politics may be influential, US Colombians may also derive their political attitudes primarily from their experiences once they arrive in the United States, which would result in greater similarities to other Latinas/os/xs. Scholars suggest that policies that restrict the rights of immigrants can spur political involvement in Colombian American communities (Guarnizo et al. 1999). However, other research suggests that compared to other Latina/o/x groups, US Colombians have yet to achieve a significant political force, with other Latina/o/x communities such as Dominicans having more success organizing political coalitions and electing members into office (Escobar 2004, 2010). Even when Colombians form organizations, they are not very likely to be involved in US-based political activities (Portes et al. 2008). This suggests that although US Colombians are less likely to be politically involved, there may be potential for mobilization.

However, the attitudes of Colombian Americans toward US politics have not been systematically studied, which may underestimate our understanding of Colombian American participation in US electoral politics. The scholarship that has examined US Colombians and political attitudes is generally largely qualitative and uses convenience samples. A quantitative study with a nationally representative sample comparing US Colombians to other subgroups has yet to be conducted.

Toward understanding US Colombian political attitudes and behavior

This article argues that an assessment of the political opinions and behavior of “new Latina/o/x” groups, which in this case consists of US Colombians, is essential for understanding points of commonality and difference under the umbrella term of “the Latina/o/x vote.” Prior work provides a limited picture on how Latinas/os/xs other than Mexicans, Puerto Ricans and Cubans are incorporating into US politics. Furthermore, in evaluating Latina/o/x political preferences, this article also considers how other sources of diversity within the Latina/o/x community, such as socioeconomic status, nativity, and the perceived importance of certain political issues, may contribute to differences across subgroups and for US Colombians more specifically. Although the heterogeneity within the Latina/o/x population is recognized among scholars, many studies generally lack a comparative perspective in examining particularities within the group. Examining US Colombians relative to other subgroups is of particular importance, because of the significant growth of this group over the past several decades, as well as its demographic similarities to and differences from other Latina/o/x groups. This article addresses these shortcomings and provides a nuanced understanding of US Colombians’ political attitudes, an understudied group within the Latina/o/x umbrella. By examining the political opinions of Colombian Americans in comparison to those of other Latina/o/x groups, we critically engage with the idea of a unified “Latina/o/x vote.” In order to do this, this article compiles several datasets to investigate Colombian Americans’ political attitudes at the US national level with a representative sample, and compares these attitudes to those of other Latina/o/x national-origin groups.

This article posits several hypotheses drawing from the existing literature. We expect that US Colombians will be in the center of the political ideological scale—in other words, more moderate than other subgroups and more likely to identify as Republicans. Since US Colombians occupy a middle position on the socioeconomic spectrum (Marrow 2005), they should have greater likelihoods of being conservative relative to other Latina/o/x groups with lower levels of socioeconomic status, but more liberal than those with higher levels (Verba and Nie 1987). The existing literature on US Colombians also suggests that they exhibit high levels of distrust in government, which may result in a lower likelihood of electoral participation (Guarnizo et al. 1999). However, contrary to this, we posit that Colombians are not likely to exhibit major differences in political participation compared to other Latina/o/x subgroups. Immigration is a key mobilizing issue for Colombian Americans (Guarnizo et al. 1999), therefore we expect to see fewer differences between US Colombians and other Latina/o/x groups with similar histories of migration and contexts of reception (Cain et al. 1991; Garza 2004) and for this group to be equally likely to vote in the 2016 election, where immigration was a key political and mobilizing issue (Gutierrez et al. 2019).

Data and methods

To investigate the attitudes and political involvement of Colombian Americans in the US, this article compiles six national surveys of Latinas/os/xs conducted during the 2016 presidential election to construct a large and nationally representative sample of US Colombians (n = 531). This dataset compiled by the authors is the largest nationally representative sample of Colombian Americans. We use this data to study the political preferences of US Colombians, including their partisanship, evaluation of candidates during the 2016 US presidential election and their likelihood of voting. We also rely on the entire dataset (n = 16,910), or subsets of it, to examine how Colombian Americans’ attitudes relate to those of other Latina/o/x subgroups.Footnote 3 Appendix 1 contains detailed information on each one of the samples, when the surveys were conducted, and who they surveyed. All of these samples, with the exception of the Collaborative Multiracial Post-Election Survey, surveyed Latina/o/x registered voters of all nationalities.Footnote 4 Appendix 2 contains information regarding the number of respondents by national-origin group in the pooled dataset.

Prior national studies of the political attitudes of US Colombians have not been feasible, given how few Colombians have been surveyed as part of large-scale nationally representative surveys of the Latina/o/x population. For example, the 2006 Latino National Survey, one of the most comprehensive studies of US Latinas/os/xs, included only 139 Colombian Americans (Fraga et al. 2007). On average, the number of Mexican, Cuban, Puerto Rican and Dominican respondents is much larger in national surveys because their populations are much larger. By pooling six surveys from the 2016 US presidential election, we have compiled a large enough number of Colombian American respondents to run statistical analyses and be confident about the robustness of our results.Footnote 5

Throughout several analyses, this article relies on questions that were asked in all or most of the surveys. For political party identification, the surveys first asked respondents which party they identified with the most. Respondents who did not identify with one of the two major parties were asked a follow-up question and were asked to pick a party if they had to choose one. They were allowed to choose Independent in the follow-up and were coded as such. The favorability questions asked respondents to evaluate their favorability of the 2016 presidential candidates Hillary Clinton and Donald Trump, as well as then-president Barack Obama. Each of these favorability items is measured on a five-point scale, ranging from very unfavorable to very favorable. Lastly, self-reported electoral turnout is used to measure levels of political engagement. This item was coded as 1 (yes) if respondents reported that they had or would turn out to the polls, and 0 (no) if they reported that they had not or would not vote in the presidential election.Footnote 6

Our models also account for other factors that are known to influence political attitudes and participation. We account for respondents’ age, gender and nativity (whether or not the respondents were born in the United States or abroad). For socioeconomic status, we include a categorical measure for income, wherein the first category represents an income of less than $20,000 and the eighth category represents an income of above $150,000. The measure for education includes a seven-point scale capturing highest level of education earned by the respondent, ranging from having earned a first- to eighth-grade education (1) to having earned a postgraduate education (7). In addition, we also examine whether the survey respondent elected to take the survey in Spanish (1) or in English (0), which we use as a measure of acculturation. We also examine attitudes toward the most important political issue facing the Latina/o/x community that individuals felt either the president or Congress needed to address. Respondents could identify up to two items. Some of these issues identified by the respondents included immigration, the economy, foreign policy, education, housing, healthcare, anti-Latina/o/x sentiment, climate change, and taxes, among others. Based on the distribution of these variables, we include the economy, education, discrimination, healthcare and immigration, but group other issues under “other.”

Results

Demographically, US Colombians generally occupy higher socioeconomic statuses relative to Latinas/os/xs as a whole, as they generally have higher incomes and are more likely to have college degrees. Consistent with other work on US Colombians, our sample indicates 68.4% of Colombian Americans were born abroad, compared to 32.2% of other Latinas/os/xs who were foreign-born.

When considering partisan affiliation of Colombian Americans relative to the largest Latina/o/x groups (Mexicans, Cubans, and Puerto Ricans, shown in Fig. 1), descriptive findings suggest that Colombian Americans are likely to identify as Democrats (68.9%), at a slightly lower rate than Mexicans but a slightly higher rate than Puerto Ricans. On the other hand, only 38.1% of Cubans identify with the Democratic Party, whereas 52% identify with the Republican Party. Only 21.1% of Colombian Americans identify as Republican, which is slightly higher than Mexicans (16.7%) and Puerto Ricans (18.3%). Colombian Americans are also marginally less likely to identify as Independent than both Cubans and Puerto Ricans, but more likely than Mexicans. Colombian Americans, despite their higher level of socioeconomic status and conservative political views in their home country, are not much more likely to identify as Republicans. In fact, their levels of partisan identification are very similar to those of other Latina/o/x origin groups, except for that of Cubans, who continue to show higher levels of preference for the Republican Party.

Fig. 1
figure 1

Partisan identification by national origin

Full size image

Figure 2 presents descriptive statistics for favorability of the 2016 presidential candidates, Trump and Clinton, as well as Obama, among Colombian Americans. US Colombians have very positive views of Obama, with 46.3% and 28.1% reporting that they view him very favorably and somewhat favorably, respectively. Although Colombian Americans also report favorable views of Clinton, these are substantially lower than their ratings of Obama, with only a small percentage (19%) reporting that they view her very favorably and 31.5% viewing her somewhat favorably. A sizable 36.6% of Colombian Americans view Clinton very unfavorably. Most Colombian Americans (55.1%) also see Trump very unfavorably, with smaller shares of the group viewing him somewhat favorably (15.6%) and very favorably (8.2%).

Fig. 2
figure 2

Candidate favorability among US Colombians

Full size image

Figure 3 examines the policy opinions of Colombian Americans and shows which political issue respondents believe is the most important. The majority of Colombian Americans reported that immigration (27.3%) and the economy (26.5%) are the most pressing issues. Discrimination, education and healthcare are all somewhat important, with less than 10% of the sample rating these as their top issues.Footnote 7

Fig. 3
figure 3

Most Important Issues for US Colombians

Full size image

Although the descriptive results provide an important avenue to understand the average political attitudes of US Colombians, accounting for other potential confounding characteristics through multivariate statistical models is essential for considering whether the differences between Colombian Americans and other groups can primarily be explained by other factors, such as different socioeconomic statuses. For country of origin, the models examine US Colombians (the reference categoryFootnote 8) in comparison to other large Latina/o/x origin groups. Because of small sample sizes, respondents from South American and Central American countries are combined into larger categories. In the analysis that follows, we use coefficients from logistic regression models representing different variables.Footnote 9 Given that the coefficients of logistic regression models are directly uninterpretable in terms of their magnitude (King 1998; King et al. 2000), sample predictions are drawn to examine the predicted probability of a respondent reporting a particular attitude or having voted. These predicted probabilities are calculated by holding all the other variables constant at their means.

Table 1 presents nested logistic regression models predicting respondents’ party identification. Colombian Americans share some similarities and differences in partisan identification with other national-origin groups. To interpret these differences, Fig. 4 shows predicted probabilities indicating the magnitude of country-of-origin effects, derived from models 2 and 4, which predict the likelihood of identifying as Democrat and as Republican with all relevant predictors held at their means, respectively. US Colombians have a moderately high likelihood (59%) of identifying as Democrat, yet this is significantly lower than Dominicans (68%), as well as higher than Cubans. On the other hand, they are not very likely to identify as Republicans (21%), which is significantly lower than Cubans (52%) but significantly higher than Dominicans (13%). In addition, US Colombians are more likely to identify as Republican compared to Mexicans, but this effect disappears after accounting for socioeconomic status and demographics. Socioeconomic status and other demographic characteristics operate in the expected direction, where age, education and income are all positively associated with being more conservative, meaning that older, more highly educated respondents and those with higher incomes are more likely to identify as Republican. Generally, these results show that US Colombians are different from Dominicans, who are more likely to identify as Democrats, and Cubans, who are more likely to identify as Republicans. At the same time, these findings show that US Colombians share similarities with most Latina/o/x groups in their relatively high likelihood of identifying with the Democratic Party.

Table 1 Predictors of partisan affiliation (Logit Models)
Full size table
Fig. 4
figure 4

Predicted probabilities of partisan identification

Full size image

Table 2 presents six models that assess favorability toward Obama, Clinton, and Trump. For each candidate, we present logistic regression models predicting the likelihood that they are seen very unfavorably, as well as models predicting the likelihood that respondents see them very favorably.Footnote 10 Generally, there are no country-of-origin differences in predicting favorability toward Obama, whereas there are differences in support for Clinton and Trump. Figure 5 presents predicted probabilities for the likelihood of viewing Clinton favorably and the likelihood of viewing Trump unfavorably (models 4 and 5), in order to interpret the magnitude of the coefficients for country of origin. US Colombians are less likely to have a very favorable view of Clinton, in comparison to other national-origin groups. Their likelihood of seeing Clinton as very favorable is around 16%, which is significantly lower than other groups such as Mexicans (23%), Puerto Ricans (26%), and even Cubans (21%). These differences remain after controlling for demographic characteristics and their political party identification. US Colombians are also more likely to view Trump unfavorably (64%) relative to other South Americans (54%). In analyses not shown, US Colombians are also more likely to view Trump unfavorably relative to Cubans prior to controlling for party identification. These results suggest that US Colombians have more negative views of Clinton in comparison to other national-origin groups. They are also somewhat different in their favorability rating of Trump, in that they are less likely to view him favorably compared to other South Americans.

Table 2 Predictors of favorability for sitting president and 2016 candidates (logit models)
Full size table
Fig. 5
figure 5

Predicted probabilities of viewing Clinton and Trump favorably and unfavorably

Full size image

When considering demographic characteristics, results are generally in the expected direction. Women, foreign-born respondents, and those who took the survey in Spanish are more likely to have favorable views of Obama and Clinton, but have unfavorable views of Trump. It appears that issue importance is an important set of predictors of favorability toward Obama, with respondents who identified education, discrimination, healthcare or immigration as a top issue being less likely to view him very unfavorably. On the other hand, those who think that immigration and education are the most important issues have lower likelihoods of viewing Clinton favorably. As expected, respondents who view discrimination and immigration as the most important issue are less likely to express favorable views of Trump.

Table 3 presents coefficients from logistic regression models showing candidates’ likelihood of reporting that they would vote in the 2016 election.Footnote 11 When considering electoral turnout, US Colombians are just as likely to vote as other Latinas/os/xs, rather than being less politically engaged as the literature suggests. The effect of demographic characteristics is in the expected direction, with older, more highly educated individuals with higher incomes expressing a greater likelihood of voting. Interestingly, Democrats also report a higher likelihood of voting relative to Republicans. In addition, it appears that immigration is a mobilizing force for Latinas/os/xs, who view it as the top issue, as they report greater likelihoods of voting in the 2016 election.

Table 3 Predictors of participation in the 2016 election (logit models)
Full size table

Last, we examine the determinants of attitudes among US Colombians alone, to assess whether patterns differ from those found among the Latina/o/x population as a whole. Table 4 shows political party affiliation among US Colombians. Colombian Americans who are older are more likely to identify as Republican and less likely to identify as Democrat, which is similar to the pattern found for Latinas/os/xs as a whole. Interestingly, neither education nor income appears to shape partisanship differently among US Colombians the way they do among all Latinas/os/xs, since these are not significant predictors of political party affiliation. Colombian Americans who took the survey in Spanish are more likely to identify as Democrat, which is another pattern that is not seen for Latinas/os/xs as a whole. Furthermore, women are more likely to not have any party identification.

Table 4 Predictors of party identification among US Colombians (logit models)
Full size table

Table 5 examines favorability of Obama, Clinton and Trump among US Colombians. As expected, partisanship is a very influential factor, with Colombian Americans who identify as Democrats expressing high levels of favorability for Obama and Clinton. Similarly, US Colombians who identify as Republican are more likely to have favorable views of Trump. Interestingly, compared to Republicans, Independents are more likely to express high levels of favorability for Obama and are more likely to express unfavorable views of Trump, but they are not more or less likely to view Clinton positively. Respondents who selected immigration as their most important issue have more positive views of Clinton and Obama, and the importance of healthcare was also associated with higher support for Clinton. Regarding other characteristics, there is evidence that higher levels of education are associated with a greater likelihood of viewing Clinton unfavorably.

Table 5 Predictors of favorability of sitting president and 2016 candidates among US Colombians (logit models)
Full size table

US Colombians and the nuance of the Latina/o/x vote

Through an examination of US Colombians and their political attitudes compared to those of other Latina/o/x groups, this article’s findings suggest that, although the “Latina/o/x vote” still represents common interests, there are important areas of heterogeneity within the umbrella category that warrant closer examination. Although socioeconomic status and views of trust toward the Colombian government would suggest that US Colombians are generally more conservative and would be less likely to be politically engaged in the US, our results show that they exhibit similarities with most other Latina/o/x groups on their political attitudes, indicating the influential nature of experiences in the United States. At the same time, US Colombians are most distinct from other Latina/o/x groups in their evaluation of presidential candidates. Whether these differences stem from the influence of their home country politics or their particular experiences upon arrival in the United States is unclear.

Colombian Americans exhibit various similarities on partisan affiliation with most other Latina/o/x groups, as they are more likely to identify as Democrats. Although we find evidence that US Colombians are more likely to be Republican in comparison to Mexicans, this effect is eliminated after including variables that consider demographic characteristics. This suggests that their middle position on the socioeconomic-status spectrum is partly responsible for their greater likelihood of identifying as Republican relative to groups who are from lower socioeconomic statuses. In addition, Colombian Americans are less likely to identify as Democrats relative to Dominicans, but more likely to do so in comparison to Cubans, even after considering socioeconomic characteristics. Cubans’ higher likelihood of being Republican is consistent with previous findings that compare them to groups such as Mexicans and Puerto Ricans (Garza 2004). Dominicans’ higher likelihood of identifying as Democrats may be due to their political socialization from their home country, where they are very likely to be involved in politics (Itzigsohn 2009). Upon arrival to the United States, Dominicans establish political organizations that not only work toward transnational goals, but also have helped galvanize Dominican support for the Democratic Party in places like Rhode Island (Itzigsohn 2009). On the other hand, research also shows that political parties are less likely to galvanize immigrant voters who are not already mobilized (Jones-Correa 1998), which may be a reason why US Colombians are less likely to identify with the Democratic Party relative to a highly mobilized group like Dominicans. This suggests that there are many commonalities within Latinas/os/xs, as they are mostly likely to identify as Democrats, despite key differences within the umbrella group. At the same time, the important sources of heterogeneity that exist within the Latina/o/x vote should not be overlooked when considering the preferences of this population.

When considering opinions of specific politicians and candidates, US Colombians exhibit many similarities with most other Latinas/os/xs on their ratings of Obama, but they view Clinton much less favorably. From examining Colombian Americans’ preferences of Clinton, it appears that those who view healthcare as their top issue are less likely to view Clinton favorably. However, US Colombians who view immigration as their top issue are more likely to view her favorably. There has not been much scholarly work on the specifics of why US Colombians may be less likely to favor Clinton relative to Obama compared to other Latinas/os/xs, especially since Clinton was secretary of state during the Obama administration. Although a higher likelihood of rejecting a woman candidate is a possible explanation for their lower support of Clinton, previous studies find that Latinas/os/xs from South America, in comparison to Latinas/os/xs from other countries (Fraga et al. 2010), are less likely to perceive men as more qualified to be political leaders. As such, more scholarly work should be conducted to examine the intricacies among different candidates within the Democratic Party, as it appears that not all Latinas/os/xs will support Democratic candidates equally.

However, the findings do not suggest that US Colombians were more likely to support Trump over Clinton in the 2016 election. Cubans view Trump more favorably than Colombian Americans, but this effect is eliminated once party identification is taken into consideration. South Americans are also more likely to view Trump favorably relative to US Colombians, which suggests differences within “new Latinas/os/xs” with regard to their voting preferences. This suggests that, despite the many similarities that emerge within the context of the Latina/o/x vote, nuances still exist within the category, as some groups are more likely to support some candidates over others even candidates within the same party.

Although the literature posits that US Colombians are not as politically engaged as other Latinas/os/xs because of their distrust of the Colombian government, this article finds that Colombian Americans are equally likely to vote as other Latina/o/x groups, even those groups that the literature argues are highly mobilized, such as Dominicans (Itzigsohn 2009; Guarnizo et al. 1999; Escobar 2004). This may be due to the important role that immigration played in the 2016 election (Gutierrez et al. 2019), which scholars have argued can be a strong mobilizing force for Colombian Americans (Guarnizo et al. 1999). Our results confirm this, as respondents who viewed immigration as their top issue were more likely to vote in the 2016 election. In addition, a substantial percentage of US Colombians (27%) identified immigration as their top issue, which was rated similarly high as the economy. With regard to the future of the Latina/o/x vote, the role of immigration may result in greater likelihoods of Latinas/os/xs turning out to the polls, as it is likely that this issue will remain salient in future elections.

Among all Latinas/os/xs, demographic characteristics predict partisan affiliation. Respondents who have higher socioeconomic statuses and are older are more likely to identify as Republicans. Similar results are evident when considering candidate favorability. These demographic differences echo patterns found in the literature, which argue that lower socioeconomic status is generally associated with support for Democratic candidates (Cain et al. 1991; Garza 2004). Respondents who are Spanish-dominant are also more likely to identify as Democrat and less likely to identify as Republican or Independent. This runs counter to the literature that suggests that less-acculturated Latinas/os/xs are more likely to be independent and that they do not identify with a party until later (Hajnal and Lee 2011). This suggests that Latinas/os/xs are socialized into US politics quickly, rather than adopting a more neutral position at the beginning. The findings in this article support other recent work that demonstrates that most Latinas/os/xs have crystalized attitudes about parties and their candidates (Sears et al. 2016).

Interestingly, Latinas/os/xs who indicated immigration as their top issue are less likely to have favorable views of both Clinton and Trump, but more positive views of Obama. As such, this article suggests that although Latinas/os/xs generally support Democratic candidates more on average, important nuances remain in candidate preferences within the Democratic Party. This will have important implications in the 2020 election and beyond, as the electoral power of the Latina/o/x vote will continue to increase in magnitude. Even if Latinas/os/xs continue to identify mainly as Democrats, these findings suggest that Latinas/os/xs will play an important role because of the increasing divisions between moderate and progressive wings of the Democratic Party and the candidates that represent these alignments.

Even though this article has disaggregated Latinas/os/xs into national-origin categories at levels not done by many quantitative researchers, we are still constrained by issues of sample size for some groups, which resulted in larger categories, such as “Central American” and “South American.” To fully understand how new Latinas/os/xs are shaping the Latina/o/x vote, more scholars should disaggregate Latinas/os/xs, as well as explore differences beyond national origin. Even within specific national-origin groups, differences such as class and race are bound to influence the political attitudes of each subgroup.

The findings in this article highlight that, although many similarities are apparent across Latina/o/x groups, a number of important differences also complicate our understanding of political preferences subsumed within the Latina/o/x umbrella category. This article offers new insights on US Colombians, which challenge many of the previous findings regarding their political socialization. Furthermore, our ability to make claims based on a nationally representative sample illuminates the experiences of a wide range of US Colombians beyond just a few cases in a restricted geographic location. By examining an understudied group like US Colombians, we seek to broaden the scope of Latina/o/x studies by including subgroups that are traditionally not seriously considered in empirical studies. The experiences of US Colombians, as highlighted in this article, represent almost one million people, the study of whom allows us to understand Latina/o/x communities in more complex ways. As this study is one of the first to examine the political preferences of a wide range of Latina/o/x groups, we expect that what the “Latina/o/x vote” exactly means will become clearer only through a more thorough investigation of different subgroups.