Abstract
For undocumented immigrants, processes of integration are contingent on the qualities of their local context. A lack of legal status may require them to strategically manage their presence in order to avoid detection that could lead to deportation. The authors ask how the need to mask one’s legal status affects the civic integration of undocumented immigrants. Drawing primarily on data from the California Health Interview Survey, the authors estimate the probability of participation in voluntary associations for undocumented immigrants. They naturalized immigrants and find that undocumented immigrants exhibit a lower rate of participation but that this low rate of participation is unlikely the result of their legal status. The findings also show that undocumented immigrants are less likely to participate in voluntary associations if they live in counties where large shares of voters cast votes for Donald Trump in the 2016 presidential election.
The scholarship on immigrant integration suggests that a lack of legal status can affect the everyday behaviors of undocumented immigrants. On the one hand, some researchers find that living without legal status forces immigrants to avoid behaviors that are perceived to increase the risk of interacting with law enforcement, such as driving, applying for a new license, taking children to medical appointments, or attending church (Aranda and Vaquera 2015; Menjívar 2011). To the extent that undocumented immigrants limit behaviors that are commonplace for other members of their communities, integration might be inhibited. On the other hand, some researchers find that a lack of legal status catalyzes certain types of social behaviors, especially political participation (Burciaga and Martinez 2017). For example, undocumented immigrants may channel frustration about policies that hinder their inclusion into participation in political campaigns, labor movements, protests, or organizations aimed at shaping immigration policy (Avalos 2019; Bada, Fox, and Selee 2006; DeSipio 2011; Gonzales and Chavez 2012; Gordon 2005; Milkman 2000). These forms of political participation might hasten the pace of integration.
How is it that a lack of legal status can both inhibit and facilitate integration? One possibility is that legal status becomes salient only in certain contexts or that its effects are contingent on the characteristics of the places where immigrants live (Portes and Rumbaut 2001). Recent research suggests that in the United States, the effects of being undocumented may vary according to subnational immigration policies, the demographic composition of the host community, and the institutional resources available to immigrants (Ayón 2018; García and Schmalzbauer 2017; Portes and Rumbaut 2014; Ramakrishnan and Bloemraad 2008; Schildkraut et al. 2019; Wong and García 2016). In some contexts, undocumented immigrants may try to pass as American in order to avoid detection that could lead to deportation (García 2014). They may also manage their precarious status by developing a “hyper awareness” of the law (Menjívar 2011) and by pursuing political knowledge (García 2020) that can help them avoid unwanted attention. These strategies are used to calculate the risks of engaging in routine activities and to avoid exposure to law enforcement. Building on these findings, we ask how the need to strategically manage one’s legal status shapes civic integration. Specifically, we consider what role, if any, the context of reception plays in shaping the willingness and ability of undocumented immigrants to participate in local voluntary associations.
By focusing on voluntary associations, we expand our knowledge of civic integration in at least three ways. First, most studies of civic integration focus on political behaviors or how undocumented immigrants get involved in the political process by influencing voters and legislation (Alba and Foner 2015; Bloemraad 2006; Bloemraad, Sarabia, and Fillingim 2016; Fiorito and Nicholls 2016; Gast and Okamoto 2016). Less is known about the relationship between legal status and other forms of civic engagement such as participation in community-based organizations, nonprofits, religious groups, educational institutions, or other types of voluntary associations. Our focus on voluntary associations broadens the scope of civic integration and builds on the work of scholars who study civic engagement outside of politics (e.g., Ramakrishnan and Viramontes 2006; Terriquez 2012). Second, few studies compare rates of civic participation among immigrants with different legal statuses, because survey data on undocumented immigrants are sparse. The lack of comparisons limits our understanding of whether civic integration varies by legal status. Here we advance the scholarship on civic integration by using data that are well suited for identifying the legal statuses of immigrants. We analyze participation in voluntary associations by comparing rates of participation among immigrants with and without full legal status, and we consider whether this participation is contingent on the organizational, social, and political characteristics of the communities where these immigrants live. And finally, we build upon previous scholarship that focused on civic participation and immigrant rights protests in the early 2000s (Lai 2021; Voss and Bloemraad 2011) by analyzing the current political climate. Given the increased involvement of localities in immigrant politics (Steil and Vasi 2014), analyzing present contextual determinants provides a more comprehensive understanding of the effects of legal status and local contexts on civic participation.
Our analysis relies on several sources of data. We draw primarily upon the California Health Interview Survey (CHIS 2015-2016). The data from this survey allow us to identify immigrants living in California, their legal statuses, and their participation in voluntary associations. For data describing the contexts in which these individuals live, we merge data from the CHIS with data from the American Community Survey, the National Center for Charitable Statistics, and voting results from the 2016 presidential election. Our findings suggest that although undocumented immigrants participate in voluntary associations at lower rates than other groups of immigrants and nonmigrants living in California, their lower rate of participation is unlikely the result of their legal status. Our findings further suggest that undocumented immigrants may be less likely to participate in voluntary associations if they live in places where voters support candidates with anti-immigrant political agendas. We argue that the context of reception may shape the ability of undocumented immigrants to become civically engaged members of their communities but that a lack of legal status is not the primary determinant of their willingness or ability to get involved.
Legal Status and Civic Integration
In today’s polarized political climate, undocumented immigrants are commonly portrayed in the media as dangerous and undeserving of citizenship, which can lead to exploitation and other forms of harsh treatment (Massey 2007). From family separations at the U.S.-Mexico border to sweeping enforcement raids, undocumented immigrants are subjected to various forms of legal violence (Menjívar and Abrego 2012). This violence heightens fears of deportation (De Genova 2002), and it shapes the ways in which undocumented immigrants go about their daily activities. 1
For example, some undocumented immigrants choose not to ride in the same car as their spouses because of the fear of leaving their children behind if both parents are deported (Menjívar and Abrego 2012). Others cease driving altogether because they are concerned about being racially profiled (Aranda and Vaquera 2015). The lack of legal status and the corresponding fear of legal violence also shape larger decisions, such as where to live. Some Hispanic undocumented immigrants feel safer living in neighborhoods where the majority of the population is also Hispanic, and others choose to live in predominantly Black neighborhoods because it makes them feel invisible to law enforcement (Asad and Rosen 2019). There is also evidence that legal violence extends into the workplace. Undocumented immigrants may choose not to report workplace abuses out of fear that their employers might contact Immigration and Customs Enforcement (Gleeson 2010). If a lack of legal status forces undocumented immigrants to avoid behaviors that might unwittingly place them in the public eye, then it might also inhibit various forms of civic engagement that require them to be present in public spaces or to interact with strangers.
Yet despite the risks for legal violence, researchers find evidence that undocumented immigrants actively engage in civic affairs. For instance, Abrego (2011) showed how 1.5-generation immigrants were motivated by the stigma surrounding illegality to voice their concerns, speak about their contributions to society, and participate in collective political activities. This might be due in part to their subjective identities as U.S.-raised children who hold feelings of invisibility and deep-rooted ideals of being culturally American (Aranda, Vaquera, and Sousa-Rodriguez 2015; Negrón-Gonzales 2014). Another motivation for undocumented youth participation is prior engagement in youth organizations or activist groups (Terriquez 2017). Undocumented youth use civil rights discourse and cultural and symbolic capital to make claims for full incorporation into the nation-state or to demand full legal rights (Aranda et al. 2015; Nicholls 2013). For instance, undocumented youth organized and participated in the Coming Out of the Shadows Week campaign, in which students shared stories and revealed their undocumented status in front of crowds and media representatives (Preston 2011). Beyond these types of claims-making activities, undocumented immigrants also engage in acts of civil disobedience. At the height of the debate over Arizona’s SB 1070, undocumented youth staged a sit-in at Senator John McCain’s Arizona office, calling for an end to the criminalization of immigrants and advocating for the passage of the DREAM (Development, Relief, and Education for Alien Minors) Act (Preston 2010). A few months later, undocumented students protested against Georgia’s HB 87, a bill with similarly restrictive measures as Arizona’s SB 1070.
Civic participation among undocumented immigrants is not limited to youth. For example, Varsanyi (2005) showed that undocumented immigrants in Los Angeles, despite an inability to cast formal ballots, attended campaign rallies and participated in “get out the vote” drives, which were intended to persuade registered voters to support candidates and legislation that advanced the needs of the undocumented community. Similarly, Gordon (2005) reported on the ways undocumented immigrants in New York mobilized to demand fair wages and safe working conditions, ultimately winning a 30 percent raise for day laborers and a domestic workers’ bill of rights. These forms of organization do not necessarily need to be in defense of social justice initiatives; in Prieto’s (2018) study, he identified “instrumental activists” who mobilize to “secure the quotidian makings of a dignified and decent life” (p. 32). And even when they are not actively participating in public acts of civic engagement, undocumented immigrants educate themselves on political processes and issues in order to strategically navigate their environments (García 2020). These examples demonstrate how a lack of legal status may encourage civic engagement, and they stand in contrast to findings suggesting that lacking legal status causes undocumented immigrants to withdraw from public life in order to avoid law enforcement.
Building on this body of research, we ask whether being undocumented encourages or inhibits participation in local voluntary associations. Because the answer to this question might be contingent on the places where undocumented immigrants live, we consider several ways in which the context of reception might shape the willingness and ability of undocumented immigrants to get involved in local associational life.
The Context of Reception
Several qualities of the local contexts where undocumented immigrants live may shape civic engagement. The first is social capital. Social capital takes on a number of meanings, but most scholars agree that it refers to the resources that inhere in relationships rather than individuals (Coleman 1988; Portes and Sensenbrenner 1993; Putnam 1993). Social capital among local residents, particularly among coethnics, can help immigrants find jobs or learn useful skills for thriving in a new society (Portes 1998). It is also useful for accessing material support and information (Portes and Manning 1986; Zhou 1992). One of the central mechanisms facilitating the flow of resources in local support networks is trust (Stack 1974). For undocumented immigrants, trust in one’s neighbors might be particularly important for learning about, and getting involved in, local voluntary associations. They might also be less concerned about hiding their legal status if they trust the members of their communities who might participate in the same associations.
A second quality of the context of reception that may shape civic engagement is ethnic diversity. Neighborhoods where social capital is highest might be those that exhibit a similar set of structural characteristics such as ethnic homogeneity (Sampson, Morenoff, and Earls 1999; Sampson, Raudenbush, and Earls 1997). Scholars have long suggested that ethnic enclaves ease the transition of settling in a new place because high levels of social capital help newcomers quickly find jobs (e.g., Portes and Manning 1986; Wilson and Portes 1980). Enclaves may also facilitate integration, for example, by helping immigrants acquire the resources necessary to buy a home in a middle-class neighborhood, helping immigrants amass the resources necessary for providing their children with access to higher quality educational institutions (Zhou 1992), and serving as conduits for information about naturalization (Abascal 2017). Beyond a shared culture, a common language among coethnics is especially important in helping immigrants with employment prospects (Le 2000).
Although coethnics can serve as crucial support networks for newly arrived migrants, there is evidence that reciprocity in these networks can break down in the context of a weakened labor market, when the capacity of local charities and organizations is diminished, and when government policies do not provide opportunities for migrants to work in the formal economy (Menjívar 2000). Ethnic networks can be a hindrance during the integration process, as individuals may be subject to misinformation and exploitation from their community (Rosales 2020). Moreover, the networks themselves may be too bounded in nature so as to offer individuals meaningful opportunities outside the ethnic community. Nonetheless, most studies of ethnic enclaves and support networks suggest that they ease the transition into a new society. If this is true, then they might also lower barriers to participation in voluntary associations that an immigrant might otherwise face when living among others who speak different languages, share fewer experiences in common, or are less likely to provide support.
A third source of variation in the context of reception is the concentration of nonprofit organizations that might provide volunteering opportunities. Local nonprofits may be among the first organizations immigrants encounter when they arrive to the United States. As such, community-based organizations are crucial institutions that help immigrants bridge cultural gaps in understanding a host community’s institutions, norms, processes, and programs. For example, large hometown associations with nonprofit status can provide access to established networks while also facilitating integration by directly providing services such as language classes (Somerville, Durana, and Terrazas 2008). Access to nonprofits may also be crucial for immigrants as they try to create change in the communities where they live. Nonprofit organizations can educate immigrants on their eligibility for certain public services and encourage immigrants to advocate for themselves if the organizations providing critical services are threatened with funding cuts (de Graauw 2008). Nonprofit organizations are especially critical during election seasons in getting immigrants to register to vote and offering other volunteering opportunities for immigrants who are not eligible to vote. Nonprofits may also be adept at providing linguistically appropriate services, while also emphasizing the importance of English proficiency that is needed for engaging with public officials (de Graauw 2008). Such organizations may include neighborhood associations, educational groups, faith-based groups, ethnic organizations, and hometown associations (Ramakrishnan and Viramontes 2010). Organizations in which immigrants can speak their first language may better communicate information about opportunities to get involved in community activities, provide resources, and foster trust. Therefore, undocumented immigrants may have more opportunities to volunteer if they live in a community with a high density of nonprofits or community-based organizations.
A fourth source of variation in the context of reception that might shape civic engagement is the political climate. Whether undocumented immigrants participate in local voluntary associations might be contingent on the political attitudes of the people living around them, as the political climate can significantly shape whether they feel included in their new communities (Portes and Rumbaut 2014). Participation may also be contingent on the level of enforcement of restrictive immigration policies, which itself may be a reflection of the political climate. García’s (2014) research suggests that although national-level policies are typically restrictive, there is greater variation in policies and enforcement at the local level. In localities with more restrictive immigration policies and enforcement, civic participation may decline. For example, in the face of increased enforcement activity in North Carolina, Hagan, Rodriguez, and Castro (2011) identified a decline in immigrants’ participation at community events and willingness to visit public places, with some immigrants responding to the increased enforcement by returning to their home countries. Therefore, in places with a restrictive political climate, undocumented immigrants may be less likely to participate in voluntary associations.
Data
Our analysis considers how participation in voluntary associations varies by legal status and the qualities of the context of reception described above. We analyze participation in voluntary associations located in California where undocumented immigrants make up 6 percent of the state’s total population and are 25 percent of all immigrants living in the state (Hayes and Hill 2017). California is one of five states that is majority minority (Panzar 2015), and with the rest of the country projected to reach majority-minority status by 2044 (Frey 2014), lessons about how the racial and ethnic composition of California’s communities affect civic participation could be indicative of trends occurring throughout the country. Within California, opportunities for civic engagement vary from one city to the next (Pastor et al. 2012). Despite being known as an “integration paragon” (Rodríguez 2018) and enacting a statewide sanctuary law (SB 54) in 2018, there is variation in the protections that California’s cities and counties provide to immigrants (Hayes and Hill 2017). In sum, California’s political, racial, and ethnic diversity make it a suitable place to study civic integration.
We draw on pooled data from the 2015–2016 CHIS to explain the relationship between legal status, the context of reception, and participation in voluntary associations. The CHIS is a random-dial telephone survey of Californians that reports on the health and well-being of the state’s residents. The survey also includes a series of questions on civic engagement. With replicate weights that adjust for the complex survey design, the data are representative of California’s population. The confidential files of the CHIS are particularly useful for our purposes because they provide sufficient information for identifying immigrants, their legal status, and the census tracts where they live. We identify undocumented immigrants as those immigrants who have not naturalized, who do not hold green cards, and who do not hold valid documents that allow a temporary stay in the United States, such as a visa.
We use the following CHIS survey question to construct a dichotomous dependent variable: “In the past 12 months, have you done any volunteer work or community service that you have not been paid for?” (1 = yes). Our dependent variable shares similarities with those used in other studies of immigrant volunteerism and civic engagement. For instance, Lai (2021:206) collapsed multiple categories of volunteerism to create a measure of “general civic participation,” and Ishizawa (2015:269) relied on a measure of time spent “volunteering or performing community service.” Although our dependent variable is similarly broad, a notable difference is that we do not collapse multiple questions on civic participation and volunteering to create it; instead, we rely on a general measure of volunteerism provided by the CHIS. One limitation of using data from the CHIS is that we cannot determine the specific types of volunteer work respondents have completed (e.g., participation in a neighborhood group vs. participation in a religious organization), nor can we determine whether a respondent participated in formal or informal associations. The latter problem is one that other scholars have encountered when analyzing survey data on immigrant volunteerism (e.g., see Ramakrishnan and Viramontes’s [2006] analysis of the Current Population Survey), and it may be relevant to our study if respondents in the CHIS underreport volunteer work that occurs outside of formal organizations. This may be particularly important if undocumented immigrants are more likely to participate in volunteer opportunities that do not require formal membership with an organization.
We also use the CHIS to construct control variables for age, gender, race, ethnicity, education, household income, employment status, marital status, household composition, housing tenure, time in residence (years), and English-speaking ability (1 = none/not well, 2 = well/very well, 3 = only English). Several independent variables of interest come from the CHIS data as well. These indicate whether immigrants are naturalized citizens, green card holders, undocumented, or living in the United States with some form of temporary status such as a work visa. Finally, the CHIS data provide us with data for constructing a measure of social capital. Because social capital cannot be observed directly, we use data describing trust among neighbors to understand the role of social capital in the places where immigrants live. Specifically, we use responses to the following question to construct an ordinal measure of trust: “People in this neighborhood can be trusted” (1 = strongly disagree, 2 = disagree, 3 = agree, 4 = strongly agree).
The remaining independent variables come from sources of data that we merge with the CHIS. First, we identify the share of coethnics in a respondent’s neighborhood by drawing on the five-year 2012–2016 American Community Survey’s estimates of race and ethnicity by census tract (U.S. Census Bureau 2012–2016). Although census categories of race and ethnicity are broad, they are the best measures available for estimating the number of coethnics in an immigrant’s neighborhood or, in other words, for operationalizing a defining characteristic of ethnic enclaves. Second, we use data from the National Center for Charitable Statistics to create a count variable of community-serving nonprofit organizations by ZIP codes. These are nonprofits classified under the National Taxonomy of Exempt Entities as “community improvement and capacity building” nonprofits. 2 Third, we use voting results from the 2016 presidential election, reported by California’s secretary of state, to estimate the political climate or the political attitudes of residents in counties where the survey respondents live (Figure 1). This variable indicates the share of voters in each of California’s counties who voted for the 2016 Republican presidential candidate (see Okamoto and Ebert 2010 for a similar use of voting results as a proxy for the political context). An important limitation of using counties as the unit of analysis for measuring the political context is that it may not capture local variation in the political attitudes of California’s residents. However, counties are the smallest geographies at which we can operationalize the political climate given the available data.
Share of voters who voted for Donald Trump in the 2016 presidential election by county.
We provide weighted descriptive statistics in Table 1. We divide our sample into five groups according to migration status and legal status: (1) nonmigrants; (2) naturalized immigrants; (3) permanent residents; (4) immigrants with some form of temporary status, such as a valid visa; and (5) undocumented immigrants. Nonmigrants in our sample exhibit the highest rate of participation in voluntary associations (46 percent), followed by naturalized immigrants (30 percent) and immigrants with temporary status (30 percent), permanent residents (21 percent), and undocumented immigrants (17 percent). Table 1 also displays differences in the places where the survey respondents live. Compared with all other groups, undocumented immigrants are less likely to live in places where they trust their neighbors and more likely to live among coethnics. Undocumented immigrants in our sample live in counties where an average of 31 percent of voters cast a vote for the 2016 Republican presidential candidate, and they live in ZIP codes with an average of six community-serving nonprofit organizations.
Descriptive Statistics.
Analytic Strategy and Hypotheses
Our strategy is to model participation in voluntary associations by fitting a logistic curve to our weighted data. We specify several models and use them to estimate the probability of participation for undocumented immigrants and naturalized immigrants. 3 We draw comparisons between undocumented immigrants and naturalized immigrants to test our hypotheses about how legal status shapes participation in voluntary associations. Naturalized immigrants are a suitable comparison group because, in contrast to undocumented immigrants, they have more secure access to the rights guaranteed by the state. To isolate the relationship between legal status and participation in voluntary associations, we control for many of the other ways in which undocumented immigrants might systematically differ from naturalized immigrants such as educational attainment, household income, household composition, housing tenure, and fluency in English.
On the basis of prior research describing the need for undocumented immigrants to “pass” in the host community (García 2014) and their “hyper awareness” of the law when conducting routine activities (Menjívar 2011), we expect their participation in voluntary associations to be relatively low compared with naturalized immigrants, who have full legal status.
Hypothesis 1: Undocumented immigrants are less likely to participate in voluntary associations than naturalized immigrants, all else equal.
Undocumented immigrants could be less likely to participate in voluntary associations if they live in contexts that inhibit participation. The places where undocumented immigrants live could inhibit participation in at least four ways. First, if undocumented immigrants live in contexts in which they are less likely to trust their neighbors, this lack of social capital could account for differences in participation across the two groups. Second, if undocumented immigrants are less likely to live among coethnics who might lower the barriers to participation (e.g., by providing information about opportunities to get involved) then they might be less likely to join voluntary associations. Third, if undocumented immigrants have less access to volunteer opportunities because there are fewer voluntary associations in their communities, such as community-serving nonprofit organizations, then this could explain differences in the probability of participation as well. Fourth, differences in the probability of participation could result from differences in the political context. If undocumented immigrants tend to live in places where attitudes and laws toward immigration are more restrictive and less welcoming, then we might observe less participation in voluntary associations. We expect the gap in the probability of participation between undocumented and naturalized immigrants to shrink after controlling for differences in the context of reception.
Hypothesis 2: Differences in the likelihood of participation between undocumented and naturalized immigrants will attenuate after controlling for the social, organizational, and political qualities of the context of reception.
Finally, we expect undocumented immigrants to be particularly sensitive to the political climates of their communities given that immigration generally, and unauthorized immigration in particular, was at the center of debates leading up to the 2016 presidential election (Nowicki 2015), which overlaps with the period of time when the CHIS survey was administered. Because the Republican presidential candidate, Donald Trump, embraced an overtly anti-immigrant platform, we would expect undocumented immigrants to be less likely to participate in voluntary associations in places where support for Donald Trump was widespread. We test this hypothesis using an interaction term that allows us to compare participation among undocumented immigrants living in counties with different levels of support for Donald Trump.
Hypothesis 3: Undocumented immigrants are less likely to participate in voluntary associations if they live in counties where large shares of voters supported the 2016 Republican presidential candidate.
Findings
We report our findings from the logistic regressions in Table 2. The first model of participation in voluntary associations includes individual- and household-level controls. We find evidence that participation is significantly and positively associated with education, household income, homeownership, the presence of children in the household, and being female. We find a significant and negative association between participation and age, and we find that Hispanic and Asian respondents are less likely to participate in voluntary associations than white respondents, all else equal. Most important for our hypotheses, we find a positive and significant relationship between fluency in English and participation in voluntary associations. In other words, we are more likely to observe participation among those who are more comfortable speaking English, net of all controls. This suggests that language barriers may play a role in shaping one’s ability to participate in at least some voluntary associations.
Participation in Voluntary Associations.
Note: Household income is measured in dollars, and values in parentheses are jackknife standard errors. All models are weighted to account for the California Health Interview Survey 2015–2016 survey design.
p < .05 and **p < .01 (two-tailed tests).
In the second model, we include controls for legal status and test our first hypothesis. Here we find that the coefficient for fluency in English is no longer statistically different from zero. We also find evidence that participation in voluntary associations varies by legal status. Immigrants who are naturalized citizens and permanent residents are significantly less likely to participate in voluntary associations compared with nonmigrants, net of all controls. We find no evidence that undocumented immigrants are significantly more or less likely to participate in voluntary associations than nonmigrants. We use the coefficients in model 2 to estimate the probability of participation for naturalized immigrants and undocumented immigrants, and we display the results in Figure 2. We estimate these probabilities assuming the characteristics of the average immigrant in our sample: this is an individual who is female, Hispanic, employed, married, and speaks English well or very well. This individual is also a renter who lives in a household that earns a yearly income of roughly $56,000. 4 In Figure 2, we report the predicted probabilities that this individual participates in voluntary associations given two different legal statuses: undocumented status and naturalized status.
Participation in voluntary associations by legal status.
As displayed in Figure 2, we estimate the probability of participation for this individual to be greater with undocumented status (24.6 percent) than with naturalized status (22.0 percent). Although our estimates suggest that, all else equal, undocumented immigrants are more likely to participate in voluntary associations than naturalized immigrants, the overlapping confidence intervals indicate that we cannot be confident that the difference between these two estimates is not actually zero. Therefore, Figure 2 suggests that hypothesis 1 is not supported; there is no evidence that undocumented immigrants are less likely to participate in voluntary association than naturalized immigrants. Although prior research suggests that undocumented immigrants make efforts to mask their legal status and manage their public self-presentation, we find no evidence that a lack of legal status inhibits participation in voluntary associations.
In model 3, we control for the contextual determinants of participation, including trust in one’s neighbors, the share of coethnics in one’s census tract, the concentration of nonprofits in one’s ZIP code, and the political climate of one’s county. All of the coefficients for these variables are positively signed, but none are significantly greater than zero. This suggests that the probability of participation may not significantly differ across the different contexts considered here. In model 3, we again find evidence that immigrants who are naturalized citizens and permanent residents are significantly less likely to participate in voluntary associations than nonmigrants, all else equal. We also find no significant difference in the probability of participation between undocumented immigrants and nonmigrants.
We use model 3 to estimate probabilities of participation for undocumented and naturalized immigrants. As in Figure 2, our point estimates continue to suggest that undocumented immigrants might be more likely to participate in voluntary associations than naturalized immigrants, all else equal. For instance, in Figures 3 to 5, we display predicted probabilities of participation for undocumented and naturalized immigrants given changes in context. Our estimates suggest that as trust increases, as the share of coethnics increases, and as the number of nonprofits increases, so too does the likelihood of participation for both undocumented and naturalized immigrants. In each case we estimate the probability of participation to be higher for undocumented immigrants. However, the overlapping confidence intervals in each figure indicate that we cannot be confident that the probability of participation is any different between immigrants who are undocumented and those who are naturalized citizens. We therefore find no evidence to support hypothesis 2: there is no significant difference in the probability that naturalized and undocumented immigrants participate in voluntary associations and thus no evidence that controlling for qualities of the local context attenuates a gap in the probability of participation.
Participation in voluntary associations by legal status and trust in neighbors.
Participation in voluntary associations by legal status and share of coethnics.
Participation in voluntary associations by legal status and nonprofits.
In model 4, we introduce interaction terms that allow us to compare participation for immigrants with different legal statuses living in counties where varying shares of voters cast votes for Donald Trump in the 2016 presidential election. The coefficient for the interaction between undocumented status and the share voting for Donald Trump is significantly less than zero. This suggests that as county-level support for Donald Trump increases, undocumented immigrants are less likely to participate in voluntary associations. The interaction between naturalized status and the share voting for Donald Trump is negatively signed but not statistically significant, suggesting that naturalized immigrants may not be any more or less likely to participate in voluntary associations given residency in counties with different levels of support for the Republican candidate.
In Figure 6, we plot predicted probabilities of participation for undocumented and naturalized immigrants assuming residency in counties with increasing levels of support for Donald Trump. The overlapping confidence intervals indicate that undocumented immigrants may not be any more or less likely to participate in voluntary associations than naturalized immigrants. However, we do find significant differences in the probability of participation when comparing estimates of participation among undocumented immigrants living in counties with different levels of support for Donald Trump. For instance, we estimate the probability of participation to be 37 percent for an undocumented immigrant living in a county such as San Francisco, where approximately 9 percent of voters cast votes for Donald Trump in the 2016 election. In contrast, we estimate the probability of participation to be 9 percent for an undocumented immigrant living in a place like Lassen County, where approximately 72 percent of voters cast votes for Donald Trump. This difference in context creates a gap of 28 percentage points in the likelihood of participation. The nonoverlapping confidence intervals surrounding the two estimates above indicate a significant difference in the likelihood of participation across political contexts, providing evidence to support hypothesis 3.
Participation in voluntary associations by legal status and political context.
Conclusion
Although undocumented immigrants in California participate in voluntary associations at lower rates than other groups of immigrants and nonmigrants, we find no evidence to suggest that this gap in participation is the result of differences in legal status. While holding constant individual- and household-level attributes that might shape one’s willingness and ability to participate in voluntary associations, we find no significant difference in the probability of participation between undocumented immigrants and either naturalized immigrants or nonmigrants living in California. We also find little evidence to suggest that differences in the likelihood of participation between undocumented immigrants and naturalized immigrants result from contextual effects such as the presence of local nonprofits, the concentration of coethnics in one’s neighborhood, trust among neighbors, or the political climates of counties. Instead, variation in rates of participation between undocumented immigrants and all other Californians may result from differences other than legal status. It may be that undocumented immigrants less often participate in voluntary associations because they have fewer opportunities, or less ability, to do so. For example, fewer opportunities could result from having to work more hours to make ends meet, having less access to transportation that might be necessary for meeting with others, or having less information about voluntary associations due to language barriers.
We do find evidence that among undocumented immigrants, some contexts may facilitate participation in voluntary associations more than others. Specifically, our findings suggest that undocumented immigrants are less likely to participate if they live in counties that voted heavily for Donald Trump in the 2016 presidential election, which tended to be inland counties and those in the northeastern corner of the state (see Figure 1). These are California’s less densely populated areas, with smaller shares of immigrants. These findings do not necessarily mean that undocumented immigrants are disengaged from politics. Indeed, the Public Policy Institute of California found that nearly half (49 percent) of California’s immigrants report paying more attention to politics since the 2016 presidential election (Baldassare et al. 2018), which is in line with García’s (2020) findings that undocumented immigrants pursue political knowledge in order to keep themselves and their loved ones out of danger. Nonparticipation may itself be a strategic response to a hostile political context (Prieto 2018). These findings may mean that, to the extent undocumented immigrants living in these counties perceived widespread support for Donald Trump—in campaign flyers, posters, speeches, rallies, and so on—and connected those campaign efforts to Donald Trump’s anti-immigrant platform, they may have been deterred from participating in associations with other community members whose political views and views on immigration were unknown. It is also possible that counties where Donald Trump captured a larger share of the vote might be those with greater enforcement of restrictive immigration policies, which could increase the need for undocumented immigrants to mask their legal status by avoiding interactions with others. 5
Studying the context of reception in California might seem counterintuitive. Since the early 2000s, the state has been considered a defender of immigrant rights and a vanguard in the immigrant rights movement. The state has passed policies allowing driver’s licenses and in-state tuition for undocumented immigrants, expanded Medi-Cal to cover individuals under the age of 26 regardless of their immigration status, and established a statewide director of immigrant integration to oversee and lead integration efforts in California. Progressive statewide efforts, however, do not always trickle down to the local level. When California passed SB 54, a statewide sanctuary law, at least 14 cities and two counties passed resolutions opposing the legislation. This contradiction between state and local policies leads undocumented immigrants to be less trustful in the efficacy of sanctuary policies (Wong, Shklyan, and Silva forthcoming). Similarly, our findings describe nuances in the context of reception in California, particularly at the county and local levels. As California is considered a bellwether for other states (FitzGerald and Skrentny 2021), it will be important for future scholarship to consider local dynamics in other states when assessing integration outcomes.
Our finding regarding the association between support for Donald Trump and lower levels of civic participation among undocumented immigrants speaks to a broader literature that links the political context to various immigrant experiences and integration pathways. The political climate in which immigrants live can affect homeownership rates (Allen and Ishizawa 2015), graduation patterns (Filindra, Blanding, and Coll 2011), and naturalization rates (Cort 2012), among other markers of integration. Even more “subjective” experiences, such as reports of discrimination (Ebert and Ovink 2014) and fear of deportation (Amuedo-Dorantes, Puttitanun, and Martinez-Donate 2013) are affected by the political climate. In sum, there is an abundance of evidence to suggest that policies and the political context trickle down to affect the integration experience. Our study adds to this body of scholarship by demonstrating a new way in which the political climate shapes the integration process. Our findings suggest that participation in voluntary associations may hinge, in part, on the political climate in which undocumented immigrants live. We would therefore expect integration to occur more slowly in places where anti-immigrant sentiment was politically salient.
Our findings also contribute to a growing body of scholarship on the relationship between legal status and the context of reception. First, our findings suggest that a lack of legal rights may not inhibit all forms of civic integration. This finding is consistent with recent studies showing that undocumented immigrants participate in the political process despite an inability to vote (e.g., Alba and Foner 2015; Bloemraad 2006; Bloemraad et al. 2016; Fiorito and Nicholls 2016; Gast and Okamoto 2016). Second, the findings advance our understanding of the context of reception. Scholars recognize that integration processes unfold unevenly as a result of variation in the places where immigrants settle. For example, we know that subnational immigration policies, varying levels of enforcement of immigration policies, different geographic landscapes, and varying levels of access to institutional resources can all shape the experience of settling in a new place (Ayón 2018; García and Schmalzbauer 2017; Portes and Rumbaut 2014; Ramakrishnan and Bloemraad 2008; Schildkraut et al. 2019; Wong and García 2016). We contribute to these studies by testing a series of hypotheses about the contextual determinants of participation in voluntary associations. Our findings suggest that for undocumented immigrants, the political climate in which they live may shape their willingness to engage, particularly during times when immigration policy is a salient topic of public debate and infused with overtly xenophobic rhetoric.
Finally, our findings advance the literature by broadening our understanding of civic integration to include participation in voluntary associations. The civic integration scholarship is typically located within research on political participation such as signing petitions, participating in protests, voting in elections, and becoming a naturalized citizen (for example, see DeSipio 2011). When viewed solely through the lens of political participation, we may fail to perceive some of the ways in which other forms of participation can serve as stepping-stones toward integration.
The limitations of our study pave the way for future research. Our findings were unable to capture variation in subnational immigration policies. Scholars have demonstrated some of the complicated ways in which federal, state, and local policies shape the daily activities of immigrants. Many of these policies and their effects are difficult to quantify, especially at the local level. Much of the difficulty is that data describing the effects of local policies are not always publicly available and may not be easily aggregated at smaller units of analysis, such as at the level of neighborhoods or counties. Also, we were unable to observe how the context of reception changes over time. Researchers have shown that voluntary organizations and civic engagement in the United States have evolved over the past century, in that immigrants are integrating into a society in which civic engagement is, on average, declining (Putnam 2000; Skocpol 2003). Therefore, future work might focus on the ways in which integration is shaped by broader trends of civic engagement in U.S. society. These shortcomings of our work provide opportunities for innovative research on immigration federalism and integration.
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the California Immigration Research Initiative at the University of California San Diego’s Center for Comparative Immigration Studies.
