Abstract
In response to an increase in undocumented immigrants in the United States, several states have proposed and passed ballot measures that are intended to target immigrants, particularly Latinos. Yet, previous research has often failed to focus on Latino viewpoints in relation to these measures and the restrictions therein. As such, I utilize the 2006 Latino National Survey to examine how different sectors of the U.S. Latino population view policies that have been the focus of ballot initiatives targeting immigrants. I first make the distinction between native-born and foreign-born Latinos to determine whether opinions concerning proposed restrictions on immigrants differ between these groups. Furthermore, I examine whether foreign-born Latinos who are more assimilated into U.S. society differ in their attitudes on these measures than more recent immigrants. Results indicate a clear distinction with U.S.-born Latinos exhibiting more nativist views toward immigrants and restrictive measures than foreign-born Latinos.
In recent decades, there has been a noted increase in the number of undocumented immigrants entering and residing in the United States, particularly Mexican immigrants crossing the U.S.-Mexico border. As this population has grown, so has hostility toward immigrants along with demands for immigration reform (Binder, Polinard, & Wrinkle, 1997; Citrin, Green, Muste, & Wong, 1997; de la Garza, Falcon, Garcia, & Garcia, 1990; Lee, Ottati, & Hussain, 2001). Several state legislatures have addressed undocumented immigration by passing legislation that targets those in the United States without proper authorization. There have also been efforts to target these immigrants through ballot measures, particularly in border states. These initiatives, which are proposed by citizens of the state, have denied rights allowed to U.S. citizens to undocumented immigrants and, in some cases, have even impacted the lives of Latinos who are here legally.
Thus, this backlash from the sharp increase in immigration has broad implications for Latinos in the United States. Yet, research has not focused much attention on how Latinos view these ballot initiatives and the restrictions inherent therein. This, in turn, raises questions as to how a group as diverse as the Latino population may differ in their stances concerning nativist ballot initiatives, particularly native-born versus foreign-born Latinos. Previous research has demonstrated divergent views between these two groups concerning immigration with native-born Latinos demonstrating more nativist attitudes (Abrajano & Singh, 2009; Binder et al., 1997; Branton, 2007; de la Garza et al., 1990; Hood, Morris, & Shirkey, 1997; Newton, 2000; Sanchez, 2006). However, it is unclear whether these differences are seen in relation to ballot measures that are intended to target Latino immigrants. I, therefore, use the 2006 Latino National Survey (LNS) to examine how nativity and degree of assimilation may influence stances on policies that have been the focus of ballot measures targeting undocumented immigrants. These issues, which include bilingual education and English-only measures along with denying public services and in-state tuition to undocumented immigrants, reveal an important divide among Latinos based on nativity.
Ballot Measures Targeting Immigrants
Ballot propositions directed at those in the United States illegally are often seen in states with a large Latino immigrant population. These propositions such as California’s Proposition 187 are frequently approved with considerable support from voters. However, the question arises as to what leads an individual to support or oppose such measures. Several studies have stressed the differences that exist between immigrants and the native population of the United States in relation to race in particular along with the language spoken and citizenship status (King, 2000). These differences are argued to produce a prejudice against Latinos that is an important factor driving opposition to undocumented immigrants, particularly among whites (Citrin et al., 1997; Kinder, 2003). Closer proximity to the border and immigrants also appears to increase hostility toward this group (Branton, Dillingham, Dunaway, & Miller, 2007).
Beyond this prejudice, other studies have focused on the economic impact of immigration. For instance, research concerning immigration reform reveals that those who want to restrict immigration are often motivated by economics and the perception that immigrants will compete with them for jobs and place a burden on the government that will lead to increased taxation (Citrin et al., 1997; Polinard, Wrinkle, & de la Garza, 1984). Furthermore, those who have less education hold more restrictive views concerning immigration, which suggests that these individuals may feel that there is greater job competition from immigrants than those who are educated and likely have better employment opportunities (Citrin et al., 1997).
Yet, fewer studies have focused on the views of Latinos toward undocumented immigrants and policies directed at these individuals. Even so, it is clear that ethnicity is an important factor influencing attitudes toward immigration policy due to an in-group-out-group bias (Lee & Ottati, 2002). This in-group bias results in Latinos holding more favorable views of Latino immigrants, who are members of the same ethnic groups, than whites. Yet, native-born Latinos, who do not have a shared immigration experience, may take on a nativist stance in terms of seeing undocumented immigrants as outsiders (Abrajano & Alvarez, 2010; Branton, 2007). These more nativist views have implications for policies targeting undocumented immigrants. In fact, many Latinos assert that immigration should not be increased and that there should be stricter enforcement of immigration laws (Binder et al., 1997; de la Garza et al., 1990). Furthermore, a majority of Latinos surveyed felt that granting a pathway to citizenship for undocumented immigrants would reward illegal behavior and lead to an increase in undocumented immigration (Lopez, Taylor, Funk, & Gonzalez-Barrera, 2013).
Yet, a study by Michelson (2001) found that Mexican Americans felt that there was already too much enforcement along the U.S.-Mexico border. Furthermore, findings of the Pew Hispanic Center indicate that a majority of Latinos favor increasing legal immigration into the United States through temporary work visas and also emphasize the importance of relief from deportation for undocumented immigrants (Lopez et al., 2013). More importantly, the anti-immigrant bias exhibited in relation to immigration policy serves to heighten a sense of commonality among Latinos who are often targeted by these policies (Masuoka, 2006; Sanchez, 2006).
Even so, studies have found that foreign-born Latinos, particularly recent immigrants, have more positive attitudes toward immigration than native-born Latinos (Abrajano & Singh, 2009; Branton, 2007; Hood et al., 1997; Rouse, Wilkinson, & Garand, 2010). In addition, immigrants who are less assimilated into U.S. culture appear to be more supportive of policies that benefit undocumented immigrants than those who more recently immigrated (Abrajano & Singh, 2009; Branton, 2007; de la Garza, Falcon, Garcia, & Garcia, 1993). For instance, when it comes to extending public services to undocumented immigrants, foreign-born Mexican Americans supported this action, whereas native-born Mexican Americans were largely opposed (Polinard et al., 1984). Furthermore, English proficiency, which is tied to assimilation, along with citizenship status also appear to influence support for immigration policies, particularly immigration levels (Branton, 2007; Hood et al., 1997; Newton, 2000; Sanchez, 2006).
Latinos and Ballot Measures
As for ballot measures, the few studies conducted demonstrate that Latinos as a group tend to withhold support from measures that target immigrants. For example, California’s Proposition 187, which sought to deny public services to undocumented immigrants and to also require public officials to report those suspected of being undocumented, saw Latinos vote against the measure by a margin of over three to one (Chang, 2012). While economic conditions and partisanship influenced vote choice, ethnicity, citizenship status, and English fluency were key in support for this measure (Hero, 1998; Lee et al., 2001; Newton, 2000).
When it comes to English immersion initiatives, research on California’s Proposition 227, which ended bilingual education, found that only 37% of Latinos supported such action (Holland, 1998). Indeed, when it comes to policies impacting the usage of Spanish, there is a high level of support for both bilingual education and retaining Spanish (Abrajano & Alvarez, 2010; Hero, 1992; Uhlaner & Garcia, 2002). However, there is also evidence to suggest that native-born Latinos tend to demonstrate less support for bilingual education than their foreign-born counterparts, which indicates that this group may prefer English immersion for primarily Spanish-speaking students (Abrajano & Alvarez, 2010). Yet, Sanchez (2006) found that those born in the United States were more supportive of bilingual education than foreign-born Latinos. Furthermore, those who were citizens, English proficient, and longer residents of the United States were more likely to support bilingual education than their counterparts. Sanchez argues that this may indicate a desire for children of recent immigrants to learn English quickly due to the perceived benefits.
As for initiatives focused on the education of undocumented immigrants, early research demonstrated that Mexican Americans, particularly those who were foreign born, favored states educating children of undocumented immigrants (Polinard et al., 1984). Furthermore, recent efforts to deny in-state tuition to undocumented immigrants through legislation and ballot measures have led to opposition from many within the Latino community. In fact, a majority of Latinos indicate that they support in-state tuition for undocumented students, whereas less than half of the general population state the same (Lopez, Gonzalez-Barrera, & Motel, 2011).
In terms of English-only propositions, these initiatives arguably have a broader impact on the Latino community. Several states have considered or adopted measures making English the state’s official language. However, Hero (1992) pointed out that most Latinos did not support these measures. In fact, research on initiatives in California demonstrate that Latinos opposed making English the state’s official language by 61% to 39% (Citrin et al., 1990). Yet, for an earlier proposition in California that called for English-only ballots, a slight majority of Latinos, 52%, actually voted in favor of this restriction (Hajnal, Gerber, & Louch, 2002).
Thus, there appear to be some important divisions within the Latino population concerning measures that target immigrants. Indeed, questions remain as to how native and foreign-born Latinos vary in their support of these policies recently placed on ballots within several states. With that in mind, this study examines four policy issues—the denial of services to undocumented immigrants, bilingual education, denial of in-state tuition, and English-only measures—to determine where support may exist among Latinos for these propositions.
Hypotheses
Based on the literature above, there are three hypotheses that will be tested across the four issues. First, stemming from the literature demonstrating that native-born Latinos hold less favorable views of undocumented immigrants and show less support for policies benefiting this group, I hypothesize the following:
I also test two hypotheses focused on the foreign-born population. As the literature has indicated that Latinos who are more assimilated into U.S. culture demonstrate less support for immigrants, I expect that indicators of assimilation including length of time in the United States along with English proficiency may influence support for policies targeting undocumented immigrants. Accordingly, I hypothesize the following:
Data and Method
In order to test these hypotheses, I utilize the 2006 LNS, which includes interviews with both native-born and foreign-born Latinos. I then employ eight logit models focused on four issues that have been subjects of ballot propositions directed at undocumented immigrants. I also provide predicted probabilities to parse out how significant key variables are on support for some of the measures under analysis. Four of these models will examine any possible differences between native-born and foreign-born Latinos. The other four models will look solely at the foreign-born population and what within-group differences may exist based on how assimilated into American society a respondent may be. In all eight models, the dependent variable is whether or not the respondent supports the policy issue under consideration. Those who strongly supported or somewhat supported were coded as supporting the issue. All other responses were coded as not supporting the issue.
Among the policy issues considered are views concerning an upcoming actual English-only ballot proposition in Arizona, denial of services to undocumented immigrants in Arizona, English immersion after 1 year of bilingual education, and the denial of in-state tuition for undocumented students who graduated high school in the state. The English immersion and in-state tuition questions were asked of all respondents in the survey. However, questions concerning the denial of public services to undocumented immigrants and Arizona’s English-only ballot proposition were only asked of residents of Arizona.
As for the independent variables, the models focused on the entire Latino population utilize a dichotomous variable indicating whether respondents are born in the United States or not. For those models focused on foreign-born Latinos, I also look at how time spent in the United States, which is measured by years living in the United States, may influence support for these issues. I likewise examine how reported levels of English proficiency may be influencing support for the issues. English proficiency is a binary variable measured as either being fluent in English or not fluent. Length of time in the United States along with English proficiency were chosen because these factors are indicators of how assimilated an immigrant is into American society and may, thereby, help determine whether these individuals are holding more nativist viewpoints toward undocumented immigrants. I also control for naturalization status of the foreign-born population, as some of the foreign-born population are not U.S. citizens. Furthermore, in those models not focused on Arizona, I control for the state’s context by utilizing a variable that captures the percentage of Latinos residing in the states in 2006. This variable was added because more nativist attitudes may be influenced by the size of the Latino population. It is coded as three categories including a low percentage of Latinos or less than 15%, a moderate percentage of Latinos or 15% to less than 30%, and a high percentage of Latinos or 30% and above.
I likewise use a host of sociodemographic variables such as age, level of education, income, and sex. Age is a continuous variable measured in years and extends from 18 to 90. Level of education and income are both categorical variables with each category representing an increase in education level or income level for respondents. A respondent’s sex is the sole dichotomous variable among these demographic variables and is coded as being female or not.
Finally, I also include variables for partisanship with Republican, Democrat, and Nonpartisan being utilized. Nonpartisans, which includes almost half of Latinos in the survey, are coded as those who identified as Independents, with another third party, or did not affiliate with any party and are, therefore, nonidentifiers. Because Republicans have been shown to hold more restrictive views of immigration and immigrants than those who are not Republicans (Citrin et al., 1997), the reference category is comprised of Republicans.
Results
Looking over the models, a distinct pattern emerges between native-born and foreign-born Latinos. However, there are important distinctions found based on the specific issue under consideration. For the first two models, the issue focused on is an actual ballot proposition in Arizona that would make English the official language of the state. The model with all Latinos indicates that native-born Latinos were in fact significantly more likely at the p < .05 level to support making English the state’s official language than were foreign-born Latinos. Odds ratios demonstrate that native-born Latinos are 1.8 times, or 80%, more likely to support this initiative than foreign-born Latinos, all else equal. These findings are particularly noteworthy considering that the control variables of sex, age, education, and income did not reach significance in the model. The only other variable to reach significance in the model was Nonpartisans, who were significantly less likely than Republicans to support this proposition. This suggests that Latinos in Arizona who are nonpartisans view the issue of undocumented immigration differently than Latinos who identify with one of the major parties. However, this may point to the fact that almost half of Latinos in the survey did not identify as Democrats or Republicans, which is in keeping with partisanship among the Latino population (Hajnal & Lee, 2011; Jones, 2012).
Even so, the second model focused on foreign-born Latinos did not demonstrate any significance for partisanship variables. Differences in assimilation were also not found to influence support for this proposition. Furthermore, Latinos who had been in the United States longer and were proficient in English were not more likely to support making English the official language of Arizona than were more recent immigrants and those who were not English proficient (see Table 1).
Support for Arizona’s English-Only Measure Among Latinos Residing in Arizona.
Note. Standard errors are in parentheses.
p < .10. **p < .05. ***p < .01.
The third and fourth models focused on the likelihood of supporting initiatives that would deny public services to undocumented immigrants in Arizona. For the model with all Latinos, native-born Latinos were not significantly more likely to support denying services to undocumented immigrants than were foreign-born Latinos, which may point to the harshness of denying services as compared to a more symbolic issue like making English the official language of the state. Yet, older Latinos in Arizona were significantly more likely to support prohibiting services to these immigrants than younger Latinos, while Democrats were significantly less supportive of denying services than Republicans were, all else equal. This seems to indicate that a more conservative group of voters may be driving support for such initiatives among Latinos.
As for the model focused on foreign-born Latinos, the only significance found was with Democrats, who once again were significantly less likely to support the denial of services than Republicans. However, none of the assimilation variables reached significance in this model. Latinos who immigrated to Arizona at an earlier time and were more proficient in English were not more inclined to support the denial of services to undocumented immigrants than those who immigrated more recently and were not proficient in English (see Table 2).
Latino Support for Prohibiting Public Services to Undocumented Immigrants Among Latinos Residing in Arizona.
Note. Standard errors are in parentheses.
p < .10. **p < .05. ***p < .01.
The fifth and sixth models, which feature a policy question asked of the entire sample, demonstrated markedly different results. These models, which focus on English immersion in schools after 1 year of bilingual education, revealed significance for nativity and one of the assimilation variables. Specifically, the model encompassing both native-born and foreign-born Latinos revealed that native-born Latinos were significantly less likely to support ending bilingual education than foreign-born Latinos at the p < .01 level. The odds ratio of .764 for nativity indicates that the odds that a native-born Latino will support this measure is 23.6% less than for a foreign-born Latino. While this may seem counterintuitive, Sanchez (2006) found similar results, which implies that Latino immigrants see benefits in learning English as quickly as possible. Beyond this variable, older Latinos were significantly more likely to support English immersion than younger Latinos. However, Democrats and Nonpartisans were significantly less likely to support this measure than Republicans. The more educated were also significantly less likely to support English immersion than those who are less educated.
In the model focused solely on foreign-born Latinos, those who were English proficient were significantly more likely to support English immersion than those who were not proficient at the p < .05 level (see Table 3). The odds that those who were English proficient would support this measure is 1.48 times, or 48%, greater than for those who were not English proficient. However, time spent in the United States was not significant in the model indicating a lack of an assimilation effect. Even so, those who were more educated and wealthier were found to be significantly more likely to support ending bilingual education than those with less education and those who were poorer.
Latino Support for English Immersion in Schools.
Note. Standard errors are in parentheses.
p < .10. **p < .05. ***p < .01.
The last two models focused on support for denying in-state tuition to undocumented immigrants also demonstrated important distinctions. Once again, native-born Latinos were more supportive of denying in-state tuition to undocumented students at the p < .01 level than foreign-born Latinos. The odds ratio indicates that a native-born Latino is almost two times, or 95%, more likely to support this measure, all else equal. Wealthier and older Latinos were also significantly more likely to support this measure than poorer and younger Latinos. However, women and the more educated along with Democrats and Nonpartisans were significantly less likely to support this measure than men, the less educated, or Republicans.
As for the foreign-born model, time spent in the United States once again failed to reach significance. Yet, those who were English proficient were significantly more likely to support denying in-state tuition to undocumented students than those who were not English proficient at the p < .01 level (see Table 4). According to the odds ratio, those who are English proficient are 1.86 times, or 86%, more likely to support this policy than those who are not. Furthermore, Democrats were significantly less likely to support denying in-state tuition than Republicans. However, the more educated were significantly less likely to support this measure than those who had less education.
Latino Support for Denying In-State Tuition to Undocumented Immigrants.
Note. Standard errors are in parentheses.
p < .10. **p < .05. ***p < .01.
Discussion
Taken together, there was a statistically significant difference seen between native and foreign-born Latinos in every model except the denial of services to undocumented immigrants. Furthermore, time spent in the United States was not significant in any of the models, while English proficiency was only significant in the models concerning educational policies, which would include English immersion and the denial of in-state tuition to undocumented immigrants. This indicates that assimilation into U.S. society is not as influential in explaining support for these ballot propositions as was expected.
However, the strong significance for party identification and nativity for the last two policy issues warranted further analysis. As such, predicted probabilities were calculated for the policy issues involving English immersion and denial of in-state tuition to undocumented immigrants for the models including all Latinos. In doing so, it is evident that although partisanship variables are very influential in terms of support for these measures, nativity still plays an important part in whether an individual supports these policies.
In fact, there is a clear distinction between native-born and foreign-born Latinos on support for English immersion. Regardless of whether a foreign-born Latino is a Republican, Democrat, or Nonpartisan, they still demonstrated more support for English immersion than native-born Latinos of the same party identification. At the upper end, a foreign born, male, Republican of median age, education, and income would have a 53% probability of supporting English immersion. In contrast, a native born, male, Democrat of median age, education, and income had just a 38% probability of supporting English immersion. Thus, while Republicans were more likely to support this measure, about a 6% to 7% difference was seen between native-born and foreign-born Latinos across all three party identification categories.
A similar pattern is seen with predicted probabilities for support for denying in-state tuition to undocumented immigrants. Those who were native born demonstrated more support for this policy than foreign-born Latinos of the same party. Looking at male Latinos, a native-born Republican of median age, education, and income would have a 29% probability of supporting the policy. In contrast, a foreign-born, male, Democrat would have an 11% probability of supporting the denial of in-state tuition to undocumented students. There is also a distinct difference in terms of support for this measure between native-born and foreign-born Latinos regardless of party affiliation. As such, it is evident that although party is exerting a significant influence on support for these issues, nativity is a key factor driving support.
Indeed, there are statistically significant differences between native-born and foreign-born Latinos in these models, which supports the hypothesis. All else equal, those who are native born appear to exhibit more nativist views toward immigrants and policies that target these individuals than Latinos who are foreign born. As for those born outside the United States, there is limited evidence to suggest that those who are more assimilated into U.S. society take on more nativist viewpoints concerning policies directed at undocumented immigrants. Instead, they show a greater affinity with immigrants than is seen with the native-born population. Time spent in the United States was not statistically significant in any of the models and English proficiency only reached significance in relation to educational policies. Thus, it would seem that affinity with and support for immigrants is much stronger with foreign-born Latinos than with Latinos who are native born.
There are, nevertheless, some limitations with this research that should be pointed out. Specifically, only one question addressed an actual ballot proposition, and it was only focused on residents of Arizona. However, the other policies examined have been placed on ballots in several states. Length of time in the United States along with English proficiency were likewise not exact measures of assimilation but merely proxies. Even so, these factors are arguably the best indicators available to determine how immersed foreign-born Latinos are into American. society.
Furthermore, even with limitations, this study serves as an important starting point to examining how native-born and foreign-born Latinos differ in relation to initiatives that demonstrate an anti-Latino bias. Future research could also focus on how these subgroups may differ on other ballot initiatives that are not directed at immigrants. Perhaps there would likewise be differences by nativity on policy issues that impact those outside the group. Furthermore, it would be informative to examine whether Latinos in states with a history of initiatives targeting undocumented immigrants take a decidedly different view of these policies than Latinos in states that do not have the same history with initiatives. Perhaps the former group would be more hostile to these initiatives than Latinos who have not been repeatedly exposed to these kinds of initiatives. Thus, more research in this area is important to further our understanding of attitudes associated with this growing sector of the electorate in relation to ballot measures.
Footnotes
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.