Department of Political Science
San Antonio, TX 78212
firstname.lastname@example.org James A. McCann
Department of Political Science
West Lafayette, IN 47907
The Brian Lamb School of Communication
West Lafayette, IN 47907
Abstract: Immigrants in the United States frequently express a desire to return to their country of birth, an attitude that could undermine political acculturation and democratic inclusion in the American context. This mindset has been labeled an “ideology of return.” Drawing from original surveys of the Mexican-born population conducted during the 2006 and 2008 elections, we show that partisan competition during major national campaigns has the potential to pull immigrants towards U.S. civic life and make the prospect of remaining in the country over the long-term a more attractive option.
Prepared for presentation at the annual meeting of the Western Political Science Association, San Antonio, TX, April 21-23, 2011.
Approximately one-eighth of the current U.S. resident population is foreign-born, a nearly unprecedented proportion that stands in marked contrast to the 1960s and 1970s, when immigrants numbered fewer than one in twenty (Batalova and Terrazas 2010). As the foreign-born population has grown, social scientists have devoted increasing attention to questions of inclusion in American society and public affairs. One recurring theme that emerges in field interviews and opinion surveys is the ambivalence that immigrants often feel about living in the United States. Many of the foreign-born carry with them a deep sense of nostalgia and admiration for their country of birth, work hard to keep the country’s distinctive traditions alive in their households and communities, and hope one day to return to it (Arthur 2000; Jones-Correa 1998a, 1998b; Massey and Sánchez 2010; Portes and Rumbaut 2006; Wampler, Chávez and Pedraza 2009; Wong 2006). As one author put it, it is common for immigrants to refer to the United States as “this country,” and the country from which they came as “my land” or “my country” (Duany 1994, referenced in Jones-Correa 1998a, 91). Such a mindset, which has been labeled an “ideology of return,” could depress naturalization rates and limit political engagement in the U.S. (Jones-Correa 1998a; DeSipio 2006; McCann, Cornelius and Leal 2009; Pantoja 2005).
In this paper, we consider the effects of political parties on this “ideology of return.” Can the signals sent by the Democrats and Republicans during contemporary national election campaigns cause immigrants to update their views of the U.S. and make the prospect of remaining in the country over the long-term more desirable? Can partisan mobilization influence attitudes that are closely tied to residential preferences, such as general feelings towards the American people and U.S. governing institutions, beliefs regarding personal competence and influence in the American context, and a sense of immigrant group consciousness and solidarity? Are immigrants who are less acquainted with the United States – relative newcomers, the non-naturalized, and those who live outside of traditional settlement areas – more responsive to major party campaign stimuli?
When examining the institutions that pull immigrants towards American civic life and encourage them to put down roots, scholars give great weight to a variety of nonpartisan organizations – labor unions and worker centers, social service agencies, ethnic associations, and churches, among others (e.g., Andersen 2008a and 2008b; Andersen and Cohen 2005; Bada, Fox, and Selee 2006; de Graauw 2008; Hamlin 2008; Jones-Correa 2005; Verba et al. 1995; Wong 2006; Wong, Rim and Perez 2008). The two major parties figure far less prominently in these accounts. The powerful urban machines that selectively incorporated immigrant groups through the distribution of goods and services in exchange for support during the first half of the twentieth century have all but disappeared. In this post-reform era, parties seek first and foremost to serve the short-term interests of office-seekers. This means targeting established constituencies during electoral campaigns, typically through short candidate-centered advertisements and photo-opportunities, rather than cultivating personal ties with immigrants, many of whom are ineligible to vote or do not participate regularly for other reasons. “The replacement of the patronage system with the merit system,” write Scott and Hrebenar (1984, 15-16; quoted in Wong 2006, 58), has “reduced the parties’ opportunities to function as socializers. . . . Are any of today’s immigrants introduced to American politics and political traditions through the medium of the Democratic or Republican parties?”
While the historic movement away from patronage-based party politics is beyond dispute, and for most observers would be a sign of health in American democracy, we do not see the question that Scott and Hrebenar raise as purely rhetorical. Even in a political environment where the foreign-born are not the primary subjects of partisan mobilization, and a good many lack formal standing to take part, it is still possible for immigrants to be exposed to much of the fanfare of campaign events and appeals during major national elections in the United States. Drawing from an original panel survey of the Mexican-born population over the course of the 2006 midterm elections and a randomized survey experiment conducted within this same population in September 2008, we demonstrate that election-year political messages have the potential to counteract the “ideology of return.” In this regard, the party system of the United States may continue to function, at least after a fashion, as a conduit for immigrant inclusion.
Much is known about the factors that are responsible for the recent surge in migration from the lesser developed nations of Latin America, Asia, and Africa to the United States and other industrialized democracies. For many migrants, economic calculations, specifically the expected difference in wages and opportunities in the sending versus the host country, carry the most weight in the decision to relocate. Quality of life matters a great deal as well, with crime, social unrest, or political violence driving many to seek refuge and stability abroad. Transnational family connections and social networks further facilitate movement from one country to another; it is far easier to contemplate leaving one’s native land if family members and friends have previously done so (Castles and Miller 2003; Massey and Sánchez 2010; Portes and Rumbaut 2006).
Substantially less research has been conducted on what causes migrants to return to the country of birth once they have settled abroad. Even less is known about the desire to repatriate in the first place, a preference that has been likened to an ideology because it surfaces often and forcefully in discussions with immigrants. The limited work in this area stresses economic considerations. Immigrants who are dissatisfied with their employment status or anxious about long-term job prospects, for example, are more likely than the financially secure to wish to return, even if their current living arrangements would not permit this (Waldorf 1995). Socio-psychological factors, however, might also play an important part in shaping views on repatriation among the foreign-born. Empirical investigations based on interviews with immigrants in the United States point to three interrelated types of attitudinal variables that shape residential preferences. The first of these is the degree of affect felt towards Americans. Quite understandably, immigrants who hold the U.S.-born in high esteem and are more trusting of Americans tend to be less inclined to want to repatriate at some point in the future (Massey and Sánchez 2010; Wampler, Chávez, and Pedraza 2009).
Perceptions of status and efficacy in the United States might also have an effect on long-term residential preferences. Aguilera (2004) finds, for example, in an examination of Mexican migrants who originally entered the U.S. without papers that self-described proficiency in English is a major predictor of planning to stay in the country through retirement. More generally, DeSipio (2006) reports that Latino immigrants who believe that they have more influence in the United States vis-à-vis the country of origin are much more likely to wish to remain rather than repatriate (see also Jones-Correa 1998b).
In addition to these beliefs, group consciousness and identities have been found to bear on dispositions towards staying or returning. Immigrants typically identify with their native country; such self-designations may remain salient long after settlement in the United States. This does not prevent the foreign-born, however, from developing an understanding of their distinctive social position and group interests in the American context, a recognition that is an important step in civic acculturation. Massey and Sánchez (2010, Chapter 1) note that becoming an “immigrant” in this sense – as opposed to a “sojourner” or “temporary guest” – could result, somewhat counterintuitively, in identifying less as an “American” even as familiarity with the U.S. and the desire to remain in the country grow.
Since most of the foreign-born living in the United States today spent their formative childhood years outside of the country, they may not possess the repertoire of fundamental core beliefs about American politics and civic involvement that native citizens have (Abrajano and Alvarez 2010; Easton and Dennis 1969; Hajnal and Lee 2011; Krosnick 1991; Sears 1983). This could make the political attitudes of immigrants less stable compared to the American-born, and more responsive to current events and to signals sent from mainstream society and governing institutions (Massey and Sánchez 2010). In the United States and other established democracies, such signals are most pronounced during major national election campaigns. In these periods, parties conduct an internal dialogue to define their positions and strategies, and then turn their attention to reaching out to the country. Whether a particular party is successful or not in persuading voters to back its candidates, a great deal of research on the effects of the contemporary electoral process itself shows that spirited conflict and mass mobilization can pay democratic dividends within the mass public.
As parties seek to sell their programs and attract followers, they also implicitly reinforce a sense of political legitimacy, civic competence and duty, interpersonal trust, and group attachments – all essential ingredients in the maintenance of a democratic system (see, e.g., Anderson and Paskeviciute 2009; Banducci and Karp 2003; Clarke and Acock 1989; Conover 1988; Finkel 1985; Kam 2007; Mansbridge 1999; Rahn et al. 1999; Valentino 2001; Verba et al. 1995). These effects are not necessarily limited to those who turn out to vote or work in a political campaign. Rahn et al. (1999) show, for instance, that Americans who are merely “psychologically involved” in electoral politics – that is, they are attentive to campaigns – become more trusting of their fellow citizens and feel more efficacious.
Expanding on this work, electoral politics in the United States might likewise make long-term settlement a more attractive proposition for that segment of the public for whom living in another country is a realistic option. Messages of democratic inclusion, solidarity, and representation during a campaign could cause immigrants to see the U.S. in a more positive light, feel more confident in their ability to make a difference in the country, and become more aware of particular group identities – all of which should lessen the desire to repatriate and lead to more substantive democratic engagement within the U.S. Of course, this could only happen if the foreign-born are inclined to follow public affairs and be exposed to campaign stimuli. Recent large-scale surveys of Latin American and Asian immigrants suggest that this is generally the case (Fraga et al. 2007; Junn et al. 2008; McCann, Cornelius and Leal 2009). To illustrate, in the 2006 Latino National Survey, over 80 percent of the foreign-born respondents stated that they watch television news “daily” or “most days,” as opposed to “once or twice a week” or “never”; over two-thirds of these immigrants were “somewhat” or “very” interested in politics. It appears safe to assume that immigrants in the United States develop some awareness of American campaign politics each time a major national election takes place.2
In the following section, we probe the impact of electoral outreach and advertising more deeply with a focus on the Mexican-born population. Mexicans are by far the largest group of immigrants in the United States; approximately one-third of the current foreign-born population emigrated from Mexico. Given the country’s close proximity to the U.S. and the considerable differences between the two nations, a number of authors, notably Samuel P. Huntington (2004), see the ultimate incorporation of Mexican immigrants into American civic life as particularly challenging. If the ebb and flow of partisan competition in the United States has a salutary effect on long-term residential preferences for this population, such concerns might well become less pressing.
As explained in greater detail below, we pay particularly close attention to potential variations in campaign effects, since the Mexican immigrant population is strikingly diverse with respect to length of time in the United States, place of residence, and naturalization status (Massey 2008). Much research on public opinion in the U.S. and elsewhere demonstrates that attitudes across a wide array of domains (e.g., partisanship, self-positioning along a left-right ideological continuum, and political values) become more stable as individuals gain experience in public affairs. In keeping with this literature, it could be the case that Mexicans who have lived in the U.S. for many years, have settled in locations that are rich in Mexican-American culture and civic organizations (e.g., Texas or southern California), or have become naturalized American citizens, a long process that requires a great deal of commitment, have a relatively well-developed attitude towards staying or repatriating. Campaign communication could have far less of an effect for these immigrants compared to relative newcomers, noncitizens, or Mexicans who live outside of traditional settlement destinations and would not be exposed as regularly to political stimuli.
Research Design and Findings
The potential effects of American political campaigns on immigrants’ long-term residential preferences are examined through two surveys. The first bracketed the historic midterm elections of 2006, which brought about a transition in the U.S. Congress from Republican to Democratic control. The second was fielded in September of 2008, when the political climate was quite different but equally historic.
In June of 2006, telephone interviews were conducted with 753 Mexican immigrants across three sampling sites, Dallas, San Diego, and north-central Indiana, including Indianapolis but excluding the Chicago region. Respondents were recruited randomly through records obtained from a well established marketing research firm specializing in the Latino community.3 Immediately after the November elections, as many immigrants as possible were contacted for another interview (N = 264).4 Panel data such as these allow researchers to track changes in attitudes with a great deal of precision (Bartels 2006; Finkel 1995).
The three sampling areas were selected to maximize variation in settlement areas (King, Keohane and Verba 1995). Dallas and San Diego are major traditional destinations for migrants, with a current combined Mexican-born population of over one million (Batalova 2008). North-central Indiana is typical of “new” settlement destinations for Mexicans and other immigrant groups. Between 2000 and 2004, the number of Indiana-based Mexicans rose by approximately 60,000. Out of all metropolitan areas in the United States, Indianapolis had the fifth-highest rate of Latino population growth during this period (Menéndez Alarcón and Novak 2010; Sagamore Institute for Policy Research 2006). While the number of Mexicans now living in Indiana is much smaller than in California or Texas, this rapid expansion of immigrant communities was unprecedented in the Hoosier State.
The item used to gauge long-term residential preferences was comparable to what other scholars have employed (e.g., DeSipio 2006; Wampler, Chávez, and Pedraza 2009; Waldorf 1995): “Do you want to return to Mexico permanently to live one day, or are you planning to remain in the United States?” Table 1 reports the frequency of responses in each of the survey waves, with study participants in the three regions pooled.5 In June, two out of three immigrants stated a preference for returning. This percentage varied somewhat across sampling regions, with Mexicans in Indiana being more inclined to repatriate. In each location, however, a solid majority wished to return.
[Table 1 about here]
When interviewed five months later, the thought of remaining in the U.S. had become more attractive. In this instance, the pooled sample was nearly divided in half, with only 56 percent now wishing to repatriate. As noted at the bottom of this table, this increase is not likely to have come about by chance; a paired-sample t-test is highly statistically significant. At the individual-level, responses to this item changed a fair amount. The test-retest correlation is moderately-sized at .35, underscoring that for Mexican immigrants, there is a measure of uncertainty or flexibility in long-term residential preferences. The overall gain in desire to remain suggests that the signals and outreach of contemporary electoral campaigns have the potential to pull immigrants – many of whom may see themselves as living “in between” two countries (Jones-Correa 1998a) – more closely towards the United States.
Immigrants with less experience living in the U.S., those who have not become American citizens, or those who have settled in regions nontraditional destinations such as Indiana may have been particularly sensitive to campaign effects. The regression analysis in the lower part of Table 1 examines these differences. In this model, residential preferences as measured in the post-election wave (dummy-coded, with 1 = “remain permanently”) are regressed on preferences from the June wave plus sampling region, naturalization status (1 = citizen), and years spent in the United States.6 By including the lagged dependent variable as a predictor, we are modeling changes in settlement preferences (Finkel 1995). The findings from this analysis indicate that the rise in desire to remain in the country did not vary with respect to sampling location, length of time in the U.S., and formal civic status. If the dynamics of the midterm elections were indeed responsible for making the United States a more attractive option for long-term settlement, even immigrants with a great deal of prior exposure to American politics and culture were affected at the same rate.
Other explanations for this change in attitude are certainly possible, however. One plausible rival interpretation is that the turn towards wishing to live permanently in the United States between June and November reflects a particular type of transnational demobilization effect in the Mexican context. The data collection for the first survey wave took place on the eve of the July 2 presidential election in Mexico. Even though Mexican political parties are legally barred from canvassing voters outside of the country and the costs of casting an international absentee ballot in this election were inordinately high, many of the immigrants in the study closely followed this election and were aware of the main themes of the campaigns (McCann, Cornelius, and Leal 2009). Large numbers engaged in informal participation in this context, such as encouraging friends and family in Mexico to turn out to vote. By the time of the November survey wave, mass electoral mobilization within Mexico had long since ceased.7 This dynamic, more than the fanfare and pull of American electoral politics, could have tilted immigrants away from Mexico.
To delve more deeply into the effects of U.S. campaigns within this population, and to assess whether changes in residential preferences coincide with changes in attitudes in the three closely related domains discussed above – general feelings towards Americans, beliefs about one’s ability to succeed in the United States, and a sense of group consciousness -- we conducted a follow-up study during the American presidential election two years later.
As in 2006, randomly selected Mexican immigrants were interviewed by telephone. Nearly all surveys again were in Spanish. In total, 1,023 respondents took part in the study, which commenced after the Labor Day weekend, the traditional start of fall presidential campaigns. Sampling this time took place in two regions rather than three: north-central Indiana, again excluding the Chicago area, and San Antonio, TX. The reduction in survey sites raised the number of respondents in each area (N = 501 in Indiana and 522 in San Antonio), which serves to increase the statistical power for regional comparisons. San Antonio was chosen in place of Dallas because its Mexican population is larger and somewhat better established.9 The city has long been recognized as a leading center of Mexican-American culture. For this reason, candidates wishing to show their commitment to the Mexican community frequently campaign in San Antonio.10 These features offer a more advantageous contrast with the emerging Mexican immigrant population of Indiana.
Researchers investigating the impact of campaign stimuli on political attitudes and perceptions via conventional survey methods frequently wrestle with thorny issues concerning measurement error and causal inference. Respondents may be unable to state how many commercials they have seen or heard (Ansolabehere, Iyengar and Simon 1999). Furthermore, even if they can accurately report exposure, evaluations of the candidates or identification with a party may prompt individuals to watch particular television programs or listen to radio stations where exposure to political information is likely to occur. In the 2006 panel study, long-term residential preferences for immigrants were seemingly prone to updating, and partisan mobilization signals could have exerted strong effects on these attitudes. Here we put a finer point on this claim.
A few minutes into the interviews, survey participants were randomly divided into three groups. Approximately half were chosen to listen either to an ad for Barack Obama (N = 262, Treatment Group 1), or an ad for John McCain (N = 250, Treatment Group 2). These were actual ads produced for Spanish-language radio stations. The third group (N = 511) served as the control. For the treatment groups, exposure was prefaced by the remark, “Recently political parties in the U.S. began their presidential campaigns. I am going to play a brief ad from Senator [Obama / McCain]. You may have already heard it on the radio. After listening to it, I will have a couple questions for you about the ad.” Following exposure, respondents were queried on whether they remembered hearing such a commercial before, whether they believe advertisements influence voters in general, and whether the ad gave a positive or negative impression of the candidate. The lead-in to the sound clip and questions afterwards were intended to supply a plausible rationale for the treatments.11
Only a very small number of subjects in each treatment group stated that the advertisement left a bad impression. Overwhelmingly, the ads were seen as showing the candidate in a positive light. This was our intention. In selecting political spots for the treatments, the goal was to expose respondents to that most ubiquitous form of campaign communication in the Latino media market: the upbeat personal endorsement from someone who is a part of that ethnic community.12 Wordings for the Obama and McCain ads are given in Table 2. Each was played to the randomly selected respondents in its entirety.13
[Table 2 about here]
If short candidate-centered partisan messages implicitly draw immigrants closer to American society over the course of a campaign cycle, we should see evidence of this effect in attitudes that were gauged several minutes after subjects in the treatment groups listened to an ad from Obama or McCain. Using the same instrumentation as in the 2006 study, all immigrants stated towards the end of the interview whether they wish to return to Mexico one day or remain in the U.S. for the rest of their lives. Subjects also offered summary affective ratings of the American people, U.S. governing institutions, and Mexicans; judgments concerning their personal competence and efficacy in the American context; and responses to several items pertaining to “immigrant group consciousness.” All of these attitudes are elements in an “ideology of return.” If the residential preferences that study participants express are more than top-of-the-head responses – that is, if they are indicative of a coherent mindset – responses across these multiple domains should move in sync following exposure to campaign stimuli. The following wordings were used to capture the different attitude elements:
Affective Ratings. On a scale of 0 to 10, where 0 means that your opinion is very bad and 10 means that it is very good, where would you place Americans? The U.S. government in Washington, DC? Democrats? Republicans? Mexicans?
Agree or disagree: American politics is so complicated that I sometimes do not understand it well.
Do you feel that your English is good, so-so, or not so good?
Immigrant Group Consciousness.
Have you been badly treated in the U.S. because of being an immigrant? If so, where was this? Finding employment? Arranging housing? Seeking medical attention? Shopping? Just going about personal business in the street?
Which of these words would you use to describe yourself? American? Mexican-American? Hispanic? Latino?
How well do the two parties represent the interests of Latinos? Average American voters? Large corporations? Immigrants?
The zero to ten-point scale for affective ratings is comparable to those that have been used in public opinion surveys in Mexico (e.g., Lawson et al. 2007).14 The first item to tap into personal competence is adapted from the American National Election Study series; it is a decades-old measure of personal political efficacy. The second would be relevant only within immigrant populations. Language skills are a key marker of social incorporation in the United States and an important determinant of status among the foreign-born. Appraisal of language ability, however, is inherently inexact. Skills can fluctuate over time and depend on one’s general level of confidence in a given context. If political campaigns can instill a sense of accomplishment and empowerment (Clarke and Acock 1989; Finkel 1985; Mansbridge 1999), these effects for this population could extend beyond political efficacy to judgments about nonpolitical skills.
The concept of group consciousness appears frequently in the scholarly literature on African-American participation and has been fruitfully applied to a wide range of ethnic groups (Sanchez 2006). The three sets of items in this category are derived from this expansive literature: interpreting the treatment one receives in the course of daily activities as a consequence of membership in a specific social category, namely “immigrant”; reluctance to identify oneself as “American,” but instead to prefer labels that are tied to ethnicity; and evaluating the major parties based on how well they represent immigrants, as opposed to Latinos, average American voters, and business corporations.
Table 3 shows the impact of exposure to one of the campaign ads on long-term residential preferences. Two effects are apparent; listening to either advertisement reduced uncertainty about living in the United States versus returning to Mexico, and increased the attractiveness of remaining in the U.S. In the control group, 47 percent stated a preference for remaining. This figure rises eight points for immigrants in the “Obama ad” group and 13 for those exposed to the McCain commercial. This change is highly significant (p < .01) and comports with the trend from 2006. The kinds of mobilization messages that are aired during American campaigns can orient Mexican immigrants towards life in the United States. There is little difference in effect between the two ads. Outreach for the Republicans counted as much as messages from the Democrats.15
[Table 3 about here]
The campaign ads further affected summary evaluations of the two major parties, Americans, and the U.S. government. In the case of the parties, a distinction emerges. Listening to the Obama advertisement significantly improved assessments of the Democrats, even though the party was not specifically mentioned in the commercial. The McCain ad did not have a noteworthy impact on ratings of the Democrats, but it did cause immigrants to give Republicans somewhat more positive ratings relative to both the “Obama ad” and the control groups (p = .088). Respondents in the two treatment groups, however, evaluated both “Americans” and the “U.S. government” more highly compared to those who had not heard an ad (p < .05 in each case). On the other hand, exposure to an ad had no appreciable impact on affect towards “Mexicans” as a group.
These shifts in opinions further underscore the potential for campaign communication to contribute to the social incorporation of immigrants, even if the goal of the candidates is first and foremost to move an audience towards one party in particular. Changes in the two efficacy items are presented in the lower part of Table 3. In the case of political efficacy, the ad treatments did not significantly alter appraisals. For respondents in all experimental groups, the consensus was that American politics can sometimes be too complicated to understand. Such a breakdown in responses typically surfaces in ANES surveys of the mass public. Exposure to the McCain advertisement boosted political efficacy to some extent, but compared to the control group, this is not a significant change. On the other hand, there is some evidence that after hearing one of the campaign commercials, immigrants felt more confident in their English language skills, a key indicator of perceived status in the United States (p = .088). As with the zero to ten-point evaluation scales and the question of whether to remain in the country, the McCain and Obama ads had comparable effects.
The impacts of the commercials on the multiple indicators of immigrant group consciousness are presented in Table 4. Perceptions of group-based mistreatment were tallied and coded into three categories (no discrimination, fewer than three types of mistreatment, or more than five types). The pattern that emerges after breaking down this tally by experimental treatments demonstrates that political outreach to the Latino community can cause the foreign-born to interpret social interactions in a way that reinforces identification as an immigrant. At the same time, the campaign messages significantly decreased the odds of identifying as an “American.” In the control group, 36 percent designated themselves as such, as opposed to 28 percent in both the Obama and McCain treatment groups (p < .01). No similar reductions emerged for other labels (Mexican-American, Latino, and Hispanic). The results for these latter terms imply that ethnic group attachments are more strongly felt among the foreign-born and are less bound to particular contexts, a point that has been raised in previous work on immigrant identities (e.g., Jones-Correa 1998a).
[Table 4 about here]
The regression analyses in the lower half of the table speak to another facet of campaign effects, the priming of group-based considerations when expressing summary evaluations of the two major parties (Iyengar and Kinder 1987). As shown in Table 3, study participants who listened to an advertisement updated their general impressions of the parties to a significant degree, with those in the “Obama” treatment group liking the Democrats more and those in the “McCain” group rating Republicans more highly. These evaluations were differenced (i.e., the rating for Republicans was subtracted from that for the Democrats) and regressed on four items that touched on group representation within the party system: whether the Democratic Party, relative to the Republican Party, represents the interests of “Latinos,” “average American voters,” “business corporations,” and “immigrants in the United States.” Each judgment was measured on a three-point scale (-1 = only the Republican Party represents the particular group, +1 = only the Democratic Party represents the group, 0 = all other responses).
In the “no ad” control group, the most relevant constituency groups for respondents when evaluating the parties were “Latinos” and “average American voters.” Both regression slopes are highly significant (p < .01) and run in the expected direction: perceptions that the Democratic Party is more representative than the Republican Party are linked to higher general evaluations of the Democrats relative to the Republicans. The impact of perceptions of “immigrant” representation is also positive, and the coefficient for “business corporations” has a negative sign. These signs would have been expected, but neither slope is large or statistically significant. Turning to the two treatment groups, a much different pattern appears. Whether subjects listened to an ad for Obama or McCain, perceptions of how well immigrants were represented took on much greater weight and became the largest coefficient out of the set.16
Taken together, the findings in Tables 3 and 4 provide a framework for interpreting the change in long-term settlement preferences over the course of the 2006 campaign. Conventional political outreach during elections can shape basic orientations towards U.S. society and sharpen group attachments that are relevant in the American context, with the ultimate effect of making immigrants feel more at home in the country. For each of the items in these two tables, subsequent models were estimated to examine variations in causal outcomes: advertising treatments by sampling region, by length of stay in the United States, or by naturalization status. In no case did significant interactions appear.17 As in the panel survey two years earlier, it appears that Mexican immigrants are sensitive to the cues and signals from partisan elites even if they have resided in the country for many years, live in an environment that is rich in bicultural civic organizations and Latino leaders, or have become American citizens.