Introduction

Deportation is arguably one of the more conflicted and contentious aspects of contemporary United States policy discourse. Public preferences for strict enforcement of immigration laws and increased control of our borders are offset by humanitarian concerns for immigrants. Likewise, concern for immigrant competition for jobs and fear of depressed wages are countered by the need for immigrant labour and talent to fuel our national economy. These competing perspectives about immigration are further shaded by fears resulting from the post-9/11 threat of terrorism and the spectre of international crime, especially the illegal drug trade. The policy context is so challenging that comprehensive immigration reform has become difficult. President George W. Bush was unsuccessful with his legislative efforts for immigration reform, and President Barack Obama has not pursued legislative approaches, but has instead used executive directives and selective enforcement of existing immigration laws. With so many differing and competing viewpoints toward immigration, an understanding of public policy preferences and the underlying influences shaping them becomes important for policy formation.

This study provides additional information about public opinion toward the deportation of Hispanics and Latinos as a component of US immigration policy, which is important to research for several reasons. First, Hispanic and Latino immigrants, as the largest and most visible immigrant groups, have become the ‘face of immigration’ and the centre of the debate for reform of US immigration policy. Consequently, an understanding of public support for immigration policies is rooted in an understanding of US perceptions about the merits of both documented and undocumented Hispanic immigration. Second, there is surprisingly little research on the perceptions about, and preferences for, deportation as a policy option, which is perhaps the most controversial and contentious aspect of immigration reform.

In this research note, we hope to contribute to this discussion through the investigation of public opinion in a time of economic stress in the US. A recent national survey of attitudes and opinions concerning the topic of deportation was undertaken by the PEW Research Center, with its results first published in 2006. At the time of that survey, it is expected that anti-immigrant sentiment and perceived economic competition among ethnic groups were noticeably lower than one might expect in our current economic climate. Given that expectation, the Social Climate Survey of Hispanic Immigration (SCSHI) was conducted in February and March of 2009, consisting of a nationally representative sample of 1,510 respondents. The current investigation into the SCSHI is aimed at our interest in the effect of ethnic prejudice and perceptions of economic competition on public support for the deportation of undocumented Hispanic immigrants. Improved knowledge about these relationships engenders an understanding of the public opinion psychology integral to immigration policy formation.

Defining the policy question

Extant public opinion surveys often ask a question similar to: ‘Do you think immigration to the US should be: increased, left the same, or decreased?’ as a proxy for policy preferences about immigration (e.g., Citrin et al. 1997). This question provides a measure of whether people desire more or less immigration, but fails to directly assess specific policy options that would achieve their immigration goal. In our survey, we used questions that more closely approximated actual policy options. For example, ‘Do you think that undocumented Hispanic or Latino immigrants should be: required to return home, granted temporary worker status, or allowed to stay permanently?’ was asked to obtain public opinion measures for three distinct policy options: (1) deportation, (2) temporary worker programs, and (3) amnesty programs. We explicitly identified undocumented immigrants as a focus of the study, because the contemporary policy debate centres on issues related to undocumented or unauthorized immigration. Furthermore, our study emphasized undocumented immigrants, because the undocumented status is a ‘flashpoint’ issue that can elicit highly emotional opposition (Huber and Espenshade 1997). Our study is based on the supposition that the determinants of public opinion supporting or opposing undocumented immigration are critical for policy formulation.

To date, the most recent examination of attitudes concerning support for deportation in the US was undertaken by the Pew Hispanic Center, before the downward trend of the American economy.Footnote 1 Our initial examination of the SCSHI results revealed a substantial departure from the pattern of public opinion found in prior surveys. Previously, deportation had been a policy option that was endorsed by a relatively small percentage of the US population. For example, in the Pew Hispanic Center survey conducted in 2006, it was reported that 27 % of respondents preferred undocumented immigrants to be ‘required to return home’ (Pew Hispanic Center 2006, p. 20). In the SCSHI questionnaire, however, the same question was asked with the results depicting support for enforced deportation almost doubling to 51 %. The shift in support of deportation was broad-based; the percentage of respondents preferring deportation increased in every political category, jumping from a minority amongst all parties and ideologies to a majority in nearly every grouping (Pew Hispanic Center 2006).

While there is currently a lack of research specifically focusing on public opinion supporting or opposing the deportation of undocumented immigrants, the magnitude of the issue and the importance of the topic to US policy makers and rights advocates have never been more publicly relevant. To that point, a change in US immigration policy toward aggressively pursuing deportation is by no means a trivial issue. There are profound demographic, economic, and social implications of attempting to forcibly remove such a large and important part of the population. The Pew Hispanic Center estimates that in 2008 there were approximately 11.9 million unauthorized immigrants in the US, of whom 76 %, or slightly more than 9 million, are Hispanic (Passel and Cohn 2009, i). This is an incredibly large population to target for deportation, especially when it is considered that they are so widely dispersed geographically and integrated into so many aspects of our society. For example, unauthorized immigrants are estimated to constitute 4 % of the population and 5.4 % of the work force. Children of undocumented immigrants, both those who are themselves unauthorized and those who are US citizens, make up 6.8 % of the student population in primary and secondary schools (Passel and Cohn 2009, i). Clearly, the size of the unauthorized immigrant population makes a comprehensive deportation policy a massive, nationwide undertaking. To what degree such an undertaking is supported and to what degree factional different opinions exist remains to be settled. What we know from our preliminary analysis is that attitudes and beliefs about deportation are probably quite shallow and tend to ebb and flow along with the strength of the US economy. Ultimately, we expect that these shifts in support for deportation are most likely to be driven by a cross-sectional measurement of ethnic prejudice and perceived economic competition, which we expect to have risen during the US’s most recent economic downturn.

Method

Data

The SCSHI is a national public opinion survey (N = 1,510) that elicited responses about beliefs, attitudes, and policy preferences about Hispanic immigration. The survey was conducted in February and March of 2009 by computer-assisted telephone interviewing techniques (CATI) and enhanced Random Digit Dialing (RDD) sampling. Whenever possible, the survey instruments included questions that had been asked in previous national opinion surveys (e.g., the Pew Hispanic Center, the Gallop Organization). The CASRO response rate for the survey was 53.8 %, and the co-operation rate for the survey was 81.7 %, and only 3.8 % explicitly refused to participate. To address unrepresentative bias, the survey data were weighted according to the 2008 US Census figures on ethnicity, sex, age, and education. The sampling error for the total data set (binomial response option with 50/50 split) is no larger than ± 2.5 % (95 % confidence level).

Measures

Public support for deportation is measured as a discrete categorical indicator; respondents were asked their opinion on what actions should be taken in the handling of illegal immigrants. The respondents were provided with three potential responses: (1) deportation, (2) allow them to stay temporarily on a work permit, and (3) allow them to stay permanently. Figure 1 shows that 51 % of the respondents of the nationally representative SCHIS supported the deportation option in early 2009. In direct comparison, the Pew Hispanic Center reported that only 27 % of their sample supported the same option from their report published in 2006. Also, in Fig. 1, 32 % of the respondents supported the ‘allow them to stay temporarily on a worker’s permit’ and 11 % supported allowing them to stay permanently.

Fig. 1
figure 1

Public opinion concerning policy options of US handling of undocumented immigrants

Full size image

Of primary interest in understanding the deviations in responses presented in Fig. 1 is the strength of the SCSHI questionnaire in its ability to capture many aspects of ethnic prejudice and perceptions of economic competition. The results of a principal components exploratory factor analysis provide support for our inclusion of a ten-item scale, to which we refer as a measure of ethnic prejudice and perceived economic competition. The ten items that make up the scale include measures of opinions and attitudes concerning the effect Hispanic immigrants are having on the national economy and the ethnic groups themselves.Footnote 2 Ethnic prejudice items included questions about whether Hispanic immigrants (1) ‘threaten American culture,’ (2) ‘significantly increase the crime rate,’ (3) ‘often end up on welfare,’ (4) ‘do not do well in school,’ (5) ‘have strong family values (reverse coded),’ and (6) ‘do not pay their fair share of taxes.’ Likewise, there were four items included in the SCSHI that provided potential measures of perceived economic competition. These items asked whether Hispanic immigrants (1) ‘offer a great deal to the economy (reverse coded),’ (2) ‘decrease the wages and salaries of native born Americans,’ (3) ‘strengthen our country due to their hard work and talents (reverse coded),’ and (4) ‘threaten the US economy.’ The results of the factor analysis are presented in Fig. 2.

Fig. 2
figure 2

Factor analysis: ethnic prejudice and perceived competition scale

Full size image

Once the factor extraction procedure was complete, we obtained a single factor solution. This single factor was the result of high loadings from all 10 survey items (0.542 or greater) and indicates the existence of sentiments associated with high levels of ethnic prejudice and economic competition as they pertain to attitudes and opinions toward Hispanics. The scree-plot in Fig. 2 highlights the unidimensionality of the underlying factors associated with the ten items. Furthermore, the high level of scale reliability, as measured by a Cronbach’s Alpha score of 0.84, indicates that the combination of the ten items reliably measures a single concept. The factor loadings are also presented in the bottom half of Fig. 2. The rotated factor loadings range from 0.542, Hispanics ‘threaten American culture’ to 0.714, Hispanics ‘offer a great deal to the US economy (reverse coded).’ Given the results of the factor analysis, an additive scale was calculated, ranging from zero to ten, with a higher score indicating an increasing level of ethnic prejudice and perceived economic competition.Footnote 3 A series of variables were also included in this investigation as controls, in order to better understand the independent effects of ethnic prejudice and perceived economic competition as predictors of respondent support for deportation in reference to the two other policy options. These controls include measures of age (in years), employment (full-time indicator), sex (male indicator), nativity (native-born indicator), race (black and Hispanic dummies in reference to whites), education (high-school graduate, some college, and college graduate indicators in reference to non-high-school graduates), and political party affiliation (Democrat and Independent indicators in reference to Republicans).

There is a significant amount of research suggesting that these indicators are significantly associated with migration preferences, and they are drawn from research directly focused on political ideology (Espenshade and Hempstead 1996), party affiliation (Espenshade and Hempstead 1996), socio-economic status and education (Ayers et al. 2009; Citrin et al. 1997; Espenshade and Hempstead 1996; Scheve and Slaughter 2001), race (Citrin et al. 1997; Espenshade and Hempstead 1996), social distance (Ayers et al. 2009; Branton et al. 2007; Esses et al. 2001; Hood and Morris 1998), types of contact with Hispanic immigrants (Hood and Morris 1998), ethnic prejudice (Burns and Gimpel 2000; Citrin et al. 1997; Ayers et al. 2009), perceived economic competition (Citrin et al. 1997; Esses et al. 2001; Huber and Espenshade 1997; Kessler 2001; Lapinski et al. 1997; Scheve and Slaughter 2001), and sources of news or information (Branton and Dunaway 2008; Burns and Gimpel 2000; Espenshade 1995; Huber and Espenshade 1997; Morris 2007).

Analysis

Given the limited scope of this research note, the analytic approach taken here is relatively simple and straightforward. Here we employ a single multinomial regression model to test the odds of the SCHIS respondent’s likelihood of supporting the policy options of ‘granting temporary worker’s status’ and ‘allowing them to stay permanently.’ The results are presented in the form of odds ratios for ease of interpretation, with a specific focus on the effect of the ethnic prejudice and perceived economic competition scale in relation to the likelihood of supporting policy alternatives to deportation.

Findings

The results of the multinomial regression model are presented in Table 1 and are organized to present the odds of supporting temporary and permanent status, respectively, in reference to the policy option of deportation. In reference to the primary relationship in this analysis, we see that each additional agreement with the statements indicating ethnic prejudice or perceived economic competition results in the survey respondent having about 35 % lower odds of supporting temporary status in reference to the option of deportation. Likewise, each additional increase in the ethnic prejudice and perceived economic competition scale is associated with 43 % lower odds of supporting the amnesty option in reference to the deportation option. Other significant relationships in the model show that the native-born are less likely to support the temporary status option than the deportation option (74 % lower), Hispanics are more likely to support both temporary and permanent options in relation to deportation (5 and 6.3 times more likely, respectively), and black respondents are about 51 % less likely to support temporary status and 39 % more likely to support permanent status, both in reference to the deportation option.

Table 1 Multinomial logistic regression results comparing odds of policy options concerning handling of undocumented immigrants
Full size table

In regard to the effects of education, one can see that all higher educational categories are at least three times more likely to support the temporary worker status option than are non-high-school graduates in reference to the deportation option. In contrast, all three of the higher educational categories are incrementally less likely to support the permanent status option than are non-high-school graduates in reference to the deportation option. The final set of variables, political party affiliation, provides further significant associations to these policy options. Independents report 55 % lower odds of supporting the temporary status policy option in reference to the deportation option, while Democrats report 74 % higher odds of supporting the permanent status option in reference to deportation.

Conclusions and implications

A number of potentially important implications for immigration policy can be gleaned from this research. In addressing the differences between our results and those of previous studies, we favour an explanation that is rooted in the dynamics of a nation experiencing economic distress. Just 3 years before our survey (SCSHI), the Pew Hispanic Center reported that 27 % of US citizens favoured deportation. When our data were collected in February and March of 2009, the US economy was in free fall. Economic fears and uncertainty were dominating both the media and public discourse. In such a gloomy economic climate, fears about immigration could be expected to have a powerful influence on public opinion. A comparison of the two polls resulted in a sharp increase from 27 % supporting deportation in 2006 to 51 % in 2009. The increase in support for deportation in such a short time frame is most likely to be a consequence of dramatically changing economic circumstances.

The rapid shift in support for deportation between 2006 and 2009 suggests that deportation of undocumented immigrants has moved from a fringe issue to a policy option with mainstream support. Not only has deportation sentiment increased over time, it has also become a majority preference in many political categories. In the 2006 Pew Hispanic Center report, every political category, including liberals, moderates, conservatives, Democrats, Independents, and Republicans, expressed only minority or fringe support for deportation of undocumented Hispanics. In fact, there was less than 35 % support for deportation in any of the ideological and party groups. By 2009, the majority of moderates and conservatives, as well as the majority of Independents and Republicans, was endorsing deportation. Even among respondents identifying themselves as liberal or Democratic, a substantial 40.4 and 46.1 %, respectively, indicated support for deportation. This pattern of results is consistent with the interpretation that support for deportation had become a mainstream feature of public opinion that transcends ideology and party. The data set, however, has a major limitation in that it provides no basis for estimating the ‘depth’ or realism of the policy preference. It is one thing to maintain that illegal immigrants should be deported as an opinion; it is entirely another to deal with the consequences of deporting 9–11 million individuals from the United States. The historical examples of the US deportation experiments have typically been short- lived and apparently failed to maintain public support. Given the modern media’s capacity for instantaneous dissemination of information and visual depictions of events, the reality of mass deportation would be transparent and probably repugnant to the US public.