Determinants of Attitudes toward Immigrants and Refugees – The Role of Perceived Threat and Cultural Values


Adriana Espinosa (The City College of New York), & Glen Milstein (The City College of New York)

Analyzing the United States’ public attitudes toward immigration policy using latent class analysis

For the United States the year 2016 is a presidential election year in which the two major candidates have starkly disparate views on immigration policy, restrictions on immigration and deportation of undocumented persons. Media reports have suggested the candidate who supports more restrictive policies is mainly supported by people with less education, who are male, European-American, and financially vulnerable. Heading into the 2016 election, we sought to measure a nationwide baseline of attitudes toward immigration from before the current election cycle. We sought to see if there was then a discernable set of attitudes associated with individual characteristics and perceptions that would both explain today’s different levels of support for the two major candidates, as well as serve as predictors of future voting patterns.

In this vein we used latent class analysis to determine the structure of attitudes about immigration policy within the American public, using the 2012 Time Series Study from the American National Election Studies (ANES) (N = 5,280), and multivariate logistic regression to investigate factors that relate to the probability of group membership. The 2012 ANES study has the advantage that unlike in previous releases, which obtained the public’s opinion about only one aspect of immigration policy, the study from 2012 asked five different questions about this topic, thereby allowing for the use of statistical methods to ascertain the taxonomy of public opinion regarding immigration policy in the US. Latent class analysis (LCA) uses conditional probability to classify respondents into various mutually exclusive groups or categories. Individual members of any given class have similar response patterns to a series of different categorical questions (Hagenaars & McCutcheon, 2002; Muthen, 2001). The questions analyzed via LCA gauged the public’s perceptions and preferences about immigrants taking away jobs from natives, laws to allow immigrant status checks on suspects, proposals to allow citizenship to illegal aliens who entered the country as children, US government policy towards unauthorized immigrants, and levels of legal immigration in the US. Five latent classes of attitudes toward immigration policy were identified. According to inter class probabilities, classes were labelled as [1] “High Restrictionist” (14.1%, n = 745), [2] “Moderate Restrictionist” (16.4%, n = 867), [3] “Minimal Acceptance” (28.0%, n = 1,484), [4] “Low Acceptance” (14.2%, n = 748) , and [5] “Moderate Acceptance” (27.1%, n = 1,436).

We considered ethnicity, education, income, social trust, perceptions about personal finances, and the importance of being American to one’s identity as predictors of latent class membership. The main results indicate Black and Hispanic respondents are as high as 15 times more likely than White respondents to belong to the moderate acceptance class relative to the high restrictioninst class. Also, individuals with at least a college degree are as high as 6 times more likely than their no-college degree counterparts, to belong to the moderate acceptance class relative to the high restrictionist class. Similarly, individuals who considered their American identity as extremely or very important, and individuals whose financial situations had gotten worse over last year, were about 3 times more likely than individuals who did not view an American identity as important or whose finances had not gotten worse than last year, to belong in the high restrictionist class relative to the moderate acceptance class. Respondents who have little to no social trust are also about 2 times more likely than their higher trusting counterparts to belong in the high restrictionist class than in the moderate acceptance latent class. Finally, males are approximately 1.5 times more likely than females to be in the high restrictionist class than in the less restrictive ones, and neither age nor income were significantly related to the probability of belonging to any individual latent class.

These findings suggest messages regarding immigration policy, which are framed to exacerbate the public’s general fear of people, loss of their economic stability, and their American identity, in particular among those with low degrees of education, will lead to higher resistance to amend immigration policy towards the inclusion of newcomers in the US, and bring support to the more restrictionist candidate. The latent class structure also identified a middle group, the minimal acceptance group, who represented 28% of the population, are more likely to be educated and have higher social trust and less worry about their personal finances than the most restrictive class, but lower than the highest accepting class. For this group candidates and other policy makers will need to direct their efforts to seek support for immigration policy change.



Ravini Abeywickrama (University of Melbourne), & Simon Laham (University of Melbourne)

“How do you feel about immigrants?” Distinct threat and emotion profiles characterize prejudice towards economic migrants, refugees and asylum seekers

The world faces a migration crisis. At least 60 million people are refugees, seeking asylum or internally displaced (UNHCR, 2014). Understanding attitudes towards migrants is essential for ensuring that these millions of displaced persons are welcomed and integrated into host societies. A complexity that is often overlooked both in research and in the media are the subtle, yet important distinctions that exist between different types of migrants. While some arrive through sanctioned channels (economic migrants and refugees) others may not (asylum-seekers), and so are frequently perceived to be “underserving” or “illegal” (Sales, 2002; Haslam & Holland, 2012; Hartley & Pederson, 2015).

Differences in migrant status are further important in informing policy, educating electorate and for fostering integrative host communities. This is particularly pertinent to the Australian context, given Australia’s long history of diverse and continuing migration, and the increasing need to facilitate multiculturalism.

Despite substantial differences in migrant status and accompanying implications, no research to our knowledge has concurrently examined the correlates of attitudes, action tendencies and policy endorsement towards three important migrant groups: economic migrants, refugees and asylum seekers. The current study used the Socio-Functional Account (SFA) of prejudice (Cottrell & Neuberg, 2005) to explicate the origins of migrant-related prejudice in the Australian context. This theory posits that specific groups pose distinct threats and, thus, distinct profiles of affective and behavioural responses, aimed at mitigating those threats. A consequence of this is that prejudicial attitudes towards different migrant groups might have importantly different consequences for behaviour and policy endorsement. The current study extended SFA by comparing how these threat-affect-behaviour relationships differ for three important migrant groups. Further, we incorporated a range of policy measures to assess how different threat-affect profiles relate to pertinent political behavioural intentions.

Participants (N=119, 93 female, Mage = 19.1, SDage = 3.17) completed an online survey measuring threat perceptions, emotions, general attitudes, stereotypes, action tendencies and endorsement of policies for all three migrant groups (economic migrants, asylum seekers, refugees).

Results showed that although the pattern of general attitudes was consistent with previous research – participants were most strongly prejudiced towards economic migrants, followed by asylum-seekers and refugees – each group elicited different threat and emotion profiles. Refugees and asylum seekers were seen as more threatening (especially to physical safety, health, in-group morality and reciprocity), than economic migrants. Consistent with this, asylum-seekers and refugees elicited more fear, compared to economic migrants; these groups also elicited more pity and guilt. Asylum-seekers further elicited significantly higher levels of these threats than refugees, and as a result, greater feelings of anger.

Consistent with this pattern of threat perceptions and emotional responses, compared to economic migrants, these two groups elicited less aggression, and also less support for restrictive economic policies (such as lower endorsement of placing caps on jobs and restriction of annual intake). This is also consistent with greater expression of guilt and pity towards asylum seekers and refugees compared to economic migrants. In addition, consistent with greater threat perceptions and negative emotions, asylum-seekers elicited more aggressive action tendencies and less respectful action tendencies than refugees. This was reflected in greater endorsement of integrative policies toward refugees (e.g., provision of subsidised utilities and greater willingness to engage in rallies to support this group.

Overall, participants perceived and intended to act most favourably toward refugees, while perceiving economic migrants and asylum seekers more negatively and seeking to endorse harsher policies towards them. These findings generally corroborate prevailing attitudes towards these groups, however, the current study is distinct from previous research in that it highlights the affective complexities underlying prejudice towards these three groups.

An implication of these results is that attitude and behaviour change interventions aimed at promoting acceptance and tolerance of migrants may need to be tuned to the specific migrant group under consideration. For example, in the context of asylum seekers, health and physical safety threat perceptions coupled with feelings of anger and fear seem to underpin prejudice and thus may be specifically targeted to combat negative attitudes towards this group. On the other hand, a different set of threat perceptions (such as perceived threat to economic opportunities) may need to be addressed to shape attitudes towards economic migrants. Relevant stakeholders may therefore have to utilise different integration methods to accommodate different migrant groups into host communities.

These findings have important implications for the economic, social and cultural integration of migrants, and provide preliminary insights into the ways in which to facilitate acceptance in host communities.


Hyeyoung Shin (Yale University), & John F. Dovidio (Yale University)

The role of cultural values in anti-immigrant attitudes

The present research investigated the role of cultural values in anti-immigrant attitudes. The basic premise of this research is that core values of a culture influence individuals’ perceptions of themselves and others and shape the nature of attitudes toward in/outgroup members. Although there is only limited work investigating cross-cultural variations in intergroup attitudes, previous research has shown some empirical evidence of cultural variations in the degree of prejudice (e.g., Dunham, Baron, & Banaji, 2006; Lee, Pratto, & Li, 2007; Sidanius & Pratto, 1999). Further research revealed that cultural differences in the degree of prejudice are related to core values of the culture to which people belong. For example, prejudice toward non-normative groups (e.g., different race or homosexuals) was greater in East Asian countries than in Northern European-heritage countries, due in part to the value of individual uniqueness in individual-oriented Northern European-heritage cultures and the emphasis on behavioural conformity in group-oriented East Asian cultures (Shin, Dovidio, & Napier, 2013). Also, economic competitiveness, that is related to a central value of Northern European-heritage cultures (the emphasis on capitalistic/meritocratic achievements), predicted prejudice toward immigrants and foreign workers more strongly in Northern European-heritage than in East Asian cultures (Shin & Dovidio, accepted with minor revisions). In addition, the value of individual achievements was associated with social hierarchy beliefs (the endorsement of social order based on achieved status or social reputations, e.g., beliefs in school rankings) in the US but not in South Korea (Shin, Dovidio, Napier, Stangor, & Uleman, in preparation). These cross-cultural differences in dynamics of intergroup attitudes suggest that understanding the role of cultural values in interracial/interethnic attitudes, including anti-immigrant attitudes, is important, particularly in major immigrant-receiving countries in which the chance of interracial/interethnic contacts (between the host-immigrant groups or between the immigrant-immigrant groups) is high.

We thus examined factors that represent cultural values relating to anti-immigrant attitudes (economic competitiveness, the value of recognition/achievements, perceived controllability in one’s life, and perceived distance to different race, religion, & language) and their associations with anti-immigrant attitudes, controlling for gender, age, education, income, and concerns for economic stability and safety from crime. We tested both within- and between-cultural variations on these associations in order see how these associations are similar and/or different within- and between-cultures, which would also show attitudes of not only the host groups (Whites in the major immigrant-receiving countries) but also the immigrant groups (individuals with Latino or East Asian cultural backgrounds). We used nationally representative samples drawn from the World Values Survey ( through face-to-face interviews between 2010 and 2015. In Study 1 (N=4570; US whites, n=1613; Australia whites, n=1216; Germany whites, n=1741), the proportions of negative responses to immigrants and foreign workers were higher in Germany and in the US, compared to in Australia. However, as hypothesized, economic competitiveness and the value of recognition/achievements, that are related to core values of Northern European-heritage cultures (based on the Protestant Ethic and meritocracy beliefs), commonly predicted anti-immigrant attitudes among individuals with Northern European-heritage cultures, in addition to perceived distance to difference. Perceived distance to different race, religion, and language predicted anti-immigrant attitudes particularly strongly in Germany. In Study 2 (N=9,319), we additionally considered ecological diversity of each country (based on the degree of ethnic & linguistic fractionalization) using data from US (Whites, n=1,613), Mexico (n=2,000), and China (n=2,300) as countries with a high ecological diversity, and Sweden (n=1,206), Chile (n=1,000), and South Korea (n=1,200) as countries with a low ecological diversity. Anti-immigrant attitudes were stronger for East Asian groups compared to Northern European-heritage and Latino groups. Also, perceived distance to different race, religion, and language was a common predictor of anti-immigrant attitudes across cultures, except for the perceived distance to different religion in Northern European-heritage group. However, perceived distance to differences predicted anti-immigrant attitudes particularly strongly in Latino and East Asian groups based on that group-oriented cultures tend to emphasize conformity with ingroup norms rather than individual uniqueness and salience. Also, as predicted, economic competitiveness and the value of recognition/achievements predicted anti-immigrant attitudes for Northern European-heritage groups, whereas perceived controllability in one’s life was negatively associated with anti-immigrant attitudes for East Asian groups (stronger beliefs in no choice in one’s life, stronger anti-immigrant attitudes) reflecting the holistic/circular world view and the emphasis on predetermined qualities in group-oriented East Asian cultures (group-oriented cultures prefer relatively more fixed and well-organized relations to maintain group-oriented society more stably). Results reveal that understanding the role of cultural values is important to understand anti-immigrant attitudes more comprehensively and to suggest both general and culture-specific interventions.