Hierarchies of hostility: Why do citizens prefer certain immigrants? Authors: Judit Kende, Fouad Bou Zeineddine, Jessica Gale, Eva G. T. Green Published: 2026 ABSTRACT A reanalysis of 100 studies examining which immigrants citizens prefer highlights why some immigrants are favored and what creates dichotomies of “good” and “bad” immigrants. Studies examining attitudes toward immigrants across countries often focus on immigrants as a monolithic group. Such research asks, for example, whether citizens find immigrants threatening to their economy or whether they support admitting more immigrants, in general, to a country (1). In this issue of Science Advances, Aviña et al. (2) move beyond this approach, offering new insights by highlighting which immigrants citizens prefer across different countries. In their study, Aviña et al. reanalyze datasets from 100 individual studies covering 142,817 survey respondents from 36 countries. This is the largest dataset collated so far to examine the impact of specific attributes of hypothetical immigrants on citizen preferences, like their gender or country of origin. All studies examined by Aviña et al. use conjoint experiments, a method that allows the isolation of which attributes of fictitious immigrants influence their evaluations. The results show that some preferences are broadly similar across 36 countries such that economic, cultural, humanitarian, and legal attributes matter, although citizens’ preferences differ across political and party lines. In addition to these key contributions, the authors report several other findings that are highly relevant to current theoretical and policy debates. Our focus article highlights the findings that lead us to ask why some immigrants are preferred over others. To answer this question, we describe how such dichotomies of “good” and “bad” immigrants are created and reinforced by public policies and political narratives. The authors provide valuable and perhaps surprisingly positive conclusions about citizens’ overall evaluation of immigrants. Despite divisive rhetoric and violent border control in many countries, overall, citizens are more likely to accept immigrants into a country than not and more likely to grant them citizenship than not. While this positive evaluation might seem astonishing, it is in line with trends from other large-scale studies showing that in the US and most of Europe, attitudes toward immigrants and immigration have improved over the past decades and are overall rather positive (3). The pattern of preferences also shows some puzzling contradictions about which immigrants are preferred. For example, citizens favor immigrants fleeing violence. At the same time, immigrants should not have a physical disability or suffer from posttraumatic stress disorder. Immigrants thus face unreasonable expectations, as the very violence they flee could have handicapped or traumatized them. Similarly, citizens’ preferences about immigrants’ motivations contradict their preferences regarding their contributions: Although citizens frown at migration for economic reasons, they nonetheless prefer immigrants with high levels of education, employment status, and skills. These qualities help immigrants find work and thus contribute economically, even though many wealthy countries are also in dire need of so-called low-skilled labor (4). Immigrants are thus expected to contribute to the economy but not to migrate to improve their own financial situation. While these contradictions are partly due to the design of the studies, they also reveal that citizens often think of immigrants in an ambivalent and utilitarian, even exploitative, manner. Paraphrasing the famous quote by Swiss writer Max Frisch, citizens ask for grateful refugees and workers, but they are surprised when they get people instead (3). At first glance, based on Aviña et al., race, ethnicity, and cultural characteristics of immigrants do not matter much for citizens’ preferences, especially for left-leaning citizens. More specifically, in their study, citizens’ preferences do not differ or only differ slightly based on whether immigrants come from a country viewed as having a shared culture, democratic political norms, or higher economic development status and whether immigrants come from a predominantly white, predominantly Black, or Muslim-majority country, even though many of these attributes matter slightly more for right-leaning citizens. This result contradicts established findings showing that white citizens often prefer white immigrants and those immigrants that they expect to be culturally similar (1). However, although less obvious at first glance, a closer look suggests that some of the most strongly penalized or preferred attributes are attributes that are racialized in a more hidden or subtle way. For example, one of the most penalized attributes is irregular entry into the country or undocumented status. While these attributes could be interpreted as simply a question of legal status, legal entry or status is strongly tied to race and ethnicity. Migrants who enter without a visa or stay without a permit are far more likely to be racialized immigrants. The reason for this is that the migration policies of highly developed countries provide easier access to visas and residence permits (or other forms of legal entry and stay) to citizens of other highly developed, predominantly white countries, for example, of European citizens to the US (3). Moreover, when white US citizens think of undocumented immigrants, they actually think of racialized immigrants (5), and when white citizens think of highly educated, highly skilled immigrants, they often think of white immigrants (6). Thus, while it could appear that citizens do not evaluate immigrants by their race, ethnicity, or culture, a more careful unpacking of immigrant characteristics reveals racialized evaluations. Many of the findings are not generally applicable “across countries,” as the authors sometimes imply. Twenty-nine of the 36 countries were North American or European, and 82% of the observations were from the Global North (meaning, from Aviña et al., Europe without Turkey, plus Australia, Canada, Japan, South Korea, and the US). Moreover, several of the strongest preferences in the overall sample, for instance, irregular entry, were less researched in Global South countries, inhibiting comparisons between countries. Looking at the results where comparisons between different regions were possible, several preferences observed among Global North citizens disappeared among Global South citizens, such as the penalty against immigrants from Afrodescendant countries. Preferences are far less consensual in the Global South than in the Global North, as indicated by the substantially wider confidence intervals. Given these constraints, it is not possible to draw far-reaching conclusions about consensual global preferences based on the available data. Aviña et al. provide rich insights regarding which immigrants are preferred, raising the question of why such immigrants are preferred. More specifically, why do citizens think in dichotomies of good and bad immigrants, such as economic migrants versus refugees or low-skilled versus highly skilled immigrants? Let us look at how such evaluations are driven by public policies and political narratives (among other factors) reinforcing racialized hierarchies of immigrant groups (6–8). Legal-political policies of migration and integration create categories of immigrants and endow different categories of immigrants with distinct rights. These frameworks sort migrants into different categories such as involuntary (e.g., political asylum seekers) versus voluntary (e.g., economic migrants) or as highly skilled versus low skilled. These almost universally applied binaries often do not correspond to migrants’ lived realities, yet the legal-political categorization determines which migrants are allowed to stay, for how long, and with what rights. For example, are they are allowed to settle without further procedures or are they granted the right to vote? Immigrants sorted into these different legal-political categories are also endowed with different rights in the socioeconomic domain, for instance, whether they have access to health care or education. On this politico-legal foundation, political and media narratives deploy categories of “good” and “bad”’ immigrants and assign higher or lower status to different immigrant groups (7). Policies and narratives produce, in tandem, a hierarchy of different migrant groups that reinforces and legitimizes uneven distribution of rights, resources, and status. An influential body of work shows that policies and narratives matter for the evaluation of immigrants (Fig. 1) (8, 9). For example, research by our group shows that inclusive integration policies can shift citizens’ immigration attitudes, with more inclusive integration policies sometimes leading to a 30% improvement in attitudes (10). Citizens’ preferences for immigrant groups. Public policies and political narratives drive the creation and reinforcement of racialized hierarchies of immigrant groups. Credit: Ashley Mastin/Science Advances. Past work on the impact of policies and narratives on immigration attitudes has largely focused on immigrants as a monolithic group (8, 9). This work is similar to the studies mentioned at the outset that ask whether immigrants are accepted (1). Future research should build on the valuable findings and methods of Aviña et al., showing which immigrants are accepted, and explicitly test why certain immigrants are accepted, focusing on the role played by policies and political and media narratives. Moreover, many of the attributes used by Aviña et al. have layered semantic associations and often racialized connotations. One such example is that people think of racialized immigrants when they think of undocumented immigrants (5). Therefore, future research should unpack what the attributes mean to citizens, as in, who do citizens think about when they imagine immigrants with these attributes. It is also appropriate to reflect on the risk of legitimizing racialized hierarchies of immigrant groups scientifically if the focus remains on group-based categorical ranking, especially given the advancement of the far right globally today. Even if the work itself clearly does not intend to legitimize hierarchies, a high risk exists of misuse by bad actors. The consequences could be severe, given historical experiences in which groups were socially ranked in such formalized ways institutionally. While we acknowledge the scientific value of the work, we lament that the paper neglects these thorny questions.