National Identity and Preferences for Chinese Inclusion in Mexico

Abstract

In this paper, we explore the bases of Mexican national identity construction and use an array of conceptions of nationhood to study contemporary attitudes towards foreigners’ sociopolitical rights in Mexico. Rarely is the study of national identity connected with immigration policy preferences in general, and even less so outside advanced countries. We explore the content of Mexicanness and use this content to understand public opinion preferences towards the integration of diverse groups of foreigners in Mexico. We employ 2016 survey data and a survey experiment and find the persistence of xenophobic attitudes towards the Chinese community in Mexico. We also show that civic conceptions of nationhood cannot counter contemporary anti-Chinese sentiment, in great part because the civic belonging of the Chinese was defined on racial terms. Lastly, we show that these processes of national identity construction, based on the marginalization of certain groups, are persistent and shape todays’ attitudes and preferences towards the incorporation of different groups of foreigners. It remains to be explored whether material interests associated with the recent Chinese “going out” policy may be able to counter deep-seated anti-Chinismo

Introduction

In 2010, a national government agency in charge of preventing discrimination stated that “the prevalent image of Mexico as a country in solidarity with and open toward foreigners is debatable” (Buchenau 2014, 68). We explore contemporary attitudes towards immigrant socio-political rights using the process of Mexican national identity construction and different conceptions of nationhood as the bases for substantiating contemporary discriminatory attitudes towards foreigners.

Rarely is the study of national identity connected with immigration policy preferences in general, and even less so outside the developed world (Foote and Goebel 2014; Simonsen and Bonikowski 2020; López 2014; Freier et al. 2020, 144; Mylonas and Tudor 2023, 6). In this research, we explore the content of Mexican identity using survey data. We then use this content to study public opinion preferences towards the rights of diverse groups of foreigners in Mexico. We associate sentiments about the Chinese community in Mexico with the construction of a mestizo (mixed) Mexican identity in the post-Revolution period. The construction of Mexicanness was overtly anti-foreigner and strongly anti-Chinese (Chang 2017; Chao Romero 2011; Delgado 2012; Rénique 2015; Yankelevich 2012). In this regard, this study suggests that the content of national identity shapes in part how people feel about contemporary migrants’ rights. Further, we show that civic conceptions of nationhood cannot counter contemporary anti-Chinese sentiment, in large part because in Mexico, the civic belonging of the Chinese community was shaped by racial discrimination (Chang 2017; Chao Romero 2011; Delgado 2012).

Under the impulse of the Porfiriato regime (1876–1911) and following the 1882 Chinese Exclusion Act in the United States, a large number of Chinese nationals settled in Mexico. Their numbers quickly increased from barely 916 in 1895 to almost 20,000 in 1930 (Chang 2017, 13), becoming the second-largest group of immigrants in Mexico (Chang 2017, 11; Chao Romero 2011, 175). The size of this group continued to increase until the early 1930s, even amid intense anti-Chinese campaigns that culminated in the expulsion of the Chinese from Sonora in 1931. A decade later, the number of Chinese nationals fell to over 6,000 and continued to decline in the second half of the twentieth century (Chang 2017, 13; Rodríguez García, 2017).

In 2015, one year before the survey we use in this study, the Chinese population stood at 8,860, with 25% holding Mexican citizenship (Rodríguez García, 2017; INEGI, 2015). There are about 5,500 Japanese and a similar number of South Koreans (INEGI, 2020). Chinese foreign-born populations, while small, have increased since the 2000s (Lisbona and Rodríguez 2018). According to the most recent Census (INEGI 2020), the estimated number of Chinese immigrants has grown to 10,547 in 2020. This is still a very small proportion of the approximately 1.2 million foreigners with residence in Mexico (INEGI 2020). ¹ This increase has intensified in recent years, associated with China’s “going out” investment strategy (Armony and Velásquez 2015). According to recent data, temporary visas for Chinese workers have increased by 157% between 2019 and 2023, and China has become the third country of origin of temporary immigrants behind the United States and Colombia. ²

Our study shows that in comparison to other groups of foreigners and despite their modest presence, in 2016, there was little support for foreigners of Asian origin to accessing socio-political rights. Based on the historical account we offer and the contemporary numbers we provide, we propose that understanding today’s attitudes toward Asians in Mexico requires considering the central role that hostility against the Chinese community played in the construction of Mexican identity. Our results suggest that the process of national identity construction, based on the marginalization of this community, appears to persist, shaping today’s attitudes and preferences towards foreigners’ rights in distinctive ways.

A caveat should be in place from the outset. A wealth of historical evidence justifies that in this article, rather than referring to a generic anti-Asian sentiment, we focus on anti-Chinismo. Not only do the Chinese continue to be the largest Asian community in Mexico; but the role and visibility of this community during the process of national identity construction justifies our focus on this group in the discussion that follows. Equating the categories Chinese and Asian seems to be part of the collective Mexican—and not only Mexican—imaginary. For instance, discussing the Chilean case, Chan and Montt Strabucchi (2021, 377) state, “(I)n Chile and in many Latin American countries, the term for Chinese persons is a catch-all that also refers to persons of East Asian origin: ‘chino.’ This homogenizes and racializes persons of East Asian appearance under the unstable category of ‘chino’.” Lim (2020, 443) discusses the portrayal of Koreans in Mexico as “just a type of chinos.” And as Chang (2017, 10) puts it “[I]n Mexico, Chinese populations—in fact all Asians—are known as los chinos. This term is commonly heard across Latin America in reference to all Asians as a racial category carried over from the colonial era…” With this in mind, we ground our discussion of contemporary attitudes towards foreigners’ rights on the construction of Mexican national identity in the post-Revolution period and the role that anti-Chinismo played in that process.

Our article proceeds as follows. In the first section, we discuss the content of different conceptions of national identity. While we acknowledge that this is a complex debate with multiple overlapping, unstable, and often controversial definitions of nationhood, we use these conceptions in a pragmatic way to shed light on the content of Mexican national identity (Theiss-Morse 2009; Schildkraut 2011, 2014; Bertossi and Duyvendak 2012; Huddy 2016; Ariely 2020; Piwoni and Mußotter 2023; Miller 1995; Mylonas and Tudor 2021). Second, we argue that understanding national membership in Mexico and its consequences today requires understanding the history and context that shaped the process of Mexican identity construction. Third, we use the survey México y el Mundo (2016) (Maldonado et al. 2018) to study different dimensions of Mexican identity; we then show that these dimensions are relevant to understanding contemporary preferences towards migrant rights. An embedded survey experiment asking about attitudes towards the rights of different foreigners suggests that the content of national identity has heterogeneous effects that may depend on the particular ethnic origin (Simonsen and Bonikowski 2020; Konitzer et al. 2019; Hainmueller and Hopkins 2014). While our analysis has limitations, and is exploratory in many respects, we make a contribution to debates on national identity and immigration policy preferences in the Global South. We conclude with some reflections on policy implications and suggestions for future research in the context of the growing geopolitical relevance of China in the region.

1. National Identity and Migrants’ Rights

Despite the academic interest in the study of nationhood, little attention has been paid to national identity as an explanatory variable; that is, to conceptions of national identity as input to explanations of immigration policy preferences (Foote and Goebel 2014; Green et al. 2011; Simonsen and Bonikowski 2020; Lindstam, Mader, and Schoen 2021). Given “the Eurocentric bias of mainstream nationalism scholarship” (Mylonas and Tudor 2023, 6), studies of this relationship outside the developed world are even less common.

We follow the social identity theory approach (Tajfel 1981; Tajfel and Turner 1986) as a starting point. According to this perspective, national identity is a social identity, which derives from a person’s knowledge of their membership in a social group, together with the emotional significance attached to that belonging (Tajfel 1981; Theiss-Morse 2009). How attached to their group people feel and how narrowly they set boundaries around the ingroup shapes their attitudes and behaviors towards outgroup members (Theiss-Morse 2009, 8). In this regard, we claim that the content of national identity can shape how people feel about migrants’ rights. In fact, previous studies have shown that conceptions of national identity influence opinions about language policy, racial profiling, and immigrant integration policies (Citrin, Wong, and Duff 2001; Schildkraut 2005, 2011; Theiss-Morse 2009).

One key element for our study is the “constitutive norms” of national identity, meaning “the rules that define group membership” (Abdelal et al. 2009, 19). Constitutive norms inform people how to behave as group members and guide attitudes and behaviors within a group (Theiss-Morse 2009). So, adherence to these rules helps to set boundaries between those who belong to the ingroup and those who are excluded from the “imagined community” (Anderson 1983).

When people think about the constitutive norms of their national identity, they think about the features they share with their community. The most relevant conceptual building blocks in the study of national identity are the following dimensions: (1) the ethnic vs. civic dimension; and (2) the nationalism vs. patriotism dimension (Brubaker 1992; Schildkraut 2014, 447; Theiss-Morse 2009; Green et al. 2011; Wright et al. 2012; Huddy 2016; Lindstam et al. 2021; Piwoni and Mußotter 2023). We use these categories as a heuristic device in our empirical research, but we acknowledge the complex, polymorphous character of this classification. These dimensions should be regarded as a continuum rather than starkly different, immutable categories (Piwoni and Mußotter 2023).

First, an ethnic view of national identity emphasizes the heritage of the past, the ancestral character of identity, and the role of racial, ethnic, religious, and cultural factors. Among the latter, ethnic views focus on common language and collective memory in the creation and reproduction of nations (Smith 1991). In ethnic terms, boundaries are determined in part by the shared memories of their members (Anderson 1983; Miller 1995) and in part by the boundaries purposedly crafted by political and intellectual elites (Theiss-Morse 2009; Miller 1995). Although some of the ethnic characteristics can be acquired, such as learning a new language, most of them are ascriptive elements that are transmitted through the generations (Schildkraut 2014, 447). So, these elements are considered strong identifiers that are more likely to set rigid boundaries regarding who belongs to the main group (Theiss-Morse 2009, 13; Schildkraut 2014). Boundaries based on ethnic features demarcate who the prototypical member is.

Second, a civic conception of the national identity focuses on the idea of a political community derived from citizenship, entailing a series of laws, institutions, and shared values (Greenfeld 1992). The content of civic approaches to national identity is multidimensional. Civic concepts of national identity emphasize actions—such as willingness to actively contribute to public life—and identities—such as seeing oneself as and feeling part of the national group (Schildkraut 2011, 2014). As long as these behaviors and identities are embraced, civic conceptions of national identity do not, in principle, exclude. In Schildkraut’s studies, civic identity—that is, endorsing the idea that to belong to the national group one must think and feel as a member of it—is even more determinant of policy preferences than civic doing.

Some studies suggest that different conceptions of nationhood explain attitudes towards immigrants and their inclusion (Schildkraut 2005, 2011, 2014; Theiss-Morse 2009; Wong 2010; Ariely 2020; Lindstam et al. 2021). Boundaries based on civic terms are considered weaker and more permeable than ethnic identifiers that can exclude (Theiss-Morse 2009; Lindstam et al. 2021). When national identity puts conceptions of citizenship rights and responsibilities at the center of its definition, this is likely to lead to more open policy preferences towards immigrants and their rights, as long as foreigners fulfil their duties as committed citizens willing to “do their part” for the common good (Ariely 2020).

The differentiation between ingroup and outgroup based on ethnic features is often conducive to discriminatory and hostile attitudes towards those perceived as not belonging to the ingroup (Schildkraut 2005, 2011, 2014; Theiss-Morse 2009; Wong 2010). For instance, research shows that those who define American identity in ethnic terms are less supportive of open immigration, more opposed to extending citizenship rights, or more in favor of selecting immigrants on the basis of language rather than education or skills (Schildkraut 2005, 2011; Wong 2010; Wright 2011; Wright et al. 2012).

While there is evidence that embracing a civic dimension of national identity tends to correlate with more open immigration policy preferences (Green et al. 2011), some suggest a nuanced picture. For example, Schildkraut (2005, 2011, 2014), in her research on what being American means, shows that the feeling (identity) aspect of civic nationhood is associated with endorsement of the assimilation principle that immigrants should make the effort to integrate. Moreover, the impact of the civic content of national identity on policy preferences seems to be conditional on the specific policy analyzed. This is well illustrated in the case of language policies, partly because language straddles the categories of ethnic trait, cultural trait (Kymlicka 2001; Shulman 2002), and civic duty, the last one if learning the local language is considered a civic obligation (Schildkraut 2011, 48; Reeskens and Hooghe 2010, 591).

There also seems to be heterogeneity regarding the immigrant group (Simonsen and Bonikowski 2020): if the immigrant group is perceived as not being compatible with the values that define civic belonging, embracing a civic conception of nationhood may not imply more support for foreigners’ rights (Reeskens and Hooghe 2010; Lindstam et al. 2021). In other words, the impact of civic conceptions of national membership on immigration policy preferences is heterogeneous in terms of which policies and in terms of which immigrant groups. Besides, the co-existence of the two dimensions (ethnic and civic) possibly pushing in different directions may lead to ambivalent, unstable outcomes regarding conceptions of nationhood and preferences for migrant integration.

Lastly, beliefs about nationhood are also part of constitutive norms of national identity (Schildkraut 2011, 6). These beliefs fall under two main categories: nationalism and patriotism (de Figueiredo and Elkins 2003; Wong 2010; Huddy 2016; Theiss-Morse 2009; Piwoni and Mußotter 2023). Nationalism tends to have a negative connotation since it involves an “uncritical and blind attachment to the nation” and feelings of superiority with respect to other countries and nationalities. ³ In contrast, patriotism implies pride and love for the country and its institutions (Green et al. 2011, 369; Theiss-Morse 2009, 24; Piwoni and Mußotter 2023). ⁴ Patriotism, or love for the country, does not have any comparative or competitive implication. In their study, de Figueiredo and Elkins (2003) contrasted nationalism with patriotism, finding that more nationalist individuals tend to be “bigots” (see also Green et al. 2011), but those who are more patriotic are not.

All considered, it seems that different conceptions of national identity play a multifaceted role in the attitudes of natives toward migrants’ rights. Using the Mexican case, we propose that historical processes of national identity construction solidified certain beliefs about what a national of a country should be in ethnic, civic (acting and feeling), and nationalist terms. As we show, the sedimentation of these beliefs helps to explain contemporary attitudes toward the socio-political inclusion of foreigners of Asian origin (popularly referred to as los chinos).

2. The Chinese in Mexico: Conceptions of National Identity and Socio-political Exclusion

We offer a brief account of the socio-political exclusion to which the Chinese community was subjected in the process of defining the Mexican identity. We argue that today’s socio-political exclusion cannot be detached from the anti-Chinismo that was central to the construction of the Mexican national identity, around the “mestizaje” project.

The history of the Chinese in Mexico is one of permanent harassment and recurrent attempts at social exclusion and eventually expulsion (Delgado 2012; Rénique 2015; Craib 1996; Chao Romero 2011; Fitzgerald and Cook-Martin 2014; Chang 2017; 2019). ⁵ This history became deeply ingrained in the concept of what being Mexican means, and in the role assigned to the Chinese in the crafting of that identity. We propose that these constitutive norms linger in today’s attitudes of the Mexican public toward the Chinese, which continue to be marked by high levels of anti-Chinese sentiment (Acevedo and Meseguer 2022).

Although the Chinese presence in Mexico is closely related to Porfiriato policies to attract cheap labor around the turn of the nineteenth century (1876–1911), including low-wage Chinese workers, our review of hostilities puts the focus on the period in which those hostilities heightened (during the Revolution and after), spreading to the rest of the country until they became a national anti-Chinese crusade (Craib 1996; Chao Romero 2011; Delgado 2012; Rénique 2015; Chang 2017; 2019). ⁶ Many acts of exclusion were motivated by the perception that the Chinese were racially and culturally unfit to contribute in a positive way to the vision of the nation that was embraced after the Mexican Revolution (1910–1920).

First, on ethnic terms, Sinophobia cannot be understood without reference to the revolutionary and the post-revolutionary construction of a Mexican racial identity around the ideal of mestizaje. Virtually since their arrival, the Chinese were regarded as spreaders of disease, a source of vice and germs, deficient in moral and physical integrity, and “racial poisons” (Craib 1996, 17; Delgado 2012, 162, 189; Rénique 2015, 95). For this reason, the Chinese were often ostracized, if not openly harassed both informally and through explicit legislation, as we explain below. After the Revolution, the state building process embraced national eugenics and social hygiene movements as the model for building a new mestizo race that could homogenize and unify the country. Prominently after 1924, the mestizo project, inspired by José de Vasconcelos and Manuel Gamio among others, envisaged a new “cosmic race,” prescribing the “correct” miscegenation of indigenous peoples and Mexican-born Whites to produce a mestizo race (Rénique 2015, 106; Chang 2019). In this context, Chinese miscegenation with Mexican women turned out to constitute the main threat to the new mestizo identity. Every union between Chinese men and chineras (Mexican women who married Chinese men) meant “the injection of sickly yellow blood,” which conflicted with the mestizo project of national state building (cited in Delgado 2012, 158). This justified all sorts of attacks on the Chinese community and indeed a politically orchestrated plan aimed at racial cleansing (Chang 2017, 17, 25).

Second, on civic terms, the narrow ethnic vision of the new Mexican state that emerged from the Revolution was reflected in numerous anti-Chinese laws affecting all aspects of this group’s daily life, in both the public and private spheres. As Delgado aptly summarizes (2012, 189), confining “civic belonging to narrow concepts of mestizaje” implied that despite the Chinese insistence on upholding their political and constitutional rights, the Chinese were “unmade as Mexicans.” In other words, civic identity and civic belonging was shaped by ethnicity, to the extreme of undoing or ignoring legally granted civic and social rights to this group.

Notably, anti-Chinese legislation openly contradicted the protection of civic rights granted to the Chinese, in existing legislation such as the 1917 Constitution, and even through their naturalization as Mexican citizens. In fact, naturalization often proved insufficient to protect the Chinese against anti-Chinista hatred. Several fronts on which anti-Chinese legislation conflicted with civic rights was in the regulation—indeed penalization—of mixed marriages, relocation of the Chinese to restricted parts of cities or sections of states to minimize the probability of miscegenation, and the limited value afforded to naturalization in terms of guaranteeing civic rights (Chao Romero 2011; Delgado 2012; Chang 2017). For instance, barrioization, while often justified on the basis of sanitation concerns, had more to do with preventing miscegenation, and involved isolating mixed families and Chinese businesses by confining them to specific barrios (Chinatowns) or territories (Rénique 2015; Gómez Izquierdo 2019).

As for naturalized Chinese, while several pieces of legislation, prominently the 1917 Constitution, granted equal rights and full civic belonging, the daily life of Chinese Mexicans was often marked by xenophobic local laws. Anti-Chinistas vocally opposed the naturalization of the Chinese. A combative leader of the Sonoran anti-Chinese movement, José Angel Espinoza, stated that “every naturalization certificate [that] our government extends to a Chinese citizen is like a rattlesnake placed at the bosom of the motherland” (cited in Delgado 2012, 163). This statement tellingly reflects the strong opposition to granting the Chinese full civic rights. Their demands to be recognized as Mexican citizens and granted the same rights were occasionally heard, but only because national leaders were concerned about international criticism and reputation.

Finally, while xenophobic nationalism permeated both the definition of ethnic and civic identities, exclusionary nationalism (“Mexico for the Mexicans, China for the Chinese”) was most evident in the economic sphere. In our view, economic nationalism as described here is closely related to the civic conception of nationhood; that is, how a “true” member of the community must act. The perception of the Chinese as “ruthless business competitors,” “tax evaders,” and “an insatiable yellow hydra” (Delgado 2012, 109; Chang 2017, 87; Gómez Izquierdo 2019, 208) raised continual demands to regulate their economic activities. On the economic front, the Chinese community suffered attacks on business ownership and severe labor laws. For instance, the so-called 80% law (1919) forced businesses to hire at least 80% Mexicans, defined in terms of ethnicity and birth. The Chinese appealed the law and invoked Mexican constitutional tradition, demanding the “full enjoyment of social and private rights” (Delgado 2012, 157). But as Delgado puts it, the Chinese learned the hard way that “the post-revolutionary project was Sinophobia and not the Constitution” (p. 159). In sum, ethnically, politically, and economically, the Chinese were regarded as “pernicious individuals” to be excluded from any constitutive norm (ethnic but also civic) of Mexicanness.

3. Hypotheses

Although the picture is attenuated by time, respect for legally granted civic and social rights, and considerable levels of political correctness, at least in the official discourse, ⁷ the Chinese in Mexico continue to be portrayed today as an alien race, economically predatory, and reluctant to integrate (Armony and Velásquez 2015). Anti-Chinismo is still evident on two fronts: first, when it comes to referring to China and the Chinese community’s contribution to the Mexican economy; and second, when the malleability and capacity to integrate of foreigners with Chinese ancestry is portrayed in contemporary media. Examples abound. For instance, the significant commercial deficit with China arouses frequent complaints among local producers about unfair Chinese competition practices. Chinese businesses are often charged with allegations of violating labor and wage standards, engaging in piracy, and smuggling (Armony and Velásquez 2015; Cornejo et al. 2013, 64). The import of products from China, such as prickly pears, which are regarded as national symbols of Mexico, has raised misgivings and provoked emotional pleas in written and digital media to protect local production (Lutz and Padilla 2012). Chinese foreign investment often arouses mistrust about quality product and spurs demands for more employment of Mexican workers—which the National Immigration Institute forces to be 90% of the labor force per plant (Ortiz and Prudencio 2022). Headlines of the sort “China, a threat,” “Mexico at war against China in the WTO,” or “Chinese pirates invade Chiapa” (cited in Cornejo et al. 2013) suggest that China continues to be a “favourite villain” for business, media, and political elites in Mexico.

On the second front, that of the unmalleability of the Chinese, illustrations that suggest the endurance of negative stereotypes also abound (Lisbona Guillén and Balam 2018). For example, the potential arrival of 1,000 Chinese families in the Caribbean associated with an investment project (Dragon Mat) raised concerns due to the “endogamy and resistance to integration in the host culture […]” (Cornejo 2019, 890). The argument that the arrival of Chinese nationals posed a threat to community co-existence is aligned with the deeply ingrained perception that the Chinese are “clannish” and not trustworthy (Cornejo, Haro Navejas, and León-Manríquez 2013; Foote and Goebel 2014; Cornejo 2019). Racism continues to affect second, third, and fourth generation Chino-mexicanos, who continue to report bullying and forms of symbolic violence (Manzano-Munguía and Juárez Palomino 2020). More recently, the revival of racist discourse associated with the COVID-19 pandemic has shown that anti-Chinese racism is alive and well in Mexico (Lim 2020; Sanchez-Rivera 2020; Manzano-Munguía and Juárez Palomino 2020) and elsewhere (Yiu 2022; Chan and Montt Strabucchi 2021). Sinophobia appears to be inspired by factors similar to those that led to the ethnic, civic, and nationalist exclusion of the Chinese a century ago (Yiu 2022; Chan and Montt Strabucchi 2021). For instance, in Mexico, the media, songs, and public figures associated the virus with Chinese eating customs and poor hygiene (Lim 2020, 443; Sanchez-Rivera 2020, 35–36). ⁸

With this in mind, we hypothesize that the long history of discrimination reflected in so many instances of civic harassment and deprivation of rights has fed a deeply exclusionary national identity that regards the Chinese as an inferior, unsuitable, and un-civic outgroup. If the construction of Mexicanness succeeded in solidifying an image of the Chinese as undesirable foreigners, then even today, endorsement of a (racialized) civic definition of national membership, as well as ideas of national superiority will likely be reflected in little support for foreigners of Asian ancestry to enjoy access to socio-political rights. Therefore,

(H1) Natives who support an ethnic conception of Mexican identity will exhibit less support for granting foreigners of Asian origin socio-political rights.

(H2) Natives who support a civic conception of Mexican identity will exhibit less support for granting foreigners of Asian origin socio-political rights.

(H3) Natives who think that Mexico is a superior country will exhibit less support for granting foreigners of Asian origin socio-political rights.

(H4) The above expectations do not apply to other groups of resident foreigners.

4. Data and Method

We use the 2016 Mexico application of Las Americas y el Mundo to analyze whether and how the content of Mexican national identity shapes mass public preferences regarding foreigners’ access to socio-political rights (access to public education, family reunification, voting, and forming political organizations). The survey data is from 2,400 respondents in a nationally representative sample (Maldonado et al. 2018).

Recall that our main research question is whether and how these policy preferences are shaped by ethnic (ascriptive), civic (permeable), and nationalistic (exclusionary) constitutive norms of Mexican national identity. The survey features an embedded experiment on the origin of foreigners which by itself allows us to explore the importance of migrant national origins on policy preferences. Prior to the questions on migrant access to the above policies, and without adding any other information, respondents were primed with an image of a hypothetical foreigner in Mexico. There are three treatment groups in addition to the control group, which did not receive a priming image. Respondents were randomly assigned an image of an immigrant. The three treatments vary the ethnic profile of the immigrant, while gender is held constant (males): White Anglo-Saxon (Picture B), Asian (Picture C), and indigenous Central American (Picture D). ⁹ As interviewers show the image of the migrant (Figure 1), respondents are asked to keep “the image of the foreigner in mind” when they answer the above questions about migrant rights. ¹⁰

Figure 1. Survey Treatments

The experimental treatments, photos of hypothetical foreigners, seek to reveal how implicit biases interact with different framings of Mexican identity. The experimental cues from the photographs were not accompanied by any other additional information regarding migrants. Respondents were not told about the racial or ethnic background of the person in the photo. Therefore, the priming is a suitable way to reveal respondents’ implicit biases against these groups. Considering the previous historical account about the presence of the Chinese in Mexico and how the process of national identity construction developed, we think it reasonable to assume that the Asian phenotype likely prompted respondents to think of foreigners of Chinese origin. The results must be interpreted with this assumption in mind.

The survey features a battery of 11 questions asking respondents what makes someone Mexican. Importantly, these questions were asked before the experimental assignment and before the series of questions on support for migrants’ rights. We use these 11 questions to identify relevant dimensions of Mexicanness, which we then use as independent variables to tackle the question of whether endorsing ethnic, civic, and nationalist dimensions of national identity influences opinions regarding the socio-political rights of foreigners.

Table 1 shows the responses to the questions. Importantly, some of the items included in the Mexico y el Mundo survey are standard questions in survey modules about national identity (for example, “respecting laws and institutions,” or “language” and “place of birth”). Others, however, are specific to the Mexican context (for instance, “defending Mexico in war”). Also, some questions typically included in widely used national identity modules in surveys were not asked in ours, most important among them, a question on citizenship. ¹¹ This omission is particularly important because access to citizenship can transcend the limitations to accessing rights that the ascriptive trait “place of birth” may entail.

Table 1. Importance for Mexican Identity

With these limitations in mind, it is interesting to observe that some traits show greater importance than others. The three most important constitutive norms for respondents are being born in Mexico, which is supported by a solid majority of Mexicans (77.27%); feeling proud of being Mexican (72.59%); and considering Mexico as the best country (62.68%). These top three traits correspond to what the literature describes as ascriptive norms (being born in Mexico); civic norms in the identity dimension (feeling proud of being Mexican); and a nationalistic, comparative belief (belief that Mexico is the best country) (Smith 1991; Schildkraut 2011; Reeskens and Hooghe 2010). In contrast, there are traits that respondents find “not at all” important, such as being Catholic (20%) and not supporting the United States (22%).

We use factor analysis to consolidate the 11 items into three dimensions of Mexican identity. Factor analysis is a tool widely used in the study of national identity. Confirmatory factor analysis enabled us to test the usefulness of the conceptual building blocks to address national identity in the case of Mexico. In our view, the use of factor analysis reduces arbitrariness when it comes to classifying the above items into conceptual dimensions. For the most part, the dimensions that this exercise returns make sense to us. Following Schildkraut’s (2011) procedure to produce variables that capture different constitutive norms of Mexican identity, the factor analysis justifies three latent variables on Mexican identity. Roughly, they correspond to ethnic, civic, and nationalist content (see the scree plot in the Appendix, Figure B3). The rotated factor loadings in Table 2 show which of the traits fall into each dimension with a 0.4 loading as the cutoff. ¹² Only two traits fall into the ethnic factor: being born in Mexico and speaking Spanish. ¹³ Five traits contribute the most to the civic dimension of Mexican identity: feeling proud of being Mexican, respect for patriotic symbols, defending Mexico in war, paying taxes, and respect for Mexican laws and institutions. Note that under this civic umbrella, most of the items refer to how a true Mexican should act (respect for patriotic symbols, defending Mexico in war, paying taxes, and respecting Mexican laws and institutions); but there is also one item concerning how a true Mexican should feel (being proud of being Mexican). ¹⁴ The nationalist factor groups the traits that allude to Mexican superiority and are explicitly comparative: thinking that Mexico is the best country in the world, preference for Mexican arts over those of other countries, and not supporting the United States. Catholicism contributes to the nationalistic factor, more so than to the ethnic dimension where a priori it would be expected to belong. Catholicism has a factor loading of 0.447 for the ethnic factor but a higher loading (0.511) on the nationalist factor. ¹⁵ Remember that as an individual item, being Catholic receives little support as constitutive norm of Mexicanness (Table 1).

Table 2. Factor Analysis

Next, for each dimension of national identity (ethnic, civic, nationalist), we created an additive index based on the items that contribute to each factor. Following Schildkraut (2011), we summed the responses and divided by the number of items for that factor. For example, the civic factor added together the responses from the five items that contribute to the factor (Table 2). The variables are scaled from 0 to 1, where 0 means that respondents said that all the items for the civic factor are not at all important, while 1 means very important. Table 3 presents the summary statistics for each national identity dimension. ¹⁶ The mean level of the ethnic factor is 0.85 and of the civic factor 0.82. Therefore, large majorities of Mexicans place a high importance on the ethnic and civic dimensions of what defines someone as Mexican. For the average Mexican, the nationalist dimension, while still important, weighs less than the civic and ethnic dimensions (0.61).

Table 3. Summary Statistics for Factors

The dependent variables, in turn, are measured using questions capturing levels of support for a number of rights, namely, (1) family reunification; (2) access to public education; (3) migrants’ right to form organizations; (4) and the right to vote. Table 4 presents the responses to these four questions. The responses are on a four-point scale from “strongly disagree” to “strongly agree.” The first question asks respondents to what extent they agree that migrants should have the right to bring their family over to live with them. Almost 80% of respondents either somewhat or strongly agree with family reunification. The second question addresses whether foreigners should have access to public education. Similar to family reunification, solid majorities of Mexican respondents show a high support for granting migrants access to education. The third question asks whether migrants should have the right to form organizations to defend their rights. The responses show greater variation than the previous two questions, with 66% of respondents showing support for foreigners having the right to form organizations. The final question asks whether foreigners should have the right to vote. Even though a majority of respondents support migrants having the right to vote, this question yielded the lowest support among the four policies under scrutiny (60%). Therefore, the average Mexican shows less support for granting migrants political rights than they do for granting access to social rights and policies, such as access to public education and family reunification. We rescaled the responses from zero to one, and standardized them to ease the interpretation.

Table 4. Support for Migrant Rights

In the regression analysis, we interact the indices of ethnic, civic, and nationalist dimensions of Mexicanness with the Asian treatment. This allows us to test whether and how the relationship between ethnicity (being primed with an Asian profile) and natives’ preferences for immigrants’ access to rights is moderated by the three constitutive norms of Mexican identity.

5. Results

5.1 Baseline Results: Race and Integration Policy Preferences

Figure 2 (based on results in Table B3) presents the baseline results from the survey experiment and the effect of migrant origin on support for migrant rights after random assignment to different ethnic treatments. Exposure to the Asian treatment reduces support for migrant rights and the effect is statistically significant for three of the four policies. Being primed with the Asian treatment reduces support for family reunification by nearly 0.2 standard units. The Asian treatment reduces support for migrants forming organizations and voting by 0.11 and 0.16, respectively. The results are robust to state fixed effects and to controlling for socio-demographic factors (see Tables B4–B5). In contrast, the White treatment, while negative in sign, does not have a statistically significant effect on support for migrant rights. ¹⁷ Exposure to the indigenous treatment reduces support for family reunification only. ¹⁸ In short, foreigners of an Asian profile are the only ones that do not receive public support when it comes to granting foreigners socio-political rights. Being exposed to one or another treatment does not explain preferences for access to public education, this being the only policy in which we do not observe statistically significant opposition (but not support either).

Figure 2. Treatment Effects by Migrant Rights

Therefore, the experiment provides valuable information in relation to an ascriptive aspect of Mexican national identity, namely, race. Consistent with the historical review in the previous section, there seems to be a hierarchy of ethnic origins for the Mexican respondent (Freier et al. 2020; Freier and Bird 2021). This hierarchy points to less support for ethnic Asians to enjoy socio-political rights in 2016 Mexico.

We next explore whether different dimensions of Mexican identity moderate the effect of the Asian treatment or not. In particular, does endorsing a civic approach to nationhood increase support for ethnic Asians enjoying rights or, on the contrary, does it reinforce opposition? Does the racial dimension render other ethnic traits, such as language and place of birth, irrelevant? And does exclusionary nationalism carry any weight in strengthening opposition to incorporation of Asians? Shedding light on these questions helps us to better understand the roots of the apparent anti-Asian sentiment reflected in the results above.

5.2 Interaction Results: Do Constitutive Norms Moderate Sinophobia?

The interaction models should reveal how different dimensions of Mexican identity moderate the effect of being exposed to specific profiles, and the support for migrant rights. We tested the interaction of the Asian treatment with each dimension of Mexican identity (ethnic, civic, and nationalist). Our argument proposes that opposition to Asians’ incorporation into Mexico is at least partly rooted in their exclusion (evident in the case of the Chinese community) from civic definitions of national identity. In other words, the deep-seated un-civic and derogative portrayal of the Chinese community, central to the crafting of the mestizo identity, suggests that endorsement of the civic constitutive norm does not necessarily translate into more support for granting immigrant rights today.

We can extract several ideas from the interactive models. ¹⁹ First, regarding the independent (not interactive) effects of treatments, being exposed to the Asian origin is the only profile consistently associated with less support for political incorporation and family reunification of immigrants (see Tables B6 and B7). Second, high levels of endorsement of the ethnic factor, which features place of birth and language as the main components, imply even less support for migrant rights to organize and vote, and for family reunification; but the interaction effects are not statistically significant. Figure 3 shows that at low levels of endorsement of the ethnic dimension of Mexicanness, the Asian treatment effect is not statistically significant for migrants’ right to organize, vote, and access public education. With respect to the right to create organizations and to vote, the interaction between the Asian treatment and the ethnic factor weakly reinforces the unconditional negative effect of origin. Support for family reunification is consistently negative and substantively similar across all levels of endorsement of the ethnic dimension of nationality (see Table B8 and B9 in the Appendix).

Figure 3. Marginal Effects for Asian Treatment by Degree of Endorsement of the Ethnic Factor

Third, while endorsement of the civic factor (not interacted) correlates with more support for migrants’ rights (Tables B6 and B7), ethnic Asians appear to be penalized by those embracing a civic conception of Mexicanness. Unconditionally, a one standard unit increase in the civic factor raises support for each of the four migrant rights by more than 0.10 standard units (Tables B10 and B11). The mean level of support for the civic factor was 0.82, which shows that Mexicans highly value civic norms of Mexican identity. Yet the negative interaction effect between the civic factor and the Asian treatment suggests that despite the great importance Mexicans place on the civic content of Mexican identity, endorsement of a civic conception of Mexican identity reduces, rather than increases, support for migrant rights. This effect is statistically significant in the case of family reunification (see Tables B10 and B11 in the Appendix).

Figure 4 shows that the Asian treatment effect remains negative even at one standard deviation above the mean level of endorsement of the civic factor. The fact that endorsement of civic norms, when interacted with the Asian treatment, is statistically significant in reducing support for granting family reunification rights, suggests the perception that these foreigners are perceived as not meeting civic standards of national membership. In all, this finding is aligned with the argument that in Mexico, civic nationhood and civic belonging were heavily racialized.

Figure 4. Marginal Effects for Asian Treatment by Degree of Endorsement of the Civic Factor

Finally, as Figure 5 shows, the effect of the Asian treatment does not change across different levels of support for the nationalist factor. Recall that this dimension appeared as the least relevant in the definition of Mexicanness. The nationalist factor only has a statistically significant relationship with decreasing support for access to public education (see Table B12 and B13). Yet, strongly endorsing this dimension somewhat increases the baseline support for granting family reunification rights to foreigners when interacted with the Asian treatment; but the effect does not reach statistical significance (see Table B12 and B13 in Appendix B).

Figure 5. Marginal Effects for Asian Treatment by Degree of Endorsement of the Nationalism Factor

In sum, race, typically included in ethnic dimensions of nationality, but addressed here with a survey experiment, shows that Sinophobia is a candidate for predicting less support for foreigners of Asian ancestry enjoying a number of important socio-political rights. The ethnic factor, which includes place of birth and language, weakly contributes to making the Asian treatment stronger. To us, the most important finding from a policy point of view is the fact that despite the substantial and positive independent effect that endorsing a civic conception of Mexicanness has on supporting foreigners’ socio-political rights, when prompted to think about foreigners of Asian profile, the civic dimension seems not to reduce the negative impact of racial discrimination. On the contrary, it reinforces it.

Finally, a legitimate question is whether the above patterns pertain to the Asian origin only. And indeed, this seems to be the case. As shown in Figure 6, the legacies of historically constructed Sinophobia appear to be a convincing explanation for contemporary anti-Asian sentiment. Consistent with the historical review, neither indigenous nor White ethnicities raise similar misgivings today. Recall that in the construction of the mestizo project, indigenous peoples were vindicated and therefore incorporated into the mestizo project as “subjects of improvement” (Chang 2017, 10). White ethnicities, in turn, always ranked high in the hierarchy of desirable races to mix with (Freier and Bird 2021). As Chang puts it, “anti-Americanism was anti-Imperial […] but did not constitute a racial threat” (p.123) that could endanger the post-Revolution racial state.

Figure 6. Sinophobia, Xenophobia, and Preferences for Socio-Political Rights.

Figure 6 shows the interaction effects between different race treatments and the ethnic factor of national identity based on the results in Table B8. The marginal effects for the White and Central American treatments are not statistically significant across different levels of endorsement for the ethnic factor. ²⁰ In other words, when it comes to granting immigrants socio-political rights, only the Asian profile is penalized; and contemplating other ethno-cultural traits (place of birth and language) strengthens this finding only in the case of being exposed to the Asian treatment. ²¹

To give further validity to our argument, the online Appendix discusses and Tables B14–B19 show that our findings are robust to a number of alternative hypotheses of determinants of attitudes towards immigrant integration, such as the respondent’s occupation (to capture fears of labor market competition) and geographic location (to capture historical settlement patterns); respondents’ perception of their own and the national economy; respondents’ first-hand knowledge of foreigners; political ideology; and respondents’ residence in a state with varying presence of Chinese foreign investment. Most of these variables are not statistically significant when interacted with the Asian treatment. High levels of Chinese foreign investments, when interacted with the Asian treatment, if they have any effect, appear to increase support for foreigners’ access to rights.

6. Conclusion

We studied how different conceptions of national identity can explain contemporary preferences towards immigrants’ access to socio-political rights in Mexico. The article is novel in that to our knowledge, it is the first one to explore the content of Mexican identity empirically using a survey experiment; it traces back the content of constitutive norms to the historical process that led to an exclusionary, overtly racial construction of Mexicanness; and it shows how these norms appear to be consequential to understanding today’s unenthusiastic support for granting socio-political rights to particular groups of foreigners.

To summarize, in contemporary Mexico, both ethnicity and behaviors matter to explain natives’ attitudes towards the Chinese. Asian ancestry, an ascriptive trait, is associated with less support for granting foreigners political rights and the right to bring their families with them. Since civic belonging was racialized in the construction of Mexican identity—the Chinese were considered unmalleable, untrustworthy, and a threat to community life—the positive impact that civic norms often have on natives’ support for immigrant integration does not show for this outgroup. Interestingly, other groups of foreigners do not evoke the same misgivings. We argue that in contemporary Mexico, there is a rationale to interpret these findings under the lenses of the convulsive process and period in which Mexican identity was crafted.

The paper has, of course, limitations. It does not intend to provide a definitive answer to contemporary Sinophobia; but rather suggests interesting paths to explore. A logical extension to this work would be to test whether other origins, such as Black, would be the object of a similar racism. During the period of national identity construction, Blacks, Jewish, Syrians, and other nationalities and ethnicities were considered equally pernicious and undesirable from the point of view of eugenics (Fitzgerald and Cook-Martin 2014, 219; Chang, 2017, 22). While our expectation would a priori be a similar unenthusiastic response to the socio-political inclusion of these other groups, a different finding would point to a distinctive, deeply ingrained dislike of Asian ancestry, which would reinforce our argument. We shall address this hypothesis in future research.

A line of research that tentatively emerges from this paper is that contrary to elites’ nationalist rhetoric, the average Mexican respondent may not perceive new Chinese investment as threatening (Maldonado et al. 2018). Our robustness tests suggest that in those states where Chinese investment increased in the most recent years, there is more support for granting this community socio-political rights. Sinophobia may thus be more an elite rhetoric with little echo among the public, who welcomes the cheaper imports, investment projects, and jobs that Chinese economic involvement brings. Whether material interests may counter identity concerns is certainly another line that deserves further scrutiny.

The policy implications of our findings are important. When identities continue to uphold racial discrimination, we are likely to observe high levels of resentment, which harms social cohesion (Schildkraut 2011, 2014; Theiss-Morse 2009). While in other contexts, civic acting and feeling may counter the resistance that ascriptive conceptions of nationhood may arouse (Green et al. 2011), this is unlikely to be the case in Mexico. If civic constitutive norms of nationhood may not counter ethnic discrimination, what can contribute to combating it? From a policy perspective, while the existence of laws can indeed protect certain groups, it may be necessary to remain vigilant with respect to their actual implementation, as well as to proactively protect the rights of foreigners who appear more vulnerable to xenophobia.

Finally, the story that we unpack here is relevant for other Latin American cases where the construction of national identity similarly revolved around the explicit exclusion of specific racial minorities (Foote and Goebel 2014; López 2014; Fitzgerald and Cook-Martin 2014; Chan and Montt Strabucchi 2021). Two other recent trends will contribute to making the study of Latin American national identities a research priority: the increase in immigrant presence—particularly forcibly displaced population—in the region and the resulting ascendance of immigration as a salient political issue; and the controversial, growing influence of Chinese economic presence in countries such as Peru, Brazil, Colombia, and Chile (Armony and Velásquez 2015; Liang 2019; Ratigan 2021). This work contributes to paving the way for novel empirical research on these pressing topics and their connection with deep-rooted conceptions of national membership.