The Integration Paradox

Abstract

The integration paradox refers to the phenomenon of the more highly educated and structurally integrated immigrants turning away from the host society, rather than becoming more oriented toward it. This article provides an overview of the empirical evidence documenting this paradox in the Netherlands. In addition, the theoretical arguments and the available findings about the social psychological processes involved in this paradox are considered. The existing evidence for the integration paradox and what might explain it form the basis for making suggestion for future theoretical work and empirical research, and for discussing possible policy implications.

The so-called “immigrant paradox” has received considerable attention in the literature on immigrant integration. This paradox concerns the notion that first-generation immigrants and less acculturated individuals are sometimes found to outperform second generations and more acculturated individuals on a variety of adaptation outcomes, such as academic achievement and health behaviors (see Garcia Coll et al., 2012). While originally documented in research in the United States (e.g., Fuligni, 1998; Palacios, Guttmanova, & Chase-Lansdale, 2008), studies also have shown these differences with respect to sociocultural adaptation outcomes in several European countries that have experienced large-scale immigration in recent decades (Sam, Vedder, Liebkind, Neto, & Virta, 2008; van Geel, & Vedder, 2011).

Another phenomenon regarding immigrant integration that has received much less attention in the research literature is the integration paradox. This paradox describes the situation of the economically more integrated and highly educated immigrants turning psychologically away from the host society, instead of becoming more oriented toward it (Buijs, Demant, & Hamdy, 2006; Entzinger & Dourleijn, 2008). Media reports in Western Europe have suggested that policy makers should be concerned about relatively highly educated immigrants becoming disenchanted and disengaged from the host country. Education and structural integration would not always lead to a stronger orientation on the host country but rather can have the opposite effect. The current article provides an overview of the empirical research on the integration paradox conducted in the Netherlands. To my knowledge, there is no other country in which this paradox has been investigated systematically.

Theoretically, the paradoxical nature of highly educated immigrants’ disengaging from society is important and interesting. Classical immigration theories suggest that structural integration (improving one’s educational and economic position) will be conducive to other forms of integration, such as developing a sense of belonging and a more positive attitude toward the host society (Alba & Nee, 2003; Esser, 2001; Gordon, 1964). Yet education could also be an obstacle for developing positive attitudes toward natives and the host society. A key reason for this might be that higher educated immigrants feel relatively deprived.

Relative Deprivation

Relative deprivation concerns the perception that oneself or one’s group is at an unfair disadvantage in comparison with others (Pettigrew et al., 2008; Smith, Pettigrew, Pippin, & Bialosiewicz, 2012). Feelings of relative deprivation contain three aspects (Smith et al., 2012). First, there must be comparisons made at the individual or group level. Second, the comparison must lead to the perception that one is at a relative disadvantage with respect to other individuals or groups. Third, the perceived disadvantage should be seen as being unfair. There are a number of reasons to expect that feelings of relative deprivation are higher among immigrants who are more integrated structurally, especially in relation to their level of education.

First, within the relative deprivation framework, it has been argued that the more advantaged members of disadvantaged groups are most likely to engage in intergroup comparisons (Taylor & Moghaddam, 1994). Thus, higher educated immigrants may feel more deprived because they increasingly compare their situation and opportunities with majority members. Higher education has been found to increase immigrants’ contact opportunities and actual contacts with majority members (Kalmijn & van Tubergen, 2006; Martinovic, 2013), which makes the majority a more relevant comparison group. In addition, in most Western countries, the level of unemployment is much higher among immigrant-origin groups compared with majority members, independently of the educational level. Furthermore, compared with similarly educated majority members, immigrants tend to have lower level employment and more temporary jobs (Alba & Nee, 2003; Hall & Farkas, 2008; Kogan, 2006). Thus, higher educated immigrants may feel more deprived because the relevant comparison with similarly educated majority members turns out unfavorably.

Second, the theory of rising expectations suggests that immigrants who pursue higher education and try to attain a good position in society develop higher expectations. They, therefore, are more strongly disappointed about unequal opportunities and treatment, whereby their higher expectations are not met with equal rewards (Entzinger & Dourleijn, 2008). The higher educated tend to be more sensitive to acceptance and rejection by the majority population. In contrast to the lower educated, they can more confidently claim that a lack of opportunities and discrimination, rather than a lack of efforts and skills, prevents them and members of their group from gaining economic parity with natives.

Third, higher education implies higher cognitive sophistication which can mean that the higher educated immigrants are more aware of, and have a better understanding of, processes of discrimination and reduced opportunities in society (Kane & Kyyro, 2001; Wodtke, 2012). Education enables immigrants to become more informed social critics who can seek to challenge discrimination and advocate policies that redress group disadvantages.

Education and Perceived Discrimination

The “integration paradox” implies that the higher educated feel relatively deprived and as a result distance themselves psychologically from the host society (see Figure 1).

Figure 1.

The relations involved in the integration paradox.

The first link in this process involves the relation between education and relative deprivation (Path 1 in Figure 1). This link has mainly been examined in terms of perceived discrimination, opportunities, and feelings of acceptance. For immigrants, perceptions of discrimination and lack of opportunities in the host society combine the different aspects of relative deprivation. When immigrants have the sense of being discriminated against, they will feel that they have an unfair disadvantage relative to members of the majority group, either personally or as a group. And when immigrants have the sense that their group lacks opportunities to succeed economically and to freely enjoy their social and cultural life, they compare their position with the opportunities that are open to other groups in society and the majority in particular. This means that it can be expected that in particular experiences and perceptions of nonacceptance and discrimination, despite one’s efforts and achievements, lead immigrants to distance themselves from society. Thus, the more successful ones would be more sensitive to ethnic acceptance and equality, which in turn would drive their reactions to the host society.

In support of Figure 1 (Path 1), research has found a positive relation between level of education and perceived discrimination among immigrants in, for example, North America and New Zealand (Organization for Economic Co-operation and Development [OECD], 2012; Sigelman & Welch, 1991; Sizemore & Milner, 2004; Wodtke, 2012). Research in the Netherlands has also provided empirical evidence that the higher educated tend to perceive more discrimination and lower societal acceptance. This has been found in large samples of the four main immigrant-origin groups (Turks, Moroccans, Surinamese, and Antilleans; Gijsberts & Vervoort, 2007; Tolsma, Lubbers, & Gijsberts, 2012), four refugee groups (Muller, 2011), and a sample of smaller immigrant groups (Afghani, Iraqi, Irani, Somali, Polish, and Chinese; van Doorn, Scheepers, & Dagevos, 2013). Thus, there is among a variety of groups clear empirical evidence for the higher educated perceiving more discrimination and less societal acceptance of immigrants.

The higher perception of discrimination among higher educated immigrants is not only in the “eye of the beholder” but can be based on actual experiences. The theory of exposure suggests that the higher educated are actually experiencing more discrimination and lower acceptance in everyday life. The higher compared with the lower educated more often use host country media and tend to have more negative contacts with majority members on the labor market and in associations, and therefore are more exposed to discrimination and derogating messages. For example, in a study among the four largest immigrant-origin groups in the Netherlands, it was found that the higher educated are more interested in the often quite negative Dutch political immigration and integration debate (van Heerden, de Lange, van der Burg, & Fennema, 2014; Vasta, 2007) and therefore have a less positive attitude toward Dutch society (Gijsberts & Vervoort, 2007). In another study among smaller immigrant groups, it was found that the link between educational attainment and discrimination can be explained by the higher educated participating more in associations, being exposed more to Dutch politics in the media, and experiencing that their foreign education is not valued and that their job is below their educational level (van Doorn et al., 2013).

Discrimination and Host Society Disengagement

As shown in Figure 1, the second link in the processes involved in the integration paradox concerns the association between relative deprivation and host society disengagement. It has been shown that individuals who perceive more group relative deprivation have more prejudiced attitudes toward out-groups (Pettigrew et al., 2008; Pettigrew & Meertens, 1995), and less positive attitudes toward the social system, such as reduced just-world beliefs, reduced confidence in political institutions, and higher support for political protest (Corning, 2000; Klandermans, Roefs, & Olivier, 2001; Pettigrew et al., 2008).

Among immigrant-origin groups, various studies have demonstrated that experiences with discrimination and low public acceptance instigate processes of disengagement. For example, the “rejection-identification model” argues that being a target of discrimination leads individuals to identify more strongly with their ethnic minority group and distance themselves from the majority (Branscombe, Schmitt, & Harvey, 1999). Experimental and longitudinal evidence has shown that perceived threats can indeed increase minority group identification and lead to greater emotional attachment to one’s minority group (e.g., Jetten, Branscombe, Schmitt, & Spears, 2001). Furthermore, research among immigrant-origin groups shows that increased perceptions of discrimination predicts increased ethnic group identification. This has been found, for example, in three studies among Turkish Dutch people (Verkuyten & Yildiz, 2007), in a longitudinal study among immigrants in Finland (Jasinskaja-Lahti, Liebkind, & Solheim, 2009), and in a study among Latino students in the United States (Cronin, Levin, Branscombe, van Laar, & Tropp, 2012). Furthermore, this type of research has shown that higher perceived discrimination also is associated with weaker identification with the host society (Jasinskaja-Lahti et al., 2009). For example, in a large-scale research among the four main immigrant-origin groups in the Netherlands, it was found that perceived discrimination was the strongest negative predictor of host national identification (de Vroome, Verkuyten, & Martinovic, 2014; see also Tolsma et al., 2012).

Yet low host national identification or little commitment to the host nation does not have to indicate host society disengagement, whereby there is an adversarial stance in which the host society is subjectively rejected. Low national identification implies that aspects of the host society are not strongly connected to oneself but does not have to consist of disconnecting these aspects from oneself (Kreiner & Ashforth, 2004). Disidentification is not merely the opposite of identification. The former consists of immigrant minorities reacting against the host society and developing a so-called reactive or oppositional identity in which people actively separate their minority identity from the culture and defining aspects of the dominant group (Ogbu, 1993; Portes & Zhou, 1993). Furthermore, studies in organizational contexts have shown that disidentification is a different psychological state than identification (e.g., Elsbach & Bhattacharya, 2001; Kreiner & Ashforth, 2004). Hence, perceived social rejection and devaluation might not only result in decreased identification with the host society but also increased disidentification with the nation. In a survey study among Turkish Dutch people and using factor analysis, we found that the participants made a clear empirical distinction between the construct of Dutch identification (e.g., “I identify with the Dutch,” “I feel connected to the Dutch”) and Dutch disidentification (e.g., “I would never say ‘we Dutch’,” “I certainly do not want to see myself as Dutch”; Verkuyten & Yildiz, 2007). Furthermore, a stronger Turkish identity was associated with a weaker Dutch identification and independently also with a stronger disidentification with the Dutch.

In a recent study among Turkish immigrants in Germany and the Netherlands, we again found a clear empirical distinction between the two constructs of host society identification and disidentification. Furthermore, higher perceived discrimination was associated with stronger disidentification in Germany (N = 363, r = .19, p < .001) and in the Netherlands (N = 425, r = .15, p = .003), while discrimination was also negatively and independently associated with identification in Germany (r = −.16, p = .002; Netherlands, r = −.06, p > .05). Thus, perceived discrimination is not only associated with a lower sense of belonging to the host society but is also related to stronger psychological disengagement from the host society (Path 2, Figure 1).

Education, Discrimination, and Disengagement

The integration paradox implies a relation of mediation, whereby the higher educated feel relatively deprived and as a result disengage psychologically from the host society. Thus, there should be an indirect negative relation between education on the one hand and favorable attitudes toward the host society and natives on the other, via indicators of relative deprivation (Figure 1).

We investigated these associations in a study among the four main immigrant-origin groups in the Netherlands (de Vroome, Martinovic, & Verkuyten, 2014). The samples were randomly drawn from the population registry and we analyzed the data of 3,981 first- and second-generation participants. We examined feelings of relative deprivation, in terms of the perception of discrimination and subgroup respect. Discrimination implies unfair treatment and such treatment tells immigrants that they are not equal members of society. Subgroup respect refers to immigrants’ sense of whether the majority population recognizes and values their minority group and immigrants more generally (Huo & Molina, 2006). Thus, we focused on perceived negative reactions (discrimination) and perceived positive reactions (respect) toward immigrants as two indicators of relative deprivation. Furthermore, we also tested whether education in the host country and in the country of origin are differently related to perceived discrimination and respect. According to the integration paradox, immigrants perceive that they themselves or members of their group get lower returns for exactly the same educational investments in the host society. The native population is a relevant comparison group for immigrants who are educated in the host society (Zagefka & Brown, 2005) and probably less so for those who were educated in their country of origin.

In the analysis, we first tested the measurement models, and confirmatory factor analyses indicated that perceived discrimination and subgroup respect were two separate latent constructs. In addition, tests of cross-group measurement equivalence indicated that the two latent constructs were sufficiently comparable across the four immigrant-origin groups and the two generations. So the hypothesized relations could be validly compared across groups and generations (see de Vroome, Martinovic, et al., 2014, for details).

In the second step, we tested the structural model with the expected relations and investigated the robustness of the model across immigrant groups and generations. In doing so, we controlled statistically for age, years since migration, gender, and occupational status. The findings supported the indirect relationships proposed in the mediation model (for details, see de Vroome, Martinovic, et al., 2014). Higher education was positively related to perceived discrimination, and perceived discrimination, in turn, was negatively related to favorable attitudes toward the host society (“level of satisfaction with Dutch society”) as well as the native population (“general feeling”). In addition, higher education was negatively related to perceived respect which, in turn, was positively related to favorable attitudes toward the host society and the native population. The proposed mediation model was found to be similar for the two generations and found support among all four immigrant-origin groups. Furthermore, additional analyses (“curve estimation”) did not show significant quadratic or other linear effects for education on the attitude toward the host society and native population. Thus, there was no evidence that, for example, the lowest and highest educated are more negative toward the Dutch than those with a moderate level of education.

Importantly, the findings further showed that higher education was associated with lower perceived respect among immigrants who were educated in the Netherlands, and not among those educated in the country of origin. Thus, the level of host country education rather than origin country education was associated with a perceived lack of respect. Origin and host country education were similarly related to perceived discrimination. However, compared with perceived lack of respect, discrimination was, in turn, less strongly associated with favorable attitude toward the host society. Thus, the integration paradox seems most applicable to immigrants who have invested in host country education. The education of majority members is probably a less relevant standard of comparison for immigrants who are educated in the country of origin.

Education, Discrimination, and Positive Social Contacts

Higher education has been found to increase immigrants’ contact opportunities and actual positive contacts with majority members (Kalmijn & van Tubergen, 2006; Martinovic, 2013) which might lead to developing a more positive attitude toward the host society and majority group members (Pettigrew & Tropp, 2006). The integration paradox, however, implies that higher educated immigrants will have more negative attitudes toward the host society and the majority population. This might mean that there are two pathways. On the one hand, higher educated immigrants have more frequent positive contacts with the majority population than lower educated immigrants, and more contact is associated with a more favorable attitude toward the majority population. On the other hand, higher educated immigrants will feel less accepted in the country and perceive more discrimination (negative contact) than lower educated immigrants, resulting in a less favorable attitude toward the majority population.

We examined these two pathways in another study among different large samples (N = 4,199) of the same four immigrant-origin groups in the Netherlands (ten Teije, Coenders, & Verkuyten, 2013). Thus, we tried to replicate the findings of the previous study by using a different attitude measure (extent to which participants considered the native Dutch as honest, polite, helpful, hospitable, neat, tolerant, and friendly; Cronbach’s alpha = .76) and by taking positive social contacts into account. For these samples, there was only information available about education in the Netherlands and not in the country of origin. Participants were again randomly selected from registers of various municipalities. As indicators of relative deprivation, we focused on perceived acceptance and perceived discrimination, against oneself as a minority member (personal discrimination) and against one’s group in general (group discrimination).

In a first step of the analysis, we confirmed (confirmatory factor analysis) the empirical distinction between these constructs, established measurement equivalence across the four immigrant groups, and found no significant quadratic or other nonlinear effects of education on the attitude toward the native Dutch. In the second step, we tested the proposed structural model that is shown in Figure 2. We controlled statistically for age, years since migration, generation, gender, occupational status, and Dutch language proficiency. The findings supported the indirect relationships proposed in the mediation model (Figure 2; for details, see ten Teije et al., 2013). For the first pathway, the findings demonstrated that higher educated migrants indeed had more contacts with the native Dutch and that contact was associated with more positive attitude toward the native Dutch. More important, there was also empirical evidence for the second pathway. Higher educated immigrants perceived lower acceptance of ethnic minority groups in Dutch society and more group discrimination than lower educated migrants. These perceptions were related to less positive attitudes.

Figure 2.

Path diagram with standardized direct and indirect effects of education on immigrants’ attitude toward the native Dutch.

Note. Circles indicate latent variables and squares manifest variables.

**p < .01. ***p < .001.

In addition, the unfair treatment of co-ethnics (group discrimination), rather than personal experiences with discrimination, was found to be important for the attitude toward the Dutch. This supports the idea that the more advantaged members of disadvantaged groups tend to engage in group comparisons and develop more negative attitudes toward the advantaged group (Taylor & Moghaddam, 1994). Higher education might increase one’s awareness of and concerns about the vulnerable and relatively marginal position of immigrants in society. In addition, when they perceive and experience ethnic discrimination, higher status minority members might be more assertive (Baumgartner, 1998). Although we did not explicitly test these interpretations, we did find that the higher educated perceived more group discrimination than the lower educated, whereas there was no difference for personal discrimination. In turn, and independently of personal discrimination, higher perceived nonacceptance of migrant groups and of group discrimination were associated with a less favorable attitude toward the native Dutch. The proposed mediation model was found to be quite similar for the four immigrant-origin groups with most of the relations being comparable in direction and also in size.

Future Directions

The “integration paradox” indicates that the more highly educated immigrants will show relatively high levels of dissatisfaction with the host society and more negative attitudes toward the native majority. Although research in the Netherlands has made clear progress in documenting and understanding the integration paradox, many important questions remain. A first question is whether the processes involved in the integration paradox are limited to host society attitudes and immigrants’ feelings of belonging. There are reasons to expect that this is not the case and that it also involves, for example, behavior. A study showed that Moroccan Dutch youngsters who are on custody awaiting trial are much better integrated in Dutch society than their Moroccan Dutch noncriminal peers (Stevens, Veen, & Vollebergh, 2009). Relative to the latter, the former group more often speaks the Dutch language fluently, has more contacts with Dutch people, and self-identify more often as Dutch. Their relatively strong orientation on Dutch society makes them extra sensitive to inequalities and negative stereotypes. This increases the likelihood of feelings of relative deprivation with the associated negative emotions of anger, resentment, and frustration that can lead to crime (Agnew, 2001; Smith et al., 2012). Similarly, several studies in the United States have found relative high levels of delinquency among ethnic minority youth who have a strong orientation on American society (Samaniego & Gonzales, 1999; Vega, Gil, Warheit, Zimmerman, & Apospori, 1993).

A second question relates to the underlying processes. There can be various reasons for why immigrants’ higher structural integration is associated with weaker feelings of identification and belonging in the host society. I focused on one important reason and that is that immigrants who have achieved higher levels of education also develop higher expectations and therefore are more susceptible to feelings of relative deprivation. However, the research discussed did not directly examine these feelings but used perceptions of discrimination, acceptance, and subgroup respect as indicators. This means that the role of feelings of relative deprivation should be examined more fully and systematically. Feelings of relative deprivation depend on the individual and group comparisons that people make and typically involve anger and resentment (Smith et al., 2012). This means that future studies could examine justice-related feelings and emotions directly and examine whether these more fully explain the association between education and host society disengagement.

A third issue is that the research discussed relied on cross-sectional data and therefore we do not know the causal directions of the proposed relations. The path model (Figure 1) was theoretically derived and other research has shown, for example, that perceived discrimination drives immigrants’ attitudes toward the host society (Jasinskaja-Lahti et al., 2009), but different and mutual directions of influence are possible. Longitudinal panel data are needed for examining the integration paradox more fully. These kind of data are increasingly collected in North America and many European countries and it might be possible to investigate the over-time associations involved in the integration paradox by using existing panel data, or to incorporate some measures in new data collections.

The research discussed is from a single country in Western Europe. Therefore, we do not know the generalizability of the findings to other countries. Although there is some qualitative evidence for similar processes in small-scale studies among minority groups in the U.S. context (e.g., Baumgartner, 1998; Ogbu, 1993), there might be country differences that moderate the associations found. For example, large-scale research has shown that in the Netherlands, Canada, and New Zealand, discrimination is more often reported by the highly educated immigrants compared with the low educated, while in other European countries, the low educated report more discrimination (OECD, 2012). The reasons for these differences are not clear but might have to do with the fact that European countries do not consider themselves as immigration societies such as Canada, the United States, and New Zealand. Yet this cannot explain why there is evidence for the integration paradox in the Netherlands but less so in other European countries (see Schneider, 2013).

Not only country characteristics are important to investigate in future studies on the integration paradox but also differences between and within immigrant-origin groups. In the Netherlands, evidence for the processes behind the paradox was found among quite different groups, including relatively large groups that have a history of migrant labor, ex-colonial groups, and more recent and smaller groups that migrated for various reasons. This indicates that the phenomenon is quite robust and not specific for one particular group. Yet there might be meaningful differences between immigrant-origin groups within other countries that have an impact on the integration paradox. For example, it might be the case that feelings of entitlements and relative deprivation are stronger among immigrant-origin groups that have a long historical relationship with the country of settlement, such as Pakistani and Bangladeshi in the United Kingdom. Additionally, the economic and cultural (mis)match between the host country and country of origin might be important for how immigrants perceive their situation, feel about the reactions of the majority population, and emphasize the importance of education.

There are important differences within each particular immigrant-origin group. The fact that people originate from the same country tell us something, but not a whole lot. Although we found no generational differences in the integration paradox processes, there are other demographic differences that might be important, such as gender, age, length of residence, and region (city, village) of origin and settlement. There also are important differences in the immediate social context (neighborhood, community organizations, schools) and the family context in which immigrants from a particular country of origin are embedded. And there are differences in political orientations, ideological beliefs, and religiosity that can be relevant for how immigrants perceive and evaluate their situation and the opportunities and barriers in the host society. Future research could examine whether and how these kinds of differences within an immigrant-origin group are relevant for understanding the processes and conditions involved in the integration paradox.

Conclusion

In addition to the “immigrant paradox” (Garcia Coll et al., 2012), the integration paradox presents another surprising and troublesome phenomenon. This paradox relates to what happens when immigrants and their children advance their educational and socioeconomic position. Classical immigration theories as well as policy makers and politicians typically argue that education is the road to other forms of integration, including a sense of belonging and a positive attitude toward the host society. There are many important reasons for being concerned about educational arrears or underachievement of immigrants and for putting every effort in trying to improve their educational outcomes. Yet educational investments without paying close attention to processes of discrimination and feelings of relative deprivation can lead to a distancing or disengagement from society. Integration policies that focus on structural integration will not necessarily be successful in developing a sense of belonging and a positive attitude toward a host society. Questions of relative deprivation, and of ethnic discrimination and subgroup respect in particular, also need to be addressed. Higher education and structural integration more generally make immigrants more aware of discrimination in society and more sophisticated social critics. There is a serious problem regarding the perceived (equality of) opportunities in society for those immigrants who meet the educational requirements and make a strong effort to integrate, but find that key positions in society are not open to them. Not really being accepted or being treated as a second-class citizens despite one’s educational investments and efforts undermines the development of a sense of belonging to the country of settlement and a positive attitude toward the dominant majority. The lack of feelings of belonging can put a strain on the cohesion of society because such feelings are generally considered a prerequisite for national solidarity, a unified society, and effective democracy. Conditions that hamper immigrants’ orientation to the host society should therefore be addressed and research on the integration paradox can make a contribution for doing so.

Author Biography

Maykel Verkuyten is a professor in Interdisciplinary Social Science at the Faculty of Social and Behavioral Sciences, Utrecht University, the Netherlands. He is also the academic director of the European Research Centre on Migration and Ethnic relations at Utrecht University. He has published widely on questions related to ethnic identity and cultural diversity.