Discrimination and Reactive Transnationalism Among Recent Italian, Polish, Turkish, and Syrian Immigrants in Germany

A reactive form of transnationalism

Several recent articles examine the possibility of a reactive transnationalism (AlMasarweh 2022; Herda 2025b; Itzigsohn & Saucedo 2002, 2005; Redclift & Rajina 2021; Snel, Hart & Van Bochove 2016). This involves transnational practices being triggered or accelerated after immigrants encounter a hostile context of reception. In other words, those experiencing negative or unfair treatment respond by exhibiting greater engagement with their origin society. The assumption is that the unfairness of these experiences causes immigrants to question their place in the receiving society and whether they want to participate fully in it. Decades of research document a long list of emotional, physical, and behavioral consequences for discrimination victims in general (Mays et al. 2007; Pascoe & Richman 2009; Smith, Allen & Danley 2007; Williams et al. 2003). The impact should certainly be enough to motivate immigrant disengagement from their receiving society and re‑engagement with their origin.

The reactive transnationalism concept brings together transnationalism with the idea of reactive ethnicity from Segmented Assimilation Theory (Haller, Portes & Lynch 2011; Portes & Rumbaut 2001; Portes & Zhou 1993; Zhou 1997). This focuses on how a hostile receiving society context can cause immigrants to rethink their identities. The more discrimination that they encounter, the more likely they will identify with their origin society or develop identities in opposition to the receiving society (Celik 2015; Chaney 2022; Doering & Peker 2022; Herda 2018; Herda 2025a; Jiang et al. 2016; Verkuyten & Yildiz 2007). Reactive transnationalism takes this a step further by considering a behavioral consequence. Transnationalism, of course, can exist without discrimination. It is only reactive if it develops in response to receiving societal hostility.

The recent reactive transnationalism research considers multiple samples and datacollection methods, each time finding evidence for the hypothesized pattern. Itzigsohn and Saucedo (2002, 2005) conducted early survey research among US‑based Colombian, Salvadoran, and Dominican immigrants. Reported discrimination experiences were significantly associated with more remittances sent to the origin, increased involvement in origin culture organizations, and more travel to the origin. Snel et al. (2016) found a reactive transnationalism pattern among middle‑class Turkish, Moroccan, and Surinamese respondents in the Netherlands: Those reporting discrimination experience were more likely to maintain contacts with their origin society, remit money, and participate in political action connected to their origin society. In South Korea, Yoo and Jang (2024) found reactive transnationalism among immigrants who perceive their workplace conditions as unfair relative to native‑born Koreans. Specifically, perceived unfair treatment moderated the association between human capital and remittances. Those with higher education and income are more likely to remit when they considered their workplace as unfair.

Herda (2025b) confirmed similar patterns among Latin American immigrants in the US. However, this study also provided evidence linking the rise of the anti‑immigrant Trump administration to both increased remittances and increased transnational contacts. The results showed that discrimination existing at the structural level can generate reactive transnationalism independently. An interpersonal experience with discrimination was not required.

Qualitative research provides further evidence for reactive transnationalism. In AlMasarweh’s (2022) sample of 32 second‑generation Muslims in the US, more discrimination was related to greater involvement in local mosques, which connected them to their origin culture. Studies by Redclift and Rajina (2021) and Dorzadeh (2022) considered responses to discrimination among Bangladeshi immigrants in the UK and Baloch immigrants in Qatar, respectively. Both found that the hostility they encountered motivated property investments in their origin. The two studies understood these purchases as a protective contingency, allowing migrants to eventually leave the hostile host country.

The current study replicates this research in the German context, which is an important site for understanding integration patterns. Moving beyond replication, this study also considers different types of discrimination—experienced and perceived—as possible motivators. This permits consideration of whether experience is necessary for reactive transnationalism.

Is experience needed for reactive transnationalism?

Most reactive transnationalism research focuses only on first‑hand discrimination ex‑periences. The current study assumes that these are important but questions whether they are required. This expectation follows research using General Strain Theory to understand the connection between discrimination and criminal behavior (Agnew 1992; Herda 2016, 2021; Herda & McCarthy 2018; Unnever & Gabbidon 2011). The theory acknowledges that discrimination can impact individuals in indirect ways, even when one is not a first‑hand victim. For instance, anticipated discrimination involves the fear that one may fall victim to discrimination. This indirect variety has proven consequential. In Gordon et al.’s (2020) research, they found that, when respondents expected discrimination in a hypothetical scenario, it was enough to worsen sleep quality as measured through the Pittsburgh Sleep Quality Index. While discrimination experience mattered, the anticipation effect also held net of controls, suggesting that anticipated discrimination is independently associated with sleep‑related consequences.

Also important is vicarious discrimination, which involves witnessing or hearing about the experiences of others. For instance, Wofford, Defever, and Chopik’s (2019) research on romantic couples finds that the partner’s discrimination experiences are associated with lower self‑reported health, higher depression, and greater relationship strain for the respondent of focus. These associations existed net of discrimination experienced by the respondent, suggesting that experience is not needed to bring about emotional and relational consequences. However, is the mere perception that discrimination exists associated with transnational behaviors, even without direct experience?

Notable among the reactive transnationalism research is Itzigsohn and Saucedo’s (2002) article focusing on Latin American immigrants. Like the current study, this research considers discrimination experience as well as the perception that one’s group is victimized in the receiving society. They measured the latter through an index, which included questions asking whether relations between White Americans and one’s own group are friendly, cold, or distant. These questions did not measure discriminatory encounters but rather their perception that overall relations are hostile. Their index also contains measures of perceived general economic discrimination—not necessarily toward one’s group—and whether the American lifestyle strengthens families. The inclusion of these other components complicates interpretation, as the variable measures something beyond what the current study seeks to isolate. Regardless, in their multivariate models, both experienced and perceived group discrimination were associated with sociocultural transnationalism (Itzigsohn & Saucedo 2002). Significant coefficients for both net of one another suggest that reactive transnationalism may not require a first‑hand encounter with receiving‑society hostility.

Regardless, simultaneous examinations of experienced and perceived discrimination are rare. Thus, the current analysis includes a measure of perceived unfairness in the receiving society directed toward one’s group as an independent variable alongside discrimination experience. Including these simultaneously allows consideration of whether one or both are linked to transnational behaviors. Both measures are also used in interactions with immigrant origins to determine if the reactive transnationalism association varies across groups.

Is reactive transnationalism consistent across origin groups?

The ENTRA data include four immigrant origin groups: Italians, Poles, Turks, and Syrians. Each presents important cultural differences compared to native‑born Germans. They also vary in status, with only Italians and Poles being EU citizens within the Schengen Zone. This brings extensive rights and privileges, including the freedom of movement. Combined with geographic proximity, it further advantages Italians and Poles in terms of entry and exit, while Turks and Syrians are subject to greater distance and immigration controls. Variation in origin circumstances means that Italians, Poles, and Turks can normally choose to emigrate, whereas Syrians are more likely forced to seek asylum. Any of these differences may produce different potential for reactive transnationalism.

Turkish migrants have a highly transnational reputation (Abadan‑Unat 2011; Duru, Favell & Varela 2022; Klok et al. 2020; Snel, Hart & Van Bochove 2016). This may be due to the Turkish government’s efforts to foster a passion for Türkiye among its expatriates. Controversially, in 2008 and 2011, President Erdogan himself visited Germany to a ‘rockstar welcome’ (Ataman 2008; Gezer & Reimann 2011). Facing crowds of more than 10,000, he urged Turkish expatriates, ‘Yes, integrate yourselves into German society but don’t assimilate yourselves. No one has the right to deprive us of our culture and our identity’ (Gezer & Reimann 2011). During the more recent rally, he also announced upgrades to the Blue Card program, which allows former Turkish citizens to maintain many of their former rights (Ministry of Foreign Affairs 2022). In destination countries like Germany that do not permit dual citizenship, it allows members of the diaspora to obtain a receiving‑society passport but classify as something other than aliens within Türkiye.

Thus, Turkish immigrants might be the most transnational group considered here. Snel et al. (2016) found this, with their Turkish respondents out‑pacing Moroccan and Surinamese immigrants in both transnational identification and transnational activities. However, their study did not test Turks’ response to discrimination. Their unique transnational infrastructure can be relied upon to easily engage with the origin country both financially and politically if the German society appears unwelcoming. Certainly, the same exists to some degree for Italian, Polish, and Syrian immigrants, but the intentional work done by Türkiye to promote and maintain these connections may provide a unique circumstance. Therefore, it is possible that Turkish respondents in particular will turn toward transnationalism when they experience or perceive discriminatory treatment in the receiving society. This possibility of Turkish uniqueness will be tested through interactions between immigrant origin and both discrimination experience and perceived group unfairness.

Data, variables, and methods

The current study analyzes the ENTRA survey, whose stated goal is to ‘collect data on recent immigrants in Germany, covering immigration and settlement dynamics, as well as integration trajectories’ (Diehl, Koenig & Kristen 2024; Seuring et al. 2023). Respondents were recruited with a two‑stage sampling design. First, a sampling frame was developed using information from local immigrant registration offices from the five cities¹ with the largest populations of each group—Italians, Poles, Turks, and Syrians. Second, a random sample of each was selected separately, with the goal of obtaining at least 1000 participants per group. The researchers recruited immigrants between 18 and 40 years old who settled in Germany between 2015 and 2019 and did not have German citizenship. Data collection occurred between May and November of 2019.² The final analytical sample includes 4448 respondents.

Dependent variables: transnationalism

The current study considers two outcomes that measure economic and political transnationalism. The transfer of remittances to the origin country represents a common form of economic transnationalism (Carling 2008). This dichotomous variable is built from two questions: ‘Since you moved to Germany this time, have you (or your partner) given or sent any money to [origin country]’ 1) relatives or 2) friends? Responses are coded as 1 if they answered ‘yes’ to either³ and 0 otherwise. Table 1 displays sample means and percentages for this variable and all others considered. Across the full sample, 23.71% of respondents have sent money to people in their origin, ranging from 14.21% among Italians to 33.73% among Syrians. Both values differ significantly from the sample average. Turkish and Polish respondents remit at levels no different from the average.

Political transnationalism is an index of three items capturing political interest and engagement in the origin country. The questions read: 1) ‘How interested would you say you are in [origin country] politics?’; 2) ‘How well do you understand the important political issues facing [origin country]?’; and 3) ‘In the last 12 months, how often did you typically discuss major political issues facing [origin country] with others?’ Response options range from either 1 to 4 or 1 to 5 with higher values indicating greater engagement frequency. Each variable was standardized to have a mean of zero and a variance of one and then combined into a mean index of political transnationalism. The items have a Cronbach’s alpha of .81, indicating high reliability. Across groups, Syrians fall significantly below the mean and Turks rise significantly above. The high level of political transnationalism among Turks also significantly exceeds that of Italians (t = 10.798; P < .000) and Poles (t = 12.609; P < .000).

Independent variables

Confirming and expanding the reactive transnationalism correlation

Table 2 displays multivariate analyses predicting political transnationalism and remittance behaviors. The control variables exhibit similar associations for both outcomes in some instances, like men and older respondents being significantly more transnational. However, other variables exhibit opposing associations. Years of education and education as a migration motivator are both related to greater political transnationalism but fewer remittances. Conversely, work as a migration motivator is related to less political transnationalism but more remittances. In other cases, variables matter for one outcome, but not the other. Income, single status, and the safety motivation are all related to greater political transnationalism but have no impact on remittances. Conversely, family migration is related to greater remittances but unrelated to political transnationalism. Thus, while both outcomes are forms of transnationalism, they are clearly distinct.

Transnationalism associated with any of these factors is not considered reactive. This occurs when it is correlated with discrimination variables, which are the primary focus of the analysis. Starting with direct experience in the political transnationalism model, for each unit increase in discrimination, political transnationalism is predicted to be .031 units higher. Across the entire discrimination scale, this corresponds to about 15% of a standard deviation in transnational political interest. Likewise, in the remittances model, the coefficient of .018 indicates that respondents are significantly more likely to send money back to the origin when exposed to discrimination.

Similarly, the belief that one’s group experiences discrimination in Germany correlates significantly with transnational behaviors. For each unit increase in the perceived unfairness, political transnationalism is predicted to be .055 points higher. Across the entire length of the independent variable scale, this corresponds to about 25% of a standard deviation in political transnationalism. Likewise, the coefficient of .025 indicates that those believing that their group is treated unfairly are significantly more likely to send money to their country of origin.

In both models, the perceived group unfairness association is larger than reported discrimination experience. However, overlapping 95% confidence intervals indicate that the slopes are not significantly different. Thus, both experienced and perceived discrimination exhibit independent associations with transnational practices of comparable magnitude.

Differential effects across immigrant origins

Table 2 shows that both outcomes vary across origin groups. All remit significantly more than Italians. This is especially true for Syrians, who follow the bivariate patterns in Table 1. Turks are significantly more politically transnational than those of the other origins (versus Poland: F = 107.36, P < .000; versus Syria: F = 225.26, P < .000)—also following the patterns in Table 1. Both Poles and Syrians are significantly less politically transnational than Italians.

These origin indicators are included to determine if the reactive transnationalism associations differ across groups. This requires interactions with the discrimination variables, which are displayed in Table 3. For the remittances model, none of the interactions reach statistical significance. These non‑significant patterns are displayed graphically in Figure 1. The significant country main effects are reflected in the different heights of each line. The non‑significant discrimination experience main effect suggests a flat trajectory for Italians. The non‑significant interactions indicate that the other origin group trajectories are equivalent to parallel with that of Italians. These results suggest that experiencing discrimination and perceiving one’s group as being victimized have consistent, positive correlations with remittance behaviors.

Table 3 includes interactions between discrimination experience and each origin. The large and positive interaction for Turks is statistically significant. The analogous interactions for Poland and Syria are half the size and non‑significant. Figure 2 charts the predicted values. The trajectory for Italians is relatively flat, reflected by the near‑zero discrimination experience main effect. Discrimination has little impact on Italians’ political transnationalism. The trajectories for Poles and Syrians are statistically equivalent to that of Italians.

Conversely, Turks exhibit a more dramatic upward slope. The interaction of .089 suggests that their trajectory is significantly different from the reference category. Turks’ higher starting point reflects their higher average levels of political transnationalism in Table 1. The steep slope demonstrates a stronger reactive transnationalism association. In fact, the flat lines for the other groups suggest that a reactive transnationalism is unique to Turkish immigrants for the political outcome.

The final model of Table 3 considers the interaction between origin and perceived group unfairness. Like the remittances models, none of these reach statistical significance. Thus, while perceiving that one’s group is treated unfairly is associated with transnational practices, these patterns are consistent across immigrant groups. The associations are displayed graphically in Figure 2. While the interaction patterns are similar to those of discrimination experience, the lack of significant associations indicates that the trajectories are statistically equivalent to parallel.

What is next for reactive transnationalism research?

The current analysis adds Germany to the list of countries where evidence for reactive transnationalism has been found. There are other contexts that might be worth exploring, including Nordic countries, but we now have strong evidence that this pattern exists. An important next step is to move beyond the simple correlation and begin examining other aspects of the phenomenon.

The old adage about correlation not implying causation is relevant here, as well as with previous reactive transnationalism research. Since these studies rely on cross‑sectional evidence to examine an inherently longitudinal process—experience and perception causing transnational behavior—there is a risk of reverse causality and unobserved confounders. It is possible that highly transnational migrants more readily interpret interactions as discriminatory, which is a reverse of the assumed causal sequence. They are more oriented toward their origin and may view the receiving society with more skepticism. With ENTRA’s longitudinal structure, it is possible to partially test causality with discrimination experience through a difference‑in‑differences approach. Discrimination experience can be grouped into four categories indicating whether an incident was reported at the initial and subsequent waves. Those reporting discrimination for the first time prior to the second wave (but not the first) are of particular interest. Their change in status could be evidence for a causal pattern should their discrimination be associated with a change in remittance or political transnationalism across the waves. However, no significant changes in transnational behaviors were observed (see Appendix⁶), leading to an inconclusive result. Regardless, future studies should consider the causal sequence of reactive transnationalism through longitudinal and experimental approaches.

Experience is not needed for reactive transnationalism

The finding that non‑experienced discrimination is also related to transnational behaviors has significant implications for future research. It suggests that an unwelcoming receiving society reputation alone may influence immigrants’ behavior. Simply hearing about discrimination happening to other people may be consequential. It also suggests that any intentional act of anti‑immigrant discrimination will have ripple effects beyond the intended victim.

This finding provides opportunities for future research. There is nuance to the unexperienced forms of discrimination that are not captured by single‑item measures. More intentional measures of vicarious and anticipated discrimination can show which form is more consequential. Further, the frequency and intensity of either type may also be important (Gonzales, Lee & Marchiondo 2019).

Nuance in reactive transnationalism across origin groups

We now have evidence indicating that the reactive transnationalism association differs across immigrant origin groups. The fact that Turkish immigrants focus on origin country politics more when they experience receiving‑society hostility may result from the transnational infrastructure that is accessible to them. Efforts by the Turkish government to keep its diaspora engaged are likely influential. Turkish politics can serve as an accessible refuge if immigrants see a need to disassociate from German society. While opportunities to connect with Italy and Poland also exist, these countries are easier to return to if hostility becomes unbearable. This discussion is speculative since survey data do not provide sufficient context. Future research should continue to examine reactive transnationalism with qualitative approaches to understand immigrants’ feelings and explanations behind their actions.

At the same time, there was no nuanced pattern for remittances. It may not be the case that previous research has missed variation. Rather, it likely depends on the specific outcomes and circumstances under consideration. Nevertheless, future research should distinguish between groups when reporting reactive transnationalism. The current study likely only scratches the surface regarding group differences. For instance, diversity within the category of ‘Turkish’—Kurds, Assyrians, etc.—may hide additional patterns. Certain groups may be immune to discrimination, or certain transnational behaviors may be especially triggered by hostile encounters. Exploring this nuance is an important task moving forward.

Additional limitations and directions for future research

While the current study offers some useful findings, there are additional limitations worth noting. For instance, the ENTRA data focuses only on recent immigrants, specifically those present in Germany for under five years. This includes respondents living in the receiving society for less than one year. Such recent immigrants may be more prone to send remittances or focus more on origin politics, and this tendency might wane with more time in the destination country. Whether the current findings reflect patterns among more established immigrants is unknown, but it is something that future research should consider.

Also notable is ENTRA’s focus on destination cities, which does not produce a representative sample. Smaller cities, towns, and rural areas are excluded. Whether these individuals are different than those residing in Berlin or Hamburg is unknown, but this would be an avenue for future research.