This is an open access article distributed under the terms of the Creative Commons Attribution License (The migration of refugees into Germany and other European countries has become a major challenge for politics and societies. Has the culture of welcome entered troubled waters (Becker, 2022; Dinkelaker et al., 2021; Kober, 2017; Vorländer, 2022)? Anti-immigration politicians and parties are gaining more applause and support from voters (Dennison & Geddes, 2019). Since Russia began this aggressive war against Ukraine, European countries are confronted with an influx of new refugees. Are they still welcome? Attitudes in the population toward immigrants and refugees play an increasingly important role and should be better investigated and better understood. While anti-immigrant attitudes appear to apply to immigrants in general, an in the population widespread distinction is less considered: that between attitudes toward refugees conceived as victims seeking refuge from war, on the one hand, and refugees supposedly arriving for better living conditions, on the other hand. This study speaks to the questions related to the evolution of the ‘culture of welcome’ over time, analyzes the differences in attitudes toward war refugees and ‘economic’ refugees,1 and presents a model to understand the effects of othering and prejudice, on the one hand, and responsiveness to the Other, on the other hand, on these attitudes toward refugees.
Attitudes Toward Refugees in Europe
Regarding attitudes toward refugees in Europe, there exists a substantial body of research based on large survey data reflecting the situation before the war against Ukraine (for selected recent contributions, see Abdelaaty & Steele, 2022; Becker, 2022; Benoit, 2021; Czymara, 2021; Gerhards & Dilger, 2020; Polak & Schuster, 2023; Reinhardt & Frings, 2020; Streib & Klein, 2018b). For contextualizing our study, we highlight some recent contributions that include findings about the attitudes toward refugees from Ukraine.
De Vries and Hoffmann (2022) report results from a survey of nearly 12,000 European citizens conducted between March and June 2022. Their results suggest that a majority of 83% Germans agreed to the question “Should your country accept refugees from Ukraine?” Dražanová and Geddes (2022) surveyed people in eight European countries (Austria, Czech Republic, Germany, Hungary, Italy, Poland, Romania and Slovakia) between May 15 and June 6, 2022 with nationally representative samples of approximately 1000 respondents; they report that 86.8% Germans agreed to allow “some” or “many” Ukrainian refugees, an additional 9.0% would allow “a few,” while only 4.2% would “allow none.”
Herold and colleagues (2022) report from their survey of ten European countries in fall of 2022 (MIDEM Study), which included 2,091 Germans, a difference in attitudes according to the refugees’ region of origin. To the question “Refugees from the Ukraine integrate more easily in the German society than those from the Near East and Africa,” 49% Germans agreed, while 32% disagreed. This indicates that about half of the German respondents tend to accept war refugees from Ukraine, rather than refugees who come from the Near East or Africa. Another differential perspective comes into view, when religious culture is considered: to the question whether immigration of Muslims into Germany should be restricted, 54% German respondents agreed, while 27% disagreed. While Herold and colleagues (2022) report frequencies only, their results raise questions for understanding the difference between different types of refugees. Are Ukrainian refugees more welcome because they are expected to integrate better into the German society, are not coming from Near East or Africa, and having mostly no Muslim religious-cultural background?
One of the recent studies on right-wing and anti-democratic attitudes in Germany (the so-called 'Mitte-Studie', Zick et al., 2023) included questions on attitudes toward refugees from Ukraine. Küpper and Hellmann (2023) report findings how the attitudes toward refugees vary according to the region of origin: While 34.8% approved “rather” or “fully” that Germany accepts refugees from Africa, and 38.5% approved the acceptance of refugees from Syria, accepting refugees from Ukraine is supported by 61.5%.
Based on their analyses of the European Social Survey (ESS) 2002 and 2014, Abdelaaty and Steele (2022) suggest distinguishing between the attitudes toward refugees and the attitudes toward immigrants more clearly, and call for more research on attitudes toward refugees, which they see as rather a neglected field of research. It is one of their key findings that the “individual-level support for immigrants is more consistently related to the financial well-being of households while support for refugees is more robustly associated with country-level political circumstances” (Abdelaaty & Steele, 2022, pp. 124-125). While their analysis of the ESS data does not address the role of other factors such as personality, worldview, or religiosity, their approach of distinguishing between immigrants and refugees is important.
Taken together, extant research reports high agreement to welcoming war refugees from Ukraine, but also suggests attending to the difference between attitudes. In this study, we attend to the difference between refugees who supposedly come for reasons of better living conditions and refugees who flee from war in their home countries seeking shelter and security; therefore, we included in our questionnaire items that ask for (dis)agreement to accepting these people who come to Germany. And we sought to document the difference in attitudes between 2015/2016 and 2022 with two repeated surveys.
Prejudice Research and Attitudes Toward Immigrants and Refugees
This study intends to open a perspective on understanding the attitudes toward refuges, thus, it speaks to research on prejudice and prejudice reduction. One of the limitations of previous research on prejudice consists in a rather exclusive attention to the pathogenic effects: the development of prejudice. In contrast, less attention has been paid to the salutogenic potentials that may foster prejudice reduction.2 Thus, especially when intervention strategies are at stake, research on prejudice should attend to the questions of prejudice reduction.
And looking into extant research here, progress in results about prejudice reduction appears to slowly develop and improve. From their evaluation of almost a thousand reports on prejudice reduction interventions, Paluck and Green (2009, p. 360) conclude that “psychologists are a long way from demonstrating the most effective ways to reduce prejudice.” Paluck (2016) continuously discussed the question of how to overcome prejudice, and more recently (Paluck et al., 2021) note some progress and identify landmark studies, among them, for example, Broockman and Kalla’s (2016; Kalla & Broockman’s 2020, 2023) field experiments about the reduction of prejudice against transgender and undocumented immigrants by perspective-taking and perspective-getting interventions.
The theoretical framework for this study has been developed also in response to the limitations of extant research. While the assumptions about the salutogenic effect of perspective-taking and interpersonal encounter with the strange, was included in our past conceptualizations (Streib, 2018; Streib & Klein, 2014), it is only recently that we developed, with reference to Waldenfels (e.g., 2011), the explication of Other-orientation and responsiveness with greater philosophical conciseness (Streib, 2024a, 2024b). Thus, the basic theoretical model for this study assumes a polarity between othering and responsiveness to the other as part of prejudice research with a clear focus on prejudice reduction and focusing on the specific case of attitudes toward refugees and immigrants. The specific research question is this: What are the factors that contribute to the reduction of prejudice against war and economic refugees? Components of the proposed model need to be discussed in more detail.
Group-Based Enmity vs. Other-Oriented Responsiveness: A Hypothetical Model
Klein and colleagues (2018, p. 25) note that, from the beginning of prejudice research (e.g. Adorno et al., 1950; Allport, 1954/1979), it has repeatedly been noticed that several types of prejudice co-occur—somebody holding xenophobic attitudes is also likely to hold anti-Semitic, racist, sexist, or homophobic attitudes (Zick et al., 2008). This has been described and investigated as “generalized prejudice” (Asbrock et al., 2010; Bergh et al., 2016; Bergh & Brandt, 2023). Thus, Heitmeyer (2002) has proposed to understand distinct manifestations of prejudice to be symptoms of a superordinate syndrome that he calls “Group-focused Enmity” (GFE). The GFE syndrome includes a variety of prejudices, such as xenophobia, racism, anti-Semitism, Islamophobia, sexism, homophobia (Küpper & Zick, 2014). In the GFE, xenophobia is assessed with items such as “There are too many immigrants in Germany” and “We need to protect our culture from foreign influences.” In the assessment of GFE, xenophobia is included as only one component that taps, together with the other components, into the syndrome of prejudice and hostility toward out-groups and minorities. Using the construct of GFE in this study, we aimed to examine whether and how the attitudes toward war refugees and ‘economic’ refugees relate to the syndrome of GFE to approach the question of how we can explain the motivation of the attitudes of welcoming or rejection of refugees.
We find support for our assumption of how the GFE is related to attitudes toward immigrants and refugees from previous research. In 2008/2009 Zick and colleagues (2011) investigated the prejudice towards immigrants in a survey of eight European countries (France, Germany, Great Britain, Hungary, Italy, Netherlands, Poland, Portugal). Across these eight European countries the (negative) attitudes toward immigrants correlated with several GFE elements (racism, anti-Semitism, anti-Muslim attitudes, sexism, and sexual prejudice) between r = .30 (homophobia) and r = .59 (anti-Muslim attitudes) (Zick et al., 2011, p. 70); this can be understood as support for the coherence of the GFE as a syndrome. For Germany, the anti-immigrant attitudes received the strongest agreement of the six GFE elements, only topped by anti-Muslim attitudes (Küpper & Zick, 2014, p. 249). Attitudes toward refugees were also included in a previous Mitte-Studie (Zick & Küpper, 2021). The study was conducted in early 2021, thus before the war against Ukraine. Interestingly, self-reported ‘very negative’ and ‘rather negative’ attitudes toward refugees (21.7%) are only topped by those against Muslims (22.6%) and Sinti and Roma (31.7%), while the negative attitudes toward black people (9.0%) or Jews (8.7%) were lower (Zick & Küpper, 2021, Table A.6.1). This indicates that, in the Mitte-Studie 2021, the negative attitudes toward refugees stand out compared to other GFE aspects.
As the opposite pole to GFE we include in our hypothetical model the construct of Other-oriented Responsiveness (OOR), which combines three components: two subscales of the Religious Schema Scale (Streib et al., 2010), xenosophia/inter-religious dialog and fairness, tolerance and rational choice, and the items for pluralism from the Religion Monitor (Bertelsmann-Foundation, 2009; Huber, 2009). Certainly, several items in these three component scales explicitly address the encounter with other religions and religious truth claims, but they assess an attitude of non-egocentric, non-sociocentric responsiveness that embraces also the domain of intersubjectivity. And in fact, previous research with the two RSS subscales demonstrated their effect on the mitigation of intergroup, intercultural, and interreligious prejudice (Ardi et al., 2021; Kamble et al., 2014; Lühr et al., 2018; Melles & Frey, 2018; Streib & Klein, 2014, 2018a; Tekke et al., 2015; Watson et al., 2015). Thus, we assume that these variables may relate positively with the acceptance of war refugees.
As noted, the OOR is inspired by Waldenfels’ (2011, 2016) philosophy of the alien and its recent interpretation for psychology and wisdom research (Streib, 2024a, 2024b). The polarity between GFE and OOR is reflected in the model of styles of responsiveness (Streib, 2024a), whereby the egocentric/sociocentric style involves othering based on an in-group vs. out-group perspective, while OOR features negotiatory and xenocentric responsiveness—including, in the context of this study: responsiveness for the individual who comes to us seeking help, security, and solidarity.
This Study
This study offers insights into how war refugees, in comparison to the attitudes toward ‘economic’ refugees, are accepted under the unexpected conditions of high immigration to Germany. The data allow for an investigation of the relations of the attitudes toward refugees with demographics, personality characteristics, religious and worldview styles, and a variety of attitudes.
Furthermore, this study aims to explain the attitudes of accepting refugees and thereby attending to the difference between attitudes toward war refugees and ‘economic’ refugees. To this end, we have constructed a model (Figure 1) that works with the two opposed constructs: We posit that rejection of immigrants and refugees belongs to an attitude of othering. The GFE syndrome is a way of combining a variety of prejudices; and the interpretation as 'othering syndrome' appears justified. Thus, the GFE syndrome is included as a whole syndrome to test the assumption that othering strongly relates to the rejection of refugees. In a similar way, welcoming refugees can be combined with other positive, other-oriented attitudes in a cluster of responsiveness. The factor of Other-oriented Responsiveness (OOR) is empirically a latent variable that is constructed from three scales in our data as described above.
The model in Figure 1 works with the polarity between GFE and OOR that are assumed to correlate negatively with each other: GFE features egocentric or sociocentric othering and is assumed to promote prejudice and hostility against refugees, while OOR features the attitudes of non-egocentric and non-sociocentric Other-orientated responsiveness and may promote greater openness to welcoming refugees.
Figure 1
Model How Group-Focused Enmity and Other-Oriented Responsiveness Relate to Accepting Refugees
Taken together, we have two specific aims and assumptions for this study. First, we describe the two samples and show the differences in attitudes toward refugees and other variables in the two cross-sectional samples. The primary hypothesis is that there is the stability of the agreement to accepting war refugees into Germany on a high level. Second, we test our hypothetical model and explore how GFE and OOR relate to attitudes toward both types of refugees in their unique ways. We hypothesize that GFE would be negatively associated with both attitudes, whereas OOR will be positively associated with attitudes toward war refugees, while association with attitudes toward ‘economic’ refugees is an open question.
Method
Participants
Data were two national representative samples surveyed by an online-questionnaire and recruited by an opinion research service in August of 2015 to March of 2016 (n = 1,262), and in June of 2022 (n = 1,339). To reflect distribution in the German population, quota were set for age groups, gender, region (Germany-East, Germany-West), and education levels. Additional demographic variables included measures for per-capita income, religious affiliation, and religious/spiritual identification. Personal characteristic variables included self-report of the extent to which one identifies as religious, spiritual, and atheist, using questions of “How religious would you describe yourself?”, “How spiritual would you describe yourself?”, and “Would you call yourself an atheist?” on five-point scales ranging from 1 (Not at all) to 5 (Very much). Ethical approval was granted from Bielefeld University. There was no informed consent, because the questionnaire was anonymous.
Study Variables
Attitudes Toward Refugees
Items for the attitudes toward refugees were created for the first survey in 2015. For war refugees, the item said, “War refugees should be accepted into Germany.” For ‘economic’ refugees, the item said, “Refugees who only come to Germany because of the better living conditions should directly be deported to their home countries,” which has been reverse scored with higher values indicating more positive attitudes. All items for attitudes were measured on a 1-to-4 Likert scale (completely disagree to completely agree).
Group-Focused Enmity (GFE)
Four variables assessed GFE (Küpper & Zick, 2014; Zick et al., 2011). Xenophobia is calculated from three items: “There are too many immigrants in Germany;” “The government is doing too little against foreign infiltration;” “We need to protect our own culture from the influence of other cultures.” The estimated internal consistency is α = .83 for the 2016 sample and α = .88 for the 2022 sample. Racism consists of two items: “There is a natural hierarchy between people from different cultures;” “People from different cultures should not get married.” Homophobia also is based on two items: “There is nothing immoral about homosexuality;” “It is a good thing to allow marriages between two men or two women.” Finally, sexism includes two items: “Women should take their role as wives and mothers more seriously;” “When jobs are scarce, men should have more right to a job than women.” All items for attitudes were measured on a 1-to-4 Likert scale (completely disagree to completely agree).
Other-Oriented Responsiveness (OOR)
Three variables assessed OOR. Two variables came from the Religious Schema Scale (RSS; Streib et al., 2010): fairness, tolerance & rational choice, is a schema that features rational exchange of arguments (e.g., “We should resolve differences in how people appear to each other through fair and just discussion”). The estimated internal consistency is α = .88 for the 2016 sample and α = .86 for the 2022 sample. Xenosophia/inter-religious dialog, is a schema reflecting an appreciation of difference, of the other, and of dialog (e.g., “We need to look beyond the denominational and religious differences to find the ultimate reality”). The estimated internal consistency is α = .77 for the 2016 sample and α = .79 for the 2022 sample. Pluralism is a three-item measure assessing openness to truths in various traditions (sample item “For me every worldview and religion has a core of truth”), adopted from the Religion Monitor (Bertelsmann-Foundation, 2009; Huber, 2009). Items were rated from 1 (Definitely not true) to 5 (Definitely true). The estimated internal consistency is α = .81 for both the 2016 and the 2022 sample.
Religious Worldviews
Two variables assessed different flavors of religious worldviews. The RSS 5-item subscale truth of texts and teachings features an authoritative understanding of one’s own sacred texts (e.g., “What the texts and stories of my religion tell me is absolutely true and must not be changed”), but does not assume moral positions against other traditions. The estimated internal consistency is α = .90 for the 2016 sample and α = .89 for the 2022 sample. The 9-item fundamentalism scale, adopted from the Religion Monitor (Bertelsmann-Foundation, 2009; Huber, 2009) assumes ethical positions against things that one’s religion sanctions (e.g., “For my religion or worldview it is important that I resolutely fight against evil”). Items were rated on a five-point scale ranging from 1 (Definitely not true) to 5 (Definitely true). The estimated internal consistency is α = .87 for the 2016 sample and α = .88 for the 2022 sample.
Personality
We assessed personality traits using the 10-item Big Five Inventory (BFI-10; Rammstedt & John, 2007). The BFI-10 assesses neuroticism (“I see myself as someone who get nervous easily.”), extraversion (“I see myself as someone who is outgoing, sociable.”), openness to experience (“I see myself as someone who has an active imagination.”), agreeableness (“I see myself as someone who is generally trusting.”), and conscentiousness (“I see myself as someone who does a thorough job.”).
Analysis
Across all demographic and study variables, 2.9% individuals had missing values, and most missed on only one variable. The missing pattern did not fail the Little’s test, χ2 = 577.71, p = .402, indicating missing completely at random. Analysis proceeded with all available data. We first computed the mean differences across all study variables between the two samples of 2016 and 2022. Thereby, we report the effect size as indicator of magnitude of differences. Thereby, Cramer’s V was used for categorical variables, with values less than .30 indicating small effect and .50 indicating medium effect. For continuous variables, Cohen’s d was used, and the value less than .20 indicating small and .50 medium effect. For the analysis of the hypothetical model, composite scores for GFE and OOR were computed using principal component analysis. Xenophobia, Racism, Homophobia, and Sexism were used to compute the composite variable GFE. The RSS subscales fairness, tolerance & rational choice and xenosophia/inter-religious dialog, as well as pluralism were used as variables for the composite variable OOR. Using the combined sample, we then used multiple regression to examine associations of GFE and OOR with the two attitudes toward refugees, controlling for demographic and characteristic variables.
Results
Description of the 2016 and 2022 Samples
Description of demographics and study variables for both samples 2016 and 2022 are presented in Table 1. As the column for effect size demonstrates, there is not much difference between the two samples. Cohen’s d was mostly lower than 0.20, considered small. Exceptions are xenophobia and fundamentalism that decreased more greatly, and the 2022 sample was older.
Table 1
Description of Demographic and Study Variables
| Variable | 2016 (n = 1,262) | 2022 (n = 1,339) | Effect size |
|---|---|---|---|
| Age, M (SD) | 43.84 (13.92) | 50.24 (17.07) | 0.41 |
| Sex, % | 0.00 | ||
| Male | 48.9% | 49.7% | |
| Female | 51.1% | 50.3% | |
| Region, % | 0.03 | ||
| West Germany | 79.7% | 76.4% | |
| East Germany | 20.3% | 23.6% | |
| Education, % | 0.17 | ||
| Below college | 78.0% | 62.1% | |
| College or above | 22.0% | 37.0% | |
| Income, % | 0.02 | ||
| Low (below $26,825) | 59.0% | 52.5% | |
| Medium or above ($26,825 or above) | 41.0% | 40.4% | |
| Religious affiliation, % | 0.10 | ||
| No religious affiliation | 38.4% | 47.8% | |
| Protestant | 29.6% | 25.5% | |
| Catholic | 25.4% | 19.9% | |
| Other religious affiliation | 5.9% | 6.6% | |
| Religious/spiritual identification, % | 0.03 | ||
| Neither religious nor spiritual | 53.2% | 49.4% | |
| Equally religious and spiritual | 12.1% | 12.8% | |
| More religious than spiritual | 16.5% | 19.7% | |
| More spiritual than religious | 18.2% | 17.8% | |
| Religious/spiritual self-report, M (SD) | |||
| Atheistic | 2.45 (1.49) | 2.68 (1.51) | 0.15 |
| Religious | 2.04 (1.05) | 2.13 (1.14) | 0.19 |
| Spiritual | 2.09 (1.13) | 2.13 (1.20) | 0.03 |
| Religious worldviews, M (SD) | |||
| Truth of texts and teachings | 2.40 (0.96) | 2.28 (1.00) | -0.12 |
| Fundamentalism | 2.57 (0.82) | 2.24 (0.84) | -0.39 |
| Big Five personality, M (SD) | |||
| Neuroticism | 2.69 (0.92) | 2.73 (0.99) | 0.04 |
| Extroversion | 3.14 (0.97) | 3.05 (0.96) | -0.09 |
| Openness to experience | 3.41 (0.92) | 3.41 (0.97) | -0.00 |
| Agreeableness | 2.91 (0.85) | 2.85 (0.87) | -0.07 |
| Conscientiousness | 3.69 (0.81) | 3.78 (0.80) | 0.11 |
| Accepting attitudes toward refugees, M (SD) | |||
| War Refugees | 3.21 (0.90) | 3.29 (0.72) | 0.11 |
| ‘Economic’ refugees | 2.02 (0.98) | 2.20 (0.97) | 0.19 |
| Group-focused enmity, M (SD) | |||
| Xenophobia | 2.70 (0.88) | 2.48 (0.92) | -0.24 |
| Racism | 1.68 (0.77) | 1.81 (0.77) | 0.17 |
| Homophobia | 1.95 (0.91) | 1.83 (0.91) | -0.13 |
| Sexism | 1.84 (0.77) | 1.75 (0.71) | -0.12 |
| Other-oriented responsiveness, M (SD) | |||
| Fairness, tolerance & rational choice | 3.82 (0.79) | 3.91 (0.84) | 0.11 |
| Xenosophia/inter-religious dialog | 3.15 (0.75) | 3.07 (0.85) | -0.11 |
| Pluralism | 3.44 (0.97) | 3.32 (1.02) | -0.11 |
Note. Percentages may not add up to 100% due to missing. For the 2016 sample, n ranged from 1,253 to 1,262; for the 2022 sample, n ranged from 1,244 to 1,339 due to missing. Effect size for categorical variables used Cramer’s V and for continuous variables used Cohen’s d. Attitudes toward refugee and GFE variables were measured on a 4-point scale ranging from 1 to 4 while the other study variables were measured on a 5-point scale ranging from 1 to 5.
The sample characteristics were also quite similar with Cramer’s V less than 0.30. Nevertheless, there are interesting changes between 2016 and 2022, such as the proportion in religious non-affiliation (increasing over time), the corresponding self-rating as being atheist that also increased over time, and the scores for truth of texts and teachings and fundamentalism, and their decrease over time.
Besides the differences between samples 2016 and 2022, Table 1 presents interesting differences in one of the central variables for this study: the contrast between attitudes toward war refugees and ‘economic’ refugees. The scores for accepting war refugees into Germany were significantly higher than the scores for accepting ‘economic’ refugees, t = 55.06, p < .001. The two attitudes correlated at r = 0.32, 95% CI [0.28, 0.35].
To detail and visualize this difference between the attitudes toward war and ‘economic’ refugees, the frequencies for accepting war refugees in comparison with accepting ‘economic’ refugees into Germany are presented in Figure 2. This visualization of frequencies indicates that in 2016 a majority of 81.1% tend to agree or fully agree that war refuges should be accepted into Germany. But with 89.8% the acceptance for war refugees was even higher in 2022.
Figure 2
Attitudes Toward War Refugees and ‘Economic’ Refugees 2016 and 2022
In contrast to accepting war refugees, accepting ‘economic’ refugees into Germany is strongly rejected by the majority and the percentage of rejections of ‘economic’ refugees has declined between 2015/2016 (70.0%) and 2022 (60.8%)—or on the positive side: the readiness for accepting ‘economic’ refugees into Germany has increased somewhat between 30.0% in 2016 and 39.2% in 2022. But there is across both study years a portion of over 50% who agree or tend to agree to the acceptance of war refugees, but not ‘economic’ refugees.
Attitudes Toward Refugees as Outcomes
Because there was not much difference between the samples 2016 and 2022 (as noted, Cohen’s d was mostly lower than 0.20; sample characteristics were also quite similar with Cramer’s V less than 0.30), we used the combined sample (n = 2,601) for the multiple regression analyses. The combined sample lends more power to the analyses. Table 2 presents the results.
Table 2
Regression Analyses With Attitudes Toward Refugees as Outcomes
| Target Variable | Accepting attitudes toward … | |
|---|---|---|
| War refugees | ‘Economic’ refugees | |
| β [95% CI] | β [95% CI] | |
| Demographic and characteristic covariates | ||
| Sample 2022 (vs. 2016) | 0.07* [0.01, 0.13] | 0.22** [0.15, 0.29] |
| Age | 0.01 [-0.02, 0.04] | -0.13** [-0.17, -0.1] |
| Female (vs. Male) | -0.09** [-0.15, -0.03] | 0.00 [-0.07, 0.07] |
| East Germany (vs. West Germany) | -0.02 [-0.09, 0.05] | -0.10* [-0.19, -0.01] |
| College education (vs. below college) | -0.01 [-0.08, 0.05] | 0.05 [-0.03, 0.13] |
| Medium or above income (vs. low income) | -0.03 [-0.09, 0.03] | -0.04 [-0.11, 0.04] |
| Protestant (vs. no religious affiliation) | 0.08* [0.00, 0.16] | 0.01 [-0.09, 0.10] |
| Catholic (vs. no religious affiliation) | 0.03 [-0.06, 0.11] | -0.01 [-0.11, 0.10] |
| Other affiliation (vs. no religious affiliation) | 0.04 [-0.09, 0.17] | 0.16* [0.00, 0.32] |
| Equally religious and spiritual (vs. neither) | -0.02 [-0.16, 0.12] | 0.03 [-0.14, 0.19] |
| More religious than spiritual (vs. neither) | -0.09 [-0.22, 0.03] | -0.07 [-0.22, 0.09] |
| More spiritual than religious (vs. neither) | 0.01 [-0.12, 0.14] | -0.11 [-0.27, 0.05] |
| Atheistic self-rating | 0.03 [-0.00, 0.07] | 0.00 [-0.05, 0.04] |
| Religious self-rating | 0.09* [0.02, 0.15] | 0.03 [-0.05, 0.11] |
| Spiritual self-rating | -0.03 [-0.10, 0.03] | 0.02 [-0.06, 0.09] |
| Neuroticism | 0.01 [-0.02, 0.05] | -0.02 [-0.06, 0.02] |
| Extroversion | -0.02 [-0.05, 0.01] | 0.00 [-0.04, 0.03] |
| Openness to experience | 0.02 [-0.01, 0.05] | 0.03 [-0.00, 0.07] |
| Agreeableness | 0.02 [-0.01, 0.05] | 0.05** [0.01, 0.08] |
| Conscientiousness | 0.04* [0.01, 0.07] | -0.10** [-0.13, -0.06] |
| Truth of texts and teachings | -0.03 [-0.07, 0.01] | 0.06* [0.01, 0.11] |
| Fundamentalism | 0.02 [-0.02, 0.06] | 0.06** [0.02, 0.11] |
| Group-focused Enmity and Other-oriented Responsiveness | ||
| Group-focused enmity | -0.31** [-0.34, -0.27] | -0.49** [-0.53, -0.45] |
| Other-oriented responsiveness | 0.23** [0.20, 0.27] | 0.00 [-0.04, 0.05] |
*p < .05. **p < .01.
The 2022 sample, compared to the 2016 sample, is slightly higher in accepting war refugees into Germany (β = .07), and considerably higher in accepting ‘economic’ refugees (β = .22); this corroborates results presented in Table 1 and Figure 2. Another interesting result is that age makes a difference, but only for the ‘economic’ refugees (β = -.13): The older the participant the lower the agreement to accept ‘economic’ refugees into Germany. Female participants have slightly lower (β = -.09) openness for accepting war refugees into Germany than males. Results also indicate a difference between Germany-East and Germany-West: participants in the eastern states are slightly more reluctant (β = -.10) accepting ‘economic’ refugees into Germany. Education and income, to conclude the effects of the demographic variables, appear to make no significant difference.
Results for the role of religious affiliation is somewhat unexpected: While Protestants are slightly more open (β = 0.08) to accept war, but not ‘economic’, refugees compared to religiously unaffiliated respondents, there is no such effect for Catholic participants; and members of other, non-Christian religions are more open (β = .16) to accepting ‘economic’ refugees into Germany. The latter may indicate a kind of solidarity with ‘economic’ refugees by the part of culturally and religiously diverse minorities including perhaps people with immigration backgrounds themselves.
Spiritual or atheist self-identifications appear to have no differential effect, except the self-rating as religious that is related (β = .09) to higher acceptance of war refugees. Two variables, truth of texts and teachings and fundamentalism are slightly associated (β = .06 for both) with accepting ‘economic’ refugees. For an explanation we could speculate that these results for truth of texts and teachings and fundamentalism indicate a stronger obligation of fundamentalist and highly committed religious people to follow prescriptions for sharing one’s ‘economic’ prosperity with those who are less privileged. Table 2 also indicates that conscientiousness, a personality trait that favors efforts and diligence, is negatively associated (β = -.10) with accepting ‘economic’ refugees, but positively associated with war refugees.
So far, the results presented in Table 2 reveal some details about the effects of demographics, religiosity and personality variables on the attitudes of accepting war refugees in comparison with ‘economic’ refugees. But these effects are small, marginal, and many are insignificant (which is important information nevertheless). This is considerably different for the two composite variables that are part of the hypothetical model presented above. This is the focus of the following section.
Effects of Group-Focused Enmity and Other-Oriented Responsiveness
The principal component GFE explained 56% variance of the four independent variables whose loadings ranged from .61 to .82. The principal component OOR explained 74% variance of the three independent variables whose loadings ranged from .84 to .86.
The strongest regression weights in Table 2 are those from GFE and OOR on the attitudes toward refugees: GFE was negatively associated with attitudes of accepting war refugees into Germany (β = -.31), and even much more strongly of accepting ‘economic’ refugees (β = -.49). OOR, by contrast, was positively associated with accepting war refugees (β = .23); and not associated with ‘economic’ refugees. These results are visualized in Figure 3. The analysis of these effects has held demographic and personal characteristic variables and religious variables equal.
Figure 3
Regression Weights of Group-Focused Enmity and Other-Oriented Responsiveness on Accepting War and ‘Economic’ Refugees
The effects suggest several interesting observations. First and corroborating results reported above, ‘economic’ refugees are, by the part of the participants, not considered the same as war refugees. Another interesting observation is that OOR is more selective of the type of others – it’s able to accept the war refugees while at the same time remains neutral to ‘economic’ refugees – while on the other hand, GFE rejects both groups, however ‘economic’ refugees more strongly.
Discussion
As expected, the attitudes of Germans for accepting war refugees remained on a high level between 2015/2016 and 2022, it has not declined in 2022, but instead we see some increase from 81.1% in 2016 to 89.9% in June 2022. The high agreement to accepting war refugees from Ukraine in our study dovetails with the high level of acceptance found in other research (de Vries & Hoffmann, 2022; Dražanová & Geddes, 2022).
But while a majority agree that war refugees should be accepted into Germany, also a majority suggests that ‘economic’ refugees should be immediately returned to their home country. Anyway, also the accepting attitudes toward ‘economic’ refugees have somewhat increased between 2016 (30.0%) and 2022 (39.2%). This gap between the attitudes toward the two types of refugees is stronger than we had expected: about half of the German population appears to accept war refugees, but not ‘economic’ refugees in both 2016 and 2022. Interestingly, this result appears in line with the findings of the MIDEM Study (Herold et al., 2022), where results to the question “Refugees from the Ukraine integrate more easily in the German society than those from the Near East and Africa” indicate that 49% of the German respondents accept war refugees from Ukraine, but rather not other refugees who come from the Near East or Africa. This is rather close to our results. Likewise, Küpper and Hellmann (2023) report findings how attitudes toward immigrants vary according to the region of origin: While 34.8% approved “rather” or “fully” that Germany has accepted refugees from Africa and 38.5% approve the acceptance of refugees from Syria, accepting refugees from Ukraine is supported by 61.5%. This is also comparable with what we found. However, to be sure, on basis of our data we cannot answer the question whether Ukrainian refugees are more welcome, because they are expected to integrate better into the German society, are not coming from Near East or Africa, and having mostly no Muslim religious-cultural background. We only have results about participants’ (more or less prejudiced) attribution of the reasons of refugees for coming to Germany. But the difference is extraordinarily high. This, in turn, may put our results in profile and raise additional questions for future research: Are the high scores for welcoming war refugees in 2022 and likewise in 2016 due to the actual situation when the surveys were online—which were situations of high media presence of war refugees? Were there perhaps declines after some time to give way to less welcoming attitudes? Küpper and Hellmann’s (2023) finding about the currently rather low welcome for Syrian refugees appear to support this assumption.
Beyond frequencies, our data allowed the identification of potential motivations to the attitudes toward refugees. Results are summarized in Figure 3. Here, it is one of our most important findings that negative attitudes toward war refugees and ‘economic’ refugees are related to the GFE syndrome, while the relation of GFE with the attitudes toward ‘economic’ refuges is much stronger. Rejecting ‘economic’ refugees can be seen as part and parcel of the GFE syndrome. In contrast, the acceptance of war refugees finds strong support in the OOR, which is composed of the components of xenosophia/inter-religious dialog (xenos, RSS), fairness, tolerance and rational choice (ftr, RSS), and pluralism.
The results for our model entail some interesting implications for religiosity and worldview: It is less the general religiosity, but the kind of religiosity and worldview that is open to tolerance, dialog, and pluralism, which positively relates to accepting war refugees. This result corroborates previous findings that revealed the effects of the ftr and xenos subscales of the Religious Schema Scale on moderating prejudice (Streib & Klein, 2014) and accepting war refugees (Streib & Klein, 2018a). These results also reflect a body of research that used the RSS and indicated that the two RSS subscales ftr and xenos relate to the mitigation of intercultural and inter-religious prejudice (Ardi et al., 2021; Kamble et al., 2014; Lühr et al., 2018; Melles & Frey, 2018; Tekke et al., 2015; Watson et al., 2015).
But beyond or beneath religiosity and worldview, these findings speak to the domain of intersubjectivity, indicating that specific versions of relating to the Other—Other-orientation in way of perspective-getting (Streib, 2024a)—positively relate to welcoming war refugees. The model of this study works with the opposition between GFE and OOR. It reflects the polarity of othering vs. responsiveness, which aligns (in our case: the attitudes toward refugees) with the polarity of focusing on in-group and out-groups vs. attending to the individual. Therefore, in our study higher agreement to the latent variable OOR (and the three scales that relate to this component) is largely associated with responsiveness to another person who is seen as individual, rather than part of an out-group. In contrast, the GFE syndrome works with out-group categorizations, cliches, othering, and prejudice.
Since our results indicate that both targets in our model, accepting war refugees and accepting ‘economic’ refugees, are negatively related to GFE (the attitudes toward ‘economic’ refugees more negatively), while OOR positively relates to accepting war refugees (and is insignificant for ‘economic’ refugees), we may conclude that perspective-taking and, perhaps even more effective, perspective-getting through the responsiveness to another individual support the openness to accepting war refugees.
This echoes a great number of studies and field experiments that demonstrate the effectiveness of perspective-getting for prejudice reduction, including the one by Broockman and Kalla (2016; Kalla & Broockman, 2020, 2023) mentioned above, and many others discussed by Paluck and colleagues (2021). To be sure, our study presents no field experiment, but it clearly confirms the effectiveness of perspective-getting on the attitudes toward refugees, war refugees in particular. This would, in terms of practical implications, suggest interventions that invite identification with and a response to the individual person (refugee), who may appear rather unexpected and strange.
Limitations
Our study has limitations that need to be noted: First, because our data consist of repeated surveys and do not include longitudinal responses, we only can present findings based on cross-sectional analyses, which should be read as only preliminary estimates of change over time, or, regarding the model, as correlations, rather than predictions. Longitudinal data are recommended for any future research in this area. Second, we cannot answer the question whether the culture of welcome has decreased in the meantime, because we do not have a repetition of the survey in the most recent past. Third, the data are self-reported attitudes, thus are perhaps influenced by social desirability. Fourth, the construction of the component of OOR was constructed from what we had in our data and has not been validated previously. For future research, it would be optimal to have a more specific measure, quantitative or qualitative, of OOR. These limitations need to be considered when interpreting our results. But we claim that the limitations do not call our basic findings and conclusions into question.
Conclusion
Our study indicates that the culture of welcoming war refugees was very high in 2022: Nine of ten respondents agree to accept war refugees into Germany. This welcome has not declined, on the contrary, it has slightly increased between the first survey in 2015/2016 and the second survey in 2022—which both were times of high immigration of war refugees into Germany. Also surprising: the openness for accepting ‘economic’ refugees into Germany has increased even somewhat stronger between 2016 and 2022. Nevertheless, the attitudes of accepting ‘economic’ refugees were very low: About half of the Germans appears to accept war refugees, but not ‘economic’ refugees in both 2016 and 2022.
This study is the first to test the hypothetical model which relates the opposed components of GFE and OOR to the divergent attitudes toward refugees. Attitudes toward refugees appear to be part of an antagonistic competition between egocentric/sociocentric othering that features prejudice and reveals to be, especially for ‘economic’ refugees, part and parcel of the GFE syndrome, on the one hand, and, on the other hand, responsiveness to the Other and perspective-getting, which rather attends to the other as individual who, in case of war refugees, needs shelter and deserves compassion. This opens a perspective on a variety of avenues for future investigation. This study is well understood as a first step testing the role and effect of OOR that was recently proposed (Streib, 2024a, 2024b) based on Waldenfels’ philosophy of the alien. Thus, the results of this study reflect Waldenfels’ (2017) proposal for a “responsive policy of the stranger.”