Socio-psychological integration from the perspective of receiving communities: a cross-country comparison between Sweden, Germany, Croatia and Jordan

Abstract

The socio-psychological dimension of integration is based on relations between the refugees and receiving community members revealed through intergroup thoughts, perceptions, emotions and behaviours. This study aimed to investigate and interpret the differences in the indicators of socio-psychological integration among the receiving communities of Sweden, Germany, Croatia and Jordan—countries with diverse socio-economic, socio-political and cultural contexts, histories of inward migration, as well as differently preferred destinations in the migration of refugees from Syria in the 2010s. The contextual differences are reflected in the attitudes of the members of receiving communities towards refugees from Syria, perceptions of realistic and symbolic threats posed by refugees, frequency and valence of intergroup contact, support for the rights of refugees and readiness to assist them, social proximity, perception of refugees’ exposure to discrimination, and the assessment of how much the refugees are a part of the respective local communities.

Introduction

In the last decade, the integration of refugees and receiving communities has become an increasingly important topic in academic research, especially due to the need to study the consequences of the flight of a large part of the Syrian population to neighbouring countries and Europe due to the civil war. Acculturation is a process in which individuals learn the norms characteristic of the other culture, and integration is a type of acculturation process (Berry et al., 1986). Within the European Union's Framework of Integration, it is defined as a "dynamic, two-way process of mutual accommodation by all immigrants and residents of Member States" (European Commission, 2020). The degree of integration can be assessed using the evidence-based Indicators of Integration framework, which includes fourteen indicators belonging to four dimensions of integration: Markers and Means, which parallel the socio-economic dimension of integration; Social Connections, a parallel to the socio-psychological dimension of integration; Facilitators, elements that ease the integration process; and the Foundation—Rights and Responsibilities (Ndofor-Tah et al., 2019). Integration is dynamic because it is multi-directional between immigrants and the local residents, involves the assumption of responsibilities by both groups, and is context-specific (Ndofor-Tah et al., 2019). The socio-psychological dimension of integration encapsulates the intergroup relations between newly arrived immigrants and members of the receiving community. It is based on the social relations governed by the psychological processes that occur before, during, or after intergroup contact.

In general, research on socio-psychological integration has focused primarily on exploring the interdependence of various constructs related to intergroup relations, such as attitudes (e.g., Ajduković et al, 2019; Croucamp et al, 2017; Pedersen et al., 2005), perceptions of intergroup threat (Haase et al., 2019; Schweitzer et al., 2005; Stephan et al., 2005), intergroup contact (Barlow et al., 2012; De Tezanos-Pinto et al., 2017; Saab et al., 2017), social distance (Bruneau et al., 2018; Koc & Anderson, 2018), support for refugee/asylum seeker rights and support for integration policies (Hartley & Pedersen, 2007; Verkuyten et al, 2018), emotions and solidarity (Bracic, 2018; Pawlicka et al., 2019; Pedersen & Thomas, 2013; Verkuyten, 2004), behavioural intentions (Badea et al., 2017; Yitmen & Verkuyten, 2018), political orientation and related personality traits such as right-wing authoritarianism and social dominance orientation (Anderson, 2016, 2018; Trouson et al., 2015), perceptions of different types of immigrant groups (Abeywickarma et al., 2018; Bansak et al, 2016; Gregurović et al., 2016). Some studies focused on the effect of an intervention on attitudes toward refugees and asylum seekers (Berndsen et al., 2018; Cameron et al., 2007; Crowell, 2000).

These constructs have been explored primarily by focusing on receiving community members for several reasons, most notably an underlying assimilationist perspective on newcomer acculturation, but also because of the dynamics of majority and minority groups themselves, with the receiving community having the power and responsibility to ensure that refugees' human and legal rights and entitlements are respected and enforced. Looking more closely at the ways in which the receiving community can accommodate newcomers, Phillimore (2019) has defined five areas in which receiving community activities can have a positive impact on two-way integration opportunities: Influence at the local community (locality) level, affirmative political and media discourse, fostering comfortable and encouraging intergroup relations, developing and implementing structures to facilitate integration, and initiatives and support for the refugee integration process. The way in which the receiving community welcomes and accommodates newcomers, in both formal and informal settings, provides the framework for the process of acculturation.

The goal of this study was twofold. First, we aimed to examine the current sentiments of members of the receiving community towards refugees from Syria by studying a set of indicators of socio-psychological integration, including intergroup attitudes, perceptions of realistic and symbolic intergroup threat, intergroup contact, support for refugees' rights, willingness to assist them, social proximity, perceptions of the frequency of refugees' experiences of discrimination, and perceptions of the extent to which refugees are members of society in the receiving countries (society membership). Most of these indicators have previously been shown empirically to be relevant to intergroup relations between the receiving community and refugees (e.g., positive intergroup attitudes), and some of these indicators have now been introduced to expand the picture of intergroup relations and perception of the receiving community of refugees (e.g., perceptions of refugees' society membership). The relationships among these indicators are illustrated in Fig. 1. They show that a number of them are interrelated, reflecting the complexity of different aspects of the relationships between the groups.

Fig. 1

Relationship between socio-psychological indicators of integration

Second, we aimed to contribute to previous empirical work by comparing the indicators of socio-psychological integration between four countries with different migration histories, macroeconomic situations, cultures, and roles in receiving Syrian refugees in the 2010s. The main advantage of comparing the data between these countries is to contextualize the findings and understand possible reasons for their patterns. Sweden and Germany have a long history of receiving migrants and a stable and prosperous macroeconomic situation. They are also very secular countries with multi-ethnic populations. They were the most attractive European destination countries for refugees in the 2010s. Croatia is more culturally homogeneous and the vast majority of its population are Croats and Catholics. Despite a relatively stable macroeconomic situation, the country has a much lower standard of living compared to Sweden and Germany. It is a transit country for refugees, although the number of refugees from Syria is small compared to other countries. Jordan is a Muslim country bordering Syria and is the first country of asylum for refugees from Syria. It has also been a transit country for many traveling to other destinations. At the end of 2011, Jordan ranked sixth on the list of the top 10 host countries for refugees in the world, with 451,000 refugees living there (UNHCR, 2012). By the end of 2013, this number had increased to 641,900 (UNHCR, 2014), and Jordan had the second highest number of refugees per 1000 inhabitants in the world (UNCHR, 2014). The macroeconomic situation and standard of living in Jordan are low compared to the selected European countries. Each of these four countries has at least one element in common with another study country (e.g., level of ethnic and religious diversity, macroeconomic situation), but also has a unique context. These overlaps can help us better understand intergroup relations as reflected by domicile community members.

The main research question we ask in this study is, 'How does the socio-psychological integration of refugees differ from the perspective of the receiving community members in Sweden, Germany, Croatia, and Jordan?' We aim to answer this question by comparing a set of indicators of socio-psychological integration related to thoughts, emotions and feelings, and behavioural intentions of the receiving community members in the context of integration, thus providing a broader framework for examining intergroup relations than in previous studies. Moreover, by comparing differences across the selected countries, we can draw parallels between contexts of integration and potentially extrapolate key elements common to receiving communities that influence refugee socio-psychological integration through an impact on the receiving community members.

While cross-national comparisons are not a novelty in research on socio-psychological elements of integration (e.g., Bansak et al., 2016; Bruneau et al., 2018; De Coninck, 2019; Hasbún López et al., 2019; Koos & Seibel, 2019; Maggini & Fernández, 2019; Van der Linden et al, 2017), this study is, to our knowledge, the first to compare a variety of indicators of socio-psychological integration in these countries and the first of its kind conducted in the European context.

Method

The data were collected as part of a broader project aimed at analysing the relationship between newly arrived refugees from Syria and long-term residents of Sweden, Germany, Croatia, and Jordan. Survey data on socio-economic and socio-psychological aspects of integration were collected from both populations. This paper focuses on the responses of receiving community members on the main socio-psychological aspects of integration, which are described below.

Respondents

Receiving community members' (RCMs) were defined for the European study sites as individuals who have citizenship or permanent residence in the respective European country and have lived in the same receiving country for at least two years prior to the arrival of refugees from Syria in 2015. In Jordan, RCMs were defined as Jordanian citizens (in Jordan, foreigners cannot obtain citizenship or permanent residence) and had lived continuously in Jordan for at least two years prior to the peak arrival of refugees from Syria in 2011. These restrictions limited the sampling to those RCMs who resided in the country and could develop an in-depth experience of living in the country. Qualifying criteria also included age: respondents had to be between 18 and 65 years old.

Data collection

Survey data were collected from RCMs in Sweden, Germany, Croatia, and Jordan in cities with the largest proportion of refugees from Syria (Gothenburg, Malmoe, and Stockholm in Sweden; Berlin, Hamburg, and Leipzig in Germany; Zagreb, Sisak, and Karlovac in Croatia; Amman, Zarqa, Irbid, and Mafraq in Jordan). These cities were selected to maximize the likelihood that refugees and RCMs would have first-hand experience with refugees. The sample size in each city was proportional to the population of the target area. Two probabilistic sampling techniques were used to select respondents. In Sweden, respondents were approached randomly using the national population register, with questionnaires sent by mail to respondents who could also answer them online. In Germany, Croatia, and Jordan, respondents were approached using probabilistic sampling of geographic clusters and Random Walk Technique.

Data collection began in November 2019 and ended in October 2020, with interruptions due to the outbreak of COVID -19 In Sweden, data were collected from June to September 2020. In Germany, data were collected in two phases—from December 2019 to March 2020 and from July 2020 to October 2020. Data were collected prior to COVID -19, from November 2019 to January 2020 in Croatia, and in January 2020 in Jordan.

Instruments and measures

Data collection materials consisted of a detailed letter with information about the research, an informed consent form, and a survey questionnaire. All materials were translated into Swedish, German, Croatian, and Arabic using a standard back-translation method and piloted on smaller, independent samples of RCMs with a total of 78 participants in the pilot study. The aim of the pilot study was to assess the comprehension, acceptability, feasibility, and applicability of the instruments.

The socio-psychological part of the questionnaires was designed to measure the following concepts.

Intergroup attitudes

Attitudes are complex organizations of beliefs, emotions, and behavioural tendencies directed toward someone or something that is relevant to us (Hogg & Vaughan, 2005), and include the way we think, feel, and behave or intend to behave toward others. They greatly influence how RCMs and refugees build and maintain relationships. Positive attitudes at the personal and whole-group levels promote positive interactions, rapprochement, and bonding and are at the very centre of socio-psychological integration. Attitudes are important not only for openness and facilitation of positive interactions at informal levels (neighbourhoods, schools, workplaces, public spaces), but also in formal settings such as institutions and services. Receiving community members who work in these settings and provide services to refugees, approach them based on their personal attitudes. Therefore, forming and promoting positive attitudes in all areas of life is beneficial to both groups.

The attitudes of RCMs towards refugees from Syria were measured using an adapted version of the Attitudes Towards Refugees Scale (Ajduković et al., 2019), which originally consisted of 19 items but was used in its shortened version with 6 items selected based on their saturation with the general factor. In this study, the measurement of intergroup attitudes focused on the cognitive and emotional components, while the behavioural intentions of RCMs toward refugees were explored with readiness to assist refugees and social proximity toward refugees, which are described in more detail below.

Respondents answered the questions using a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). All items were formatted so that higher agreement indicated more positive attitudes toward refugees, e.g., 'If a /country national/ and a refugee do equal work, it is fair that they receive the same wage.' The total score for each respondent was calculated as the mean of all responses, following the principle of 'using all available data' (Newman, 2014; Schafer & Graham, 2002). The scale showed good reliability in all countries (k = 6; ω = 0.717 to 0.861; α = 0.708 to 0.857).¹

Perception of intergroup threat

Perception of intergroup threat is a set of negative feelings and expectations about the way members of another group behave or influence one's own group. In their revised Integrated theory of intergroup threat, Stephan et al. (2009) distinguish between perceptions of realistic and perceptions of symbolic intergroup threat. Stephan et al. (2009) also argue that both minority and majority groups may perceive the threat posed by the other group that it manifests in the same way, but has a different origin due to power dynamics.

The Realistic and Symbolic Threat Scale (Ajduković et al., 2019) was used as a measure of perceptions of two types of threat-realistic threat to one's socio-economic resources and physical wellbeing, and symbolic threat to one's culture, norms, and way of life. A short form with six items was used, three of which measured the perception of a realistic threat, e.g., 'I am afraid that the crime rate in /Country/ might increase because of refugees; and three of which measured the perception of a symbolic threat, e.g., 'Refugees might threaten our values and way of life.' Respondents indicated their agreement with each statement on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). The total score per respondent was calculated separately for each type of threat as the mean of the responses. The subscale for perception of realistic threat showed reliability ranging from almost acceptable to good (k = 3; ω = 0.683 to 0.820; α = 0.645 to 0.800), as did the subscale for perception of symbolic threat (k = 6; ω = 0.680 to 0.896; α = 0.676 to 0.895).

Support for the rights of refugees

RCMs' support for refugee rights is related to the 'social connections' of the Indicators of Integration framework (Ndofor-Tah et al., 2019), as they represent supporting the refugees in exercising their rights and entitlements when communicating with institutions and services.

RCMs' views on refugees' rights and entitlements in their countries were measured using the Support for Refugees' Entitlements Scale (Ajduković et al., 2019). The scale consisted of statements that represented the legally guaranteed rights of refugees in the receiving country, e.g. 'Refugees in /Country/ should be allowed to get a job.' These statements were constructed by adapting the syntax of laws and regulations common to all four study countries into items used in this scale. Respondents indicated how much they agreed with each statement on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Because the scale depends on the rights recognized by law in each country, it was adapted for use in each of the four study countries. A total of 11 rights that were identical in all countries were included in the cross-country analysis, and the total score per respondent was calculated as the mean of the responses. It showed good reliability for all countries (k = 11; ω = 0.733 to 0.910; α = 0.728 to 0.906).

Intention of prosocial behaviour towards refugees

Based on the Theory of planned behaviour (Ajzen & Fishbein, 2000), behavioural intention directly precedes the behaviour and drives behaviour. Based on this theory, we would expect that individuals who express a willingness to help members of the other group would be more likely to provide assistance when the opportunity presents itself.

Intentions of prosocial behaviour toward refugees were measured using the Readiness to Assist Refugees Scale (Ajduković et al., 2019), which involves being in contact with refugees and actively helping them by using personal resources such as attention, property, time, and food to help refugees. The scale consisted of four items, e.g., 'I would bring food and/or other supplies to refugees,' to which respondents answered on a 5-point Likert scale (1 = Definitely not, 5 = Definitely yes). The total score per respondent was calculated as the mean. The scale showed good reliability in all samples (k = 4; ω = 0.833 to 0.878; α = 0.728 to 0.879).

Social proximity toward refugees

Social proximity is a measure of willingness to engage in various types of relationships with members of the other group. The measure was first introduced by Bogardus (1933) as "social distance" and has been used repeatedly in social research on intergroup relations (see Parillo & Donaghue, 2005 for a review of the use of the social distance scale).

It was measured using the adapted version of the Social Proximity to Refugees Scale (Ajduković et al., 2019) with five items that assessed respondents' willingness to accept a romantic relationship/marriage with a refugee, to accept a refugee as a family member, as a friend, as a neighbour and as a work colleague, e.g., 'I would accept a refugee as a family member.' Respondents answered 'Yes' or 'No' to each statement. The total score was determined based on the highest level of intimacy to which the respondent answered 'Yes'. Each level of intimacy was defined with a value, with the highest level of intimacy (love/marriage relationship) coded as 5 and the lowest (work colleague) coded as 1. If the respondent indicated that they were not willing to engage in any type of relationship with refugees, their total score was marked as 0. Due to the nature of the scale and the method used to create the total score, the reliability of the scale was not calculated.

Perceptions of discrimination against refugees

The perception of RCMs of the frequency with which refugees are treated unequally in the receiving country was measured using an adapted scale based on the Longitudinal Survey of Immigrants to Canada (Statistics Canada, 2003). Though this survey does not primarily address discrimination, this particular scale was incorporated to capture perceived discrimination among immigrants in Canada because it is considered an important dimension of their lives. The instrument is based on similar constructs that have been used in international studies (Liebkind et al., 2000; Slobodin et al., 2021) and in international surveys such as the SOEP. As described by Quillian (2006), the goal of this instrument is to measure "the difference between the treatment a target group actually receives and the treatment they would receive if they were not members of the target group but were otherwise the same" (Quillian, 2006, p. 302; see also National Research Council, 2004; Pager, 2007). Our adaptation of the scale involved adjusting the wording to reflect the RCM's views on discrimination against refugees in seven areas of life (i.e., 'To what extent do you believe Syrian refugees are treated unequally compared to /Country nationals/ in a store/restaurant/bank; when applying for a job or promotion; when dealing with the police or courts; at school; when looking for housing, in sports/leisure activities; in hospitals, or by health care workers'). Respondents indicated on a 5-point Likert scale (1 = never, 5 = very often) how often refugees are treated unequally compared to the country nationals. The total score per respondent was calculated as the mean. The scale showed excellent reliability in all samples (k = 7; ω = 0.844 to 0.919; α = 0.843 to 0.918).

Perception of refugees' belonging to society

RCMs' perceptions of the extent to which refugees are part of the society in which they live were measured with an item: 'How much do you feel that refugees are part of the /country's/ society' on a 5-point Likert scale ranging from 1 = not at all to 5 = very much.

Quantity and quality of intergroup contact

Intergroup contact is a set of interactions between people who perceive themselves as part of different groups. The contact hypothesis and Theory of intergroup contact (Allport, 1954) states that pleasant contact with members of the other group has a positive effect on attitudes toward members of the other group, and scholarly research has confirmed this in the context of refugee-community relations (Barlow et al., 2012; Healy, et al., 2017; Turoy et al., 2013). Superficial contact, on the other hand, is not always correlated with attitudes or may even be negatively associated, especially among individuals with unpleasant contact experiences, while the correlation between closed relationships and attitudes is positive (Rafiqi & Tomsen, 2021).

For this study, a measure of the quantity and quality of contact between groups was developed. It included ten items—five for quantity and five for quality of contact in different contexts—public transportation or streets and market, neighbourhood, work, school and educational institutions, and public events. These contexts were chosen because they are part of daily life and provide space and opportunity for individual and group interactions.

Respondents rated both characteristics of intergroup contact using 5-point Likert scales (for quantity-1 = never, 5 = very often; for quality-1 = very negative, 5 = very positive). Because each of these items measures the quality or quantity of contact in a very different context, it is unlikely that respondents would answer uniformly, as is the case, for example, with attitudes between groups, so they are not tau-equivalent. In other words, the frequency with which a person encounters members of the other group in the neighbourhood is not necessarily related to the frequency with which he or she encounters them in another context (school, work, market), so there is no consistency to be expected between items that would allow a single score to be determined for each respondent. Instead, each item was treated as a separate measure of intergroup contact, which allowed us to examine each context in detail, so the reliability of the scale was not calculated.

Results

Socio-demographic and socio-economic characteristics of the samples

A total of N = 3025 RCM individuals from all four countries participated in the survey. The majority of the sample was from Sweden (42.2%), followed by Jordan (20.6%), Croatia (19.8%), and Germany (17.3%). In the overall sample, gender was balanced with 50.9% females. The average age of RCMs in the study countries was 42.6 years. The majority of the total sample reported having a tertiary level of education (52.9%), followed by a secondary level of education (42.0%) and a primary level of education (4.6%), but the proportion of respondents with different levels of education varied across the samples. The proportion of respondents with a migrant background (i.e., previous migrant cohorts) also varied across countries: Sweden had the highest proportion (34%), followed by Germany and Jordan (24%), while Croatia had the lowest proportion (17%).

The majority of all RCMs were employed at the time of data collection (65.4%), although there were significant differences between countries.

Differences can also be seen in the mean values of the total monthly income of the respondents' households, with the highest income in Sweden (M = 5151.63€. SD = 2881.99€, N = 1230) and the lowest in Jordan (M = 499.88€. SD = 320.18€, N = 624). These differences are consistent with the macroeconomic situation of the countries.

Political orientation was measured only in the European context, and respondents were mostly left-leaning (43.9% of the total sample of Sweden, Germany, and Croatia). The percentage of left-leaning respondents was highest in Germany (61.3%) (in Sweden 40.8%, in Croatia 33.1%). Due to the nature of the discourse on migration and displacement in Germany, RCMs with more positive attitudes toward refugees were likely more willing to participate in the study, possibly leading to positively skewed responses on measures of intergroup relations.

Finally, RCMs in Sweden and Germany indicated on average that religion was "not at all" or "slightly" important in their lives, while it was "moderately" important on average in Croatia and "quite" important in Jordan.

In interpreting the results of the cross-country comparisons, we took into account the specifics of each study country (e.g., the macroeconomic situation) and the sample surveyed (e.g., the proportion of left-leaning respondents). Table 1 shows the basic socioeconomic characteristics of the samples per country.

Table 1 Socio-demographic and socio-economic characteristics of the samples across four study countries

Comparison of socio-psychological integration indicators across countries

Intergroup attitudes RCMs across all four countries differed in their attitudes toward refugees (F(3, 3018) = 158.31, p < 0.01). Respondents in Germany (M = 4.20, SD = 0.667, n = 523) had the most positive attitudes toward refugees from Syria, followed by respondents in Sweden (M = 3.82, SD = 0.858, n = 1275) and Jordan (M = 3.49, SD = 0.844, n = 624). RCMs in Croatia reported the lowest scores and showed neutral attitudes toward refugees from Syria in Croatia (M = 3.21, SD = 0.873, n = 600). The differences between all countries were significant.

Perception of intergroup threat

Because the scales measuring this construct did not have sufficient reliability levels, statistical comparisons were not made between countries. Instead, the results for each country can be viewed separately and in comparison to the midpoint of the response scale. Respondents in the different countries perceived the realistic threat posed by refugees differently (F(3, 3018) = 191.51, p < 0.01). RCMs in Jordan perceived the realistic threat posed by refugees from Syria most strongly compared to the other countries, slightly above the midpoint of the scale (M = 3.43, SD = 1.171, n = 624). The same was true for respondents in Croatia, who also had a mid-range level of realistic threat perception (M = 3.16, SD = 0.960, n = 600). The overall scores of RCMs in Jordan and Croatia were in the middle of the scale. Respondents in Sweden (M = 2.49, SD = 1.100, n = 1275) and Germany (M = 2.20, SD = 0.876, n = 523) showed lower levels of realistic threat perception, with scores varying in the lower half of the scale.

In terms of symbolic threat perception, respondents in Croatia scored above the middle of the scale and generally perceived refugees as somewhat threatening to their culture and customs (M = 3.42, SD = 0.960, n = 600). Respondents in Sweden (M = 2.78, SD = 1.212, n = 1275) and Jordan (M = 2.73, SD = 1.100, n = 624) were slightly below the mid-point of the scale. Finally, RCMs in Germany reported a symbolic threat from Syrian refugees that was in the lower half of the scale (M = 2.53, SD = 0.977, n = 523).

Support for refugee rights

RCMs in four countries differed in their level of support for refugee rights (F(3, 3019) = 181.43, p < 0.01). Respondents in Germany (M = 4.46, SD = 0.502, n = 523) showed the highest level of support. Swedish (M = 3.82, SD = 0.836, n = 1276) and Jordanian RCMs (M = 3.73, SD = 0.733, n = 624) did not differ significantly but also supported refugee rights. However, respondents in Croatia showed the least support for refugee rights (M = 3.41, SD = 0.845, n = 600), but similar to other indicators of socio-psychological integration, the level of support indicated that respondents were neutral.

Readiness to assist refugees

RCMs in the four study countries differed significantly in their readiness to assist refugees (F(3, 3016) = 113.37, p < 0.01). Respondents in Jordan showed the highest willingness to assist refugees (M = 4.02, SD = 0.937, n = 624), followed by respondents in Germany (M = 3.62, SD = 0.909, n = 523), Sweden (M = 3.37, SD = 1.083, n = 1274), and Croatia (M = 2.98, SD = 1.045, n = 600).

Perceptions of discrimination against refugees

RCMs in four countries differed significantly in their perceptions of the frequency with which refugees are discriminated against (F(3, 3007) = 543.666, p < 0.01). Respondents in Sweden (M = 3.41, SD = 0.882, n = 1265) and Germany (M = 3.23, SD = 0.882, n = 522) had the same perception of this frequency and believed that refugees were discriminated against more often than respondents in Croatia (M = 2.33, SD = 0.941, n = 600) and Jordan (M = 1.89, SD = 0.871, n = 624), who also differed significantly among themselves.

Perceptions of society membership of refugees

Respondents in the four countries differed significantly in their perceptions of the extent to which refugees from Syria are a part of the community in which they live in the receiving country (F(3, 3006) = 211.173, p < 0.01). Respondents in Jordan had the highest perception of Syrian refugees as part of the community compared to the other countries (M = 3.41, SD = 1.071, n = 624), followed by respondents in Sweden (M = 2.88, SD = 1.021, n = 1263), Germany (M = 2.55, SD = 0.835, n = 523), and Croatia (M = 2.05, SD = 0.893, n = 600). The differences between all countries were significant.

Figure 2 presents the means of the indicators of socio-psychological integration across four study countries.

Fig. 2

Means of indicators of socio-psychological integration per study country

Quantity of intergroup contact

Statistically significant differences in the quantity (frequency) of RCMs' contact with refugees from Syria were found between the four study sites for all five contexts—in public transport/on the street/at the market (F(3, 2758) = 485.345, p < 0.01), in the neighborhood (F(3, 2824) = 227.554, p < 0.01), at work (F(3, 2291) = 144.687, p < 0.01), at school/college/educational institution (F(3, 1348) = 217.780, p < 0.01), and at public events (F(3, 2516) = 165.558, p < 0.01). While contact occurred in Sweden, Germany, and Jordan, with mean scores for each context varying between 'rarely' and 'frequently', it was very rare in Croatia, with scores varying between 'never' and 'rarely', which was related to the very low number of refugees in the communities. In all countries, contact was rarest in the workplace and in education, while it was more frequent but superficial in public institutions. This resulted in many cases missing from the data in all countries. The percentage of missing data ranged from 0.33% for RCMs' estimates in Croatia on the frequency of contacts in the neighborhood to 77.40% of RCMs' missing estimates in Jordan on contacts in educational institutions.

Quality of intergroup contact

Similarly to the quantity, there were differences between countries in the quality (degree of pleasantness) of contact between RCMs and refugees in all contexts—in public transport/on the street/at the market (F(3, 2512) = 52.877, p < 0.01), in the neighborhood (F (3, 2130) = 37.604, p < 0.01), at work (F(3, 1332) = 6.433, p < 0.01), at school/college/educational institution (F(3, 776) = 3.501, p < 0.05), and at public events (F(3, 1984) = 59.359, p < 0.01). In general, the average responses of RCMs in all countries ranged from "neither positive nor negative" to "positive." Interpreted alongside the quantity of contacts, this indicates that RCMs tend to have infrequent contact with refugees, but when they do, they appear to be neutral to positive. Because the number of valid responses varied significantly between and within countries, particularly in Croatia, where valid Ns ranged from 39 (93.5% missing data) to 374 (37.6% missing data) per context, and between contexts-particularly for work and education activities, where valid Ns declined in all countries, caution is warranted when interpreting results on the quantity and quality of contacts.

The means of quantity and quality of contact per country are presented in Fig. 3.

Fig. 3

Means of quantity and quality of contact for each context per study country

Discussion

Key socio-psychological indicators of refugee integration from Syria and receiving communities were studied in Croatia, Germany, Jordan, and Sweden, using the same operationalization of constructs and measures. Our aim was to discover patterns in the cross-national results that could help us understand which context-specific characteristics might be related to the integration of the receiving community and refugees at the level of their everyday interactions. This study found differences in the socio-psychological integration of refugees from Syria across the four countries with different histories of immigration, macroeconomic situations, cultural and religious backgrounds, and experiences with refugees from Syria.

Among the four countries, most indicators of socio-psychological integration were more favourable in Germany, followed by Sweden and Jordan, while they were least favourable in Croatia. RCMs' attitudes toward refugees from Syria were most positive in Germany, followed by Sweden and Jordan, while they were neutral in Croatia. RCMs' support for refugees' rights and entitlements and RCMs' social proximity towards refugees showed similar results. While respondents in Germany, Sweden, and Jordan showed readiness to engage in closer relationships such as family and love relationships, respondents from Croatia remained mostly neutral in their sentiments or preferred to remain only at the level of friendship with refugees from Syria.

The overall neutral or reserved attitude of Croatian participants could be related to the lack of information or interest in refugee integration. This, in turn, is likely a result of the very rare opportunity to interact with Syrian refugees, as their total number at the time of data collection was around 600, including children (Croatian Ministry of Interior, 2022). It seems likely that the number of refugees in the receiving country influences the general sentiment in the receiving society, even though the challenges of integrating refugees in various receiving countries are rather the same. This seems consistent with an overall finding that interaction with refugees correlates with positive attitudes toward refugees (see Additional file 1 for more information).

RCMs in Germany clearly stand out from RCMs in other countries, showing very positive attitudes toward refugees from Syria, possibly due to the higher proportion of politically left-leaning respondents in the sample. Left-leaning individuals have been shown to have more positive attitudes toward migrants and refugees (Anderson, 2018; Cowling et al., 2019; Graf & Sczesny, 2019; Gregurović et al., 2016; Pedersen et al., 2005; Sunhan et al., 2012), stronger support for their rights (Canetti et al., 2016; Hercowitz-Amir & Raijman, 2020), and greater social closeness to them (Bruneau et al., 2018). Consistent with this trend, RCMs in Germany showed the lowest levels of perceived realistic threat, followed by RCMs in Sweden, Croatia, and Jordan. Perceived realistic threat is defined as the assessment of danger to a group or individual's resources, such as jobs, school/university quota, and personal safety and physical integrity. This has previously been shown to be negatively related to attitudes toward refugees and migrants (Berndsen et al, 2018; Cowling et al, 2019; Geschke, 2007; Mancini, 2018; Murray & Marx, 2013; Schweitzer et al, 2005) and the readiness to assist refugees or engage in other types of prosocial behavior (Ajduković et al, 2019; Mancini, 2018). RCMs in Jordan perceived refugees from Syria as a greater realistic threat than respondents in other countries, which could be explained by the general macroeconomic situation in Jordan. The poor job market, economic situation with a high deficit, a high percentage of unemployment, and a low supply of jobs may have led to the perception of a potentially greater negative impact of refugee empowerment in this country. Respondents in Croatia also showed a moderate perception of a realistic threat, likely due to the country's worse socioeconomic situation and the problem of unemployment in the labor market, although potential competitors among refugees were very few. Nevertheless, the results in both countries fluctuate around the neutral position of respondents ('neither agree nor disagree'), indicating some degree of uncertainty or reluctance to perceive refugees as a potential threat.

Having a more stable socio-economic situation and a long history of receiving migrants, RCMs in Germany and Sweden showed lower levels of perception of a realistic threat, with respondents mostly disagreeing with the statements describing the potentially negative impact of the refugee arrival and integration in the receiving country. This trend supports the notion that RCMs in these two countries, which have more prosperous economies and higher living standards, are less likely to believe that their resources will be unjustly allocated to aid the newcomers and threaten their socio-economic wellbeing. It also shows that the factual number of refugees in the receiving country is not necessarily related to the strength of these negative perceptions: For instance, Croatia has a very low presence of refugees from Syria versus Jordan.

The perception of symbolic threat was quite prominent in Croatia, showing lower levels of threat to the RCMs’ way of life, norms and culture. This is of no surprise taking into account that Croatia is substantively a mono-cultural country with a very high proportion of ethnic Croats, who mostly self-declare as Christians. The prominent religion is a relatively conservative national variant of Catholicism in essence promoting ethnic closeness. Furthermore, the war following the breakup of former Yugoslavia also in-part involved conflict with Muslims (Bosniaks) in Bosnia and Herzegovina. In some national narratives, this was perceived as a conflict with Islam, which is the dominant religion of refugees from Syria. Moreover, Croatia has a limited experience of inward migration from other cultures. These factors could have led to a greater perception of symbolic threat in the Croatian receiving community than in other countries, although the size of the overall refugee community in Croatia is very small.

RCMs in Sweden and Jordan both showed moderate and similar levels of perceived symbolic threat. Even though the results are numerically similar, the dynamic of the perception of symbolic threat may differ between these countries. In Sweden, it could be related to the widespread liberal attitudes, but also the increase in cultural diversity resulting from migration over many years. In Jordan, a traditionally culturally more conservative society, the cultural difference between the receiving and refugee communities is smaller, but could still be viewed as threatening. Finding the perception of symbolic threat least expressed by RCMs in Germany corresponds with the positive attitudes towards refugees from Syria. The perception of symbolic threat is related to several socio-psychological indicators of integration and is especially relevant in the multicultural context in which the differences in culture and religion are perceived as important. It is negatively related to attitudes (Cowling et al., 2019; Geschke, 2007; Mancini, 2018; Murray & Marx, 2013; Schweitzer et al., 2005), readiness to assist refugees and behave in a prosocial way (Ajduković et al., 2019; Mancini, 2018), and social proximity (Ajduković et al., 2019). Previous research has also demonstrated that the perceptions of the realistic and symbolic threat posed by refugees and migrants are significantly correlated (Geschke, 2007; Mancini, 2018), which is further supported by this study (see Additional file 1 for detailed information on the correlations between the studied constructs).

The readiness of the RCMs to assist refugees shows the degree to which the RCMs estimate they are prepared to provide concrete help to the refugees. RCMs in Jordan showed the highest levels of readiness to assist refugees from Syria, followed by the RCMs in Germany, Sweden and Croatia. This could be related to the historical relations between Jordan and Syria, the similarities in religion and traditions. Even though the differences between the countries are significant, it is worth noting that the results fall between the neutral and somewhat positive levels for all respondents.

The reluctance of RCMs in Croatia to assist the refugees corresponds to their neutral attitudes and support for the rights of the refugees, as they likely did not have the opportunity to interact with the refugees, and form bonds or a strong opinion.

In comparison to Croatia and, in particular, Jordan, the RCMs in Sweden and Germany reported believing that the refugees from Syria experience frequent discrimination. It is important to emphasize that this was not a measure of the actual experiences of discrimination of refuges, nor the measure of the frequency with which the RCMs treat (or believe they would treat) refuges unequally, but a measure of their belief of refugee exposure to discrimination in their country. Therefore, the differences between the countries could be due to the sensitivity of the RCMs to social justice and equality, and their understanding of discrimination or unequal treatment. As noted before, this could also be related to the characteristics of the samples: in Germany, more participants were politically left-oriented, and in Sweden, the participants had mostly a tertiary level of education. In line with the previous research that showed that the level of education and left-wing political orientation are related to positive indicators of intergroup relations (e.g. Cowling et al., 2019; Hartley & Pedersen, 2015; Pedersen et al., 2005, Sunhan et al., 2012), probably, these persons are also more aware of the potentially discriminatory practices and sentiments in their community.

Interestingly, RCMs in Jordan estimated that the refugees are a part of the society they live in more than did RCMs in Germany, Sweden and especially Croatia, though the responses of the RCMs in Jordan varied between ‘moderately’ and ‘quite’. These differences may originate from the number of refugees in each country and the intercultural similarities between the receiving and refugee communities. Jordan has hosted the greatest number of refugees from Syria since the start of the Syrian civil war (UNHCR, 2017), and is culturally more similar to Syria than are European countries. This interpretation is supported by the (opposite) results from Croatia. Nevertheless, perception of the frequency of discrimination of refugees in the respective receiving countries was consistently low across the study countries, ranging from “rarely” in Jordan and Croatia to “sometimes” in Germany and Sweden. It is questionable whether the receiving community is aware of the actual discrimination and racism experienced by the refugees, further emphasized by the previously empirically presented tendency of refugees to downplay the severity of their experiences (Parker, 2018).

Intergroup contact, particularly the quality of it, was previously shown to be highly impactful on attitudes toward refugees and migrants (Graf & Sczesny, 2019; Healy et al., 2017; Schultz & Taylor, 2018; Turoy et al., 2013), support for the rights of refugees (Schulz & Taylor, 2018), and the perception of threat (Hercowitz-Amir et al., 2017; Turoy et al., 2013).

The quantity of contact was also found significantly related to other socio-psychological indicators of integration such as attitudes (Barlow et al., 2012), and readiness to assist refugees (Ajduković et al., 2019), and social proximity (Bagci et al., 2020). This study has shown that the quantity of contact varies strongly across the countries but is not as frequent as we would expect even in the countries where the number of refugees is large, such as Jordan, where the frequency was estimated between ‘sometimes’ and ‘frequently’. Furthermore, RCMs and refugees across all countries more often came into contact in public places than at work or at an educational facility which does not allow for deeper and more significant interaction. It seems, rather, that these encounters remain on a superficial level. Intergroup contact is the rarest and most negative in Croatia in comparison to other countries, though the actual perception of the RCMs in Croatia is that the contact, when it happens, is neutral—neither positive nor negative. Contact was generally most positive in Jordan, except for the workplace where the RCMs from Sweden, Germany and Jordan provided similar estimations of its degree of pleasantness and the educational context for which the RCMs in Germany provided the most positive perceptions. There are much greater differences between the countries in the actual frequency with which the refugees and RCMs come into contact than in the perception of the quality of that contact. Still, two methodological questions on the study of contact remain. First, can RCMs differentiate between refugees from Syria and refugees of other ethnicities and provide accurate estimations of their contact with refugees from Syria alone? And secondly, how did the respondents understand the answering options? In other words—how often is ‘rarely’ or ‘frequently’ and how positive is ‘positive’?

In general, several patterns emerged throughout different socio-psychological indicators of the integration of refugees and RCMs. Germany and Sweden showed a positive stance towards the refugees including positive attitudes, support for the rights of refugees, closer social proximity and awareness of discrimination against refugees. Consequently, they reported lower perception of realistic and symbolic threat. Jordan showed attitudes approaching the positive pole, relatively neutral level of perceived symbolic threat, highest readiness to assist refugees, and a somewhat high view that the refugees are a part of the Jordanian society. But, similarly to Croatia, RCMs in Jordan perceived a higher realistic threat than in Germany and Sweden, social proximity mostly at the level of friendship, and considered the refugees to be rarely exposed to discrimination. Croatia stood out as a country with the lowest indicators of socio-psychological integration, the highest perception of symbolic threat and the belief that the refugees from Syria are not a part of the Croatian society. Intergroup contact was relatively rare in all countries, though, as expected, more prevalent in Jordan, Germany and Sweden. At the same time, these experiences were heterogeneous, ranging from neutral to positive depending on the context and country.

In general, these findings have several potential implications for the integration practice. Cultural similarity between the receiving society and the incoming refugees may facilitate overall positive intergroup relations, as shown in Jordan where most integration indicators were good (positive attitude towards refugees, readiness to assist them, high social membership, pleasant contact experiences, support for their rights and entitlements, low intergroup symbolic threat, close social proximity). However, competition for resources given the unfavourable economic conditions in a country increases the perception of realistic threat from refugees and hinders integration as a two-way process. On the other hand, in socio-economically prosperous societies, such as Sweden and Germany, with high awareness of human rights, almost all integration indicators are very good, the RCMs feel neither symbolic nor realistic threat from refugees. Nevertheless, they do not perceive refugees as members of the common society, notwithstanding considerable history of immigration and the need for workforce. In societies where inward immigration is a new experience, such as Croatia, the refugee integration is an uncharted notion, resulting in neither positive nor negative views towards it, and some apprehension about impact on the own culture. Under such circumstances the concern about competition for resources increases perception of realistic threat which impairs the integration process. Thus, different obstacles to smooth socio-psychological integration were identified in the four countries, indicating that flexible approaches to its facilitation should be used and that socio-economic dimensions of integration have to be taken into account.

Receiving community members’ part of the two-way integration process should be tackled through systematic efforts to disperse concerns about the intergroup threat from the incoming refugees in countries where it is present (Croatia, Jordan). This can be done thru transparent information aimed at the general public regarding allocation of resources earmarked for the integration and paying attention within public campaigns to the feelings of their fair distribution. Intergroup positive attitudes of RCMs could be encouraged by promoting the fact that refugee rights do not endanger the rights of RCMs, since support for the rights and positive attitudes are closely linked. Moreover, creating opportunities for more substantial intergroup contact may be a channel to promote social membership (Sweden, Germany). In addition to the contact offered by educational and work related settings, public events that provide opportunity for more interaction under pleasant circumstances, such as music and food festivals, could foster intercultural communication and intergroup relations. In turn, such intergroup experiences have a potential to contribute to building trust and reciprocity (IFRC PS Centre, 2022), foster positive mutual perceptions and attitudes, and advance social membership.

Limitations of the study

Studies of intercultural differences are prone to limitations, and this investigation of cross-country socio-psychological integration conducted in four very different socio-economic and political contexts is no exception. Firstly, the participants were approached using one of the two probabilistic sampling techniques—the Random Walk Technique or randomly from a national population register. Two main issues are related to the sampling methods: the risk of self-selection bias and the difference in approaching the participants. This could have impacted the sample characteristics, such as the overrepresentation of the left-wing participants in the German sample, or the high number of participants with tertiary education in the Swedish sample. Therefore, it is crucial to interpret the findings by taking into account the potential impact of the characteristics of the sample.

Secondly, the technique of data collection differed among the countries: Germany, Croatia and Jordan used the Computer-Assisted Personal Interviewing technique (CAPI), while a pen-and-paper and online questionnaire were used in Sweden. The answers of the respondents to a postal or online survey were perhaps less under the influence of a socially-desirable answer than the responses collected in-vivo by an interviewer but may have affected the sample educational composition. In the personal interviewing approach, we aimed to mitigate such effects by training the interviewers to remain neutral concerning the answers provided by the respondents. Additionally, the information letter with details on privacy and data protection was handed to and read by the respondent before they granted a written consent to partake in the study, thus following principles of ethical science and data protection protocols.

Thirdly, the data were collected at different times largely due to the restrictions imposed by the response to the COVID-19 pandemic. In cross-site research, strictly aligning the timing of the data collection is improbable due to the specifics of the study sites (e.g. the time needed to obtain ethical clearance or to collect the targeted number of questionnaires at different sites).

Nevertheless, the sample size in all study sites had sufficient power to allow for legitimate comparisons, except for contact. Using the same set of questionnaires with validated scales that have been carefully translated into four languages also enabled legitimate comparison across countries.

Another potential limitation are interconnections among different socio-psychological indicators of integration. Intergroup relations depend on many cognitive, emotional and behavioural factors which are complex and multidimensional, but also interrelated. This complexity of psychological processes can blur the separate relationships between the particular constructs resulting in high intercorrelations between them. In socio-psychological research, such interdependency of factors is expected, and should be addressed and further explored.

Conclusions

The goal of this cross-site study was to detect differences in the socio-psychological indicators of integration across four study countries and interpret them keeping in mind the context of each study site. These differences between the countries are reflected in the findings of the cross-country comparisons of attitudes, perceptions of realistic and symbolic threat, support for the rights of the refugees, readiness to assist refugees, social proximity, perception of discrimination against refugees, perception of society membership of refugees and contact.

Though focused on comparisons between the countries, this study also points to the country-specific results and offers practitioners and scientists in the study countries the baseline findings regarding indicators of socio-psychological integration between RCMs and refugees as seen by RCMs, upon which further actions can be taken.

Germany and Sweden are multi-cultural, and economically stable countries with a long history of inward migration, and well-developed integration policies. Croatia and Jordan are rather mono-cultural and economically less stable. Croatia has a very short history of inward migration, primarily related to the migration of the 2010s, and a much greater experience of outward migration, and only a very small number of refugees which drastically limits the contact between the RCMs and refugees. Jordan has a very high number of refugees from Syria.

Related to these country characteristics, RCMs in Germany and Sweden generally showed a more positive stance towards refugees from Syria, and, in comparison to the macroeconomically less stable Croatia and Jordan, and trends of lower levels of perception of realistic threat. Rather than the actual macroeconomic situation of the receiving country, perhaps the perception of the RCMs of said macroeconomic situation could be very influential in predicting the perception of threat to one’s resources and physical safety.

Additionally, the RCMs in Sweden and Germany were particularly sensitive to the question of unequal treatment of the refugees, which could be the result of sampling and/or the socio-political situations in those countries—the German sample had a disproportionate number of politically left-wing oriented respondents, while the Swedish sample had a majority of respondents with a tertiary level of education. It is possible that the rise and vocality of the right-wing political parties on the topic of migration led to a greater proportion of respondents who wished to provide their (positive) opinion on the matter of integration of refugees from Syria in their countries.

Croatian RCMs showed the most neutral stance, which is most likely related to the overall low number of migrants and refugees in Croatia, which did not allow the RCMs to develop strong opinions, or were reluctant to provide them honestly. Still, RCMs in Croatia displayed the highest levels of perception of symbolic threat, which could also be related to the lack of intercultural contact, and the fact that the population in Croatia is rather homogenous in ethnicity, culture and religion. At the same time, Croatia is culturally more dissimilar to Syria than other countries. In comparison, Jordan, too, is a largely culturally homogenous country, but more similar to Syria, which could explain the country differences in the perception of symbolic threat.

It seems that the intercultural experience and cultural and religious composition of the receiving community, alongside the socio-economic situation of the country, play an important role in the stance of RCMs towards refugees. The characteristics of the receiving country should be taken into account when considering the socio-psychological integration of the RCMs and refugees, as shown through this study of differences between Sweden, Germany, Croatia and Jordan.