Introduction

When addressing multiplicity and plurality in a society, the term diversity has been coming to the front with increasing popularity in the scientific literature for more than a decade. Relatedly, as fights against exclusion and discrimination, and attempts to resolve conflicts between social groups have gained prominence, inclusion has been suggested as a strategy to manage diversity (e.g., Syed & Özbilgin, 2020). An ample of empirical studies have focused on how diversity is perceived and managed to ensure inclusiveness, especially in the contexts of education (e.g., Petriwskyj, 2010) and work (e.g., Farndale et al., 2015; Kirton & Greene, 2022). However, inclusion in the context of urban space is still a less traveled road, although everyday usage of urban space has been a well-studied topic since the 1970s in social sciences (e.g., Burgers, 2000; Çağlar & Glick-Schiller, 2018). Thus, the question of whether social groups (including immigrants in general and refugees in particular) expand their daily life to the whole city (i.e., usage and enjoyment of urban space), or whether they practice their daily routines in a restricted space (i.e., segregated) is worth asking. We aimed to answer this question in terms of inclusion by inquiring about the extent to which it is accessed by social groups in urban spaces where diversity is one of the main attributes of contemporary urban life. To reach this goal, we adopted a homemaking approach (see e.g., Boccagni, 2022a; Boccagni & Brighenti, 2017; Boccagni & Duyvendak, 2021) that provides us a theoretical framework to better understand the phenomenon of inclusion and managing diversity in cities. Looking at the homemaking activities of people as a non-detachable process in human life (see Beeckmans et al., 2022) informs us about whether and to what extent diverse social groups are included in the context of urban space.

Diversity in cities in the contemporary era is largely due to migration across the globe (e.g., Bauman, 2016; Glick-Schiller & Çağlar, 2010). Immigrants and non-immigrant groups cohabitate in cities (Bauman, 2018; see also Bauman, 2013 for glocalization). Therefore, in the current research, we focus on the inclusion of immigrants (specifically refugees) in the urban space by comparing their homemaking activities to their non-immigrant counterparts (i.e., nativesFootnote 1). Furthermore, today’s scientific knowledge has mainly concentrated on the Global North, but in this study, we focused on the issue in a country outside this geographical and cultural scope. We conducted our research in a Mediterranean country, i.e., Turkey. There is no consensus in the literature on how to locate Turkey in the globe, whether as a Middle Eastern or European. We prefer the term Mediterranean to refer to its in-betweenness in terms of the cultural construction of self (see Uskul et al., 2023). This region of the world has been understudied regarding diversity and inclusion. Finally, we concentrated on the expectations and experiences of both sides of migration (i.e., refugees and natives in the current work) to better understand the paradox of diversity and inclusion that is observed in their homemaking activities.

Homemaking

Human beings are endowed with the capacity of homemaking (e.g., Bonfanti et al., 2022), have spatial agency (Beeckmans et al., 2022), and make homes everywhere (e.g., Ahmed et al., 2003) independent of whether they claim it willfully or not (see also Gezici Yalçın et al., 2023). Homemaking is a spatial term since it needs space to be practiced. It is practiced in the whole urban space spanning from dwelling to the city center (see Fig. 1). Instead of the traditional Western split between public and private spheres, we adopt the perspective of scholars who argue that homemaking connects domesticity to urban public spaces (see e.g., Beeckmans et al., 2022). Homemaking is realized through daily life activities (see e.g., Boccagni, 2017; Boccagni & Brighenti, 2017) that can be exemplified by working, schooling, shopping, and visiting friends and relatives.

Fig. 1
figure 1

The Scale of Homemaking (Spatial Practices)

Several terms have been suggested in the literature to address homemaking activities such as the appropriation of space (Boccagni, 2022b; Kalandides & Vaiou, 2012), use of the public sphere or ownership of public space (Boccagni, 2022a), spatial practice (Beeckmans et al., 2022), and homemaking practices (Bilecen, 2017). In this research, we define homemaking as the usage of urban space for a particular period to carry out daily activities resembling the account of the action space approach by Dangschat et al. (1982). Whenever we talk about homemaking we also talk about the use and ownership of the public sphere. Furthermore, we categorize urban space in the spatial scales like dwelling, neighborhood, and city center as suggested by Ünlü Yücesoy (Ünlü Yücesoy, 2006: 133): “Each urban public space in these spatial scales marks different facilities and activities, as well as distinct social and spatial qualities. Therefore, the control, conduct, and experience of these urban public spaces are diverse.” On the other hand, we do not disregard the complexity of daily life and the transcendental aspect of homemaking between public and private spheres (Beeckmans et al., 2022) as in the cases of, for example, establishing personal relationships in the city center or establishing impersonal social relations at home through online shopping or social media usage. In addition, the use of urban space refers also to legal access of all persons living in cities to public spaces like streets, sidewalks, parks, buildings, plazas, etc. (see Lofland, 1973). More straightforwardly, we argue that each level of categorization of urban space refers to different modes of social encounters and interactions.

Urban spaces bring people from diverse backgrounds with the possibility to contact others (Low et al., 2005). It ordinarily creates abundant opportunities for interacting with people (Demerath & Levinger, 2003). Time spent in urban public spaces has been found to increase attachment and sense of community among people, as well as higher levels of perceived health and lower levels of loneliness (Cattell et al., 2008). Three visits to urban public spaces per week was associated with having a network of contacts of a one-half standard deviation higher in diversity when compared with the average (Hampton et al., 2011).

Nevertheless, different social actors and groups have unequal access to and use of urban public space and therefore have different opportunities to make themselves at home (Boccagni, 2022a). Previous research shows that spatial usage differs between genders (e.g., Yon & Nadimpalli, 2017; see also Fenster, 2005; Scraton & Watson, 1998 for the term gendered cities). Furthermore, due to residential segregation, immigrant groups have little access to urban public space compared to natives (e.g., Castenada, 2018; Huttman et al., 1991). In the following, we will outline the segregation in urban space and its impact on the homemaking of immigrants and natives.

Paradox of Homemaking

Migration usually generates a conflict in the arrival country which reveals itself in the paradox between keeping of existing home of natives versus searching for a new home of immigrants. That is, natives usually perceive threat from the newcomers and are motivated to seclude their houses, neighborhoods, and cities from immigrants (e.g., Abeywickrama et al., 2018; Göregenli & Karakuş, 2014). On the other hand, immigrants are searching for a new place to settle down, be safe and secure, and fulfill their needs through daily life activities (e.g., Kaya, 2017; Kılıçarslan, 2016; Mazumdar et al., 2000). The use of places where these daily activities are realized informs us about the level of inclusion in the city. Research shows that most encounters between natives and immigrants take place at work or in the streets whereas interactions outside work and school are limited (e.g., Avenarius, 2014; Burgers & Zuijderwijk, 2016; Ünlü Yücesoy, 2006).

On the other hand, despite residential segregation of immigrant groups (e.g., Blokland et al., 2016; Bozok & Bozok, 2019; Castenada, 2018; Huttman et al., 1991), immigrants pursue homemaking and reshape urban spaces (e.g., Boccagni & Brighenti, 2017; Kalandides & Vaiou, 2012). Hondagneu-Sotelo (2015) argues that when focused on the daily life activities of immigrants, one can gain an insight into the ‘right to homemaking’ of immigrants (resembling the term ‘right to the city’ by Harvey, 2003). Homemaking is a fundamental human right according to the Human Rights Declaration, even though not addressed by the same term. In that sense, a good number of articles directly address the right to homemaking, the right to the city, and the right to a decent social life. These include life, liberty, and security (Article 3), freedom of movement and residence (Article 13), social security (Article 22), work (Article 23), rest and leisure (Article 24), health and well-being (Article 25), education (Article 26), participation in the cultural life of the community (Article 27). Research findings show that immigrants around the globe face, although to varying extend, difficulties in fulfilling their basic need of homemaking.

In this light, we argue that assessing the spatial usage of immigrants would inform us whether they enjoy their right to homemaking, and to what extent they are included in the urban space. Inquiry in terms of spatial inclusion is mostly ignored in social sciences. Most research conducted on the issue of migration has primarily focused on people’s attitudes and behaviors regarding discrimination, social exclusion, and violence without considering the spatial aspect. Yet, revealing the level of using the right of homemaking per se shows the degree of inclusion of diverse social groups having unequal access to urban space (Boccagni, 2017; Moore, 2000).

Overview of Studies and Hypotheses

Homemaking and its relationship with spatial satisfaction and feelings have not received much attention in social sciences, although place-related issues (place attachment, place identity, urban identity, and social representations of places) have long been focused on (see Moore, 2000). In this research, we aim to explore the paradox that emerges when looking at the homemaking activities of immigrants (refugees in the current work) and non-immigrants (natives) simultaneously. To do that we ask research questions separately for each group of participants due to the status and motivational differences between them (refugees vs. natives): On the one side, natives perceive threat from refugees and are prone to seclude their homes by excluding newcomers from urban spaces (Study 1). On the other side, refugees are striving to make a new home (seeking to regain homely feelings such as security, comfort, and access to amenities) and homemaking (Study 2). Therefore, we formulated our hypotheses specific to the populations as below.

We argue that natives will perceive threat from refugees and, therefore will not be willing refugees to be included in society. Refuge of people elicits different types of threats among locals (e.g., Abeywickrama et al., 2018; Landmann et al., 2019; Tartakovsky & Walsh, 2016; see also Stephan et al., 1999). A recent qualitative study revealed that mostly cultural and religious differences and similarities between refugees and natives were narrated in Mardin (the current research is conducted in the same city) as well as other social comparisons between “us” versus “they” were made (Gezici Yalçın et al., 2022). In addition, locals did not express only burdens of refugees on the economic system, but also their contributions to the economy (see also Nshom et al., 2022 for both perceived threat and benefit). Therefore, we only focused on perceived symbolic threat in the current research which is a fear that arises from perception of differences between values, cultures, language, religion, morals, worldview, and way of life (Stephan & Stephan, 1996). The fear is that the ingroup’s meaning system could be negatively impacted or destroyed by the outgroup (Stephan et al., 2009). Thus, we hypothesize that the higher the symbolic threat perceived from refugees the lesser inclusion of this group in society is desired (Study 1, H1).

Yet, when refugees’ social reputation is positively evaluated, we expect that the willingness to include them in society will increase (Study 1, H2). At the bottom of our argument lies the fact that social reputation plays an important role in how people are perceived and treated in the Mediterranean context due to honor logic (e.g., Kirchner-Häusler et al., 2022; Uskul et al., 2023). Following this reasoning, we assume that the link between perceived symbolic threat and willingness for inclusion of refugees will be mediated by the ascribed social reputation of refugees (Study 1, H3). We also assume that the homemaking of natives plays a significant role in the willingness to include refugees by moderating the effects of perceived threat from refugees (Study 1, H4), and social reputation ascribed to refugees on willingness to include refugees (Study 1, H5). Consequently, we suggest a conceptual model as presented in Fig. 2 to be tested in Study 1.

Fig. 2
figure 2

A Conceptual Model for Natives (Study 1)

In the second study, our hypotheses are different (see Fig. 3): First, we expect that perceived inclusion by refugees will play a significant role in spatial satisfaction. Therefore, we hypothesize that the more inclusion is perceived by refugees the more spatial satisfaction will be reported (Study 2, H1). Furthermore, we will explore the space (dwelling, neighborhood, city center, none of them, or all of them) where refugees mostly feel sheltered, secure, comfortable, alien, and lonely (i.e., homely feelings). In addition, the relationship between the experiential element of space (perceived inclusion) and (spatial) satisfaction (see Moore, 2000) will be investigated. Previous research suggests that using urban public space and interacting with others within space has several positive outcomes (e.g., Cattell et al., 2008; see also Gezici Yalçın et al., 2022; Gezici Yalçın et al., 2023), yet the relationship has not been explored in terms of usage of urban space. We hypothesize that there is a significant relationship between spatial satisfaction and homely feelings (Study 2, H2). However, the direction of the interrelationship will be different for positive versus negative feelings. Put straightforward, we hypothesize a positive association between spatial satisfaction and feelings of comfort (Study 2, H2a), security (Study 2, H2b), and sheltered (Study 2, H2c), but a negative association of spatial satisfaction and feelings of alienated (Study 2, H2d) and loneliness (Study 2, H2e). Furthermore, we expect that spatial satisfaction will mediate the link between perceived inclusion and homely feelings (Study 2, H3). On the other hand, homemaking will moderate the interrelationship between perceived inclusion and homely feelings (Study 2, H4) as well as the interrelationship between spatial satisfaction and homely feelings (Study 2, H5).

Fig. 3
figure 3

A Conceptual Model for Refugees (Study 2)

Research Context

This research was conducted in a city in Turkey, a country at the crossroads of continents that its in-between geolocation makes it difficult to classify under one region. In some sources, Turkey is seen as a part of the Middle East whereas in others it has been seen as a part of Europe. In any way, it is not considered to belong to the Global North and is not counted among WEIRD (Western, educated, industrialized, rich, democratic; Henrich et al., 2010) societies. Recently, it has been categorized as part of the Mediterranean region which is a context neither Eastern nor Western regarding cultural comparison of self-construal (see Uskul et al., 2023). Doing research in this region provides us an opportunity to expand the Western perspective, or perhaps challenge it since it dominates the production of scientific knowledge in the Global North.

Furthermore, we conducted our research in a city that has distinguishing characteristics due to its historicity and location (see Fig. 4). Mainly, this is a city very close to the Syrian border in Turkey and has hosted diverse ethnonational (Kurds, Arabs, Turks), and religious (Muslim, Christian) groups for centuries. After the Syrian proxy war, the city hosted refugees who belonged to the same ethnocultural group, but this time they encountered as immigrants and non-immigrants in the same urban space. In a nutshell, Kurds and Arabs living in Mardin (Turkey) met with immigrant Kurds and Arabs from Syria after several years of cohabitation in the same villages and towns until the border split in 1928. Due to the ethnocultural ties between immigrants (i.e., refugees) and non-immigrants (i.e., natives), we expect more inclusion of newcomers in this city.

Fig. 4
figure 4

Research Sites in Mardin (Kızıltepe and Artuklu) in Turkey

Study 1

Based on the conceptual model we suggested for natives (see Fig. 2), we hypothesize that the symbolic threat perceived from refugees (H1), and social reputation (H2) ascribed to refugees significantly predict the willingness for inclusion of refugees. Moreover, ascribed social reputation to refugees mediates the interrelationship between perceived symbolic threat and willingness for inclusion of refugees (H3) whereas homemaking moderates this relationship (H4). Moreover, homemaking moderates also the relationship between ascribed social reputation to refugees and the willingness for inclusion of refugees (H5).

Methods

Participants and Procedure

Out of 493 locals who were reached, data from participants whose parents were born in other cities (n = 43) or whose mother tongue was Turkish (n = 29) were excluded to eliminate representing the nonlocal population in the sample. Thus, responses of 421 participants (aged 18 to 85; Mage = 37.1, SD = 14.3; 50% women) who reported that their ancestors were locals of the city remained in the data set. According to self-reports, more than half of the participants (58.4%, n = 246) speak Kurdish, less than a quarter of the participants (21.9%, n = 92) speak Arabic, and 19.7% (n = 83) speak both Kurdish and Arabic at home. Participants were reached out from two districts: 44.4% of all participants (n = 185) live in Kızıltepe (a recently developed town where mainly Kurdish villagers were settled), whereas 55.6% of participants (n = 232) live in Artuklu (the historical old town mainly Christian and Muslim Arabic communities reside).

The sampling technique involved quota determination based on a random selection of neighborhoods, followed by participant limitation in each street (n = 2) and neighborhood (n = 50) in two districts largely populated by refugees. Twenty interviewers (10 women) who are fluent in Kurdish, Arabic, and Turkish participated in data collection after receiving preparatory training about the study by a field coordinator. All interviews were made by teams consisting of a man and a woman team member speaking Turkish because, although our sample consisted of Kurds and Arabs, they are bilingual and speak Turkish competently. Interviewers read questions to participants and recorded their responses to eliminate the confound of nonexperience in filling out the survey data. The data were collected roughly in a month (from May to June) in 2017.

Research Instruments

Willingness for Inclusion of Refugees

A meta-analysis showed that cross-group friendships especially among school children who are provided with equal status reduce negative outgroup bias (e.g., Pettigrew & Tropp, 2006). A similar approach has also found a place in the 2023 UNESCO Report on refugee inclusion (Sunny & Borkowski, 2023). Informed by these sources, it appeared that we have two items at our disposal to assess participants’ attitudes towards the inclusion of refugee children: “Children of the Syrian refugees should have equal conditions with Turkish children”, and “Being friends with Syrian children is a chance for the Turkish children to learn living with people from different cultures”. Both items had a 5-point Likert-type scale (1 = “I totally disagree”, 5 = “I totally agree”; r = .29, p < .01).

Homemaking

Participants were asked to respond to the question obtaining the duration of time spent on their daily life activities (work, education, health, sport, entertainment, cultural and artistic activities, shopping, taking a trip, visiting relatives and friends) via scalar of space (1 = “my house”, 2 = “my neighborhood”, 3 = “city center”, and 4 = “other”).

Perceived Symbolic Threat from Refugees

The original scale with a two-factorial structure (realistic versus symbolic threats) was developed by Stephan et al. (1999). However, the internal consistency of perceived symbolic threat was found to be relatively low in this original study. The scale was first adapted to Turkish in 2013 (Balaban, 2013), yet the two-factorial (realistic and symbolic threats) structure was not obtained. Some of the symbolic threat items loaded on the factor of realistic threat. Later, the 3-item scale was used in many studies in the Turkish context with the conclusion that three items tap symbolic threat most distinctively (e.g., Demirdağ, 2021; Yanbolluoğlu, 2018). Although item formulations differed, these were related to ways of living, cultural values, customs, religion, and family relations. We, therefore, used three items via 5-point Likert-type scales (1 = “I totally disagree”, 5 = “I totally agree”. Due to a low inter-item correlation (.14), we dropped one item from the scale. The remaining two items (“The religious and moral values of Syrian refugees do not comply with that of Turks”, and “The culture of Syrian refugees is a threat to the Turkish culture”) were aggregated to assess perceived symbolic threat from refugees (r = .32, p <. 01).

Ascribed Social Reputation to Refugees

To assess participants’ ascription of characteristics to refugees, eight positive attributes (e.g., honorable, respectful, reliable, hard-working) were used (1 = “I totally disagree”, 5 = “I totally agree”; α = .79). The scale was developed based on previous studies which show that moral behaviors and respect (social and self) are central in Turkish culture (e.g., Uskul & Cross, 2019). In addition, the in-group social image measure developed by Rodriguez-Mosquera et al. (2017) was adapted in a way to assess out-group social image based on the most prevalent stereotypes about Syrian refugees in Turkish society (Gezici Yalçın et al., 2022).

Demographics

The age, gender, mother tongue, marital status of participants, and district they live in were asked.

Results

Descriptive statistics and bivariate correlation coefficients can be seen in Table 1.

Table 1 Descriptive Statistics and Bivariate Correlations for Study 1

To test whether gender (men or women) and district (Artuklu or Kızıltepe) have a significant effect on research variables two-way MANOVA was run. Since Levene’s test of equality of error variances showed that variances were not homogenous for homemaking (p = .000), this variable was excluded from MANOVA. This variance between the two districts in terms of homemaking might be due to differences in the distribution of resources and infrastructure facilities in urban space.

In the second run, the rest of the research variables (perceived symbolic threat from refugees, ascribed social reputation to refugees, willingness for inclusion of refugees) were included in the two-way MANOVA. According to the results, the main effects of gender and district were significant on different outcome variables, but no interaction between gender and district was found on any outcome variable. It was found that gender had a significant effect on ascribing social reputation to refugees [F(1, 413) = 11.30, p = .000, η2 = .04, multivariate F(3, 411) = 6.00, p = .000, Wilk’s Λ = .95, η2 = .05]. Women (n = 208, M = 2.76, SD = .73) ascribed more social reputation to refugees compared to men (n = 209; M = 2.43, SD = .79). On the other hand, the district had a significant effect on the perceived symbolic threat from refugees [F(1, 413) = 6.07, p = .048, η2 = .01, multivariate F(3, 411) = 3.18, p = .024, Wilk’s Λ = .97, η2 = .02]. According to pairwise comparisons, participants who lived in Kızıltepe (n = 232; M = 3.46, SD = 1.25) reported significantly more symbolic threat than participants in Artuklu (n = 185; M = 3.22, SD = 1.24).

To test the interrelations between research variables, more specifically, whether the perceived symbolic threat from refugees (H1) and ascribed social reputation to refugees (H2) predict the willingness for inclusion of refugees, we ran a linear regression analysis. Results showed that both variables predicted willingness for inclusion significantly [F(2, 418) = 33.55, p = .000]. More specifically, perceived threat negatively (ß = -.12, t = -2.40, p = .017), but social reputation positively (ß = .32, t = 6.52, p = .000) predicted willingness for inclusion.

Mediation Analysis

We tested whether social reputation to refugees plays a mediation role in the link between the perceived threat from and the willingness for inclusion (H3). We ran a mediation analysis (see Hayes, 2013) with 1000 Bootstrap samples. Results showed that the effect of perceived threat on the willingness for inclusion was significantly mediated by the social reputation (see Table 2). That is, indirect effect (a x b) was significant (E = -.11, SE = .02, [-.16, -.07], Z = -4.96, p < .001). All path coefficients were significant as well. That is the path from perceived threat to social reputation (a = -.21, SE = .03, [-.27, -.16], Z = 7.79, p < .001), from social reputation to willingness for inclusion (b = .52, SE = .08, [.37, .69], Z = 6.40, p < .001), from perceived threat to willingness for inclusion (c = -.12, SE = .05, [-.22, -.01], Z = 2.22, p = .026) were all significant.

Table 2 Mediation Estimates for The Mediator of Social Reputation

Moderation Analysis

To test the interaction between the perceived threat and homemaking on willingness for inclusion (H4), a moderation analysis was run (see Hayes, 2013). Results showed that the interaction between these two variables was not significant (p = .14). Similarly, the moderation effect of homemaking on the interrelationship between social reputation and willingness for inclusion (H4) was not significant (p =.12).

Discussion

The study conducted with natives of the city showed the importance of perceived symbolic threat from refugees playing an important role in willingness to include them in society as reflected by supporting equal rights among children and encouraging inter-group friendship (H1 was confirmed). This finding is in line with literature showing the negative link between perceived threat and intergroup hostile behaviors (e.g., Abeywickrama et al., 2018; Landmann et al., 2019; Tartakovsky & Walsh, 2016; see also Stephan et al., 1999). That is, the lesser the extent that threat is perceived the higher the willingness to include refugees. Moreover, the social reputation ascribed to refugees predicts the willingness to include them in society (H2 was confirmed). This finding also concurs with the research line showing the role of social reputation in attitudes and behaviors (e.g., Cross et al., 2014; Rodriguez-Mosquera, 2016). Moreover, our research contributes to the literature by showing the mediation effect of social reputation on the link between perceived symbolic threat and willingness for inclusion (H3 was confirmed). Unfortunately, the moderating effect of homemaking was not found (H4 and H5 were not confirmed). As can be seen from Table 1, the homemaking of natives was not found to be related to research variables but only to demographic variables (age and gender). We conclude that further research is needed to better understand the relationship between the homemaking of natives and their willingness to include refugees.

Study 2

To test the interrelationships between perceived inclusion, homely feelings, spatial satisfaction, and homemaking we developed a conceptual model for refugees as presented above (see Fig. 3). According to this model, we hypothesize that perceived inclusion (H1), and spatial satisfaction (H2) significantly predict homely feelings (feeling comfortable, secure, alienated, sheltered, lonely). Moreover, spatial satisfaction mediates the interrelationship between perceived inclusion and homely feelings (H3), whereas homemaking moderates this relationship (H4), and the relationship between spatial satisfaction and homely feelings (H5), too.

Methods

Participants and Procedure

In this survey, internationally displaced people from Syria living in the city of Mardin were reached as a part of a larger study designed to examine additional research questions concerning group differences in acculturation. A total of 913 internationally displaced people aged between 18 and 80 years old were reached. However, based on missing information in basic demographics yielded 889 displaced participants (M = 36.59; SD = 12.51; 62% women). Among them, participants reported that they speak either Kurdish (545 participants; M = 36.41; SD = 12.28; 69% women), or Arabic (344 participants; M = 36.87; SD = 12.87; 56% women). Kurdish-speaking participants migrated mostly to the city district called Kızıltepe (78%) where most of the old established inhabitants speak Kurdish in this community, too. On the other hand, most of the Arabic-speaking participants migrated to the city center called Artuklu (72%) where most of the old established inhabitants speak Arabic as well. Most of the participants reported being married (91% of Kurdish-speaking participants and 88% of Arabic-speaking participants).

The same sampling technique used in Study 1 was applied. In addition, while refugees do not speak Turkish, the questionnaires were translated into Kurdish and Arabic by bilingual team members via the committee method. All interviews were made by teams consisting of a man and a woman who were fluent in Kurdish, Arabic, and Turkish. Thus, the language barrier was surmounted by the bilingual researcher teams. All teams were designed in this way to overcome the difficulties due to gender-specific complexities (refugee women avoided contacting a male interviewer). Survey participants were interviewed in places they consider as safe, mostly at their homes. The data were collected in June 2017.

Research Instruments

Perceived Inclusion from Natives

Based on the previous work (Gezici Yalçın et al., 2023), we asked participants about their perception of how they are treated during their social interactions with natives. Thus, participants’ perception of inclusion was assessed based on their experiences (“How people treat you when you socialize in Mardin?”, and “How your colleagues treat you at work?”). They were provided options to respond as “No interaction” (0), “They verbally express that we are unwanted here” (1), “They keep starring us” (2), “Sometimes good, sometimes bad” (3), “They treat us as the way they treat others” (4), “They welcome us” (5), “Other” (99). These two items were aggregated, and the total score (Min. = 2, Max. = 9) was included in further analyses. The higher score represents the higher level of perceived inclusion from natives.

Homemaking

Like Study 1, participants rated their daily life activities via scalar of space (1 = “my house”, 2 = “my neighborhood”, 3 = “city center”, and 4 = “all”). That is, ten spatial practices were assessed in a way that higher value represents access to the city center.

Spatial Satisfaction

This scale was adapted from the original Satisfaction with Life Scale (Diener et al., 1985) to living in an urban space. Thus, five items measuring participants’ satisfaction with life on the streets, in the neighborhood, and the city on a 5-point Likert-type scale (1 = “not satisfied at all”, 5 = “very satisfied”; α = .91) were used.

Homely Feelings

Relying on the Feeling-at-Home Scale (Gezici Yalçın et al., 2023), we assessed participants’ homely feelings via space. Put straightforwardly, participants were asked to rate their feelings regarding spatial radius: “The place you feel most comfortable/secure/alien/sheltering/lonely.” Participants were provided options to select as “none” (0) “at my home” (1), “in my neighborhood” (2), “in the city center” (3), and “all” (4). Since items were formulated in a way to assess their homely feelings in terms of space, we used each item separately in the analyses to see their unique effect and direction.

Demographics

Age, gender, mother tongue, marriage status, and also the district they live in were asked.

Results

First, descriptive statistics for variables (minimum and maximum values, means, and standard deviation) were checked (see Table 3). Since homely feelings (comfortable, secure, alienated, sheltered, lonely) were assessed through the places it is felt mostly, frequencies were checked for these variables as presented in Table 3. Second, to see whether gender (men vs. women) and district (Kızıltepe vs. Artuklu) have a significant effect on research variables (perceived inclusion, spatial satisfaction, and homemaking), a two-way MANOVA was run. No significant effect was found; hence, they were not included in further analyses.

Table 3 Descriptive Statistics, and Bivariate Correlations for Study 2

When homely feelings were examined, it was seen that there were significant differences in terms of reported feelings. Participants reported that they feel themselves mostly comfortable (86.1%) and secure (84.6%) in the dwelling, but unexpectedly and contrary to feeling comfortable and secure they also reported in greater numbers that they feel themselves alien at home (70.5%). On the other hand, in terms of feeling sheltered and lonely, the answers were more diverse: 39.5% of all participants reported that they do not feel sheltered at all, whereas 36.8% of participants feel sheltered in the city center and only 14.6% feel sheltered at the dwelling. Regarding feeling lonely, 38.5% of participants reported that they do not feel lonely at all, whereas 32.4% feel lonely in the city center, and 15.6% of them feel lonely in all places.

When we ran regression analysis, we found that perceived inclusion [ß = -.18, t = -4.73, p = .000, F(2, 797) = 20.06, p = .000] and spatial satisfaction [ß = -.11, t = -3.01, p = .008, F(2, 797) = 20.07, p = .000, R2 change = .22] significantly predicted feeling of sheltered. Similarly, perceived inclusion [ß = -.17, t = -4.73, p = .000, F(2, 797) = 23.47, p = .000] and spatial satisfaction [ß = -.13, t = -3.74, p = .000, F(2, 797) = 23.48, p = .000, R2 change = .24] significantly predicted feeling lonely. However, feeling secure [ß = .07, t = 1.50, p = .05, F(2, 797) = 3.07, p = .05], and feeling comfortable [ß = .12, t = 3.33, p = .001, F(2, 797) = 7.70, p = .000] were significantly predicted only by spatial satisfaction.

Mediation Analyses

We ran a mediation analysis separate for each homely feeling (comfortable, secure, alienated, sheltered, lonely) with 1000 Bootstrap samples. Results showed that only two mediation models were significant. That is, spatial satisfaction mediated the link from perceived inclusion to feeling sheltered as well as to feeling lonely. The results of mediation analyses were presented separately in the following. First, according to the results, the effect of perceived inclusion on feeling sheltered was significantly mediated by spatial satisfaction (see Table 4). That is indirect effect (a x b) was significant, E = -.02, SE = .01, [-.03, -.01], Z = -2.81, p = .005. All path coefficients were significant as well: path from perceived inclusion to spatial satisfaction (a = .13, SE = .02, [.08, .17], Z = 5.72, p < .001), from spatial satisfaction to feeling sheltered (b = -.17, SE = .05, [-.28, -.05], Z = -3.05, p = .002), and perceived inclusion to feeling sheltered (c = -.15, SE = .03, [-.22, -.09], Z = -4.66, p < .001).

Table 4 Mediation Estimates on Feeling Sheltered

The second mediation analysis with 1000 Bootstrap samples for feeling lonely revealed that the effect of perceived inclusion on feeling lonely was significantly mediated by spatial satisfaction (see Table 5). Indirect effect (E = -.03, SE = .01, [-.05, -.01], Z = -3.03, p = .002), and all path coefficients were significant: the path from perceived inclusion to spatial satisfaction (a =. 13, SE = .02, [.08, .17], Z = 5.70, p < .001), from spatial satisfaction to feeling lonely (b = -.22, SE = .06, [-.35, -.10], Z = -3.47, p < .001), and perceived inclusion to feeling lonely (c = -.16, SE = .03, [-.23, -.09], Z = -4.70, p < .001).

Table 5 Mediation Estimates on Feeling Lonely

To check alternative mediation models, we repeated analyses where homely feelings were included in the model as mediators and spatial satisfaction as an outcome variable. However, indirect effects were decreased from 12.7% to 8.87% for feeling sheltered, and from 15.3% to 11.3% for feeling lonely. These results indicate that the model where spatial satisfaction is a mediator and homely feelings are outcome variables explains more variance.

Moderation Analysis

To test the interaction between perceived inclusion and homemaking on homely feelings, separate moderation analyses were run with 1000 Bootstrap samples. Results showed that the interaction between spatial satisfaction and homemaking is significant on feeling sheltered (E = -.05, SE = .01, [-.07, -.02], Z = -4.13, p < .001) as well as on feeling lonely (E = -.04, SE = .01, [-.07, -.02], Z = -3.54, p < .001). According to simple slope analyses, the effect of a high level of homemaking (+1 SD) is significant on both feeling sheltered (E = -.47, SE = .08, [-.63, -.31], Z = -5.77, p < .001), and feeling lonely (E = -.57, SE = .08, [-.74, -.40], Z = -6.53, p <.001). Simple slopes can be seen in Fig. 5.

Fig. 5
figure 5

Simple Slope Analyses for Moderation Model (Study 2)

Discussion

Our hypotheses about the significant association between perceived inclusion and homely feelings, as well as between spatial satisfaction and homely feelings were partially confirmed. More precisely, perceived inclusion did not predict feeling comfortable (H1a was not confirmed), secure (H1b was not confirmed), alienated (H1c was not confirmed), but significantly predicted feeling sheltered (H1d was confirmed), and lonely (H1e was confirmed). On the other side, spatial satisfaction significantly predicted homely feelings (H2a, H2b, H2d, and H2e were confirmed) except for feeling alienated (H2c was not confirmed). Furthermore, the mediation effect of spatial satisfaction on the link between perceived inclusion and homely feelings was obtained only for feeling sheltered and lonely (H3d and H3e were confirmed). On the other hand, homemaking did not moderate the link between perceived inclusion and homely feelings (H4 was not confirmed), but the moderation effect of homemaking was found on the link between spatial satisfaction and homely feelings yet only for feeling sheltered and feeling lonely (H5d and H5e were confirmed). In short, our assumptions were supported only by two homely feelings (sheltered and lonely), where spatial satisfaction mediated whereas homemaking moderated the link between perceived inclusion and feeling sheltered and lonely.

These findings imply that contrary to the anonymity in the city center, refugees are more familiar with their physical and social environment, have more agency and control, and initiate more close relationships in the neighborhoods (see e.g., Bozok & Bozok, 2019; Burgers & Zuijderwijk, 2016; Kalandides & Vaiou, 2012). Considering the self-reports of participants that they spend most of their lives (on many activities such as shopping, entertainment, relative visits, or cultural activities), neighborhoods are the main hubs of encounters between natives and refugees. Thus, it can be argued that those are the primary places where inclusion or exclusion is experienced (see e.g., Bilecen, 2017; Blokland et al., 2016); Castenada, 2018).

In short, our research conducted with refugees showed that there are significant differences in terms of reporting homely feelings. Most refugees feel comfortable and secure but simultaneously alienated at home. More than a quarter of the participants reported that they feel sheltered in the city center, and they also feel lonely there. These findings show the importance of place in terms of homely feelings (see Gezici Yalçın et al., 2023). It should be examined further to reveal its relations with encountering and interacting with others (ingroup vs. outgroup members). In addition, other triggering and inhibitory factors on the homely feelings and perceived inclusion are worth inquiring further. One related psychological variable might be perceived respect and dignity as inclusion is inferred from being treated with respect and dignity (e.g., Gezici Yalçın et al., 2023; Otaye-Ebede & Akobo, 2020).

General Discussion

This research is conducted to reveal a paradox of inclusion by including space as an assessable variable in the research. The homemaking approach provides a framework to examine the contrasting motivations between two sides of inclusion. On the one hand, natives are motivated to preserve their homes and daily life as it is (Gezici Yalçın et al., 2022). On the other hand, immigrants are motivated to make new homes and regain the sense of feeling at home (Gezici Yalçın et al., 2023). In two studies, we empirically showed that natives’ perception of threat and their ascription of social reputation to refugees are strong predictors of their willingness to include refugees in society, whereas interaction with natives without exclusion drives refugees to feel more sheltered and less lonely in the urban space (see Cattell et al., 2008; Hampton et al., 2011).

Thus, we conclude that the contradiction of inclusion can be best observed by looking at the inclusion of social groups in the urban space: Where they go to socialize with friends and relatives, in which places they entertain or pursue their cultural activities, and so on, in other terms cohabitation or glocalization (Bauman, 2013, 2018). The inclusion of immigrants into urban public space (see also Göregenli & Karakuş, 2014 for “integration of immigrants into space”) is not independent of the agreement or the willingness of natives (Gezici Yalçın et al., 2022). This could be regarded as an achievement by immigrants that is not a routine result ensuing the migration process automatically (see Boccagni, 2017; Boccagni & Brighenti, 2017). Instead, immigrants have to strive to reach through interaction and negotiation with natives (see Kaya, 2017; Kılıçarslan, 2016). Although homemaking is a ubiquitous human activity and thus a hard-to-deny human right, its importance can be easily ignored in the context of the movement of people (displacement and replacement) as if refugees are seen as disclaiming their right to homemaking (see Beeckmans et al., 2022; Bonfanti et al., 2022).

From the perspective of the current study, this paradox is more visible when natives encounter with refugees whom they share a comparable ethnocultural background including language and many habits of daily life (see Gezici Yalçın et al., 2022). Even then, a difference of power surfaces as the homemaking activities of refugees are affected by the agreement or willingness of the natives. On one hand, natives maintain a positive attitude towards refugees when it comes to the well-being of children and opportunities for friendship (see Cattell et al., 2008). Obviously, these two angles reflect the willingness of natives to recognize refugees’ presence in areas perceived as less threatening, possibly because of resemblances between the two groups (see Gezici Yalçın et al., 2022). On the other hand, the willingness of natives is inconclusive in areas where the activities of refugees are more concentrated on making a new home in the town where they are located. In other words, when we consider that homemaking extends beyond dwelling and includes using urban space (e.g., Boccagni, 2022b; Boccagni & Brighenti, 2017), natives are less anxious than the more settled existence of refugees due to making actual dwellings in the sense of feeling at home (see Gezici Yalçın et al., 2023).

Correspondingly, refugees, too, feel homely in urban spaces albeit accompanied by a sense of being lonely and alienated. This aspect is further confirmed by the feelings of insecurity by refugees in their dwellings. Even though refugees are more comfortable pursuing their daily activities in shared public places, that feeling does not yet extend to their dwellings.

Considering the fact that the two groups focused on in this study share a comparable ethnocultural background, we may conclude that familiarity helps both natives and refugees feel at home, in their encounter, as long as their interaction takes place in the urban space through exchange between children under equal conditions and friendship. However, anxiety develops on both sides when it is a matter of enduring presence, thus a requirement to make changes in the manner of relationships in the long run, as refugees rightfully strive to make their homes and wish to achieve a homely feeling in their new area of settlement (see Gezici Yalçın et al., 2022).

The current study points out the need for further research on the encounter of natives and refugees (or, more generally, immigrants) especially when the two groups share comparable characteristics. Seemingly, there are some facilitations for the willingness of natives to receive refugees of familiar backgrounds. Yet, the perceived threat is still there albeit for different reasons than the case the two groups in question do not have an ostensible common ground. New research should explore the variables working for and against the willingness of natives to understand better the situational and intrinsic factors that contribute to any inclusion paradox. A final word is necessary to indicate the limitations of this study. First, our study focuses on two groups sharing a comparable background. On one hand, this focus makes the distinctive aspect of this investigation as we attempted to show the facilitations and limitations of feeling at home when such groups encounter due to forced migration. On the other hand, the results of our study are not generalizable to all types of encounters, especially in the cases where groups do not share certain common attributes. Having said that, we wish our study to encourage further research comparing encounters of similar versus dissimilar background people in the context of migration.

Secondly, our study was conducted in two towns of a border province in Turkey where there were already native people coexisting for a long time despite differences in language, ethnicity, culture, and or religion. The outlook could be different from that has been shown in our study in places where natives manifest a more homogeneous composition even when they encounter immigrants or refugees of similar backgrounds.

Last but not least, intergroup encounters, interactions, exchange, and for that matter inclusion, could manifest themselves quite differently, even idiosyncratically, in metropolitan areas than cities or towns. Natives’ attitudes could likely differ in towns neighboring the country than in big cities relatively far from the immediate effects of a proxy war.

Keeping these limitations in mind, we believe we are in the face of, metaphorically speaking, the intersectionality of encounters such that dynamics and parameters of inclusion need to be treated carefully by taking into consideration the context of migration as well as the area migrated.