The Effect of Length of Residence and Geographical Origin on the Social Inclusion of Immigrants

Introduction

Migration is regarded as one of the most stressful events that people can experience that requires them to make major adaptive responses (Cakir & Guneri, 2011). The migratory process entails many simultaneous changes, which potentially have long-lasting effects (Achotegui, 2009; Teixeira de Almeida & Vaz, 2011). Migration has become an increasingly common process, and therefore intercultural adaptation processes are of increasing relevance (Barker, 2015; United Nations, 2015).

Migration involves a breakdown of social networks, loss of roots, and many changes that can lead to decreased well-being or even the immigrant syndrome of chronic and multiple stress (Achotegui, 2009; Garcini et al., 2016; Guruge, Thomson, George, & Chaze, 2015). Acculturation stress refers to the consequences of the process of adaptation to a new culture and is associated with multiple stressors, such as culture shock, communication barriers, economic challenges, loss of support, and unemployment. These stressors promote the development of various psychosomatic symptoms and ultimately lead to an increased incidence of disease (Berry, 1997; Malgesini, 2002).

From the beginning of their experience, immigrants are involved in a multidimensional process of acculturation and sociocultural adaptation, which is influenced by elements such as the development of cultural competence and proficiency in the language of the host country, age at the time of migration, and access to resources considered relevant (Amit & Bar-Lev, 2015; Birman & Trickett, 2001; Prilleltensky, 2008; Wilson, Ward, Fetvadjiev, & Bethel, 2017). Typical variables have been used to predict immigrant adaptation, such as reason for immigration, job satisfaction, marital status, religious motivation, or the development of multicultural identities (Amit & Riss, 2014; Aycan & Berry, 1996; Birman, 2011; Gove, 2018; Salo & Birman, 2015). Acculturation is affected by personal and social variables present in the origin and destination societies, and by phenomena that existed before and during the acculturation process (Berry, 1997). Thus, this multidimensional process gradually strengthens the social inclusion of immigrants in the destination over time (Berry, Phinney, Sam, & Vedder, 2006).

Social inclusion is a multidimensional and multi-levelled concept in which two complementary approaches can be distinguished: one that focuses on rights and another that emphasizes the participation of an individual in the community (Davey & Gordon, 2017). The complex concept of social inclusion includes ideas such as “social inclusiveness, cultural cohesion, communal values, a shared identity, mutual recognition, respectful dialogue, peaceful interaction, or policies of integration” (Hedetoft, 2013, p. 1).

Although length of residence has a positive influence on the inclusion of immigrants (Uña, Clemente, Espinosa, & Fernández, 2009), there is no absolute consensus on the mechanisms of intercultural adaptation. Current studies show that time is associated with increased length of residence and with the age of immigrants (Michel, Titzmann, & Silbereisen, 2012).

The influence of time of residence on adaptation among immigrants is somewhat ambiguous. Thus, when immigrants fail to adapt to specific sociocultural changes or to achieve certain goals, perceived discrimination is marked, significant social ties remain undeveloped, and certain basic needs are not met. A failure to participate in the local community weakens their initial need to adapt to the new context (Portes & Rumbaut, 2014; Viruell-Fuentes, Morenoff, Williams, & House, 2013). The process of social inclusion and growth in well-being in the host culture is non-linear. This process is affected by multiple elements, such as very long transnational family separations, economic and labour precariousness, and ethnic network or family burdens, all of which can affect positive development over time (Kim & Noh, 2015; Waters, 2011).

It is essential to understand the process by which immigrants adapt to new settings to help them in the process of social inclusion into a new community. Understanding this process has implications for the development of policies that can facilitate the process (Casado, Hong, & Harrington, 2010).

Social Inclusion: The Effect of Length of Residence and Cultural Distance

Social inclusion in the host culture can be understood as a multidimensional process, in which various personal and social factors interact. This aspect indicates the relevance of social commitment in order to foster feelings of belonging and personal growth among immigrants (Kirpitchenko & Mansouri, 2014). Time of residence in a community acts as support for the social inclusion process by allowing immigrants to take advantage of the available opportunities for social and political participation, making it possible for them to experience greater attachment and community pride and to give meaning to their life experiences (Baker & Palmer, 2006).

A positive association has been found between the social inclusion of immigrants and length of residence in a given country and in a specific area (Kearns & Whitley, 2015). In contrast to the accepted view that social inclusion is a gradual linear process by which immigrants become naturally integrated in the host culture over time (Godenau, Rinken, Martínez, & Moreno, 2014; Uña et al., 2009), a new perspective suggests that there are many types of inclusion (Bürgelt, Morgan, & Pernice, 2008; Kwok-bun & Plüss, 2013; Tartakovsky, 2009). The well-accepted U-shaped curve is one of the alternatives to such linear models; however, there is insufficient evidence to consider it a good model of the adaptation process in immigrants (Michel et al., 2012). Despite time being a positive factor in the social inclusion of immigrants, the lack of linearity can be explained by a variety of circumstances that can decrease inclusion during specific periods. Because time has been generally regarded as the control variable, there has been little research on the lack of linearity in the social inclusion process of immigrants. More specifically, the influence of length of residence on the different indicators of social inclusion has been particularly neglected, because its effects have been assumed to remain constant and equal for all immigrant groups throughout the entire process. Few studies have attempted to deeply investigate the lack of linearity in complex migratory processes. Of these studies, the most relevant are based on synthetic cohort designs. This design provides a dynamic perspective by which to analyse the social inclusion of immigrants. It uses a large amount of cross-sectional data on participants characterized by the moment of entry into a social system (Martinovic, Van Tubergen, & Maas, 2009a). In synthetic cohort studies, participants are selected according to their exposure to a specific circumstance (e.g., the migratory event) and are compared over time in relation to phenomena of interest (e.g., social inclusion). The synthetic cohort design is well suited to study linearity over time – even with the use of cross-sectional data – because it is possible to establish a time sequence from the time of exposure to a specific circumstance to the appearance of a relevant phenomenon for a group of individuals with a common characteristic. Some examples of studies that have taken this approach include Viruell-Fuentes et al. (2013), who researched the construction of social bonds in different Latin American groups living in the USA; Myers & Lee (1998), who conducted a time analysis of the residential assimilation of immigrants in the USA by home ownership; and Van Tubergen & Kalmijn (2009), who investigated the development of language skills in immigrants during the first 20 years in the receiving country.

Another key aspect of the social inclusion process is the relationship of immigrants with their environment. People are strongly influenced by the setting in which they develop, which constitutes a space-time framework that is similar to an ecosystem. The neighbourhood, understood as a territorial area of influence, is a microsystem in which interactions between people create dynamic social networks of mutual support (Bronfenbrenner, 1979).

Cultural distance is a fundamental aspect in understanding the social inclusion process of immigrants, particularly in relation to determining how different immigrant groups use spaces of coexistence and how they experience the adaptation process over time (Genkova, Trickett, Birman, & Vinokurov 2014; Weng, 2016).

Immigration fosters a heterogeneity of values, cultural identities, and lifestyles, giving rise to communities in which different cultural groups coexist among which there is more or less cultural distance. The cultural distance term refers to interreligious, inter-linguistic, and inter-ethnic diversity within the host culture and is defined as the distance between cultures regarding their socio-cultural properties, including specific cultural features and the characteristics of the population in a given area. As the cultural gap widens, intercultural contact is predicted to become increasingly difficult (Babiker, Cox, & Miller, 1980; Geeraert & Demoulin, 2013). Cultural distance strongly influences the social inclusion of immigrants. Increased cultural distance between the sending and receiving countries makes learning the host culture more challenging, thus hindering the acculturation experience and the sociocultural inclusion of immigrants (Searle & Ward, 1990; Wilson et al., 2017). The place of origin has traditionally been considered a key aspect in the analysis of cultural distance; however, what really hinders the social inclusion of immigrants is not the place of origin itself, but the implications of greater cultural distance. Studies have shown that in the face of greater cultural distance, cultural diversity management strategies have to be developed, problems of coexistence increase, and social inclusion becomes more difficult (Oppedal, Røysamb, & Heyerdahl, 2005). Proficiency in the host culture language is a fundamental aspect that improves with practice and residence time (Wilson et al., 2017). In Spain, the immigrant group with the closest cultural proximity to the host country is the Latin American group. This proximity is based on the similarity of language, religious culture, and a shared history (Arenas & Urzúa, 2016; Briones, Verkuyten, Cosano, & Tabernero, 2012). Thus, one study has suggested that the adaptation process in Spain is better among Latin American and Eastern European immigrants than among African and Asian immigrants (Checa & Monserrat, 2015). Asian immigrants in Spain tend to experience more difficulties during the sociocultural inclusion process because, despite their societies being considered collectivist, their social interaction processes tend to be predominantly hierarchical (Shavitt et al., 2016).

There are a considerable number of variables (e.g., social support) that affect the social inclusion process and that are modulated by cultural distance from the receiving society (Martínez, García, & Maya, 2002). A smaller cultural distance between the immigrant group and the receiving society favours inclusion and increases satisfaction with life (SWL) (Jasinskaja-Lahti, Liebkind, & Perhoniemi, 2006). The available evidence shows that the dynamics of social inclusion processes in immigrants are associated with the contextual circumstances of the place of reception (Hernández, Pozo, Alonso, & Martos, 2005) and with the psychological characteristics of the different immigrant groups (Tartakovsky, 2009). To a large extent, the ethnic origin of immigrants explains differences in inclusion between different groups of immigrants in the same receiving society (Martinovic et al., 2009a). Some studies have analysed the effects of the migratory process on social support and health and have found differences between collectivist and individualist cultures (Shavitt et al., 2016). It has been suggested that social policies on immigration should be based on the recognition of cultural diversity (Castles & Miller, 2004), and that further research is needed on the specific characteristics of the different immigrant groups that share the same territory (Frank, Hou, & Schellenberg, 2016).

Resilience, Sense of Community, Satisfaction with Life, and Social Support as Indicators of Social Inclusion in the Migratory Process

The majority of authors have suggested that the social inclusion of immigrants should be approached from a multidimensional perspective in which individual and contextual characteristics interact (Davey & Gordon, 2017). The individual dimension includes personal characteristics such as resilience or life satisfaction. The migratory process is one of the most adverse situations that people can experience. Thus, it is relevant to study resilience in the immigrant population, because resilient individuals have a greater capacity for integration in the host community (Barba, 2012). The life satisfaction of immigrants is another dimension that is strongly related to social inclusion (Sand & Gruber, 2018). Within the contextual dimension, social support is considered to be an interactive concept. Social support is an interpersonal transaction between the source and recipient of support that can involve emotions, material help, and information within a specific context (Wills & Shinar, 2000). Social support between immigrants and indigenous people benefits the immigrant population, because social networks facilitate social inclusion and provide resources adapted to their needs. The contextual dimension of the sense of community is particularly relevant because it integrates individual and interactional aspects which, in turn, are influenced by the physical and socio-cultural context of the immigrants. A sense of community helps people identify with the setting in which they live and facilitates interaction and social inclusion (Bathum & Baumann, 2007).

The following results have been reported in the literature: social inclusion is positively associated with SWL in immigrants (Amit & Riss, 2014; Herrero, Gracia, Fuente, & Lila, 2012); social support is a predictor variable for the inclusion of immigrants (Berry, 1997); the development of a sense of community (SOC) is essential for their social integration (Halamová, Kanovský, & Naništová, 2018; Hombrados-Mendieta, Gómez-Jacinto, Domínguez-Fuentes, & García-Leiva, 2013); resilience in immigrants is a process of restoration of sociability that reflects inclusion in a new setting (Hosseini, 2016). An association has been found between perceived social support and psychological well-being in immigrants (Ward & Kennedy, 1992), and specifically between perceived social support and satisfaction with life (SWL) (Safi, 2010). Social support promotes resilience in immigrants (Kiang, Grzywacz, Marín, Arcury, & Quandt, 2010) and has been positively associated with the sense of community (SOC) in immigrants (Tang, Chi, & Dong, 2017). Positive associations have been found between resilience and SWL (De la Paz, Mercado, & Rodríguez, 2016) and between the SOC and SWL in immigrants (Moscato, Novara, Hombrados-Mendieta, Romano, & Lavanco, 2014).

The concept of resilience is dynamic and in constant development. It has been defined as “the human capacity to face, overcome and be strengthened by or even transformed by the adversities of life” (Grotberg, 2006, p. 20). Contradictory results have been obtained regarding the association between resilience and length of residence of immigrants. Some studies have found a positive association between length of residence in a country and resilience in the immigrant population (Lee, Brown, Mitchell, & Schiraldi, 2008), whereas other studies have failed to find evidence of this association (Hosseini et al., 2017).

The concept of sense of community (SOC) refers to interdependence and perceived similarity with others (Sarason, 1974). McMillan and Chavis (1986) suggested that the sense of community is a multidimensional concept that consists of membership, influence, integration and fulfilment of needs, and shared emotional connection. The SOC is specific to the setting and its characteristics (Hill, 1996; Hombrados-Mendieta & López-Espigares, 2014; Long & Perkins, 2007). The majority of studies have found an association between an increased length of residence and an increased SOC, both in the autochthonous population (Prezza, Zampatti, Pacilli, & Paoliello, 2008) and in immigrants (Kearns & Whitley, 2015). However, it has been found that the SOC is not always associated with residential stability (Mak, Cheung, & Law, 2009).

Among other factors, subjective well-being is fundamental to social inclusion and is typically measured using SWL in the setting in which migrants reside (Sand & Gruber, 2018). Contradictory results have been obtained regarding the association between length of residence and SWL of immigrants. Some studies have found a positive association (Ullman & Tatar, 2001). However, other studies have found that the length of residence in a country is not associated with satisfaction with life among immigrants (Murillo & Molero, 2016), and others have suggested that their SWL decreases over time (Kim & Noh, 2015).

Social support has been associated with the length of residence of immigrants. In general, social support increases as the length of residence increases in the receiving country (Salinero-Fort et al., 2011). However, other studies have failed to find this association (Tonsing, 2013). Some researchers have suggested that the increase in social support is associated with the length of residence in a specific area or neighbourhood, rather than in the receiving country (Kearns & Whitley, 2015). Some studies have drawn attention to the lack of linearity in the association between the length of residence in a given country and the creation of social bonds (Viruell-Fuentes et al., 2013). It has been suggested that standard methodological approaches are inadequate to investigate the dynamic process of social support (Liu, Wu, & Chen, 2016), given that time is not a suitable control variable to study a potentially non-linear association (Wasserman et al., 2006). The various sources of support also play differential roles in the inclusion process. Fundamental sources include family (García-Ramírez, Martínez-García, & Albar-Marín, 2002), the support network of the autochthonous population (Domínguez-Fuentes & Hombrados-Mendieta, 2012), and links with other immigrants (Hernández et al., 2005).

The Present Study

This study used a synthetic cohort design to analyse the effects of length of residence and place of origin on the social inclusion of different migration cohorts into the metropolitan community of Málaga (Spain). Social support, resilience, SOC, and SWL have typically been studied separately. This study uses these variables as indicators of the social inclusion of immigrants. The aim was to investigate whether an increased length of residence promotes the social inclusion of immigrants as indicated by increased resilience, sense of community and satisfaction with life, and perceived social support. We hypothesised that place of origin would moderate the positive effect of length of residence on social inclusion, because the place of origin of a migration cohort and sociocultural characteristics of members of that cohort were expected to generate different patterns of social inclusion. Thus, we examined the inclusion trajectories of four socioculturally distinct migration cohorts: (a) Africans, (b) East Europeans, (c) Latin Americans, and (d) Asians. As noted earlier, the results reported in the literature remain contradictory, and the use of standard methodological approaches is considered insufficient to clarify this association. Figure 1 shows the conceptual model.

Figure 1

Conceptual Model.

Method

Participants and Procedure

The study participants were 2,374 immigrants residing in Málaga (Spain) (age range: 16-80 years, M = 32.44, SD = 11.45). In total, 48.3% were men and 51.7% were women. The places of origin of the members of these four migration cohorts were: Africa (n = 597), East Europe (n = 565), Latin America (n = 652), and Asia (n = 560). This ratio is representative of the distribution of immigrants in Málaga (Instituto Nacional de Estadística, 2018). Participants were selected based on the largest groups residing in Málaga by continent of origin. European participants were mainly from the Ukraine, Romania, Bulgaria, and Russia; all the African participants were from the Maghreb; Latin American participants were mainly from Paraguay, Argentina, Colombia, and Venezuela; and all Asian participants were from China.

These four categories are similar to those used in studies which analysed differences in social inclusion patterns between immigrant groups according to geographic region (Basabe & Bobowik, 2013; Berry et al., 2006; Checa & Monserrat, 2015; Zolkowska, Cantor-Graae, & McNeil, 2001).

All participants were volunteers and were selected according to quota sampling by sex/gender and place of migratory origin. They were contacted through associations and clubs, businesses, meeting places, and so on. The questionnaires administered to the non-Spanish-speaking participants were translated into their language of origin by native speakers who had a full command of Spanish.

Research Design

Few studies have analysed the social inclusion process among immigrants in different time periods and over the long term. However, the adaptation process among immigrants is a matter of years rather than months (Michel et al., 2012; Uña et al., 2009). The psychosocial adaptation of immigrants is strongly influenced by elements that are consolidated over the long term (e.g., mastery of the language, social support network, and sense of community) (Zhang & Goodson, 2011). During the first decade of initial resettlement, levels of depression, anxiety, and psychological vulnerability tend to be high. Thus, it takes more than a decade for immigrants who do not speak the language to adapt psychologically (Tran, Manalo, & Nguyen, 2007). Nevertheless, it has been found that 5 years after immigration, there is a favourable change in objective parameters of the absorption of immigrants, but no decrease in psychological distress (Lerner, Kertes, & Zilber, 2005). For this reason, studies on the social inclusion process have tended to compare immigrants who have been in the host country for around 5 years to those who have been there for longer periods (Viruell-Fuentes et al., 2013; Wilson et al., 2017). It is known that immigrants experience their first year in the host country as a period of shock and uncertainty in which they have strong difficulties in sociocultural and psychological adaptation. However, the inclusion of recent arrivals could introduce bias because of their tendency to return to their country of origin or to use countries as bridges between their place of origin and final destination due to the aforementioned initial difficulties and uncertainty (Chib, Wilkin, & Hua, 2013; Gouin, Zhou, & Fitzpatrick, 2015).

Berry et al., (2006) used temporal cohorts consisting of three categories: 0-6 years, 6-12 years, and 12-18 years. Similarly, studies on the risk of experiencing depression during the migratory process have used temporary cohorts of less than 7 years, 7 to 14 years, and more than 14 years to categorize residence time in the destination country (Hwang, Chun, Takeuchi, Myers, & Siddarth, 2005).

In general, the use of groups with fewer than 5 years or 5 to 10 years of residence in the host country is adequate to compare variations in social inclusion of immigrants over time (Martinovic, Van Tubergen, & Maas, 2009b; Martinovic, Van Tubergen, & Maas, 2015). In the present study, the synthetic cohort was constructed using a stratification period of 5 years. Specifically, the data were modelled in temporal terms to investigate their effects on the dependent variables. The main independent variable was residence time in Malaga, which was used to construct the synthetic cohort (< 5 years, from 5 to 10 years, and > 10 years), with the objective of analysing the influence of residence time on the social inclusion of immigrants (resilience, SOC, SWL, and social support). We also analysed whether place of origin (Africa, Eastern Europe, Latin America, or Asia) would moderate the positive effect of residence time on social inclusion (see Figure 1). We used the three temporal strata to analyse the linear and non-linear effects of each of the four places of origin on social inclusion. Thus, this synthetic cohort design provides a dynamic perspective and is able to separate the effects of time of residence in Malaga from the effects of place of origin on the social inclusion of immigrants (Viruell-Fuentes et al., 2013). It has been shown that synthetic cohort designs overcome the limitations of cross-sectional studies that typically use time as a control variable. Time as an independent variable can be used to analyse the influence of specific personal and social circumstances that can modulate the social inclusion of immigrants during the migration process. This strategy can help to analyse potential non-linear patterns of social inclusion in the different groups of immigrants, given that social inclusion can accelerate or decelerate at different times and may not be a totally linear and homogeneous process (Martinovic et al., 2009a).

Measures

Sociodemographic variables

Length of residence. Length of residence was the first independent variable used in the study. The participants were asked to provide information on their length of residence in Málaga. The average length of residence in Málaga was 8.66 years (SD = 6.37, range = 0-50 years). Immigrant populations are associated with high residential mobility. Thus, length of residence in Málaga was used instead of length of residence in a specific neighbourhood as a more suitable variable to assess the effect of territorial setting on the different indicators of social inclusion used in this study.

This variable was used to construct the synthetic cohort. In this study, exposure to the specific circumstance was length of residence in the receiving location from the time of arrival. The participants were compared over time to investigate its effect on the process of social inclusion, particularly in relation to its linear effect, thereby allowing us to refine our understanding of this phenomenon. Data were grouped into three periods (< 5 years, from 5 to 10 years, and > 10 years) based on cross-sectional data from a representative sample of immigrants living in Málaga. These three time cohorts reflect the evolution of migratory processes in Spain and patterns of development of linguistic competence in the host language (Michel et al., 2012).

Place of origin. Place of origin was the second independent variable used in the study. It was used to measure the cultural distance of immigrants, which is a key element in understanding social inclusion processes. Four categories were established according to place of origin: Africa (25.1%), East Europe (23.8%), Latin America (27.5%), and Asia (23.6%).

Dependent variables. Resilience, SOC, SWL, and social support were used as dependent variables and indicators of social inclusion.

Resilience. The 10-item version of the Connor-Davidson Resilience Scale (CD-RISC 10) (Davidson & Connor, 2018) was used to measure resilience. (e.g., “I am able to adapt when changes occur”). In the CD-RISC 10, resilience is considered to be a mediating element in recovery processes following various kinds of harm. It promotes better adaptation to life challenges and thus helps maintain confidence in the face of hostile circumstances. It also refers to having clear objectives and high perceived cognitive and emotional self-control. Resilience contributes to modulating negative emotions and acts as a protective factor against stressful stimuli. The scale used ranges between 1 and 5, where 1 equals never and 5 equals almost always. The scale has a Cronbach’s α = .89.

Sense of community. The Sense of Community Index (SCI-2) (Chavis, Lee, & Acosta, 2008) was used to investigate SOC, which was defined across four dimensions: membership, influence, integration and fulfilment of needs, and shared emotional connection (McMillan & Chavis, 1986):

• Membership is the feeling of belonging or of sharing a sense of personal relatedness (e.g., “Being a member of this community is a part of my identity.”). Cronbach’s α = .78.
• Influence is a sense of mattering, of making a difference to a group and of the group mattering to its members (e.g., “I have influence over what this community is like.”). Cronbach’s α = .80.
• Integration and fulfilment of needs is the feeling that members’ needs will be met by the resources received through their membership in the group (e.g., “I get important needs of mine met because I am part of this community.”). Cronbach’s α = .84.
• Shared emotional connection is the commitment and belief that members have shared and will share history, common places, time together, and similar experiences (e.g., “Members of this community care about each other.”). Cronbach’s α = .86.

The questionnaire consists of 24 items measured on a Likert-type scale, where 1 equals not at all, 2 equals somewhat, 3 equals mostly, and 4 equals completely.

Satisfaction with Life Scale. The Satisfaction with Life Scale (SWLS) (Pavot & Diener, 1993) was used to measure SWL. The scale assesses a person’s satisfaction with life as a whole, based on comparing perceived life circumstances to self-imposed standards. The scale focuses on the positive elements of an individual’s experiences. The questionnaire consists of 5 items (e.g., “In most ways my life is close to my ideal”), which are answered on 7-point Likert-type scale, where 1 equals completely unsatisfied and 7 equals completely satisfied. The scale has a Cronbach’s α = .90.

Social support. Social support was assessed using the Questionnaire on the Frequency of and Satisfaction with Social Support (García-Martín, Hombrados-Mendieta, & Gómez-Jacinto, 2016). This instrument measures the frequency of and satisfaction with the emotional, instrumental, and informational support received from different sources. It consists of 35 items (e.g., “Your family is loving and affectionate and listens to you when you want to talk and express your feelings”). Frequency of support is measured on a 5-item Likert-type scale, where 1 equals rarely and 5 equals always. Satisfaction with received support is also measured on a Likert-type scale, where 1 equals dissatisfied and 5 equals very satisfied. The following sources of support were analysed: family (Cronbach’s α = .92), Spanish friends (Cronbach’s α = .94), immigrant friends (Cronbach’s α = .95), neighbours (Cronbach’s α = .94), and community members (Cronbach’s α = .96) (i.e., members of associations, volunteer groups, religious organizations, etc.). Each of the sources received an index.

Statistical Analysis

All analyses were performed using the IBM SPSS Statistics 20 software package. Several ANOVAs were conducted using length of residence in the city (3 cohorts: < 5 years, 5-10 years, and >10 years) and place of origin (4 groups: Africa, East Europe, Latin America, and Asia) as independent variables, and using resilience, satisfaction with life, and each of the sources of social support as dependent variables. Regarding sources of social support, ANOVA was employed because different samples were used and a MANOVA would have limited the sample to the common cases. A MANOVA was conducted using length of residence (3 periods) and place of origin (4 groups) as independent variables, and using the four dimensions of SOC as dependent variables. In the ANOVAs and MANOVA, a p value of 1% was used as a cut-off for statistical significance and the Bonferroni test was applied to make multiple comparisons between the groups.

Results

Table 1 shows the descriptive statistics of resilience and satisfaction with life. Regarding resilience, statistically significant differences were found between the three periods of residence (F = 19.42, p < .001, η² = .017) and between the groups by place of origin (F = 10.74, p < .001, η² = .014). However, no significant interaction effect was found between the two variables (F = 1.54, p = .16, η² = .004). Having found differences between length of residence, the Bonferroni test was applied to make multiple comparisons between the three groups. It was found that resilience increased as a function of increased length of residence. Statistically significant differences were found between the first and third periods of residence (p < .001), and between the second and third periods of residence (p < .001), but not between the first and second periods of residence (p = .335). Having found differences between groups by place of origin, the Bonferroni test was used to conduct multiple comparisons of the four groups. Latin Americans were more resilient than Asians (p < .01) and Africans (p < .01). Regarding resilience, this result was the only one that reached statistical significance. No other statistically significant differences were found between the four groups.

Table 1

Descriptive Statistics of Resilience and Satisfaction with Life by Place of Origin and Length of Residence

Regarding SWL, univariate analysis showed statistically significant differences between the three periods of residence (F = 33.83, p < .001, η² = .029) and between groups by place of origin (F = 17.19, p < .001, η² = .022). In addition, a significant interaction effect was found between length of residence and place of origin (F = 3.34, p = .003, η² = .009). It was found that there was an increase in SWL as a function of increased length of residence. A statistically significant difference was found in SWL between the first and third periods (p < .001) and between the second and third periods (p < .001); however, a lower statistically significant difference was found between the first and second period (p < .125). The Bonferroni test showed that Latin Americans had the highest scores in SWL. The scores were statistically significantly different between this group and Africans (p < .001). East Europeans had the next highest scores, which were much higher than those of Africans (p < .001). However, no statistically significant difference was found between East Europeans and Asians (p = .29). Asians had significantly higher scores than those of Africans (p = .004).

The Bonferroni test showed an interaction effect between SWL and length of residence in the African, Latin American, and East European groups. The strength of the interaction effect was greatest in the African group followed by the Latin American and East European groups. No significant interaction effect was found in the Asian group.

The interaction between the two independent variables can be clearly observed in Figure 2: an increase in the length of residence increased the levels of SWL in Africans, East Europeans, and Latin Americans. Growth in SWL was slow among Latin Americans and particularly slow among East Europeans. Overall, growth was slow between 5 years and 10 years, but accelerated after 10 years. Although Africans began their period of residence with the lowest levels of SWL, growth in this variable was rapid and remained almost constant over time, and whereas there was a decrease in SWL among Asians over the medium term, there was an increase over the long term.

Figure 2

Interaction between the two Independent Variables.

Table 2 shows the descriptive statistics of the four dimensions of SOC. The multivariate analysis (Pillai trace) found statistically significant differences between the three periods of residence (F = 17, p < .001, η² = .029) and between the four origin groups (F = 2.53, p = .003, η² = .004); however, univariate analysis found no statistically significant differences in the components of the SOC between the four groups by place of origin. No significant interaction was found between length of residence and place of origin (F = 1.4, p = .091, η² = .004). A statistically significant effect was found between length of residence and all four dimensions of the SOC. It was found that there was an increase in all four dimensions as a function of increased length of residence. The Bonferroni test found statistically significant differences in fulfilment of needs between the first and third periods (p < .001) and between the second and third periods (p < .001); however, a lower statistically significant difference was found between the first and second period (p < .001). Statistically significant differences were found in membership between the three periods (p < .001). It was found that there was an increase in membership as a function of increased length of residence. A statistically significant difference in influence was only found between the first and third periods (p < .001). Finally, a statistically significant difference was found in emotional connection between the first and third periods (p < .001) and between the first and second periods (p < .001).

Table 2

Descriptive Statistics of the Four Components of Sense of Community by Place of Origin and Length of Residence

Table 3 presents the descriptive statistics of social support. Univariate analysis showed statistically significant differences in family social support between the three periods of residence (F = 10.75, p < .001, η² = .010) and between the groups by place of origin (F = 7.64, p < .001, η² = .010). However, no statistically significant interaction effect was found between both variables (F = .79, p = .58, η² = .002). It was found that there was an increase in family social support as a function of increased length of residence. The Bonferroni test showed a statistically significant difference in family social support between the first and third period (p < .001) and between the second and third period (p = .002); however, no statistically significant difference was found between the first and second period (p = .50). The Bonferroni test was used to conduct multiple comparisons between the four groups by place of origin. Latin Americans had the highest scores in family social support. The scores were significantly different between this group and Africans (p < .001) and Asians (p = .004); however, no statistically significant difference was found between Latin Americans and East Europeans (p = .017). East Europeans had the next highest scores, although no statistically significant difference was found between this group and Africans and Asians. No significant difference were found between Africans and Asians (p = .99)

Table 3

Descriptive Statistics of the Different Types of Social Support by Place of Origin and Length of Residence

ANOVA found statistically significant differences in social support from native friends between the three resident periods (F = 27.12, p < .001, η² = .024) and between the four groups by place of origin (F = 15.85, p < .001, η² = .021). However, no statistically significant interaction effect was found between both variables (F = 2.5, p = .020, η² = .007). The Bonferroni test was used to conduct multiple comparisons between the three periods of residence. It was found that there was an increase in social support from native friends as a function of increased length of residence. Statistically significant differences were found in social support from native friends between the first and third periods (p < .001) and between the second and third periods (p < .001); however, a lower statistically significant difference was found between the first and second period (p = .001). Latin Americans had the highest scores in social support from native friends. The scores were significantly different between this group and Africans and Asians (p < .001); however, no statistically significant difference was found between Latin Americans and East Europeans (p = .181). East Europeans had the next highest scores, which were much higher than those of Asians (p = .001) and slightly higher than those of Africans (p = .016). No statistically significant difference was found between Africans and Asians (p = .99).

Univariate analysis found no statistically significant differences in social support from immigrant friends between the three periods of residence. No statistically significant interaction effect was found. However, there were statistically significant differences in social support from immigrant friends between groups by place of origin (F = 4.42, p = .004, η² = .006). The Bonferroni test showed that Latin Americans had the highest scores in social support from immigrant friends. The scores were statistically significantly different between this group and Asians (p = .005). No significant differences were found between the Latin American group and the East European (p = .059) and African groups (p = .99). Africans had the next highest scores. No statistically significant difference was found in scores between the African and Asian groups (p = .18) or in scores between the East European and Asian groups (p = .99).

ANOVA found statistically significant differences in social support from neighbours between the three periods of residence (F = 16.33, p < .001, η² = .015). No statistically significant differences were found between groups by place of origin, and no statistically significant interaction effect was found. It was found that there was an increase in social support from neighbours as a function of increased length of residence. The Bonferroni test found statistically significant differences in social support from neighbours between the first and third periods (p < .001) and between the second and third periods (p = .002); however, a lower statistically significant difference was found between the first and second period (p = .03).

ANOVA found no statistically significant differences in social support from the community between the three periods of residence. No statistically significant interaction effect was found. However, statistically significant differences were found in social support from the community between groups by place of origin (F = 6.83, p < .001, η² = .010). The Bonferroni test was used to conduct multiple comparisons between the four groups by place of origin. Africans had the highest scores in social support from the community. Significantly different scores were found between this group and those of the Latin Americans (p = .005) and Asians (p < .001). No significant difference in scores was found between Africans and East Europeans (p = .180). East Europeans had the next highest scores. No statistically significant difference was found in scores between this group and Asians (p = .21). No statistically significant differences were found in scores between Latin Americans and Asians (p = .99).

Discussion and Conclusions

From the multidimensional perspective of social inclusion (Baker & Palmer, 2006; Hedetoft, 2013; Kirpitchenko & Mansouri, 2014), this article examined the social inclusion of immigrants in Málaga using the dimensions of sense of community, resilience, satisfaction with life, and perceived social support from family, native friends, immigrant friends, and neighbours and community.

The aim of this study was to investigate whether increased length of residence promotes the social inclusion of immigrants. The results support this association. Regardless of place of origin, length of residence in Málaga had a statistically significant effect on almost all of the dimensions of the sense of community, resilience, satisfaction with life, and perceived social support from family, neighbours, and natives. The only exceptions were support from immigrant friends and from the community. The values of the dependent variables increased as a function of increased length of residence in Málaga. This result is in line with those of other studies on human migration (Uña et al., 2009). The greatest increase was observed between participants who had been resident for fewer than 5 years and those who had been resident for more than 10 years. This result suggests that appropriate levels of social inclusion develop over the long term (Michel et al., 2012). This suggestion is reinforced by the fact that the group that had been resident in Málaga for more than 10 years included individuals who had resided there for several decades. Significant changes in support as a function of length of residence can be understood as a reflection of the development of social networks, which is a key aspect for increasing perceived support over time.

The analysis of the different sources of support over time shows that social support can be studied on various levels (Lin, Dean, & Ensel, 1986; Trejos-Herrera, Bahamón, Alarcón-Vásquez, Vélez, & Vinaccia, 2018). In particular, community support (i.e., parish, associations) showed little variation over time. This result may due to the fact that support from close relationships and social networks increases over time. However, it has been shown that community support has a lower positive effect on well-being than other forms of support. This aspect calls into question whether community support resources can fulfil the support needs of the immigrant population. We suggest that community support could be improved by identifying the resources that would fulfil the needs of this group (Hombrados-Mendieta, García-Martín, & Gómez-Jacinto, 2013). The results also show that support from other immigrants does not increase over time. This is explained by the fact that immigrants initially seek contact and support from their compatriots, but as they settle and become integrated in the receiving country, they encounter more contact and support from the native population. This aspect has been shown to be a good indicator of the social integration of immigrants (García-Cid, Hombrados-Mendieta, Gómez-Jacinto, Palma-García, & Millán-Franco, 2017), although we should continue to analyse this issue.

It has been shown that cultural proximity has a positive influence on almost all of the variables of social inclusion analysed in this study (Martínez et al., 2002). The only exception was SOC. Although the results suggest that SOC increases as a function of increased length of residence, no association was found between SOC and place of origin or cultural distance. These results suggest that the processes of immigration and adaptation to a new country involve a loss of SOC with the culture of origin and the development of a new SOC as part of the integration process in the new culture (Bathum & Baumann, 2007). Immigrants have to develop new adaptation strategies (Downie et al., 2007). The present study shows that this process was similar in all the immigrant groups. People identify more strongly with the setting in which they live as a function of an increase in SOC, which occurs with increased length of residence.

Latin Americans had significantly higher scores in resilience, family social support, support from native and immigrant friends, and SWL, followed in descending order by East Europeans, Africans, and Asians. However, Africans obtained the highest scores on perceived community support. In addition, more support from immigrant friends was received by Africans than by the East European and Asian groups.

Overall, the Latin American group had higher indexes on the variables under study. The Latin American group is culturally closer to Spain than the other three groups. The literature suggests that a common language facilitates adjustment to a new country (Arenas & Urzúa, 2016). Some studies have shown the relevance of cultural distance, and that Latin American and East European immigrants find it easier to adapt to Spanish culture than African immigrants (Checa & Monserrat, 2015). It has been suggested that Latin Americans perceive more support because they belong to horizontal collectivist cultures. This type of culture encourages sociability among peers in contrast with the vertical collectivist cultures common in Asia, where social interaction processes tend to be hierarchical (Shavitt et al., 2016). The results suggest that the Latin American group was the most resilient and had the highest levels of perceived SWL. Some studies have shown that Latin America immigrants develop more resilience, followed by Europeans and Africans (González, Vázquez, & Álvarez, 2013). It has also been shown that proximity to the receiving culture is associated with increased SWL (Ayyash-Abdo & Alamuddin, 2007).

Given that Latin Americans and East Europeans are closer to the culture of Spain than Asians and Africans, the results support the hypothesis that cultural proximity influences the processes of social inclusion. However, a statistically significant interaction was found between length of residence and place of origin only in the case of SWL, which suggests that place of origin barely moderates the positive influence of length of residence on social inclusion. Analysis of the results suggests that length of residence had a uniform effect on resilience across all immigrant groups. In contrast, the effect of length of residence on SWL was heterogeneous across groups. This effect was greater among Africans than among Latin Americans, which may be partially explained by the fact that Africans had the lowest levels of SWL during the initial period after migration. It is therefore worth highlighting that the African immigrants obtained the most benefit from the effect of the passage of time on SWL. Ethnic origin and cultural distance can explain a large part of the differences in cultural inclusion that occur over time between different immigrant groups who share the same receiving location (Martinovic et al., 2009a).

Thus, different immigrant groups may have different needs; in the present case, Africans took longer to become integrated, but when they did, they had greater SWL than other groups. This result underscores the relevance of modelling the effects of time as a variable in this type of study because its lack of inclusion leads to contradictory results.

Asian immigrants had the lowest scores on all the variables under study, whose result is in line with those of other studies (e.g., Wang & Lau, 2015). The present study found differences between the immigrant groups. Thus, further studies are needed to determine which variables promote inclusion in each group, although it should be emphasised that there was a significant increase in SOC in all four groups over time.

In relation to SOC and sources of social support, no significant interaction was found between length of residence and place of origin, which suggests that time had a uniform effect across different immigrant groups.

This study shows the relevance of length of residence as a variable to explain social inclusion processes in immigrants. It also shows that linear models are not the only viable alternative in the study of social inclusion, despite the positive trend over time (Bürgelt et al., 2008; Kwok-bun & Plüss, 2013; Tartakovsky, 2009). The social inclusion of immigrants should be discussed in terms of multiple specific processes rather than a single specific process and should be based on recognition of cultural diversity while taking into account the characteristics of place of reception. The overall similarity of the results regarding the effect of length of residence in Málaga and place of origin on the dependent variables suggests that these variables are highly interrelated and are reliable indicators of social inclusion processes. A similar pattern of effects was observed in relation to each indicator of social inclusion, thus confirming their interrelationship with the migratory process.

This study investigated the factors that contribute to the social inclusion of immigrants over time. A synthetic cohort design was used with the aim of overcoming the challenges involved in obtaining a representative longitudinal sample. Given that the sample was very large and comprised four different groups by place of origin, the results of the study may well be applicable to other settings. The results of research on migration processes are heavily influenced by the community into which immigrants enter and reside for varying periods of time (Hill, 1996; Long & Perkins, 2007), and so generalizations should be made with caution. However, we suggest that our results are broadly applicable to other settings and that they provide data of interest for planning interventions for social inclusion.

In general, the participating immigrants had high social inclusion rates, which indicates the relevance of designing interventions addressing the development of the indicators used (social support, SOC, etc).

The increase in social inclusion indicators among participants who had resided in the city for fewer than 5 years or more than 10 years should be considered a key to psychosocial intervention given the need to analyse the needs of immigrants with 5 to 10 years of residence and to plan interventions based on these needs. These results should help make all professionals working in this field aware of the need to prolong measures over time. Studies should also include immigrants already settled in the destination country, based on the understanding that the inclusion process is not linear or similar in all immigrant groups.

It was found that African and Asian immigrants are the most vulnerable groups, and so all professionals involved should pay particular attention to these groups. The identification and definition of the characteristic elements of the different ethnic communities would substantially improve the effectiveness of the interventions conducted (Robinson & Liu, 2015). Regardless of the place of origin or cultural distance, time was the only variable associated with an increased SOC. These findings demonstrate the need to promote policies and intervention measures that encourage the participation of immigrants in relevant community issues in order to increase their potential for integration in the community and their sense of belonging both from the beginning of the migratory process and continuously thereafter (Itzhaky, Zanbar, Levy, & Schwartz, 2015; Talò, Mannarini, & Rochira, 2014). Community support has not improved over time, and thus it is relevant to plan and optimize community resources (e.g., associations) aimed at the immigrant population and to improve the accessibility and availability of these resources. A very positive step would be to develop programs in different settings (e.g. neighbourhoods, schools) to strengthen inter-ethnic support networks and systematize the participation of immigrants in Spanish language courses in the first 4 years of migration given the long-term positive impact of language skills, cultural competence, development of social ties, and social integration itself (Geurts & Lubbers, 2017; Hoehne & Michalowski, 2016). The benefits of intervention programs have also been shown to improve resilience among immigrants in the first phase of migration and that their positive effects are both visible in the short term and can persist over time (Yu, Lam, Liu, & Stewart, 2015).

Some limitations of the study should be noted. The data were collected using self-report questionnaires. When self-report questionnaires are applied, the researcher makes the assumption that participants’ responses accurately reflect their feelings (Heppner, Kivlighan, & Wampold, 1992). In addition, these results may not adequately reflect the association between these variables in other countries, and thus it would be of interest to replicate these results in countries other than Spain.

It should also be noted that in the present study the temporal variable is constructed from cross-sectional data. It would be useful if future studies collected information at different times.

Conclusions

In conclusion, a multidimensional approach should be adopted to deepen the understanding of the social inclusion process of immigrants. It is assumed that the positive effect of length of residence on social inclusion is not linear in the short term, but manifests over the long term and is strongly associated with the receiving setting. These results should be of assistance in guiding social policies that recognise cultural diversity and guarantee and accelerate the positive dynamics of migratory processes in a residential setting.