Partner Type Attitudes of Parents and Adolescents: Understanding the Decline in Transnational Partnerships among Turkish Migrants in Flanders

Abstract

This article describes an unprecedented decline in transnational partnerships among Turkish migrants in Flanders, using population data on all marriages between 2001 and 2008. Studying parental preferences regarding partner selection, we examine attitudinal mechanisms behind this decline. Based on a representative survey, our first result is that (direct) parental involvement in partner selection is lower among the more recent marriage cohorts. Second, parents and adolescents have moved away from a focus on the origin country in partner selection, while ethnic homogamy remains preferred. Third, openness toward mixed partnerships is found among a small but salient proportion of parents and associated with the religious attendance of male parents. We conclude that an attitudinal shift has occurred from a focus on the origin country to an orientation toward the local (ethnic) community. This decline in transnational partnerships is more a product of intense attitudinal change than a reflection of a policy change in the direction of discouraging partner migration and has implications for the integration and demographic characteristics of Turkish ethnic minorities in Flemish society. Additionally, international migration patterns are affected as the character of long-lasting migration from Turkey to Europe is changing and partner migration, one of the most accessible channels to enter Europe, is rapidly decreasing.

Introduction

Flanders, the Dutch-speaking northern part of Belgium, is characterized by a large Turkish ethnic minority, which, as is the case in Wallonia (the southern part of Belgium), the Netherlands, and Germany, originated in the context of labor migration in the 1960s (Atalik and Beeley 1993). Despite a moratorium on labor migration in 1974, immigration from Turkey to Flanders continued, driven by family reunification and, more importantly, partner migration (Lievens 2000). A preference for country-of-origin partners over local co-ethnic partners has been observed among first- and second-generation Turkish migrants for several decades. In his analysis of population data, for example, Lievens (1999) described the high percentage (around 70%) of transnational marriages among Turkish migrants in 1991. Similar results have been reported more recently in 2004, with approximately 60 percent of second-generation migrants choosing a transnational partner (Yalcin et al. 2006), as has been found in several other Western European countries as well (Hooghiemstra [2003] for the Netherlands, Baykara-Krumme and Fuß [2009] for Germany, Milewski and Hamel [2010] for France).

More recent studies, however, have shown that the prevalence of transnational partnerships among Turkish migrants in Belgium has been declining, making local co-ethnic partnerships the most preferred (Lievens et al. 2013). This trend echoes a similar decline in the Netherlands (Loozen, De Valk, and Wobma 2012; CBS 2015), Sweden (Carol, Ersanilli, and Wagner 2014), and Germany (Aybek, Straßburger, and Yüksel-Kaptanoğlu 2015), although the predominance of transnational partnerships in Germany has been lower than in other countries (Carol, Ersanilli, and Wagner 2014). Recent policy changes implemented throughout Europe to reduce immigration, especially partner migration, may partially explain this decline. Although migration policies hindering transnational partnerships can have a clear influence on partner type preferences (Leerkes and Kulu-Glasgow 2011; Carol, Ersanilli, and Wagner 2014), we cannot ignore the possibility that attitudinal changes may also contribute to this decline. This may be especially true in Belgium (and Flanders), where the decline in transnational partnerships began in 2006 and, thus, predates the emergence of stricter migration policies in 2011 (Lievens et al. 2013).

Although national register data are necessary to determine these particular partner selection trends, the data do not provide the in-depth information needed to discover possible attitudinal mechanisms behind these changes. Van Kerckem et al.’s (2013) qualitative research, however, provides a first insight into the attitudinal mechanisms behind the recent decline in transnational partnerships, suggesting three possible mechanisms. First, adolescent migrants tend to prefer local co-ethnic partners because they recognize the risks and downsides of transnational partnerships and evaluate the dependence of newly immigrated partners negatively. Second, the fact that premarital relationships are more often allowed may also enable the increase in local (co-ethnic) partnerships. Third, lower levels of parental involvement among the more recent marriage cohorts could also contribute to the decline in transnational partnerships, as parents are believed to be more traditional and to prefer transnational partnerships for their children.

The current study aims to further clarify the attitudinal dynamics behind recent trends in partner selections made by Turkish migrants¹ in Flanders. The qualitative sociological literature suggests that transnational partnerships may be declining partially because migrant parents are taking less initiative when selecting partners for their children (Van Kerckem et al. 2013). Parents are supposed to have a stronger preference for transnational partnerships than their children do. Hence, when parental involvement decreases, partnerships that are romantic matches increase and are more likely to occur in the local ethnic community instead of the origin country.

However, three observations should be made when considering parental influence. First, parental influence remains relevant, despite increasing autonomy, because the partner selection process of Turkish migrants has evolved from being initiated by parents and family to being initiated by partners with parental consent (Descheemaeker et al. 2009; Hooghiemstra 2003). Parental approval, thus, is still important and a well-accepted condition for getting married (Milewski and Hamel 2010; Huschek, de Valk, and Liefbroer 2012). Second, the literature attributing the decline in transnational partnerships among Turkish migrants to changes in parental involvement discusses the extent of their influence but overlooks preferences concerning specific partner types. The assumption is made that migrant parents prefer transnational partnerships for their children without researching specific parental preferences (e.g., Huschek, de Valk, and Liefbroer 2012; Van Zantvliet, Kalmijn, and Verbakel 2014). Third, less parental involvement does not necessarily lead to fewer transnational partnerships, as the prevalence of this partner type could also be a result of a match between the interests and objectives of all parties involved — parents and adolescents (Reniers and Lievens 1999). Therefore, the question arises whether the decline in transnational partnerships could be associated with a change in attitudes and preferences of adolescents and parents in addition to policy changes. To what extent do Turkish migrants consider transnational partnerships the ideal partner type, and to what extent are they more likely to prefer local co-ethnic or possibly even ethnically mixed partnerships?

To address these questions, we start by discussing several mechanisms behind Turkish migrants’ transnational partnerships, with a focus on parental involvement and preferences. A first dataset is used to describe the recent trends in partner selection of Turkish migrants in Flanders. Next, a second dataset is used to unravel the mechanisms behind these recent trends and to determine whether parental involvement is actually lower in the most recent marriage cohorts. Then, we compare the distinct preferences migrant parents have concerning ideal partner types for daughters versus sons, especially regarding transnational partnerships. Furthermore, in view of possible attitudinal changes, we consider ethnicity’s central role in partner selection and determine which parents show more openness to mixed partnerships. Finally, we discuss adolescents’ preferences concerning their own future partners. We conclude by outlining the implications of our findings.

Mechanisms behind Transnational Partnerships

Several factors influence the complex process of partner selection: the preferences of individuals, the influence of third parties, and the constraints of the marriage market in which one searches for a partner (Kalmijn 1991). The mechanisms behind transnational partnerships can therefore be considered on three different levels: micro, macro, and meso. The micro-level includes individual preferences concerning ideal partners. On the one hand, several studies note traditional motives among Turkish migrants for choosing transnational partnerships (Lievens 1999; Timmerman, Lodewyckx, and Wets 2009), which are associated with higher levels of religiosity, maintaining and strengthening ethnic identity, and stronger ties with the origin country (Carol, Ersanilli, and Wagner 2014). On the other hand, migrant adolescents looking for a partner report a scarcity of eligible partners in the local marriage market (Hooghiemstra 2003).

Two factors at the core of this reported scarcity are the negative view men and women have of each other and the minority group’s increasing diversification. First, migrants often have a negative view of potential local co-ethnic partners and an idealized image of potential partners from the origin country (Hooghiemstra 2003; Sterckx and Bouw 2005; Straßburger 2005). Second, the Turkish ethnic minority in Flanders is rather homogeneous in terms of a lower educational level. Hence, because of increasing diversification, highly educated migrants may have a harder time finding an equally educated partner in the local ethnic community, motivating them to seek an eligible partner in Turkey (Timmerman, Vanderwaeren, and Crul 2003). This practice is more common among more highly educated women than men (Autant 1995; Lievens 1999; Timmerman, Lodewyckx, and Wets 2009; Liversage 2012). Lievens (1999) concludes that by choosing a transnational partner, migrant women may be able to gain more autonomy and “power” within the relationship because they are not subject to the generally strong influence of their in-laws and because their partner is new to the resident country. Hence, women may choose this partner type to satisfy modern goals, whereas men search within the origin country for more traditional spouses.

Evidence for this hypothesis has been found mainly in qualitative studies (Autant 1995; Timmerman, Lodewyckx, and Wets 2009; Liversage 2012). In quantitative studies, González-Ferrer (2006) and Milewski and Hamel (2010) found no support for this interaction of gender and educational attainment of Turkish migrants. Carol, Ersanilli, and Wagner (2014) did find support for this interaction but questioned educational attainment as a proxy for traditional orientation, as the interaction remained significant while controlling for religiosity. Hence, the choice for a transnational partnership could also be the result of a lack of appropriate partners in the country of residence (Straßburger 2003). While highly educated women may need to turn to their origin country to find co-ethnic partners with similar levels of education, this is less true for highly educated men, who are more likely to marry women with lower education.

The macro-level includes structural factors such as transnational networks, socio-economic conditions in the origin country, and migration policies. Turkish migrants from the same Turkish region often find themselves living in the same communities in Flanders (Surkyn and Reniers 1996). These networks preserve social and cultural structures from the region of origin and enable transnational partnerships, as migrants’ marriage market transcends national borders (Timmerman 2006). Furthermore, the literature discusses the importance of socio-economic conditions as push factors for migration (Timmerman, Lodewyckx, and Wets 2009). Interviews conducted in 2005 with partner migrants still living in Turkey reveal that motives for immigrating are often related to socio-economic factors rather than to the partnership itself. Socio-economics combined with European policies restricting migration opportunities from outside Europe make marriage one of most accessible channels of migration to Europe. This situation generates a large pool of possible partners in Turkey and increases pressure on Turkish migrants to marry a partner from the origin country. Migrants in turn become more attractive marriage partners, potentially giving them a better chance of finding a suitable partner in the origin country than in the local ethnic community (Van Kerckem et al. 2013).

However, the recent implementation of stricter requirements for migration throughout Europe has created barriers to choosing a transnational partnership, even for nationals (Beck-Gernsheim 2007). The policy changes, which establish a minimum age and include income, language, and housing requirements, have been implemented throughout Western Europe in an attempt to reduce immigration in general and transnational partnerships in particular, as studies in the Netherlands and Sweden illustrate (Leerkes and Kulu-Glasgow 2011; Carol, Ersanilli, and Wagner 2014). They are also partially a result of policymakers’ concerns about ethnic minorities’ level of integration in the face of a constant influx of immigrants (Schmidt et al. 2009). Perhaps most importantly for our study, these policies create socio-economic and gender inequalities in the freedom to choose a partner (Van Kerckem et al. 2013; Aybek 2015). Only those with a higher socio-economic status can freely choose a partner (Leerkes and Kulu-Glasgow 2011), and women may have more trouble meeting the (income) requirements to marry a partner living in Turkey (Kraler 2010).

These policies also make transnational partnerships more stressful because the migrating partner’s dependency is very high, and traditional gender roles may shift when the partner migrant is a man (Kraler 2010). Under them, transnational partnerships become riskier because the level of uncertainty about whether the union can be formed is higher (Aybek 2015; Aybek, Straßburger, and Yüksel-Kaptanoğlu 2015). When it is not possible, consequences may be either divorce or relocation to Turkey. Hence, these policies may lead to cancelation or postponement of the partnership until the requirements can be met. It is also possible that the restrictions will foster a change in attitudes toward transnational partnerships themselves (Carol, Ersanilli, and Wagner 2014). In all these ways, then, whereas in the past, migration policies made marriage one of the most accessible channels of migration to Europe, today, the recently reformed policies may have the opposite effect.

The meso-level includes third parties such as peers, the local ethnic community, parents, and extended family members. Partner selection is seldom an entirely individual choice, as people strive to gain social approval and adhere to group norms (Ajzen 1985). For ethnic minorities, strict norms of endogamy, for example, can be important as they pursue group identity (Kalmijn 1998). Third parties are important in the partner selection process as they transmit values and norms during socialization and act as role models (Youniss and Smollar 1985). Because Turkish family culture has a strong influence on the partner selection process and because parental preferences are the focus of this article, we assess the role of family and parents in particular.

In Turkish culture, marriage is seen as a bond between families and individuals (Timmerman 2006), and the reputation of potential partners is essential in the preservation of family honor. Young adults’ behavior is determined by an honor and shame system accompanied by a virginity norm and a strong preference for ethnic homogamy (Esveldt et al. 1995; Hooghiemstra 2003).² The social network of migrant adolescents uses these norms to regulate the partner selection process and exerts a high level of social control, especially on daughters. Because adhering to these social norms leads to certain levels of ethnic and gender segregation and, thus, limits opportunities to meet potential partners, parents and close relatives can play active roles in the selection process. We discuss migrant parents’ involvement in the partner selection process and the link with transnational partnerships more thoroughly in the next section.

Parental Involvement and Preferences

Involvement

As indicated above, parental involvement in partner selection of Turkish migrants is motivated by the central role marriage plays in the preservation of family honor (Esveldt et al. 1995; Hooghiemstra 2003). Finding a suitable partner is essential, and parents are generally trusted to have reliable insight and to offer the best guidance in the partner selection process. Parental involvement is especially high in transnational partnerships (Huschek, de Valk, and Liefbroer 2012), as partner compatibility can be evaluated beforehand, which is important given the greater uncertainties and risks of this partner type (Aybek, Straßburger, and Yüksel-Kaptanoğlu 2015). Parents and family members often serve as matchmakers between two partners living in different countries; however, parental involvement is also present in the formation of local partnerships (Hense and Schorch 2013). Adolescents often accept parental and family involvement in response to possible family pressure, as well as their own desire for family cohesion and solidarity (De Valk and Liefbroer 2007). Therefore, the practice of arranged partnerships is frequently accepted because it is based on a supportive network and on the compatibility of the partners rather than on emotions alone (Aybek 2015).

Nevertheless, there is variation in the degree of parental involvement, as described by two models of family values: the traditional model of interdependence and the model of psychological interdependence (Kâğıtçıbaşı 1996). The former is characterized by a collective focus and by children’s dependence on parents. Thus, arranged partnerships are more frequent (Hortaçsu and Oral 1994; Kagitcibasi and Ataca 2005). Parents who have lower levels of educational attainment, traditional attitudes regarding gender roles, religious commitment, and rural origins are more likely to subscribe to these family values. The psychological interdependence model, by contrast, is characterized by a strong conformity to parental preferences and expectations; however, children have more autonomy in life-course decisions, thus making couple-initiated romantic partnerships more common (Hortaçsu and Oral 1994). This model prevails among parents with higher educational attainment, fewer children, and less religious commitment (Kagitcibasi and Ataca 2005).³

Kagitcibasi and Ataca (2005) describe an evolution in family values, facilitated by socio-economic development, from the traditional model to the model of psychological interdependence between generations. Nevertheless, this evolution does not mean that adolescents become completely autonomous in their partner selection. The psychological interdependence model differs from the Western independence model in that the former assumes emotional dependence between parents and child instead of emotional separation. Even when parents do not propose a specific partner, they still may influence their children’s partner selection process in other ways (Hooghiemstra 2003; Sterckx and Bouw 2005). For example, parents can create opportunities for their children to meet potential partners during holidays in Turkey while restricting social contacts in the resident country. Additionally, they might explain how a transnational marriage would please them or encourage other family members to influence their child’s choice.

Preferences

Qualitative and anthropological studies from the late 1980s and early 1990s indicate that parents of Turkish descent have distinct preferences for transnational partnerships (De Vries 1987; Van der Hoek and Kret 1992; Callaerts 1997). Several factors explain this orientation. First, migrant parents often have strong ties to their families in Turkey through transnational networks characterized by high levels of solidarity and pressure or a sense of obligation to help kin that stayed behind (Sterckx and Bouw 2005; Timmerman, Lodewyckx, and Wets 2009). Transnational partnerships can help maintain and strengthen those transnational networks. Second, parents belonging to the first generation may find themselves living in a largely unknown society and culture, which may lead to preferences for transnational partnerships, making them adhere more rigidly to their traditions, customs, and ethnic identity (Timmerman 2006). Third, parents generally believe that partners from the origin country are more compatible (e.g., sharing norms and values and being a better cultural fit) than local co-ethnics, who often have a bad reputation (Callaerts 1997; Sterckx and Bouw 2005). Several studies, in fact, indicate that both parents and children have an idealized view of transnational partners and attribute to them characteristics that they do not find among local co-ethnics (Hooghiemstra 2001; Timmerman 2006).

However, qualitative research from the mid-2000s on does not report an explicit preference for partners from the origin country (Yalcin et al. 2006; Descheemaeker et al. 2009). Turkish migrants in a transnational partnership, as well as Turks who migrated as newlyweds, are careful in recommending transnational partnerships because of the limited time partners have to get to know each other beforehand and the difficulties they themselves encountered during the partnership (Yalcin et al. 2006; Descheemaeker et al. 2009; Aybek, Straßburger, and Yüksel-Kaptanoğlu 2015). Although transnational partnerships are often idealized because partners are expected to share religion, norms, and values, as well as an ethnic-cultural identity (Zemni, Casier, and Peene 2006; Descheemaeker et al. 2009), many parents and adolescents claim to have changed their minds about this partner type after witnessing relationship difficulties in transnational partnerships (Zemni, Casier, and Peene 2006; Descheemaeker et al. 2009). While shared ethnicity and religion remain important, there is no report of a distinct preference for a partner from Turkey among Turkish migrants, possibly indicating an evolution in attitudes regarding transnational partnerships, as has been hypothesized in other research (Esveldt et al. 1995; Lievens 1999; Huschek, de Valk, and Liefbroer 2012).

Recent Trends in Partner Selection of Turkish Migrants in Flanders

The trends in partner selection of Turkish migrants examined here are based on data extracted⁴ from the Belgian National Register. We include all first marriages of second-generation Turkish migrants in Flanders conducted between 2001 and 2008 (N = 7,274) that meet the following conditions: there is at least one partner who (1) is a resident of Flanders, (2) was born with Turkish nationality, and (3) either immigrated to Belgium before age 16 or was born in Belgium. We distinguish between three different partner types: (1) a transnational partnership with a partner living in Turkey, (2) a local co-ethnic partnership, and (3) a mixed partnership with a native Belgian. Although it is increasingly important to include cohabitation when researching Turkish migrants’ partner selection (De Valk and Liefbroer 2007), we choose to exclude it, as only 1.3 percent of first partnerships in our analyses were cohabitations. Among the Turkish population, living together without being married is (still) frowned upon, and people who choose this partner type are usually less religious and more highly educated (Kalmijn and Van Tubergen 2006; Huschek 2011). We analyze the number of marriages each year, regardless of their possible dissolution afterwards. Fifteen percent of Turkish migrants in a transnational marriage were divorced by 2008, compared to 17 percent of migrants in a mixed partnerships and 10 percent of migrants with a local co-ethnic partner.

As seen in Figure 1, the prevalence of transnational partnerships among Turkish men declined from 52.9 percent in 2001 to 33.8 percent in 2008, making partnerships with local co-ethnics the most common choice in 2008 (an increase from 44% to 59.9%). Mixed partnerships remain the least common partner type, although the percentage doubled from 3.1 percent to 6.3 percent. We note similar trends for women. A distinct decline in transnational partnerships (57.3% to 38.4%) is mostly absorbed by partnerships with a local co-ethnic (41.2% to 58.8%), with the percentage of mixed partnerships remaining low (1.5% to 2.9%).

Figure 1.

Trends in partner types of marriages formed between 2001 and 2008.

When comparing trends for men and women, the percentage of transnational partnerships among men was slightly lower than that among women in both 2001 (52.9% vs. 57.3%) and 2008 (33.8% vs. 38.4%). However, the decline in the prevalence of this partner type amounts to 19 percentage points for both sexes. The trend in local co-ethnic marriages within this time frame is similar for men and women (approximately 40% in 2001 and 60% in 2008). In contrast, mixed partnerships are more common among men than among women (approximately 6.3% and 2.9%, respectively). A preference for ethnically homogeneous partnerships seems to be more pronounced for women than for men (Esveldt et al. 1995; Hooghiemstra 2003).

To conclude, we observe a steep decline in transnational partnerships, making local co-ethnic partnerships the most popular partner type in 2008, while mixed partnerships remained the least preferred. In the following sections, we try to better understand this change in partner selection by Turkish migrants by analyzing distinct preferences among both parents and adolescents concerning ideal partner types.

Understanding Recent Trends in the Partner Selection of Turkish Migrants

Research Questions and Hypotheses

This article focuses on the possibility that attitudinal changes contribute to the decline in transnational partnerships among Turkish migrants in Flanders. We state that the influence of parental involvement’s decline on the prevalence of transnational partnerships is re-enforced by attitudinal changes regarding partner types, namely, transnational partnerships being less idealized by parents and partners.

Our first research question (Q1) is, To what extent does parental influence differ in partner selection across marriage cohorts? Among both Turks in Turkey (Kagitcibasi and Ataca 2005) and Turkish migrants in Belgium (Loobuyck 2005; Yalcin et al. 2006), an evolution toward more individualization in partner selection is visible. Hence, we hypothesize that parental involvement will be lower among more recent marriage cohorts (H1). Our second research question (Q2) addresses our main focus: Which partner types do migrant parents prefer for their children, and is there a difference for daughters versus sons? After evaluating qualitative research from recent decades (De Vries 1987; Van der Hoek and Kret 1992; Callaerts 1997; Yalcin et al. 2006; Descheemaeker et al. 2009), we hypothesize that parental attitudes toward ideal partner types have, in fact, changed and that the distinct preference for a partner from the origin country may have diminished over time (H2).

In its place, partnerships with local co-ethnics are becoming predominant, perhaps because ethnic homogamy is still preferred (H3). Social groups often enforce homogamy norms to protect group cohesion and maintain group values and traditions (Clark-Ibáñez and Felmlee 2004) and because mixed partnerships can be seen as a threat to group identity and solidarity (Sniderman and Hagendoorn 2007). Additionally, within Turkish culture, ethnic homogamy has a religious dimension. Islam does not consider children of a Muslim woman and non-Muslim man to be Muslim, although this norm is less severe for children of Muslim men in mixed marriages (De Vries 1987). Other gender dynamics may be relevant as well. For example, while gender equality in Turkish society is increasing, family honor is still largely dependent on women’s sexuality, which results in stricter social control of women’s behavior and a certain gender hierarchy (Esveldt et al. 1995; Hooghiemstra 2003). Hence, a preference for ethnically homogeneous partnerships may be more pronounced for daughters than for sons (H4).

Furthermore, in view of possible attitudinal changes, our third research question (Q3) is, What characterizes migrant parents who are more open to mixed partnerships with native Belgians? We suspect that openness to mixed partnerships may increase over successive generations (Lieberson and Waters 1988) (H5). As an ethnic minority group’s duration of stay and size increase, transnational networks between relatives may decrease in intensity, especially for second-generation migrants, potentially reducing the strength of emotional ties and sensitivity to kin obligations (Esveldt et al. 1995; Huschek, de Valk, and Liefbroer 2012). The resident country’s culture could also influence the ethnic identity of second-generation parents and possibly result in less emphasis on ethnic homogamy. Furthermore, we suspect that more highly educated parents will be more open to mixed partnerships (H6), since highly educated migrants are known to hold less traditional norms and values concerning partner selection (Huschek, de Valk, and Liefbroer 2012). Moreover, highly educated migrants are extensively exposed to the resident country’s values system during their education and have more opportunities to meet non-migrant peers (Kalmijn 1998; Baykara-Krumme and Fuß 2009). Higher educational attainment is also believed to weaken attachments to the origin community and to reduce cultural barriers to ethnically heterogeneous partnerships (Hwang, Saenz, and Aguirre 1997). Furthermore, we expect individuals who only occasionally attend religious services to be more open to mixed partnerships (H7), since they may attribute less significance to religious norms that prescribe ethnic and religious homogamy when selecting a partner (Kalmijn 1998; Hooghiemstra 2001). Additionally, some scholars find less ethnic distance and more openness to mixed partnerships among less religious people (Scheepers, Gijsberts, and Hello 2002).

Our final research question (Q4) is, What are adolescents’ preferences about the ethnicity of their future partners? Including adolescents’ attitudes is essential to obtaining a comprehensive view of recent partner selection dynamics of Turkish migrants, since their role in the process may become more important over time (Yalcin et al. 2006; Van Zantvliet, Kalmijn, and Verbakel 2014). The inclusion of their preferences, thus, may provide additional insight into future trends in Turkish migrants’ partner selection. We hypothesize that parents’ attitudinal changes concerning ideal partner types may also be present among adolescents, thereby resulting in a shared preference for local co-ethnic partnerships (H9).

Methods — Data Sample

For the following analyses, we use a second dataset — a subsample of the Sexpert survey, which consists of detailed and extensive data on the sexual health of Turkish migrants in Flanders and on its biomedical, psychological, and sociocultural correlates. Data collection took place between 2012 and 2013 through face-to-face interviews conducted by interviewers belonging to the Turkish ethnic minority, using a combination of CAPI (computer-assisted personal interviewing) and CASI (computer-assisted self-interview). Data were gathered in two stages to construct a population-based probability sample. The first stage included the selection of primary sampling units, that is, Flemish municipalities. By ordering and systematic sampling, a municipality’s chance of being selected was proportional to the number of inhabitants meeting eligibility criteria (14–59 years old, Belgian nationality, and at least one parent born with Turkish nationality). In the second stage, respondents were selected randomly from the Belgian National Register. The final sample contains 430 respondents (response rate of 57%). Data were weighted by gender and age to make the sample representative of the population of Flemish residents of Turkish descent aged 14 to 59. Respondents could choose between a Dutch and Turkish questionnaire; 36.4 percent answered in Turkish.

Parental Involvement

Our first analysis discusses the extent to which parents influence their children’s partner selection across marriage cohorts (Q1). To answer this first question, we selected all respondents from the Sexpert survey who were in a partnership (N = 263). Parental involvement in the formation of their partnerships was captured by asking, “To what extent did your parents influence the formation of your current partnership?” The degree of freedom in choosing their current partner was captured by asking, “Were you able to choose your partner freely?” Both were measured on a 5-point Likert scale ranging from 1 (absolutely not) to 5 (completely) and recoded into three categories for ease of interpretation: 1 and 2 into “little to none,” 3 into “some,” and 4 and 5 into “high.”

The first part of Table 1 describes parental involvement across marriage cohorts. The percentage of respondents whose partner choice was only slightly influenced by their parents has clearly increased in more recent marriage cohorts — from 15.1 percent of all partnerships formed before 1992 to 53.6 percent of all partnerships formed after 2006. Likewise, 77.4 percent of all partner choices made before 1992 were highly influenced by parents, compared to 32.1 percent of all partnerships formed after 2006. Hence, we can observe an evolution toward individualization and individual-initiated partnerships among more recent marriage cohorts, confirming H1. Nevertheless, 32.1 percent of respondents in the most recent cohort stated that their parents had a high degree of influence on the formation of their partnerships. However, this does not mean they felt they had no freedom of choice. Table 1’s second part shows that parental influence did not prevent feeling free to choose a partner. Of respondents who entered into a union after 2006, 85 percent felt they had a lot of freedom in choosing their current partner, although parents still had a strong influence on the formation of 32.1 percent of partnerships in this cohort.

Table 1.

Parental Involvement and Degree of Freedom in Partner Selection, by Marriage Cohort and by Sex.

		Parental Involvement (%) (N = 174)				Degree of Freedom in Partner Selection (%) (N = 255)
		Little to None	Some	High	Total	Little to None	Some	High	Total
Marriage cohorts	2007–2012	53.6	14.3	32.1	100	5	10	85	100
	2001–2006	42.5	12.5	45	100	16.7	11.7	71.7	100
	1992–2000	34	11.3	54.7	100	14.8	15.9	69.3	100
	Before 1992	15.1	7.5	77.4	100	29.9	25.4	44.8	100
		Parental Involvement (%) (N = 210)				Degree of Freedom in Partner Selection (%) (N = 296)
		Little to None	Some	High	Total	Little to None	Some	High	Total
Sex	Man	46.7	13.1	40.2	100	10.5	15.7	73.9	100
Sex	Woman	33	8.7	58.3	100	25.2	13.3	61.5	100

Parental involvement in Turkish migrants’ partner selection is known to differ according to their children’s gender (Baykara-Krumme 2015).⁵ Hence, Table 1’s lower part distinguishes between men and women. As expected, parents influence men’s partner choice less than they influence women’s choice (40.2% vs. 58.3%, respectively). The degree of freedom in choosing a partner also differs according to gender: 73.9 percent of men felt they had a lot of freedom in choosing their partner, compared to 61.5 percent of women. Similarly, more women felt very little freedom of choice: 25.2 percent versus 10 percent of male respondents.

Attitudes concerning Ideal Partner Types

Parental Preferences

Our second question assesses parents’ distinct preferences concerning ideal partner types for their children and whether parental preferences differ between daughters and sons (Q2). All Sexpert respondents who were asked about their preferences concerning their children’s future partner were either older than 25 (regardless of their relationship status) or younger and already married (N = 305). If respondents were childless (N = 173), they were asked to imagine which partner type they would want for their children if they had any.⁶

Six variables are used as indicators of parental preferences concerning child(ren)’s partner type. All are measured using a 5-point Likert scale ranging from 1 (very unimportant) to 5 (very important). The variables were obtained from the following question, asked separately for male and female children: “How important is it to you that the future marriage partner of your child is (1) of Turkish descent and currently lives in Turkey, (2) of Turkish descent and currently living in Belgium, or (3) of Belgian descent?” These six variables were recoded from five categories into three: 1 = “unimportant,” 2 = “in between,” and 3 = “important.”

We analyze parental preferences, using three-way cross-tabs that enable us to tease out the distinct preferences concerning different partner types (Tables 2 and 3). We identify four different partner types that are the most pronounced and show them in the shaded cells: a transnational partnership, a mixed partnership with a native Belgian, a partnership with a local co-ethnic, and an ethnically homogeneous partnership without a preference regarding the potential partner’s place of residence. Specific partner type preferences are not found in the shaded cells only; however, we believe that they identify the most pronounced preferences.

Table 2.

Parental Preferences for Future Partner Type for Daughters (N = 255, 100%).

		Local Co-ethnic
Native Belgian	Partner Living in Turkey	Unimportant	In Between	Important
Unimportant	Unimportant	22 (8.63%)	2 (0.78%)	50 (19.61%)
	In between	1 (0.39%)	27 (10.59%)	15 (5.88%)
	Important	2 (0.78%)	1 (0.39%)	46 (18.04%)
In between	Unimportant	6 (2.35%)	3 (1.18%)	14 (5.49%)
	In between	0 (0%)	15 (5.88%)	6 (2.35%)
	Important	0 (0%)	0 (0%)	10 (3.92%)
Important	Unimportant	4 (1.57%)	0 (0%)	6 (2.35%)
	In between	2 (0.78%)	1 (0.39%)	5 (1.96%)
	Important	0 (0%)	1 (0.39%)	16 (6.27%)
The shaded cells identify the most pronounced partner type preferences:
Transnational	Mixed	Local co-ethnic	Ethnic homogeneous

Table 3.

Parental Preferences for Future Partner Type for Sons (N = 251, 100%).

		Local Co-ethnic
Native Belgian	Partner Living in Turkey	Unimportant	In Between	Important
Unimportant	Unimportant	17 (6.77%)	3 (1.2%)	32 (12.75%)
	In between	0 (0%)	18 (7.17%)	11 (4.38%)
	Important	5 (1.99%)	1 (0.4%)	30 (11.95%)
In between	Unimportant	1 (0.4%)	1 (0.4%)	16 (6.37%)
	In between	0 (0%)	21 (8.37%)	9 (3.59%)
	Important	0 (0%)	0 (0%)	16 (6.37%)
Important	Unimportant	3 (1.2%)	0 (0%)	14 (5.58%)
	In between	0 (0%)	4 (1.59%)	10 (3.98%)
	Important	0 (0%)	0 (0%)	39 (15.54%)
The shaded cells identify the most pronounced partner type preferences:
Transnational	Mixed	Local co-ethnic	Ethnic homogeneous

We conclude that the number of parents that prefer transnational partnerships (who consider a native Belgian or local co-ethnic partner unimportant, but a partner living in Turkey important) is small. Only 0.78 percent of parents preferred this partner type for daughters, 1.99 percent for sons, confirming H2. Our findings concerning a preference for mixed partnerships (a partner who is local and co-ethnic or living in Turkey seen as unimportant, a native Belgian as important) are similar: 1.57 percent of parents preferred this partner type for daughters, 1.2 percent for sons. The majority of respondents preferred a partnership with either a local co-ethnic (a partner who is a native Belgian or living in Turkey is considered unimportant, a local co-ethnic important) or a co-ethnic regardless of the place of residence (a native Belgian is considered unimportant; a partner living in Turkey or a local co-ethnic important), confirming H3. We find this perspective to be the case for daughters especially, since 19.61 percent of parents prefer a local co-ethnic, and 18.04 percent prefer a co-ethnic regardless of place of residence. Concerning the ideal partner type for sons, 12.75 percent preferred a local co-ethnic and 11.95 percent a co-ethnic regardless of place of residence. As these numbers show, the preference for ethnic homogamy is more pronounced for daughters than for sons, confirming H4. Finally, there are a significant number of respondents who did not view any of these choices as important (8.63% concerning daughters and 6.77% concerning sons) or found all three equally important (6.27% concerning daughters and 15.54% concerning sons). We categorize these respondents as having no distinct preference for a particular partner type since they find either none or all of the ethnic characteristics important. These respondents may have moved away from ethnicity’s central role in the partner selection process and be more open to mixed partnerships. In the following analysis, we determine which factors differentiate parents with a distinct preference concerning ethnicity from those without.

Parents Considering Ethnicity Unimportant

We compare parents with no distinct preference for partners of Turkish descent, who consider ethnicity unimportant in their children’s partner choice, to parents who found the ethnicity of potential partners important (Q3) by constructing two dichotomous variables (one concerning daughters’ partners; one concerning sons’ partners) from the six variables used earlier. Table 4 describes the specific categorization of these variables. The same selection of respondents from the previous descriptive section is used.

Table 4.

Operationalization of “Considering Ethnicity Unimportant.”

		Local Co-ethnic
Native Belgian	Partner Living in Turkey	Unimportant	In Between	Important
Unimportant	Unimportant	1	0	0
	In between	0	0	0
	Important	0	0	0
In between	Unimportant	1	0	0
	In between	0	0	0
	Important	0	0	0
Important	Unimportant	1	1	1
	In between	1	1	1
	Important	1	1	1

Note. 1 = Not Important, 0 = Important.

Independent variables

Three variables are included to explain differences in the importance parents attribute to ethnicity as a characteristic of their children’s future marriage partner: migration generation, educational attainment, and religious attendance. We operationalize migration generation based on the socialization stage at which one migrates, which plays an important role in the development of attitudes and values (Bronfenbrenner 1986; Kalmijn and Kraaykamp 2018). A distinction is made between first-generation migrants, who were almost exclusively socialized in a Turkish context (migrated at age 15 or older); and second-generation migrants, who are mostly socialized in a Belgian context (migrated before 15 or born in Belgium). Additionally, it is important to note that only 4.9 percent of the first generation in this dataset migrated within the context of labor migration. Due to the age selection criterion (14–59 years old), most respondents belonging to the first generation migrated through family reunification (29.6%) or as partner migrants (54.2%).

Educational attainment is measured according to three categories: “primary school and lower secondary,” “higher secondary,” and “tertiary education” based on the highest diploma obtained (regardless of where it was obtained). Religious attendance is measured by the item, “In the past 6 months, how often did you attend religious gatherings or services?” Possible responses on a 6-point scale are never, only on special occasions, monthly, more than once a month, weekly, and more than once a week. The six categories are recoded into three for ease of interpretation: “never or on special occasions,” “at least once a month,” and “at least weekly.” The first part of Table 5 presents the univariate distributions of the variables described above.

Table 5.

Univariate Distributions of (In)dependent Variables and Binomial Logistic Regressions Considering Ethnicity Unimportant.

		Univariate Distribution		Binomial Regression Models
		Total Sample		Considering Ethnicity Unimportant (partner of daughters)		Considering Ethnicity Unimportant (partner of sons)
		N	%	b (SE)	b (SE)	b (SE)	b (SE)
Considering ethnicity unimportant (partner of daughters)	Not important	64	25.3
	Important	189	74.7
Considering ethnicity unimportant (partner of sons)	Not important	88	35.7
	Important	163	64.3
Intercept				–0.39 (0.86)	–0.06 (0.89)	0.33 (0.76)	0.56 (0.80)
Age				–0.01 (0.02)	–0.01 (0.02)	–0.01 (0.02)	–0.01 (0.02)
Sex
	Woman	216	50.3	–0.46 (0.33)	–0.77 (0.40)	–0.23 (0.30)	–0.50 (0.41)
	Man	214	49.7
Migration generation
	First	142	33
	Second	288	67	0.24 (0.34)	0.22 (0.34)	–0.45 (0.30)	–0.48 (0.31)
Educational attainment
	Primary and lower secondary	176	41.3	0.31 (0.37)	0.24 (0.37)	–0.35 (0.33)	–0.40 (0.33)
	Higher secondary	202	47.4
	Tertiary	48	11.2	0.70 (0.45)	0.67 (0.45)	–0.13 (0.43)	–0.16 (0.44)
Religious attendance
	Never, or on special occasions	202	47.2
	At least monthly	97	22.7	–1.01 (0.46)*	–1.68 (0.80)*	–0.61 (0.39)	–0.91 (0.62)
	At least weekly	129	30.1	–0.96 (0.38)**	–1.25 (0.47)**	–0.01 (0.33)	–0.27 (0.43)
Gender × religious attendance
	Woman × at least monthly				1.09 (0.97)		0.47 (0.80)
	Woman × at least weekly				0.77 (0.79)		0.62 (0.66)
–2 loglikelihood				263.57	261.74	315.64	314.69

*p ≤ 0.05. **p ≤ 0.01.

The predictors’ effect is estimated in binomial logistic regression models. The aim is to distinguish between parents who show more openness to mixed partnerships and parents who consider ethnicity important in this matter. In a first step, the control variables (sex and age) were included, with the remaining predictors added successively to build explanatory models. Only complete models are reported in Table 5’s second part, as there were no cases of suppression or redundancy while building the model. We do, however, include interaction terms between gender and religious attendance as a last step, since bivariate analyses show a large difference in religious attendance according to gender (58.8% of female parents never attended religious services, compared to 38.1% of male parents). This is not surprising as the religious practices of men and women are substantially different. For example, Islam compels men to go to the mosque each Friday, while this is not compulsory for women (Breuilly, O’Brien, and Palmer 1997). We analyze attitudes about partner choices for daughters and sons separately. In the results, the effects are recalculated to odds ratios for ease of interpretation.

With regard to the ethnicity of daughters’ partners, the effect of religious attendance indicates that the odds of considering ethnicity unimportant are 2.75 and 2.61 times lower for parents who attend religious services at least monthly or weekly, respectively, compared to parents who never attend religious services (1/[exp(–1.01)] = 2.75; 1/[exp(–0.96)] = 2.61), confirming H7. The interaction terms added in a subsequent step are not statistically significant. However, the main effects remain significant, indicating that the effect of religious attendance is only significant for men: the odds of considering ethnicity unimportant are respectively 5.36 and 3.49 times lower for men who attended religious services at least monthly or weekly than for men who never attended religious services (1/[exp(–1.68)] = 5.36; 1/[exp(–1.25)] = 3.49). Contrary to H7, H5 and H6 cannot be confirmed as no other significant effects are found. Similarly, regarding the ethnicity of sons’ partners, none of the predictors explained the differences in finding ethnicity unimportant.

Adolescent Preferences

Our final research question concerns adolescents’ preferences regarding their future partnerships (Q4). Sexpert respondents who are younger than 26 and unmarried were asked the same questions as “adult” respondents, with the understanding that the questions referred to their future partners (N = 123). The same approach was used to obtain six variables based on the following questions, asked separately of girls and boys: “How important is it to you that your future marriage partner is of Turkish descent and currently living in Turkey, of Turkish descent and currently living in Belgium, or of Belgian descent?” As with our analysis of parent preferences, we analyze adolescents’ attitudes in three-way cross-tabs (see Table 6).

Table 6.

Preferred Future Partner Type of Adolescents.

Preferred Future Partner Type of Girls (N = 61, 100%)
		Local Co-ethnic
Native Belgian	Partner Living in Turkey	Unimportant	In Between	Important
Unimportant	Unimportant	3 (4.92%)	1 (1.64%)	26 (42.62%)
	In between	0 (0%)	3 (4.92%)	5 (8.20%)
	Important	1 (1.64%)	0 (0%)	6 (9.84%)
In between	Unimportant	0 (0%)	2 (3.28%)	5 (8.20%)
	In between	1 (1.64%)	1 (1.64%)	1 (1.64%)
	Important	0 (0%)	0 (0%)	0 (0%)
Important	Unimportant	1 (1.64%)	1 (1.64%)	2 (3.28%)
Important	In between	0 (0%)	0 (0%)	0 (0%)
	Important	1 (1.64%)	0 (0%)	1 (1.64%)
Preferred Future Partner Type of Boys (N = 62, 100%)
		Local Co-ethnic
Native Belgian	Partner Living in Turkey	Unimportant	In Between	Important
Unimportant	Unimportant	2 (3.23%)	1 (1.61%)	10 (16.13%)
	In between	0 (0%)	2 (3.23%)	2 (3.23%)
	Important	0 (0%)	0 (0%)	7 (11.29%)
In between	Unimportant	1 (1.61%)	3 (4.84%)	8 (12.90%)
	In between	0 (0%)	4 (6.45%)	4 (6.45%)
	Important	0 (0%)	0 (0%)	0 (0%)
Important	Unimportant	1 (1.61%)	1 (1.61%)	7 (11.29%)
	In between	0 (0%)	0 (0%)	1 (1.61%)
	Important	0 (0%)	0 (0%)	8 (12.9%)
The shaded cells identify the most pronounced partner type preferences:
Transnational	Mixed	Local co-ethnic	Ethnic homogeneous

We conclude that only a small number of adolescents prefer a transnational partnership (considering native Belgian and local co-ethnic partners unimportant, a partner living in Turkey important), with only one girl preferring this partner type (1.64%). Similarly, we find no distinct preference for a mixed partnership (considering a partner living in Turkey and a local co-ethnic partner unimportant, a native Belgian important), with only one girl (1.64%) and one boy (1.64%) specifically preferring a native Belgian for a future partner. The only distinct preference observed was for a local co-ethnic partnership (considering a native Belgian partner and a partner living in Turkey unimportant, a local co-ethnic important). This confirms our hypothesis that attitudinal changes similar to those of parents are visible in adolescents, thus resulting in distinct preferences for local co-ethnics (H9). This preference is more pronounced among girls (42.62%) than among boys (16.13%), which was to be expected because the preference for ethnically homogeneous partnerships is more pronounced for girls than for boys, as indicated earlier.

Contrary to the previous multivariate analyses considering parental attitudes, we do not build multivariate models to assess which adolescents show more openness toward mixed partnerships, due to the small sample sizes. Exploratory analyses, however, show similar results to the analyses of parental attitudes — namely, the importance of religious attendance.

Discussion

Turkish migrants’ partner choices are changing rapidly in several European countries, after having been consistent for decades. Recent studies describe a decline in the prevalence of transnational partnerships of Turkish migrants in the Netherlands (Loozen, De Valk, and Wobma 2012; CBS 2015), Sweden (Carol, Ersanilli, and Wagner 2014), and Belgium (Lievens et al. 2013). This decline has been partially ascribed to recent policy changes implemented throughout Europe to discourage partner migration (Leerkes and Kulu-Glasgow 2011; Carol, Ersanilli, and Wagner 2014) and other changes such as a higher prevalence of premarital relationships and declining parental involvement in children’s partner choice (Huschek, de Valk, and Liefbroer 2012; Van Kerckem et al. 2013; Van Zantvliet, Kalmijn, and Verbakel 2014). Authors describe the possible influence of decreasing parental involvement on the prevalence of transnational partnerships but overlook the preferences themselves for specific partner types. This article adds to the existing literature by focusing on these preferences, both of parents and of young people. Based on the recent literature, we hypothesize that parental involvement, although decreasing, is still important in the partner selection of young Turkish adolescents and that parents’ attitudes are shifting away from a preference for a partner from the origin country (Yalcin et al. 2006; Descheemaeker et al. 2009). This shift could also explain why a sharp decrease in the prevalence of transnational partnerships was observed in Belgium years before the implementation of stricter migration policies (Lievens et al. 2013).

Our results confirm a decline in the degree of parental involvement in partner selection over time (Milewski and Hamel 2010; Van Zantvliet, Kalmijn, and Verbakel 2014), as has been observed in Turkey as well (Baykara-Krumme 2015). Nonetheless, parental involvement has not disappeared, as parents remain highly involved in the formation of a third of the partnerships in the most recent cohort, echoing earlier research (Milewski and Hamel 2010; Baykara-Krumme 2015). Moreover, we find that high levels of parental involvement do not contradict freedom of choice. Parental involvement is inherent to the process and therefore generally well accepted and appreciated (Loobuyck 2005; Aybek 2015).

Furthermore, our analyses reveal that in 2013, parents’ influence on the partner selection of their children favored local (co-ethnic) partners over transnational partners. This is a striking observation given that research from only two decades earlier reported a strong and pronounced preference for a partner from the origin country (De Vries 1987; Van der Hoek and Kret 1992; Callaerts 1997). This rapid attitudinal change undoubtedly is reflected in the recent decline in transnational partnerships, which is shaped by far more than a policy change in the direction of discouraging partner migration.

Several elements could be underlying this attitudinal shift. First, such a change may result from growing awareness of the possible risks associated with transnational partnerships, such as higher divorce rates and decreasing social support (Eeckhaut et al. 2011; Van Kerckem et al. 2013; Aybek, Straßburger, and Yüksel-Kaptanoğlu 2015). Residents of Turkey could be motivated to marry a migrant primarily for the opportunity it provides to settle legally in a European country. Moreover, transnational partnerships formed by second-generation migrants are known to be less stable due to cultural differences (Eeckhaut et al. 2011). Since they are born and/or raised and educated in Belgium, their cultural frame of reference is a mixture of Belgian and Turkish cultures. Other reported complications and risks include poor language skills, unemployment and financial troubles, contradictory expectations, and social isolation of the marriage migrant (Van Kerckem et al. 2013). Second, transnational family networks may decrease in intensity, especially for the second generation, as the duration of stay and size of the ethnic minority group increase. This could reduce the strength of the emotional ties and sensitivity to kin obligations in Turkey (Straßburger 2005), as well as the opportunity and ability to negotiate a transnational partnership (Esveldt et al. 1995; Huschek, de Valk, and Liefbroer 2012).

Besides the absence of a preference for transnational partners among parents in 2013, our analyses also show that an important proportion (25.3% for daughters and 35.7% for sons) of parents moved away from the religious norm of ethnic homogamy. These parents considered potential partners’ ethnicity less important and, consequently, may show more openness to mixed partnerships. This is a remarkable observation, given the religious motivation of resistance toward mixed partnerships, especially of female Muslims (De Vries 1987). This religious dimension, however, still is noticeable in two observations from our analyses. First, parents show more openness to mixed partnerships for sons than for daughters. Second, religious attendance is negatively associated with openness to mixed partnerships. Male migrants who never attended religious services were more likely to consider ethnicity unimportant concerning the partners of their daughters.

This study shows that the recent trends in partner selection of Turkish migrants can be understood as an attitudinal shift, which discloses much about the orientation of the Turkish ethnic minorities. Turkish migrants are becoming less oriented toward the origin country and more toward the local (ethnic) community. Furthermore, an openness to strengthening connections with the majority population — with regard to intimate relations — is found among a specific group. After being oriented toward the origin country for decades, the shift toward the local (ethnic) community is particularly relevant, as partner migration has recently been the focus of migration policies and public debates in several European countries (Wray 2009; Kraler 2010; Schmidt 2011; Jørgensen 2013; Van Kerckem et al. 2013). Marrying transnationally is believed to hinder the integration process (Surkyn and Reniers 1996; Hooghiemstra 2001; Lichter, Carmalt, and Qian 2011) because it is considered a sign of segregation, as migrants isolate themselves from the culture of destination and maintain their cultural praxis from the origin country (Berry 1997; Ward, Furnham, and Bochner 2005). Otherwise, partnerships with local partners and native partners especially are seen as manifestations of integration and assimilation (Lieberson and Waters 1988; Waters and Jiménez 2005). A growing focus on the local community and majority population could contribute to a decrease in ethnic differences, improve social integration, and diminish cultural distance (Gordon 1964).

With regard to the future, we assume the recent decline in transnational partnerships and increase in local co-ethnic partnerships will continue because of both parents’ and adolescents’ preference for (local) co-ethnic partnerships over transnational or mixed partnerships. Future research could obtain population data and analyze whether this trend continued to decline after 2008. Until now, emigration from Turkey was considered self-perpetuating, as the majority of both the first and second generations chose transnational partners. However, if the decline in the prevalence of transnational partnerships continues, it could both influence the characteristics of Turkish immigration to Flanders — currently defined by family migration — and significantly alter the structure of the Turkish ethnic minority in Flanders.

Studying Turkish migrants in Flanders is the equivalent of studying one ethnic minority group in one federal state of a small country. Notwithstanding our sample’s specificity, however, its relevance lies in the fact that we identify a trend in partner selection that is also present among both Turkish migrants in other countries and Moroccan migrants in Belgium, as shown by Lievens et al. (2013). The Moroccan migrant group in Belgium is quite similar to the Turkish minority group in Flanders (Beck-Gernsheim 2007; Huschek, de Valk, and Liefbroer 2012; Lievens et al. 2013; Carol, Ersanilli, and Wagner 2014). Our research, thus, provides greater insight into current partner selection decisions among a wide group of young adolescent migrants. Although partner selection decisions are key markers of entering adulthood and strongly related to choices in other domains of young adults’ lives, relatively little is known about different aspects of these decisions among the second generation. Increasing our knowledge about partner selection choices and the factors that shape them, then, is important, as a growing proportion of young adults in Europe have a migrant background.

Of course, this study is not without limitations. First, the Sexpert dataset’s sample size is rather small, limiting us to descriptive analyses and causing us to interpret the findings with caution. A larger and more representative dataset would enable us to test hypotheses in multivariate designs and reach more generalizable results. Second, we have considered possible explanations for the recent decline in transnational partnerships among Turkish migrants; however, the explanations still need verification using the appropriate analytical methods. Future research could use trend analyses to further test the various hypotheses. In spite of these limitations, our study adds to the existing literature by showing that there has been an attitudinal change among Turkish adolescents and parents regarding ideal partner types. This change is reflected in the recent decline in transnational partnerships, which is shaped by more than stricter migration policies as often described.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Ajzen

1985. “From Intentions to Actions: A Theory of Planned Behavior.” In Action Control, edited by Kuhl

Beckmann

, 11–39. Berlin: Springer.

Atalik

Beeley

. 1993. “What Mass Migration Has Meant for Turkey.” In Mass Migration in Europe: The Legacy and Future, 156–73. London: Bellhaven Press.

Autant

1995. “La tradition au service des transitions. Le mariage des jeunes Turcs dans l’immigration.” Migrants-Formations 101:168–79.

Aybek

C. M.

2015. “Time Matters: Temporal Aspects of Transnational Intimate Relationships and Marriage Migration Processes from Turkey to Germany.” Journal of Family Issues 36(11):1529–49.

Aybek

C. M.

Straßburger

Yüksel-Kaptanoğlu

İ.

. 2015. “Marriage Migration from Turkey to Germany: Risks and Coping Strategies of Transnational Couples.” In Spatial Mobility, Migration, and Living Arrangements, edited by Aybek

C. M.

Huinink

Muttarak

, 23–42. Cham, Switzerland: Springer.

Baykara-Krumme

2015. “Three-Generation Marriage Patterns: New Insights from the ‘Dissimilation’ Perspective.” Journal of Ethnic and Migration Studies 41(8):1324–46.

Baykara-Krumme

Fuß

. 2009. “Heiratsmigration nach Deutschland: Determinanten der transnationalen Partnerwahl türkeistämmiger Migranten.” Zeitschrift für Bevölkerungswissenschaft 34(1–2):135–63.

Beck-Gernsheim

2007. “Transnational Lives, Transnational Marriages: A Review of the Evidence from Migrant Communities in Europe.” Global Networks 7(3):271–88.

Berry

J. W.

1997. “Immigration, Acculturation, and Adaptation.” Applied Psychology 46(1):5–34.

10.

Breuilly

O’Brien

Palmer

. 1997. Religions of the World. London: Transedition Limited.

11.

Bronfenbrenner

1986. “Ecology of the Family as a Context for Human Development: Research Perspectives.” Developmental Psychology 22(6):723.

12.

Callaerts

1997. “Import huwelijken bij jonge Turkse migranten in Vlaanderen en Brussel: een kwalitatieve benadering.” Working Papers Etnische Minderheden in België, Vrije Universiteit Brussel.

13.

Carol

Ersanilli

Wagner

. 2014. “Spousal Choice among the Children of Turkish and Moroccan Immigrants in Six European Countries: Transnational Spouse or Co-ethnic Migrant?” International Migration Review 48(2):387–414.

14.

CBS, Dutch Statistics Office. 2015. http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=71888NED&D1=a&D2=0&D3=5&D4=l&D5=a&HDR=G1, G3, T&STB=G2, G4&VW=T.

15.

Clark-Ibáñez

Felmlee

. 2004. “Interethnic Relationships: The Role of Social Network Diversity.” Journal of Marriage and Family 66(2):293–305.

16.

De Valk

H. A. G.

Liefbroer

A. C.

. 2007. “Parental Influence on Union Formation Preferences among Turkish, Moroccan, and Dutch Adolescents in the Netherlands.” Journal of Cross-Cultural Psychology 38(4):487–505.

17.

De Vries

1987. Ogen in je rug. Turkse meisjes en jonge vrouwen in Nederland. Alphen aan den Rijn, the Netherlands: Samson.

18.

Descheemaeker

Wets

Heyse

Timmerman

. 2009. Partnerkeuze en huwelijkssluiting van allochtone mannen. Een kwantitatieve en kwalitatieve analyse van het partnerkeuzeproces en het huwelijk van Marokkaanse, Turkse en sikhmannen. Brussels: Instituut voor de gelijkheid van vrouwen en mannen.

19.

Eeckhaut

Lievens

Van de Putte

Lusyne

. 2011. “Partner Selection and Divorce in Ethnic Minorities: Distinguishing between Two Types of Ethnic Homogamous Marriages.” International Migration Review 45(2):269–96.

20.

Esveldt

Kulu-Glasgow

Schoorl

Van Solinge

. 1995. Migratiemotieven, migratienetwerken en partnerkeuze van Turken en Marokkanen in Nederland. The Haague: Stichting Nederlands Interdisciplinair Demografisch Instituut (NIDI).

21.

González-Ferrer

2006. “Who Do Immigrants Marry? Partner Choice among Single Immigrants in Germany.” European Sociological Review 22(2):171–85.

22.

Gordon

M. M.

1964. Assimilation in American Life: The Role of Race, Religion and National Origins. Oxford: Oxford University Press.

23.

Hense

Schorch

. 2013. “Arranged Marriages as Support for Intra-ethnic Matchmaking? A Case Study on Muslim Migrants in Germany.” International Migration 51(2):104–26.

24.

Hooghiemstra

2001. “Migrants, Partner Selection and Integration: Crossing Borders?” Journal of Comparative Family Studies 32(4):601–26.

25.

Hooghiemstra

. 2003. Trouwen over de grens. Achtergronden van partnerkeuze van Turken en Marokkanen in Nederland. Den Haag: Sociaal Cultureel Planbureau.

26.

Hortaçsu

Oral

. 1994. “Comparison of Couple-and Family-Initiated Marriages in Turkey.” Journal of Social Psychology 134(2):229–39.

27.

Huschek

2011. “Union Formation and Partner Choice of the Second Generation of Turkish Origin in Europe: The Influence of Third Parties and Institutional Context.” Dissertation, Free University Amsterdam.

28.

Huschek

de Valk

H. A. G.

Liefbroer

A. C.

. 2012. “Partner Choice Patterns among the Descendants of Turkish Immigrants in Europe.” European Journal of Population 28(3):241–68.

29.

Hwang

S.-S.

Saenz

Aguirre

B. E.

. 1997. “Structural and Assimilationist Explanations of Asian American Intermarriage.” Journal of Marriage and the Family 59(3):758–72.

30.

Jørgensen

M. B.

2013. “Danish Regulations on Marriage Migration: Policy Understandings of Transnational Marriages.” In Transnational Marriage: New Perspectives from Europe and Beyond, edited by Charsley

, 67–85. New York: Routledge.

31.

Kâğıtçıbaşı

Ç.

1996. Family and Human Development across Cultures: A View from the Other Side. Hillsdale, NJ: Lawrence Erlbaum.

32.

Kagitcibasi

Ç.

Ataca

. 2005. “Value of Children and Family Change: A Three-Decade Portrait from Turkey.” Applied Psychology 54(3):317–37.

33.

Kalmijn

1991. “Shifting Boundaries: Trends in Religious and Educational Homogamy.” American Sociological Review 56(6):786–800.

34.

Kalmijn

. 1998. “Intermarriage and Homogamy: Causes, Patterns, Trends.” Annual Review of Sociology 24:395–421.

35.

Kalmijn

Kraaykamp

. 2018. “Determinants of Cultural Assimilation in the Second Generation. A Longitudinal Analysis of Values about Marriage and Sexuality among Moroccan and Turkish Migrants.” Journal of Ethnic and Migration Studies 44(5):697–717.

36.

Kalmijn

Van Tubergen

. 2006. “Ethnic Intermarriage in the Netherlands: Confirmations and Refutations of Accepted Insights.” European Journal of Population 22(4):371–97.

37.

Kraler

2010. Civic Stratification, Gender and Family Migration Policies in Europe. Vienna: International Centre for Migration Policy Development.

38.

Leerkes

Kulu-Glasgow

. 2011. “Playing Hard (er) to Get: The State, International Couples, and the Income Requirement.” European Journal of Migration and Law 13(1):95–121.

39.

Lichter

D. T.

Carmalt

J. H.

Qian

. 2011. “Immigration and Intermarriage among Hispanics: Crossing Racial and Generational Boundaries.” Sociological Forum 26:241–64.

40.

Lieberson

Waters

M. C.

. 1988. From Many Strands: Ethnic and Racial Groups in Contemporary America. New York: Russell Sage Foundation.

41.

Lievens

1999. “Family-Forming Migration from Turkey and Morocco to Belgium: The Demand for Marriage Partners from the Countries of Origin.” International Migration Review 33(3):717–44.

42.

Lievens

. 2000. “The Third Wave of Immigration from Turkey and Morocco: Determinants and Characteristics.” In Communities and Generations: Turkish and Moroccan Populations in Belgium. Brussels: NIDI-CBGS/VUB-press.

43.

Lievens

Van de Putte

Van der Bracht

Caestecker

. 2013. Trends in partnerkeuze van eerste en tweedegeneratiemigranten in België: partnerkeuze van personen woonachtig in België met een migratieachtergrond, 2001-2008. Met bijzondere aandacht voor de bevolkingsgroep van Marokkaanse en Turkse herkomst in België. Gent, Belgium: Universiteit Gent.

44.

Liversage

2012. “Gender, Conflict and Subordination within the Household: Turkish Migrant Marriage and Divorce in Denmark.” Journal of Ethnic and Migration Studies 38(7):1119–36.

45.

Loobuyck

2005. “De sociale dynamiek van huwelijksmigratie bij etnische minderheden in West-Europa.” In Huwelijksmigratie, een zaak voor de overheid?, edited by Caestecker

, 23–35. Leuven, Belgium: Acco.

46.

Loozen

De Valk

H. A. G.

Wobma

. 2012. “Demografie.” In Jaarrapport integratie 2011, edited by Gijsberts

Dagevos

. Den Haag: SCP.

47.

Milewski

Hamel

. 2010. “Union Formation and Partner Choice in a Transnational Context: The Case of Descendants of Turkish Immigrants in France.” International Migration Review 44(3):615–58.

48.

Reniers

Lievens

. 1999. “Stereotypen in perspectief: de evolutie van enkele aspecten van het huwelijk bij de Turkse en Marokkaanse minderheden in België.” Migrantenstudies 15(1):28–44.

49.

Scheepers

Gijsberts

Hello

. 2002. “Religiosity and Prejudice against Ethnic Minorities in Europe: Cross-National Tests on a Controversial Relationship.” Review of Religious Research 43(3):242–65.

50.

Schmidt

2011. “Law and Identity: Transnational Arranged Marriages and the Boundaries of Danishness.” Journal of Ethnic and Migration Studies 37(2):257–75.

51.

Schmidt

Graversen

B. K.

Jakobsen

Jensen

T. G.

Liversage

. 2009. New Regulations on Family Reunification. What Have the New Regulations Meant for Patterns in Partner Selection by Ethnic Minorities? Copenhagen: SFI.

52.

Sniderman

P. M.

Hagendoorn

. 2007. When Ways of Life Collide: Multiculturalism and Its Discontents in the Netherlands. Princeton, NJ: Princeton University Press.

53.

Sterckx

Bouw

. 2005. Liefde op maat: partnerkeuze van Turkse en Marokkaanse jongeren. Amsterdam: Het Spinhuis.

54.

Straßburger

2003. Marriage Behavior and Partner Choice in the Immigrant Context - Marriages of Second Generation of Turkish Origin. Würzburg, Germany: Ergon Verlag.

55.

Straßburger

. 2005. “Transnational Social Spaces.” In Comparative European Research in Migration, Diversity and Identities, edited by Husband

Bosswick

. Bilbao, Spain: University of Deusto.

56.

Surkyn

Reniers

. 1996. “Selecte gezelschappen: over de migratiegeschiedenis en de interne dynamiek van migratieprocessen.” Bevolking en Gezin 25(themanummer):41–72.

57.

Timmerman

2006. “Gender Dynamics in the Context of Turkish Marriage Migration: The Case of Belgium.” Turkish Studies 7(1):125–43.

58.

Timmerman

Lodewyckx

Wets

. 2009. “Marriage at the Intersection between Tradition and Globalization: Turkish Marriage Migration between Emirdag and Belgium from 1989 to Present.” History of the Family 14(2):232–44.

59.

Timmerman

Vanderwaeren

Crul

. 2003. “The Second Generation in Belgium.” International Migration Review 37(4):1065–90.

60.

Van der Hoek

Kret

. 1992. Marokkaanse tienermeisjes: gezinsinvloeden op keuzen en kansen: verschillen in de voorbereiding op de deelname aan onderwijs en arbeidsmarkt. Utrecht, the Netherlands: Uitgeverij Jan van Arkel i.s.m. Rijksuniversiteit Leiden.

61.

Van Kerckem

Van der Bracht

Stevens

P. A. J.

Van de Putte

. 2013. “Transnational Marriages on the Decline: Explaining Changing Trends in Partner Choice among Turkish Belgians.” International Migration Review 47(4):1006–38.

62.

Van Zantvliet

P. I.

Kalmijn

Verbakel

. 2014. “Parental Involvement in Partner Choice: The Case of Turks and Moroccans in the Netherlands.” European Sociological Review 30(3):387–98.

63.

Ward

Furnham

Bochner

. 2005. The Psychology of Culture Shock. London: Routledge.

64.

Waters

M. C.

Jiménez

T. R.

. 2005. “Assessing Immigrant Assimilation: New Empirical and Theoretical Challenges.” Annual Review of Sociology 31:105–25.

65.

Wray

2009. “Moulding the Migrant Family.” Legal Studies 29(4):592–618.

66.

Yalcin

Lodewyckx

Marynissen

Van Caudenberg

Timmerman

. 2006. Verliefd, verloofd…gemigreerd: een onderzoek naar Turkse huwelijksmigratie in Vlaanderen. Antwerpen: Steunpunt Gelijkekansenbeleid.

67.

Youniss

Smollar

. 1985. Adolescent Relations with Mothers, Fathers, and Friends. Chicago: University of Chicago Press.

68.

Zemni

Casier

Peene

. 2006. Studie naar de factoren die de vrijheid van keuze van een echtgenoot beperken, bij bevolkingsgroepen van vreemde oorsprong in België. Ghent, Belgium: CIE.