Patriarchal Attitudes in Turkey 1990–2011: The Influence of Religion and Political Conservatism

Patriarchal Attitudes in Turkey 1990–2011: The Influence of Religion and Political Conservatism Abstract While contemporary theories associate gender equality with the process of modernization and economic development, we find that Turkey does not follow this pattern. We investigate changes in patriarchal attitudes from 1990 to 2011 in Turkey, and how the country has been influenced by religiosity and political conservatism. By studying how institutional changes affect gender inequality in Turkey, we find that Turkey has become more patriarchal since the 1990s. Moreover, the effect of religion on patriarchal attitudes has diminished over time while the effect of political conservatism has remained stable. The effect of religion is strongest for economic and weakest for educational patriarchy, while the effect of political conservatism does not vary. Introduction According to The Global Gender Gap Report (World Economic Forum 2015), gender inequality is highest in the Middle East, Africa, and South Asia while Scandinavian countries have the smallest gender gap in income, education, labor force, and political participation. Modernization scholars argue that a global trend toward gender equality is closely related to economic development and democratization of societies (Esmer 2008; Inglehart and Norris 2000, 2003; Inglehart, Norris, and Welzel 2002). They argue that the process of modernization results in cultural and normative changes that promote both democratization and an increase in the proportion of women in public life. For instance, high-income, economically advanced countries have a higher percentage of women in parliament and more egalitarian values of gender than low-income countries (Inglehart and Norris 2003). While contemporary theories may elucidate the revolutionary change in gender values in the Western World, we feel that they fail to explain gender value change in Middle Eastern societies. We find that the trend of increasing gender equality coterminous with the process of industrialization may not be applicable to Middle Eastern regions even when countries are governed through democratic elections. Turkey, a country with a predominantly Muslim population and, throughout the twentieth century, a secular, democratically elected government, is a preeminent example of this. While Turkey’s economic development has been on the rise since the turn of the century, women’s inequality has lagged in every measure (Hansen 2013). In 2006, Turkey was ranked 105th among 115 countries in terms of gender inequality (the 91st percentile). Turkey is ranked 130th among 145 countries in the most current report in 2015 (the 89th percentile)—with Iceland and Norway in the top two positions, and Yemen and Pakistan in the bottom two positions. We know very little about changes in patriarchal values over time in predominantly Islamic societies. Modern Turkey is not comparable with other Islamic nations because of its Westernization and its movement toward secularization since its establishment. Nevertheless, an overwhelming majority of the Turkish population are devout Muslims who practice religion on a regular basis (Esmer 2008). Despite its historical differences with much of the rest of the Islamic world, by studying patriarchal attitudes in Turkey, we can test theories about attitudes toward gender issues in other non-Western societies. Previous research finds that both religiosity and political conservatism have a negative impact on gender inequality (Civettini and Glass 2008; Davis and Greenstein 2009; Frejka and Westoff 2008). In this paper, we test the effect of two types of conservatism, political and religious, on attitudes related to gender. Our first research question asks, how have patriarchal attitudes in Turkish society changed from 1990 to 2011? Second, how does religiosity impact individuals’ patriarchal values? Lastly, how does political conservatism influence patriarchal attitudes in Turkey? We first provide a historical discussion of the feminist movement in Turkey. We then explicate institutional changes in four spheres of patriarchal values (economic, political, familial, and educational) that affect gender inequality in contemporary Turkey. Applying Integrated Threat Theory, we argue that different types of patriarchal values do not share the same trajectory over time and are influenced differently by religiosity and political conservatism. To examine the effect of religion and political conservatism on patriarchal values, we use data from five waves of the World Values Survey from 1990 to 2011 (before the current dramatic changes in Turkish governance). The World Values Survey (2017) measures the changing values, beliefs, and attitudes about the 3 social, economic, and political life of over 9,000 respondents in Turkey. Applying institutionalization of gender in the context of Turkey, we hypothesize that levels of patriarchal attitudes have been increasing in Turkey in that period. We argue that both religiosity and political conservatism are positively associated with patriarchal values. More specifically, we focus on four types of patriarchal attitudes: political, economic, familial, and educational. We posit that, based on Integrated Threat Theory, the effects of religion and political conservatism’s on political and economic gender attitudes are stronger than the effects of educational and familial patriarchal values. Moreover, our discussion on the institutionalization of economic, political, familial, and educational patriarchy complements our findings. In the next section, we will offer a brief history of women’s rights in Turkey, followed by a discussion of the institutionalization of gender, and how it impacts women’s educational, economic, familial, and political outcomes. Feminist Movements in Turkey Turkey was established in 1923 after the fall of the Ottoman Empire as a secular democratic state. Under the leadership of Mustafa Kemal Atatürk, Turkey underwent drastic economic, political, and social changes between 1920 and 1930 that aimed to modernize Turkish society and remove the power of religion from state institutions. These modernization efforts, also referred to as the Kemalist movement, involved adopting a multi-party system which abandoned the sultanate and the caliphate system; implementing secular public education; adopting the Latin alphabet; changing the official national language to Turkish; and promoting European dress codes. Religious attire such as headscarves and fezzes were first discouraged and later prohibited in government institutions (Eligür 2010; Pope and Pope 2011; Saktanber 2002; Saktanber and Çorbacıoğlu 2008). Improving the position of women in society was an explicit objective of the newly formed Turkish Republic. Between 1930 and 1934, the parliament implemented policies that created unprecedented civil rights for women, including the right to vote and the right to be elected to public office. This took place at a time when women in most Western and non-Western countries could not vote. Under the Civil Code of 1926, polygyny was banned, and women were given equal access to divorce and matters related to inheritance and property management. Women’s entrance into the public sphere was also facilitated by expanding educational opportunities and promoting their participation in the work force (Pope and Pope 2011). The goal of these institutional gender reforms was to remove Islam from public affairs through state control (Eligür 2010; Pope and Pope 2011). However, Islam remained an active force in Turkish culture, ingrained in most of the country’s social institutions (Saktanber 2002). The secularization experiment in Turkey was a drastic top-down change and the emancipation of women was revolutionary for its time. Emancipation of women in Turkey was dissimilar to women’s liberation movements in other developing countries (Erhart and Eslen-Ziya 2014; Eslen-Ziya and Korkut 2010). Some scholars argue that women’s emancipation played a strategic role in the Turkish government claiming its rightful place among other Western democratic nations (Kandiyoti 1987; Saktanber and Çorbacıoğlu 2008). However, Turkey did not have a strong women’s rights movement. The first wave of feminism was limited to a small number of women’s organizations initiated by educated and privileged women and did not recognize the plurality of experiences of women (Kandiyoti 1987). Instead, rights for women were granted by enlightened governing elite men and thus, emancipation efforts and reforms remained limited to elite Turkish bourgeoisie women while excluding the vast majority of rural and poor women at the margins of society (Kandiyoti 1987). Religious marriages, polygamy, the demand for başlık parası (bride-price), child marriages, virginity examinations, honor killings, and denial of girls’ rights to education remained continuing signs of gender inequality in Turkey (Kandiyoti 1988). The reforms in the newly formed country left crucial issues such as male privilege, heteronormativity, and traditional gender norms untouched. In this sense, women’s status was emancipated but unliberated, especially because of the absence of an autonomous women’s movement with consciousness-raising outreach efforts. The majority of women in Turkey remained oppressed, subordinated, and marginalized after the secularization experiment (Arat 2000). The first wave of feminism ended after the government closed down the Turkish Women’s Union in 1935. The Turkish Republican elite propagated the myth that, because the government granted equal legal and political rights with men, women’s organizations were no longer needed (Arat 2000). It was not until the 1980s that the feminist movement in Turkey brought up issues similar to the second wave of feminism in the West. These issues included sexual freedom, the oppression women experience in the family, and the elimination of violence against women (Aldıkaçtı Marshall 2005; Diner and Toktaş 2010). A new wave of feminism was initiated in the 1990s when Islamist, Kurdish nationalist, and LGBT movements began to challenge the earlier waves of feminism with the rise of identity politics. Kurdish women criticized the dominant Turkish feminist movement for othering them and ignoring issues related to the Kurdish conflict (Arat 2000). Activist Islamic women demanded more religious freedom by protesting the headscarf ban in government institutions (Aldıkaçtı Marshall 2005; İslam 2010; Saktanber and Çorbacıoğlu 2008). Islamic women were marginalized both by their male counterparts who supported political Islam and by the mainstream feminist movement. Islamic men stressed that Islam treats women equally and that no further questioning was necessary. Similarly, the dominant feminist discourse excluded Islamic feminists because they perceived Islam as a patriarchal religion that promoted the subservience and oppression of women (Erhart and Eslen-Ziya 2014). Some feminists, however, developed a more critical approach to Islamic feminism, which endorsed social agency in defining freedom and supported Islamic women’s right to cover their heads (Diner and Toktaş 2010). Overall, even though the three waves of feminist movements in Turkey have had the same goal of improving the position of women in society, they failed to coalesce into a unified voice. Institutional Changes in Patriarchal Values Affecting Gender Inequality in Contemporary Turkey Despite the laws and regulations implemented by Turkey to improve the rights and position of women in the past, gender inequality and a patriarchal value system continue to be a persistent problem. Turkey through the twentieth century was recognized as a modern society that embraced a neoliberal economic model and moderate Muslim values (Hansen 2013). However, while economic development was on the rise, the status of women lagged behind and the regulation of the private lives of women increased (Hansen 2013). Significant gender inequality persists in education, employment, and politics in Turkey (Arat 2010; Dedeoğlu 2012; The World Bank 2012). In the next section, we will discuss the institutional changes in four spheres of patriarchy—educational, familial, economic, and political—and how they influence the status of women. Institutional Changes in Educational Patriarchy Until the late 1990s, Turkey’s education system included a five-year compulsory primary education, followed by a non-compulsory three-year secondary and a three-year high school education (Caner et al. 2016). In 1997, the Turkish Ministry of National Education along with the Turkish Higher Education Council implemented an education reform that integrated primary and secondary schooling, extending primary education to a compulsory eight years (Caner et al. 2016). This reform increased the school enrollment of children from 86 percent in 1997 to 96 percent in 2002 (Dulger 2004). This institutionalization of education made an impressive impact on girls’ enrollment, particularly in rural areas. In the first year of the program, enrollment of girls in rural areas increased by 160 percent (Dulger 2004). Before this reform, parents were free to enroll their children in imam-hatip schools (religious schools that provide trainings for imams and preachers) after the five-year primary schooling (Kamal 2017). Girls composed almost half of the enrollments of imam-hatip schools albeit they could become neither imams nor hatips. The 1997 reform prevented children from enrolling in imam-hatip schools until the age of fourteen (Kamal 2017), and thus decreased the number of religious schools and their attendees. After the election of the Justice and Development Party in 2002, there has been an exponential increase in imam-hatip schools. Enrollees increased from 65,000 in 2002 to 658,000 in 2013 (Cornell 2015). According to the Global Gender Gap Report, conducted by the World Economic Forum (2015), Turkey is ranked 105 out of 145 countries in education. However, the gender parity index for women’s participation in tertiary education has gradually increased from 50 percent in 1990 to about 85 percent in 2013 (see figure 1, dashed line). A recent study conducted by Gumus and Chudgar (2016) shows that with the exception of the Northern Turkey, being female has a significant negative impact on school participation. This negative impact is highest for the less developed Eastern region of Turkey and lowest for the more developed Western region (Gumus and Chudgar 2016). Other studies (see El-Sanabary 1993; Rankin and Aytaç 2006; Smits and Hosgör 2006) find that Muslim women and women in general tend to be less educated than their male counterparts. El-Sanabary (1993) points out that girls’ educational and occupational aspirations are often silenced as a result of traditional norms that promote women’s sole duty in marriage and family. Accordingly, investment in girls’ education is discouraged. Figure 1. View largeDownload slide Female labor force participation and gender parity index in tertiary education, Turkey 1990–2014. Source:World Development Indicators (2016). Figure 1. View largeDownload slide Female labor force participation and gender parity index in tertiary education, Turkey 1990–2014. Source:World Development Indicators (2016). Institutional Changes in Familial Patriarchy Institutionally, family law and policies have been patriarchal until recently (Sunar and Fişek 2005). Until 2001, the husband was considered the head of the family. Abolishing the husband as the head of the family and allowing women to keep their maiden name after marriage are among the most important legal changes that impact women’s status in the family (Sunar and Fişek 2005). The old healthcare law positioned women as “dependents” of their husbands or fathers (Şanli 2016). Women were insured under their father’s social security by the government as long as they were not married or employed regardless of age while sons were subjected to age limits, which contributed to existing gender norms. In 2006, the Social Security and Health Care Law limited certain entitlements for women in order to equalize welfare policies, promoting the same treatment of men and women. However, this institutional change worked against women because it was implemented without changing the existing traditional gender ideologies (Şanli 2016). Traditional family roles also contribute to strong son preference in education (Sunar and Fişek 2005). Analyzing father-reported data from 1988, Rankin and Aytaç (2006) find that girls were specifically disadvantaged when fathers preferred their wives to hold domestic roles, and when they lived in sex-segregated households. Examining mother-reported data from a 1998 survey, Smits and Hosgör (2006) demonstrate that girls’ educational attainment was lower when mothers agreed with phrases such as, “important decisions should be made by men” and “it is better for male children to be educated.” Moreover, in addition to many unpaid household duties, the burden of domestic care work rests on women (Şanli 2016). There is no or very little public support for childcare and no systematic policy or incentive to facilitate women’s participation in the formal sector. Institutional Changes in Economic Patriarchy Shortly after the establishment of the Turkish Republic in 1923, several institutional changes encouraged women’s participation in the workforce. However, these reforms remained accessible predominantly to upper-class women. Until the 1990s, married women needed their husband’s approval to work (Dedeoğlu 2012). Although this law was found unconstitutional in 1992, the civil code continued to disadvantage women. For instance, divorcees could only keep the assets acquired in their name (Başpınar 2003; Dedeoğlu 2012). This left women vulnerable because cultural norms dictate men’s ownership of economic gains (Dedeoğlu 2012). Moreover, the strict division of labor by gender has been a structural barrier perpetuating women’s economic inequality. Women are in agricultural and unskilled service jobs in Turkey while men are distributed more equally across the occupational strata (Şanli 2016). The 2003 Labor Law banned discrimination in employment on the basis of gender. However, this did not change the preference of males for traditionally “male” jobs, such as engineering, construction, mining, and even civil service (Şanli 2016). Additionally, women are often asked about their marriage and childbearing intentions during job interviews (Şanli 2016). Formal labor force participation of females has actually decreased from 1990, when almost 35 percent of women over age fifteen participated in the labor force, to about 23 percent in 2005, then rising again to almost 30 percent by 2012 (see figure 1, solid line). However, in addition to unpaid household duties, many economically disadvantaged women participate in the informal sector. Their lack of education is a driving factor for participation in the informal sector in low-paid domestic jobs without benefits (Şanli 2016). In addition, Özyeğin (2001) reports that husbands still exert a great deal of surveillance and control over their wives’ participation in the workforce. Boratav, Fişek, and Eslen-Ziya (2014) find, for example, that husbands may find their wives’ work unproblematic as long as it does not interfere with their household duties. These structural barriers prevent Turkish women from achieving economic equality. Islam’s disapproval of women’s labor in the urban economy is one of the primary reasons for low female workforce participation (Moghadam 2003). The percentage of economically active females in non-Muslim countries is twice that of Muslim countries (Moghadam 2003). Institutional Changes in Political Patriarchy Women achieved their right to vote in 1934, and their right to be elected to public office in 1935. In 1935, seventeen women were elected to the Turkish parliament, but for the next five and a half decades, the number of female parliamentary members remained under twenty (Turkish Statistical Institute 2012). In 1993, the Turkish government elected its first and only female prime minister, Tansu Çiller, at a time when even most developed nations did not have a female head of state (Republic of Turkey Ministry of Foreign Affairs 2017). While the election of Çiller was a profound achievement for its time, women’s participation in politics has been considerably lower compared to men (Turkish Statistical Institute 2012). Parliamentary seats occupied by women increased from twenty-three to seventy-seven from the 1990s to 2015. Females currently hold 79 parliamentary seats in the Grand National Assembly of Turkey compared to 469 seats for males—just below 15 percent (Türkiye Büyük Millet Meclisi 2017). Institutional changes that promote women’s participation in the political arena remain limited. Some scholars argue that the headscarf ban in public institutions in Turkey (a society in which half of the women wear headscarves) dramatically lowered the participation of religious women in political work, in addition to their exclusion from high-status private- and public-sector jobs (Demir 2017). Demir (2017) argues that the amendment of the headscarf law in 2013 should ameliorate religious women’s decades-long exclusion from political, educational, and economic circles. Gender inequality issues have been more visible in Turkish politics since the 1990s. The ratification of CEDAW (Convention on the Elimination of All Forms of Discrimination against Women) drew more attention to gender inequality issues. However, gender inequality in education and employment received more attention compared with political inequality. Other issues experienced by women, such as domestic violence, sexual harassment, and virginity examinations, continue to be a prominent problem in Turkish society. According to Turkey’s International Strategic Research Center, 42 percent of Turkish women experience domestic violence in their lifetimes (Hansen 2013). Many cases of honor killings and forced marriages, especially in low-income areas of Turkey, have also been reported by non-governmental organizations (The World Bank 2012). The global trend toward equality of men and women would predict that most regions of the world have experienced a trend toward a reduced level of patriarchy. However, Turkey appears to counteract this global trend toward a reduction in patriarchy. Thus, in our study, we anticipate that the level of patriarchal attitudes in Turkey has either stayed the same or increased over the study period, 1990–2011 (H1). More recently, women in Turkey have been subject to an even greater backlash in regard to gender issues. Political leaders have publicly stated that they do not believe in the equality of men and women. In 2011, the name of the Women’s Ministry was changed to the Ministry of Family and Social Policies. In 2012, a law was drafted that would ban abortion, although this law was later repealed because of the efforts of women’s rights activists. Public figures have regularly referred to abortion as murder and birth control as a form of treason (Ahmadi 2012). In 2013, the country’s leader claimed that he has a natural right to advise women to have at least three children (Hansen 2013). As late as in 2016, there was an attempt to legalize statutory rape in the Turkish parliament if the offender married the victim (Weise 2016). Moreover, the promotion of sex segregation of boys and girls by the Ministry of Education continues (Ertürk 2016). The new education reform of 2012, also termed “4 + 4 + 4,” changed the structure of the Turkish education system (Cornell 2015; Kamal 2017), which, among other changes, increased the attendance of religious schools (Cornell 2015) and reduced formal schooling of children, particularly girls, in rural areas (Kamal 2017). Institutionalization of gender differences, thus, is associated with increasingly patriarchal attitudes dominating Turkish society. We suspect that patriarchal values in Turkey have increased even more since 2011. However, our data only allow us to examine patriarchal values until 2011. Theoretical Perspectives We employ two theories in an attempt to explain why patriarchal values have been increasing in Turkey. The first is Blalock’s (1967),Power Threat Theory and the second is Stephan and Stephan’s (2000),Integrated Threat Theory. According to Blalock (1967), two main types of threats lead to negative attitudes toward minority populations: competition over economic resources and competition over political power. Similarly, Stephan and Stephan (2000) identify realistic threats (economic, physical, and political) that can produce prejudice among the dominant group toward out-groups. When the dominant group perceives a competition over limited resources, it results in attitudes and behaviors that aim to marginalize minority groups (Stephan and Stephan 2000). While both of these theories focus on the threats that minority groups pose to the dominant group, we can apply them to the case of Turkish women since they may also be perceived as posing a similar threat to the patriarchal system by competing over economic resources and political power. Stephan et al. (2000) apply Integrated Threat Theory to gender, examining the association of specific types of threats (realistic, symbolic, intergroup anxiety, and negative stereotypes) with women’s attitudes toward men. They posit that limited resources result in the conflict of men and women’s interests within the economic and political domains. Tougas et al. (1995) find that the economic threat women pose to men is significantly related to men’s attitudes toward women. Despite the gains in educational outcomes, women experience barriers in the political arena and the labor force (Seguino 2016). While women are closing the educational gap globally, they still lack equal access to employment and government representation. The glass-ceiling that women experience is a result of centuries-old gender norms and hierarchies. While educational equality is the initial barrier to reverse the gender hierarchy, gender reversal in education does not necessarily lead to gender parity in other areas. Accordingly, we expect that the trend in patriarchal attitudes over time depends on the type of patriarchal attitudes. We expect attitudes that are directly related to the societal power structure (economic and political patriarchal attitudes) to have become stronger in Turkey over the study period (H2a). On the other hand, we hypothesize that patriarchal attitudes that are not directly related to the power structures in Turkish society (such as patriarchal attitudes about education) will follow the global trend toward gender equality, and thus show a decline over those twenty years (H2b). Religion, Conservatism, and Patriarchal Attitudes Empirically, the association between religion and patriarchal attitudes has been well-documented in both historic and contemporary societies (Civettini and Glass 2008; Davis and Greenstein 2009; Frejka and Westoff 2008; Goldscheider 2006; Goldscheider, Goldscheider, and Rico-Gonzalez 2014; Klingorova and Havlicek 2015; Pearce and Thornton 2007). For instance, both Frejka and Westoff (2008) and Goldscheider (2006) find that as individuals’ level of religiosity increases, their familial patriarchal values also increase. Similarly, a recent study conducted by Goldscheider, Goldscheider, and Rico-Gonzalez (2014) finds that religiosity increases patriarchal family values among younger Swedish adults. The same study demonstrates that respondents report less patriarchal family values when their religious denomination supports a more egalitarian distribution of gender work balance in the family. Other research also links religiosity with disapproval of divorce, cohabitation, and premarital sex (Goldscheider, Goldscheider, and Rico-Gonzalez 2014; Pearce and Thornton 2007). Drawing on a state-level analysis examining religion’s effect on women’s social, political, and economic status, Klingorova and Havlicek (2015) find a positive association between religious practice and gender inequality. They find that countries where a majority of inhabitants are without a major religion report the least gender inequality. Predominantly Christian and Buddhist societies express average levels of gender inequality, while gender inequality is the highest among predominantly Islamic and Hindu states. Similarly, traditional attitudes toward gender roles have also been linked to political conservatism. Characteristics of radical right parties are associated with conservative family values and traditional gender norms that promote women’s roles as wives and mothers (Buğra 2014; Chong 2006; Inglehart and Norris 2000; Korkut and Eslen-Ziya 2011). Turkey has been influenced by both European values and traditional conservative values. In this paper, we hypothesize that religiosity and conservatism increase patriarchal attitudes in Turkish society (H3 and H4). However, we expect this effect to differ by the type of patriarchal attitude. Drawing from Power Threat Theory (Blalock 1967) and Integrated Threat Theory (Stephan and Stephan 2000), we hypothesize that both religiosity and political conservatism have stronger effects on patriarchal attitudes associated with power at the societal level. Alternatively, we expect the influence of religion and political conservatism on patriarchal attitudes that are not directly related to power at the societal level to be significantly weaker (H3a and H4a). We also test the effect of religiosity and conservatism on patriarchal values over time. Data and Methods We use five waves (1990, 1996, 2001, 2007, and 2011) of the World Values Survey (WVS) data. The data are cross-sectional, meaning that a different set of respondents was interviewed in each wave. The Turkish Statistics Institute conducted this survey using multi-stage full probability sampling, under the supervision of Professor Yılmaz Esmer. Respondents were interviewed regardless of their language, citizenship, and immigration status, though the surveys were all conducted in Turkish. The questionnaire for each wave consists of over 200 closed-ended questions that measure Turkish individuals’ attitudes toward social, political, economic, health, and environmental issues (WVS 2011). A total of 9,289 individuals have been interviewed in Turkey for the WVS over these five waves. In our analysis, we use listwise deletion which reduces the sample size for the regression analyses to the range from 6,190 to 7,171, depending on the dependent variables used. Based on bivariate analysis of the missing data, individuals who are less patriarchal in terms of economic, political, and educational patriarchy were more likely to be in the missing data group. Moreover, individuals who are younger, in a higher income scale, less religious, and less conservative were more likely to be in the missing data group. Females, those who are not married, and those who are not working also have a slightly higher chance of being in the missing data group. Operationalization of Variables We analyze four different types of patriarchal attitudes: economic, familial, political, and educational patriarchy. We initially attempted to combine these four variables into a single index. However, this was not possible due to both low eigenvalues (<1) and factor loadings suggesting that they clearly measure different aspects of patriarchy. That is, those who hold patriarchal values in one aspect (i.e., political patriarchy) do not necessarily hold patriarchal values in another (i.e., educational patriarchy). Both threat theories suggest that competition over economic resources and political power threatens the status quo, and thus results in negative attitudes toward out-groups. Accordingly, we consider political and economic patriarchal attitudes to be directly related to the power structure at the societal level. We assume that educational patriarchal values are not directly related to the power structure at the societal level. Lastly, we consider familial patriarchal attitudes to be associated with power structure in the institution of the family. For our study, economic patriarchy is operationalized by individuals’ level of agreement with the statement: “When jobs are scarce, men should have more right to a job than women.” Political patriarchy is measured by the agreement with the statement: “On the whole men make better political leaders than women do.” Familial patriarchy is operationalized by the agreement with the statement: “Being a housewife is just as fulfilling as working for pay.” Lastly, individuals’ educational patriarchy is operationalized by their level of agreement with the following statement: “A university education is more important for a boy than for a girl.” Response choices for these questions range on a scale from strongly agree (4) to strongly disagree (1). Economic and familial patriarchal attitudes are available for years 1990 through 2011, while political and educational patriarchal values are only available from 1996 to 2011. On a scale from one to four, mean familial and political patriarchy is close to three, and mean educational patriarchy is 1.99. On a scale from one to three, mean economic patriarchy is 2.29 (see table 1 for descriptive statistics). Thus, the patriarchal values are equally strong for family, political, and economic participation, and weakest for education. Table 1 Descriptive statistics   Mean  SD  Minimum  Maximum  Dependent variables  Patriarchy: Economic  2.29  0.90  1  3  Patriarchy: Familial  2.99  0.83  1  4  Patriarchy: Political  2.76  0.94  1  4  Patriarchy: Educational  1.99  0.97  1  4  Independent variables  Religiosity  0.04  0.73  –3.27  0.61  Conservatism  6.01  2.58  1  10  Control variables  Years of education  9.5  3.81  0  16  Employment status           Full-time working (ref)           Part-time working  0.06  0.24  0  1   Not working  0.55  0.50  0  1  Female  0.49  0.50  0  1  Age  37.69  13.80  17  91  Marital status           Married (ref)           Previously married  0.05  0.22  0  1   Single  0.23  0.42  0  1  Number of children  2.08  1.97  0  8  Income levels  3.94  2.19  1  10    Mean  SD  Minimum  Maximum  Dependent variables  Patriarchy: Economic  2.29  0.90  1  3  Patriarchy: Familial  2.99  0.83  1  4  Patriarchy: Political  2.76  0.94  1  4  Patriarchy: Educational  1.99  0.97  1  4  Independent variables  Religiosity  0.04  0.73  –3.27  0.61  Conservatism  6.01  2.58  1  10  Control variables  Years of education  9.5  3.81  0  16  Employment status           Full-time working (ref)           Part-time working  0.06  0.24  0  1   Not working  0.55  0.50  0  1  Female  0.49  0.50  0  1  Age  37.69  13.80  17  91  Marital status           Married (ref)           Previously married  0.05  0.22  0  1   Single  0.23  0.42  0  1  Number of children  2.08  1.97  0  8  Income levels  3.94  2.19  1  10  Source: World Values Survey, 1990–2011. Valid N varies from 6,190 to 7,107. Table 1 Descriptive statistics   Mean  SD  Minimum  Maximum  Dependent variables  Patriarchy: Economic  2.29  0.90  1  3  Patriarchy: Familial  2.99  0.83  1  4  Patriarchy: Political  2.76  0.94  1  4  Patriarchy: Educational  1.99  0.97  1  4  Independent variables  Religiosity  0.04  0.73  –3.27  0.61  Conservatism  6.01  2.58  1  10  Control variables  Years of education  9.5  3.81  0  16  Employment status           Full-time working (ref)           Part-time working  0.06  0.24  0  1   Not working  0.55  0.50  0  1  Female  0.49  0.50  0  1  Age  37.69  13.80  17  91  Marital status           Married (ref)           Previously married  0.05  0.22  0  1   Single  0.23  0.42  0  1  Number of children  2.08  1.97  0  8  Income levels  3.94  2.19  1  10    Mean  SD  Minimum  Maximum  Dependent variables  Patriarchy: Economic  2.29  0.90  1  3  Patriarchy: Familial  2.99  0.83  1  4  Patriarchy: Political  2.76  0.94  1  4  Patriarchy: Educational  1.99  0.97  1  4  Independent variables  Religiosity  0.04  0.73  –3.27  0.61  Conservatism  6.01  2.58  1  10  Control variables  Years of education  9.5  3.81  0  16  Employment status           Full-time working (ref)           Part-time working  0.06  0.24  0  1   Not working  0.55  0.50  0  1  Female  0.49  0.50  0  1  Age  37.69  13.80  17  91  Marital status           Married (ref)           Previously married  0.05  0.22  0  1   Single  0.23  0.42  0  1  Number of children  2.08  1.97  0  8  Income levels  3.94  2.19  1  10  Source: World Values Survey, 1990–2011. Valid N varies from 6,190 to 7,107. The main independent variables are the level of religiosity and conservatism. There are three questions in the data that measure the intensity of respondents’ religious attitudes. These questions ask individuals’ frequency of attending religious services (seven-point scale), how important God is in their life (ten-point scale), as well as how important religion is in their life (three-point scale). Around 35 percent of the respondents attend religious services at least once a week. Over 75 percent of Turkish individuals said that God is very important in their life. In addition, 73 percent of respondents claimed that religion is very important in their life. Using factor analysis, we combined these three religion variables into a single religiosity variable (the eigenvalue from one-factor solution to two-factor solution drops from 1.02 to –0.06). Factor loadings for the three religiosity variables were 0.39, 0.65, and 0.67. The range for the religiosity index is from –3.27 to 0.61, with a mean of 0.04 (standard deviation [SD] 0.73). A further test of reliability (α = 0.64) in addition to similar item-test correlation scores show that the variables fit well in the scale and sufficiently measure the same underlying concept. To measure political conservatism, respondents place their views on political matters on the scale ranging from one (liberal) to ten (conservative). The average level of conservatism is 6.01 (SD = 2.58). All of the analyses are controlled for respondents’ gender, age, educational attainment, employment status, marital status, number of children, and income. The sample consists of 49 percent females. Respondents’ ages range from 17 to 91 years, with a mean of 37.7 years. Educational attainment was initially a categorical variable, operationalized by asking individuals’ highest educational level (nine different levels). The response choices ranged from no formal education to university-level education with degree. We recoded education into years of schooling in order to treat it as a continuous variable. Individuals’ years of education range from zero to sixteen, on average they have 9.6 years of education—equivalent to incomplete secondary education. While the number of children varies from zero to eight, on average respondents have 2.08 children. Income is measured on a subjective scale of one to ten, in which one indicates the lowest income group and ten indicates the highest income group. The mean level of income is 3.95. Categories of employment status (part-time working and not working) are coded as dummy variables, where the reference category is working full-time. About 55 percent of the sample is not working, 6 percent works part-time, and 49 percent works full-time. Marital status is measured by two dummy variables—single and previously married—where the reference category is married. About 72 percent of the sample is currently married, 5 percent are previously married, and 23 percent of the sample is single, never married. Method We use ordered logistic regression to test the effect of religiosity and conservatism on patriarchal attitudes. Including the year of survey as a categorical variable allows us to measure the change in patriarchy over time. It also allows us to model the change over time as a non-linear change. Predicted values from these regressions allow us to plot the trend of patriarchy over time. To investigate the change in the effect of religion and conservatism on patriarchy, we use interaction terms between year dummies and the main independent variable (either religion or conservatism). Again, using year dummies, instead of continuous measure of year, allows us to model the change in the effect of religiosity and conservatism for each time period separately instead of forcing a linear trend. To better understand the interaction effect results, we plot the predicted values from the interaction effects. Findings Table 2 presents the odds ratios from the ordered logistic regression testing the main effects of religiosity and conservatism on economic, familial, political, and educational patriarchy in Turkey from 1990 to 2011. We should note that the reference category for year in the models predicting economic and familial patriarchy is 1990 while it is 1996 for the models predicting political and educational patriarchy since political and educational patriarchy were not measured in 1990. Table 2 Odd ratios from ordered logistic regression: the effect of religiosity and conservatism on patriarchal attitudes in Turkey (1990–2011)   Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)           1996  1.47***  0.90       2001  1.49***  0.89  0.90  0.66***   2007  1.45***  0.77**  0.88†  0.49***   2011  2.28***  0.87  1.74***  1.19*  Independent variables          Religiosity  1.50***  1.33***  1.33***  1.14***  Conservatism  1.10***  1.04***  1.07***  1.06***  Control variables          Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.48***  0.71***  0.65***  0.56***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)           Part-time working  0.98  0.74**  1.12  0.82†   Not working  1.05  1.02  1.03  1.08  Marital status(1)           Single  0.88†  0.95  1.06  1.20*   Previously married  0.94  0.90  0.79*  0.92  Number of children  1.11***  1.04*  1.01  1.06***  Income levels  0.92***  0.96**  0.96***  0.94***  Likelihood ratio χ2  1123.45***  405.56***  562.21***  638.39***   Degrees of freedom  15  15  14  14  Pseudo R2  0.09  0.03  0.04  0.04  Observations  7,107  7,003  6,190  6,301    Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)           1996  1.47***  0.90       2001  1.49***  0.89  0.90  0.66***   2007  1.45***  0.77**  0.88†  0.49***   2011  2.28***  0.87  1.74***  1.19*  Independent variables          Religiosity  1.50***  1.33***  1.33***  1.14***  Conservatism  1.10***  1.04***  1.07***  1.06***  Control variables          Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.48***  0.71***  0.65***  0.56***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)           Part-time working  0.98  0.74**  1.12  0.82†   Not working  1.05  1.02  1.03  1.08  Marital status(1)           Single  0.88†  0.95  1.06  1.20*   Previously married  0.94  0.90  0.79*  0.92  Number of children  1.11***  1.04*  1.01  1.06***  Income levels  0.92***  0.96**  0.96***  0.94***  Likelihood ratio χ2  1123.45***  405.56***  562.21***  638.39***   Degrees of freedom  15  15  14  14  Pseudo R2  0.09  0.03  0.04  0.04  Observations  7,107  7,003  6,190  6,301  * p < 0.05; **p < 0.01; ***p < 0.001; †p < 0.10. Source: World Values Survey, 1990–2011. Note. (1) Reference category for year is 1990 for economic and familial patriarchy, and 1996 for political and educational patriarchy. Reference category for employment status is full time and for marital status it is married. Table 2 Odd ratios from ordered logistic regression: the effect of religiosity and conservatism on patriarchal attitudes in Turkey (1990–2011)   Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)           1996  1.47***  0.90       2001  1.49***  0.89  0.90  0.66***   2007  1.45***  0.77**  0.88†  0.49***   2011  2.28***  0.87  1.74***  1.19*  Independent variables          Religiosity  1.50***  1.33***  1.33***  1.14***  Conservatism  1.10***  1.04***  1.07***  1.06***  Control variables          Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.48***  0.71***  0.65***  0.56***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)           Part-time working  0.98  0.74**  1.12  0.82†   Not working  1.05  1.02  1.03  1.08  Marital status(1)           Single  0.88†  0.95  1.06  1.20*   Previously married  0.94  0.90  0.79*  0.92  Number of children  1.11***  1.04*  1.01  1.06***  Income levels  0.92***  0.96**  0.96***  0.94***  Likelihood ratio χ2  1123.45***  405.56***  562.21***  638.39***   Degrees of freedom  15  15  14  14  Pseudo R2  0.09  0.03  0.04  0.04  Observations  7,107  7,003  6,190  6,301    Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)           1996  1.47***  0.90       2001  1.49***  0.89  0.90  0.66***   2007  1.45***  0.77**  0.88†  0.49***   2011  2.28***  0.87  1.74***  1.19*  Independent variables          Religiosity  1.50***  1.33***  1.33***  1.14***  Conservatism  1.10***  1.04***  1.07***  1.06***  Control variables          Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.48***  0.71***  0.65***  0.56***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)           Part-time working  0.98  0.74**  1.12  0.82†   Not working  1.05  1.02  1.03  1.08  Marital status(1)           Single  0.88†  0.95  1.06  1.20*   Previously married  0.94  0.90  0.79*  0.92  Number of children  1.11***  1.04*  1.01  1.06***  Income levels  0.92***  0.96**  0.96***  0.94***  Likelihood ratio χ2  1123.45***  405.56***  562.21***  638.39***   Degrees of freedom  15  15  14  14  Pseudo R2  0.09  0.03  0.04  0.04  Observations  7,107  7,003  6,190  6,301  * p < 0.05; **p < 0.01; ***p < 0.001; †p < 0.10. Source: World Values Survey, 1990–2011. Note. (1) Reference category for year is 1990 for economic and familial patriarchy, and 1996 for political and educational patriarchy. Reference category for employment status is full time and for marital status it is married. The results show that patriarchal economic values have significantly increased since the 1990s. Compared with 1990, the odds of holding patriarchal economic attitudes are 2.28 times greater in 2011. The increase over this time is highest for attitudes of economic patriarchy. Even after the government’s implementation of the new Labor Law in 2003 in order to ameliorate economic gender inequality, the labor force participation of women is lower in 2011 than in 1990. Individuals also hold stronger political and educational patriarchal attitudes in 2011 than they did in 1990. Compared with 1990, while the odds of holding patriarchal political values were not significantly different in 2001 and 2007, these odds had increased by 74 percent by the year 2011. This is an expected increase considering the lack of institutional reforms or incentives to promote women’s political participation. The odds of possessing educational patriarchal values were actually significantly lower in 2001 and 2007 than in 1990. This is consistent with the 1997 education reform that extended education of children to a compulsory eight years, which had a positive impact on girls’ education. However, these odds had increased by 19 percent as compared with 1990 by the year 2011. Institutional changes in educational patriarchy are helpful in explaining this predicament. The reversal of this progressive trend could be associated with the Turkish government’s conservative education policies, such as a sharp increase in imam-hatip schools since 2002 that disadvantage women and girls. While the odds of holding patriarchal family values were lower in 2007 than in 1990, by 2011, they had increased to their previous 1990 level. Accordingly, the odds of patriarchal familial values in 2011 are not significantly different than they were in 1990 (see table 2). The Turkish government implemented institutional family laws and policies to improve the position of women in society. These include permitting women to be the head of the family, and to keep their maiden name after marriage (both in 2001), and changing women’s “dependent” social security status in 2006 (Sunar and Fişek 2005). However, we find that empirically these institutional changes did not have an impact on familial patriarchal values. Overall, the results demonstrate that economic, political, and educational patriarchal attitudes have increased significantly from 1990 to 2011. Meanwhile, familial patriarchy has not changed significantly during the same time span. For all types of patriarchal attitudes, the increase is most prominent in the most recent years, from 2007 to 2011. Figure 2 presents predicted probabilities of different types of patriarchal attitudes over time based on the models listed in table 2. The graphs indicate that, in general, the largest increase in patriarchal attitudes has occurred in recent years, from 2007 to 2011. Economic patriarchy increased also from 1990 to 1996. Educational patriarchy, on the other hand, decreased from 1996 to 2007. But even educational patriarchy saw an increase from 2007 to 2011. Figure 2. View largeDownload slide Predicted probability of different types of patriarchal attitudes, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N varies from 6,190 to 7,107. Figure 2. View largeDownload slide Predicted probability of different types of patriarchal attitudes, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N varies from 6,190 to 7,107. In addition to indicating the trend of patriarchal attitudes over time, table 2 demonstrates the general effect of independent variables on patriarchal attitudes. Both religiosity and conservatism have a positive association with the four measures of patriarchy. As individuals’ level of religiosity increases, their odds of having patriarchal economic, political, familial, and educational attitudes also increase by 50, 33, 33, and 14 percent, respectively (p<0.001). Similarly, as individuals’ conservatism increases, their odds of expressing economic, political, familial, and educational patriarchal attitudes increase by 10, 4, 7, and 6 percent, respectively (p<0.001). Females, respondents with more years of education, and those with higher income have significantly lower odds of holding patriarchal attitudes. To understand if the effect of religiosity or conservatism has changed over time, we test two sets of interaction effects: (i) between year dummies and religiosity, and (ii) between year dummies and conservatism. Because of the added interaction effects, the coefficient for religiosity now represents the effect of religiosity in 1990 for economic and familial patriarchy and in 1996 for political and educational patriarchy (the earliest measure for these types of patriarchal values). For all of the patriarchal values, religiosity starts out with a strong positive effect: each additional unit of religiosity increases the odds of higher patriarchal values by 27 percent (familial patriarchy) to 53 percent (educational patriarchy). The mainly insignificant interaction effects for familial patriarchy and political patriarchy indicate that the effect of religiosity has not changed much over time for familial and for political patriarchy. The effect of religiosity on economic patriarchy increased from 1990 to 1996, but then again, decreased in 2011. The effect of religiosity on educational patriarchy decreased over time. To better understand the results of the interaction effects, we calculated the predicted probability of strongly agreeing with the patriarchal statements as a function of religiosity over time and graphed it in figure 3. As confirmed from the interpretation above, the effect of religiosity remains the same across time for political patriarchy and for familial patriarchy. Figure 3. View largeDownload slide Predicted probability of strongly agreeing with the patriarchal attitude as a function of religiosity, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N varies from 6,190 to 7,107. Figure 3. View largeDownload slide Predicted probability of strongly agreeing with the patriarchal attitude as a function of religiosity, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N varies from 6,190 to 7,107. However, much more movement over time is seen for economic patriarchy and for educational patriarchy. The effect of religiosity on economic patriarchy becomes sharply stronger from 1990 to 1996 and then starts slowly reducing over time. The effect of religiosity on economic patriarchy in 2011 is significantly weaker than the effect of religiosity in 1990. For educational patriarchy, the effect of religiosity decreased slowly from 1996 to 2007. From predicting the most patriarchal attitude for educational patriarchy (strongly agreeing that university is more important for boys than for girls), religiosity reverses its effect in 2011: this means, as religiosity increases, the probability of strongly agreeing with this statement decreases (the opposite is also true: as religiosity increases, the probability of strongly disagreeing with this statement increases). One reason why the effect of religion on educational and economic patriarchy has been decreasing may be due to the increasing numbers of Islamic schools and institutions in Turkey over the past decade. Eligür (2010) points out that with the election of the Justice and Development Party in 2002, there has been an upsurge in businesses, schools, and other public institutions that are run by Islamists. Thus, individuals may be feeling more comfortable allowing their children to attend and work at these religious institutions, regardless of their gender. Table 3 also presents the results of the interaction effects between year and conservatism. First, these results indicate that conservatism does not really affect familial patriarchy. This confirms that traditional gender norms persist in Turkish society. For all other types of patriarchy, conservatism increases patriarchal attitudes during the earliest year of measurement in this data. However, the effect of conservatism changes over time only for economic patriarchy and not for other types of patriarchy. The effect of conservatism for economic patriarchy is lower in 2001 and 2007 than in 1990. Thus, there seems to be a temporary decrease in the effect of conservatism on economic patriarchy. However, by 2011 the effect of conservatism on economic patriarchy had returned to its 1990 level. Table 3 Odd ratios from ordered logistic regressions with interaction effects: the over-time effect of religiosity and conservatism on patriarchal attitudes in Turkey (1990–2011)   Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)   1996  1.63†  0.96       2001  2.48***  0.78  1.11  0.65*   2007  2.74***  0.78  1.15  0.46***   2011  3.18***  0.65†  1.88**  1.63*  Religiosity  1.45***  1.27**  1.37***  1.53***  Conservatism  1.18***  1.02  1.10***  1.06**  Interaction year × religiosity   1996  1.42**  1.21†       2001  1.11  1.01  0.99  0.83†   2007  0.90  1.18  1.05  0.65**   2011  0.76*  0.95  0.82†  0.44***  Interaction year × conservatism   1996  0.98  0.99       2001  0.91*  1.03  0.97  1.01   2007  0.89*  1.00  0.96  1.02   2011  0.94  1.05  0.98  0.95  Control variables  Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.47***  0.70***  0.64***  0.55***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)   Part-time working  0.97  0.74**  1.12  0.82†   Not working  1.05  1.02  1.04  1.08  Marital status(1)           Single  0.85*  0.95  1.05  1.16*   Previously married  0.96  0.90  0.80*  0.94  Number of children  1.11***  1.04*  1.01  1.06**  Income levels  0.92***  0.96**  0.96**  0.95***  Likelihood ratio χ2  1171.29  414.87  570.39  724.83   Degrees of freedom  23  23  20  20  Pseudo R2  0.09  0.03  0.04  0.05  Observations  7,107  7,003  6,190  6,301    Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)   1996  1.63†  0.96       2001  2.48***  0.78  1.11  0.65*   2007  2.74***  0.78  1.15  0.46***   2011  3.18***  0.65†  1.88**  1.63*  Religiosity  1.45***  1.27**  1.37***  1.53***  Conservatism  1.18***  1.02  1.10***  1.06**  Interaction year × religiosity   1996  1.42**  1.21†       2001  1.11  1.01  0.99  0.83†   2007  0.90  1.18  1.05  0.65**   2011  0.76*  0.95  0.82†  0.44***  Interaction year × conservatism   1996  0.98  0.99       2001  0.91*  1.03  0.97  1.01   2007  0.89*  1.00  0.96  1.02   2011  0.94  1.05  0.98  0.95  Control variables  Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.47***  0.70***  0.64***  0.55***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)   Part-time working  0.97  0.74**  1.12  0.82†   Not working  1.05  1.02  1.04  1.08  Marital status(1)           Single  0.85*  0.95  1.05  1.16*   Previously married  0.96  0.90  0.80*  0.94  Number of children  1.11***  1.04*  1.01  1.06**  Income levels  0.92***  0.96**  0.96**  0.95***  Likelihood ratio χ2  1171.29  414.87  570.39  724.83   Degrees of freedom  23  23  20  20  Pseudo R2  0.09  0.03  0.04  0.05  Observations  7,107  7,003  6,190  6,301  * p < 0.05; **p < 0.01; ***p < 0.001; †p < 0.10. Source: World Values Survey, 1990–2011. Note. (1) Reference category for year is 1990 for economic and familial patriarchy, and 1996 for political and educational patriarchy. Reference category for employment status is full-time, and for marital status it is married. Table 3 Odd ratios from ordered logistic regressions with interaction effects: the over-time effect of religiosity and conservatism on patriarchal attitudes in Turkey (1990–2011)   Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)   1996  1.63†  0.96       2001  2.48***  0.78  1.11  0.65*   2007  2.74***  0.78  1.15  0.46***   2011  3.18***  0.65†  1.88**  1.63*  Religiosity  1.45***  1.27**  1.37***  1.53***  Conservatism  1.18***  1.02  1.10***  1.06**  Interaction year × religiosity   1996  1.42**  1.21†       2001  1.11  1.01  0.99  0.83†   2007  0.90  1.18  1.05  0.65**   2011  0.76*  0.95  0.82†  0.44***  Interaction year × conservatism   1996  0.98  0.99       2001  0.91*  1.03  0.97  1.01   2007  0.89*  1.00  0.96  1.02   2011  0.94  1.05  0.98  0.95  Control variables  Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.47***  0.70***  0.64***  0.55***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)   Part-time working  0.97  0.74**  1.12  0.82†   Not working  1.05  1.02  1.04  1.08  Marital status(1)           Single  0.85*  0.95  1.05  1.16*   Previously married  0.96  0.90  0.80*  0.94  Number of children  1.11***  1.04*  1.01  1.06**  Income levels  0.92***  0.96**  0.96**  0.95***  Likelihood ratio χ2  1171.29  414.87  570.39  724.83   Degrees of freedom  23  23  20  20  Pseudo R2  0.09  0.03  0.04  0.05  Observations  7,107  7,003  6,190  6,301    Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)   1996  1.63†  0.96       2001  2.48***  0.78  1.11  0.65*   2007  2.74***  0.78  1.15  0.46***   2011  3.18***  0.65†  1.88**  1.63*  Religiosity  1.45***  1.27**  1.37***  1.53***  Conservatism  1.18***  1.02  1.10***  1.06**  Interaction year × religiosity   1996  1.42**  1.21†       2001  1.11  1.01  0.99  0.83†   2007  0.90  1.18  1.05  0.65**   2011  0.76*  0.95  0.82†  0.44***  Interaction year × conservatism   1996  0.98  0.99       2001  0.91*  1.03  0.97  1.01   2007  0.89*  1.00  0.96  1.02   2011  0.94  1.05  0.98  0.95  Control variables  Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.47***  0.70***  0.64***  0.55***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)   Part-time working  0.97  0.74**  1.12  0.82†   Not working  1.05  1.02  1.04  1.08  Marital status(1)           Single  0.85*  0.95  1.05  1.16*   Previously married  0.96  0.90  0.80*  0.94  Number of children  1.11***  1.04*  1.01  1.06**  Income levels  0.92***  0.96**  0.96**  0.95***  Likelihood ratio χ2  1171.29  414.87  570.39  724.83   Degrees of freedom  23  23  20  20  Pseudo R2  0.09  0.03  0.04  0.05  Observations  7,107  7,003  6,190  6,301  * p < 0.05; **p < 0.01; ***p < 0.001; †p < 0.10. Source: World Values Survey, 1990–2011. Note. (1) Reference category for year is 1990 for economic and familial patriarchy, and 1996 for political and educational patriarchy. Reference category for employment status is full-time, and for marital status it is married. Figure 2 shows us that over time different types of patriarchy have increased in Turkey (except for familial patriarchy). What contributes to that increase? The interaction effects between year and conservatism showed that the effect of conservatism has not increased (rather, it diminished in some years for economic patriarchy). Similarly, the effect of religiosity has either remained the same or decreased. However, if we look at the overall level of religiosity and conservatism in Turkey, we see that both average religiosity and average conservatism have increased (see figure 4). Figure 4 presents a bivariate graph between year and religiosity and between year and conservatism. Clearly, both religiosity and conservatism overall have increased over this time period (and based on ANOVA and Bonferroni tests the increase is significant). We see a gradual increase in conservatism over time. Starting with 1996, religiosity is higher than in 1990, though there has not been a significant increase since. Instead, there is a slight decrease in the level of religiosity in 2011. Nevertheless, the level of religiosity in 2011 is significantly higher than it was in 1990. Thus, perhaps the overall increase in the level of religiosity and conservatism explain the increase in patriarchal values as there is a clear positive relationship between religiosity and patriarchal attitudes and between conservatism and patriarchal attitudes. Figure 4. View largeDownload slide Mean level of religiosity and conservatism over time, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N ranges from 6,190 to 7,107. Figure 4. View largeDownload slide Mean level of religiosity and conservatism over time, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N ranges from 6,190 to 7,107. Conclusion In this paper, we examine the effect of religiosity and conservatism on patriarchal attitudes in Turkey using the five waves of the WVS (1990, 1996, 2001, 2007, and 2011). Modernization theorists argue that gender equality increases with industrialization and economic development (Inglehart and Norris 2000, 2003; Inglehart, Norris, and Welzel 2002). However, we see that Turkey does not follow this pattern. In fact, patriarchal values have been on the rise. Most Turkish individuals still agree with the statements: “When jobs are scarce, men should have more right to a job than women” (economic patriarchy); “On the whole men make better political leaders than women do” (political patriarchy); “A university education is more important for a boy than for a girl” (educational patriarchy), and “Being a housewife is just as fulfilling as working for pay” (familial patriarchy). Our first hypothesis (H1) is supported: despite Turkey’s economic progress, the level of patriarchal attitudes in Turkey increased in recent years. The increase is highest during the latest years analyzed (2007–2011). Another important question we address is why Turkey does not follow the pattern of Western societies even though it has had a secular, democratically elected government for most of twentieth century. To explain this, we apply Power Threat Theory and Integrated Threat Theory to our findings. We argue that realistic threats (such as political and economic power) posed to the in-group (men) by the out-group (women) influence specific threat perceptions. We find that economic, political, and educational patriarchal attitudes are significantly higher in 2011 than in the previous years of the study period. For all types of patriarchal attitudes, the increase is most prominent in the most recent years, from 2007 to 2011. In fact, educational patriarchy decreased in 2001 and in 2007, but then increased again when measured in 2011. On the other hand, familial patriarchy has not significantly changed since the 1990s. Drawing from Power Threat Theory and Integrated Threat Theory, we hypothesized that patriarchal attitudes that are directly related to the power structure in society (economic and political) would grow stronger in Turkey during the last twenty years and that patriarchal attitudes that are not directly related to societal power structure (familial and educational) would follow the global trend toward equality (H2). Indeed, we see this with educational patriarchy. Educational patriarchy decreased from 1996 to 2001 and from 2001 to 2007. However, these patriarchal attitudes gained strength again from 2007 to 2011 and surpassed the previous levels reported in 1996. Thus, we are able to identify a trend toward more egalitarian attitudes, but then we see a reverse of this progressive trend in more recent years. The reversal of this progressive trend can be explained by the shift in Turkey’s political system. Since 2002, Turkey has been run by a conservative party that has been following an Islamist agenda, aiming to weaken the power of secularism and democracy (Eligür 2010). In our earlier discussion, we also noted that attendance in imam-hatip (religious) schools has increased sharply since 2002. With the launch of the new education reform in 2012, we predict even further increase in educational patriarchy after 2011. Interestingly, the familial patriarchal attitudes have not changed over time. Familial patriarchal attitudes deal with power structures within the family and not necessarily within society. However, family power structure is even more personal than societal-level power structures. Thus, we should expect the least change in attitudes toward family power structure. We find that institutional changes in family law and policy do not impact familial patriarchal values. Neither the global trend toward egalitarianism nor the increase in conservativism has changed the level of familial patriarchal attitudes in Turkey over time. The descriptive findings of patriarchal attitudes in Turkey confirm our historical discussion on the failure of the feminist movement in Turkey (Arat 2000; Kandiyoti 1987). Its lack of a collective voice to integrate individuals from all backgrounds is reflected in the increase in patriarchal values over time. We find a significant and positive effect of both religion and conservatism on patriarchal values. As individuals’ level of religiosity increases, they are much more likely to possess patriarchal attitudes toward economy, family, politics, and education (supporting hypothesis H3). Similarly, as Turkish individuals’ level of political conservatism increases, they are more likely to express patriarchal attitudes (supporting hypothesis H4). The overall effect of conservatism with pooled data across five waves of the study does not vary based on the type of patriarchal attitude (thus, no support for hypothesis H4a). The effect of religion seems to be strongest for economic patriarchy and weakest for educational patriarchy, following our hypothesis H3a: religion’s effect is stronger for patriarchal attitudes associated with the power structure in the society and weaker for patriarchal attitudes that are not directly related to the power structure in the society. Religion’s effect on patriarchal economic and educational attitudes diminished significantly since 1990 and 1996. It is interesting to note that while overall patriarchal educational and economic values were increasing, religion’s effect on these values was diminishing. This may be associated with the contemporary upsurge in Islamic schools and institutions. Religious families may be more inclined to allow their children to attend these institutions. On the other hand, we find that religion’s effect on patriarchal family values has not significantly changed over time. This demonstrates that the traditional view of women’s role in the family remains ingrained in Turkish society. We also find that religion’s effect on patriarchal political values has remained the same (significant and positive). This also confirms our theoretical framework. That is, since political patriarchy is directly related to the power structure, its effect would be the hardest one to reverse. Lastly, we see that the positive effect of conservatism on all forms of patriarchal attitudes has remained the same over time. Patriarchal attitudes play a central role in the continuing gender inequality in economic, familial, political, and educational domains of Turkish society. However, the overall picture of the Turkish case is more complex than a one-size-fits-all solution. We find that modernization theories fail to account for the change in values in Turkey, a society with a predominantly Muslim population that has experienced Westernization and democratic elections. Emancipation of women in Turkey was used as a political agenda and women were treated as tools of modernization and Westernization (Eslen-Ziya and Korkut 2010). In line with Moghadam’s (2003) discussion of neo-patriarchal states, we find that gender equality does not occur as an automatic outcome of economic development. Instead, in neo-patriarchal states such as Turkey “religion is bound to power and state authority; moreover, the family, rather than the individual constitutes the universal building block of the community” (Moghadam 2003, 11). Thus, both the patriarchal family and neo-patriarchal state influence one another, buttressing “normative views of women and the family, often but not exclusively through the law” (Moghadam 2003, 130). Our theoretical framework supports most of our predictions. Future research should specifically explore other potential causes of the recent upsurge in patriarchal attitudes in Turkish society. Also, more research is needed on how religion and political conservatism influence actual gender gaps in education, occupation, employment, and political participation as opposed to attitudes. Future research should explore whether patriarchal attitudes impact ethnic minority women (i.e., Kurdish women), rural, or economically disadvantaged women differently. Finally, our paper focuses on the period from 1990 to 2011. More recent political developments in Turkey have led the country away from a democratically elected government and have stalled its Westernization. Thus, Turkey has become more comparable with other Islamic nations in allowing religion to infuse with governance, limiting political freedoms and decreasing rights for women. Once the data beyond 2011 becomes available, an entirely new look at Turkey is warranted. Notes Ceylan Engin is a doctoral candidate in the Department of Sociology at Texas A&M University. She is an interdisciplinary scholar of sociology, demography, and women’s and gender studies with a specific focus on Turkey. Her current research focuses on the intersections of religion, gender, and sexuality. Heili Pals is an Assistant Professor in the Department of Sociology at Texas A&M University. Her current research focuses on the social psychology of the life course, the social psychological explanations behind the development of deviance, inequality due to sexual orientation, and determinants of attitudes toward homosexuality. She works in collaboration with the Howard B. Kaplan Laboratory for Social Science Research and is an associate editor of the journal Population Review. She has published in journals such as Social Psychology Quarterly, American Journal of Criminal Justice, Deviant Behavior, and Journal of Adolescence. References Ahmadi Arash. 2012. Turkey PM Erdogan sparks row over abortion. BBC News, June 1, accessed July 15, 2015, http://www.bbc.com/news/world-europe-18297760. Aldıkaçtı Marshall Gül. 2005. Ideology, progress, and dialogue. Gender & Society , 19 ( 1): 104– 20. Google Scholar CrossRef Search ADS   Arat Yeşim. 2000. From emancipation to liberation. Journal of International Affairs , 54 ( 1): 107– 23. Arat Yeşim. 2010. Religion, politics and gender equality in Turkey. Third World Quarterly , 31 ( 6): 869– 84. Google Scholar CrossRef Search ADS PubMed  Başpınar Veysel. 2003. Türk Medeni Kanunu ile Aile Hukukunda Yapılan Değişiklikler ve Bu Konuda Bazı Önerilerimiz. AHFD , 52 ( 3): 79– 101. Blalock Hubert M. 1967. Toward a theory of minority-group relations . New York: Capricorn Books. Boratav Hande, Fişek Güler, Eslen-Ziya Hande. 2014. Unpacking masculinities in the context of social change. Men and Masculinities , 17 ( 3): 299– 324. Google Scholar CrossRef Search ADS   Buğra Ayşe. 2014. Revisiting the Wollstonecraft dilemma in the context of conservative liberalism. Social Politics , 21 ( 1): 148– 66. Google Scholar CrossRef Search ADS   Caner Asena, Güven Cahit, Ökten Çağla, Sakallı Seyhun Orcan. 2016. Gender roles and the education gender gap in Turkey. Social Indicators Research , 129: 1231– 1254. Google Scholar CrossRef Search ADS   Chong Kelly H. 2006. Negotiating patriarchy. Gender & Society , 20 ( 6): 697– 724. http://dx.doi.org/10.1177/0891243206291111 Google Scholar CrossRef Search ADS   Civettini Nicole H.W, Glass Jennifer. 2008. The impact of religious conservatism on men's work and family environment. Gender & Society , 22 ( 2): 172– 93. http://dx.doi.org/10.1177/0891243207310714 Google Scholar CrossRef Search ADS   Cornell Svante E. 2015. The Islamization of Turkey. The Turkey Analyst, September 2, accessed July 3, 2017, https://www.turkeyanalyst.org/publications/turkey-analyst-articles/item/437-the-islamization-of-turkey-erdo%C4%9Fan%E2%80%99s-education-reforms.html. Davis Shannon N., Greenstein Theodore N.. 2009. Gender ideology. Annual Review of Sociology , 35: 87– 105. Google Scholar CrossRef Search ADS   Dedeoğlu Saniye. 2012. Equality, protection or discrimination. Social Politics , 19 ( 2): 269– 90. Google Scholar CrossRef Search ADS   Demir Zekiye. 2017. Women’s participation in politics and strategies to raise the presence of women within politics, accessed July 3, 2017, http://kadinarastirmalari.kadem.org.tr/kadinlarin-siyasete-katilimi-ve-katilimi-artirmaya-yonelik-stratejiler/. Diner Cagla, Toktaş Şule. 2010. Waves of feminism in Turkey: Kemalist, Islamist and Kurdish women's movements in an era of globalization. Journal of Balkan and Near Eastern Studies , 12 ( 1): 41– 57. Google Scholar CrossRef Search ADS   Dulger Ilhan. 2004. Turkey: rapid coverage for compulsory education. The World Bank, accessed July 3, 2017, http://siteresources.worldbank.org/INTTURKEY/Resources/361616-1142415001082/Compulsory_Education_by_Dulger.pdf. Eligür Banu. 2010. The mobilization of political Islam in Turkey . New York: Cambridge University Press. Google Scholar CrossRef Search ADS   El-Sanabary Nagat. 1993. Middle East and North Africa. In Women’s education in developing countries: Barriers, benefits, and policies , ed. King E., Hill M., 136– 74. Baltimore, MD: Johns Hopkins University Press. Erhart Itır, Eslen-Ziya Hande. 2014. Janissary: An orientalist heroine or a role model for Muslim Women. In Heroines of comic books and literature , ed. Bajac-Carter Maja, Jones Norma, Batchelor Bob, 95– 106. Lanham, MD: Rowman & Littlefield. Eslen-Ziya Hande, Korkut Umut. 2010. Political religion and politicized women in Turkey. Totalitarian Movements and Political Religions , 11 ( 3–4): 311– 26. Google Scholar CrossRef Search ADS   Esmer Yılmaz. 2008. Islam, gender, democracy and values. In Changing values, persisting cultures , ed. Petterson Thorleif, Esmer Yılmaz, 283– 99. Leiden and Boston: Brill. Google Scholar CrossRef Search ADS   Etürk Ali Ekber. 2016. Karma eğitime darbe hazırlığı. Sözcü. June 1, accessed July 3, 2017, http://www.sozcu.com.tr/2016/egitim/karma-egitime-darbe-hazirligi-1256364/. Frejka Tomas, Westoff Charles F.. 2008. Religion, religiousness and fertility in the US and in Europe. European Journal of Population , 24: 5– 31. Google Scholar CrossRef Search ADS   Goldscheider Calvin. 2006. Religion, family, and fertility. In Religion and the decline of fertility in the Western world , ed. Derosas R., van Poppel F., 41– 57. Dordrecht, Netherlands: Springer. Google Scholar CrossRef Search ADS   Goldscheider Frances, Goldscheider Calvin, Rico-Gonzalez Antonio. 2014. Gender equality in Sweden. Journal of Family Issues , 35 ( 7): 892– 908. Google Scholar CrossRef Search ADS   Gumus Sedat, Chudgar Amita. 2016. Factors affecting school participation in Turkey: an analysis of regional differences. Compare: A Journal of Comparative and International Education , 46 ( 6): 929– 51. http://dx.doi.org/10.1080/03057925.2015.1095073 Google Scholar CrossRef Search ADS   Hansen Suzy. 2013. Turkey’s women strike back. New York Review, June 16, accessed July 15, 2015, http://www.nybooks.com/daily/2013/08/19/turkeys-women-strike-back/. Inglehart Ronald, Norris Pippa. 2000. The developmental theory of the gender gap. International Political Science Review , 21 ( 4): 441– 63. http://dx.doi.org/10.1177/0192512100214007 Google Scholar CrossRef Search ADS   Inglehart Ronald, Norris Pippa. 2003. Rising Tide . New York and Cambridge: Cambridge University Press. Google Scholar CrossRef Search ADS   Inglehart Ronald, Norris Pippa, and Christian Welzel. 2002. Gender equality and democracy. Comparative Sociology , 1 ( 3/4): 321– 46. Google Scholar CrossRef Search ADS   İslam Merve Kavakçı. 2010. Headscarf politics . New York: Palgrave Macmillan. Google Scholar CrossRef Search ADS   Kamal Kevin. 2017. Education in Turkey. World Education News and Reviews, April 4, accessed July 3, 2017, http://wenr.wes.org/2017/04/education-in-turkey. Kandiyoti Deniz. 1987. Emancipated but unliberated? Feminist Studies , 13 ( 2): 317– 38. Google Scholar CrossRef Search ADS   Kandiyoti Deniz. 1988. Bargaining with patriarchy. Gender & Society , 2 ( 3): 274– 90. http://dx.doi.org/10.1177/089124388002003004 Google Scholar CrossRef Search ADS   Klingorova Kamila, Havlicek Tomáš. 2015. Religion and gender inequality. Moravian Geographical Reports , 23 ( 2): 2– 11. Korkut Umut, Eslen-Ziya Hande. 2011. The impact of conservative discourses in family policies, population politics, and gender rights in Poland and Turkey. Social Politics , 18 ( 3): 387– 418. http://dx.doi.org/10.1093/sp/jxr014 Google Scholar CrossRef Search ADS PubMed  Moghadam Valentine. 2003. Modernizing women: Gender and social change in the Middle East . Boulder, CO: Lynne Rienner Publishers. Özyeğin Gül. 2001. Untidy gender . Philadelphia: Temple University Press. Pearce D. Lisa, Thornton Arland. 2007. Religious identity and family ideologies in the transition to adulthood. Journal of Marriage and Family , 69: 1227– 1243. http://dx.doi.org/10.1111/j.1741-3737.2007.00443.x Google Scholar CrossRef Search ADS   Pope Nicole, Pope Hugh. 2011. Turkey unveiled . New York: The Overlook Press. Rankin H. Bruce, Aytaç Işık A.. 2006. Gender inequality in schooling. Sociology of Education , 79 ( 1): 25– 43. Google Scholar CrossRef Search ADS   Republic of Turkey Ministry of Foreign Affairs. 2017. Sayın Prof. Dr.Tansu Çiller'in Özgeçmişi (Background of Dr. Tansu Ciller), accessed July 3, 2017, http://www.mfa.gov.tr/sayin-prof_dr_tansu-ciller_in-ozgecmisi_-_haziran-1996_.tr.mfa. Şanli Şölen. 2016. Women and cultural citizenship in Turkey . New York: I.B. Tauris & Co. Ltd. Saktanber Ayşe. 2002. Living Islam . New York: I.B. Tauris & Co. Ltd. Saktanber Ayşe, Çorbacıoğlu Gül. 2008. Veiling and headscarf-skepticism in Turkey. Social Politics , 15 ( 4): 514– 38. Google Scholar CrossRef Search ADS   Seguino Stephanie. 2016. Global trends in gender equality. Journal of African Development , 18: 1– 30. Smits Jereon, Hosgör Ayse Gündüz. 2006. Effects of family background characteristics on educational participation in Turkey. International Journal of Educational Development , 26: 545– 60. Google Scholar CrossRef Search ADS   Stephan G. Walter, Stephan Cookie W.. 2000. An integrated threat theory of prejudice. In Reducing prejudice and discrimination , ed. Oskamp S.. 23– 46. Mahwah, NJ: Lawrence Erlbaum. Stephan Cookie White, Stephan Walter G., Demitrakis Katherine M., Yamada Ann Marie, Clason Dennis L.. 2000. Women's attitudes toward men. Psychology of Women Quarterly , 24 ( 1): 63– 73. Google Scholar CrossRef Search ADS   Sunar Diane, Fişek Güler Okman. 2005. Contemporary Turkish families. In Families in global perspective , ed. Gielen U., Roopnarine J., 169– 83. Boston, MA: Allyn and Bacon/Pearson. Tougas Francine, Brown Rupert, Beaton Ann M., Joly Stéphane. 1995. Neosexism: Plus ça change, plus c'est pareil. Personality and Social Psychology Bulletin , 21: 842– 49. Google Scholar CrossRef Search ADS   Turkiye İstatislik Kurumu (Turkish Statistical Institute). 2012. Milletvekili genel seçimleri 1923-2011 (Deputy General Elections 1923-2011) . Ankara: TUIK. Türkiye Büyük Millet Meclisi (TBMM). 2017. Türkiye büyük millet meclisi milletvekilleri dağılımı [List of deputies of the grand national assembly of Turkey], accessed June 25, 2017, https://www.tbmm.gov.tr/develop/owa/milletvekillerimiz_sd.dagilim. Weise Zia. 2016. Turkey bill to pardon statutory rape if perpetrators marry victim. The Telegraph, November 18, accessed July 3, 2017, http://www.telegraph.co.uk/news/2016/11/18/turkey-bill-to-pardon-statutory-rape-if-perpetrators-marry-victi/. World Bank. 2012. World development report 2012 gender equality and development, accessed September 5, 2015, https://openknowledge.worldbank.org/handle/10986/4391. World Development Indicators. 2016. Labor force participation rate, female (% of female population ages 15+) (Modeled ILO Estimate), accessed May 15, 2016, http://data.worldbank.org/indicator/SL.TLF.CACT.FE.ZS?page=5. World Economic Forum. 2015. The global gender gap report, accessed December 12, 2015, http://reports.weforum.org/global-gender-gap-report-2015/economies/#economy=TUR. World Values Survey. 2017. Fieldwork and sampling, accessed July 5, 2017, http://www.worldvaluessurvey.org/WVSContents.jsp?CMSID=FieldworkSampling. © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Social Politics Oxford University Press

Patriarchal Attitudes in Turkey 1990–2011: The Influence of Religion and Political Conservatism

Social Politics , Volume Advance Article – Jan 23, 2018

Loading next page...
 
/lp/ou_press/patriarchal-attitudes-in-turkey-1990-2011-the-influence-of-religion-qLea64YUu7
Publisher
Social Politics LLC
Copyright
© The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com
ISSN
1072-4745
eISSN
1468-2893
D.O.I.
10.1093/sp/jxx021
Publisher site
See Article on Publisher Site

Abstract

Abstract While contemporary theories associate gender equality with the process of modernization and economic development, we find that Turkey does not follow this pattern. We investigate changes in patriarchal attitudes from 1990 to 2011 in Turkey, and how the country has been influenced by religiosity and political conservatism. By studying how institutional changes affect gender inequality in Turkey, we find that Turkey has become more patriarchal since the 1990s. Moreover, the effect of religion on patriarchal attitudes has diminished over time while the effect of political conservatism has remained stable. The effect of religion is strongest for economic and weakest for educational patriarchy, while the effect of political conservatism does not vary. Introduction According to The Global Gender Gap Report (World Economic Forum 2015), gender inequality is highest in the Middle East, Africa, and South Asia while Scandinavian countries have the smallest gender gap in income, education, labor force, and political participation. Modernization scholars argue that a global trend toward gender equality is closely related to economic development and democratization of societies (Esmer 2008; Inglehart and Norris 2000, 2003; Inglehart, Norris, and Welzel 2002). They argue that the process of modernization results in cultural and normative changes that promote both democratization and an increase in the proportion of women in public life. For instance, high-income, economically advanced countries have a higher percentage of women in parliament and more egalitarian values of gender than low-income countries (Inglehart and Norris 2003). While contemporary theories may elucidate the revolutionary change in gender values in the Western World, we feel that they fail to explain gender value change in Middle Eastern societies. We find that the trend of increasing gender equality coterminous with the process of industrialization may not be applicable to Middle Eastern regions even when countries are governed through democratic elections. Turkey, a country with a predominantly Muslim population and, throughout the twentieth century, a secular, democratically elected government, is a preeminent example of this. While Turkey’s economic development has been on the rise since the turn of the century, women’s inequality has lagged in every measure (Hansen 2013). In 2006, Turkey was ranked 105th among 115 countries in terms of gender inequality (the 91st percentile). Turkey is ranked 130th among 145 countries in the most current report in 2015 (the 89th percentile)—with Iceland and Norway in the top two positions, and Yemen and Pakistan in the bottom two positions. We know very little about changes in patriarchal values over time in predominantly Islamic societies. Modern Turkey is not comparable with other Islamic nations because of its Westernization and its movement toward secularization since its establishment. Nevertheless, an overwhelming majority of the Turkish population are devout Muslims who practice religion on a regular basis (Esmer 2008). Despite its historical differences with much of the rest of the Islamic world, by studying patriarchal attitudes in Turkey, we can test theories about attitudes toward gender issues in other non-Western societies. Previous research finds that both religiosity and political conservatism have a negative impact on gender inequality (Civettini and Glass 2008; Davis and Greenstein 2009; Frejka and Westoff 2008). In this paper, we test the effect of two types of conservatism, political and religious, on attitudes related to gender. Our first research question asks, how have patriarchal attitudes in Turkish society changed from 1990 to 2011? Second, how does religiosity impact individuals’ patriarchal values? Lastly, how does political conservatism influence patriarchal attitudes in Turkey? We first provide a historical discussion of the feminist movement in Turkey. We then explicate institutional changes in four spheres of patriarchal values (economic, political, familial, and educational) that affect gender inequality in contemporary Turkey. Applying Integrated Threat Theory, we argue that different types of patriarchal values do not share the same trajectory over time and are influenced differently by religiosity and political conservatism. To examine the effect of religion and political conservatism on patriarchal values, we use data from five waves of the World Values Survey from 1990 to 2011 (before the current dramatic changes in Turkish governance). The World Values Survey (2017) measures the changing values, beliefs, and attitudes about the 3 social, economic, and political life of over 9,000 respondents in Turkey. Applying institutionalization of gender in the context of Turkey, we hypothesize that levels of patriarchal attitudes have been increasing in Turkey in that period. We argue that both religiosity and political conservatism are positively associated with patriarchal values. More specifically, we focus on four types of patriarchal attitudes: political, economic, familial, and educational. We posit that, based on Integrated Threat Theory, the effects of religion and political conservatism’s on political and economic gender attitudes are stronger than the effects of educational and familial patriarchal values. Moreover, our discussion on the institutionalization of economic, political, familial, and educational patriarchy complements our findings. In the next section, we will offer a brief history of women’s rights in Turkey, followed by a discussion of the institutionalization of gender, and how it impacts women’s educational, economic, familial, and political outcomes. Feminist Movements in Turkey Turkey was established in 1923 after the fall of the Ottoman Empire as a secular democratic state. Under the leadership of Mustafa Kemal Atatürk, Turkey underwent drastic economic, political, and social changes between 1920 and 1930 that aimed to modernize Turkish society and remove the power of religion from state institutions. These modernization efforts, also referred to as the Kemalist movement, involved adopting a multi-party system which abandoned the sultanate and the caliphate system; implementing secular public education; adopting the Latin alphabet; changing the official national language to Turkish; and promoting European dress codes. Religious attire such as headscarves and fezzes were first discouraged and later prohibited in government institutions (Eligür 2010; Pope and Pope 2011; Saktanber 2002; Saktanber and Çorbacıoğlu 2008). Improving the position of women in society was an explicit objective of the newly formed Turkish Republic. Between 1930 and 1934, the parliament implemented policies that created unprecedented civil rights for women, including the right to vote and the right to be elected to public office. This took place at a time when women in most Western and non-Western countries could not vote. Under the Civil Code of 1926, polygyny was banned, and women were given equal access to divorce and matters related to inheritance and property management. Women’s entrance into the public sphere was also facilitated by expanding educational opportunities and promoting their participation in the work force (Pope and Pope 2011). The goal of these institutional gender reforms was to remove Islam from public affairs through state control (Eligür 2010; Pope and Pope 2011). However, Islam remained an active force in Turkish culture, ingrained in most of the country’s social institutions (Saktanber 2002). The secularization experiment in Turkey was a drastic top-down change and the emancipation of women was revolutionary for its time. Emancipation of women in Turkey was dissimilar to women’s liberation movements in other developing countries (Erhart and Eslen-Ziya 2014; Eslen-Ziya and Korkut 2010). Some scholars argue that women’s emancipation played a strategic role in the Turkish government claiming its rightful place among other Western democratic nations (Kandiyoti 1987; Saktanber and Çorbacıoğlu 2008). However, Turkey did not have a strong women’s rights movement. The first wave of feminism was limited to a small number of women’s organizations initiated by educated and privileged women and did not recognize the plurality of experiences of women (Kandiyoti 1987). Instead, rights for women were granted by enlightened governing elite men and thus, emancipation efforts and reforms remained limited to elite Turkish bourgeoisie women while excluding the vast majority of rural and poor women at the margins of society (Kandiyoti 1987). Religious marriages, polygamy, the demand for başlık parası (bride-price), child marriages, virginity examinations, honor killings, and denial of girls’ rights to education remained continuing signs of gender inequality in Turkey (Kandiyoti 1988). The reforms in the newly formed country left crucial issues such as male privilege, heteronormativity, and traditional gender norms untouched. In this sense, women’s status was emancipated but unliberated, especially because of the absence of an autonomous women’s movement with consciousness-raising outreach efforts. The majority of women in Turkey remained oppressed, subordinated, and marginalized after the secularization experiment (Arat 2000). The first wave of feminism ended after the government closed down the Turkish Women’s Union in 1935. The Turkish Republican elite propagated the myth that, because the government granted equal legal and political rights with men, women’s organizations were no longer needed (Arat 2000). It was not until the 1980s that the feminist movement in Turkey brought up issues similar to the second wave of feminism in the West. These issues included sexual freedom, the oppression women experience in the family, and the elimination of violence against women (Aldıkaçtı Marshall 2005; Diner and Toktaş 2010). A new wave of feminism was initiated in the 1990s when Islamist, Kurdish nationalist, and LGBT movements began to challenge the earlier waves of feminism with the rise of identity politics. Kurdish women criticized the dominant Turkish feminist movement for othering them and ignoring issues related to the Kurdish conflict (Arat 2000). Activist Islamic women demanded more religious freedom by protesting the headscarf ban in government institutions (Aldıkaçtı Marshall 2005; İslam 2010; Saktanber and Çorbacıoğlu 2008). Islamic women were marginalized both by their male counterparts who supported political Islam and by the mainstream feminist movement. Islamic men stressed that Islam treats women equally and that no further questioning was necessary. Similarly, the dominant feminist discourse excluded Islamic feminists because they perceived Islam as a patriarchal religion that promoted the subservience and oppression of women (Erhart and Eslen-Ziya 2014). Some feminists, however, developed a more critical approach to Islamic feminism, which endorsed social agency in defining freedom and supported Islamic women’s right to cover their heads (Diner and Toktaş 2010). Overall, even though the three waves of feminist movements in Turkey have had the same goal of improving the position of women in society, they failed to coalesce into a unified voice. Institutional Changes in Patriarchal Values Affecting Gender Inequality in Contemporary Turkey Despite the laws and regulations implemented by Turkey to improve the rights and position of women in the past, gender inequality and a patriarchal value system continue to be a persistent problem. Turkey through the twentieth century was recognized as a modern society that embraced a neoliberal economic model and moderate Muslim values (Hansen 2013). However, while economic development was on the rise, the status of women lagged behind and the regulation of the private lives of women increased (Hansen 2013). Significant gender inequality persists in education, employment, and politics in Turkey (Arat 2010; Dedeoğlu 2012; The World Bank 2012). In the next section, we will discuss the institutional changes in four spheres of patriarchy—educational, familial, economic, and political—and how they influence the status of women. Institutional Changes in Educational Patriarchy Until the late 1990s, Turkey’s education system included a five-year compulsory primary education, followed by a non-compulsory three-year secondary and a three-year high school education (Caner et al. 2016). In 1997, the Turkish Ministry of National Education along with the Turkish Higher Education Council implemented an education reform that integrated primary and secondary schooling, extending primary education to a compulsory eight years (Caner et al. 2016). This reform increased the school enrollment of children from 86 percent in 1997 to 96 percent in 2002 (Dulger 2004). This institutionalization of education made an impressive impact on girls’ enrollment, particularly in rural areas. In the first year of the program, enrollment of girls in rural areas increased by 160 percent (Dulger 2004). Before this reform, parents were free to enroll their children in imam-hatip schools (religious schools that provide trainings for imams and preachers) after the five-year primary schooling (Kamal 2017). Girls composed almost half of the enrollments of imam-hatip schools albeit they could become neither imams nor hatips. The 1997 reform prevented children from enrolling in imam-hatip schools until the age of fourteen (Kamal 2017), and thus decreased the number of religious schools and their attendees. After the election of the Justice and Development Party in 2002, there has been an exponential increase in imam-hatip schools. Enrollees increased from 65,000 in 2002 to 658,000 in 2013 (Cornell 2015). According to the Global Gender Gap Report, conducted by the World Economic Forum (2015), Turkey is ranked 105 out of 145 countries in education. However, the gender parity index for women’s participation in tertiary education has gradually increased from 50 percent in 1990 to about 85 percent in 2013 (see figure 1, dashed line). A recent study conducted by Gumus and Chudgar (2016) shows that with the exception of the Northern Turkey, being female has a significant negative impact on school participation. This negative impact is highest for the less developed Eastern region of Turkey and lowest for the more developed Western region (Gumus and Chudgar 2016). Other studies (see El-Sanabary 1993; Rankin and Aytaç 2006; Smits and Hosgör 2006) find that Muslim women and women in general tend to be less educated than their male counterparts. El-Sanabary (1993) points out that girls’ educational and occupational aspirations are often silenced as a result of traditional norms that promote women’s sole duty in marriage and family. Accordingly, investment in girls’ education is discouraged. Figure 1. View largeDownload slide Female labor force participation and gender parity index in tertiary education, Turkey 1990–2014. Source:World Development Indicators (2016). Figure 1. View largeDownload slide Female labor force participation and gender parity index in tertiary education, Turkey 1990–2014. Source:World Development Indicators (2016). Institutional Changes in Familial Patriarchy Institutionally, family law and policies have been patriarchal until recently (Sunar and Fişek 2005). Until 2001, the husband was considered the head of the family. Abolishing the husband as the head of the family and allowing women to keep their maiden name after marriage are among the most important legal changes that impact women’s status in the family (Sunar and Fişek 2005). The old healthcare law positioned women as “dependents” of their husbands or fathers (Şanli 2016). Women were insured under their father’s social security by the government as long as they were not married or employed regardless of age while sons were subjected to age limits, which contributed to existing gender norms. In 2006, the Social Security and Health Care Law limited certain entitlements for women in order to equalize welfare policies, promoting the same treatment of men and women. However, this institutional change worked against women because it was implemented without changing the existing traditional gender ideologies (Şanli 2016). Traditional family roles also contribute to strong son preference in education (Sunar and Fişek 2005). Analyzing father-reported data from 1988, Rankin and Aytaç (2006) find that girls were specifically disadvantaged when fathers preferred their wives to hold domestic roles, and when they lived in sex-segregated households. Examining mother-reported data from a 1998 survey, Smits and Hosgör (2006) demonstrate that girls’ educational attainment was lower when mothers agreed with phrases such as, “important decisions should be made by men” and “it is better for male children to be educated.” Moreover, in addition to many unpaid household duties, the burden of domestic care work rests on women (Şanli 2016). There is no or very little public support for childcare and no systematic policy or incentive to facilitate women’s participation in the formal sector. Institutional Changes in Economic Patriarchy Shortly after the establishment of the Turkish Republic in 1923, several institutional changes encouraged women’s participation in the workforce. However, these reforms remained accessible predominantly to upper-class women. Until the 1990s, married women needed their husband’s approval to work (Dedeoğlu 2012). Although this law was found unconstitutional in 1992, the civil code continued to disadvantage women. For instance, divorcees could only keep the assets acquired in their name (Başpınar 2003; Dedeoğlu 2012). This left women vulnerable because cultural norms dictate men’s ownership of economic gains (Dedeoğlu 2012). Moreover, the strict division of labor by gender has been a structural barrier perpetuating women’s economic inequality. Women are in agricultural and unskilled service jobs in Turkey while men are distributed more equally across the occupational strata (Şanli 2016). The 2003 Labor Law banned discrimination in employment on the basis of gender. However, this did not change the preference of males for traditionally “male” jobs, such as engineering, construction, mining, and even civil service (Şanli 2016). Additionally, women are often asked about their marriage and childbearing intentions during job interviews (Şanli 2016). Formal labor force participation of females has actually decreased from 1990, when almost 35 percent of women over age fifteen participated in the labor force, to about 23 percent in 2005, then rising again to almost 30 percent by 2012 (see figure 1, solid line). However, in addition to unpaid household duties, many economically disadvantaged women participate in the informal sector. Their lack of education is a driving factor for participation in the informal sector in low-paid domestic jobs without benefits (Şanli 2016). In addition, Özyeğin (2001) reports that husbands still exert a great deal of surveillance and control over their wives’ participation in the workforce. Boratav, Fişek, and Eslen-Ziya (2014) find, for example, that husbands may find their wives’ work unproblematic as long as it does not interfere with their household duties. These structural barriers prevent Turkish women from achieving economic equality. Islam’s disapproval of women’s labor in the urban economy is one of the primary reasons for low female workforce participation (Moghadam 2003). The percentage of economically active females in non-Muslim countries is twice that of Muslim countries (Moghadam 2003). Institutional Changes in Political Patriarchy Women achieved their right to vote in 1934, and their right to be elected to public office in 1935. In 1935, seventeen women were elected to the Turkish parliament, but for the next five and a half decades, the number of female parliamentary members remained under twenty (Turkish Statistical Institute 2012). In 1993, the Turkish government elected its first and only female prime minister, Tansu Çiller, at a time when even most developed nations did not have a female head of state (Republic of Turkey Ministry of Foreign Affairs 2017). While the election of Çiller was a profound achievement for its time, women’s participation in politics has been considerably lower compared to men (Turkish Statistical Institute 2012). Parliamentary seats occupied by women increased from twenty-three to seventy-seven from the 1990s to 2015. Females currently hold 79 parliamentary seats in the Grand National Assembly of Turkey compared to 469 seats for males—just below 15 percent (Türkiye Büyük Millet Meclisi 2017). Institutional changes that promote women’s participation in the political arena remain limited. Some scholars argue that the headscarf ban in public institutions in Turkey (a society in which half of the women wear headscarves) dramatically lowered the participation of religious women in political work, in addition to their exclusion from high-status private- and public-sector jobs (Demir 2017). Demir (2017) argues that the amendment of the headscarf law in 2013 should ameliorate religious women’s decades-long exclusion from political, educational, and economic circles. Gender inequality issues have been more visible in Turkish politics since the 1990s. The ratification of CEDAW (Convention on the Elimination of All Forms of Discrimination against Women) drew more attention to gender inequality issues. However, gender inequality in education and employment received more attention compared with political inequality. Other issues experienced by women, such as domestic violence, sexual harassment, and virginity examinations, continue to be a prominent problem in Turkish society. According to Turkey’s International Strategic Research Center, 42 percent of Turkish women experience domestic violence in their lifetimes (Hansen 2013). Many cases of honor killings and forced marriages, especially in low-income areas of Turkey, have also been reported by non-governmental organizations (The World Bank 2012). The global trend toward equality of men and women would predict that most regions of the world have experienced a trend toward a reduced level of patriarchy. However, Turkey appears to counteract this global trend toward a reduction in patriarchy. Thus, in our study, we anticipate that the level of patriarchal attitudes in Turkey has either stayed the same or increased over the study period, 1990–2011 (H1). More recently, women in Turkey have been subject to an even greater backlash in regard to gender issues. Political leaders have publicly stated that they do not believe in the equality of men and women. In 2011, the name of the Women’s Ministry was changed to the Ministry of Family and Social Policies. In 2012, a law was drafted that would ban abortion, although this law was later repealed because of the efforts of women’s rights activists. Public figures have regularly referred to abortion as murder and birth control as a form of treason (Ahmadi 2012). In 2013, the country’s leader claimed that he has a natural right to advise women to have at least three children (Hansen 2013). As late as in 2016, there was an attempt to legalize statutory rape in the Turkish parliament if the offender married the victim (Weise 2016). Moreover, the promotion of sex segregation of boys and girls by the Ministry of Education continues (Ertürk 2016). The new education reform of 2012, also termed “4 + 4 + 4,” changed the structure of the Turkish education system (Cornell 2015; Kamal 2017), which, among other changes, increased the attendance of religious schools (Cornell 2015) and reduced formal schooling of children, particularly girls, in rural areas (Kamal 2017). Institutionalization of gender differences, thus, is associated with increasingly patriarchal attitudes dominating Turkish society. We suspect that patriarchal values in Turkey have increased even more since 2011. However, our data only allow us to examine patriarchal values until 2011. Theoretical Perspectives We employ two theories in an attempt to explain why patriarchal values have been increasing in Turkey. The first is Blalock’s (1967),Power Threat Theory and the second is Stephan and Stephan’s (2000),Integrated Threat Theory. According to Blalock (1967), two main types of threats lead to negative attitudes toward minority populations: competition over economic resources and competition over political power. Similarly, Stephan and Stephan (2000) identify realistic threats (economic, physical, and political) that can produce prejudice among the dominant group toward out-groups. When the dominant group perceives a competition over limited resources, it results in attitudes and behaviors that aim to marginalize minority groups (Stephan and Stephan 2000). While both of these theories focus on the threats that minority groups pose to the dominant group, we can apply them to the case of Turkish women since they may also be perceived as posing a similar threat to the patriarchal system by competing over economic resources and political power. Stephan et al. (2000) apply Integrated Threat Theory to gender, examining the association of specific types of threats (realistic, symbolic, intergroup anxiety, and negative stereotypes) with women’s attitudes toward men. They posit that limited resources result in the conflict of men and women’s interests within the economic and political domains. Tougas et al. (1995) find that the economic threat women pose to men is significantly related to men’s attitudes toward women. Despite the gains in educational outcomes, women experience barriers in the political arena and the labor force (Seguino 2016). While women are closing the educational gap globally, they still lack equal access to employment and government representation. The glass-ceiling that women experience is a result of centuries-old gender norms and hierarchies. While educational equality is the initial barrier to reverse the gender hierarchy, gender reversal in education does not necessarily lead to gender parity in other areas. Accordingly, we expect that the trend in patriarchal attitudes over time depends on the type of patriarchal attitudes. We expect attitudes that are directly related to the societal power structure (economic and political patriarchal attitudes) to have become stronger in Turkey over the study period (H2a). On the other hand, we hypothesize that patriarchal attitudes that are not directly related to the power structures in Turkish society (such as patriarchal attitudes about education) will follow the global trend toward gender equality, and thus show a decline over those twenty years (H2b). Religion, Conservatism, and Patriarchal Attitudes Empirically, the association between religion and patriarchal attitudes has been well-documented in both historic and contemporary societies (Civettini and Glass 2008; Davis and Greenstein 2009; Frejka and Westoff 2008; Goldscheider 2006; Goldscheider, Goldscheider, and Rico-Gonzalez 2014; Klingorova and Havlicek 2015; Pearce and Thornton 2007). For instance, both Frejka and Westoff (2008) and Goldscheider (2006) find that as individuals’ level of religiosity increases, their familial patriarchal values also increase. Similarly, a recent study conducted by Goldscheider, Goldscheider, and Rico-Gonzalez (2014) finds that religiosity increases patriarchal family values among younger Swedish adults. The same study demonstrates that respondents report less patriarchal family values when their religious denomination supports a more egalitarian distribution of gender work balance in the family. Other research also links religiosity with disapproval of divorce, cohabitation, and premarital sex (Goldscheider, Goldscheider, and Rico-Gonzalez 2014; Pearce and Thornton 2007). Drawing on a state-level analysis examining religion’s effect on women’s social, political, and economic status, Klingorova and Havlicek (2015) find a positive association between religious practice and gender inequality. They find that countries where a majority of inhabitants are without a major religion report the least gender inequality. Predominantly Christian and Buddhist societies express average levels of gender inequality, while gender inequality is the highest among predominantly Islamic and Hindu states. Similarly, traditional attitudes toward gender roles have also been linked to political conservatism. Characteristics of radical right parties are associated with conservative family values and traditional gender norms that promote women’s roles as wives and mothers (Buğra 2014; Chong 2006; Inglehart and Norris 2000; Korkut and Eslen-Ziya 2011). Turkey has been influenced by both European values and traditional conservative values. In this paper, we hypothesize that religiosity and conservatism increase patriarchal attitudes in Turkish society (H3 and H4). However, we expect this effect to differ by the type of patriarchal attitude. Drawing from Power Threat Theory (Blalock 1967) and Integrated Threat Theory (Stephan and Stephan 2000), we hypothesize that both religiosity and political conservatism have stronger effects on patriarchal attitudes associated with power at the societal level. Alternatively, we expect the influence of religion and political conservatism on patriarchal attitudes that are not directly related to power at the societal level to be significantly weaker (H3a and H4a). We also test the effect of religiosity and conservatism on patriarchal values over time. Data and Methods We use five waves (1990, 1996, 2001, 2007, and 2011) of the World Values Survey (WVS) data. The data are cross-sectional, meaning that a different set of respondents was interviewed in each wave. The Turkish Statistics Institute conducted this survey using multi-stage full probability sampling, under the supervision of Professor Yılmaz Esmer. Respondents were interviewed regardless of their language, citizenship, and immigration status, though the surveys were all conducted in Turkish. The questionnaire for each wave consists of over 200 closed-ended questions that measure Turkish individuals’ attitudes toward social, political, economic, health, and environmental issues (WVS 2011). A total of 9,289 individuals have been interviewed in Turkey for the WVS over these five waves. In our analysis, we use listwise deletion which reduces the sample size for the regression analyses to the range from 6,190 to 7,171, depending on the dependent variables used. Based on bivariate analysis of the missing data, individuals who are less patriarchal in terms of economic, political, and educational patriarchy were more likely to be in the missing data group. Moreover, individuals who are younger, in a higher income scale, less religious, and less conservative were more likely to be in the missing data group. Females, those who are not married, and those who are not working also have a slightly higher chance of being in the missing data group. Operationalization of Variables We analyze four different types of patriarchal attitudes: economic, familial, political, and educational patriarchy. We initially attempted to combine these four variables into a single index. However, this was not possible due to both low eigenvalues (<1) and factor loadings suggesting that they clearly measure different aspects of patriarchy. That is, those who hold patriarchal values in one aspect (i.e., political patriarchy) do not necessarily hold patriarchal values in another (i.e., educational patriarchy). Both threat theories suggest that competition over economic resources and political power threatens the status quo, and thus results in negative attitudes toward out-groups. Accordingly, we consider political and economic patriarchal attitudes to be directly related to the power structure at the societal level. We assume that educational patriarchal values are not directly related to the power structure at the societal level. Lastly, we consider familial patriarchal attitudes to be associated with power structure in the institution of the family. For our study, economic patriarchy is operationalized by individuals’ level of agreement with the statement: “When jobs are scarce, men should have more right to a job than women.” Political patriarchy is measured by the agreement with the statement: “On the whole men make better political leaders than women do.” Familial patriarchy is operationalized by the agreement with the statement: “Being a housewife is just as fulfilling as working for pay.” Lastly, individuals’ educational patriarchy is operationalized by their level of agreement with the following statement: “A university education is more important for a boy than for a girl.” Response choices for these questions range on a scale from strongly agree (4) to strongly disagree (1). Economic and familial patriarchal attitudes are available for years 1990 through 2011, while political and educational patriarchal values are only available from 1996 to 2011. On a scale from one to four, mean familial and political patriarchy is close to three, and mean educational patriarchy is 1.99. On a scale from one to three, mean economic patriarchy is 2.29 (see table 1 for descriptive statistics). Thus, the patriarchal values are equally strong for family, political, and economic participation, and weakest for education. Table 1 Descriptive statistics   Mean  SD  Minimum  Maximum  Dependent variables  Patriarchy: Economic  2.29  0.90  1  3  Patriarchy: Familial  2.99  0.83  1  4  Patriarchy: Political  2.76  0.94  1  4  Patriarchy: Educational  1.99  0.97  1  4  Independent variables  Religiosity  0.04  0.73  –3.27  0.61  Conservatism  6.01  2.58  1  10  Control variables  Years of education  9.5  3.81  0  16  Employment status           Full-time working (ref)           Part-time working  0.06  0.24  0  1   Not working  0.55  0.50  0  1  Female  0.49  0.50  0  1  Age  37.69  13.80  17  91  Marital status           Married (ref)           Previously married  0.05  0.22  0  1   Single  0.23  0.42  0  1  Number of children  2.08  1.97  0  8  Income levels  3.94  2.19  1  10    Mean  SD  Minimum  Maximum  Dependent variables  Patriarchy: Economic  2.29  0.90  1  3  Patriarchy: Familial  2.99  0.83  1  4  Patriarchy: Political  2.76  0.94  1  4  Patriarchy: Educational  1.99  0.97  1  4  Independent variables  Religiosity  0.04  0.73  –3.27  0.61  Conservatism  6.01  2.58  1  10  Control variables  Years of education  9.5  3.81  0  16  Employment status           Full-time working (ref)           Part-time working  0.06  0.24  0  1   Not working  0.55  0.50  0  1  Female  0.49  0.50  0  1  Age  37.69  13.80  17  91  Marital status           Married (ref)           Previously married  0.05  0.22  0  1   Single  0.23  0.42  0  1  Number of children  2.08  1.97  0  8  Income levels  3.94  2.19  1  10  Source: World Values Survey, 1990–2011. Valid N varies from 6,190 to 7,107. Table 1 Descriptive statistics   Mean  SD  Minimum  Maximum  Dependent variables  Patriarchy: Economic  2.29  0.90  1  3  Patriarchy: Familial  2.99  0.83  1  4  Patriarchy: Political  2.76  0.94  1  4  Patriarchy: Educational  1.99  0.97  1  4  Independent variables  Religiosity  0.04  0.73  –3.27  0.61  Conservatism  6.01  2.58  1  10  Control variables  Years of education  9.5  3.81  0  16  Employment status           Full-time working (ref)           Part-time working  0.06  0.24  0  1   Not working  0.55  0.50  0  1  Female  0.49  0.50  0  1  Age  37.69  13.80  17  91  Marital status           Married (ref)           Previously married  0.05  0.22  0  1   Single  0.23  0.42  0  1  Number of children  2.08  1.97  0  8  Income levels  3.94  2.19  1  10    Mean  SD  Minimum  Maximum  Dependent variables  Patriarchy: Economic  2.29  0.90  1  3  Patriarchy: Familial  2.99  0.83  1  4  Patriarchy: Political  2.76  0.94  1  4  Patriarchy: Educational  1.99  0.97  1  4  Independent variables  Religiosity  0.04  0.73  –3.27  0.61  Conservatism  6.01  2.58  1  10  Control variables  Years of education  9.5  3.81  0  16  Employment status           Full-time working (ref)           Part-time working  0.06  0.24  0  1   Not working  0.55  0.50  0  1  Female  0.49  0.50  0  1  Age  37.69  13.80  17  91  Marital status           Married (ref)           Previously married  0.05  0.22  0  1   Single  0.23  0.42  0  1  Number of children  2.08  1.97  0  8  Income levels  3.94  2.19  1  10  Source: World Values Survey, 1990–2011. Valid N varies from 6,190 to 7,107. The main independent variables are the level of religiosity and conservatism. There are three questions in the data that measure the intensity of respondents’ religious attitudes. These questions ask individuals’ frequency of attending religious services (seven-point scale), how important God is in their life (ten-point scale), as well as how important religion is in their life (three-point scale). Around 35 percent of the respondents attend religious services at least once a week. Over 75 percent of Turkish individuals said that God is very important in their life. In addition, 73 percent of respondents claimed that religion is very important in their life. Using factor analysis, we combined these three religion variables into a single religiosity variable (the eigenvalue from one-factor solution to two-factor solution drops from 1.02 to –0.06). Factor loadings for the three religiosity variables were 0.39, 0.65, and 0.67. The range for the religiosity index is from –3.27 to 0.61, with a mean of 0.04 (standard deviation [SD] 0.73). A further test of reliability (α = 0.64) in addition to similar item-test correlation scores show that the variables fit well in the scale and sufficiently measure the same underlying concept. To measure political conservatism, respondents place their views on political matters on the scale ranging from one (liberal) to ten (conservative). The average level of conservatism is 6.01 (SD = 2.58). All of the analyses are controlled for respondents’ gender, age, educational attainment, employment status, marital status, number of children, and income. The sample consists of 49 percent females. Respondents’ ages range from 17 to 91 years, with a mean of 37.7 years. Educational attainment was initially a categorical variable, operationalized by asking individuals’ highest educational level (nine different levels). The response choices ranged from no formal education to university-level education with degree. We recoded education into years of schooling in order to treat it as a continuous variable. Individuals’ years of education range from zero to sixteen, on average they have 9.6 years of education—equivalent to incomplete secondary education. While the number of children varies from zero to eight, on average respondents have 2.08 children. Income is measured on a subjective scale of one to ten, in which one indicates the lowest income group and ten indicates the highest income group. The mean level of income is 3.95. Categories of employment status (part-time working and not working) are coded as dummy variables, where the reference category is working full-time. About 55 percent of the sample is not working, 6 percent works part-time, and 49 percent works full-time. Marital status is measured by two dummy variables—single and previously married—where the reference category is married. About 72 percent of the sample is currently married, 5 percent are previously married, and 23 percent of the sample is single, never married. Method We use ordered logistic regression to test the effect of religiosity and conservatism on patriarchal attitudes. Including the year of survey as a categorical variable allows us to measure the change in patriarchy over time. It also allows us to model the change over time as a non-linear change. Predicted values from these regressions allow us to plot the trend of patriarchy over time. To investigate the change in the effect of religion and conservatism on patriarchy, we use interaction terms between year dummies and the main independent variable (either religion or conservatism). Again, using year dummies, instead of continuous measure of year, allows us to model the change in the effect of religiosity and conservatism for each time period separately instead of forcing a linear trend. To better understand the interaction effect results, we plot the predicted values from the interaction effects. Findings Table 2 presents the odds ratios from the ordered logistic regression testing the main effects of religiosity and conservatism on economic, familial, political, and educational patriarchy in Turkey from 1990 to 2011. We should note that the reference category for year in the models predicting economic and familial patriarchy is 1990 while it is 1996 for the models predicting political and educational patriarchy since political and educational patriarchy were not measured in 1990. Table 2 Odd ratios from ordered logistic regression: the effect of religiosity and conservatism on patriarchal attitudes in Turkey (1990–2011)   Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)           1996  1.47***  0.90       2001  1.49***  0.89  0.90  0.66***   2007  1.45***  0.77**  0.88†  0.49***   2011  2.28***  0.87  1.74***  1.19*  Independent variables          Religiosity  1.50***  1.33***  1.33***  1.14***  Conservatism  1.10***  1.04***  1.07***  1.06***  Control variables          Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.48***  0.71***  0.65***  0.56***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)           Part-time working  0.98  0.74**  1.12  0.82†   Not working  1.05  1.02  1.03  1.08  Marital status(1)           Single  0.88†  0.95  1.06  1.20*   Previously married  0.94  0.90  0.79*  0.92  Number of children  1.11***  1.04*  1.01  1.06***  Income levels  0.92***  0.96**  0.96***  0.94***  Likelihood ratio χ2  1123.45***  405.56***  562.21***  638.39***   Degrees of freedom  15  15  14  14  Pseudo R2  0.09  0.03  0.04  0.04  Observations  7,107  7,003  6,190  6,301    Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)           1996  1.47***  0.90       2001  1.49***  0.89  0.90  0.66***   2007  1.45***  0.77**  0.88†  0.49***   2011  2.28***  0.87  1.74***  1.19*  Independent variables          Religiosity  1.50***  1.33***  1.33***  1.14***  Conservatism  1.10***  1.04***  1.07***  1.06***  Control variables          Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.48***  0.71***  0.65***  0.56***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)           Part-time working  0.98  0.74**  1.12  0.82†   Not working  1.05  1.02  1.03  1.08  Marital status(1)           Single  0.88†  0.95  1.06  1.20*   Previously married  0.94  0.90  0.79*  0.92  Number of children  1.11***  1.04*  1.01  1.06***  Income levels  0.92***  0.96**  0.96***  0.94***  Likelihood ratio χ2  1123.45***  405.56***  562.21***  638.39***   Degrees of freedom  15  15  14  14  Pseudo R2  0.09  0.03  0.04  0.04  Observations  7,107  7,003  6,190  6,301  * p < 0.05; **p < 0.01; ***p < 0.001; †p < 0.10. Source: World Values Survey, 1990–2011. Note. (1) Reference category for year is 1990 for economic and familial patriarchy, and 1996 for political and educational patriarchy. Reference category for employment status is full time and for marital status it is married. Table 2 Odd ratios from ordered logistic regression: the effect of religiosity and conservatism on patriarchal attitudes in Turkey (1990–2011)   Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)           1996  1.47***  0.90       2001  1.49***  0.89  0.90  0.66***   2007  1.45***  0.77**  0.88†  0.49***   2011  2.28***  0.87  1.74***  1.19*  Independent variables          Religiosity  1.50***  1.33***  1.33***  1.14***  Conservatism  1.10***  1.04***  1.07***  1.06***  Control variables          Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.48***  0.71***  0.65***  0.56***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)           Part-time working  0.98  0.74**  1.12  0.82†   Not working  1.05  1.02  1.03  1.08  Marital status(1)           Single  0.88†  0.95  1.06  1.20*   Previously married  0.94  0.90  0.79*  0.92  Number of children  1.11***  1.04*  1.01  1.06***  Income levels  0.92***  0.96**  0.96***  0.94***  Likelihood ratio χ2  1123.45***  405.56***  562.21***  638.39***   Degrees of freedom  15  15  14  14  Pseudo R2  0.09  0.03  0.04  0.04  Observations  7,107  7,003  6,190  6,301    Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)           1996  1.47***  0.90       2001  1.49***  0.89  0.90  0.66***   2007  1.45***  0.77**  0.88†  0.49***   2011  2.28***  0.87  1.74***  1.19*  Independent variables          Religiosity  1.50***  1.33***  1.33***  1.14***  Conservatism  1.10***  1.04***  1.07***  1.06***  Control variables          Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.48***  0.71***  0.65***  0.56***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)           Part-time working  0.98  0.74**  1.12  0.82†   Not working  1.05  1.02  1.03  1.08  Marital status(1)           Single  0.88†  0.95  1.06  1.20*   Previously married  0.94  0.90  0.79*  0.92  Number of children  1.11***  1.04*  1.01  1.06***  Income levels  0.92***  0.96**  0.96***  0.94***  Likelihood ratio χ2  1123.45***  405.56***  562.21***  638.39***   Degrees of freedom  15  15  14  14  Pseudo R2  0.09  0.03  0.04  0.04  Observations  7,107  7,003  6,190  6,301  * p < 0.05; **p < 0.01; ***p < 0.001; †p < 0.10. Source: World Values Survey, 1990–2011. Note. (1) Reference category for year is 1990 for economic and familial patriarchy, and 1996 for political and educational patriarchy. Reference category for employment status is full time and for marital status it is married. The results show that patriarchal economic values have significantly increased since the 1990s. Compared with 1990, the odds of holding patriarchal economic attitudes are 2.28 times greater in 2011. The increase over this time is highest for attitudes of economic patriarchy. Even after the government’s implementation of the new Labor Law in 2003 in order to ameliorate economic gender inequality, the labor force participation of women is lower in 2011 than in 1990. Individuals also hold stronger political and educational patriarchal attitudes in 2011 than they did in 1990. Compared with 1990, while the odds of holding patriarchal political values were not significantly different in 2001 and 2007, these odds had increased by 74 percent by the year 2011. This is an expected increase considering the lack of institutional reforms or incentives to promote women’s political participation. The odds of possessing educational patriarchal values were actually significantly lower in 2001 and 2007 than in 1990. This is consistent with the 1997 education reform that extended education of children to a compulsory eight years, which had a positive impact on girls’ education. However, these odds had increased by 19 percent as compared with 1990 by the year 2011. Institutional changes in educational patriarchy are helpful in explaining this predicament. The reversal of this progressive trend could be associated with the Turkish government’s conservative education policies, such as a sharp increase in imam-hatip schools since 2002 that disadvantage women and girls. While the odds of holding patriarchal family values were lower in 2007 than in 1990, by 2011, they had increased to their previous 1990 level. Accordingly, the odds of patriarchal familial values in 2011 are not significantly different than they were in 1990 (see table 2). The Turkish government implemented institutional family laws and policies to improve the position of women in society. These include permitting women to be the head of the family, and to keep their maiden name after marriage (both in 2001), and changing women’s “dependent” social security status in 2006 (Sunar and Fişek 2005). However, we find that empirically these institutional changes did not have an impact on familial patriarchal values. Overall, the results demonstrate that economic, political, and educational patriarchal attitudes have increased significantly from 1990 to 2011. Meanwhile, familial patriarchy has not changed significantly during the same time span. For all types of patriarchal attitudes, the increase is most prominent in the most recent years, from 2007 to 2011. Figure 2 presents predicted probabilities of different types of patriarchal attitudes over time based on the models listed in table 2. The graphs indicate that, in general, the largest increase in patriarchal attitudes has occurred in recent years, from 2007 to 2011. Economic patriarchy increased also from 1990 to 1996. Educational patriarchy, on the other hand, decreased from 1996 to 2007. But even educational patriarchy saw an increase from 2007 to 2011. Figure 2. View largeDownload slide Predicted probability of different types of patriarchal attitudes, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N varies from 6,190 to 7,107. Figure 2. View largeDownload slide Predicted probability of different types of patriarchal attitudes, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N varies from 6,190 to 7,107. In addition to indicating the trend of patriarchal attitudes over time, table 2 demonstrates the general effect of independent variables on patriarchal attitudes. Both religiosity and conservatism have a positive association with the four measures of patriarchy. As individuals’ level of religiosity increases, their odds of having patriarchal economic, political, familial, and educational attitudes also increase by 50, 33, 33, and 14 percent, respectively (p<0.001). Similarly, as individuals’ conservatism increases, their odds of expressing economic, political, familial, and educational patriarchal attitudes increase by 10, 4, 7, and 6 percent, respectively (p<0.001). Females, respondents with more years of education, and those with higher income have significantly lower odds of holding patriarchal attitudes. To understand if the effect of religiosity or conservatism has changed over time, we test two sets of interaction effects: (i) between year dummies and religiosity, and (ii) between year dummies and conservatism. Because of the added interaction effects, the coefficient for religiosity now represents the effect of religiosity in 1990 for economic and familial patriarchy and in 1996 for political and educational patriarchy (the earliest measure for these types of patriarchal values). For all of the patriarchal values, religiosity starts out with a strong positive effect: each additional unit of religiosity increases the odds of higher patriarchal values by 27 percent (familial patriarchy) to 53 percent (educational patriarchy). The mainly insignificant interaction effects for familial patriarchy and political patriarchy indicate that the effect of religiosity has not changed much over time for familial and for political patriarchy. The effect of religiosity on economic patriarchy increased from 1990 to 1996, but then again, decreased in 2011. The effect of religiosity on educational patriarchy decreased over time. To better understand the results of the interaction effects, we calculated the predicted probability of strongly agreeing with the patriarchal statements as a function of religiosity over time and graphed it in figure 3. As confirmed from the interpretation above, the effect of religiosity remains the same across time for political patriarchy and for familial patriarchy. Figure 3. View largeDownload slide Predicted probability of strongly agreeing with the patriarchal attitude as a function of religiosity, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N varies from 6,190 to 7,107. Figure 3. View largeDownload slide Predicted probability of strongly agreeing with the patriarchal attitude as a function of religiosity, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N varies from 6,190 to 7,107. However, much more movement over time is seen for economic patriarchy and for educational patriarchy. The effect of religiosity on economic patriarchy becomes sharply stronger from 1990 to 1996 and then starts slowly reducing over time. The effect of religiosity on economic patriarchy in 2011 is significantly weaker than the effect of religiosity in 1990. For educational patriarchy, the effect of religiosity decreased slowly from 1996 to 2007. From predicting the most patriarchal attitude for educational patriarchy (strongly agreeing that university is more important for boys than for girls), religiosity reverses its effect in 2011: this means, as religiosity increases, the probability of strongly agreeing with this statement decreases (the opposite is also true: as religiosity increases, the probability of strongly disagreeing with this statement increases). One reason why the effect of religion on educational and economic patriarchy has been decreasing may be due to the increasing numbers of Islamic schools and institutions in Turkey over the past decade. Eligür (2010) points out that with the election of the Justice and Development Party in 2002, there has been an upsurge in businesses, schools, and other public institutions that are run by Islamists. Thus, individuals may be feeling more comfortable allowing their children to attend and work at these religious institutions, regardless of their gender. Table 3 also presents the results of the interaction effects between year and conservatism. First, these results indicate that conservatism does not really affect familial patriarchy. This confirms that traditional gender norms persist in Turkish society. For all other types of patriarchy, conservatism increases patriarchal attitudes during the earliest year of measurement in this data. However, the effect of conservatism changes over time only for economic patriarchy and not for other types of patriarchy. The effect of conservatism for economic patriarchy is lower in 2001 and 2007 than in 1990. Thus, there seems to be a temporary decrease in the effect of conservatism on economic patriarchy. However, by 2011 the effect of conservatism on economic patriarchy had returned to its 1990 level. Table 3 Odd ratios from ordered logistic regressions with interaction effects: the over-time effect of religiosity and conservatism on patriarchal attitudes in Turkey (1990–2011)   Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)   1996  1.63†  0.96       2001  2.48***  0.78  1.11  0.65*   2007  2.74***  0.78  1.15  0.46***   2011  3.18***  0.65†  1.88**  1.63*  Religiosity  1.45***  1.27**  1.37***  1.53***  Conservatism  1.18***  1.02  1.10***  1.06**  Interaction year × religiosity   1996  1.42**  1.21†       2001  1.11  1.01  0.99  0.83†   2007  0.90  1.18  1.05  0.65**   2011  0.76*  0.95  0.82†  0.44***  Interaction year × conservatism   1996  0.98  0.99       2001  0.91*  1.03  0.97  1.01   2007  0.89*  1.00  0.96  1.02   2011  0.94  1.05  0.98  0.95  Control variables  Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.47***  0.70***  0.64***  0.55***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)   Part-time working  0.97  0.74**  1.12  0.82†   Not working  1.05  1.02  1.04  1.08  Marital status(1)           Single  0.85*  0.95  1.05  1.16*   Previously married  0.96  0.90  0.80*  0.94  Number of children  1.11***  1.04*  1.01  1.06**  Income levels  0.92***  0.96**  0.96**  0.95***  Likelihood ratio χ2  1171.29  414.87  570.39  724.83   Degrees of freedom  23  23  20  20  Pseudo R2  0.09  0.03  0.04  0.05  Observations  7,107  7,003  6,190  6,301    Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)   1996  1.63†  0.96       2001  2.48***  0.78  1.11  0.65*   2007  2.74***  0.78  1.15  0.46***   2011  3.18***  0.65†  1.88**  1.63*  Religiosity  1.45***  1.27**  1.37***  1.53***  Conservatism  1.18***  1.02  1.10***  1.06**  Interaction year × religiosity   1996  1.42**  1.21†       2001  1.11  1.01  0.99  0.83†   2007  0.90  1.18  1.05  0.65**   2011  0.76*  0.95  0.82†  0.44***  Interaction year × conservatism   1996  0.98  0.99       2001  0.91*  1.03  0.97  1.01   2007  0.89*  1.00  0.96  1.02   2011  0.94  1.05  0.98  0.95  Control variables  Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.47***  0.70***  0.64***  0.55***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)   Part-time working  0.97  0.74**  1.12  0.82†   Not working  1.05  1.02  1.04  1.08  Marital status(1)           Single  0.85*  0.95  1.05  1.16*   Previously married  0.96  0.90  0.80*  0.94  Number of children  1.11***  1.04*  1.01  1.06**  Income levels  0.92***  0.96**  0.96**  0.95***  Likelihood ratio χ2  1171.29  414.87  570.39  724.83   Degrees of freedom  23  23  20  20  Pseudo R2  0.09  0.03  0.04  0.05  Observations  7,107  7,003  6,190  6,301  * p < 0.05; **p < 0.01; ***p < 0.001; †p < 0.10. Source: World Values Survey, 1990–2011. Note. (1) Reference category for year is 1990 for economic and familial patriarchy, and 1996 for political and educational patriarchy. Reference category for employment status is full-time, and for marital status it is married. Table 3 Odd ratios from ordered logistic regressions with interaction effects: the over-time effect of religiosity and conservatism on patriarchal attitudes in Turkey (1990–2011)   Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)   1996  1.63†  0.96       2001  2.48***  0.78  1.11  0.65*   2007  2.74***  0.78  1.15  0.46***   2011  3.18***  0.65†  1.88**  1.63*  Religiosity  1.45***  1.27**  1.37***  1.53***  Conservatism  1.18***  1.02  1.10***  1.06**  Interaction year × religiosity   1996  1.42**  1.21†       2001  1.11  1.01  0.99  0.83†   2007  0.90  1.18  1.05  0.65**   2011  0.76*  0.95  0.82†  0.44***  Interaction year × conservatism   1996  0.98  0.99       2001  0.91*  1.03  0.97  1.01   2007  0.89*  1.00  0.96  1.02   2011  0.94  1.05  0.98  0.95  Control variables  Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.47***  0.70***  0.64***  0.55***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)   Part-time working  0.97  0.74**  1.12  0.82†   Not working  1.05  1.02  1.04  1.08  Marital status(1)           Single  0.85*  0.95  1.05  1.16*   Previously married  0.96  0.90  0.80*  0.94  Number of children  1.11***  1.04*  1.01  1.06**  Income levels  0.92***  0.96**  0.96**  0.95***  Likelihood ratio χ2  1171.29  414.87  570.39  724.83   Degrees of freedom  23  23  20  20  Pseudo R2  0.09  0.03  0.04  0.05  Observations  7,107  7,003  6,190  6,301    Economic patriarchy  Familial patriarchy  Political patriarchy  Educational patriarchy  Year(1)   1996  1.63†  0.96       2001  2.48***  0.78  1.11  0.65*   2007  2.74***  0.78  1.15  0.46***   2011  3.18***  0.65†  1.88**  1.63*  Religiosity  1.45***  1.27**  1.37***  1.53***  Conservatism  1.18***  1.02  1.10***  1.06**  Interaction year × religiosity   1996  1.42**  1.21†       2001  1.11  1.01  0.99  0.83†   2007  0.90  1.18  1.05  0.65**   2011  0.76*  0.95  0.82†  0.44***  Interaction year × conservatism   1996  0.98  0.99       2001  0.91*  1.03  0.97  1.01   2007  0.89*  1.00  0.96  1.02   2011  0.94  1.05  0.98  0.95  Control variables  Years of education  0.92***  0.96***  0.93***  0.91***  Female  0.47***  0.70***  0.64***  0.55***  Age  0.99***  1.00  1.00  0.99**  Employment status(1)   Part-time working  0.97  0.74**  1.12  0.82†   Not working  1.05  1.02  1.04  1.08  Marital status(1)           Single  0.85*  0.95  1.05  1.16*   Previously married  0.96  0.90  0.80*  0.94  Number of children  1.11***  1.04*  1.01  1.06**  Income levels  0.92***  0.96**  0.96**  0.95***  Likelihood ratio χ2  1171.29  414.87  570.39  724.83   Degrees of freedom  23  23  20  20  Pseudo R2  0.09  0.03  0.04  0.05  Observations  7,107  7,003  6,190  6,301  * p < 0.05; **p < 0.01; ***p < 0.001; †p < 0.10. Source: World Values Survey, 1990–2011. Note. (1) Reference category for year is 1990 for economic and familial patriarchy, and 1996 for political and educational patriarchy. Reference category for employment status is full-time, and for marital status it is married. Figure 2 shows us that over time different types of patriarchy have increased in Turkey (except for familial patriarchy). What contributes to that increase? The interaction effects between year and conservatism showed that the effect of conservatism has not increased (rather, it diminished in some years for economic patriarchy). Similarly, the effect of religiosity has either remained the same or decreased. However, if we look at the overall level of religiosity and conservatism in Turkey, we see that both average religiosity and average conservatism have increased (see figure 4). Figure 4 presents a bivariate graph between year and religiosity and between year and conservatism. Clearly, both religiosity and conservatism overall have increased over this time period (and based on ANOVA and Bonferroni tests the increase is significant). We see a gradual increase in conservatism over time. Starting with 1996, religiosity is higher than in 1990, though there has not been a significant increase since. Instead, there is a slight decrease in the level of religiosity in 2011. Nevertheless, the level of religiosity in 2011 is significantly higher than it was in 1990. Thus, perhaps the overall increase in the level of religiosity and conservatism explain the increase in patriarchal values as there is a clear positive relationship between religiosity and patriarchal attitudes and between conservatism and patriarchal attitudes. Figure 4. View largeDownload slide Mean level of religiosity and conservatism over time, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N ranges from 6,190 to 7,107. Figure 4. View largeDownload slide Mean level of religiosity and conservatism over time, Turkey, 1990–2011. Source: World Value Survey, Turkey, 1990–2011. Valid N ranges from 6,190 to 7,107. Conclusion In this paper, we examine the effect of religiosity and conservatism on patriarchal attitudes in Turkey using the five waves of the WVS (1990, 1996, 2001, 2007, and 2011). Modernization theorists argue that gender equality increases with industrialization and economic development (Inglehart and Norris 2000, 2003; Inglehart, Norris, and Welzel 2002). However, we see that Turkey does not follow this pattern. In fact, patriarchal values have been on the rise. Most Turkish individuals still agree with the statements: “When jobs are scarce, men should have more right to a job than women” (economic patriarchy); “On the whole men make better political leaders than women do” (political patriarchy); “A university education is more important for a boy than for a girl” (educational patriarchy), and “Being a housewife is just as fulfilling as working for pay” (familial patriarchy). Our first hypothesis (H1) is supported: despite Turkey’s economic progress, the level of patriarchal attitudes in Turkey increased in recent years. The increase is highest during the latest years analyzed (2007–2011). Another important question we address is why Turkey does not follow the pattern of Western societies even though it has had a secular, democratically elected government for most of twentieth century. To explain this, we apply Power Threat Theory and Integrated Threat Theory to our findings. We argue that realistic threats (such as political and economic power) posed to the in-group (men) by the out-group (women) influence specific threat perceptions. We find that economic, political, and educational patriarchal attitudes are significantly higher in 2011 than in the previous years of the study period. For all types of patriarchal attitudes, the increase is most prominent in the most recent years, from 2007 to 2011. In fact, educational patriarchy decreased in 2001 and in 2007, but then increased again when measured in 2011. On the other hand, familial patriarchy has not significantly changed since the 1990s. Drawing from Power Threat Theory and Integrated Threat Theory, we hypothesized that patriarchal attitudes that are directly related to the power structure in society (economic and political) would grow stronger in Turkey during the last twenty years and that patriarchal attitudes that are not directly related to societal power structure (familial and educational) would follow the global trend toward equality (H2). Indeed, we see this with educational patriarchy. Educational patriarchy decreased from 1996 to 2001 and from 2001 to 2007. However, these patriarchal attitudes gained strength again from 2007 to 2011 and surpassed the previous levels reported in 1996. Thus, we are able to identify a trend toward more egalitarian attitudes, but then we see a reverse of this progressive trend in more recent years. The reversal of this progressive trend can be explained by the shift in Turkey’s political system. Since 2002, Turkey has been run by a conservative party that has been following an Islamist agenda, aiming to weaken the power of secularism and democracy (Eligür 2010). In our earlier discussion, we also noted that attendance in imam-hatip (religious) schools has increased sharply since 2002. With the launch of the new education reform in 2012, we predict even further increase in educational patriarchy after 2011. Interestingly, the familial patriarchal attitudes have not changed over time. Familial patriarchal attitudes deal with power structures within the family and not necessarily within society. However, family power structure is even more personal than societal-level power structures. Thus, we should expect the least change in attitudes toward family power structure. We find that institutional changes in family law and policy do not impact familial patriarchal values. Neither the global trend toward egalitarianism nor the increase in conservativism has changed the level of familial patriarchal attitudes in Turkey over time. The descriptive findings of patriarchal attitudes in Turkey confirm our historical discussion on the failure of the feminist movement in Turkey (Arat 2000; Kandiyoti 1987). Its lack of a collective voice to integrate individuals from all backgrounds is reflected in the increase in patriarchal values over time. We find a significant and positive effect of both religion and conservatism on patriarchal values. As individuals’ level of religiosity increases, they are much more likely to possess patriarchal attitudes toward economy, family, politics, and education (supporting hypothesis H3). Similarly, as Turkish individuals’ level of political conservatism increases, they are more likely to express patriarchal attitudes (supporting hypothesis H4). The overall effect of conservatism with pooled data across five waves of the study does not vary based on the type of patriarchal attitude (thus, no support for hypothesis H4a). The effect of religion seems to be strongest for economic patriarchy and weakest for educational patriarchy, following our hypothesis H3a: religion’s effect is stronger for patriarchal attitudes associated with the power structure in the society and weaker for patriarchal attitudes that are not directly related to the power structure in the society. Religion’s effect on patriarchal economic and educational attitudes diminished significantly since 1990 and 1996. It is interesting to note that while overall patriarchal educational and economic values were increasing, religion’s effect on these values was diminishing. This may be associated with the contemporary upsurge in Islamic schools and institutions. Religious families may be more inclined to allow their children to attend these institutions. On the other hand, we find that religion’s effect on patriarchal family values has not significantly changed over time. This demonstrates that the traditional view of women’s role in the family remains ingrained in Turkish society. We also find that religion’s effect on patriarchal political values has remained the same (significant and positive). This also confirms our theoretical framework. That is, since political patriarchy is directly related to the power structure, its effect would be the hardest one to reverse. Lastly, we see that the positive effect of conservatism on all forms of patriarchal attitudes has remained the same over time. Patriarchal attitudes play a central role in the continuing gender inequality in economic, familial, political, and educational domains of Turkish society. However, the overall picture of the Turkish case is more complex than a one-size-fits-all solution. We find that modernization theories fail to account for the change in values in Turkey, a society with a predominantly Muslim population that has experienced Westernization and democratic elections. Emancipation of women in Turkey was used as a political agenda and women were treated as tools of modernization and Westernization (Eslen-Ziya and Korkut 2010). In line with Moghadam’s (2003) discussion of neo-patriarchal states, we find that gender equality does not occur as an automatic outcome of economic development. Instead, in neo-patriarchal states such as Turkey “religion is bound to power and state authority; moreover, the family, rather than the individual constitutes the universal building block of the community” (Moghadam 2003, 11). Thus, both the patriarchal family and neo-patriarchal state influence one another, buttressing “normative views of women and the family, often but not exclusively through the law” (Moghadam 2003, 130). Our theoretical framework supports most of our predictions. Future research should specifically explore other potential causes of the recent upsurge in patriarchal attitudes in Turkish society. Also, more research is needed on how religion and political conservatism influence actual gender gaps in education, occupation, employment, and political participation as opposed to attitudes. Future research should explore whether patriarchal attitudes impact ethnic minority women (i.e., Kurdish women), rural, or economically disadvantaged women differently. Finally, our paper focuses on the period from 1990 to 2011. More recent political developments in Turkey have led the country away from a democratically elected government and have stalled its Westernization. Thus, Turkey has become more comparable with other Islamic nations in allowing religion to infuse with governance, limiting political freedoms and decreasing rights for women. Once the data beyond 2011 becomes available, an entirely new look at Turkey is warranted. Notes Ceylan Engin is a doctoral candidate in the Department of Sociology at Texas A&M University. She is an interdisciplinary scholar of sociology, demography, and women’s and gender studies with a specific focus on Turkey. Her current research focuses on the intersections of religion, gender, and sexuality. Heili Pals is an Assistant Professor in the Department of Sociology at Texas A&M University. Her current research focuses on the social psychology of the life course, the social psychological explanations behind the development of deviance, inequality due to sexual orientation, and determinants of attitudes toward homosexuality. She works in collaboration with the Howard B. Kaplan Laboratory for Social Science Research and is an associate editor of the journal Population Review. She has published in journals such as Social Psychology Quarterly, American Journal of Criminal Justice, Deviant Behavior, and Journal of Adolescence. References Ahmadi Arash. 2012. Turkey PM Erdogan sparks row over abortion. BBC News, June 1, accessed July 15, 2015, http://www.bbc.com/news/world-europe-18297760. Aldıkaçtı Marshall Gül. 2005. Ideology, progress, and dialogue. Gender & Society , 19 ( 1): 104– 20. Google Scholar CrossRef Search ADS   Arat Yeşim. 2000. From emancipation to liberation. Journal of International Affairs , 54 ( 1): 107– 23. Arat Yeşim. 2010. Religion, politics and gender equality in Turkey. Third World Quarterly , 31 ( 6): 869– 84. Google Scholar CrossRef Search ADS PubMed  Başpınar Veysel. 2003. Türk Medeni Kanunu ile Aile Hukukunda Yapılan Değişiklikler ve Bu Konuda Bazı Önerilerimiz. AHFD , 52 ( 3): 79– 101. Blalock Hubert M. 1967. Toward a theory of minority-group relations . New York: Capricorn Books. Boratav Hande, Fişek Güler, Eslen-Ziya Hande. 2014. Unpacking masculinities in the context of social change. Men and Masculinities , 17 ( 3): 299– 324. Google Scholar CrossRef Search ADS   Buğra Ayşe. 2014. Revisiting the Wollstonecraft dilemma in the context of conservative liberalism. Social Politics , 21 ( 1): 148– 66. Google Scholar CrossRef Search ADS   Caner Asena, Güven Cahit, Ökten Çağla, Sakallı Seyhun Orcan. 2016. Gender roles and the education gender gap in Turkey. Social Indicators Research , 129: 1231– 1254. Google Scholar CrossRef Search ADS   Chong Kelly H. 2006. Negotiating patriarchy. Gender & Society , 20 ( 6): 697– 724. http://dx.doi.org/10.1177/0891243206291111 Google Scholar CrossRef Search ADS   Civettini Nicole H.W, Glass Jennifer. 2008. The impact of religious conservatism on men's work and family environment. Gender & Society , 22 ( 2): 172– 93. http://dx.doi.org/10.1177/0891243207310714 Google Scholar CrossRef Search ADS   Cornell Svante E. 2015. The Islamization of Turkey. The Turkey Analyst, September 2, accessed July 3, 2017, https://www.turkeyanalyst.org/publications/turkey-analyst-articles/item/437-the-islamization-of-turkey-erdo%C4%9Fan%E2%80%99s-education-reforms.html. Davis Shannon N., Greenstein Theodore N.. 2009. Gender ideology. Annual Review of Sociology , 35: 87– 105. Google Scholar CrossRef Search ADS   Dedeoğlu Saniye. 2012. Equality, protection or discrimination. Social Politics , 19 ( 2): 269– 90. Google Scholar CrossRef Search ADS   Demir Zekiye. 2017. Women’s participation in politics and strategies to raise the presence of women within politics, accessed July 3, 2017, http://kadinarastirmalari.kadem.org.tr/kadinlarin-siyasete-katilimi-ve-katilimi-artirmaya-yonelik-stratejiler/. Diner Cagla, Toktaş Şule. 2010. Waves of feminism in Turkey: Kemalist, Islamist and Kurdish women's movements in an era of globalization. Journal of Balkan and Near Eastern Studies , 12 ( 1): 41– 57. Google Scholar CrossRef Search ADS   Dulger Ilhan. 2004. Turkey: rapid coverage for compulsory education. The World Bank, accessed July 3, 2017, http://siteresources.worldbank.org/INTTURKEY/Resources/361616-1142415001082/Compulsory_Education_by_Dulger.pdf. Eligür Banu. 2010. The mobilization of political Islam in Turkey . New York: Cambridge University Press. Google Scholar CrossRef Search ADS   El-Sanabary Nagat. 1993. Middle East and North Africa. In Women’s education in developing countries: Barriers, benefits, and policies , ed. King E., Hill M., 136– 74. Baltimore, MD: Johns Hopkins University Press. Erhart Itır, Eslen-Ziya Hande. 2014. Janissary: An orientalist heroine or a role model for Muslim Women. In Heroines of comic books and literature , ed. Bajac-Carter Maja, Jones Norma, Batchelor Bob, 95– 106. Lanham, MD: Rowman & Littlefield. Eslen-Ziya Hande, Korkut Umut. 2010. Political religion and politicized women in Turkey. Totalitarian Movements and Political Religions , 11 ( 3–4): 311– 26. Google Scholar CrossRef Search ADS   Esmer Yılmaz. 2008. Islam, gender, democracy and values. In Changing values, persisting cultures , ed. Petterson Thorleif, Esmer Yılmaz, 283– 99. Leiden and Boston: Brill. Google Scholar CrossRef Search ADS   Etürk Ali Ekber. 2016. Karma eğitime darbe hazırlığı. Sözcü. June 1, accessed July 3, 2017, http://www.sozcu.com.tr/2016/egitim/karma-egitime-darbe-hazirligi-1256364/. Frejka Tomas, Westoff Charles F.. 2008. Religion, religiousness and fertility in the US and in Europe. European Journal of Population , 24: 5– 31. Google Scholar CrossRef Search ADS   Goldscheider Calvin. 2006. Religion, family, and fertility. In Religion and the decline of fertility in the Western world , ed. Derosas R., van Poppel F., 41– 57. Dordrecht, Netherlands: Springer. Google Scholar CrossRef Search ADS   Goldscheider Frances, Goldscheider Calvin, Rico-Gonzalez Antonio. 2014. Gender equality in Sweden. Journal of Family Issues , 35 ( 7): 892– 908. Google Scholar CrossRef Search ADS   Gumus Sedat, Chudgar Amita. 2016. Factors affecting school participation in Turkey: an analysis of regional differences. Compare: A Journal of Comparative and International Education , 46 ( 6): 929– 51. http://dx.doi.org/10.1080/03057925.2015.1095073 Google Scholar CrossRef Search ADS   Hansen Suzy. 2013. Turkey’s women strike back. New York Review, June 16, accessed July 15, 2015, http://www.nybooks.com/daily/2013/08/19/turkeys-women-strike-back/. Inglehart Ronald, Norris Pippa. 2000. The developmental theory of the gender gap. International Political Science Review , 21 ( 4): 441– 63. http://dx.doi.org/10.1177/0192512100214007 Google Scholar CrossRef Search ADS   Inglehart Ronald, Norris Pippa. 2003. Rising Tide . New York and Cambridge: Cambridge University Press. Google Scholar CrossRef Search ADS   Inglehart Ronald, Norris Pippa, and Christian Welzel. 2002. Gender equality and democracy. Comparative Sociology , 1 ( 3/4): 321– 46. Google Scholar CrossRef Search ADS   İslam Merve Kavakçı. 2010. Headscarf politics . New York: Palgrave Macmillan. Google Scholar CrossRef Search ADS   Kamal Kevin. 2017. Education in Turkey. World Education News and Reviews, April 4, accessed July 3, 2017, http://wenr.wes.org/2017/04/education-in-turkey. Kandiyoti Deniz. 1987. Emancipated but unliberated? Feminist Studies , 13 ( 2): 317– 38. Google Scholar CrossRef Search ADS   Kandiyoti Deniz. 1988. Bargaining with patriarchy. Gender & Society , 2 ( 3): 274– 90. http://dx.doi.org/10.1177/089124388002003004 Google Scholar CrossRef Search ADS   Klingorova Kamila, Havlicek Tomáš. 2015. Religion and gender inequality. Moravian Geographical Reports , 23 ( 2): 2– 11. Korkut Umut, Eslen-Ziya Hande. 2011. The impact of conservative discourses in family policies, population politics, and gender rights in Poland and Turkey. Social Politics , 18 ( 3): 387– 418. http://dx.doi.org/10.1093/sp/jxr014 Google Scholar CrossRef Search ADS PubMed  Moghadam Valentine. 2003. Modernizing women: Gender and social change in the Middle East . Boulder, CO: Lynne Rienner Publishers. Özyeğin Gül. 2001. Untidy gender . Philadelphia: Temple University Press. Pearce D. Lisa, Thornton Arland. 2007. Religious identity and family ideologies in the transition to adulthood. Journal of Marriage and Family , 69: 1227– 1243. http://dx.doi.org/10.1111/j.1741-3737.2007.00443.x Google Scholar CrossRef Search ADS   Pope Nicole, Pope Hugh. 2011. Turkey unveiled . New York: The Overlook Press. Rankin H. Bruce, Aytaç Işık A.. 2006. Gender inequality in schooling. Sociology of Education , 79 ( 1): 25– 43. Google Scholar CrossRef Search ADS   Republic of Turkey Ministry of Foreign Affairs. 2017. Sayın Prof. Dr.Tansu Çiller'in Özgeçmişi (Background of Dr. Tansu Ciller), accessed July 3, 2017, http://www.mfa.gov.tr/sayin-prof_dr_tansu-ciller_in-ozgecmisi_-_haziran-1996_.tr.mfa. Şanli Şölen. 2016. Women and cultural citizenship in Turkey . New York: I.B. Tauris & Co. Ltd. Saktanber Ayşe. 2002. Living Islam . New York: I.B. Tauris & Co. Ltd. Saktanber Ayşe, Çorbacıoğlu Gül. 2008. Veiling and headscarf-skepticism in Turkey. Social Politics , 15 ( 4): 514– 38. Google Scholar CrossRef Search ADS   Seguino Stephanie. 2016. Global trends in gender equality. Journal of African Development , 18: 1– 30. Smits Jereon, Hosgör Ayse Gündüz. 2006. Effects of family background characteristics on educational participation in Turkey. International Journal of Educational Development , 26: 545– 60. Google Scholar CrossRef Search ADS   Stephan G. Walter, Stephan Cookie W.. 2000. An integrated threat theory of prejudice. In Reducing prejudice and discrimination , ed. Oskamp S.. 23– 46. Mahwah, NJ: Lawrence Erlbaum. Stephan Cookie White, Stephan Walter G., Demitrakis Katherine M., Yamada Ann Marie, Clason Dennis L.. 2000. Women's attitudes toward men. Psychology of Women Quarterly , 24 ( 1): 63– 73. Google Scholar CrossRef Search ADS   Sunar Diane, Fişek Güler Okman. 2005. Contemporary Turkish families. In Families in global perspective , ed. Gielen U., Roopnarine J., 169– 83. Boston, MA: Allyn and Bacon/Pearson. Tougas Francine, Brown Rupert, Beaton Ann M., Joly Stéphane. 1995. Neosexism: Plus ça change, plus c'est pareil. Personality and Social Psychology Bulletin , 21: 842– 49. Google Scholar CrossRef Search ADS   Turkiye İstatislik Kurumu (Turkish Statistical Institute). 2012. Milletvekili genel seçimleri 1923-2011 (Deputy General Elections 1923-2011) . Ankara: TUIK. Türkiye Büyük Millet Meclisi (TBMM). 2017. Türkiye büyük millet meclisi milletvekilleri dağılımı [List of deputies of the grand national assembly of Turkey], accessed June 25, 2017, https://www.tbmm.gov.tr/develop/owa/milletvekillerimiz_sd.dagilim. Weise Zia. 2016. Turkey bill to pardon statutory rape if perpetrators marry victim. The Telegraph, November 18, accessed July 3, 2017, http://www.telegraph.co.uk/news/2016/11/18/turkey-bill-to-pardon-statutory-rape-if-perpetrators-marry-victi/. World Bank. 2012. World development report 2012 gender equality and development, accessed September 5, 2015, https://openknowledge.worldbank.org/handle/10986/4391. World Development Indicators. 2016. Labor force participation rate, female (% of female population ages 15+) (Modeled ILO Estimate), accessed May 15, 2016, http://data.worldbank.org/indicator/SL.TLF.CACT.FE.ZS?page=5. World Economic Forum. 2015. The global gender gap report, accessed December 12, 2015, http://reports.weforum.org/global-gender-gap-report-2015/economies/#economy=TUR. World Values Survey. 2017. Fieldwork and sampling, accessed July 5, 2017, http://www.worldvaluessurvey.org/WVSContents.jsp?CMSID=FieldworkSampling. © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com

Journal

Social PoliticsOxford University Press

Published: Jan 23, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off