SOCIAL AND ECONOMIC IMPACT OF THE HEADSCARF BAN ON WOMEN IN TURKEY

In 1989, women in Turkey were prohibited by the Constitutional Court from participating in education and doing paid or unpaid work in public institutions whilst wearing the veil. Except for a few periods where the ruling has been relaxed, the headscarf ban has continued to intensify, engendering heated debates on public, political, and academic platforms (Aktas 2006a, 2006b; Cinar 2008; Cindoglu and Zencirci 2008; Sisman 2009). The intensification is perhaps surprising as research and extensive public opinion polls show limited support for the ban by the Turkish public: surveys demonstrate that 76 percent do not support the ban, and 72 percent want it lifted (Carkoglu and Toprak 2006). Furthermore, national and international human rights organizations have condemned the ban as a human rights violation.

Research has concentrated on the support for the ban amongst Turkish people and on the political, historical, philosophical, and religious implications of the headscarf. The headscarf has been seen in diverse ways: as a patriarchal practise that suppresses women's freedom (Pak 2006; Cinar 2008), a political symbol of protest against the secular Turkish Republic (Cinar 2008; Cindoglu and Zencirci 2008; Temelkuran 2008; Wiygul 2008), an expression of the modernizing Turkish woman (Gole 1996), and a reflection of piety (Sisman 2009). To our knowledge, except for a few noteworthy studies (Rankin and Aytac 2008), no large-scale study focuses on the implications of this ban on the women themselves. Certainly, a large number of women are impacted. Carkoglu and Toprak (2006) note that 61 percent of Turkish women cover their head with some sort of scarf when in public places. But if these women wish to actively participate in schooling, employment, politics, or the arts, they must unveil. This study considers their plight, asking how and to what extent the headscarf ban affects the social and economic well being of Turkish women.

The prohibition of the headscarf has been implemented at all educational institutions and in public sector employment and is an undeclared selection criterion for employers in the private sector. It has been well documented that educational attainment has a significant impact on the life course. Education increases opportunities in labour market, aids upward social mobility, augments earnings (Blau and Duncan 1967), and positively affects health and psychological well-being (Kenkel 1991; Meertens 2004) and political participation (Inglehart 1997). Moreover, education plays a key role in shaping social attitudes, behaviour, and beliefs regarding the meaning of citizenship and community responsibilities (Kennedy and Mellor 2003). The more education individuals receive, the more responsive they are to basic democratic values and civil liberties (Nie et al. 1996), whether in terms of abstract principles or active social duties as part of a community (Lipset 1973). Education also increases analytical ability, social skills, and cultural sophistication, resulting in the belief in freedom of speech and tolerance (Dee 2004). It breaks down barriers and opens up opportunities.¹

One may well expect that banning the headscarf hinders women's educational attainment, as many choose to forgo higher education. This, in turn, causes problems in their social and psychological progress and negatively affects their development of social and economic skills. It creates barriers to full participation in society and reduces their life chances. They become more reliant on close family members, vulnerable and dependent on male relatives.²

Scholars and officials give differing and sometimes conflicting arguments as to why the headscarf has been banned. The most commonly cited argument is that the headscarf is a symbol of resistance to the Kemalist Reforms separating state and religion which constitute the foundations of the Turkish Republic (Lewis 1961; Mardin 2006). One of these reforms concerned the wearing of Western clothing, as prescribed by the hat law (şapka Kanunu). A general ban of religious and folkloric symbols and practices was introduced in the transition period of the Republic (Aktas 2006a); it included men's fez, women's veil, religious education, Sufi practices, and Turkish folkloric music. Women were, for example, forbidden to wear religious veiling, while men were required by law to wear top hats. The hat law (şapka Kanunu) is ostensibly still in force but no man in Turkey wears the Western style hat as prescribed by this law. While most of these regulations have been relaxed, the headscarf ban has intensified since the 1980s.

However, the headscarf ban has gone beyond its origin as a ban of religious and folkloric symbols and now works to exclude women from power by keeping them poor. The Turkish elite, who flourished during the foundation period of the Republic have advantageous positions and control the social, economic, cultural, and political fields (Lewis 1961; Ozyurek 2008). As Lewis (1961): 55) points out, the Turkish elite includes military officers, bureaucrats, lawyers, and intellectuals. It is not in the best interests of this affluent upper class to open doors to the masses: it would weaken their power and threaten their lifestyle.

Bourdieu argues that blocking privileges to outsiders works not only via

Restricting women's participation and progress actually benefits the elite twice. First, it prevents the lower social classes from reaching higher status positions and thus narrowing upper class resources. The urbanization and modernization of Turkey has resulted in the mass migration of the less educated from villages and poor areas.³ The urbanization rate was 42 percent in 1975 and had increased to 67 percent by 2005.⁴ The ban slows down regional and upward social mobility by blocking women's educational attainment and labour market participation. In a related vein, it is worth emphasizing that there are either no dress regulations for low-status jobs such as cleaners, gardeners, and other labouring workers in the public and private sectors, or if there are such regulations, they are not usually put into practice. The ban has been rigorously implemented to prevent teachers from working whilst wearing the headscarf, whereas veiled cleaners in schools remain unobserved. The headscarf ban, in this sense, has served to impede social fluidity in a rapidly industrializing Turkey.

Second, women have been seen as the engine of social change and reproduction (Ruspini and Dale 2002) that could threaten the exclusive resources of the elite. That is, the strong and positive relationship between a mother's resources and her children's future prospects (Ruspini and Dale 2002) increases intergenerational social mobility and contributes to an open society. An open society threatens the elite's exclusive resources. Banning the headscarf from education and employment diminishes the possibility that women can improve their qualifications and skills to compete for better life chances, which, in turn, would work towards a more egalitarian society.

The ban has not only intensified but has been extended to include other areas of life. For example, men with veiled wives have been fired from the army, whilst other men have been prevented from entering (Yavuz 2000). Another telling example is the 2006 decision of the Turkish Council of State that the headscarf ban is not limited to state institutions. In this case, a nursery teacher who unveiled at her workplace applied for a position to become the head of the nursery. Her application was rejected with the argument that she wears a headscarf in her personal life. After a long period of discussion, the Turkish Council of State decided that she could not occupy the position because she wears a headscarf in her personal life and could therefore not be a model citizen for her pupils. At this point, the ban is spreading to include women's personal lives (Benli 2005).

In accordance with Bourdieu's (1984) theory, these extensions of the headscarf ban create dress codes which state that the model female citizen is unveiled. By implication, veiled women are situated outside the high-status fields of social life, and this exclusion includes their husbands, friends, and relatives. Furthermore, the widening of the headscarf ban provides a guideline for private sector institutions when selecting their employees, customers, or students. Consequently, the ban and its extensions make it hard for women who unveil in places where the ban applies to build a career. This ban worsens the traditionally poor conditions of women in Turkey and provides a means for patriarchal Turkish men to control their daughters, sisters, wives, and other close female relatives (Kilic 2008; Rankin and Aytac 2008). It both directly and indirectly weakens the already fragile social position of women and breaks down women's protests against and resistance to these practices.

We use the Headscarf Ban Survey initiated by the HAZAR Education and Solidarity Association and held between 5 January and 15 February 2007 from nine Turkish cities: Istanbul, Ankara, Izmir, Bursa, Erzurum, Adana, Diyarbakir, Konya, and Samsun. The sample sizes for each city appear in the Appendix. The survey collects data from women who have faced the headscarf ban and who are between the ages of 18 and 50. Since the data include only women with experience of the consequences of the ban, participants are mainly skilled and qualified women with career aspirations. Some now wear the headscarf; others once wore it but later unveiled. Veiled women who stay at home are less likely to experience the ban and therefore are less represented in the sample. The snowball sampling method was employed to collect data, and networks and databases of civil society institutions⁵ were used to recruit participants who fit the criteria. Women in the first sampling points were screened with regard to whether they faced or have been facing the headscarf ban in Turkey in any way. A confirmatory answer made them eligible for interview. Furthermore, eligible and ineligible women were asked to give names of other women who would meet the criteria. This method provided 1,206 semi-structured face-to-face interviews.

Snowball data have some limitations compared to probability sampling data. It has been argued that the former are not suitable for statistical inferences because they do not meet the assumptions of probability (Agresti and Finlay 1997; Bryman 2008). Nonetheless, snowball sampling may be a better method to reveal relations and associations in specific populations (Goodman 1961). We think that the data may have some constraints in generalizing results; nevertheless, they are unique because they represent detailed and large-scale information collected from a seldom-sampled population. To some extent, they show the impact of the ban on the social and economic conditions of women who wear or have worn the headscarf.

Table 1 shows the level of education of the women in the sample. The Turkish educational structure is straightforward; it used to have five years of primary school, three years of secondary school (orta okul), three years of higher secondary school (lise), which also contains vocational tracks, two years of higher school, and four years of university. Recently, the primary and secondary schools were combined and made compulsory. Eight percent of the sample have secondary education or lower; 41 percent have high secondary qualifications; 48 percent have university education or higher; and 3 percent dropped out of education. The last column of Table 1 shows the percentages of the highest level of education of women in Turkey in the 2008 Census. As these figures make clear, the level of education of women in our sample is higher than in the general female population. In the general population, the overwhelming majority (64 percent) have secondary education or lower, 16 percent have high secondary qualifications, and 6 percent have university education or higher. A reason for the overrepresentation of more educated women in the data is, as noted, that these are the women who face the headscarf ban. Further, the ban is not strictly implemented in the jobs that require little or no education, as noted above.

Table 2 presents the results for the question ‘What did you do when you faced the ban?’ The largest single group of women in the sample did not unveil after facing the ban (43 percent); those who took their scarves off in places where the ban applies represent the second largest group in the study (33.7 percent). Nineteen percent used a wig or hat instead of the traditional or religious headscarves to avoid the ban. Finally, 1.2 percent of women in the sample unveiled completely.⁶

Those who unveiled in one way or another were asked why they did so; the results are presented in Table 3. The most often reported reasons for women to unveil are social and economic concerns: Women unveiled to complete their education (73 percent), to keep their jobs (22.5 percent), because of their worries about future prospects (19.7 percent), and to complete their academic career (11.1 percent). As the table shows, women also remove their scarves because of pressure from others.

Table 4 shows how women feel when they have to unveil. Although 15 percent reported that they believed in what they were doing and therefore were not badly affected, the rest were all affected in some way or another. An overwhelming majority said their personality was adversely affected (70.8 percent) and they suffered an affront (63.2 percent) by unveiling. Others were ashamed, thought they were sinners, lost self-respect, experienced a decrease in their educational performance, and were full of anger towards those who forced them to unveil. Women mainly unveil to remove barriers to social and economic achievement, resulting in poor performance and anger.

Table 5 shows responses to a question about conditions at the workplace because of the headscarf. Women reported that they could not find jobs, were forced to work where they could not be seen by others, had to work in jobs other than in their own specialization, and made less money; some suffered decreased work performance and were fired. As expected, these results show that veiled women are penalized in their work careers.

Table 5 also deals with conditions outside the workplace. It shows that women meet with negative attitudes and behaviour in their social lives. For one thing, the headscarf ban restricts women's behaviour: 51 percent avoid places where problems might occur, and 18 percent have withdrawn from social life. Moreover, women were married quite young after they were forced to leave school (2.3 percent).

The last part of Table 5 shows the psychological conditions women have experienced because of their headscarf. Four percent have not experienced any psychological problems, but all others have been badly affected. The headscarf ban has damaged their trust in state institutions (63.9 percent) and the law (60.6 percent), key institutions of a democratic society.

Thus far, we have discussed the self-reported experiences of women. To reveal social and economic disparities, we need to compare veiled women with women who have never worn or aspired to wear the headscarf. Although the data do not contain such a comparison group, we have information on whether women have unveiled or used a wig or hat to camouflage their veil. Therefore, we compare the level of education and employment amongst women who have not unveiled, those who have used a wig or hat, and those who have unveiled either permanently or in places where the ban applies. If women unveil, the obstacle to participate in education and labour markets is partly removed. Consequently, one would expect the women who unveil permanently or in places where the ban applies to have a higher level of education; they are also more likely to be employed than those who do not unveil.

Table 6 ⁷ shows the level of education for women who unveil entirely or in places where the ban applies, women who use a wig or hat, and veiled women. There are significant differences between these groups of women, with veiled women having the lowest education. Drop-out rates are highest among veiled women. Thirteen percent of the veiled women have secondary or lower education, compared to 4 percent of the unveiled, and 1 percent of those using a wig. Sixty-three percent of women using a wig or hat, 52 percent of the unveiled women, and 40 percent of veiled women have or higher educational attainment. We cannot fully explain why women wearing a wig have a higher level of education than unveiled women. Possibly, wig wearers are more motivated than those who are unveiled because the former aim to optimize their chances to advance. To sum up, there are statistically significant differences between veiled and unveiled women, supporting the expectation that the headscarf ban in Turkey negatively punishes veiled women in the area of educational attainment.

Table 7 shows the differences in employment status between veiled women and those who unveil or use a means of camouflage. The differences in employment status are not significant between women who have unveiled permanently or in places where the ban applies and women who use a wig or hat to avoid the ban. There are, however, substantial differences between veiled and unveiled women in employment status: 73 percent of unveiled women are active in the labour market whereas this share decreases to 55 percent for veiled women, a drop of (72.8–54.5) 18 percent. Furthermore, 24 percent of the unveiled women are unemployed; this increases to 39 percent for veiled women, showing a disparity of (39.3–24.0) 15 percent. These differences remain significant after controlling for age, education, and attitudes towards women's labour market participation (not shown here).⁸ These findings support the expectation that veiled women are disadvantaged in labour market outcomes.

To compare how different cities accommodate veiled women's educational and economic attainment, we distinguished between cities with ruling parties that represent the interests of the upper classes, the Turkish elite, and parties that advocate the interests of the lower classes. The local governments had been in power for more than three years at the time of data collection in the cities.

The ruling party in Izmir was the Republican People's Party (CHP)⁹ which was and still is the main opposition party in the parliament, and is fundamentally opposed to the headscarf in the public domain. In 2008, CHP applied to the Constitutional Court to cancel constitutional amendments that would have allowed women to wear a headscarf in universities. The ruling party in Diyarbakir was the Social Democratic People's Party (SHP) which also supports the headscarf ban. In 1990, the leader of the former SHP submitted an application to the Constitutional Court to terminate a newly accepted law that aimed to lift the ban in universities. The CHP and SHP are closely linked: the founding leader of SHP returned to his former party CHP in the 2009 local elections and was a mayoral candidate in Ankara. The Justice and Development Party (AKP) was in power in the other seven cities in the sample. This Islamist party is against the headscarf ban and has made several attempts to lift it.

CHP and SHP predominantly promote the interests of people from the urban areas and the upper middle classes while AKP primarily advocates the interests of people from the lower social classes and the rural regions (Sayari and Esmer 2002; Carkoglu and Kalaycioglu 2007). CHP was established during the foundation of the Republic of Turkey (Lewis 1961; Sayari and Esmer 2002; Carkoglu and Kalaycioglu 2007) and is the bulwark of the Kemalist elite (Ozyurek 2008).

We expect that the veiled women in our sample will have lower levels of education and labour market participation in the cities with CHP and SHP governance than where AKP rules. Parties against the ban could create opportunities for veiled women in their municipalities by partly or fully ignoring the ban whereas parties supporting the ban are more likely to strictly implement it. But managers in state and private sector institutions in cities with CHP and SHP governance may implement the ban more tightly, resulting in lower educational and economic attainments for veiled women.

In fact, Table 8 shows that the educational attainment of women in the sample is lower in CHP and SHP cities than where AKP is in power. The proportion of women with a university education is substantially higher under AKP rule (49.3 percent) than under CHP or SHP governance (34.8 percent). The dropout rate is also lower in AKP cities (2.4 percent) than in CHP and SHP cities (6.7 percent).

In terms of employment, Table 9 shows that veiled women in cities with CHP or SHP governance are significantly worse off than those under AKP governance: 63 percent of veiled women in AKP cities are active in the labour market; this figure drops to 41 percent in CHP and SHP cities, and overall unemployment is 12 percent higher in CHP and SHP cities (Izmir and Diyarbakir) than in AKP cities. These results support our claim that veiled women are obstructed in their bid to climb the social ladder by the higher social classes who seek to protect their privileged positions.

This study seeks to determine how and to what extent the headscarf ban affects the social and economic conditions of women in Turkey. To this end, we have used self-reported information about the impact of the ban and analysed the indicators of women's educational and labour market outcomes. The data support our expectation that the social and economic conditions of women facing or having faced the ban are negatively affected. More specifically, unveiled women have a significantly higher level of education than veiled women, and veiled women have difficulties in the labour market.

Since it only targets women, the ban has been an effective means by which to reproduce social, economic, regional, and mainly gender inequalities. We argue that the ban on the headscarf serves multiple mechanisms that benefit the elite and exclude lower classes from resources and privileged positions. This claim is supported by the data which show lower educational and economic attainment in cities with a ruling party supported mainly by the highly educated upper classes and against the veiling of women in the public domain (CHP in Izmir and SHP in Diyarbakir) than in cities with a ruling party supported by less well educated lower classes and against the ban (AKP in all other cities).

The data have some limitations. For example, they do not include the women from villages, towns, and cities who have not had an opportunity to participate in educational and economic activities (Rankin and Aytac 2008), and this constitutes the majority of women in Turkey. The ban may contribute to the disinclination of parents in the outlying regions to invest in the education of their daughters. Rankin and Aytac (2008) show that in Turkey, fathers expecting their daughters to wear a headscarf are less likely to send them to school. Even if families from the villages without a secondary or even a primary school spend time, effort, and money sending their daughters to school in towns and cities, the labour market prospects for these girls are not promising; it is even worse if they refuse to unveil. Parents are not encouraged to invest in their daughters when their future prospects are not good. They are kept out of the educational system and married off at an early age.

Nor is there information on women who have fully withdrawn from society. Women in the sample indicated that they are tired of the headscarf issues and have lost faith in institutions, as well as friends and family. Arguably, recruiting respondents via civil society organizations means that we underestimate the negative effect of the ban, and this effect might have been larger or stronger had we used a probability sample. We cautiously argue that women in this study are more engaged in life and take more part in social interactions (as do respondents to all surveys), and therefore experience lighter penalties as a result of the ban or can better cope with the consequences of the ban than those who withdraw from social life. That is, the negative effects of the headscarf ban might actually be greater than is reported. Nevertheless, these large-scale data and study are unique in that they give insight into the lives of the women themselves.

The negative effects of the ban have remained covert, partly because of the lack of large-scale data and partly because of the disinclination of social scientists to tackle this politically-loaded issue. In 2005, a report was accepted by the European Parliament that investigated the barriers to women's participation in social, economic, cultural, and political life in Turkey. This report makes no reference to the headscarf ban, nor does it question its effects on women's activities. After protests from women's organizations, a later European Parliament report suggests examining the effects of the headscarf ban on women in Turkey, thus tacitly admitting the possibility of a negative relationship between the headscarf ban and veiled women's social and economic conditions.

In recent years, some European societies, including France and Germany, have implemented a headscarf ban, while others are considering it. Their prohibition of the headscarf has attracted much attention, resulting in a voluminous literature (Judge 2004; Giannone 2005; Thomas 2006; Body-Gendrot 2007; Joppke 2007; Bruck 2008; Hancock 2008; Rottmann and Ferree 2008; Wiygu 2008). Interestingly, to our knowledge, none of these studies focuses on the impact of the ban on the women. We encourage scholars to investigate the effects of the ban on women in Europe, and we hope that policy makers find enough evidence in this study to reject proposals to prohibit the veil.

We are grateful to HAZAR, Education and Solidarity Association, for making use of their data.

Ayse Guveli holds a RCUK Academic Fellowship on migration, work and life course in the Department of Sociology and Institute for Social and Economic Research (ISER) Centre, University of Essex. Before joining the University of Essex, she studied at ICS in The Netherlands. She developed a project on the changing class structure in post-industrial societies, which resulted in a number of articles articulating the significance and value of the ‘new’ classes for stratification research, as well as a book on the subject. In recent research her interests have focused on religion and the social mobility of immigrants. Dr Guveli is PI for a major study of the Turkish diaspora in Europe funded by NORFACE.