Gender impartiality of public institutions, religiosity, and satisfaction with democracy: findings from Turkey

ABSTRACT Previous literature demonstrated that gender inequality is a major challenge for democratic consolidation. However, research that studies the effect of gender inequality on citizens’ satisfaction with democracy is limited. This study contributes to this literature by exploring the relationship between citizens’ perceptions of gender impartiality of public institutions and satisfaction with democracy in Turkey, where gender inequality is an acute problem. Analysis of a nationally representative survey showed that the perception of gender impartiality of public institutions is a major factor explaining citizens’ satisfaction with democracy. Results also revealed that perceptions of gender impartiality of public institutions affect citizens’ evaluations of the long-term performance of democracy. Another finding is that religiosity moderates the effect of perceptions of gender impartiality of public institutions on citizens’ satisfaction with democracy. We conclude that gender inequality is not a peripheral issue to democratic consolidation in Turkey but a social problem that remains at the heart of it.


Introduction
Impartiality of public institutions is a measure of the quality of a democracy and, therefore, is an important determinant of citizens' satisfaction with as well as their support for democracy. 1 Previous research studying the ways in which impartiality of public institutions affects citizens' attitudes towards democracy does not distinguish among different types of impartiality. 2 In particular, the relationship between perceptions of gender impartiality of public institutions and satisfaction with democracy has been ignored. This study fills this gap and explores whether the extent to which public institutions such as the state, courts, and the police treat men and women equally shapes citizens' satisfaction with democracy in Turkey. It also investigates whether perceptions of gender impartiality of public institutions influence citizens' comparisons of the performance of democracy between the past and the present. This comparison will contribute to our understanding of the relationship between gender equality and democracy by revealing the significance that citizens attach to gender impartiality of public institutions in their evaluations of the long-term performance of democracy.
Gender impartiality of public institutions is an aspect of gender equality, which is, according to many scholars, significant for consolidation of democracy. 3 Hence, studying gender impartiality of public institutions and democracy in tandem is important for countries where gender inequality is a persistent problem. In Turkey, gender relations have been dramatically asymmetric, and democracy has failed to consolidate despite the country's long history with the multiparty system. Turkey is one of the most gender unequal countries in the world, 4 and the majority of Turkish citizens think that men and women do not have equal rights in the country. 5 Since the foundation of the Republic in 1923, women in Turkey have been exposed to inequalities in social, economic, and political life that are mostly, but not always, produced and reproduced by the state. Turkey is thus a good case to explore how citizens' perceptions of gender impartiality of public institutions affect their satisfaction with democracy. Previous research demonstrating that gender equality attitudes are important in understanding citizens' views of democracy, especially in Muslim societies, 6 also make Turkey, which is a predominantly Muslim society, an important case to investigate. Attitudes towards gender equality and democracy are also linked to religiosity. Accordingly, our study explores the impact of religiosity on citizens' perceptions of gender impartiality of public institutions, which in turn may affect their satisfaction with democracy. Our study therefore contributes to the literature on religion, gender equality, and democracy.
Our findings suggest that perceptions of gender impartiality of public institutions are a major determinant of citizens' satisfaction with democracy as well as their evaluations of the long-term performance of democracy between 2005 and 2015 in Turkey.
In what follows, we first summarize the past research on the effects of impartiality of institutions and gender equality attitudes on citizens' evaluations of the political system, more specifically of democracy. This is followed by a review of the literature on the impact of religiosity on attitudes towards gender equality and democracy. We then give an account of the main issues in Turkey with respect to gender inequality. Lastly, we explain the research methodology and findings and end with a discussion on our findings' broader implications.

Impartiality and fairness of public institutions and attitudes towards democracy
Impartiality and fairness of public institutions is an important feature of the ideal modern state. 7 Rothstein and Teorell argued that the most important indicator of the quality of government is impartiality in the exercise of public power and defined impartiality as follows: 8 "When implementing laws and policies, government officials shall not take into consideration anything about the citizen/case that is not beforehand stipulated in the policy or the law". 9 Similarly, Galbreath and Rose suggested that impartiality is "the application of rules in the same way to everyone with the same characteristics". 10 Therefore, impartiality is a norm about the procedures followed during the implementation of the policies rather than about the content of specific policies. 11 Citizens attribute great importance to the impartiality and fairness of public institutions. 12 Indeed, there is a strong relationship between impartiality and fairness of public institutions and citizens' support for the system as well as their perceptions of system legitimacy. 13 This is especially true in democracies where policies made in a procedurally fair manner are considered more legitimate than policies that violate the principles of procedural fairness. 14 Impartiality of public institutions therefore has a positive effect on citizens' satisfaction with and support for democracy. 15 Poor evaluations of the fairness of public institutions have a negative effect on citizens' satisfaction with democracy not only in developing and newly democratized countries but also in advanced democracies. 16 In fact, impartiality of public institutions is more important than other factors such as electoral design and ideological congruence in determining citizens' satisfaction with democracy. 17 Impartiality of public institutions moderates the effect of other variables on citizens' perceptions of the political system. Positive evaluations of impartiality reduce the strength of the relationship between economic evaluations and satisfaction with democracy 18 as well as the effect of being a political loser on citizens' support for democracy. 19 The relationship between socio-tropic evaluations of economy and government approval as well as the positive effect of GDP growth on citizens' electoral support for incumbents are also moderated by their perceptions of the impartiality of public institutions. 20

Gender equality attitudes and democracy
Gender is an important factor in determining citizens' general attitudes towards democracy. There are differences between men and women in terms of the features that they emphasize in their conceptualizations of democracy 21 as well as in their satisfaction with and support for democracy. 22 In contrast to the high volume of research on how one's gender affects attitudes towards democracy, research on how gender equality perceptions affect attitudes towards democracy is limited. However, gender equality is one of the most important values of modern democracies. 23 Support for gender equality is an aspect of a broader process of cultural change that is transforming industrialized societies and contributing to democratization. 24 While the level of democracy in a country does not guarantee gender equality, the emphasis on gender equality contributes to the level of democracy. 25 Accordingly, a complete and unbiased democracy is not possible without gender equality 26 and countries that exclude women from public life are less likely to be democratic. 27 For example, there is a positive relationship between the percentage of officialdom occupied by women in a country and its Freedom House score. 28 Gender equality attitudes are also the best indicators of liberal democratic values such as tolerance 29 , and those who support gender equality are more likely to support liberal democracy. 30 Previous research showed that citizens' gender equality attitudes and their evaluations of gender equality in their country shape their attitudes towards democracy. 31 For example, in countries where gender inequality in social institutions is large, women's attitudes towards democracy is negative. 32 However, in countries where social institutions are more favourable towards gender equality, women hold more positive attitudes towards democracy. 33 Literature suggests that the effect of gender equality attitudes on attitudes towards democracy in Muslim countries is strong. 34 Indeed, the lack of gender equality has been indicated as one of the major factors in explaining the low levels of democracy in Muslim countries. 35 This is because on average, citizens in Muslim countries are less supportive of gender equality than those in Western or non-Muslim countries. 36 Accordingly, the negative relationship between democracy and Islam partly works its way through negative attitudes towards gender equality in these countries. 37 Religiosity and attitudes towards democracy and gender equality Religiosity shapes citizens' attitudes towards democracy. 38 Religiosity is considered incompatible with democracy due to its association with opposition to change, intolerance, and desire for order. However, religiosity also has a positive effect on democracy, as it encourages involvement in social religious networks. 39 Therefore, religiosity has a double-sided effect on attitudes towards democracy. 40 Indeed, whereas some research showed a positive influence of religiosity on democratic attitudes such as support for and satisfaction with democracy, others showed a negative influence of religiosity. 41 For instance, research found that religious people are more likely to be satisfied with democracy 42 while they are less likely to express support for democracy. 43 Other research however revealed that higher levels of religiosity did not result in lower levels of support for democracy among the believers of either Islam, Catholicism, Protestantism, and Orthodox Christianity. 44 Some research even showed a positive relationship between religiosity and support for democracy in some countries. 45 Religiosity is also linked to traditional gender attitudes. 46 Research demonstrated the negative influence of religiosity on citizens' attitudes towards gender equality, 47 even among women. 48 Especially, Islam is considered more patriarchal than other religions. 49 For example, in Muslim countries such as Saudi Arabia and Kuwait, religious citizens have more inegalitarian gender attitudes. 50 In Turkey, too, religious citizens hold more inegalitarian gender values and they defend more traditional roles for women. 51 Another line of research challenges this particular emphasis on Islam's negative influence on gender equality attitudes. 52 Accordingly, religiosity has a negative influence on individuals' gender equality attitudes regardless of their religious affiliation, including Buddhism and Christianity. 53 For example, religious Christians and Jews are also shown to have more inegalitarian gender attitudes. 54 Furthermore, higher level of religiosity is associated with higher levels of sexism not only for Muslims but also for Christians and Jews. 55 After the review of the relationship between religiosity and attitudes towards democracy and gender equality, the next section turns its attention to gender inequality in Turkey.

Gender inequality in Turkey
After the foundation of the Republic in 1923, the new state elite implemented a series of reforms that elevated the status of women in Turkey. However, these reforms, which included abolishing polygamy, adopting a new inheritance law, and giving women the right to elect and to be elected, did not aim at destroying the patriarchal structure in Turkey. 56 This is because the Republican elite considered women not as independent and equal citizens but as key members in the family, who would bear and educate the future generations for the nation. 57 This approach towards women has hardly changed throughout modern Turkish history. After the Justice and Development Party (JDP) came to power in 2002, the family has gained even more significance. 58 The JDP government has considered women as "devoted care providers and blessed mothers" within the framework of sacred familialism 59 and it has encouraged women to be full time stay-at-home caregivers through various policies. 60 Therefore, neither the early Republican elite nor the current political elite recognized the equality between men and women.
Women in Turkey suffer from economic, social, and political forms of inequality stemming from structural problems, the foremost of which is the patriarchal nature of the state and the society. Indeed, Turkey has remained one of the most gender unequal countries in the world. 61 This manifested inequality has increased over the years. Turkey was placed 105th in the World Economic Forum's Global Gender Gap Index in 2006, but it is currently placed 130th in the world. 62 Violence is one of the most important problems for women in Turkey. 63 Femicide in Turkey (see Figure 1 below) has persisted and worsened over the years. In 2014, 36% of women suffered physical abuse from their husbands or boyfriends while another 12% experienced some form of sexual abuse from husbands or boyfriends. 64 Despite this bleak picture, in March 2021, Turkey announced its withdrawal from the 2011 Istanbul Convention, which is an international legal framework aiming to protect women from all forms of violence.
Women in Turkey experience inequality in terms of access to social, economic, and political resources. Women's participation in the labour force stands at a mere 32.9%, whereas the average for OECD countries is above 44%. 65 Another striking figure is that 54% of women in Turkey do not have a bank account in their name, 66 an important indicator of their lack of economic independence from men. Women in Turkey suffer from inequality in terms of access to education as well; 56% of males in Turkey have a high school level education or above while this figure is 47% for females. Women's participation in politics remains quite low in Turkey. While 24.3% of the MPs in the world are women, 17.3% of the MPs in Turkey are women. This figure places Turkey 119th among 192 countries. 67 In the most recent municipal elections in March 2019, only four women mayors were elected out of 81 available provincial seats. Once they are in office, women mayors in Turkey suffer from the dominance of men in political seats in provincial councils, the dominance of masculine discourse in politics, and the burden of traditional gender roles. 68 Not surprisingly, 74% of the Turkish society thinks gender inequality is a major problem in Turkey. 69 Furthermore, gender equality is an essential component of people's attitudes towards democracy in Turkey. Research showed that 40% of people in Turkey included gender equality and women's problems such as women's freedoms, femicide, gender discrimination in the justice system, and the headscarf bans in the past in their conceptualizations of democracy. 70

Research hypotheses
In light of the literature summarized above, we suggest that whether public institutions treat men and women equally is an important aspect of gender equality, which is one of determinants of citizens' attitudes towards democracy. Impartiality of public institutions, which requires these institutions to treat citizens equally without any concern for their ethnic, religious, gender, and socio-economic characteristics, has an impact on citizens' evaluations of democracy. Given that Turkey is a country where gender inequality is a persistent problem, we explore whether perceptions of gender impartiality of public institutions, as an indicator of gender equality, influence citizens' satisfaction with democracy. We hypothesize that: H 1a: Citizens with more negative perceptions of gender impartiality of public institutions are less satisfied with democracy in Turkey.
The conventional measure for satisfaction with democracy provides only a snapshot of citizens' present satisfaction with democracy, but it does not allow citizens to express their comparative evaluations of the performance of democracy between the past and the present. This study measures citizens' comparisons of the performance of democracy in Turkey between 2005 and 2015. As the JDP has been in power uninterruptedly since 2002, our measure allows us to determine whether certain groups have better (or worse) evaluations of the present performance of democracy when compared to its performance ten years ago. Turkey has not experienced significant improvement in regard to gender equality and women have continued to suffer from social, economic and political inequalities since the JDP came to power. We therefore expect that those who have negative opinions about gender impartiality of public institutions have more negative evaluations of the performance of democracy in Turkey between 2005 and 2015. We hypothesize that: While some research showed a positive effect of religiosity on attitudes towards democracy, 71 others showed a negative effect. 72 It is therefore important to understand how religiosity affects citizens' satisfaction with democracy in Turkey, where individuals have become religiously more conservative since the 1990s. 73 The JDP has promoted a conservative religious identity 74 and implemented policies that are in line with religious people's expectations. For instance, the JDP government has heavily invested in the infrastructure of religious vocational schools (imam hatips), 75 and removed the headscarf ban, which had previously barred women from the higher education and employment in the public sector. These and similar policies have increased religious conservative people's support for the JDP. In the most recent general elections, 66% of the people defining themselves as "religious conservative" voted for the JDP while only 14% of people who defined themselves as "modern" voted for the JDP. 76 Past research demonstrated that citizens' proximity to the political elite or parties on ideological and policy positions increases their satisfaction with democracy. 77 We therefore expect religious people to be more satisfied with the current state of democracy and to have more positive evaluations of the performance of democracy in Turkey between 2005 and 2015. We hypothesize that: According to Fish, 78 low levels of gender equality in Muslim societies may account for the link between authoritarianism and Islam. Similarly, Spierings argued that the negative relationship between religiosity and support for democracy in Muslim societies works through the negative relationship between religiosity and gender equality. 79 In Turkey, increasing religiosity in both the state and society has promoted more traditional roles for women in Turkey 80 and religious citizens hold less egalitarian gender values. 81 This in turn may influence the significance that religious citizens attach to gender equality in their evaluations of democracy. We therefore explore whether religiosity moderates the effect of perceptions of gender impartiality on citizens' satisfaction with democracy. We hypothesize that: H 3a : In Turkey, the negative relationship between perceptions of gender impartiality of public institutions and satisfaction with democracy is weaker among more religious people.
A similar effect can be observed in citizens' evaluations of the performance of democracy in Turkey between 2005 and 2015. We hypothesize that: H 3b: In Turkey, the negative relationship between perceptions of gender impartiality of public institutions and evaluations of the performance of democracy between 2005 and 2015 is weaker among more religious people.

Data
This study used data from a national survey, which was conducted between March and May 2015. The survey, which measured people's perceptions of democracy and the rule of law in Turkey, included face-to-face interviews with 1,804 people who were at least 18 years old. Survey participants were randomly selected across 12 regions of Turkey in accordance with the NUTS-12 system of the Turkish Statistical Institute (TUIK). The number of participants in each region was determined in proportion to that region's population. TUIK provided the researchers with 180 randomly sampled geographical areas in Turkey. Interviewers completed 10 surveys from each sampling unit, starting with the beginning address that was randomly determined by TUIK. After completing the first survey, the interviewer visited every three addresses to the right for the next survey until 10 surveys were completed in every sampling unit.
We used two major sources to write the survey questions: (1) face-to-face interviews conducted with 60 people in five cities (Istanbul, Kayseri, Adana, Trabzon, and Diyarbakir) and (2) the following surveys: the World Values Survey and World Justice Project Survey. In face-to-face interviews, participants answered questions about their perceptions of democracy in Turkey as well as their conceptualizations of democracy. A certain number of questions in the survey were formulated based on patterns detected in the interviews. We finalized the questions after analysing the results of the pilot survey, which we conducted with 100 participants in five cities in Turkey.

Dependent variables
This study uses two dependent variables. The first dependent variable measures citizens' satisfaction with democracy during the time of the survey. We used the following question: "How well do you think democracy works in Turkey?" Participants were asked to choose a number between 1 (very bad) and 5 (very good/well). We named this variable Present SWD.
The second dependent variable measures citizens' comparative evaluations of the performance of democracy between 2005 and 2015. We used the following question that asks the participants to compare the current state of democracy with the state of democracy 10 years ago (2005) in Turkey: "Do you think that democracy in Turkey, compared to 10 years ago … ?" Participants were asked to choose a number between 1 (fares much worse) and 5 (fares much better). We named this variable Comparison SWD.

Independent variables
The first independent variable measures citizens' perceptions of gender impartiality of public institutions. We used the following question: "Some argue that women are treated unfairly compared to men. Do you think that the institutions listed below treat women as they would treat men?" Participants were asked to choose a number between 1 (no equal treatment at all) and 5 (completely equal treatment) for the following three institutions: the government, the police, and the judiciary. Reliability analysis showed that these items had a high Cronbach's alpha (0.889) and formed a single dimension when entered into factor analysis. We first reverse coded the variables so that higher scores denote more negative perceptions of gender impartiality of public institutions. Then, we implemented a simple summation procedure and divided the resulting variable by three to create a scale between 1 and 5. We named this variable Gender impartiality.
The second independent variable is religiosity. We used the following question: "How often do you pray?" Participants were asked to choose a number between 1 (every day) and 6 (never). We reverse coded this variable so that higher scores denote higher levels of religiosity. We named this variable Religiosity.
The third independent variable explores the moderation effect of religiosity on the relationship between perceptions of gender impartiality of public institutions and the two dependent variables. To this end, we used an interaction term where we multiplied the Gender impartiality by Religiosity.

Control variables
We used the following socio-demographic indicators as control variables: Age (between 18 and 91), Female (female = 1), Education (1 = no schooling; 2 = 5-year degree; 3 = 8year degree; 4 = high school degree; 5 = higher education degree), and Household expenditure (1 = under 500 Turkish Lira (TL) per month; 2 = 500-1000 TL per month; 3 = 1001-1500 TL per month; 4 = 1501-2500 TL per month; 5 = 2501-3500 TL per month; 6 = 3501-5000 TL per month; 7 = 5001-7000 TL per month; 8 = over 7000 TL per month. Previous research demonstrated that citizens' perceptions of individual economic well-being have an influence on their satisfaction with democracy. 82 To measure this effect, we used the following question: "Which income group do you think you belong to when you consider the average income level in Turkey?" (1 = lower class; 5 = upper class). We named this variable Perceptions of economic well-being.
Research showed that corruption is negatively correlated with satisfaction with democracy. 83 To measure citizens' perceptions of corruption in Turkey, we used the following statement: "In Turkey, corruption and bribery are not common problems" (1 = Completely disagree; 5 = Completely agree). We reverse coded this variable so that higher scores denote higher perceptions of corruption in Turkey. We named this variable Corruption.
The positive effect of being a political winner on one's satisfaction with democracy is a well-established finding in the literature. 84 We measured being a political winner with the following question: "If there were general elections next Sunday, which political party would you vote for?" Participants who expressed their intention, or indicated their inclination, to vote for the incumbent JDP were coded as "1". Participants who expressed their intention, or indicated their inclination, to vote for opposition parties were coded as "0". We named this variable Political winner.
Ethnic identity has an effect on citizens' satisfaction with democracy. 85 The Kurdish question is a long-standing issue in Turkey that has exacted a heavy toll on the Kurdish minority. 86 We therefore included ethnic identity as another control variable in the analyses. We used the following question: "Which ethnic group do you belong to?" Those choosing "Kurd" were coded as "1", while those choosing "Turk" were coded as "0". As this variable measures the differences between Kurds and Turks in terms of their satisfaction with democracy, participants choosing options other than Turkish or Kurdish were coded as "system missing". We named this variable Kurdish identity.
Whether citizens define democracy in substantive or procedural terms shapes their satisfaction with democracy. 87 Between 2002 and 2015, the economic growth averaged 5.94% and GPD per capita (PPP) increased from $9,215 to $25,626 in Turkey. 88 We therefore expect that citizens who prioritize outcomes rather than procedures in their understandings of democracy would be more satisfied with democracy in Turkey. To control for this effect, we used the following statement: "If a country is economically more developed compared to 10 years ago, it means that democracy has improved in this country" (1 = Completely disagree; 5 = Completely agree). We named this variable Economic outcome.

Findings
Descriptive statistics Table 1 gives the means of the variables used in the analyses. Prominent findings are as follows: The means for Present SWD and Comparison SWD are 2.73 and 2.97, respectively. The mean for Gender impartiality is 3.23 (mean for females = 3.21; mean for males = 3.26), which shows that generally speaking, in Turkey, citizens' perceptions of gender impartiality of public institutions are negative regardless of gender. The mean for Religiosity is 3.85, indicating that religiosity in Turkey is relatively high. Table 2 shows mean scores of categorical variables for dependent variables. One particular finding is the stark contrast between Kurds' and Turks' means for Present SWD (2.39 and 2.82, respectively), and for Comparison SWD (3.24 and 2.91, respectively). Kurds' comparisons of the performance of democracy between 2005 and 2015 are even higher than those of Turks, a finding with implications that will be elaborated upon in the discussion section. JDP voters' means of Present SWD (3.59) and Comparison SWD (3.79) are higher than those of the opposition party voters (Present SWD mean = 2.2; Comparison SWD mean = 2.46). Table 3 displays the findings of four different models. In models 1 and 2, the dependent variable is Present SWD whereas Comparison SWD is the dependent variable in models 3 and 4. Models 2 and 4 include the interaction term, which explores whether religiosity moderates the effect of citizens' perceptions of gender impartiality of public institutions on our dependent variables.

Multivariate regression analysis
In line with our expectations, there is a negative relationship between Gender impartiality and Present SWD in models 1 and 2. Citizens who think that public institutions are not treating women impartially are less satisfied with democracy in Turkey. This finding supports our first hypothesis (H 1a ). There is a positive relationship between Religiosity and Present SWD in model 1. However, this effect disappears after the inclusion of the interaction term in model 2. This finding therefore does not support our second hypothesis (H 2a ). The relationship between the interaction term, Religiosity*Gender impartiality, and Present SWD is positive in model 2. That is, the relationship between citizens' perceptions of gender impartiality of public institutions and their satisfaction with democracy is weaker among more religious citizens than it is among less religious citizens. This finding supports our third hypothesis (H 3a ) and suggests that religiosity moderates the effect of citizens' perceptions of gender impartiality on their satisfaction with democracy. Models 1 and 2 also demonstrate the positive relationship between Perceptions of economic well-being and Present SWD. That is, citizens' thinking that they are doing economically better than other citizens are more satisfied with democracy. We also found a positive relationship between the Political winner and Present SWD as well as between the Economic outcome and Present SWD in both models. Those who associate democratic progress with economic development and those who vote for the JDP are more satisfied with democracy in Turkey. There is a negative relationship between Corruption and Present SWD. Citizens with higher perceptions of corruption in Turkey are less satisfied with democracy. There is no relationship between Kurdish identity and Present SWD. This finding indicates that, contrary to our expectations, Kurds are not less satisfied than Turks with the current state of democracy in Turkey.
In models 3 and 4, there is a negative relationship between Gender impartiality and Comparison SDW. Citizens with negative perceptions of gender impartiality of public institutions are less likely to see an improvement in the democratic performance in Turkey between 2005 and 2015. This finding supports our first hypothesis (H 1b ). There is a positive relationship between Religiosity and Comparison SDW in model 3. Similar to model 2, this variable loses its significance after the inclusion of the interaction term, Religiosity*Gender impartiality, in model 4. This finding does not provide support for our second hypothesis (H 2b ). The relationship between the interaction term and Comparison SDW in model 4 is positive and significant. The relationship between perceptions of gender impartiality of public institutions and citizens' evaluations of the democratic performance between 2005 and 2015 is weaker among more religious citizens than it is among less religious citizens. This finding supports our third hypothesis (H 3b ) and indicates that religiosity moderates the effect of perceptions of gender impartiality on citizens' comparison of the democratic regime between 2005 and 2015.
There is negative relationship between Age and Comparison SDW. Older people are less likely to be contented with the performance of democracy in the last 10 years. The Female variable has a negative effect in models 3 and 4. Women are less likely to think that democracy has improved between 2005 and 2015. We found a positive relationship between Economic outcome and Comparison SDW while the relationship between Corruption and Comparison SDW is negative. Citizens with high perceptions of corruption in Turkey are less likely to see improvement in the performance of democracy in Turkey. Perceptions of economic well-being retain its positive relationship to Comparison SDW. Citizens thinking that they are doing economically better than others are more likely to think that democracy has improved between 2005 and 2015. The relationship between Kurdish identity and Comparison SDW is positive and significant in models 3 and 4 while it is insignificant in models 1 and 2. That is, Kurds are more likely than Turks to see an improvement in the performance of democracy between 2005 and 2015.

Discussion
Previous research extensively explored how citizens' perceptions of impartiality of public institutions shape their attitudes towards democracy. 89 However, there is no research on how citizens' perceptions of gender impartiality of public institutions affect their satisfaction with democracy. This study used Turkey as a case to explore whether citizens' perceptions of gender impartiality of public institutions shape their satisfaction with democracy and their evaluations of the long-term performance of democracy. We found that in Turkey, citizens who think that public institutions do not treat men and women equally are less satisfied with democracy. These citizens are also less likely to think that the performance of democracy has increased between 2005 and 2015. Therefore, citizens' perceptions of gender inequality, defined in terms of perceptions of gender impartiality of public institutions in this research, are a major factor behind their dissatisfaction with democracy in Turkey. Based on these findings, we argue that one of the fundamental flaws in Turkish democracy is the lack of concern for persisting gender inequality and therefore democratic consolidation in Turkey is not possible without establishing the principle of gender equality. In this respect, the government decision to withdraw Turkey from the 2011 Istanbul Convention in March 2021 is particularly concerning, as this convention is an international legal framework aiming to protect women from all forms of violence including domestic violence. Withdrawal from the Istanbul convention may exacerbate violence against women in Turkey and deepen the existing forms of gender inequality, eventually causing even more concern for the quality of democracy in Turkey.
We also found that the impact of religion on satisfaction with democracy in Turkey occurs through its effect on gender equality attitudes. That is, for religious people, whether public institutions treat women impartially is less important in determining their satisfaction with democracy and their evaluations of the performance of democracy between 2005 and 2015. We therefore suggest that not religiosity itself, but the effect of religiosity on gender equality attitudes shapes citizens' satisfaction with democracy in Turkey. This finding supports previous research arguing that religiosity and attitudes towards democracy in Muslim societies works through the negative relationship between religiosity and gender equality. 90 However, as the negative relationship between religiosity and gender equality attitudes does not pertain only to Muslim societies, 91 there is a need for research on religiosity's effect on the relationship between gender equality attitudes and attitudes towards democracy in non-Muslim societies.
Gender does not have any significant effect on citizens' present satisfaction with democracy. However, women in Turkey are more likely than men to think that the quality of democracy has declined between 2005 and 2015. The increase in the number of women murdered by their immediate family or boyfriends 92 and the government's reluctance to eliminate the inequalities between men and women in social, economic, and political institutions might have contributed to women's negative opinions about the performance of democracy in this period. This finding is in line with previous research that suggests that women have more negative attitudes towards democracy where gender inequality in social institutions is large. 93 Similarly, although there is no difference between Kurds and Turks in terms of their present satisfaction with democracy, Kurds have more positive evaluations of the performance of democracy between 2005 and 2015 than Turks. 94 This result, which is unexpected given the Kurds' problematic history with the Turkish state, probably demonstrates the effect of the JDP's reconciliatory policies, which lasted until July 2015. For example, after coming to power in 2002, the JDP abolished the state of emergency, which had paralyzed Kurdish populated cities since 1987, and removed some of the restrictions on the use of Kurdish language in broadcasting and to a limited extent in education. The Peace Process, the latest round of negotiations aiming to bring about a peaceful solution to the Kurdish question in Turkey, was still ongoing when our survey was conducted in 2015. This factor might also have caused Kurds to express more satisfaction when asked to evaluate the performance of democracy between 2005 and 2015.
Democracy does not exist if women are not equal to men in the public realm. 95 This is because the quality of democracy is not only about its institutions but also about the ability of different social groups to participate in these institutions. 96 Despite this fact, the focus of the JDP government and the opposition parties in their discussions about democracy has been mostly limited to procedural matters (e.g. election law, the constitution) while issues regarding gender inequality have remained peripheral to these discussions. Hence, the lack of concern about how women are treated by public institutions has been a major weakness of discussions on democratization in Turkey. Without problematizing gender inequality and eliminating its worst excesses in Turkey, neither democratic consolidation nor increasing citizens' satisfaction with democracy is possible. In this respect, empowering women through laws geared towards gender equality, encouraging women' economic independence, and pressuring the state to re-enact the Istanbul Convention are just some of the actions that could be taken to improve citizens' satisfaction with as well as the quality of democracy in Turkey.