Examining the relationship between women’s descriptive political representation and women’s possibility to participate in civil society across regime types

ABSTRACT

This study examines the link between women’s representation in national parliaments and their possibility to participate in civil society. Utilizing panel data from the Varieties of Democracy (V-Dem) dataset, covering 1975–2019, we investigate how the presence of women in parliaments influences the possibility for women to participate in civil society, especially in non-liberal democratic settings. The results indicate a positive association post-1990, coinciding with an international emphasis on civil society and a rise in women’s parliamentary representation in both liberal democracies and autocratic regimes. The study however highlights that in hybrid and autocratic regimes, increased female representation can also be used to project a gender-equal image without genuine democratization intentions. These findings suggest a correlation between the increase in women in parliaments and greater opportunities for their participation in civil society, but do not imply that the results reflect processes towards democracy. The findings contribute to understanding women’s possibilities in non-democratic regimes and underscores the need for further qualitative analysis to fully grasp the implications of these trends.

KEYWORDS:

• Women
• descriptive representation
• legislators
• civil society
• regime types

Introduction

Does an increase in women’s representation in parliaments matter for women’s possibility to participate in civil society? Gender equality has been purportedly considered to be an important part of broader processes of cultural change across the globe (Inglehart & Norris, 2003). While women around the world are increasingly entering parliaments as representatives, research on whether there is a link between women’s descriptive political representation and the elimination of barriers for women to participate in civil society remains inconclusive. A number of studies conclude that women’s representation in political institutions has a positive impact on varying aspects of women’s political empowerment (Alexander, 2012; Alexander et al, 2016; Barnes & Burchard, 2013; Norris & Krook, 2009; Wolbrecht & Campbell, 2007). However, other studies find little evidence of such a relationship (Lawless, 2004; Zetterberg, 2009). This indicates that the question is far from settled. The implementation of various forms of gender quotas to guarantee a certain number of parliamentary seats for women (e.g., Bjarnegård & Zetterberg, 2022) presents a new facet to consider.

In this study, our overarching aim is to contribute to this debate by examining the relationship between women’s descriptive representation of national parliaments and their ability to participate in civil society across different political regimes, time periods and existence of gender quotas. To do that we pose the following research question: To what extent does women’s representation in parliaments affect the degree to which women face barriers to participate in civil society organizations?

In the context of the present study, liberal democracies during the most recent decades are not expected to have any formal restrictions that limit the possibility for women to participate in civil society or to have restrictions against civil society organizations that represent women’s issues. Increased access to civil society here means fewer restrictions and limitations on being involved in civil society based on gender and that there are fewer restrictions for women’s organizations to promote gender issues. In fact, the definition of liberal democracy adopted in our data includes personal liberties. Therefore, we focus primarily on outcomes in regimes other than liberal democracies. In fact, some studies suggest that women’s presence in parliament and their possibility of being engaged in civil society may be especially important in contexts other than liberal democracies (Barnes & Burchard, 2013; Bjarnegård & Melander, 2011; Paxton, 1997; Stockemer, 2009; Tripp, 2001). Research on democracy and democratization has underlined the importance of women’s civil liberties for successful democratization processes (Wang et al., 2017).

Our study builds on two main arguments. First, the underlying assumption about the relationship between women’s presence in parliaments and their possibility to substantively represent women’s interests is to a large extent based on conclusions drawn from studies of longstanding Western democracies (Alexander, 2012; Karp & Banducci, 2008; Norris & Krook, 2009). Therefore, further studies that focus on political contexts other than liberal democracies, such as hybrid or authoritarian regimes, are warranted (Ertan et al., 2017; Moussawi & Koujok, 2019; Paxton et al, 2020; Wilde et al., 2018). In the present study, we direct our interest to those contexts that are referred to as hybrid regimes; that is, regimes characterized by some restrictions on core democratic principles and freedoms but with fewer and different features than closed autocracies and closed autocracies. The context of hybrid regimes may be more in flux than that of closed autocracies, and changes within these contexts may have important implications for both women’s opportunities in political decision making and civil society and for democratization processes (Bjarnegård & Melander, 2011; Bjarnegård & Zetterberg, 2014; Fallon et al., 2012; Melander, 2005). The relationship between the presence of women in parliament and substantive representation of women’s interests is especially complex in non-democratic contexts (Bjarnegård & Zetterberg, 2022; Bush et al., 2023; 2014; Krook & Zetterberg, 2014). The increase in women in parliament could both constitute a pathway to democratization and be instrumentalized by authoritarian regimes as ‘gender washing’ (Bjarnegård & Zetterberg, 2022).

Second, there is a growing academic debate on how hybrid or authoritarian regimes have adopted reforms to increase female representation in parliament (e.g., Bjarnegård & Zetterberg, 2022; Clayton et al., 2017; Krook & Zetterberg, 2014). Heightened attention has been given to women’s political representation in the decades following the collapse of communist regimes in Eastern and Central Europe. The period after the end of the Cold War was also followed by an increased focus on civil society (Diamond, 1994). Broad theories on cultural change also often link women’s rights to processes of democratization (Inglehart & Norris, 2003). The increased focus in the 1990s and onwards both on women’s rights and civil society at large. The literature on autocratic regimes have also pointed to the fact that autocratic regimes in many instances have changed their tactics to being less directly based on fear and more on manipulating public opinion (Guriev & Treisman, 2022). All in all, this leads us to anticipate varying effects of the relationship between women in parliament and women’s possibility to participate in civil society before and after the end of the Cold War.

In our analyses, we analyse a period spanning between 1975 and 2019 and use panel data from the Varieties of Democracy (V-Dem) dataset. This data set allows us to investigate within country differences across time. Our approach, which emphasizes the changes within countries over time, enriches the literature that predominantly relies on case studies or cross-sectional comparative data.

Disentangling the Link between Women’s Descriptive Representation and Women’s Participation in Civil Society

Women’s Descriptive Political Representation

Studies have shown that when women are granted civil liberties, the transition processes to democracy tend to be more stable (Wang et al., 2017). An important part of women’s civil liberties is the right to participate in elections and become elected candidates. A way to describe whether women succeed in becoming elected is through descriptive representation. The term descriptive representation derives from Pitkin’s (1967) work The Concept of Representation, which conceptualizes the idea of representation as consisting of four components that can be applied to political life. Among them, descriptive representation is the numerical representation of the population, where representatives mirror the population based on common characteristics, such as gender.

Pitkin conceptualizes descriptive representation in three ways: first, the legislature as a mirror of the population in terms of composition; second, the legislature as representative in terms of its activity instead of its composition; and third, the legislature as an exact copy of its people and acts completely on behalf of them (1967, 81f.). Several studies have investigated the outcomes of increasing women’s representation as elected politicians (e.g., Besley et al., 2017, 2008; Folke et al., 2021; Hessami & Lopes da Fonseca, 2020; Philips, 2005). The potential effects of women’s descriptive representation in politics on other realms of political life have been widely explored (e.g., Bühlmann & Schädel, 2012; Chaney, 2016; Hessami & Lopes da Fonseca, 2020). Importantly, s tudies have also underlined that women do not belong to a homogeneous group (Dovi, 2002). Previous studies point to the implications of increased female representation in many different fields, such as justice (Hughes & Paxton, 2007), development (Stockemer, 2015), political knowledge (Sanbonmatsu, 2003), and policy change (Baskaran & Hessami, 2023; Beckwith, 2007; Childs & Krook, 2009).

Studies have also noted a clear trend in recent decades with increasing shares of women in parliament (Bjarnegård & Zetterberg, 2022; Bush et al., 2023). A key point is that the drivers behind the increasing shares of women in parliament may vary according to regime type (Bjarnegård & Zetterberg, 2022). Recent findings suggest that hybrid and authoritarian regimes are prone to implementing quotas to reserve a certain number of candidates or parliamentary seats for women (e.g., Bjarnegård & Zetterberg, 2022). Concurrently autocratic regimes have been argued to rule less based on fear and more based on creating a distorted view of the political reality (Guriev & Treisman, 2022). A number of autocratic regimes are believed to use the presence of women in parliament as a tool to project a more ‘democratic’ image, particularly directed to external international observers (Bush et al., 2023; Valdini, 2019). The introduction of gender quotas for parliamentary seats in countries such as Rwanda and Saudi Arabia. Moreover, as noted by Bjarnegård and Zetterberg (2022), occurred while civil society organizations that advocated women’s rights in these countries simultaneously faced hostilities from those regimes.

Women’s Possibility of Engaging in Civil Society

Civil society is arguably a broad concept with different and contended definitions and is often defined as the sphere between the state (in a large sense), private enterprise, the family, and the polity (Bernhard et al., 2017). On the other hand, it is also recognized that the state sets preconditions for civil society (through legislation, rules, and regulations); therefore, the types of constraints that are put on civil society organizations may vary considerably (Bernhard et al., 2017).

The connection of civil society to democratization processes became especially pronounced during the 1990s in the immediate aftermath of the fall of the wall of Berlin. International aid often concentrated on projects attempting to strengthen civil society organizations with the hope of strengthening or initiating democratization processes in countries that were transitioning from authoritarian regimes (e.g., former communist countries in Eastern and Central Europe) (Diamond, 1994).

Civil society organizations can constitute an important arena, especially for women: activities in civil society that are less institutionalized are often more accessible to women; civil society organizations may facilitate the mobilization of women more broadly (e.g., Burns et al., 2001; Mudege & Kwangwari, 2013; Stokes, 2005). Particularly during processes of democratic transition, women's activism in civil society has been argued to be an important factor for transitions into stable democratic regimes (Moghadam, 2008; Wang et al., 2017; Waylen, 1994, 2007). Research findings suggest that women’s involvement in civil society organizations has been a crucial component of challenging authoritarian regimes (Wang et al., 2017). Notably, women’s organizations have played a pivotal role as democratizing forces in various countries. For instance, in Latin America, during the 1970s and onwards, women’s civil society organizations were key actors (Wang et al., 2017; Waylen, 2007). Similarly, in the 1990s, women’s organizations in African nations such as Kenya and Morocco were influential (Tripp, 2001; Wang et al., 2017).

However, it must also be acknowledged that women’s possibilities to participate in civil society may be restricted not only through the restriction of civil liberties but also, for instance, by gender norms (e.g., Coffé & Bolzendahl, 2010). In general, women, for instance, tend to have less leisure time than men, and gender norms may also limit what is deemed to be appropriate for women to carry out in the public sphere; this may also impact their possibilities to participate in civil society (Coffé & Bolzendahl, 2010). Women’s freedom to participate in civil society may therefore constitute an important part of widening the possibility for women to act in the public sphere and become politically empowered (Einhorn & Sever, 2003; Hessami & Lopes da Fonseca, 2020).

Explaining the Link between Women’s Descriptive Representation and Women’s Involvement in Civil Society

A descriptive representation of women in legislative bodies such as parliaments is often understood as linked to substantial representation of women’s rights, where elected women also advocate for the social rights of women (Clayton et al., 2019). An increased descriptive representation may therefore also affect how women’s issues are being put on the political agenda (Celis & Childs, 2008). Simply put, women in parliament are assumed to be more willing to advocate for policies that address women’s issues than men are and therefore their numerical representation in parliament matters (Clayton et al., 2019).

Other studies suggest that the presence of women in parliament has an indirect impact and that it can be important for public opinion on women in politics. The presence of women in parliament might matter for beliefs about women’s ability to rule (Alexander, 2012). However, the degree to which female legislators matter may vary across regime types (Alexander, 2012; Paxton et al, 2020 ). For instance, Zetterberg (2009) found that merely adopting gender quotas for women’s political representation was insufficient to positively affect civic engagement among women in general across seventeen Latin American countries. In a similar vein, Karp and Banducci (2008) and Lawless (2004) found only limited evidence of a positive link between increased descriptive representation of women in US Congress and women’s civic engagement. Despite this, both studies acknowledge that women’s presence in parliament may matter for public opinion, such as fostering positive evaluations of female politicians in office.

Furthermore, women’s representation in parliament can increase the legitimacy of women’s rights organizations and their ability to influence policy. This presupposition of the social representation of a group resulting in reforms benefiting the same group is, however, likely context dependent, as already mentioned above. Many of the previous studies examining the link between the descriptive representation of a group and policies benefitting the group that is being represented have been carried out in the context of representative democracies (e.g., Krook, 2010).

The role women`s presence in parliaments in hybrid regimes plays may also differ to that of women in closed autocratic regimes. Hybrid regimes often exhibit a mix of democratic and autocratic features. An increased presence of women in parliament may signal openings for women’s possibilities to become engaged in other spheres of society and hence increase the possibilities for women to be engaged in civil society and to advocate for women’s issues. In closed autocratic regimes, the room for civil society to operate is more restricted, and the interest in increasing women’s presence in parliament is often merely symbolic (e.g., Bush et al., 2023). These regimes may want to differentiate themselves from liberal democracies and signal that they are more progressive in terms of gender issues. This is exemplified by the communist regimes of Eastern and Central Europe during the Cold War, which had more gender-balanced institutions (Alexander, 2012).

Moreover, and importantly, liberal democracies do not necessarily always have higher rates of women as representatives in parliamentary institutions (Stockemer, 2009), nor is democracy necessarily a requirement for descriptive representation (Barnes & Burchard, 2013; Tripp, 2001). During the existence of the Central and Eastern European communist regimes, these regimes tended to have more gender-balanced parliaments than many Western democracies to project an image of being more progressive societies. Moreover, these parliaments had little power as institutions and were not freely elected (Bjarnegård & Zetterberg, 2022). Following the end of the Cold War, reforms that introduced mandatory quotas for women either as candidates or as representatives in parliament became exceedingly popular in autocratic and hybrid regimes (Bjarnegård & Zetterberg, 2022; Bush et al., 2023). In fact, Bjarnegård and Zetterberg (2022) showed that only approximately one-third of the countries that had adopted some form of legislated gender quotas in their parliaments belonged to the category of democratic regimes.

The representation of women in parliament can be used as an instrument by non-democratic regimes to achieve other goals, such as increasing the legitimacy of the regime both for international actors and for their own citizens (Bjarnegård & Zetterberg, 2022; Bush et al., 2023; Valdini, 2019). Hybrid or authoritarian regimes can establish the rules that govern the processes through which women are assigned seats in parliament; hence, they can also strengthen their own power by adding women candidates who are loyal to the regime (Bjarnegård & Zetterberg, 2022; Bush et al., 2023). The presence of women in parliaments of hybrid regimes may therefore be more symbolic, a way of ‘gender washing’ political institutions (Bjarnegård & Zetterberg, 2022). Hence, increased female representation may not necessarily result in reduced obstacles for women to freely participate in civil society or to form civil society organizations that address the civil rights of women.

We can therefore conclude that the increased representation of women in parliament may have varying implications for women’s opportunities to engage in civil society, depending on the type of regime. Autocratic and hybrid regimes might instrumentalize the presence of women in parliament, with little intention to enhance women’s opportunities to participate in civil society. An important caveat is that external observers often link gender representation with democratization and tend to conclude that a regime has become more democratic just by introducing gender quotas (Bjarnegård & Zetterberg, 2022; Bush et al., 2023).

That said, it is fair to conclude that the institutional context matters to a great extent for understanding the expected impact of an increased share of women as Members of Parliament (MPs) on other outcomes. In sum, the literature indicates that the relationship between women in parliament and the consequences this may have for the possibility of women participating in civil society is contingent on the context of the political regime. To reiterate, for the purpose of the present study, we do not expect any significant impact on our dependent variable, women’s participation in civil society, as liberal democracies are by definition not expected to restrict women’s participation or women’s civil society organizations. Based on our reading of the literature, we formulate the following hypotheses:

H1 Women’s representation in parliament is expected to be associated with increased possibilities for women to participate in civil society, especially in the period after 1990.

H2 The association between an increased presence of women in parliament and women’s possibility of participating in civil society is moderated by regime type.

H3 The association between an increased presence of women in parliament and women’s possibility of participating in civil society is mediated by the existence of gender quotas.

In what follows, we describe the data and empirically investigate whether there is any relationship between the share of elected women in lower chambers of parliaments and women’s engagement in civil society and whether this relationship was altered after the collapse of the Wall of Berlin. We also investigate the extent to which regime types moderate this relationship.

Research Design

Dataset and Modelling

This study follows a research design based on quantitative data analysis in which panel data are examined across countries and over time using fixed effects estimations. This study used data from version 12 (March 2022) of the V-Dem dataset, an expert-coded database measuring and conceptualizing democracy and political systems on a global scale that includes 177 countries.

The advantage of using V-Dem data is the use of unique measures of women’s political empowerment, consisting of variables such as women’s participation in civil society, which is used as the main dependent variable in this study. The operationalization of the main study variables is presented later; however, the V-Dem measures appear reasonable for the purpose of this study. As explained in the V-Dem codebook, female participation in civil society reflects whether women have the ability to express themselves and participate in groups (Coppedge et al., 2021, p. 298).

A potential critique of using V-Dem data concerns the expert coding of data, which is based on country experts’ retrospective assessment of variables. However, the comparability of expert-coded data across countries has been tested, and the V-Dem dataset has been shown to be reliable in several assessments (Coppedge et al., 2019; Coppedge et al., 2020; Maxwell et al., 2018).

Given that our data have a panel structure, one observation per country and year, we use fixed effects regression analyses, as we are interested in understanding to what extent a change in the share of women in lower chamber legislatures is associated with a change in women’s possibility to participate in civil society. We use fixed effects estimates with robust standard errors clustered at the country level, as we are interested in changes within countries. The fixed effects estimations used in this study include some strengths. This estimation technique limits the potential problems associated with correlations between unobserved variables and unobserved heterogeneity (Collischon & Eberl, 2020). We also include time fixed effects to control for underlying trends of variables that vary over time. The models are divided into two time periods (1975–1989 and 1990–2019), and the motives for this division are developed below.

Operationalization of Main Variables and Limitations

The data analysed in this study include 137 countries of the world split into two time periods (1975–1989, N = 1468; 1990–2019, N = 3836). in the appendix contains the descriptive statistics of the included variables.

Dependent Variable: Women’s Possibility to Participate in Civil Society

The dependent variable in this study is the expert-coded variable of women’s participation in civil society. Women’s possibility to participate in civil society in this study was measured by using the most relevant aspect of the women civil society participation index (Coppedge et al., 2021, p. 298), namely, the component measuring women’s participation in civil society organizations. This variable corresponds to the following question: are women prevented from participating in civil society organizations? There are five possible answers, ranging from ‘the alternative almost always’ (0) to ‘almost never’ (5) (Coppedge et al., 2021, p. 195). The variable contains the following explanation to the expert coders: ‘Please pay attention to both (A) whether women are prevented from participating in civil society organizations (CSOs) because of their gender and (B) whether CSOs pursuing women’s interests are prevented from taking part in associational life’ (Coppedge et al., 2021, p. 195). Hence, the variable focuses on the exclusion of women from civil society both as individuals and as organizations pursuing the interest of women. The variable was converted to an interval scale by the V-dem team (combining the responses from all coders of the same country) using Bayesian item response theory. The scores range from negative to positive values, thus resembling Z scores. Higher values represent lower levels of exclusion of women. We included only those observations for each country and year where more than three coders contributed data to reduce the influence of coder attrition. By doing so, an important total number of observations (20–30) were obtained from each of Barbados, Fiji, Malta, Singapore, Timor-Leste, and the United Arab Emirates. For three countries (Azerbaijan, Guinnea-Bissau, and Papua New Guinea), one to three observations per country are lost.

Key Independent Variable: Women’s Descriptive Political Representation

The explanatory variable in this study is women’s descriptive political representation in parliaments. The representation of women in parliaments was measured in the standard way by using the percentage of female legislators in the lower or unicameral chamber of national parliaments. The variable ‘lower chamber female legislators’ measures ‘what percentage of the lower (or unicameral) chamber of legislature is female’ (Coppedge et al., 2021, p. 156).

Key Independent Variables: Regime Types and Gender Quotas

Our main variable for categorizing regime typology is based on the ‘regimes in the world’ variable within the V-Dem dataset. There are four different regime categories: closed autocracies, electoral autocracies, electoral democracies, and liberal democracies. We denote electoral autocracies and electoral democracies as hybrid regimes. A closed autocracy is, according to the definition, a regime where there are ‘no multiparty elections for the chief executive or the legislature’. The variable regimes of the world define electoral autocracy as a political regime in which, despite the existence of de facto multiparty elections for the chief executive, a minimum level of the institutional prerequisites of Dahl’s (1971) polyarchy are not fulfilled. However, in electoral democracies, liberal principles – such as respect for personal liberties, the rule of law, and judicial and legislative constraints on the executive – are limited. An electoral democracy is defined as follows (Coppedge et al., 2021, p. 283):

De facto free and fair multiparty elections and a minimum level of Dahl’s institutional prerequisites for polyarchy as measured by V- Dem’s Electoral Democracy Index, but either access to justice, or transparent law enforcement, or liberal principles of respect for personal liberties, rule of law, and judicial as well as legislative constraints on the executive not satisfied.

Liberal democracy is defined as follows (Coppedge et al., 2021, p. 283):

De facto free and fair multiparty elections and a minimum level of Dahl’s institutional prerequisites for polyarchy/ … /are guaranteed as well as access to justice, transparent law enforcement and the liberal principles of respect for personal liberties, rule of law, and judicial as well as legislative constraints on the executive.

Hence, regime typologies also involve evaluations of aspects other than elections, such as the rule of law and respect for personal liberties.

Some scholars employ a dichotomous understanding of political regimes as either democracies or dictatorships, although there are variations within these categories (Alvarez et al., 1996; Cheibub et al., 2010). Others propose a more nuanced conceptualization that differentiates among various types of democracies and dictatorships or authoritarian regimes (Collier & Levitsky, 1997; Hadenius & Teorell, 2007). Based on the understanding that there are implicit variations in how transition processes occur, a strict dichotomous definition of democracy and non-democracy appears insufficient for this study. How political regimes are classified is also crucial for research on democratic transition processes to understand whether regimes transit from one regime type to another (Brooker, 2009; Levitsky & Way, 2010; Roller, 1999).

The literature on the representation of women in parliament stresses the importance of whether and, if so, what types of legislated gender quotas exist to investigate the importance of the presence of women in parliament (e.g., Zetterberg, 2009). We employ the variable v2lgqugen present in the V-Dem dataset version 12 (Hughes et al., 2019). The variable contains the following categories: no quota; yes, a statutory gender quota for all parties without sanctions; yes, a statutory gender quota for all parties with weak sanctions; yes, a statutory gender quota for all parties with strong sanctions; and yes, reserved seats in the legislature for women. Countries that have both candidate quotas and reserved seats are recorded as having reserved seats in the legislature (V-Dem, 2022). We reduce the number of categories to three, with the reference category being ‘no quota’, ‘quota with weak/no sanctions’, ‘quota with strong sanctions/reserved seats for women in the legislature’.

Temporal Concerns

The choice of analysing regimes over a specific period (1975–2019) corresponds to two important aspects: first, those suggesting increases in the level of regime ambiguity starting from the 1960s and coinciding with the third wave of democratization (Huntington, 1992; Lührmann et al., 2018). Moreover, it responds to more recent arguments of democratization scholars claiming democratic recession or a ‘third wave of autocratization’ (Bermeo, 2016; Diamond, 2015; Lührmann & Lindberg, 2019). The second refers to the effects following the initiation of the UN Decade for Women in 1976 and its potential to subsequently influence democratization processes and women’s role in politics in subsequent decades (Fallon et al., 2012). As stated, we conduct separate analyses for 1975–1989 and 1990–2019. The motive for this would be an increased focus on civil society after the breakdown of communist regimes in Eastern and Central Europe, and we would expect to find different trends after 1989 (e.g., Alexander, 2012; Diamond, 1994).

Control Variables

Furthermore, the analysis controlled for a number of other factors that might have influenced the dependent variable. In addition, the models include socioeconomic factors, such as GDP per capita, and education, as control variables. Countries with higher socioeconomic development – where GDP per capita, education, and urbanization could be incorporated into the analysis as measurements of socioeconomic development – tended to experience changes in attitudes, values, and patterns that are supportive of women’s increased participation and representation in different realms of society, including politics (Inglehart & Norris, 2003; Matland, 1998).

Results and Discussion

Changes Over Time

The initial examination of the changes in the main variables over time showed that both the dependent variable (female civic engagement) and the independent variable (women’s descriptive political representation in parliaments) increased between 1975 and 2019 (see Figure 1). In 1975, women constituted only 7% of the legislature representatives. While the variation across different regime types was minimal, closed autocracies exhibited slightly greater variation than did the other regimes. This can be attributed to the practices of communist regimes that advocated for the representation of women in legislatures without granting civil liberties. Among the liberal democracies in the period before 1989, only the Nordic countries had lower chambers, with at least 20% female representatives. However, by 2019, the representation of women had grown significantly, especially in liberal democracies and in the two categories of hybrid regimes.

Figure 1. Share of women MPs in lower chambers and women’s possibility to participate in civil society according to regime type.

Note: The left-hand figure represents women as members of parliament, and the right-hand figure represents women’s possibility to participate in civil society (V-Dem), mean values.

Data regarding women’s possibility to participate in civil society show a consistent positive trend over the years. However, what stands out most is the pronounced division based on regime type. Hybrid regimes fall somewhere between liberal democracies and closed autocracies. Notably, in the 1980s, a positive shift occurred, especially toward the end of the decade. This encourages us to conduct separate analyses on the relationship between women’s participation in parliament and civil society before and after 1990. Therefore, we examine within-country fixed effects by splitting our analysis into two distinct time periods: 1975–1989 and 1990–2019.

Regression Results

We first test the direct effect of the proportion of women in parliament on women’s possibility to participate in civil society without including any control variables. The results (Table 1, Model 1 and Model 3) indicate that there is a weak direct association between the share of women in parliament and women’s possibility to participate in civil society. This association is statistically significant only for the period after 1989, thus suggesting that the general increase in women in parliament during the period after 1989 was also associated with an increase in women’s possibility of participating in civil society. Therefore, H1 can be corroborated. However, this association does not account for differences across regime types. The results also show, as expected, that there are differences across the different regime types and that the direct effect of regime types shows that in all regimes other than our reference category, women in liberal democracies have fewer possibilities to participate in civil society before 1989 (Table 1, Model 2 and Model 4). In the period after 1989, the difference in women’s possibility of participating in civil society between liberal democracies and electoral democracies was no longer statistically significant.

Table 1. Women’s possibility to participate in civil society as the dependent variable, fixed effects with robust standard errors (at the country level).

	Model (1)	Model (2)	Model (3)	Model (4)
	1975–1989	1975–1989	1990–2019	1990–2019
Women in parliament	0.005	0.004	0.008*	0.006*
	(0.007)	(0.007)	(0.004)	(0.003)
Regime type (liberal democracy as ref.)
Closed autocratic		−0.923*		−0.524**
		(0.437)		(0.177)
Electoral autocracy		−1.009*		−0.311*
		(0.438)		(0.131)
Electoral democracy		−0.687*		−0.139
		(0.303)		(0.107)
GDP/capita		0.001		−0.013**
		(0.018)		(0.005)
Education		0.094		0.123*
		(0.060)		(0.059)
Year fixed effects	Yes	Yes	Yes	Yes
Constant	0.485***	0.731	1.025***	0.599
	(0.073)	(0.466)	(0.064)	(0.423)
N	1468	1468	3836	3836
N (countries)	114	114	135	135
R^2	0.166	0.242	0.173	0.257
adj. R^2	0.157	0.232	0.166	0.250

Standard errors in parentheses.

* p < 0.05, ** p < 0.01, *** p < 0.001.

We therefore proceed by testing whether the relationship between women’s presence in parliament and women’s possibility to participate in civil society is moderated by regime type (Table 2, Model 5 and Model 7). Figure 2 shows the marginal predicted change in women’s possibility to participate in civil society in relation to the share of women in parliament. In the pre-1989 period, none of the relationships between women in parliament and women’s possibility to participate in civil society appear to be moderated by regime type, as the interaction effect is not significant for any regime type when using liberal democracies as our reference category.

Figure 2. Predicted values for women in parliament and women’s participation in civil society moderated by regime type.

Note: Based on the results of Model 5 and Model 6 in Table 2.

Table 2. Women’s possibility to participate in civil society as the dependent variable, interaction regime types, and fixed effects with robust standard errors (at the country level).

	Model (5)	Model (6)	Model (7)	Model (8)
	1975–1989	1990–2019	1975–1989	1990–2019
Women in parliament	−0.011	0.002	−0.011	−0.001
	(0.010)	(0.004)	(0.010)	(0.005)
Regime type (liberal democracy as ref.)
Closed autocracy	−0.909*	−0.707***	−0.901*	−0.710***
	(0.435)	(0.192)	(0.436)	(0.196)
Electoral autocracy	−1.153*	−0.416***	−1.147*	−0.427***
	(0.457)	(0.115)	(0.458)	(0.118)
Electoral democracy	−0.778*	−0.194*	−0.777*	−0.208*
	(0.302)	(0.095)	(0.303)	(0.099)
Closed autocracy#women in parl.	0.008	0.015*	0.008	0.015*
	(0.013)	(0.007)	(0.013)	(0.007)
Electoral autocracy#women in parl.	0.028	0.007	0.028	0.007
	(0.018)	(0.005)	(0.018)	(0.005)
Electoral democracy#women in parl.	0.018	0.003	0.017	0.003
	(0.020)	(0.005)	(0.020)	(0.005)
GDP/capita	0.012	−0.010*	0.012	−0.009
	(0.019)	(0.005)	(0.019)	(0.005)
Education	0.099	0.104	0.099	0.091
	(0.056)	(0.058)	(0.056)	(0.056)
Gender quota (no quota as ref.)
Strongly enforced quotas			0.127	0.135
			(0.172)	(0.073)
Weakly enforced quotas			–	0.008
Year fixed effects	Yes	Yes	Yes	Yes
Constant	0.690	0.780	0.687	0.884*
	(0.447)	(0.418)	(0.446)	(0.405)
N	1468	3836	1468	3836
n (countries)	114	135	114	135
R^2	0.262	0.266	0.263	0.274
adj. R^2	0.250	0.259	0.251	0.267

Standard errors in parentheses.

* p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 2 shows that the largest difference is found in the group of closed autocracies pre- and post-1989. In the closed autocracies in the period preceding 1989, there was a negative relationship between women in parliament and women’s possibilities to participate in civil society. This result is likely driven by the relatively numerous communist regimes that had higher shares of women in parliament while maintaining restricted civil liberties. The positive relationship between women in parliament and women’s possibility to participate in civil society in the post-1989 period likely reflects the fact that both the presence of women increased (through the introduction of quotas or other measurements) and the perceived increased possibility for women to engage in civil society in a number of closed autocracies, such as Rwanda and Saudi Arabia, after 1989.

In the period past 1990, the rate of change in the possibility of participating in civil society for women is predicted to be greatest (within limits) in closed autocracies. Hence, our hypothesis H2 can only be partially corroborated as regime type significantly moderates the possibility for women to participate in civil society only in closed autocracies after 1989. We also must acknowledge that there are ceiling effects when the share of women in parliament approaches 50% and when there are no longer any perceived restrictions for women’s participation in civil society in place.

We also test whether the adoption of gender quotas mediates the association between the presence of women in parliament and women’s possibility of participating in civil society. The results show that the presence of women in parliament still has a statistically significant association with a predicted increase in the perceived possibilities for women to participate in civil society in closed autocracies post-1989 (Figure 3).

Figure 3. Predicted values for women’s possibility to participate in civil society controlling for gender quotas.

Note: Based on the results of Model 8 in Table 2.

The association is statistically significant compared to that of the reference category of liberal democracies. H3 can therefore not be corroborated. The slopes for electoral democracies and electoral autocracies are positive, indicating that an increase in the female presence in parliament is associated with a perceived increase in women’s possibility of participating in civil society even after controlling for the adoption of gender quotas. However, electoral democracies and electoral autocracies are not sufficiently different from liberal democracies to reach the threshold of statistical significance.

Conclusions

This study aimed to test the link between the share of women in parliament and women’s possibility of participating in civil society. The results confirmed that there is a positive association between women’s presence in parliament and women’s possibility to participate in civil society. An increased presence of women in parliament is associated with an increase in women's possibility to participate in civil society in the period after the fall of many communist regimes post-1990. The period after the fall of the wall of Berlin resulted in an increased focus on civil society from many different international organizations that also extended aid to civil society organizations in countries that were not liberal democracies. Moreover, in the decades following the end of the Cold War, the share of women in parliament increased, both in liberal democracies and in non-democratic regimes.

Our results also underscore that the share of women in parliament has indeed increased, especially in hybrid and closed autocratic regimes. This increase could be interpreted in line with the findings of previous studies that have argued that autocratic regimes use women’s representation to brand themselves as positive toward gender equality to enjoy the benefits especially on the international arena of such reforms without any intention to democratize their institutions (Bjarnegård & Zetterberg, 2022; Guriev & Treisman, 2022; Valdini, 2019). An increase in women in parliament that occurs during such circumstances does not necessarily result in an increase in women’s likelihood of engaging in civil society (Alexander, 2012; Wolak 2019). Our results, however, indicate that women’s possibility of participating in civil society has increased concurrently with increased female representation in parliament and especially in closed autocratic regimes. This can be interpreted as an increase in women’s political representation also can result in increased possibilities for women to participate in civil society.

We must acknowledge that there are important limitations to our findings. The key variable of women’s possibility to participate in civil society is based on expert coding, and the assessment of experts may differ somewhat from the empirical reality. Findings indicate that there is a strong tendency among external observers to more or less automatically associate an improved gender balance in governing institutions with democratization, even if no such changes have occurred (Bush et al., 2023). However, fully grasping the degree of involvement or whether there are informal barriers to participation in civil society is arguably difficult even in liberal democracies (Anheier, 2013).

We should also acknowledge that reverse causality is a potential issue in our findings. In essence, enhancing women’s opportunities to engage in civil society could increase their representation in parliament. Our research has shown the nature of the correlation between the proportion of women in parliament and the limitations on women’s ability to participate in civil society across different regime types. Despite the limitations of our study, we believe our findings suggest broader patterns and that a rise in the number of women in parliament correlates with expanded opportunities for women’s participation in civil society in regime types other than liberal democracies.

Additionally, this study contributes to the research field that understands women’s agency in non-democratic settings, such as hybrid and closed authoritarian regimes, from a gender perspective (Wilde et al., 2018). An important caveat of such regimes is, however, that they may also occupy and encourage participation in civil society organizations that are positive toward the regime. Women’s possibility to participate in such regime-friendly civil society organizations may therefore even be encouraged. Lifting existing formal or informal restrictions for women to participate in such civil society organizations may be perceived as not threatening the regime stability (Bjarnegård & Zetterberg, 2022; Valdini, 2019). We also have to take into account that women in hybrid or autocratic regimes that have their seats as MPs assigned through quotas can be more loyal to the regime and more reluctant to support women’s rights.

From the data we used in our study, it is not possible to distinguish whether women’s possibility to participate in civil society organizations is restricted mainly to organizations that are supportive of the regime. Hence, it is important to note that increased female participation in civil society organizations does not necessarily translate into processes of democratization if these increases occur mainly within regime-friendly organizations. In a broader context we have to consider that autocratic regimes also strive for regime survival (Guriev & Treisman, 2022). Consistent with the emphasis already placed by Bjarnegård and Zetterberg (2022), we would like to stress that it is problematic to automatically associate a rise in female participation in politics and in civil society with democratization without further scrutiny of both those contexts. Other studies using more detailed qualitative data are needed to further explore the relationship between women’s participation in parliament and in civil society in relation to processes of democratization.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Appendix

Table A1. Descriptive statistics of variables, within and between countries.

Variables	Mean	Std. dev.	Min	Max	Observations
Women's possibility to participate in civil society
overall	1.21	1.01	−2.66	2.76	N = 5305 (total)
between		0.87	−1.57	2.66	n = 137(countries)
within		0.51	−2.09	2.86	T = 38.72
Women in parliament
overall	14.56	11.10	0	63.75	N = 5305 (total)
between		7.30	1.94	38.14	n = 137(countries)
Within		8.17	−14.43	49.45	T = 38.72
GDP/capita
overall	12.35	12.85	0.29	75.47	N = 5305 (total)
between		12.07	0.82	60.61	N = 137(countries)
within		4.67	−9.13	54.16	T = 38.72
Education
overall	7.15	3.29	0.19	13.61	N = 5305 (total)
between		3.19	1.05	12.86	n = 137(countries)
`within		0.93	2.32	9.55	T = 38.72