Mandatory gender quotas: the impact on board and committee meetings

Gender diversity can affect both communities and companies' corporate governance effectiveness. Many earlier studies focus on female presence in companies (Nielsen and Huse, 2010; see also Burke and Mattis, 2000) and on the characteristics of women as Board of Directors (BoD) members (Huse and Solberg, 2006; Kakabadse et al., 2015; Joecks et al., 2019; Ruigrok et al., 2007). Some studies investigate why there are few women (Smith and Parrotta, 2015; Singh and Vinicombe, 2004; Hillman et al., 2007), and others investigate the link between BoD gender diversity and firm performance (Adams and Ferreira, 2009; Pucheta-Martínez and Gallego-Álvarez, 2020; Lückerath-Rovers, 2013; Rose, 2007; Liu et al., 2014; Brahma et al., 2021).

Using the framework of Minichilli et al. (2012), we contribute to the literature by analysing whether the component of females on a BoD is relevant, and the effect of a mandatory gender quota introduced in Italy by the 2011 Gender Equality Law. We test the regulative institution pillar to investigate whether the staggered mandatory introduction of a gender quota meets the regulatory goal.

Empirical studies investigate board meeting attendance (Adams and Ferreira, 2008, 2012; Lawler and Finegold, 2006). Some suggest that directors spend more time on supervisory issues, with managerial concerns also taking up a non-negligible portion of their time (Schwartz-Ziv and Weisbach, 2013). Corporate boards often delegate tasks to standing board committees. An increase in the amount of delegation reduces board tasks but increases coordination activities (Vafeas, 1999; Adams et al., 2015).

Our study tests whether the introduction of a gender quota impacts functioning of BoD and its internal committees thanks to female capacity in effort norms, cognitive conflicts and use of skills.

Firstly, we expect that gender quota will influence the organization of BoDs and internal committees. Secondly, we expect that female influence will be stronger in internal committees within BoDs, which are smaller, so that a minority group of women can raise the quality of debate and more frequent meetings (Burke and Mattis, 2000; Westphal and Milton, 2000; Forbes and Milliken, 1999). Thirdly, women are expected to improve the use of skills and knowledge and encourage cooperation and attention (Adams and Funk, 2010).

To take account of the staggered adoption, we use a difference-in-differences method, and we find that mandatory adopters have more frequent internal committee meetings after the introduction of the gender quota, which confirms that the regulation re-prioritizes work in internal committees, where the role of women has a stronger effect. We also confirm that institutional theory can be applied: a regulatory institution defined by law can explain why and how firms organise their BoD activities.

According to institutional theory, companies are influenced by social context and will thus adopt similar structures, organizations and behaviours (Krenn, 2016; Hillebrand et al., 2011). Social processes, norms and expectations of the external environment play an important role in explaining firm behaviour (Meyer and Rowan, 1977). Institutional theory explains the organization of firms and considers the context as a set of constraints/pressures to which the company must conform to gain legitimacy (DiMaggio and Powell, 1983). There are three types of institutional pillars: regulatory, normative and cognitive.

The regulatory pillar reflects regulative aspects of institutions (North, 1990; Roe, 2004). It includes the definition of rules and mechanisms to enforce compliance and control, as well as the sanction system. The normative pillar relates to the prescriptive dimension of companies, which follow certain social obligations either as a duty or to meet community expectations (Scott, 2001). Often, these pressures are related to professional circles, such as membership of associations or networks, which define shared criteria for the evaluation of behaviour (DiMaggio and Powell, 1983). The cognitive-cultural pillar concerns the way in which reality is conceived. Companies may follow models of “correct behaviour” because they represent what people believe to be correct (Scott, 2001). They tend to align with other similar, successful and prestigious companies.

Corporate governance is a reflection of regulatory, normative and cognitive pillars rather than choices made between an unlimited number of options. Iannotta et al. (2016) find that a particular configuration of country-level conditions supports the existence of a joint causal relationship between given institutional arrangements and that board gender quota legislation is not sufficient.

To investigate the effects of public policy on female representation on BoDs, we focus on the regulatory institution pillar which supports and/or allows organizational actions and decisions.

Most prior literature evaluates the effects of the percentage of women on BoDs (Adams and Ferreira, 2009; Kakabadse et al., 2015; Gordini and Rancati, 2017; Hossain et al., 2017; Abu Qa’dan and Suwaidan, 2019; Joecks et al., 2019; Hartmann and Carmenate, 2020; Sarabi and Smith, 2021), but there is little literature on its mandatory introduction. Adams and Ferreira (2009) present a brief analysis of initiatives worldwide to encourage female directors. Rigolini and Huse (2021) analyze the unexpected consequences of introducing gender quota on BoD, and Mensi-Klarbach et al. (2021) study the effectiveness of self-regulation in increasing gender diversity on BoD and show that self-regulation requires additional instruments in order to be effective.

Focussing on the staggered introduction of gender quotas in the Italian context, we investigate the effect of mandatory regulation on female representation.

Our analysis is based on the integrated model of board effectiveness (Minichilli et al., 2012), where the three determinants of board task performance are effort norms, cognitive conflicts, and skills and knowledge. Effort norms are a “group-level construct that refers to the group’s shared beliefs regarding the level of effort each individual is expected to put toward a task” (Forbes and Milliken, 1999). BoD members put a lot of effort into the preparation of meetings (Hambrick et al., 2008), but this effort varies across groups and depends on gender. Cognitive conflicts are task-oriented differences in judgment between group members, often manifested in “disagreements about the content of the tasks being performed, including differences in viewpoints, ideas and opinions” (Jehn, 1995). It is suggested that where cognitive conflict takes place, the quality of meetings is higher, and women may help in managing this conflict. The third component of the model is “the board’s ability to tap the knowledge and skills available to it and then apply them to its tasks”. This component is related to coordination of BoD members’ contributions and the division of tasks and responsibilities (Forbes and Milliken, 1999), known as female characteristics.

The underrepresentation of women at higher levels is often discussed with reference to the concepts of the glass ceiling and gender stereotyping (Mensi-Klarbach, 2014; Vasconcelos, 2018; Cimirotić et al., 2017). Gender-centred perspectives, sex-specific models and gender-specific concepts of social construction show differences between men and women (Davis et al., 2006; Harding, 2003). We connect these differences to the three pillars of the integrated model of board effectiveness (Minichilli et al., 2012).

Regarding effort norms, women on BoDs are positively associated with social responsiveness (Galbreath, 2011; Bernardi et al., 2006; Duehr and Bono, 2006; Walfisch et al., 2013). Smith et al. (2006) point to the different experiences of women in their lives compared to men. They affect time and effort used for meeting preparation. Regarding cognitive conflicts, women give greater attention to problem solving (Burke and Mattis, 2000; Bilimoria and Wheeler, 2000; Westphal and Milton, 2000; Hoffman and Maier, 1961). And in the use of skills and knowledge, females tend to be more inclusive and interactive, rely more on cooperation and collaboration and show higher scores on self-awareness, empathy and social skills (Singh et al., 2008; Adams and Funk, 2010; Jamali et al., 2008).

Meetings are the most frequent opportunities for the exchange of ideas, and meeting time gives BoD members the opportunity to discuss and monitor business (Hossain et al., 2017; De Andres et al., 2005; Cai et al., 2009; Shan, 2013). BoDs holding frequent meetings perform better and are more in tune with shareholder interests (Conger et al., 1998). Firms often react to poor performance by increasing the frequency of meetings (Vafeas, 1999); meetings make explicit the degree to which boards are motivated to participate in business operations (Wijethilake and Ekanayake, 2019). Board meetings are a key source of information for independent directors and improve efficiency (Masulis et al., 2012; Liu et al., 2016). Previous studies use the frequency of board meetings as a proxy for board activity and director involvement (Carter et al., 2003; Xie et al., 2003; Brick and Chidambaran, 2010; Ntim and Osei, 2011; Mayur and Saravanan, 2017).

More meetings do not necessarily imply better monitoring (De Andres et al., 2005). Lipton and Lorsch (1992) argue that the main problem facing directors is the lack of time to manage firms, although Conger et al. (1998) suggest that board effectiveness improves with the frequency of board meetings. Vafeas (1999) suggests that more frequent board meetings lead managers to better follow shareholder interests. However, Vafeas (1999) finds that boards are active when firms face problematic situations.

According to previous literature, we assume that BoD meetings may be a proxy of BoD functioning. On the basis of the integrated model of board effectiveness, we expect the regulatory requirement of the mandatory introduction of a gender quota to influence the organization of firm governance.

Therefore, we hypothesise as follows:

We assume that a regulatory change re-prioritizes aspects of work, and focus here on meetings of intra-board committees.

Intra-board committees make an in-depth examination of operational problems, which are subsequently considered by BoDs in plenary sessions. Some codes recommend a particular number of meetings (Méndez and García, 2007), and a positive relationship between internal board committees and accounting information quality has been identified (Xie et al., 2003; Abbott et al., 2004; Deli and Gillan, 2000). However, the frequency of the meetings is largely a matter of free choice.

We assume that with the introduction of a gender quota, the functioning of internal committees benefits from the effort made by women in preparing for and taking part in meetings in terms of higher quality of debate and levels of awareness. We thus expect a re-prioritization of work with more granulated division of work and greater use of committees. Internal committees are groups comprising different professionals facing competing demands on their time and efforts in task performance. The greater effort made by women, the higher quality decisions and capacity to use knowledge and skills thanks to women's awareness are expected to have an impact.

Table 1, presents the sample selection process. We start with 308 Italian-listed companies for the period 2010–2018. We exclude financial and insurance companies and firms with a two-tier or one-tier governance system due to differences in internal committee compositions. We exclude BoDs with retirement, death or resignations. The selection process leads to a final (unbalanced) sample of 113 companies and 952 firm-year observations. We hand-collect data from corporate governance reports and financial statement data from Compustat Global.

Table 2, presents the sample composition. Adopters are divided into three steps staggered according to the time of adoption. In Italy, BoDs are renewed every three years between March and May. Firms with BoD renewal between March and May 2012 had to comply with minimum 20% women on BoDs at the first renewal after 12 August 2012 (at BoD renewal between March and May 2015; third adopters 2015–2017 = 370 firm-year observations). Companies with BoD renewal between March and May 2013 had to comply with minimum 20% women on BoDs in 2013 (first adopters = 268 and second adopters = 314 firm-year observations). In 2018, firms could have renewed their BoDs less than three times after the Gender Equality Law of 2011. At the most, third adopters are on their second renewal in 2018–2020. Table 2 presents the resulting sample variation in the post variable.

Before the Gender Equality Law, on average, there were around 10% women on BoDs. On the first renewal, the firms met and surpassed the 20% threshold, with an average of 27%. On the second renewal, the first adopters had in general a significantly higher mean average of women on BoDs (40%) than third adopters (33%).

We follow the staggered adoption design proposed by Athey and Imbens (2018), which makes it possible to track concurrent but unrelated market-wide events, alleviating concerns that results are spuriously driven by other economic shocks or institutional changes (Leuz and Wysocki, 2016; Rauter, 2017). Literature on difference-in-differences has increasingly used staggered adoption (Borusyak and Jaravel, 2017; De Chaisemartin and d’Haultfoeuille, 2018; Strezhnev, 2018; Cao and Lu, 2019).

The research method compares changes in the outcome variables before and after the “event”. We use a difference-in-differences method, where we perform an analysis between treated and control firms in combination with the pre–post analysis. The treated firms are mandatory adopters only after the introduction of the Gender Equality Law. Control firms are voluntary adopters of a high gender quota before the introduction of the law. Given that all public limited firms were required to meet a quota of female directors, those firms that had a greater proportion of female directors before the quota faced a smaller constraint. We estimate the following regression model and expect that the introduction of gender quota in BoDs (post) will increase the outcome variable only for mandatory adopters. In other words, we expect a significant positive coefficient for post *mandatory adopters:

Appendices 1 and 2 show the variables and codes. Our dependent variables are BoD meetings, the number of board meetings in a fiscal year. AuC meeting is the number of audit committee meetings, and RemC meeting is the number of remuneration committee meetings in a fiscal year.

The test variable is mandatory adopters, a dummy variable coded 1 if a firm before quota regulation had a gender quota lower than the quota specified by the law (20%) and 0 otherwise. Next, we substitute mandatory adopters using the test variable first adopters, a dummy variable coded 1 if the firm had a BoD renewal after 12 August 2012 (after 12 August and before 31 December 2012). We use the variable post equal to 1 if the firm has a BoD renewed after 12 August 2012 and 0 otherwise.

Based on the literature (Méndez and García, 2007; Bianco et al., 2015), at the governance level, we control for the number of directors (BoD size), independent and executive directors (BoD independence and BoD executive), CEO age (CEO age) and the overlap between CEO and Chairman (CEO duality). To control at the firm level, we compute firm size as the natural logarithm of total assets, leverage as equity on total assets, Loss t−1 as equal to 1 if the firm had a loss in year t−1 and 0 otherwise, sales growth showing the sales increase (sales t–sales t−1)/sales t−1 and lastly, the level of return on asset (ROA) computed as operating profit/total assets.

Table 3 shows the descriptive statistics. On average, BoDs meet around nine times a fiscal year. Looking at internal committees, meetings are more frequent for audit committees (AuCs) than for remuneration committee RemCs (six and three times, respectively). AuCs and RemCs are mandatory for 34% of the sample but exist in about 90% of the sample.

The number of BoD members ranges from 3 to 21 (9 on average). BoDs comprise a maximum of 92.9% independent and 80% executive directors. The mean is higher for independent (45%) than executives (29.8%). CEO are 36–78 years old with average–high experience (56 years old). In 31.5% of the sample, they are also the board chair. The percentage of women in top positions is still low: 4.4% for CEOs and 8.6% for chairpersons.

Table 4 illustrates the correlation matrix. We exclude AuC and RemC sizes because they are correlated with each other and with BoD size. Moreover, firm size and BoD size, loss and return on assets show high correlation, but the variance inflation factor shows no problems of multicollinearity.

Table 5 shows a negative and significant coefficient (−0.669) for the interaction between post-mandatory adopters in relation to BoD meetings. Thus, mandatory adopters have less frequent board meetings after introduction of the regulation than before, controlling for the difference between mandatory and voluntary adopters.

Table 6 shows positive and significant coefficients (0.996) for the interaction post-first adopters for AuC and RemC meetings. Thus, first adopters in 2013 hold more frequent internal committee meetings after the gender quota, controlling for the difference between first and second–third adopters in 2014–2015. Regression results show that first adopters had fewer AuC meetings before the law came into force (−0.912); the effect of the law is thus even stronger for AuC.

In summary, the Gender Equality Law led to less frequent board meetings but more internal committee meetings. This result cannot itself be interpreted as a decline in corporate governance quality (Carcello et al., 2002; Liu et al., 2016; Min and Chizema, 2018). The law led to an increase in both AuC and RemC meetings for first adopters. The reduction in BoD meetings needs to be interpreted together with the increase in internal committee meetings.

Our findings taken together suggest that the Gender Equality Law had an important impact. It appears to enhance the role of internal committees, which are able to prepare items for discussion before BoD meetings and put forward a “semi-finished product”. This means that boards of directors can meet less frequently.

All the above may also be read together with the strand of literature on the boards' composition and its efficiency, with particular reference to studies showing that bigger groups work worse than smaller ones. Women in smaller groups like internal committees, by improving the use of skills and knowledge and encouraging cooperation and attention, can raise the quality of debate and request more frequent meetings.

So, supporting institutional theory, a regulatory institution implemented in a law has a significant effect on firm organization of BoDs and committees.

Robustness tests confirm our findings. To show that they are driven by the new law rather than firm characteristics, we repeat the analysis on a sample of similar firms. Untabulated findings show very robust results for RemC meetings. To show that findings don’t reflect the method used, we present them using a generalised difference-in-differences method (Jacobson et al., 1993). We drop the variable post and include the interactions of mandatory adopters with the full set of year-fixed effects. Untabulated results show that for RemC meetings before the regulation, the coefficients for interactions are negative, while from 2015 they become positive. For AuC meeting, before the regulation the coefficients, for the interactions are negative, while from 2014 on, they become less negative. Results confirm main results for internal committee meetings about more frequent internal committee meetings after the introduction of the gender quota: higher for RemC and less negative for AuC.

Finally, to show that our results are not affected by the smallness of the sample, we perform a three-way interaction term regression. We include additional interactions to incorporate all observations to avoid regressions on subsamples. Results show that, for RemC meetings, first adopters have a significantly higher frequency of RemC meetings after the regulation when compared to the control group of voluntary adopters before the law, although they were not significantly different before it.

We also use the percentage of women on BoD as independent variable. Mandatory adopters increase the percentage of women on BoD, decrease BoD meetings and increase AuC and RemC meetings in the sample. Data are statistically significant for RemC meetings.

Our results confirm the importance of female characteristics. They show that gender quota regulation led mandatory adopter companies to re-prioritize activities, holding fewer BoD meetings and more internal committee meetings, with very robust results for RemC meetings. We conclude that the law is a determinant of the increase in RemC meetings.

In other words, the gender quota appears to enhance the role of internal committees, the organization of meetings and the modus operandi of the BoD. The findings can usefully be linked to studies on board composition and efficiency and on smaller versus larger groups of people.

Our research has various implications. It is useful as a learning case; it shows how research can connect conceptual thinking with practice to produce stronger evidence for policy recommendations. In fact, firms may gain advantage by increasing female representation as first-movers and by creating long-term gender-diverse boards. Our study indicates that first adopters hold more frequent internal committee meetings and that moves towards gender equality improve firms’ internal functioning.

Our study is not without limitations. Characteristics are in some cases country-specific; different proxies may be chosen. The study investigates mandatory thresholds but does not consider a minimum number of women. Furthermore, the sample includes only firms making public disclosure.

Future research could investigate whether the reduction in the number of board meetings would increase their duration. Coordination and alignment of board member agendas, where a large number of positions held by women interlock (with fewer women involved), are pressing issues. It would be interesting to investigate the extent to which women decide meeting agendas and whether meeting time is wasted on unproductive tasks.

Table A1 

Table A2