Keywords

1 Introduction

In family business studies, the roles that women play is hardly researched and has become quite a new topic. In fact, only since the 90s has there been growing participation of women in family businesses, and a new line of research has begun to develop. Although there are scarcely any surveys on this issue, the available data show that, in recent years, the amount of women involved in governance is higher in family businesses than in non-family businesses (Gnan and Montemerlo 2008). Moreover, the presence of women in these businesses is increasing, although in Italy this growth has been slower and somewhat limited (Corbetta and Minichilli 2010; Corbetta et al. 2011), especially in non-listed companies. The latter, in fact, are not liable to Law n. 120/2011, which introduced the so-called “pink quotas”.

This figure is not surprising considering that Italy is one of the European countries with a much lower number of women on boards (around 6%), markedly worse than Norway (42%), Sweden (28%), Finland (27.2%) and France (22%) (Terjesen et al. 2015).

In fact, Italy is still a typically “male” country. This fact is well documented by the Global Gender Gap Report 2017, ranking Italy as 82nd out of a total of 144 countries and 118th considering women’s participation in economic activities. Moreover, Italy’s position worsened compared to 2006—the first reporting year—in which Italy ranked 77th.

What are the factors responsible for this situation? Regarding the presence of women on the boards of directors, several studies have been conducted in Italy and abroad. Various analyses have identified factors that can affect women’s participation, including ownership concentration and the family nature of the company (Gnan and Montemerlo 2008; Songini and Gnan 2009; Corbetta et al. 2011; Bianco et al. 2015; Campbell and Minguez-Vera 2008; Ruigrok et al. 2007; Nekhili and Gatfaoui 2013).

Studies concerning the presence of women on the boards have mainly considered listed and large companies. In Italy, however, the latter represent only a small percentage of existing companies. For this reason, this paper focuses on medium-sized businesses that have, in fact, been playing a leading role in the Italian economic system in the last few years (Turani 1996), thanks to their ability to contribute to innovation, job creation, exports and growth.

In the studies concerning such companies, the issue of female participation in governance roles has decidedly been neglected. Regardless, we believe that it is important to investigate this topic and speculate about the dynamics that have characterised medium-sized family businesses in recent years, considering that Italy does not represent a very favourable national context for women’s participation in economic life.

With this premise, this paper analyses data concerning a sample of medium-sized Italian family businesses, analysed from 2007–2014. Data analysis has a twofold objective:

  • understand how women’s participation in the governance of medium-sized companies has evolved in this period of time;

  • investigate the factors that can influence the presence of women on the boards of these companies.

The paper is structured as follows. The next section presents an analysis of the literature on women’s participation in corporate governance. In the third section, the research methodology is described, followed by the discussion of the main results of the analysis. Finally, the main conclusions, limitations and implications of the research are revealed and discussed.

2 Women’s Participation on the Boards of Family Firms

In recent years, the participation of women on company boards has been the subject of several studies and investigations, which can be divided into two main research areas:

  • analysis of the relationship between the presence of women on the boards and the companies’ performance;

  • analysis of factors that can favour/hinder the involvement of women.

Regarding the first aspect, many scholars have demonstrated the emergence of a series of advantages associated with boards with a high level of gender diversity (Conyon and He 2017). Such benefits would be the consequence of the different endowment of human capital of men and women and the different ways they deal with the decision-making process (Burgess and Tharenou 2002; Simpson et al. 2010; Terjesen et al. 2015). In particular, some scholars have evaluated the positive impact of gender diversity on the economic-competitive performance of companies. From this perspective, however, the results were variable and did not allow us to draw firm conclusions (Carter et al. 2003; Campbell and Minguez-Vera 2008; Adams and Ferreira 2009; Terjesen et al. 2009; Carter et al. 2010; Larcker and Tayan 2011; Jurkus et al. 2011; Ahern and Dittmar 2012; Matsa and Miller 2013; Ferreira 2015; Post and Byron 2015). Instead, other scholars have analysed how the presence of women on the board of directors affects companies’ non-financial performance, in particular, their outcome related to corporate governance.

Boards with a higher number of women are characterised by a higher number of board meetings (Bianco et al. 2015), a better quality of decision-making (Milliken and Martins 1996), severer monitoring of the CEO, more alignment with the interests of shareholders (Adams and Ferreira 2009), lower agency costs (Arfken et al. 2004), more transparency towards shareholders, higher earnings quality and lower earnings management (Gui et al. 2011).

The studies in the second line of research indicate that the participation of women in the governance of family firms has been increasing in many countries. For example, in the United States, from 2002 to 2007, the percentage of women with a CEO role in family businesses grew by 14%, although the percentage of women-led businesses was low (24%) (Glavin et al. 2007).

In Italy, the presence of women in the governance of family businesses has grown much more slowly and with different rhythms and intensities depending on the time frame considered. For example, between 2000 and 2008, the percentage of women on the boards of larger family businesses remained almost unchanged (Corbetta and Minichilli 2010). On the other hand, a very different trend has been observed in more recent years. In the listed family companies, in fact, the percentage of boards with at least one woman as the director has increased from 53.3% in 2007 to 92.7% in 2016. But in this case, it is clear that it was not a spontaneous change, but the consequence of the introduction of the so-called “pink quotas” (Law n° 120/2011). This reality is confirmed by the fact that in unlisted family businesses, which are not liable to this law, the percentage of boards with at least one woman has remained almost unaltered in the last 10 years (from 51.5% in 2007 to 55, 5% of 2016) (AUB 2017).

Regardless of the need to comply with specific legislative constraints, many scholars have questioned which factors can favour the presence of women on companies’ boards. In Italy, several studies have verified that the presence of women is higher in family businesses than in non-family businesses (Gnan and Montemerlo 2008; Songini and Gnan 2009). Similar results are provided by Corbetta et al. (2011) and by Bianco et al. (2015). They verified that in larger companies the presence of women on the board is encouraged by the existence of a family bond with the controlling shareholder. Moreover, the number of family-affiliated women is higher in smaller companies, with more concentrated ownership and larger boards (Bianco et al. 2015). Similar results have also emerged in investigations conducted outside of Italy, confirming that the family nature of businesses can favour the presence of women on boards (Campbell and Minguez-Vera 2008; Ruigrok et al. 2007; Nekhili and Gatfaoui 2013). In this case, Claessens et al. (2000) observe that if the link with the owner family is the main criterion for selecting women to be involved in the boards, the risk is that women without appropriate skills and professionalism could be involved. Nekhili and Gatfaoui (2013) also show that larger boards favour the presence of women.

Given that most of these studies focus on listed or large companies, neglecting smaller companies mostly due to the difficulty in finding data available for analysis, this work aims to fill this gap by focusing on medium-sized Italian companies. These companies have characteristics that are significantly different from larger ones, particularly listed companies, in terms of ownership structure, the involvement of members not belonging to the owner family, the degree of managerialisation and the presence of non-family managers (Palazzi 2012). As a consequence, the institutional set-up of medium-sized companies is very different from that of larger companies, and this can also affect their willingness to involve women in their corporate governance. We are aware that medium-sized firms have specific features, which leads us to believe that extending the same results from the analysis of larger companies is not justified. For that reason, we decided to take the opportunity to analyse in depth the universe of medium-sized firms.

3 Methodology

To achieve our research objectives a quantitative analysis was carried out through the study of data concerning medium-sized family businesses located in central Italy, observed from 2007 to 2014. Data were extracted from the AIDA database (Bureau van Dijk) and the historical records of the Chamber of Commerce. The following criteria were used to extract the sample:

  • to identify medium-sized companies: number of employees (50–249) and annual revenues (10–50 million euros) (in line with European Recommendation No. 2003/361/EC);

  • to identify family businesses: analysis of the family’s ownership percentage (Minichilli et al. 2010; Naldi et al. 2013). Since there is no official database of Italian family businesses, we adopted a manual procedure to thoroughly analyse the ownership structure of the selected firms and classify them as family or non-family firms. Family businesses are those in which a single individual or a family (two or more family members) holds more than 50% of equity shares. The kinship among shareholders was ascertained based on their surnames (Corbetta et al. 2015). In some cases we also used information found on the media, company websites, etc.;

  • to identify the geographical area: we referred to the notion of central Italy proposed by Istat (The Italian Institute of Statistics). Accordingly, this area corresponds to the following regions: Lazio, Tuscany, Marche and Umbria.

Drawing on the AIDA database, each year (from 2007 to 2014), all the companies corresponding to the criteria as mentioned above were extracted. As a result, over time, the number of companies under investigation were not fixed but showed slight variations, from a minimum of 214 companies in 2009 to a maximum of 233 in 2014. Historical records from the Chamber of Commerce were used to obtain information about corporate governance (board composition, the name of the chief executive officer, chairman of the board of directors, etc.).

The available data were initially used to carry out a descriptive analysis of the participation of women in the governance of medium family businesses and its evolution throughout that period.

The next step was a linear regression to identify factors that can affect the presence of women on the boards. To this end, we used the dependent variable “Women_board”, measured as the ratio between the total number of women on the board of each firm and the total number of directors. The independent variables concerning the governance of family businesses were as follows:

  • family participation in company ownership (Fam_ownership), calculated as the ratio between the number of family shareholders and the total number of shareholders;

  • the generational stage of the company (Generational_stage): we used a dummy variable, which has value 1 in the case of first generation companies and value 0 for multigenerational companies. The distinction was made by taking into account the age of the company, the threshold being the 25th year from the foundation of the company;

  • the average age of board members (Age_board), identified as the natural logarithm of the average age of members of the board of directors.

Furthermore, some control variables were considered:

  • “Size”, measures the size of the company in terms of the natural logarithm of total assets;

  • “Industry” identifies the business sector. We used a dummy, which assumes value 1 in the case of the manufacturing sector, and 0 otherwise.

4 Findings

The quantitative analysis was proposed as a twofold objective: (1) understand how the presence of women in governance roles has evolved during the period 2007–2014; (2) investigate the factors that can influence female participation on the board. The following section presents the main results.

The descriptive analysis made it possible to outline the presence of women in governance roles from 2007 to 2014. The overall picture confirms the meagre participation of women in the top positions of medium-sized family businesses. We first analysed the ownership structure. During the period under investigation, the composition of the ownership structure remained stable and highly concentrated, with an average number of shareholders-owners equal to 4 persons, about 80% represented by members of the family that hold the majority of the capital. The percentage of firms with a proprietary woman was slightly lower, falling from 52% in 2007 to 48% in 2014.

Compared to the total number of members, women accounted for a share that varies between 25% (2007) and 21% (2014). Compared to all the women involved in the companies’ ownership, those belonging to the controlling family were respectively 21% in 2007 and 18% in 2014; those outside the family, on the other hand, were 4% in 2007 and 3% in 2014.

A second element concerns the composition of the board of directors, whose structure remained stable over the analysed period. The average number of directors was 4. The percentage of directors-family members, albeit slightly decreasing, remained very high throughout the entire period of observation, with values varying between 79% in 2007 and 75% in 2014.

More than half the companies had at least one woman on the board: this condition was verified by 62% of companies in 2007 and 64% in 2014. Considering participation on the board, women represented a percentage that varied between 25% (2012, 2014) and 27% (2008, 2010) of the board members.

Further analyses were carried out to distinguish women belonging/not belonging to the owner family. In this case, our data clearly show that family firms prefer to recruit women directors within controlling families: women belonging to the controlling family represented a percentage between 76% (2012, 2014) and 74% (2008, 2010) of the total number of women involved on the boards, while the remaining 24% and 26% were women outside the family.

Only a few companies in our sample were guided by a team, with two or more persons (usually family members) who jointly share the leadership role. In fact, the percentage of companies with a team leadership ranged between 14% (2009) and 18% (2012) out of the total. Furthermore, teams were composed on average by 3 subjects, 80% of which are members of the controlling family. Among the companies with this form of leadership, a share between 46% (2007) and 38% (2014) had the presence of at least one woman involved in the team.

Women involved in the team leadership represented a percentage ranging between 23% (2008) and 17% (2013). Again, there was a significant presence of women family members, approximately 87% in 2008 and 88% in 2013.

Our analysis also concerned the office of chairman of the board of directors. In the vast majority, a person belonging to the family held this role (about 89% of cases). Moreover, this role is almost always monopolised by men, so much so that women represented only a small percentage—equal to 10%. Even in the few cases where a person from outside the family played the role (between 11 and 13% of companies), the participation of women remained low, between 19% (2007, 2008, 2014) and 28% (only in 2010).

Even in the case of companies led by a sole director, a similar situation can be observed. In fact, a family member mainly held the role of sole director, with an almost constant percentage throughout the period (2007–2014) and well above 80%. Also in this case, the involvement of women remained remarkably low over time, with percentages around 4% (2012) and 7% (2007, 2014). Finally, where the sole director is not a family member, this role is the complete prerogative of men.

Finally, we observed the figure of the CEO. The data show that this role is mainly entrusted to persons linked by kinship relationships with the controlling family (a percentage comprising 61% and 69% of the companies). The number of positions assigned to women showed a positive trend over time—although in absolute terms it was still low—rising from 12% in 2007 to 28% in 2014. Even when a non-family member held this position, the presence of women remained overshadowed, with percentages hardly exceeding 10%—except for the year 2010, which reached 14%.

In the second part of the work, a regression analysis was carried out to investigate the factors that can influence the presence of women on boards. Table 21.1 shows the correlation matrix related to the independent variables used to analyse factors that can influence the presence of women on boards. The analysis of correlations is preliminary to that of regression, as it allows us to verify the absence of multicollinearity between the variables being studied, and therefore to guarantee the adequate construction of the regression models. The analysis demonstrates the absence of multicollinearity between the explanatory variables. Moreover, the values obtained from the estimate for every single model of VIFs (variance inflation factor) are always very close to 1, a value that falls within the recommended standards (between 1 and 2). This estimation allows us to state that multicollinearity is not a problem for the estimated models.

Table 21.1 Descriptive statistics and correlation matrix
Full size table

Some initial ideas emerged regarding the correlation—positive or negative—between the dependent variable and the independent variables. There is a positive and significant correlation between the weight of the family in the ownership and the weight of women on the board. The latter variable, on the other hand, is negatively correlated with the average age of the board members, the generational stage and the sector to which they belong. There are no significant correlations between the company’s industrial sector and the participation of women on the board of directors.

Table 21.2 presents the results of the linear regression regarding the period 2007–2014. Data processing takes into account checks for heteroskedasticity, performed by applying the Breusch-Pagan test on each model. These tests proved to be significant, highlighting the presence of error heteroskedasticity, which is why the estimates were calculated with robust Huber White Sandwich errors. In addition, the models estimated for each year are characterised by a low R-squared but with significant F coefficients. This is not a problem for the purposes of our analysis, as we aimed to verify the existence of a relationship between independent and dependent variables, not to provide an exhaustive model that can adequately explain factors affecting the participation of women on the boards.

Table 21.2 Ols regression
Full size table

The overall analysis of the data revealed significant signals of the effects that some factors may have on the participation of women on the boards of medium-sized Italian family businesses. The first sign can be found in the positive and statistically significant relationship between the weight of the family in the ownership—“Fam_ownership”—and the dependent variable “Women_board” in the period 2010–2014. This means that the greater the weight of the owner family in the company ownership, the higher the participation of women on the board (2010: β = 0.136, p < 0.05; 2011: β = 0.113, p < 0.10; 2012: β = 0.115 p < 0.05; 2013 e 2014: β = 0.140, p < 0.05).

Another indication, with a contrasting sign, is found with regard to the generation that controls the company. In fact, in 2008, 2012 and 2013, the “Generational_stage” variable generated a negative impact on the presence of women on the boards (2008: β = −0.062, p < 0.10; 2012: β = −0.066, p < 0.10; 2013: β = −0.065, p < 0.10). It is, therefore, possible to deduce that first generation family businesses are less favourable to involving women on boards than those belonging to successive generations. However, the latter is more likely to accept the involvement of women on the board.

Moreover, the variable—“Age_board”—which reflects the average age of the board members—negatively affects the participation of women on the board of directors. It follows that, in the case of a board characterised by a strong component of elderly members, the tendency to involve female members will be to a lesser degree. On the contrary, younger boards will foster the presence of women. This result was strong and constant for the entire period analysed 2007–2014 (2007: β = −0.247, p < 0.01; 2008: β = −0.319, p < 0.01; 2009: β = −0.250, p < 0.01; 2010: β = −0.207, p < 0.10; 2011: β = −0.288, p < 0.01; 2012: β = −0.327, p < 0.01; 2013: β = −0.418, p < 0.01; 2014: β = −0.409, p < 0.01).

Finally, as regards the control variables included in the estimation models—“Industry” and “Size”—the results do not allow us to highlight a relationship with the dependent variable “Women_board”. That is to say that in our sample it is not possible to identify a “Size” effect, as the size of the company does not influence the weight of women on the boards. Similarly, there is no relationship between the business sector and women’s participation on boards.

5 Discussion and Conclusions

The present work complements studies on listed and/or large family businesses. In fact, it focuses on medium-sized family businesses and helps enrich existing knowledge about the participation of women in the governance of these companies. In particular, the results of the analysis presented in this paper contribute to research on women in family businesses in two ways.

Firstly, the analysis made it possible to map out the evolution of the presence of women in the governance of medium-sized family businesses over the past few years.

With regard to ownership, the results are rather discouraging. In the analysed period, in fact: (1) companies with at least one woman involved in the company ownership decreased from 52 to 48%; (2) the percentage of women out of all the company shareholders decreased by 4% (from 25 to 21%). The situation slightly improved by shifting attention to the participation of women on the boards of directors. In fact, the percentage of companies with at least one woman director rose from 62 to 64%. It is obviously a modest increase, but these percentages are much higher than those shown by previous studies concerning non-family businesses (see Terjesen et al. 2015). Our figures are also higher compared to those provided by the AUB Observatory concerning large family businesses (55.5% in 2016).

However, the weight of women on the boards remains limited. In the analysed period, in fact, women represented on average 26% of directors. This percentage is not significantly different from the threshold introduced in Italy by the “pink quotas” law (33%). Neverthless, we must consider that the size of the companies’ boards analysed in this paper was rather small, on average with four members. Therefore, even the presence of only one woman makes the percentage increase significantly. Furthermore, women on boards are almost always recruited within controlling families, with the risk that they are a purely formal presence.

Finally, these data show that the path to the desired achievement of gender balance (under consideration by the Italian government for the future), is still out of reach and much has yet to be done. Moreover, the trend for the analysed companies is even negative, as the percentage fell from 27% in 2008 to 25% in 2014. The results confirm, therefore, that in companies not liable to the law on “pink quotas”, little or nothing has changed in terms of the weight of women on boards. The results also support the importance of family ties, given that, in the vast majority of cases, the women involved in the boards are members of the owner family. In this sense, the data confirm what emerged from previous research on this subject (Gnan and Montemerlo 2008; Songini and Gnan 2009). Finally, as regards the top roles, the data indicate opposite tendencies. On the one hand, off-limit roles still exist, such as that of the sole director and chairman of the board, where the presence of women remained sporadic, and, on the other hand, there was a substantial increase of women CEOs, who in a few years more than doubled (from 12 to 28%).

Secondly, the present work contributes to improving the knowledge of the factors that can influence the presence of women on the boards of family businesses, an essential element to promote the evolution towards greater gender diversity.

Our findings are in line with results from other surveys on large firms (Corbetta et al. 2011; Bianco et al. 2015) and confirm that even in medium-sized family businesses a more significant presence of the family in the ownership favours the involvement of women on the boards. This is an important opportunity for women, who are more likely to take on a significant role on the board in family businesses. However, this opportunity risks being only “on paper” if the formal role—assumed as a consequence of family bond—does not correspond to a substantial and active role in the participation of the decisional and strategic processes of the company. On the other hand, as pointed out by Claessens et al. (2000), the importance of family ties also risks distorting the selection of board members, especially if it leads to involving women belonging to the family who lack the necessary skills to perform their role effectively. This modus operandi would nullify the benefits of gender diversity on financial and competitive performance. The involvement of women is a necessary but inadequate element to make the difference and should be accompanied by actual ability and the likelihood of carrying out their role efficiently.

Moreover, this study has the merit of having considered the influence of the generational stage, a variable that—as we currently know—has been neglected in previous studies on the factors influencing the involvement of women on boards. The results showed that the generational stage is a variable that significantly influences gender diversity. In particular, first-generation companies, compared to the companies of successive generations, are less inclined to involve women on their boards of directors. At the same time, the average age of the board members negatively affects the presence of women on the boards. The evolution towards a greater gender diversity of boards seems, therefore, to go through processes of succession that could favour the participation of women on two fronts. On the one hand, as highlighted by previous studies (Cesaroni and Sentuti 2018), it is undeniable that succession processes are becoming less oriented towards the exclusion of women, so next-generation women are more often involved in family business governance. On the other hand, it is likely that younger generations will be more open to involving women.

These results brought together, suggest the promotion of gender diversity not only in large companies, but also in medium-sized enterprises, since in the latter the presence of women is still low, and gender balance is far from being achieved.