MSc Thesis
Foreign institutional ownership and gender diversity on
corporate boards: evidence from emerging market firms.
In this research, we explore the relationship between foreign institutional investors and the representation of female directors on boards of emerging market firms, by using a sample of 548 firms in 25 emerging market economies from the period 2008-2017. Our empirical findings show that the presence of foreign institutional ownership increases the percentage of female directors. After controlling for potential endogeneity by using an instrumental variable regression, we still confirm our findings. We provide evidence the effect is explained by the country-of-origin from foreign institutional investors. Furthermore, we examine the institutional environment of the foreign institutional investors and find that the gender inequality index plays a role. We do not find evidence that the local country’s GII contribute to female directors on corporate boards.
Student name: Maartje Lensink
Student number: s2749556
Study program: International Financial Management
Supervisor: Dr. Swarnodeep Homroy
Co-assessor: Dr. Ambrogio Dalò
Date: 08-01-2021
Key words: board gender diversity, foreign institutional ownership, emerging firms,
1. Introduction
Female representation on corporate boards is at the forefront of business debates; particularly in the developed markets, the first steps towards gender parity on boards are taken (Terjesen, Aguilera & Lorenz, 2015). The numbers were slowly improving, however, due to the COVID-19 pandemic, women are negatively impacted, and the increasing number of female directors has stagnated (Coury, Huang, Kumar, Prince, Krivkovich & Yee, 2020). Although research indicates that women bring significant benefits to the board of directors, such as different cognitive frames and experiences (see e.g. Carpenter; Miller & Triana, 2009; Post & Byron, 2009) reputational benefits (Byoun, Chang & Kim, 2011) and enhance of firm performance (Adams & Ferreira, 2009), women remain underrepresented in boards of directors. Even more dramatically in emerging market firms where females represent a mere twelve percent on the board of directors (Emelianova, & Milhomem, 2019).
on corporate boards in emerging market firms. We test several hypotheses. The first hypothesis is whether foreign institutional investors have a positive effect on the percentage of female directors on corporate boards in emerging market firms. The view posits that the presence of foreign ownership in emerging markets can make socio-cultural changes to the underrepresentation of women in the boardroom. As several papers have suggested, institutional investors have several ways of a making meaningful impact on the firm’s decisions and operations, indirectly through either voting or selling shares (e.g. Shleifer & Vishny, 1986; Admati & Pfleiderer 2009; Edmans 2009) or directly, by raising their voice to the board of directors (Ferreira & Matos, 2008).
Foreign institutional investors can provide the local firm with the necessary incentives to change corporate regulations (Doige, Karolyi & Stulz, 2004). They can improve governance practices in these firms (Gillan & Starks, 2003; Aggarwal et al., 2011). As institutional investors can act as agents of societal change (Ioannou & Serafeim, 2015), they are likely to push local firms to adopt gender diversification policies within the organization. In particular, foreign institutional investors are able to exert influence because they have fewer business ties with the local firm, contrary to domestic investors. Also, foreign institutional investors encourage the local firm’s adoption of policies that enhance shareholder values (Harford, Kecskés & Mansi, 2018) such as gender inequality policies.
investors may function as the substitute for the weak governance mechanisms (Chung & Zhang, 2011). This creates the opportunity for foreign investors to push local firms to commit to newly introduced regulations (Fang et al., 2015).
Moreover, firms are embedded in the larger institutional environment and likely to reflect the regulations of this environment (Hall & Soskice, 2001), particularly because the institutional environment shapes gender parity on corporate boards (Terjesen & Signh, 2008). Country-level policies substantially contribute to the difference in gender diversity on corporate boards between countries (Iannotta et al., 2015; Grosvold, Rayton & Brammer, 2016) and will institutionalize norms in the local country that increase the female position in the boardroom (Thams, Bendell & Terjesen, 2018). Eventually, female directors are likely to appear in an environment that is supportive of gender equality as this seems to be more legitimate in these contexts (Post & Byron, 2015).
The third hypothesis is the concentration of foreign institutional investors from countries with a low (high) GII leads to a higher (lower) representation of female directors on corporate boards in emerging market firms. The effect depends on the social norms of the foreign institutional investors which they bring to the emerging market firm. Foreign institutions can transfer their social attributes (Guiso, Sapienza & Zingale, 2009) as they provide an investment channel through which society’s social norms flow into local firms. Particularly, foreign investors from high social norm countries can transfer their norms to local firms and improve the social performance of local firms (Dyck et al., 2019).
ownership stakes in countries where gender inequality is more prevalent, it may reinforce the existing local institutional environment. Therefore, they are not likely to address gender inequality as a primary concern.
To test the hypotheses, we use a sample of 548 MSCI listed firms in 25 emerging market economies over the period 2008-2017. We find that the presence of foreign institutional investors in emerging market firms leads to a higher representation of female directors. We show that this effect is partially explained by the country-of-origin of institutional investors. We also find that the firm-level characteristics number of directors and firm size and country-level characteristics strength of minority shareholder protection index and GDP per capita positively affect the percentage of female directors.
The possible issue of endogeneity of foreign institutional ownership makes it difficult to establish a causal effect. Foreign institutional investors from countries with higher gender equality may prefer to invest in firms from countries similar to their home country. To control for time-invariant unobserved firm heterogeneity, we apply firm-fixed effect regressions. We provide new evidence that the origin of foreign institutional investors considerably matters on the relationship between foreign institutional ownership and female board representation. European institutional investors are associated with more female directors, while U.S. institutional investors have no effect and institutional investors stemming from other regions lead to a lower number of female directors on corporate boards. European countries promote the importance of gender equality in the boardroom more extensively compared to other countries across the world. The European Union proposed a directive to achieve equality on corporate boards, and thus, many European countries are mandating policies already.
institutional investors investing in a local country with a high GII have a stronger effect on the percentage of female directors than investing in a local country with a low GII. To extend, we also take the institutional environment of foreign institutional investors into account. We find that foreign institutional investors from low GII countries investing in a local country with a high GII have a positive effect on the percentage of female directors. Although both effects contribute to more female directors on boards, we argue that the effect of foreign institutional from low GII countries is likely to prevail, since the largest fraction of investment is coming from these countries.
To strengthen the findings, we apply an instrumental variable regression to treat issues of endogeneity. Following the research of Li, Nguyen, Pham, and Wei (2011), we use non-family ownership as the instrumental variable to counteract the endogenous variable bias. We expect that the variable original owner of a firm is correlated with the current ownership structure of a firm, while it is unlikely that the representation of female directors is directly associated with the ultimate original owner of the firm. The instrumental variable regression depicts similar results to the baseline findings. It remains important to note that we do not fully eliminate the possibility that foreign investors prefer to invest in firms with more female directors, still, the results strongly indicate that foreign investors do not select on board characteristics in emerging market firms.
the percentage of female directors. We assume that larger differences between the indices occur when foreign institutional investors (from low GII countries) invest in countries that have a high GII. This confirms our previous finding that the effect of foreign institutional investors from low GII is likely to prevail in the relationship on female board representation.
The thesis is structured as follows. Section 2 highlights why foreign institutional investors are likely to influence the representation of female board members in emerging market firms. Section 3 describes the data collection and defines the variables. Section 4 presents the obtained regressions results. Section 5 concludes the paper and provides final remarks.
2. Literature review and hypotheses development
2.1 Board gender diversity
Over the last decades, research has paid much attention to the representation of female directors on corporate boards. Generally, research papers investigated firm and individual influences on board gender diversity, which leads to access of resources (Farrel & Hersch, 2005; Carter, Simkins & Simpson, 2003; Jurkus, Parke & Woodard, 2011), network expansion (Leung, Ricardson & Jaggi, 2014) and firm characteristics to maintain a competitive advantage and boost performance (Post & Byron, 2015; Gutiérrez-Fernández & Fernández-Torres, 2020). Furthermore, research papers also examined the pay gap between women and men (Terjesen & Singh, 2008) and the role of legal and cultural institutions in achieving gender equality on boards (see e.g., Iannotta, Gatti & Huse, 2015; Gianetti & Wang, 2020).
leads to more gender-balanced boards, particularly in firms with a more favorable attitude towards women (Gianetti & Wang, 2020).
Furthermore, the research of Knippen et al. (2019) predicts that firms are more likely to increase female directors if they experience strong external pressure. However, this effect is limited due to the existence of intergroup bias, where board members treat their ingroup members more favorably than outgroup members (Hewstone, Rubin, & Willis, 2002). The pressure of external forces increases the female representation on corporate boards. However, if the intergroup barrier is resistant enough, boards are more likely to add a new female director through the addition of a board seat instead of the substitution of a male director (Knippen, et al., 2019). This makes the male directors more likely to treat the female directors as outgroup members and heightens the salience of diversity in the boardroom. Thus, even though external pressure increases female representation in the board of directors, it remains limited in overcoming the issue of the intergroup bias.
To counter this concern, government interventions can be used to fully achieve gender balanced boards. Consistently, Terjesen et al. (2015) acknowledge that government and political institutions play a substantial role in establishing effective corporate governance policies related to gender equality. The study of Kogut et al. (2014) indicates that stipulating a quota can facilitate improvements of structural equality at the top of the firm for women and minority groups. A quota has to be implemented at the country-level to make structural changes in the number of female directors in the boardroom (Thams, et al., 2018). The establishment of mandatory gender quotas for corporate boards serves ultimately as an instrument to achieve gender equality, which was found to have a significant increase in female directors (Gianetti & Wang, 2020).
The gender quota requires that the board of directors of publicly listed and state-owned firms consists of at least 40% female directors by 2008 (Thams et al., 2018). The European Union also followed by implementing an identical gender quota and proposed a directive in 2012, which is not yet approved. India1 and Pakistan2 both introduced an act that states to have at least
one female director on board. The United States has also encouraged public awareness regarding gender equality and monitor compliance with equal employment (Terjesen et al., 2015).
Although market and legal instruments are powerful forces to change the gender inequality on corporate boards, these forces alone cannot establish a gender-parity board (Knippen, et al. 2019). Shareholder activism that requests for board gender diversity is a channel through which institutional practices of a corporate board can be changed (Perrault, 2015). The willingness of shareholders to raise their voice for activism depends on the issue, particularly when shareholders experience less responsibility for the board’s primary tasks (Gillan & Starks, 2007) or a certain degree of distrust in the firm (Daily et al., 2003) they are motivated to use their voice. Shareholders also engage in activism when they believe that they will benefit from the moral claim against the board of directors (Greenwood et al., 2002; Perrault, 2015). Hence, some shareholder activists have tried to increase the representation of female directors in a corporate board; it does not simply imply that all the stakeholders of the firm are positively associated with the appointment of female directors on corporate boards (Hillman, Shropshire & Cannella, 2007; Adams & Ferreira, 2009). Several research papers indicate negative responses from shareholders to the appointments of female CEOs and support this finding with the negative attitude from shareholders towards female directors (Lee & James, 2007; Knippen, Palar & Gentry, 2018).
1 In 2012, India introduced a gender quota for certain types of firms.
Although a negative attitude towards female board representation exists, a more gender-balanced board has positive contributions to a board, such as different cognitive values, experiences, and knowledge thereby providing unique information for better decision-making (Carter et al., 2003). The diversity of boards enhances a better understanding of the complex environments, incorporating strategic decision-making and sound risk management (Byoun, Chang & Kim, 2016). Furthermore, the ability to make better decisions is likely to improve a firm’s financial performance (Loyd, Wang, Phillips, Lount Jr., 2013). In particular, female directors have better market performance in countries with greater gender parity (Post & Byron, 2015).
2.2 Foreign institutional investors in emerging market economies
board of directors, or indirectly, by influencing directors’ actions by exiting the firm. Although research indicates the influence of foreign institutional investors on emerging market firms (see e.g., Fang et al., 2015; Bena et al., 2017; Dyck et al., 2019), there remains limited evidence that foreign institutional investors play a role in shaping a local firm’s policies toward gender equality on boards.
Not all institutional investors have the incentive or ability to influence a firm’s regulations and strategic decision-making. Controlling domestic institutional investors are generally closely tied to the firms they invest in and are likely to be more accommodating to corporate insiders (Ferreira and Matos, 2008; Fang et al., 2015; Bena et al., 2017). In contrast to foreign institutions, who are outside investors of the firm, they act as external monitors and thus can reduce managerial entrenchment (Li, Nguygen, Pham and Wei, 2011). Particularly, when an appropriate form of monitoring is in place through the board of directors, institutional investors are more committed to the firm’s international diversification strategy (Tihanyi et al. 2003). Also, the presence of foreign institutional investors in emerging market firms contributes to additional benefits to the local firm. Doidge, Karolyi, and Stulz (2004) suggest that foreign investors provide emerging market firms with the necessary tools and incentives to increase corporate governance. Moreover, large institutional owners of emerging market firms provide significant benefits to the local firm, such as monetary capital, resources, technology, and human capital, and are generally associated with long-term commitment (Li et al., 2011).Bena et al. (2017) confirm these findings and suggest that higher foreign institutional ownership leads to more long-term investment in tangible, intangible, and human capital. Furthermore, they add that foreign institutional ownership of emerging market firms positively enhances innovation output and increases the firm’s valuation.
disciplinary forces (Fang et al., 2015). Gillan and Starks (2003) document that foreign institutions play a prominent role in promoting governance changes in the local firm, generally due to their independent position and international experience. Consistently, Aggarwal et al. (2011) find that increased investment by foreign institutions leads to an improvement in a firm’s governance practices. Particularly, foreign institutional investors can exert pressure because they have fewer business ties with the local firms to jeopardize, unlike domestic institutional investors. To add, long-term foreign institutional investors encourage the local firm’s adoption of policies that enhance shareholder value and thereby influencing board outcome policies (Harford et al, 2018).
Furthermore, foreign institutional investors can significantly affect the local firm’s environmental and social performance (Dyck et al., 2019). The authors demonstrate evidence that portfolio investment provides a channel through which society’s social norms flow into firms thereby affecting economic decision-making. The results show how foreign investors from high social norm countries are successfully pushing firms to improve environmental and social performance. This pressure for foreign institutions on emerging market firms can lead to the adoption of gender diversification policies in the firm.
Also, female board representation is likely to be positively related to the fiduciary duties of the board (Post & Byron, 2015). Firms with more female directors tend to be more engaged in activities such as monitoring and strategic decision-making. Therefore, firms presumably select female and minority groups as directors as a tool to show (gender) equality within the workforce (Baysinger and Butler, 1985). A diverse board is likely to be viewed as more legitimate by the public, the media, and the government compared to less diverse boards (Adams and Ferreira, 2009). These reputational benefits of female board representation create incentives for institutional investors to pressure local firms to increase board gender diversity (Byoun et al., 2016). Thus, taking the aforementioned reasoning into account, we expect that the presence of foreign institutional ownership in emerging market firms will lead to more female directors on corporate boards in these countries. From this, we expect the following hypothesis:
H1: Foreign institutional investors are associated with a higher percentage of female
directors on corporate boards in emerging market firms.
However, it is unlikely that the positive effect of foreign institutional investors on female board representation is uniform across different environments (Pye & Pettigrew, 2005; Post & Byron, 2015). An environment where boards are motivated to consider different perspectives are expected to be present in countries where gender parity is more predominant (Post & Byron, 2015). This is in line with research that argues that a board’s attention to monitoring and strategic decision-making depends on the institutional context (Tuggle, Schnatterly & Johnson, 2010).
asymmetry between insiders and outside investors. They find that firms with high quality of governance mechanisms exhibit higher stock market liquidity and lower trading costs. Developed markets generally have strong regulatory institutions to provide a strong foundation for corporate governance. Firms in emerging markets tend to have weaker governance structures and foreign institutional investors may function as the substitute for the weak governance mechanisms, thereby providing incentives for firms to commit to newly established regulations (Fang et al., 2015). However, poorly governed firms can also hamper effective external monitoring and promote insider ownership as such that the influence of foreign investors is limited in a specific context (Kho, Stulz & Warnock, 2009).
Furthermore, as firms are embedded in the larger institutional environment; a firm’s governance practices tend to reflect the structures and regulations of this environment (Hall & Soskice, 2001). Terjesen and Singh (2008) suggest that female representation in corporate boards is shaped by the institutional environment and country-level factors contribute to a substantial influence on the difference in gender diversity on corporate boards between countries (Iannotta, Gatti & Huse, 2015; Grosvold, Rayton & Brammer, 2016). Consistently, Thams, Bendell, and Terjesen (2018) find that country-level progressive gender-related policies will institutionalize norms that ultimately lead to more female directors.
values necessary for females to occupy a board position (Wright, Baxter, & Birkelund, 1995). Female directors in relatively low gender inequality countries generally possess the education and work experiences, similar to men, that allow them to make contributions and have an influence on corporate boards (Post & Byron, 2015). Thus, we expect that if the local country is having a low gender inequality index; it positively influences the relationship between foreign institutional ownership and the representation of female directors. On the other hand, we expect that if the local country has relatively high gender inequality, the relationship between foreign institutional investors and the representation of female directors is negatively moderated. From this line of reasoning, we construct the following hypotheses:
H2a: A relatively high gender inequality index of the local country weakens the positive
effect of foreign institutional investors on the percentage of female directors.
H2b: A relatively low gender inequality index of the local country strengthens the
positive effect of foreign institutional investors on the percentage of female directors.
Finally, it is observed that the female presence in the labor market is higher in countries with low gender inequality than in countries with high gender inequality. Research has indicated that a gender diverse board is associated with higher firm valuations in lower gender inequality environments (Post & Byron, 2015). For foreign institutional investors, this might create an additional incentive to improve gender equality on boards where higher gender inequality is incorporated, in turn, the valuations of their ownership may increase.
raise capital abroad, they need to have close ties with the local country’s network and thereby comply with the local social norms (Dyck et al., 2019). In addition, these investors can transfer social attributes to local firms they invest in abroad (Guiso et al., 2009). The impact on social norms depends on foreign institutional investors’ home country. When investors originate from countries with strong norms towards social issues, they will positively influence the local firm’s social performance (Dyck et al., 2019). Powerful social norms can ultimately counter market pressures to focus exclusively on financial returns. To extend, foreign institutional investors are active in pushing firms to increase social performance, in particular, if they are stemming from countries with strong social norms (Dyck et al., 2019). Therefore, we expect that foreign institutional investors originating from countries with a relatively low gender inequality index are likely to increase the female representation on boards in emerging market firms.
In contrast, foreign institutional investors coming from countries with higher inequality and who invest in comparable countries concerning gender inequality are less likely to change the local country’s institutional environment. Since there exists a greater similarity in terms of gender inequality between the country of the foreign institutional investors and the local country where they invest. In turn, this similarity means alignment of the institutional forces and may neglect the issue of gender diversity on corporate boards. Adding to that, foreign institutional investors from high GII countries generally have fewer incentives to socio-cultural concerns, therefore they are less motivated to incorporate gender parity boards in the local country (Post & Byron, 2015).
board of directors). Ultimately, investors from high GII countries are likely to have weaker governance systems and in turn, are less motivated to act as monitoring agents as they will not be held liable for failure fiduciary responsibilities of the board (e.g. Fang et al, 2015; Post & Byron, 2015). Therefore, they are unlikely to encourage board behavior regarding the inclusion of female board members. Thus, we expect institutional investors originating from countries with a relatively high gender inequality index to be associated with a lower representation of female directors on boards. Consequently, we construct the following hypotheses:
H3a: The concentration of foreign institutional investors from countries with a high
gender inequality index is associated with a lower percentage of female directors in corporate boards in emerging market firms.
H3b: The concentration of foreign institutional investors from countries with a low
gender inequality index is associated with a higher percentage of female directors in corporate boards in emerging market firms.
3. Data description and methodology
3.1 Sample criterion
To construct the sample, we include all firms listed on the MSCI emerging markets index from 2007 to 2018. The MSCI emerging markets index is a popular benchmark for investing in emerging markets globally (Bena et al., 2017). In total, the sample represents all the firms from the 26 emerging market economies3, listed in Table 1. Furthermore, we access the BoardEx
database to retrieve data on board composition and the gender ratio within corporate boards.
The financial data of the selected sample firms is obtained from Thomson Financial Datastream. The financials are reported in U.S. dollars. We collect the data on ownership from the ThomsonOne database. This database does not contain any observations on Saudi Arabian firms and therefore this country is dropped from the sample. We collect data on country-level variables from the World Bank and the United Nations Development Program database. Furthermore, we excluded financial services firms (SIC codes 6000-6999) because the specific balance sheet structure and characteristics of these firms might affect the values of the financials and, in turn, cause biased results (see e.g., Jang, 2017; Fauver, Mingyi, Xi & Taboada, 2017). To mitigate the effect of outliers, we winsorized leverage and return on assets at a 5%-level. The final sample consists of 548 unique firms for 4,015 firm/year observations, representing 25 emerging market economies. The structure of the panel data is unbalanced, allowing firms to enter, exit, and return during the sample period(Wooldridge, 2001; Jensen & Zajac, 2004). The descriptive statistics are reported in Table 1; Panel A presents the summary statistics of the full sample and Panel B shows the summary statistics per country.
3.2 Percentage of female directors
3.3 Foreign institutional ownership
The explanatory variable describes the share of foreign institutional investors. The data on foreign institutional ownership is retrieved from the ThomsonOne database. Following the research of Chung and Zhang (2011), we define institutional ownership as the fraction of a firm’s shares that are held by institutional investors. Investors are categorized as institutional investors when ThomsonOne has labeled them as such. To separate the institutional investors, we use the following keywords for identification: “Endowment Fund”, “Foundation”, “Investment Advisor”, “Independent Research Firm”, “Investment Advisor/Hedge Fund”, “Hedge Fund”, “Pension Fund”, “Private Equity”, “Research Firm”, “Sovereign Wealth Fund” and “Venture Capital”. Then, we obtain the home country of the institutional investors. When the institutional investor is headquartered in a different country than it is investing in, the investor is defined as a foreign institutional investor. The variable domestic institutional investors is 0.23, implying that the local firms on average are largely owned for 23% by domestic institutional investors. Foreign institutional investors hold a larger fraction of local firms, indicating that on average 77% of the shares of local firms are held by foreign institutional investors.
3.4 Firm-level characteristics
are measured as the ratio of short-term debt and firm total assets (see e.g., Ferreira & Matos, 2008).
3.5 Country-level characteristics
The country-level characteristics that are included in the regressions are the strength of minority shareholders’ protection, the GDP per capita, and the local country’s gender inequality index. The strength of minority shareholders’ protection is retrieved from the World Bank database, measuring the gap between an economy’s performance and the regulatory best practice on protecting minority investors. The index ranges from 0, indicating the lowest regulatory performance score, to 100, indicating the best regulatory performance score (World Bank, 2020). The scores of the variable strength minority shareholders’ protection index range between 30 and 90. Qatar reports the lowest average score (37.50) and Malaysia the highest (85.26). A high score means that the gap is relatively small between the economy’s performance and protecting minority investors’ practices compared to a lower score.
The GDP per capita is retrieved from the World Bank database and is the gross domestic product dived by the midyear population of the local country. The GDP per capita is reported in current U.S. dollars.
Table 1: Summary statistics
This table shows the summary statistics of an MSCI emerging market sample of 4,015 observations between 2008-2017 from 25 countries. The statistics include the number of female directors, domestic (DII) and foreign institutional investors (FII) as a fraction of total investors, number of directors, firm size (in mln US$), return on assets, leverage ratio, strength of minority shareholders’ protection, GDP per capita (in US$) and the local country’s gender inequality index. Panel A shows the summary statistics for the full sample. Panel B shows the means of summary statistics per country. The variable definitions and sources are provided in Appendix 1.
Panel A: Full sample
Variables Observations Mean Standard deviation Minimum Maximum
Number of female directors 4,015 0.09 0.10 0.00 0.50
Domestic investors 4,015 0.23 0.17 0.00 1.00
Foreign institutional investors 4,015 0.77 0.17 0.00 1.00
European institutional investors 4,015 0.35 0.14 0.00 0.00
American institutional investors 4,015 0.36 0.17 0.00 0.00
Other institutional investors 4,015 0.28 0.19 0.00 0.00
Number of directors 4,015 10.46 3.01 1.00 24.00
Firm size 4,009 4.89 2.21 -6.60 9.48
Return on assets 4,009 172.35 224.70 8.25 936.30
Leverage ratio
Strength of minority shareholders’ protection 4,009 4,015 1.60 65.55 2.05 1.11 0.02 30.00 8.29 90.00 GDP per capita 4,015 8,867 7,616 998.5 85,076
Gender inequality index 4,015 0.36 0.16 0.06 0.60
Panel B: Summary statistics by country
Country name
Percentage of female
directors
FII DII No. of
directors
3.6 Correlation matrix
Appendix 2 presents the Pearson correlation matrix, which is used to assess the possibility of multicollinearity. Imperfect multicollinearity arises at levels above 0.7, and imprecise estimation of coefficients and high standard errors might occur (Brooks, 2014). The coefficients are below the absolute threshold. The variables return on assets and leverage indicate a strong relationship, since both variables are based on the firm total assets.
3.6 Research design
To analyze the impact of foreign institutional investors on the percentage of female directors on corporate boards in emerging market firms, we conduct an OLS regression. The following equation is presented in equation (1):
!"#$%"&'(")*+(,!,#= .$+ .%!00!,#+ .&1+*$%&'(")*+(,!,#+ .'!'(#2'3"!,#+ .(456!,#+ .)7"8"($9"!,#+
.*2ℎ$("ℎ+%;"(,<(+*")*'+=+,#+ .,?&<@"(A$@'*$+,#+ .-?00+,#+ .-!00!,# B ?00+,#+ C!,#+ D!,#+ E+,!,#,
(1) where i is used for an individual firm, t is used for each year, c is used for a specific country.
FemaleDirectors measures the percentage of female directors; FII is the measurement for
foreign institutional investors, which is the fraction of a firm’s shares that are held by foreign institutional investors; TotalDirectors is the total number of directors on the board; FirmSize is the size of the firm measured in the natural log of total assets; ROA defines the return on assets of the firm and is measured as the ratio of earnings before interest and taxes to total assets;
Leverage is the total leverage of the firm and measured as the ratio of short-term debt and firm
total assets; ShareholdersProtection presents the strength of minority shareholders’ protection index of the local country; GDPperCapita measures the GDP per capita of the local country;
between foreign institutional investors and the local country’s gender inequality index; C is the year dummy fixed effect; D is the industry dummy fixed effect; and E is the error term.
3.6.1 Endogeneity
The possible concern of endogeneity of foreign institutional ownership creates difficulties to establish a causal effect. In fact, foreign institutions from countries with high gender equality may prefer to invest in firms from countries similar to their home country, which is likely to explain the relationship between foreign institutional ownership and the representation of female directors. To address this omitted variable issue, we apply a firm fixed effects model that controls for time-invariant unobserved firm heterogeneity (Bena et al., 2017).
institutional investors that is not correlated with the error term is included in the second stage of the regression. The first stage of the regression is presented in equation (2):
!00F = G.,# $+ G%H+=!$#'%I!,#+ G&!00!,#+ G'1+*$%&'(")*+(,!,#+ G(!'(#2'3"!,#+ G)456!,#+
G*7"8"($9"!,#+ G,2ℎ$("ℎ+%;"(,<(+*")*'+=+,#+ G-?&<@"(A$@'*$+,#+ G/?00+,#+ C!,#+ D!,#+ 8+,!,# ,
(2) where i is used for an individual firm, t is used for each year, c is used for a specific country. !""# is the part of foreign institutional investors that is uncorrelated with v the error term and used in equation (3); NonFamily is the instrumental variable that shows whether the ultimate original of the firm is non-family owned, assuming 1 if the local firm was established by a government or widely held firms and has been nationalized in history, and 0 when the original owner was a family or individual. All control variables and year and industry dummies are kept the same according to equation (1).
The second stage of the 2-SLS instrumental variable regression is shown in equation (3):
!"#$%"&'(")*+(,!,#= .$+ .%!00F + ..,# &1+*$%&'(")*+(,!,#+ .'!'(#2'3"!,#+ .(456!,#+ .)7"8"($9"!,#+
.*2ℎ$("ℎ+%;"(,<(+*")*'+=+,#+ .,?&<@"(A$@'*$+,#+ .-?00+,#+ C!,#+ D!,#+ E+,!,#,
4. Results
In this chapter, the empirical findings of this thesis will be presented. First, a univariate analysis to indicate the differences between domestic and foreign institutional investors is performed. Then, the OLS regression and fixed-effects regression results are discussed. Furthermore, the findings on foreign institutional investors originating from different gender parity regions are discussed and followed by an analysis of the gap between the gender inequality index from the local country and the investors’ originating country. Thereafter, the instrumental variable regression findings are presented. To end, several batteries of robustness checks are performed.
4.1 Univariate analysis
In this section, a two-sample mean t-test is performed to test potential differences between domestic and foreign institutional investors. The differences in mean and standard deviation between the two types of firms are compared. This provides an impression of whether domestic institutional investors are different from foreign institutional investors.
Table 2 presents the univariate analysis of firms that are largely held by foreign institutional investors and firms that are largely held by domestic institutional investors. The sample consists of 3,714 firm/year observations that are owned by foreign institutional investors for more than 50%. Domestic institutional investors hold more than 50% of the shares in 301 firm/year observations.
the gender inequality index of the local country is lower in countries where firms held by foreign institutional investors are present. The variables firm size, return on assets, leverage, and GDP per capita are not statistically significant, implying that based on these firm-level and country-level characteristics firms owned by domestic institutional investors and firms owned by foreign institutional investors do not significantly differ from each other.
The results of the two sample mean test suggests that a difference between domestic and foreign institutional investors exists concerning female representation, number of directors, the strength of minority shareholder protection index and the gender inequality index. This might indicate that foreign institutional investors can select on preferences of the local country in which they invest.
Table 2: Univariate comparison
This table presents the results of the two sample mean tests. The sample is divided into two groups; firms that are largely owned by domestic institutional investors (firms without FII) and firms that are largely owned by foreign institutional investors (firms with FII). The difference between the means of the two groups are depicted and the t-statistics for significance are reported. The variable definitions and sources are provided in Appendix 1. The ***, **, *, indicate statistical significance at the 1%, 5%, 10% levels, respectively.
Firms without FII
Firms with FII
Difference t-statistic
Percentage of female directors Domestic investors 0.06 0.64 0.09 0.20 0.03*** -0.44*** 5.71*** -59.55*** Foreign institutional investors
European institutional investors American institutional investors Other institutional investors
0.36 0.16 0.45 0.39 0.80 0.37 0.36 0.27 0.44*** 0.21*** -0.09*** -0.11*** 59.55*** 22.43*** -5.06*** -5.65*** Number of directors 9.51 10.53 1.03*** 5.41*** Firm size 5.02 4.89 -0.13 -1.10 Return on assets 169.63 172.46 2.83 0.20 Leverage
Strength of minority shareholders protection
1.63 67.60 1.60 65.38 -0.03 -2.22*** -0.19 -4.51*** GDP per capita 8451.13 8900.35 449.22 0.96
Local country’s GII 0.44 0.35 -0.09*** -10.01***
Observations 301 3,714 4,015 4,015
4.2 Baseline regression results
Table 3 and Table 4 present the baseline results of the effect of foreign institutional investors on the representation of female directors on corporate boards in emerging markets using OLS and firm fixed-effect regressions. The regressions include year and industry dummies and control for firm-level characteristics and country-level characteristics.
Table 3 reports the OLS regression results of the percentage of female directors on foreign institutional investors. Model (1) includes the independent variable foreign institutional investors and the control variables. Model (2) – (4) specifies the regression for different origins of foreign institutional investors. Model (2) includes the foreign institutional investors from Europe, model (3) includes U.S. institutional investors; and model (4) identifies the foreign institutional investors from other countries excluding European countries and the United States. Lastly, model (5) adds the interaction term between foreign institutional investors and the local country’s gender inequality index to identify a moderating effect.
In model (1), we find that the coefficients of foreign institutional investors are positive and statistically significant at a 99% confidence interval. This finding confirms H1 and thus, foreign
a lower percentage of female directors on corporate boards. The economic effect can be measured as a one standard deviation increase in other institutional investors leads to a 0.0691 standard deviation reduction in the percentage of female directors.
In model (5), the interaction term between foreign institutional investors and the local country’s GII has a positive and significant effect (0.306 at a 1%-level). This implies that foreign institutional investors investing in a local country with a high gender inequality index have a stronger effect on the percentage of female directors than investing in a local country with a low gender inequality index. This finding rejects our H2a, which indicated that a relatively high
gender inequality index of the local country weakens the positive effect of foreign institutional investors on the percentage of female directors. Moreover, the coefficient estimate of foreign institutional investors is insignificant, which means that we have to reject our H2b as well since
we find no supportive evidence that a relatively low GII of the local country strengthens the positive effect of foreign institutional investors on the percentage of female directors. This effect might be explained by the findings of Post & Byron (2015). They suggest that factors more directly relevant to the board, such as shareholder protection regulations and laws governing board structures and behavior, provide greater motivation for directors to optimize decision-making processes rather than the influence of socio-cultural factors.
countries scoring higher on protecting minority shareholders have a higher percentage of female directors.
Weak shareholder protection generally means that there is less motivation for the board of directors to optimize their decisions as they are unlikely to be held liable for failure, whereas strong shareholder protection tends to strengthen corporate governance (La Porta, López de Silanes, & Shleifer, 1999). Therefore, foreign investors can benefit from these legal protections as it motivates directors to improve decision-making and to enhance the positive effects of female directors on boards (Post & Byron, 2015). In addition, given the lack of alternative governance mechanisms in countries, foreign institutional investors can bridge the governance differences across these countries and play the monitoring role (Ferreira & Matos, 2008; Aggarwal et al., 2011).
Table 3: OLS regressions
This table presents the OLS regression results of foreign institutional investors on the percentage of female directors using an MSCI emerging market sample of 4,015 observations between 2008-2017 from 25 countries. The dependent variable is the percentage of female directors, measured as the fraction of total directors on the board. The main explanatory variable is foreign institutional investors, measured as the share of total investors. We estimate the effect of institutional investors originating from Europe, United States and other regions excluding European countries and the United States. The firm-level control variables that are included are number of directors; firm size, measured as the natural logarithm of firm total assets; return on assets, ratio of earnings before interest and taxes, and firm total assets; leverage, ratio of short-term debt and firm total assets. The country-level control variables are strength of minority shareholders protection, GDP per capita and the local country’s gender inequality index. Model 1 includes the main explanatory variable and the control variables. Model 2-4 include the institutional investors from different regions. In model 5, the interaction term is added to the model. The regression models include year and industry dummies. The robust standard errors are reported in parentheses. The variable definitions and sources are provided in Appendix 1. The ***, **, *, indicate statistical significance at the 1%, 5%, 10% levels, respectively.
(1) (2) (3) (4) (5) Variables Percentage of female directors Percentage of female directors Percentage of female directors Percentage of female directors Percentage of female directors
Foreign institutional investors 0.0526***
(0.0167) (0.0389) -0.0622
European institutional investors American institutional investors Other institutional investors
0.109*** (0.0237) 0.00881 (0.0205) -0.0691*** (0.0192) Number of directors 0.00281*** 0.00228*** 0.00320*** 0.00355*** 0.00270*** (0.00102) (0.00104) (0.00101) (0.000983) (0.00101) Firm size 0.00489*** 0.00397*** 0.00458*** 0.00374*** 0.00541*** (0.00136) (0.00136) (0.00142) (0.00138) (0.00136)
Return on assets 1.05e-05
(1.06e-05) 1.10e-05 (1.04e-05) 1.01e-05 (1.05e-05) 1.22e-05 (1.03e-05) 1.00e-05 (1.05e-05) Leverage -0.000535 -0.000413 -0.000614 -0.000899 -0.000577 (0.00109) (0.00107) (0.00108) (0.00106) (0.00108)
Strength minority shareholders protection GDP per capita 0.00173*** (0.000311) -1.09e-06** 0.00175*** (0.000303) -1.47e-06*** 0.00158*** (0.000306) -1.44e-06*** 0.00179*** (0.000312) -1.78e-06*** 0.00172*** (0.000311) -1.39e-06***
(4.82e-07) (5.15e-07) (5.52e-07) (6.18e-07) (4.93e-07)
Local country’s GII -0.0146 0.00128 -0.0460 -0.0361 -0.259***
(0.0322) (0.0327) (0.0327) (0.0344) (0.0835)
Foreign institutional investors x Local country’s GII 0.306*** (0.100) Constant -0.126** -0.126** -0.0679 -0.0616 -0.0287 (0.0521) (0.0524) (0.0523) (0.0523) (0.0588) Observations 3,960 3,960 3,960 3,960 3,960 R-squared 0.357 0.365 0.352 0.361 0.361
Comparing Table 3 and Table 4, we observe that both tables are to some extent similar, specifically model (2) – (5). In model (1), the coefficient estimate of foreign institutional investors becomes insignificant compared to the model (1) of Table 3. In model (2), the variable European institutional investors is associated with a positive and significant coefficient (0.0384 at a 10%-level) on the percentage of female directors. This is comparable to Table 3; however, the magnitude of the coefficient is smaller. In model (4), the coefficient of other institutional investors (-0.0567 at a 1%-level) is negative and statistically significant. This variable presents a similar result compared to Table 3.
European and other institutional investors indicate significant results. The positive effect from European institutional investors and the negative effect of other institutional investors have a similar magnitude, hence in opposite direction, which means that taken these investors together; will outweigh one and another. Therefore, the coefficient of foreign institutional investors in model (1) is insignificant. This implies that even though, institutional investors are headquartered in a different country than the local country, the geographic location where the institutional investors are headquartered considerably matters. Explained by the fact that European countries emphasize the importance of gender equality in the boardroom more compared to other countries across the world. Norway mandated a gender quota since 2008 and the European Union proposed a directive to achieve gender parity boards in listed firms in 2012. Although the directive is not yet approved, the topic remains relevant, illustrated in several European countries distinctly adopting such policies already.
find supportive evidence for H2b. This effect is counterbalanced when foreign institutional
investors are investing in countries with relatively high gender inequality. Again, this is contrasting the expectations that a relative high gender inequality index of the local country weakens the positive effect of foreign institutional investors on the representation of female directors and thus, H2a is rejected.
Table 4: Firm-fixed effects regressions
This table presents the firm fixed-effects regression results of foreign institutional investors on the percentage of female directors using an MSCI emerging market sample of 4,015 observations between 2008-2017 from 25 countries. The dependent variable is the percentage of female directors, measured as the fraction of total directors on the board. The main explanatory variable is foreign institutional investors, measured as the share of total investors. We estimate the effect of institutional investors originating from Europe, United States and other regions excluding European countries and the United States. The firm-level control variables that are included are number of directors; firm size, measured as the natural logarithm of firm total assets; return on assets, ratio of earnings before interest and taxes, and firm total assets; leverage, ratio of short-term debt and firm total assets. The country-level control variables are strength of minority shareholders protection, GDP per capita and the local country’s gender inequality index. Model 1 includes the main explanatory variable and the control variables. Model 2-4 include the institutional investors from different regions. In model 5, the interaction term is added to the model. The regression models include year and firm fixed effects. The clustered standard errors are reported in parentheses. The variable definitions and sources are provided in Appendix 1. The ***, **, *, indicate statistical significance at the 1%, 5%, 10% levels, respectively.
(1) (2) (3) (4) (5) Variables Percentage of female directors Percentage of female directors Percentage of female directors Percentage of female directors Percentage of female directors Foreign institutional investors
European institutional investors American institutional investors Other institutional investors
0.00171 (0.0139) 0.0384* (0.0217) 0.0380 (0.0250) -0.0567*** (0.0192) -0.0500* (0.0297) Number of directors 0.00156 0.00147 0.00170 0.00160 0.00156 (0.00117) (0.00116) (0.00119) (0.00116) (0.00117) Firm size 0.00230* 0.00226* 0.00228* 0.00221* 0.00231* (0.00128) (0.00128) (0.00129) (0.00128) (0.00128)
Return on assets 6.25e-06 6.68e-06 6.52e-06 7.28e-06 6.49e-06
(6.86e-06) (6.83e-06) (6.87e-06) (6.84e-06) (6.86e-06)
Leverage 0.000368 0.000326 0.000320 0.000233 0.000343
(0.000704) (0.000704) (0.000703) (0.000702) (0.000702)
Strength minority shareholders protection GDP per capita -0.000512 (0.000388) 1.69e-06 -0.000508 (0.000358) 1.43e-06 -0.000557 (0.000390) 1.83e-06* -0.000579 (0.000387) 1.52e-06 -0.000451 (0.000390) 1.73e-06
(1.05e-06) (1.03e-06) (1.04e-06) (1.03e-06) (1.05e-06)
Local country’s GII 0.0333 0.0404 0.0297 0.0384 -0.0600
(0.0834) (0.0828) (0.0838) (0.0833) (0.110)
Foreign institutional investors x Local country’s GII
0.132* (0.0779)
Constant 0.0351 0.0231 0.0252 0.0575 0.0690
(0.0450) (0.0447) (0.0448) (0.0448) (0.0516)
Firm fixed effects Year fixed effects Observations Yes Yes 4,009 Yes Yes 4,009 Yes Yes 4,009 Yes Yes 4,009 Yes Yes 4,009 R-squared 0.107 0.109 0.109 0.112 0.109 Number of firms 547 547 547 547 547
4.3 Concentration of foreign institutional ownership from different GII countries
We make a distinction between the concentration of foreign ownership from relatively high gender inequality countries and relatively low gender inequality countries. We constructed the concentration of foreign institutional investors from high GII countries if the country is above the median (on average, GII between 0.058 – 0.207) of the GII distribution of foreign institutional investors for the corresponding year; and the concentration of foreign institutional investors form low GII countries if the country is below the median (on average, GII between 0.207 – 0.513) of the GII distribution of foreign institutional investors for the corresponding year. The table consists of two models; model 1 includes the concentration of foreign ownership from high and low GII countries and model 2 adds the interaction terms for the foreign ownership variables and the local country’s gender inequality index.
Table 5 shows the empirical results for the concentration of ownership from different GII countries. In model (1), the coefficient estimate of the concentration of foreign institutional investors from high GII countries is positive and statistically significant (0.0642 at a 1%-level). Hence, we reject H3a because foreign institutional investors from high GII countries are not
associated with a lower number of female directors. As anticipated, the coefficient of foreign institutional investors from low GII countries is also positive and statistically significant (0.0513 at a 1%-level). Therefore, we confirm H3b, since there is a significant relationship
between the concentration of foreign institutional investors from low GII countries on the percentage of female directors. The control variable number of directors, firm size, strength of minority shareholders’ protection, and GDP per capita are significant. This indicates that larger corporate boards, larger firms, a higher score of the minority shareholder index and, higher GDP per capita relate to a higher percentage of female directors, which is in line with the findings of Table 3.
the local country’s GII is positive (0.0231 at a 1%-level), however, its main effect of foreign institutional investors from high GII countries is insignificant. This effect suggests that foreign institutional investors from high GII countries investing in countries with high gender inequality lead to a higher percentage of female directors compared to investing in countries with a low gender inequality index. To clarify, foreign institutional investors seek similar board structures and regulations comparable to their domestic markets (Fang et al., 2015) and therefore foreign institutional investors from countries with a high GII are likely to invest in countries with a high GII.
In addition, the coefficient of the interaction term concentration of foreign institutional investors from low GII countries and the local country’s gender inequality index is positive and statistically significant (0.318 at a 1%-level) and the coefficient of the main effect of foreign institutional investors from low GII countries becomes negative (-0.0656 at a 10%-level). This result indicates that the local country’s GII counterbalances the negative effect of foreign institutional investors from countries with a low GII. This means that when foreign institutional investors from low GII countries invest in a local country with a high GII, the effect on the percentage of female directors is positive. To interpret this finding, we assume that foreign institutional investors from low GII countries act as agents in improving corporate governance practices and function as a substitute for weaker governance systems in emerging market firms (Fang et al., 2015). This allows investors to change the institutional environment, such as improving gender inequality (Dyck et al., 2019).
illustrate, 80.1% of the foreign institutional investors are from low GII countries. From this point, we can confirm the expectation that findings of the effect of foreign institutional investors from low countries will prevail.
Table 5: Concentration of foreign institutional ownership from different GII countries
This table presents the regression results of concentration of foreign institutional investors from high and low GII countries. The variable foreign institutional investors from high GII is constructed if the country is above the median of the GII distribution; and the concentration of foreign institutional investors from low GII countries if the country is below the median of the GII distribution. The firm-level control variables that are included are number of directors; firm size, measured as the natural logarithm of firm total assets; return on assets, ratio of earnings before interest and taxes, and firm total assets; leverage, ratio of short-term debt and firm total assets. The country-level control variables are strength of minority shareholders protection, GDP per capita and the local country’s gender inequality index. Model 1 includes the concentration of foreign ownership from high and low GII countries. Model 2 adds the interaction terms for the foreign ownership variables and the local country’s GII. Year and industry dummies are included. The robust standard errors are reported in parentheses. The variable definitions and sources are provided in Appendix 1. The ***, **, *, indicate statistical significance at the 1%, 5%, 10% levels, respectively.
(1) (2)
Variables Percentage of female
directors
Percentage of female directors Concentration of foreign institutional investors from
high GII countries
0.0642*** (0.0178)
-0.0239 (0.0392) Concentration of foreign institutional investors from
low GII countries
0.0513*** (0.0168)
-0.0656* (0.0397) Concentration of foreign institutional investors from
high GII countries x local country GII
0.0231*** (0.101) Concentration of foreign institutional investors from
low GII countries x local country GII
0.318** (0.101) Number of directors 0.00278*** 0.00258*** (0.00102) (0.00100) Firm size 0.00509*** 0.00524*** (0.00136) (0.00136)
Return on assets 1.09e-05 1.06e-05
(1.06e-05) (1.04e-05)
Leverage -0.000569 -0.000682
(0.00109) (0.00107)
Strength minority shareholders protection 0.00178*** 0.00179***
GDP per capita (0.000308) -1.16e-06** (4.97e-07) (0.000306) -1.48e06*** (5.04e-07)
Local country GII -0.0276 -0.250***
(0.0323) (0.0845)
Constant -0.128** -0.0351
(0.0522) (0.0590)
Observations 3,960 3,960
R-squared 0.359 0.365
4.4 Endogenous association between investors and board gender diversity
the plausible endogeneity concern, we estimate the percentage of female directors and the foreign institutional investors’ relationship using an instrumental variable regression.
The conditions for a proper instrument are that the variable should be highly correlated with foreign institutional investors and uncorrelated with the percentage of female directors. Following the research of Li et al. (2011), the instrument is a measure of the type of original ultimate controlling owner. As there is no public source on the history of ownership, we manually traced back the history of each firm to identify how the firm was established. The variable was assigned a 1 if the firm was established by a government or widely held firms and has been nationalized in history, and a 0 when the original owner was a family or individual and never has been nationalized in history. It is likely that the instrument highly correlates with foreign institutional investors, as non-family firms are more likely to sell a stake of the firm and thereby increasing the cost of foreign ownership compared to family-firms. Also, privatizations of government-owned firms are often the main targets for foreign investors to acquire significant ownership stakes in emerging market firms, which increases the likelihood of foreign ownership in non-family firms. In contrast, family-owned firms are often more resistant to share or lose control and therefore likely to have a lower share of foreign ownership (Li, et al., 2011). The instrument is based on historical information regarding the ownership foundation; thereby it is likely to be exogenous concerning the current representation of female directors on the board.
main finding that foreign institutional investors increase the representation of female directors on corporate boards in emerging market firms.
The control variables number of directors, firm size, strength minority shareholders’ protection and show similar results as our baseline findings in Table 3. GDP per capita becomes insignificant in the second stage. In addition, the local country’s GII shows a positive and significant result, which is contrasting to our findings from the main model.
Table 6: Instrumental variable regression
This table presents the instrumental variable regression results to control for the possible endogeneity issue, using an MSCI emerging market sample between 2008-2017 from 25 countries. The dependent variable is the percentage of female directors, measured as the fraction of total directors on the board. The main explanatory variable is foreign institutional investors, measured as the share of total investors. The instrument for foreign institutional investors is the dummy variable non-family ownership; a measure of the identity of the original owner of a firm, which equals 1 if the firm was established by a government or widely held firms and /or has been nationalized in history, and a 0 when the original owner was a family or individual and never has been nationalized in history. The firm-level control variables that are included are number of directors; firm size, measured as the natural logarithm of firm total assets; return on assets, ratio of earnings before interest and taxes, and firm total assets; leverage, ratio of short-term debt and firm total assets. The country-level control variables are strength of minority shareholders protection, GDP per capita and the local country’s gender inequality index. Model 1 includes the main explanatory variable and the control variables. Model 1 presents the first-stage and model 2 presents the second-stage of the instrumental variable regression. The regression includes year and industry dummies. The clustered standard errors are reported in parentheses. The variable definitions and sources are provided in Appendix 1. The ***, **, *, indicate statistical significance at the 1%, 5%, 10% levels, respectively.
(1) (2)
Variables Foreign institutional
investors
Percentage of female directors Foreign institutional investors
Non-family ownership 0.0352*** 0.473*** (0.148) (0.00734) Number of directors 0.00370*** 0.00173** (0.000969) (0.00878) Firm size -0.00706*** 0.00759*** (0.00129) (0.00131) Return on assets -0.00264** (0.00123) 0.00130 (0.000954) Leverage -0.000536 -4.45e-05 (0.00130) (0.000921)
Strength minority shareholders protection GDP per capita -0.00266*** (0.000272) -6.15e-06*** 0.00306*** (0.000452) 1.37e-06 (5.13e-07) (9.80e-07)
Local country’s GII -0.573*** 0.231**
(0.0279) (0.0907) Constant 1.205*** -0.612*** (0.0338) (0.185) Observations 3,410 3,410 R-squared 0.058 4.5 Robustness testing
The first robustness test is the exclusion of emerging market countries with less than 30 firm observations. Table 7 shows that the estimates are similar to the empirical results presented in Tables 3 – 5, hence the magnitude of the coefficients is larger. This indicates that our findings based on the full sample remain similar when we restrict the sample to emerging market countries with more than 30 individual firm observations.
In model (2), we excluded India; this country has already implemented a gender quota during the sample period. The coefficient estimate of foreign institutional investors is positive and significant, which implies that India does not influence our main finding.
The third robustness test restricts the sample to firms that have at least one female director on the corporate board. Our main explanatory variable, foreign institutional investors, remains positive and statistically significant. This suggests that our main findings based on the full sample are unchanged when we focus on at least one female director in emerging market firms. In model (4), we created the representation of female directors as a binary variable. The variable is assigned a 1 when there is at least one female director on the board, and 0 otherwise. We performed a probit regression. Foreign institutional investors still show a positive and significant coefficient (0.511 at a 10%-level), confirming our main finding.
Table 7: Alternative proxies
This table shows results of the foreign institutional investors on the percentage of female directors using an MSCI emerging market sample of 4,015 observations between 2008-2017 from 25 countries. The dependent variable is the percentage of female directors. The main explanatory variable is foreign institutional investors. The firm-level control variables that are included are number of directors; firm size, measured as the natural logarithm of firm total assets; return on assets, ratio of earnings before interest and taxes, and firm total assets; leverage, ratio of short-term debt and firm total assets. The country-level control variables are strength of minority shareholders protection, GDP per capita and the local country’s gender inequality index. Model 1 drops the countries in the sample that have less than 30 observations. Model 2 excludes the country (India) that implemented a gender quota. Model 3 includes only firms that have at least 1 female director. Model 4 measures female directors as a binary variable; 1 is assigned when the firm has at least 1 female directors, 0 otherwise. The regression models include year and industry dummies. The robust standard errors are reported in parentheses. The variable definitions and sources are provided in Appendix 1. The ***, **, *, indicate statistical significance at the 1%, 5%, 10% levels, respectively.
(1) OLS (2) OLS (3) OLS (4) Probit
Countries with more than 30 observations Exclusion of country with gender quota Firms with at least 1 female director Female directors as a binary variable Variables Percentage of
female directors female directors Percentage of female directors Number of female directors Number of
Foreign institutional investors 0.494*** 0.0489*** 0.411* 0.511*
(0.187) (0.0186) (0.219) (0.277) Number of directors 0.0913*** 0.00338*** 0.0803*** 0.144*** (0.0128) (0.00107) (0.0142) (0.0177) Firm size 0.0482*** 0.00391*** 0.0308** 0.0627*** (0.0160) (0.00136) (0.0144) (0.0207) Return on assets -0.000165 (0.000109) 1.78e-05 (1.20e-05) 4.57e-05 (0.000144) 0.000119 (0.000178) Leverage 0.00793 -0.00160 0.00139 -0.02120 (0.0111) (0.00120) (0.0130) (0.0203)
Strength minority shareholders protection 0.0148*** 0.00184*** 0.00610** 0.0266***
GDP per capita (0.00416) -1.17e-05 (7.23e-06) (0.000325) -1.39e-06*** (5.09e-07) (0.00296) 1.05e-05* (6.09e-06) (0.00485) -3.09e-05*** (1.11e-05)
Local country’s GII 0.102 0.0757* 0.503 -0.709
(0.392) (0.0408) (0.392) (0.565) Constant -1.695*** -0.145** 0.0274 -4.070*** (0.579) (0.0582) (0.630) (0.671) Observations 2,885 3,111 2,291 3,494 R-squared Pseudo R-squared 0.478 0.417 0.331 0.206
4.5.1 Gap between gender inequality indices
In model (1), we find a positive and significant coefficient for foreign institutional investors (0.0570 at a 1%-level). This confirms our main finding. The coefficient of the gap between foreign institutional investors’ GII and the local country’s GII is insignificant, which means that the difference in the indices between countries does not significantly affects the percentage of female directors.
