Relationship between Women on Board Directors and Economic Value Added: Evidence from Latin American Companies

Abstract

This study aims to evaluate the relationship between the presence of women on boards of directors and the generation of economic value added (EVA^®). For the empirical analysis, a panel data model with random effects is used, encompassing 202 Latin American companies between 2019 and 2021. The results obtained show that having women on boards of directors has a non-significant positive effect on EVA^®, which diminishes as women’s participation on the board increases. Theory suggests that more diverse boards of directors exercise better control, leading to improved financial results. However, the diversity of members has also been associated with longer decision-making processes that generate inefficiencies and increase costs. This contribution adds to the existing literature by exploring under-studied variables in the region and expanding knowledge on this topic in the Latin American context.

1. Introduction

Several studies have been conducted on corporate boards that provide a shared perspective on the director selection process and how board composition affects both organizational actions and overall performance [1]. Companies should focus on enhancing risk management effectiveness to mitigate the impact of crises, and regulators in countries with poor governance should prioritize supportive policies and long-term plans to improve governance quality and reduce the negative impact of economic turbulence [2]. The board of directors plays a crucial role in overseeing banking risk, maintaining bank stability, and market actions [3,4].

The role of women, both inside and outside companies, has evolved over the years. By the end of 2017, for the first time in Latin America, women’s labor participation was greater than 50%; however, although this was an unprecedented figure, the participation of women was still 25% lower than that of men [5]. According to a study conducted by [6], it is argued that boardrooms with greater gender diversity exhibit high-quality earnings. The study revealed that an increase in gender diversity on corporate boards is associated with an improvement in earnings quality in companies with significant power disparities. However, in societies with lower power disparities, an increase in the percentage of women in executive positions does not significantly impact the already high levels of earnings quality.

The percentages of women’s participation in corporate boards vary according to reports and studies on gender diversity. Some countries have adopted gender quota laws, such as Norway, which has had a 40% requirement since 2003, and France, which implemented a 40% quota since 2011. Other countries, like Australia and the United Kingdom, have set voluntary targets to increase female representation on these boards. Spain introduced an equality law in 2007 recommending a 40% representation of each gender on boards of directors, while Italy implemented a similar law in 2011 with a target of 30%. Both France and Italy impose financial penalties and nullification of appointments for companies that fail to meet the quotas. These legislations and voluntary goals aim to promote gender diversity and harness its benefits in corporate boards [7].

Due to its global relevance regarding women’s participation on boards of directors and the limited research available, Latin America was selected as the study area. Despite some progress in including women in high-level executive positions in the region, there remains a significant disparity compared to other regions. However, what makes Latin America particularly interesting is its wide cultural, socioeconomic, and business diversity, which provides an opportunity to examine how the presence of women on boards of directors can influence value creation in such a heterogeneous context. This diversity offers a unique opportunity to examine the impact of women’s representation on corporate boards on value creation within such a heterogeneous context. Notably, a significant disparity exists between the presence of women and men on boards of directors among the analyzed companies, with Argentina, Colombia, Peru, and Mexico exhibiting particularly pronounced gaps [8]. Consequently, it is imperative to contribute to the existing literature in this field and explore how these findings align with those of more developed nations.

Over time, this percentage has been increasing, reaching 19% for the first quarter of 2021 in the most liquid companies in the Colombian capital market, with participation rates exceeding 40% in two of these companies [9]. Although this is encouraging, there is still a considerable gap with the goal established for 2023, i.e., 30% [10].

For other Latin American countries such as Chile, Mexico, and Brazil, the percentages of women on board directors remain too low, i.e., 9%, 8%, and 12%, respectively [11], ranking well below their European counterparts. This contrasts with the fact that 50.8% of the population in the region is women [12], representing 41% of the workforce [13]. Therefore, the low percentages of women’s participation are not due to a lack of female population but to the fact that many women work in operational positions with little participation in senior management and board positions, possibly due to factors such as nepotism [14].

The difference between the percentage of women on Latin American boards and that observed in countries in other regions of the world leads to questions regarding the reasons behind these decisions pertaining to the composition of corporate governance and the imposition of quotas in some European countries such as France and Spain. Some studies have sought to test whether the presence of women on boards modifies how this body fulfills its responsibilities [15], finding a positive impact on factors such as monitoring efforts [16]. However, the empirical evidence tends towards mixed results, where the impact of women depends on the nature of the tasks performed [17]. These results are aligned with the theory in which it is proposed that heterogeneity drives strategic decision-making, creativity, and innovation but has a negative impact on group dynamics, leading to a possible decrease in value [18].

The lack of consensus between theory and empirical conclusions has led to broadening the field of study to determine whether the impacts of gender diversity on managerial tasks positively or negatively affect the financial or non-financial variables of companies, with the most concentrated analyses in the United States and Europe. In terms of profitability, market or accounting metrics have been used, finding an increase in financial outcomes [18,19], but mixed relationships have also been detected [20,21,22,23,24,25,26], with some not finding significant results [27,28,29].

The vision of value generation is currently of great importance for companies considering that the approach should focus not only on the short term but also on a long-term vision with a more inclusive perspective that leads to value generation for all stakeholders [30]. Regarding financial indicators, the one most used in empirical applications is Tobin’s Q, for which the results do not reach a consistent conclusion either [21,31,32]. Notably, the metric economic value added (EVA^®) has only been used once in a study carried out in Pakistan, for which the results were non-significant [33].

Considering that the results of empirical studies are inconsistent and that a large proportion of previous studies have been carried out in European and American countries, there is a great need to investigate the relationship between the presence of women on boards of directors and EVA^®, particularly in Latin American companies, for closing the knowledge gap identified. Thus, this research aims to evaluate the relationship between the participation of women on boards and the generation of value, as measured by EVA^®. For this, a correlational quantitative methodological design was carried out in which conclusions were drawn through the analysis of secondary sources of information. The study carried out a panel data model with random effects for a sample of Latin American companies from 2019 to 2021. This research (1) contributes to the literature on value generation, which should be an organizational priority for the sustainability of a company in the long term [26] and provides a novel indicator of interest that has not been analyzed in great detail, and (2) contributes at the regional level because developing countries in Africa [19,34] and Asia have been studied [33,35,36], but more research is needed in these types of countries, especially in Latin America, where research on this topic is scarce.

The research findings may interest managers and regulators, as they will allow these actors to understand better women’s participation on boards of directors and the financial outcomes. Likewise, evaluating whether women on boards of directors impacts the results of companies in Latin America provides a basis for decision-makers to increase the number of women in the region, ceasing to be simply an initiative of inclusion but rather a strategic move.

The rest of the article is structured as follows. After the introduction, the theoretical framework is presented, which includes an analysis of the main studies pertaining to the object of study. Subsequently, the proposed methodology and a description of the data are presented. Then, the results obtained are analyzed. Finally, a concluding section is presented with final comments, reflections, limitations, and further research.

2. Literature Review

For this study, a comprehensive, systematic, and holistic review of the literature on women on board directors was performed. Due to its excellent academic reputation, the Scopus database was searched to obtain the studies for analysis. Thus, the most relevant studies were selected in order to have the knowledge frontier on women on board.

2.1. Gender Diversity on Corporate Boards

Studies related to women on boards of directors, although not completely new, have increased in recent years, as shown in Figure 1. Of the 2534 studies retrieved, 73% have been published since 2015, peaking in 2021, the year in which 14% of the literature on the subject was generated. Figure 1 shows the number of publications per year, and Figure 2 shows the cumulative publications per year, which shows exponential growth, with the curve fitting almost perfectly to the function 14.598 × e^0.1627X. This trend reflects the growing interest in the subject and the relevance of its study in current times.

Generally, the decision process in companies includes four main steps that can be reduced into two underlying functions, i.e., decision management and decision control, with the first being a more active and purposeful role and the second being a more passive role related to monitoring [37]. According to [38], a board of directors has two main functions: (1) to monitor management on behalf of the shareholders and (2) to provide resources. Of these obligations, the first is addressed by agency theory, which includes effective monitoring as a function of board incentives [38]. The foregoing implies that this administrative body is in charge of 50% of the decision-making process; therefore, its impact on the company’s operation is evident. In addition, the board members are in charge of identifying and reinforcing value-generating activities [39], increasing their importance. The second refers to strategic resources that allow better decision-making, which can be represented in elements such as sector knowledge, relationships with the financial sector, and negotiation and persuasion skills, among others.

The following question arises: what makes boards of directors more effective? Indeed, board members’ skills, experience, expertise, and knowledge play important roles [38]. However, because boards are groups of people seeking to reach a common point, other characteristics may be more inherent to each individual, affecting the dynamics under which the company’s governance develops. In addition, because the role of boards is to direct management through low-cost mechanisms [40], they seek to maximize the benefits of the social dynamics that are generated through interactions between board members.

One of the ways through which this can be achieved is the inclusion of women on boards of directors. Studies indicate that for decision-making tasks, diversity helps to increase the number of solutions offered and the alternatives considered; therefore, mixed boards of directors have the potential to perform better than those composed of only one sex. However, diversity also presents obstacles to fluid interaction processes. Women and men interact in different ways, with the former having a more social focus and the latter being more task-oriented, potentially making them more impatient and thus affecting effective interactions. In addition, on boards with a majority male presence, women tend to feel intimidated, which reduces the potential for better results [41]. Is it thus worthwhile to include women on boards of directors? The theoretical support as well as empirical evidence on the subject is mixed.

2.2. The Impact of Gender Diversity and Corporate Boards on Firm Financial Performance

Most of the variables studied by empirical approaches have been profitability metrics such as ROE (return on equity), ROA (return on assets), and Tobin’s Q. Ref. [16] used ROA and Tobin’s Q as dependent variables to determine the impact of women on boards of directors of US companies. These authors are part of the list of researchers who obtained mixed results regarding whether women on boards of directors have a positive impact, in addition to concluding that gender quotas can negatively impact well-governed companies. Likewise, Ref. [32], through a study of events, also sought to determine the relationship between the presence of women (after the mandate in Norway that demanded a representation of women of 40%) and value from the market point of view, including the price of shares and Tobin’s Q, ultimately yielding negative results. These findings are similar to those reported by [35], who found that women on boards are significantly and negatively associated with the financial stability of a company.

Using ROA and Tobin’s Q, Ref. [27] concluded that non-family companies with women on their boards observe a decrease in their financial measures; for family-run companies, the impact is not significant. Ref. [29] studied 23 different financial indicators; 13 did not show significance, and 10 showed a negative relationship; this study used the difference-in-difference methodology and was motivated by the gender quota policy for listed companies in Belgium.

In contrast, Ref. [42] reached mixed results for ROE and Tobin’s Q; for the former, they found a positive association with the presence of at least one woman on the board of directors, and for the latter, they found that at least three women are required on boards composed of ten or more directors. Something similar was found by [43] in their study in India; these authors found that financial performance improves due to the presence of at least one woman on the board of directors. Likewise, market performance (measured by Tobin’s Q) is positively and significantly related to the percentage of women on boards, indicating that this percentage is relevant. This finding is consistent with results in the literature on critical mass and the importance of having a representative number of independent members in a corporate governance body (in this case, women), thus allowing individuals to have the confidence to communicate their ideas and opinions and have their ideas and opinions taken into account. In this sense, the results found by [44] in Spain support that women’s representation positively impacts companies’ performance by approximately 30%. However, in the study by [45], the findings contrast with this theory because although the performance of companies increases as the proportion of women on the board increases, there is a critical point (33% of female representation) beyond which the addition of women decreases the value of the company, indicating that the costs of gender diversity predicted by social identity theory (for example, lack of communication, disagreements, conflicts and loss of trust) outweigh the potential benefits of diversity predicted by the theories of agency and resource dependence.

For the same market variable mentioned previously, Ref. [46] found a non-significant negative relationship, and from the perspective of profitability variables, such as ROA and operating profit, companies with boards of directors with greater gender diversity perform better.

Non-significance was also found by [47], one of the few studies conducted in Latin America using Tobin’s Q. Nevertheless, they found that companies with at least one woman on the board outnumber those with none. However, the conclusions of [48], who, using a mix of book and market value, found non-significance, warn of the importance of carefully interpreting the results because female leadership can alter other aspects of companies. To further explain non-significant results, Ref. [36] notes that such results may occur because the number of women on boards in Malaysia, where the research was carried out, is too small, altering the statistical results, despite having board size as a control variable. A similar conclusion found by [28] was reached in Indonesia, where the authors associate non-significance with the low presence of women in the sample, with the highest percentage being 75% but the average being only 21%. Importantly, for studies in regions such as Latin America, the participation of women was found to be 7.3% [42].

On the other hand, Ref. [23] found that the presence of women on boards increases returns, such as ROA and ROE, but decreases Tobin’s Q. However, under the assumption that the relationship between women’s participation and performance is affected by certain attributes, the researchers controlled the analysis with nine variables, thus eliminating the negative effect on Tobin’s Q. In contrast, Ref. [49] conducted a study with data from 12 developing countries in Central and South East Asia and [50] with data from the United Kingdom and obtained a positive relationship with Tobin’s Q. Additionally, Ref. [51] by analyzing European countries, found an increase in value, also measured by Tobin’s Q, but they focused on women with positions of power such as CEO (Chief Executive Officer) and CFO (Chief Financial Officer). However, an interesting finding was that the positive impact is greater for countries that are governed by a woman.

In this line, Ref. [26] identified an absence of impact on ROA and ROE but positive statistical significance for price-to-book and equity returns. These researchers also highlight the sensitivity of requiring the incorporation of women into boards because they may lack the necessary characteristics to perform their roles correctly. For the subject demand, Ref. [52] came to a similar conclusion from the risk perspective. Their study of 39 countries observed that individual and systemic risk increased when banks incorporated women on their boards to comply with imposed gender quotas. Ref. [53], using a sample of 461 large banks from 24 OECD countries, found a negative relationship between the role of women on boards and indicators related to risk-taking but a positive relationship with performance metrics such as ROA, ROE, Tobin’s Q, and operating income.

In Colombia, Ref. [24] evaluated the relationship between ROA and ROE with the presence of women on boards, the participation of women in top management teams, and women as CEOs. They applied three models with inconsistent results regarding the significance of the variables of interest. For [24], the positive effect between gender diversity on the board and company performance was only significant when measured by ROE.

Ref. [54] considered a different approach and sought to relate the proportion of independent women on boards of directors with the performance of companies in corporate social responsibility, finding a positive relationship, limiting the result with the level of orientation of the company’s consumer market. On the other hand, Refs. [20,21] used both financial and non-financial metrics. Both studies reached mixed results for financial variables but did detect an improvement in non-financial outcomes. Ref. [15] studied countries in different parts of the world and evaluated financial performance (ROA and ROE, among others) and non-financial factors. Although the market metrics were not significant, they were better for companies with more women on boards in countries with higher gender parity, implying that the results depend on the characteristics of the territory where the research is carried out. In contrast, Refs. [31,55] studied Tobin’s Q in Vietnam and the United States, both reaching positive results despite being countries with such different economic and social conditions, somewhat contradicting the aforementioned.

In Tunisia, Ref. [19] used three profitability measures, i.e., ROE, ROA and Z Score, finding that all three are statistically significant and are positively related to the gender diversity of the board. Ref. [56] contributed to the literature by studying an under-studied variable: cost of debt. They found that the cost of debt is negatively related to the presence of women on boards. In addition, they emphasized that it is necessary to reach a critical mass to have a favorable impact on that cost.

On the cost side, Ref. [57] analyzed the relationship between gender diversity on boards and the rigidity of costs, finding a positive effect that was greater when companies have high agency costs, lower corporate governance, lower risk aversion, and untrustworthy managers. Ref. [30] also provided an innovation concerning evaluation methods. They compared returns by building portfolios based on board diversity, concluding that gender diversity can be associated with better share price performance. Through inferences and a literature review, Ref. [58] used data from European countries to assess ROA, ROE, and Tobin’s Q trends when women are present on boards of directors in companies in Colombia. According to their analysis, the relationship between these study metrics in this country is positive.

Ref. [59] focused on the bid/ask spread of UK companies, seeking to assess whether gender diversity on boards of directors affects the disclosure of corporate risks, obtaining positive results. In that same country, Ref. [60] used 27,352 private companies and found an inverse relationship between the presence of women on board directors and a company’s risk management, as measured by the standard deviation of EBIT (earnings before interest and taxes) deflated by total assets, the difference between maximum and minimum ROA and the standard deviation of operating cash flow deflated by total assets. An important conclusion was that a company’s risk level depends not only on whether women are on the board directors but also on whether they are local or foreign, suggesting that for private companies, knowledge of the local market is valuable.

Likewise, Ref. [61] analyzed the impact of gender diversity on board directors to explain the relationship between bank disclosure and the expected probability of banking crises. Considering a sample of companies from 42 African countries, it was found that having women on board directors leads to more prudent decision-making that positively impacts the stability of banking systems through greater disclosure, reducing the possibility of a banking crisis.

On the other hand, Ref. [62] took a different approach to profitability and focused on the return generated by mergers and acquisitions. They found a negative relationship, concluding that after the crisis, the abnormal returns generated by merger and acquisition announcements are lower for banks with at least one woman on their board directors than for those whose directors are all men. When studying the generation of value, as measured by the economic value added (EVA^®) [33], after evaluating Pearson’s correlation coefficient and performing linear regression analysis, it found no significant influence of having women on Pakistani boards.

According to the literature analyzed above, the impact of the presence of women on boards of directors can vary in significance based on the country or region being studied. The latter may be related to the theory of diversity in board directors, i.e., the possible decrease in value is related to clashes between men and women. Therefore, by studying countries with greater female empowerment, it is possible to overcome those barriers that lead to the loss of benefits of having different points of view.

2.3. Women on Boards and Firm Value in Latin America

In the Latin American case, the presence of women on boards of directors has not evolved much. Since 2008, the percentage of women on board directors has been increasing slowly, reaching 7.3% in 2018, ranking only above the Middle East. This slow evolution is common in both developed and developing countries.

However, there is momentum in countries that actively encourage the participation of women through mechanisms such as gender quotas [63]. Nevertheless, studies on the results of these impositions show that they can be counterproductive if directors do not have the necessary experience, impacting leverage and operating performance [32].

The foregoing leads us to propose that the growth in participation should be a natural result of the managerial characteristics of women. However, by not currently having these mandatory quotas, women on Latin American boards have been placed in these positions by purely managerial decisions, raising questions about whether their performance is consistent with the negative or positive side of the literature and previous research.

Within the few studies found that have examined the relationship between gender diversity on boards of directors and various aspects in Latin America, Ref. [64] found that gender diversity does not have a significant effect on corporate social responsibility related to environmental issues. However, they highlighted the inadequacy of board composition in addressing institutional gaps and promoting sustainability.

On the other hand, a previous study [42] showed a positive association between the presence of women on boards of directors and financial performance in Colombia, while [65] demonstrated that the participation of women on boards increases the likelihood of success in crowdfunding campaigns, suggesting a positive impact of gender diversity on access to financing in the region.

Using the literature reviewed above regarding the relationship between gender diversity and company performance, with the most relevant being [18] and the findings by [66] on EVA^®, a conceptual model is proposed, as observed in Figure 3, which shows the variable of interest, the independent variable and the control variables used, as well as the expected relationship for each. With this model as a reference point, the following hypothesis is proposed.

Hypothesis. 

The relationship between the presence of women on board directors and the generation of value, measured by EVA^®, is direct and significant.

To test the research hypothesis, the following model is proposed:

$$EV{A}_{i,t}={\beta }_0+{\beta }_{1}WOME{N}_{i,t}+{\beta }_{3}SIZ{E}_{i,t}+{\beta }_{4}BOAR{D}_{i,t}+{\beta }_{5}INDUSTR{Y}_{i,t}+{Ɛ}_{i,t}$$

(1)

The measurement of these variables can be observed in

Table 1. Variables in the quantitative model.

Variable	Metric	Type	Category	References
EVA	Economic value generated in the year by the company	Continuous	Dependent	[33]
MEJD	Number of women on the board directors/Total members on the board	Discrete	Independent	[19,47,57]
IMEJD	Presence or not of at least one woman on the board	Categorical	Independent	[26,44,67]
Size	Natural log of the total assets of the company in dollars	Continuous	Control	[22,24,36]
Board	Natural log of the number of members of the board directors	Continuous	Control	[18,55]

. For the estimates pertaining to women, there are two variables: MEJD and IMEJD.

3. Methodology

This quantitative correlational study sought to draw conclusions through analyses of secondary sources of information. The analysis was carried out using a panel data model with random effects for a sample of Latin American companies from 2019 to 2021. The data for the research were obtained from the ASSET4 ESG database of Refinitiv Thomson Reuters.

The panel data method has been used by other relevant studies [16,22,24,26,47,55,68,69,70]. Although other authors have preferred to use cross-sectional studies [14,27], using a panel data analysis is useful because it expands the number of observations available to the model. In addition, panel data allow more accurate inferences (higher degrees of freedom) to be obtained, control the impact of missing variables, and provide microfoundations to analyze aggregated data [71]. In this way, the following research question is posed: does the presence of women on board directors affect the generation of value?

3.1. Dependent Variable

The study variable for the research is EVA^®. This indicator is of interest because it not only evaluates the ability of a company to generate profits from its operations but also contrasts those profits with the total cost of capital used to generate them. The formula is as follows [71]:

$$Economic Value Added=NOPAT-Cost of Capita\times Invested Capital$$

(2)

NOPAT is the operating profit after taxes. Cost of capital, as measured by the WAAC (weighted average cost of capital) is a weighted average, based on the company’s financing structure, between the cost of equity and the cost of debt. For the calculation of the return to the investor, the CAPM (capital asset pricing model) is traditionally used, and the cost of debt is an explicit value that is reduced in a small proportion due to the tax shield [71]. Invested capital represents the investment in assets necessary to generate the NOPAT.

The philosophy behind the indicator is that investors should receive a return that compensates for the risk assumed when depositing their capital in the company and is at least similar to that paid by the stock market. If this condition is not met, the company decreases in value through its operations; therefore, the minimum accepted value of EVA^® is zero because this result is sufficient to avoid generating losses from the investor’s point of view [71]. Notably, this indicator involves two parts of the company that can be positively or negatively impacted by the presence of women on boards. The first is the operating side (represented by NOPAT) through, for example, greater efficiency in production or service delivery resulting from greater board control or an aggressive sales strategy that leads to more operating income (because this administrative body can influence CEO decisions through methods such as CEO compensation). The second is the ability of companies to acquire resources at a lower cost (represented by the WACC), and as shown by [56] when analyzing the cost of debt (one of the components of the WACC), it can be negatively associated with the presence of women on boards. This indicates that the greater the presence of women on board directors, the lower the debt cost. For this study, the following formula is used for EVA^®:

$$Economic Value Added=\left[EBIT\times \left(1-Tax rate\right)\right]-\left[WACC\times Invested Capital\right]$$

(3)

This transformation is possible because

$$NOPAT=EBIT\times \left(1-Tax rate\right)$$

(4)

$$EBIT=Earnings before interest and taxes$$

3.2. Independent Variable

The independent variable for this analysis is the presence of women on boards of directors. This variable was included in the model for the following two reasons:

1. Employing an indicator that takes values of 1 if there is at least one woman on the governing body and 0 if all the directors are men, this study seeks to make a distinction between groups to help reflect whether those boards that have a female perspective have a different result regarding value generation.

2. The relative weight that women have within board directors is measured as the percentage of women among the total number of board members to account for the results of [27,55,67] that indicate that the effects of relative numbers in group interactions can considerably change the interpretation of the results [72].

3.3. Control Variables

Control variables are established that help isolate external effects and allow the true impact of the independent variable on the dependent variable to be evaluated as follows.

1. Company size: In addition to the theoretical support that exists for this variable through the investigations carried out, it is considered important to include in the model, given the inherent relationship that this factor has with the dependent variable. Larger companies have competitive advantages that allow them to obtain greater market shares and cost efficiencies as well as facilitate access to the capital market [73]. With this, the two main measures of EVA^® are impacted: the operating profits, which can increase even in the face of declining income (which gives them an advantage even when there is low economic activity), and the cost of capital, which is lower when there are different sources of financing.

2. Board directors’ size: The role of board directors within companies’ operations has already been mentioned. Several authors have attempted to determine whether this control body’s size influences its functions’ performance and thereby affects company results. Some researchers, such as [74], argue that large boards may fail to control the CEO successfully, and therefore, boards should not have more than seven or eight members. Ref. [75] notes that when there are many members, it becomes challenging for everyone to express their ideas and opinions in their limited time, preventing proper communication and generating a lack of cohesion. Ref. [76] has a contrary position, based on empirical evidence that the number of members has an inverse relationship with the cost of financing, suggesting that larger boards provide greater monitoring of financial accounting processes.

Although there are discrepancies regarding whether large board directors benefit a company, it is not denied that there is an impact on financial outcomes; therefore, it is reasonable to consider this variable as a control for the model.

3. Industry: Although each company has its own factors that influence value generation, depending on the industry to which it belongs, it must face different threats and opportunities. Companies in industries with a high level of research and development have a different cost structure and capital needs with a higher level of risk than other industries where services are provided or basic products are offered that do not require innovation as a competitive factor. Factors such as economies of scale, switching costs for customers, barriers to entry, and competitive structure are determined by the characteristics of the industry [77]. Given the inherent differences between companies given by this variable, it is included as a control in the model.

3.4. Data

The years initially contemplated for this study were from 2016 to 2021. However, due to the availability of information, the years from 2016 to 2018 were excluded, leaving a window of time from 2019 to 2021. Additionally, two companies were eliminated from the sample, given that the reported values for some variables lacked financial reasonability. Thus, the data for 202 companies from Argentina, Brazil, Chile, Colombia, Mexico, Panama, Peru, Puerto Rico, and Uruguay are included. The distribution of the companies by country is provided in Figure 4. Forty-seven percent of the companies have their center of operations in Brazil, a finding that is reasonable because Brazil is the largest country in terms of area (km²) and population in the region. Chile has the third greatest number of companies, although it ranks seventh in terms of size.

Table 2. Descriptive statistics of the initial continuous variables.

	EVA	NOPAT	WACC × Capital	Company Size	MEJD	Board Size
Average	−21	551	563	15,328	0.11	10
Minimum	−5755	−573	9	274	0.00	2
25%	−150	93	114	2306	0.00	7
50%	−9	207	245	4647	0.11	9
75%	77	539	569	12,568	0.17	11
Maximum	16,974	21,672	10,133	368,523	0.50	23
Standard Dev.	994	1306	999	38,121	0.11	3

shows the descriptive statistics of the variables used before the logarithmic transformation. This table includes the values for NOPAT and WACC multiplied by capital, which are the two components of EVA^® for analysis purposes; however, they are not explicitly included in the model. These variables, together with EVA^® and company size, are expressed in millions of US dollars.

The EVA^® financial performance metric has a negative average of USD 21 million, indicating that, on average, the companies under study do not have the capacity to generate value. However, the value for the 50th percentile is close to the mean and is not far from zero. In addition, the maximum level is considerably higher than the mean and the 75th percentile, indicating the existence of extreme values, as seen in their deviations and in the breadth of the range (USD 22.729 million). The variables NOPAT and WACC × Capital have very close averages, explaining why EVA, despite being negative, has a small absolute value. The significant difference is observed in its maximums, where the cost of capital is almost half of the maximum NOPAT value.

As measured by assets, the average size of the Latin American companies in the sample is USD 15,328 million. However, as with the other variables, there are considerable changes from one company to another, with a standard deviation of USD 38,121 million. This may be due to factors such as business maturity and asset needs (because some economic activities require less investment of assets than others, such as service companies).

The board directors’ average of companies is composed of 10 directors, with the smallest consisting of 2 people and the largest consisting of 23 members. Contrary to what was previously observed, the variation, in this case, is not large; the mean is close to the 50th percentile, and the maximum is 2.5 times that value. Therefore, despite the differences in company size, the number of members seems to be more controlled, indicating that a company can grow in size while keeping the board of directors unchanged. The purpose of maintaining the board size may be to achieve efficiency in terms of costs, but it could create difficulties in monitoring tasks. The average percentage of women among board members is 11%, which indicates that there is one woman per board if related to the size of this corporate body. However, with a 50th percentile of 0%, this mean results from having a high maximum value. With a maximum of 50%, women within the boards of the sample do not exceed the number of men.

Table 3. Descriptive statistics of the transformed variables.

	EVA	NOPAT	WACC × Capital	Company Size	MEJD	Joint Size
Average	−21	551	563	9	0.11	2
Minimum	−5755	−573	9	6	0.00	1
25%	−150	93	114	8	0.00	2
50%	−9	207	245	8	0.11	2
75%	77	539	569	9	0.17	2
Maximum	16,974	21,672	10,133	13	0.50	3
Des. Standard	994	1306	999	1	0.11	0

shows the statistics after the logarithmic transformation of board and company size. With this transformation, the wide variation in assets decreases. However, the other variables remain unchanged.

3.5. Multicollinearity

Multicollinearity can represent a problem in analysis because it makes it difficult to distinguish the effect of one variable and the effect of another, which translates into inflated standard deviations that could alter the significance tests and computational complications due to the instability of the data. To determine if multicollinearity is a problem, the size must be determined. The higher the correlation of one variable with another, the more multicollinearity there is [78].

The correlation matrix of the variables used in this study is shown in

Table 4. Correlation matrix.

	EVA	MEJD	IMEJD	Size	Board	Industry
EVA	1.0000
MEJD	−0.0162	1.0000
IMEJD	−0.0055	0.7673 ***	1.0000
Size	−0.0248	0.0275	0.1344 ***	1.0000
Board	−0.0407	−0.0287	0.1701 ***	0.2813 ***	1.0000
Industry	0.0799 *	0.0050	−0.0923 **	−0.2920 ***	−0.1648 ***	1.0000

Note: *, **, *** indicate significance at 10%, 5% and 1%, respectively.

. The two variables related to gender diversity in the board directors are non-significantly correlated with value generation, and there are no high correlations with the other explanatory variables of the model. However, the two are significantly positively correlated (77%), which is to be expected because of the way IMEJD is constructed. This situation does not alter the result because MEJD and IMEJD are not evaluated together.

The control variables Size and Board are negatively correlated with the generation of value but positively correlated with each other. The latter makes sense given that as a company grows, its corporate governance needs to be more robust, which usually translates into larger board of directors. In general, no high correlations are found that interfere with the model estimates.

3.6. Heteroscedasticity

To evaluate the possible existence of heteroscedasticity, a Breusch‒Pagan test was carried out; the null hypothesis is that the errors are homoscedastic, and therefore have constant variance. The alternative hypothesis supports heteroscedasticity. Two tests are run independently, using MEJD and IMEJD. The result in both cases was a p-value greater than 5%, which does not allow rejecting the null hypothesis and indicates that there is no heteroscedasticity.

3.7. Panel Model Selection

The Hausman test was performed to decide whether to use a fixed or random effects model. The null hypothesis for the former is that the model is consistent and efficient, and the null hypothesis for the latter is that the model is not consistent. The results of the test are shown in

Table 5. Hausman test results.

	Statistical	p-Value
Model 1—IMEJD	5.1043	0.1643
Model 2—MEJD	6.6347	0.0845

for both models. Considering that the p-value is greater than the significance value of 0.05, the null hypothesis is not rejected, and a random effects panel data model is chosen.

4. Results

To meet the objective of this research, it was necessary to obtain data related to the composition of the board directors and the financial outcomes of the companies.

Table 6. Model 1 results.

	Results with Industry as a Control	Results without Industry as a Control
IMEDJ	86.916 (114.994)	13.117 (101.303)
Size	54.126 (67.093)	−11.958 (44.802)
Board	−262.142 (185.668)	−186.092 (162.330)
Industry	Yes *	No
Observations	585	585
R2	0.0462	0.002

Note: The first number in the cell is the regression coefficient, the number in parentheses is the standard deviation, and the significance is indicated by * at 10%.

shows the empirical results for the model using the presence of women on the board directors as a variable; this indicator takes values of one if there is one or more women on the board and zero otherwise. The model is run twice: including and not including the control variable Industry. The R² results obtained indicate that the models do not have a good fit for the data. For the gender variable, a positive value is obtained in both models, but it is not significant; therefore, a company’s value generation will not be significantly impacted by the presence of women on the board.

Additionally, the coefficients indicate that the effect of gender diversity is greater when the control variable Industry is included than when it is not included; however, the standard deviation will also be greater. The results using Model 2, in which gender diversity is measured by the percentage of women on the board, are shown in

Table 7. Model 2 results.

	Results with Industry as a Control	Results without Industry as a Control
MEDJ	−2.6694 (5.0190)	−3.9670 (4.4246)
Size	58.3476 (67.2728)	−10.7933 (44.8402)
Board	−236.2349 (182.8018)	−187.6337 (160.5679)
Industry	Yes *	No
Observations	585	585
R2	0.0455	0.004

Note: The first number in the cell is the regression coefficient, the number in parentheses is the standard deviation, and the significance is indicated by * at 10%.

. This new model is similar to the previous one, as the R² indicates a poor fit. Therefore, controlling by industry increases the impact of gender diversity, and the presence of women in the controlling body does not significantly impact EVA^®.

However, for this variable, the result changes the sign, indicating that the value generation decreases as the relative weight of female opinion on the board increases. Therefore, leaving aside the significance and analyzing only the meaning of the coefficients, it can be concluded that having a woman on the board can be beneficial for EVA^® but that as more women are incorporated and their percentage within the board grows, the impact on value generation becomes negative. Therefore, the hypothesis is not supported.

4.1. Control Variables

Regarding the control variables in both models, the results were consistent. For company size, measured as the natural logarithm of assets, there was a positive impact when controlling for industry and a negative impact when that variable was ignored. This is an interesting finding considering previous studies and the characteristics of large companies. Some field investigations have found a significant positive impact of this variable [32], and others have found a negative impact [18]. The reasons for the increase in value generation in large companies are mainly associated with the ability of such companies to generate economies of scale and the possibility of having lower capital costs, given that their high level of assets can serve as a guarantee and that their size facilitates access to different sources of funding. When faced with a negative impact, large companies tend to be more diversified, with higher agency costs and bureaucratic costs [79]. The models’ findings in this study indicate that higher agency costs in large companies have more relevance in some industries than the potential benefits of the operational efficiencies that can occur in large companies.

For the variable board size, both models had a negative relationship regardless of whether the model was controlled by industry. However, contrary to what occurs with the gender diversity proxy, in Model 2, the joint effect increases (becomes more negative) when the other control variable is incorporated. This relationship with board size indicates that many directors generate inefficiencies in control, and therefore, there are higher agency costs. Suppose this result is extended to the percentage of women on board directors. In that case, it can be intuited that the generation of value for the companies under study is negatively impacted by having different opinions on the board of directors (either because of a diversity issue or because of the number of members).

Comparing the variable board size with company size (something that can be carried out because both are standardized), the negative effect of very large boards is greater than the positive effect of larger companies (when controlling for the type of industry). As mentioned above, the two models were estimated twice to reflect the impact of the control variable Industry. The results show that incorporating this variable in the model improves the fit by increasing the R² and that its presence can change the direction of the effect of other variables in the models. This suggests that the relationship between value generation and gender diversity varies by Industry.

4.2. Analysis by Year

The previously presented results consider information from 2019 to 2021 because data availability for all companies did not allow for a broader time window. Given that in 2020 there was a worldwide health emergency that impacted the dynamics of all companies, it is possible that the findings from Models 1 and 2 are biased and that EVA^® is affected by external factors that do not reflect true relationships with gender diversity.

With this in mind, a regression was performed using Models 1 and 2, but omitting the time variable and using all the data available for each year independently. The results of this approximation are shown in

Table 8. Model 1 and Model 2, results by year from 2016 to 2018.

		Model 1		Model 2
Year	Variables	With Industry as a Control	Without Industry as a Control	With Industry as a Control	Without Industry as a Control
2016	Women	−4.55	45.19	−9.81	−3.56
		173.82	139.31	9.80	7.76
	Size	13.53	53.11	17.31	52.09
		70.87	48.80	70.55	48.75
	Board	−154.53	−67.59	−141.67	−42.31
		239.07	197.96	234.46	192.27
	Industry	Yes	No	Yes	No
	Observations	157	157	157	157
	R2	0.23	0.008	0.24	0.009
2017	Women	32.70	128.07	−2.01	0.90
		123.57	97.40	5.36	4.36
	Size	−26.85	−12.96	−26.03	−14.37
		49.09	33.65	49.13	33.79
	Board	−533.97 ***	−571.19 ***	−522.49 ***	−521.71 ***
		166.74	140.11	159.97	135.44
	Industry	Yes	No	Yes	No
	Observations	188	188	188	188
	R2	0.21	0.09	0.21	0.08
2018	Women	−24.15	63.94	−1.71	1.33
		125.08	98.16	6.00	4.90
	Size	−74.34	−112.30 ***	−73.79	−111.50 ***
		58.07	36.91	57.89	36.95
	Board	−568.41 ***	−528.29 ***	−572.66 ***	−508.16 ***
		173.37	143.46	166.62	139.59
	Industry	Yes	No	Yes	No
	Observations	179	179	179	179
	R2	0.29	0.15	0.29	0.14

Note: The first number in the cell is the regression coefficient, the number in parentheses is the standard deviation, and the significance is indicated by *** at 1%.

and

Table 9. Model 1 and Model 2, results by year from 2019 to 2021.

		Model 1		Model 2
Year	Variables	With Industry as a Control	Without Industry as a Control	With Industry Control	Without Industry as a Control
2019	Women	−0.95	9.14	−1.19	−1.18
		109.95	89.70	5.42	4.38
	Size	−109.59 **	−101.479 ***	−108.49 **	−100.18 ***
		46.22	31.47	46.28	31.47
	Board	−79.01	−71.08	−79.07	−69.95
		158.28	128.63	155.11	126.85
	Industry	Yes	No	Yes	No
	Observations	203	203	203	203
	R2	0.14	0.06	0.14	0.06
2020	Women	70.91	19.34	6.35	4.31
		125.93	105.89	5.74	4.84
	Size	140.76 ***	96.84 **	139.12 **	95.986 **
		53.84	38.14	53.74	38.09
	Board	64.43	−2.62	75.33	4.92
		175.54	149.17	171.41	147.15
	Industry	Yes **	No	Yes **	No
	Observations	245	245	245	245
	R2	0.17	0.03	0.18	0.03
2021	Women	131.31	9.61	−5.97	−5.91
		213.35	181.04	8.80	7.37
	Size	139.19	34.83	159.94 *	39.71
		93.15	63.40	93.51	63.37
	Board	−235.56	−241.77	−214.19	−253.43
		302.41	254.56	298.31	251.53
	Industry	Yes	No	Yes	No
	Observations	239	239	239	239
	R2	0.16	0.004	0.16	0.007

Note: The first number in the cell is the regression coefficient, the number in parentheses is the standard deviation, and the significance is indicated by *, **, *** at 10%, 5%, and 1%, respectively.

. Applying the regression by year, R² improves for Model 1 for 2016 and decreases with increasing years. For example, company size begins to be significant between 2018 and 2020, and board size becomes significant from 2017 to 2018, in addition to maintaining the sign of the relationship initially found. Regarding gender diversity, the negative effect when increasing the percentage of women is sustained; however, the impact is non-significant for the 24 regressions.

However, despite the non-significance, the participation of women within board directors has been evolving positively, as shown in Figure 5, with Panama being the country with the highest growth, increasing from an average of 0% in 2016 to 13.6% in 2021.

5. Discussion

The inclusion of women in high-level positions has been promoted through legal policies such as gender quotas. From this perspective, together with compliance with Sustainable Development Goal 5, Gender Equity, this encourages the study of the impact of gender diversity on board directors on company performance. For corporate governance, agency theory holds that diversity across board directors leads to more independent members who can better execute control and monitoring functions, reducing agency costs [31]. However, the independence of members can slow the flow of communication processes, delaying decision-making and increasing organizational costs.

The setback in potential benefits from gender diversity in board directors is presented not only in theory but also in empirical studies where there are no conclusive results. The results range from significantly positive to negative or non-significant and can change drastically depending on the variable of interest (based on the indicator one wants to use to measure financial performance) or the location of the group of companies under study. This study sought to answer whether, for 202 Latin American companies, having women on board directors between 2019 and 2021 positively impacted value generation.

The measure used in the models was EVA^®, and the approach to gender diversity was from the perspectives of absolute presence (at least one female director) and relative presence (the weight that women have within boards based on the number of total members). This approach was taken to see if it is necessary to count with a minimum percentage of women’s participation to obtain significant results or if the mere presence of a female director on the board can change the financial performance.

The panel data analysis showed that the relationship between EVA^® and gender diversity on boards is not significant, which does not allow for rejecting and accepting the proposed hypothesis. Therefore, the change in dynamics that women bring to boards is not reflected in value generation. This finding is aligned with the results reported by [33], who conducted the only study in which the same variable of interest is used. Furthermore, the non-significance of this relationship is also maintained when linear regressions are run for each year from 2016 to 2021.

There may be factors that lead to these results. In the case of highly diverse boards with an extensive decision-making and monitoring process, it may be that having female directors does not generate significant changes that translate into better performance since the benefits of the feminine vision are already incorporated from other fronts. On the other hand, in the event of hostile board environments that do not allow open discussions, having female directors (at least in low numbers) may not lead to significant change as their expertise, perspective, and values are stifled.

As mentioned in previous sections, the models include three control variables that, according to what has been found in the literature, can impact a company’s value generation. For most of the years of study, it was found that two variables (firm size and industry) had a significant impact on the economic value added. The effect of the company size depends on whether the industry variable is included or not, being positive when both are combined, which implies that a larger size can lead to better levels of profit using capital with a lower cost. The results related to board size are negative but not significant, reflecting the costs of inefficiencies generated when the decision groups are too large, affecting the process.

In general, with the results obtained, it is not possible to affirm that the presence of women on the board improves the decision-making for better financial results. Still, there is no evidence that heterogeneity leads to inefficiencies that negatively impact the generation of value.

6. Conclusions

Academic interest in the role that gender diversity plays in the financial variables of companies has been growing in various parts of the world. Developing countries have been joining this trend in recent years; therefore, studies with this regional focus help to create a literature framework that closes the existing gap. In the case of Latin America, this study provides the first conclusions on a variable not widely studied as it is economic added value, finding a non-significant positive relationship between this and the presence of women on the board of directors.

The empirical conclusions of this study should not be taken as a bad result per se. Although there are no significant results, the participation of women on board directors is highlighted, given their independence, knowledge, expertise, and risk control. Thus, continuing with policies that promote gender diversity improves companies’ operations. This is also reflected in the evolution of the percentage of women being included on board directors, which in 2016 was 8.04% but increased to 11.24% by 2021.

From a managerial perspective, these findings suggest that increasing women on board directors should not be seen as a guaranteed pathway to immediate economic value added. However, it is crucial to acknowledge the expertise and risk management skills that women bring to the table. Companies should focus on creating an inclusive and supportive environment that allows diverse perspectives and experiences to thrive. By harnessing the full potential of gender diversity, companies can benefit from enhanced decision-making processes and improved risk management strategies. Also, it is worth noting that the results of studies examining the relationship between women on boards and financial performance vary globally. While this research indicates no significance in Latin American companies, other studies have reported positive associations in different regions. Therefore, it is essential to consider contextual factors and industry-specific dynamics when evaluating the impact of women on financial performance. Despite mixed results, women’s expertise and risk management abilities serve as valuable assets that should not be overlooked in efforts to promote sustainable and effective corporate governance practices worldwide.

The limitations of the study should be considered. First, there is no large sample for the region or a consistent number of observations that allows the same companies to be studied over a long time. Second, given that the information was obtained from an international database, small companies that are not forthcoming with their information and that may present different dynamics were probably omitted, which may alter the results; therefore, the results cannot be generalized.

Further research should investigate issues such as the incidence of women in managing financial and non-financial risks, such as climate risk, as well as the impact on the cost of equity, cost of debt, and capital structure. Furthermore, new research could include additional control variables, such as expertise, corporate governance practices, and independent board members, in order to foster an all-encompassing comprehension of the paramount significance of women’s involvement in high-level positions. These topics can be of great interest given that they have not been explored in great detail and the trend of diversity research on board directors is expanding to other fields, where gender aspects and other characteristics that influence decision-making are investigated, including how the roles of the board directors’ members are exercised.