Sustainability of Financial Performance in Relation to Gender Diverse Boards: A Comparative Analysis of French and Romanian Listed Companies on Stock Exchanges

While the moral argument for gender diversity has already been made and is incontestable, and more and more economical arguments have been brought to support the business case for the presence of women on the boards of directors of publicly listed companies, the bottom-line issue of what kind of impact gender diverse boards have on firm financial performance is still unclear. The aim of this paper is to deliver a comparative analysis of the impact of gender diverse boards on firm financial performance in France and Romania. Our results do not to provide any evidence of a link between boards’ gender diversity and companies’ financial performance, but while the analysis has failed to find a positive link between female presence and firm financial performance, it has not outlined a negative one.


Introduction
The corporate board of directors and its role have received increased attention in the past years as a result of several cases of corporate failure, such as Lehman Brothers and the subprime crisis, and corporate fraud, such as Enron, WorldCom, Parmalat, and Bancorex [1][2][3]. We can also include in this list the allegations concerning Carlos Ghosn, one of France's most famous executives who was arrested in Japan. In fact, boards of directors are the ones to take the blame when companies are not performing, and they are in charge of making sure that the creation of shareholders' value is achieved [4]. Regardless of the differences that exist between European jurisdictions, all boards must perform the following range of roles: setting direction, marshalling resources, controlling and reporting, and evaluating and enhancing. Each of these roles is fulfilled by a multitude of tasks such as financial control, strategic marketing, and HR and risk management [5]. In their role of setting the direction, boards are in charge of the "big picture" issues of the firm and are the ones that make the strategic decisions and establish the strategy of the firm [6]. The structure of the board of directors, similar to its role, can also vary from one country to another. Some countries have a unitary system of corporate governance, while others have a dual system, and some, such as Romania and France, allow companies to choose between the two. Most French companies use a one-tier model, while most Romanian companies have opted for a two-tier model. In the unitary system, the shareholders elect the directors at the annual shareholders' meeting. The board of directors in the unitary system can include both executive and non-executive members, and it is responsible for all aspects of a company's activity.
Looking at the presence of women in the board room, arguments in favour fall into two categories: moral and economic [3]. The moral argument is unequivocal: gender theory framework, where we explore five different theories to explain the link between the gender diversity of boards of directors and firm financial performance: the agency theory, the resource dependency theory, the upper echelons theory, the human capital theory, and the social psychological theory. Section 3 presents the materials and explains the method, defining the regression model that we used and the financial performance measures, namely ROA (return on assets) and the Tobin's Q ratio. The next sections of the article analyse the results, followed by the discussion and conclusions.

Literature Review
With the introduction of legal requirements with respect to gender diversity on company boards of directors, research interest in the matter has increased significantly. Despite the great numbers of studies regarding this subject, empirical evidence about the effect of gender diversity of boards on financial performance appears to remain inconclusive.

Summary of Literature Review
Some studies (Dale-Olsen, et al. [13], Rose [14], Miller and Del Carmen Triana [15]) have found no link between boards' gender diversity and companies' financial performance. Post and Byron [16] meta-analysis of 140 studies offers a good overview of a multitude of studies. It finds that there is a different relationship between gender diversity and financial performance depending on the measure used and that the intensity of the link is also influenced by other factors. An insight into the difference between the results for the two measures might come from Lee and James [17]. In their study of the announcement of CEOs for all companies listed on the US stock exchanges between 1990 and 2000, they found that the fall in stock prices after the appointment of a new CEO is greater when the person appointed is a woman. Findings by Lee and James [17] are also supported by those from Haslam, et al. [18] who find that while there are no relationships between the presence of women on the board and accounting-based measures of financial performance (ROA, ROE), there is a significant negative correlation between market-based measures and the presence and percentage of women on the board. In addition, they also find that companies with all male boards had higher values of Tobin's Q than those that had any female presence. Campbell and Mínguez-Vera [3], after looking into companies listed on the Spanish stock exchange in 1995 and 2000, found that gender diversity has a positive effect on firm performance as measured by Tobin's Q. The study also uses the Blau and the Shannon index to assess the presence of women on boards of directors and shows that firm value is insignificantly affected by firm gender diversity and that the key for more effective boards is the right mix of both men and women. In addition, they conclude that Spanish investors view companies with more women on the board in a positive light and that gender diversity is likely to have a positive economic impact. Vafaei, et al. [19] in their study of the Top 500 listed firms on the Australian stock exchange find that there is a positive and significant association between female presence on the board of directors and the company's financial performance. They use accounting-based measures (ROE, ROA), a market-based measure (Tobin's Q), and an economic-based measure (CFO/TA); address endogeneity issues; and test for the influence of both female presence on the board and the percentage of females on the board. Carter, et al. [20] find no link between board gender diversity and firm performance for S&P 500 firms between 1998 and 2002 after previously having found that US firms with more women on their boards of directors have better financial performance [10]. Adams and Ferreira [21] find that gender diversity has a negative effect on average on firm performance. They use ROA and Tobin's Q as a measure of financial performance and analyse US firms between 1998 and 2003. Lückerath-Rovers [7] studies 116 publicly listed Dutch companies between 2005 and 2007 using a variety of measures and only finds a significant positive link between female board presence and ROE, but the study does not use market-based measures such as Tobin's Q.
Studies regarding board gender diversity's effect on the financial performance of French companies are more numerous in comparison to those concerning Romanian companies. Belghiti-Mahut and Lafont [22] analyse a sample of 110 companies listed on Euronext between 2003 and 2008, and the presence of women on the board is positively and significantly linked to financial performance as measured by Tobin's Q. The study does not mention adjusting the results for endogeneity. These results are in line with Ferrary [23] analysis of CAC40 (Cotation Assistée en Continue) listed companies between 2002 and 2006; however, this is a more simplistic study that does not use regression and has a smaller data sample. Studying French firms belonging to the SBF120 (Société des Bourses Françaises) stock market index over the period 2009-2011, Boubaker, et al. [24] found that board gender diversity has a negative effect on firm performance as measured by Tobin's Q. It is worth mentioning that quotas regarding boards' gender diversity were introduced in France in January 2011; thus, the study's time frame might be too short to take into account the effect of the increased presence of women on boards. Sabatier (2015) analysed French CAC40 listed companies between 2008 and 2012 and found that the increased presence of women on boards has a significant and positive effect on economic performance after accounting for the endogeneity of diversity. Bennouri, et al. [25] analysed a sample of 394 French firms between 2001 and 2010. During that period, no regulations requiring companies to meet a certain quota of female directors existed. The results of their study suggest that firms' financial performance is affected by gender diversity even after the results have been controlled for endogeneity, but that this varies depending on the measures of financial performance used. More precise, accounting-based performance measures (ROA and ROE) increase with female directorship, while the market-based performance measure (Tobin's Q) decreases with female directorship. In the most recent study, Ben Slama, et al. [26] take into account the introduction of the 40% gender quota in France, and after analysing data for 89 CAC All-Tradable listed French companies between 2008 and 2011, they find that firm performance increases with gender diversity in high-performing firms, while poorly performing firms are likely to see their financial performance decrease due to the costs of finding women to be on their boards of directors. Overall, the study concludes that complying with the 40% quota of women is likely to lead to more effective boards and, as a consequence, better firm performance. Ionascu, et al. [7] provide the only study concerning Romanian companies in line with academic standards of research. By analysing all companies listed on the Bucharest Stock Exchange from 2012 until 2016, they find that there is a significant association between gender diversity on the board and Tobin's Q only for some of the firms listed on the Bucharest Stock Exchange. Their study uses both an accounting-based financial performance measure (ROA) and market-based measures (market to book value of equity and Tobin's Q) and controls for endogeneity.
Within this framework we consider that our work brings novelty to the literature by analysing Romanian companies in direct comparison with French ones. In designing our study, we focused on ROA as a profitability indicator and Tobin's Q as a market measurement.
Our main research hypothesis is that there is a positive correlation in both countries, France and Romania, between the gender diversity of the boards of directors and the financial performance of the listed companies on the stock exchange.

Research Theory Framework
In order to explain how the composition of the boards of directors influences the performance of a company, several theoretical frameworks have been employed to explain the link between gender diversity on corporate boards and a firm's financial performance [27]. Although most existing research does not make use of any theoretical framework but is instead descriptive [28], we can identify several useful theories, such as agency theory, resource dependency theory, upper echelons theory, human capital theory, and the social psychological theory. The first two main theoretical perspectives can be identified and have been used to help analyse the link between gender diversity of boards of directors and firm financial performance, namely the agency theory and the resource dependency theory [29]. The agency theory helps us understand the monitoring role of boards of directors, while the resource dependency theory offers a theoretical perspective regarding the resource, service, and strategy roles of the board [30]. In addition, this paper explores the upper echelons theory, human capital theory, and the social psychological theory.

Agency Theory
The agency theory is the most used framework in studies that analyse the link between characteristics of boards of directors and firm financial performance [10,28] and was initially developed by [31]. According to them, the agency relationship is a contractual relationship between the principal or owner on one side and the agent or manager on the other side. The principal or principals decide to delegate some decision-making authority to the agent, which is a person they engage to perform some service on their behalf. Given that each party will most likely try to maximize their own benefit, the agent might fail to act in the best interest of the principal(s). The diverging interest between the two parties implies that the agent will act in a way that will result in the value of the firm being lower than if he/she were to act on his/her own behalf as the firm's owner and not just being employed by the firm's owner to manage the firm's activity. The tendency of the agent to act on his/her own behalf to the detriment of the principal's interest can be minimized either by establishing incentives for the agent or by monitoring the agent's activity in order to limit any harmful activities for the principal's welfare [31].
A very good example of an incentive for the agent is an employee stock option plan, which can motivate the employee to stay with the company and work harder to maximize its value. An employee stock option plan works very simply: the employee receives regular call options that will give him/her the right to buy the company's stocks at a specific price for a certain period. Therefore, the employee (the agent) has an interest in the company value growing because this means that the company's shares will most likely rise above the exercise price, and the wider the difference is between the two, the larger his/her capital gain may be when he/she exercises the options. Other incentives for the agent can be performance bonuses, profit participation in a fixed or variable percentage directly, stocks granted by the company, etc. In these ways, the agent will not only be motivated financially, but he/she will also feel that a part of the company belongs to him/her; therefore, he/she will better understand the principal's interests.
In the context of the agency theory framework, the board of directors is responsible for monitoring the agent's activity. The board exercises control over the agent by replacing him/her if he/she fails to act in the interest of shareholders and to maximize shareholders' wealth. In addition, as we just explained, the board can also ensure that the agent is rightfully incentivized so as to act in the interest of shareholders by establishing the appropriate compensation amount that the agent will receive [10]. Therefore, the board of directors' monitoring ability is paramount in ensuring that the maximization of shareholders' wealth is achieved by the agent. While the agency theory does not provide an unequivocal prediction of the way board diversity is linked with firm financial performance [10], it does offer certain insights about the way the two might interact and brings forth the question of whether certain boards or certain board members would have a higher propensity to be better monitors and, thus, to ensure that the agent acts on behalf of the principal's interest better than others. In this context, the monitoring and controlling role of a board of directors can be better realized in a more gender diverse situation, because women tend to be better monitors than men, and gender diversity increases the independence of these boards [10]. However, ownership position in the firm may have a more powerful influence on monitoring than independence [20]. Therefore, we believe that a combination of all these factors-gender diversity, independence, and equity ownership-should be taken into account when selecting an optimal board of directors. This type of board will be more balanced and impartial, offering a broader perspective while providing better controlling and monitoring of managers, ensuring that, in the end, the final goal of the maximization of shareholders' wealth is achieved by the agent, in this way solving, at least partially, the divergence from the agency theory.

Resource Dependency Theory
According to the resource dependency theory, firms need to collaborate with other actors in their environment in order to be able to acquire or exchange resources that are needed for their survival, which creates a relationship of dependency between the firm and other external actors in their environment. In order to achieve corporate success, firms seek to establish relationships through which they can obtain the most beneficial resources and choose to appoint people on their corporate boards that can best connect them with the needed resources [28]. From the perspective of the resource dependency theory, directors, especially outside directors, act as a liaison between the firm and the external environment and ease the access to resources or enable them to be obtained under more favourable conditions that would not otherwise have been available. By obtaining these resources, the functioning of the organization, the firm performance, and the survival of the company are improved [30]. According to Pfeffer and Salacik [32], the board of director's role is to link the firm to outside actors in the firm's environment and thus procure or enable access to resources the firm needs for its success. This can be done by (1) providing the firm with expertise or information the directors might themselves have, (2) opening channels of communication with important outsiders, (3) obtaining support from important actors outside the company, or (4) legitimizing the firm in the eyes of outsiders [20].Women on boards of directors can lead to diversifying the sources of information and, ultimately, enlarging the pool of customers, suppliers, etc., for the company. A more gender diverse board of directors widens the resource base and the linkage with different stakeholders.

Upper Echelons Theory
The upper echelons theory is one of the theoretical frameworks that has been used in previous studies to explain the link between the presence of women on the board and firm financial performance [33]. According to Hambrick and Mason [34], the strategic decisions made by corporate directors are influenced by their personal characteristics. These can be psychological (cognitive base, values) or observable such as age, education, previous professional experiences, socioeconomics roots, financial position, and group characteristics. Differences in preferences and behaviour between men and women that have already been well documented support the point of view of the upper echelons theory. In fact, it has been found that almost all gender differences are preserved even among females and male directors and that female directors will behave differently in the boardroom than their male counterparts. For example, female directors have been found to care more about others and less about power, which suggests that boards of directors with more female members would be more prone to consider the interests of all stakeholders more seriously and would tend to make strategic decisions that are aligned with this. Evidence also suggests that they could be more likely to make ethical decisions than their male counterparts [35]. In addition, women are less prone to corporate risk taking [36], which means that they will make strategic decisions accordingly.

Human Capital Theory
The human capital theory emphasises the importance of someone's stock of education, experiences, and abilities. The contributed capital to a company comes not only in a financial or a material form but also in the form of human capital. In fact, currently, this form of human capital can often be much more important than financial or material resources. Therefore, someone's stock of education, experiences, abilities, etc., should definitely be regarded as a form of capital and can be used to the benefit of a company [20]. Yet, the value of this human capital is not influenced just by education, experiences, abilities, skills etc., but also by gender, meaning that each director has a specific human capital [28].With regard to gender, this specificity can be translated by the fact that women are often more highly educated than men, and they also have different experiences and backgrounds, which cause them to bring unique value to companies by their presence on the board of directors.

The Social Psychological Theory
According to this theory, greater gender diversity on boards of directors will generate more diverse ideas, associations, opinions, and critical thinking that will translate to more innovative results and enhanced performance. However, in certain instances, prolonged boardroom stalemate can be a strain on time resources [3,37]. In the end, it only matters that the benefits of these inputs of divergent thinking can offset such apparent drawbacks. Carter, D'Souza, Simkins and Simpson [20] suggest that the theory and evidence of group dynamics may have both positive and negative effects. The bottom line is that a company should assemble a portfolio of directors that maximizes the firm's resources and value [6]. As in the case of any investment portfolio, we strongly believe that this portfolio of directors can benefit very well from variety and diversification.

Materials and Method
Panel data regression has been deployed for each dataset separately using Stata. In line with most of the literature, we use panel data regression as it is better in terms of controlling for unobservable heterogeneity [3].

Regression Model
The following regression model has been used to explore the link between female presence on boards of directors and firm financial performance: The dependent variable Performance stands for two measures of firm financial performance: ROA and Tobin's Q. The independent variable Diversity stands for two measures of female board representation, Female Presence and Female Proportion.
Separate regressions are used for each measure of firm financial performance. In addition, separate regressions were also used for each measure of female board representation. In the formula of the model, i represents each firm, and t represents each year.

Financial Performance Measures
The variables used to measure firm performance vary from one study to the other. Most studies, however, use a combination of accounting-based measures such as ROA or ROE and market-based measures such as Tobin's Q. Market-based measures take into account the company's market valuation, focusing on the company's outputs and performance, while accounting-based measures take into account the resources of the company, hence focusing on the inputs [38]. Accounting-based measures tend to offer a more distorted view of the firm's financial performance compared to financial-based measures since they are influenced by accounting practices and tax laws [3,39]. In addition, they represent a more objective backwards-looking way of evaluating a firm's financial performance because they are based on accounting records and take into consideration past financial results. Financial-based measures, on the other hand, are forward-looking and are influenced by the investors' expectations of the firm's future financial performance [3,40]. Tobin's Q is equal to the ratio between market capitalization and total assets. Competitive companies tend to have a Tobin's Q that is close to 1, while companies that are expected by investors to be able to generate additional value by employing available resources in an efficient manner have a Tobin's Q ratio that is greater than one. Consequently, firms with a ratio lower than 1 employ their available resources inefficiently [3,38]. Simply put, Tobin's Q measures the wealth of the company [20]. Another market ratio that we could use, instead of Tobin's Q, is the market to book value ratio (MBV). However, if we were to use MBV, this ratio would pair very well with ROE. Both MBV and ROE are related to equity and are therefore more relevant for shareholders or investors, while the ROA and Tobin's Q ratio better assesses the overall financial performance of the company. As a result, we decided to use ROA (also for additional reasons that we will explain the next paragraph), and we preferred the Tobin's Q ratio instead of the MBV. Return on assets (ROA) was chosen rather than return on equity (ROE) because of the financing of activities in the Romanian market. As this study could be expanded to more companies that are not listed, we have considered that it is easier to select ROA because the majority of companies use debt to finance their activities, so ROE could not show a clear indicator of performance. ROA shows how management is using the assets of the company to generate income, while ROE looks at the capacity of the investment of shareholders to generate income. Moreover, as Tobin's Q is also linked to assets, it made ROA the better choice as a variable. It is also important to mention that the financial performance ratios, especially (but not exclusively) the accounting-based ones, are static, showing the financial situation of the company at a specific moment in time, according to the data from the income statement, balance sheet, or cash-flow statement. Therefore, to be truly useful, they need to be analysed dynamically, over a period of time, in order to reveal accurate information, and this is why we chose this period of three years. On the other hand, horizontal analysis can also be useful, and for this, we need to compare the accounting-and market-based ratios of the analysed company with similar ratios of companies from the same area of activity. (See Table 1)

Performance variables
Return on Assets (ROA) Return on assets measured by the ratio between net income and total assets.
Tobin's Q Tobin's Q measured by the ration between market capitalization over total assets.

Female Presence
A dummy variable that is equal to 1 when at least one woman was present on board and that is equal to 0 otherwise (for Romanian firms). A dummy variable that is equal to 1 when the 40% quota was met and that is equal to 0 otherwise (for French firms).

Female proportion
The proportion of women on board measured by the number of women on board divided by the total number of board members.

Control variables
Board size Natural logarithm of the number of board members. Firm size Natural logarithm of total assets. Leverage Measured by the ratio between total debt divided by total assets.

Results
The study analyses French and Romanian companies between 2017 and 2019 separately and then tries to draw a comparison between the situations that exist in the two countries. There are several reasons why we chose to make a comparison between listed companies from France and Romania. Firstly, the authors have a thorough knowledge of the economies of these two countries, working with universities from both countries, namely Babes-Bolyai University in Romania and Ecole de Management de Normandie in France. Secondly, while in Western countries, there are numerous studies regarding the effect of gender diversity on the financial performance of companies, this is not the case for Romania. Thirdly, the reason we chose France is because the 40% gender quota on boards of directors is the first of its kind and a leading standard for other countries. Last but not least, though these two countries are situated far apart, at extremities of Europe, there is a long tradition of cooperation between them, dating from the 19th century when France helped with and influenced most of the modernization of the new Romanian state.
The population of the study is comprised of CAC 40 and CAC Next 20 French companies, on the one hand, and all companies listed on the Bucharest Stock Exchange regulated market, on the other hand. CAC 40 (Cotation Assistée en Continue 40) is the main index of the French stock market that includes 40 of the most important companies that are listed on the Euronext Paris, with a total market capitalization of about EUR 2 trillion (among these, we can find very well-known French companies such as BNP Paribas, Credite Agricole, Airbus, Alstom, Carrefour, Renault, Peugeot, LVMH, etc.). The CAC Next 20 is another stock market index of Euronext Paris, which includes 20 companies that have a market capitalization immediately after the 40 companies that compose the CAC 40 (some examples are Accor, EDF, Sodexo, Valeo, etc.). As the name suggests, these companies have the potential to replace, if necessary, companies from the composition of the CAC 40. In the case of Romania, we have two main indexes: BET (Bucharest Exchange Trading) and BET Plus. BET includes 16 of the most important companies listed on the Bucharest Stock Exchange, while BET Plus has 41 companies (but only 25 are different than those that compose the main BET index). Therefore, in the case of Romania, we have opted for including all companies because of the limited number of companies available in the local stock market and in the indexes of the Bucharest Stock Exchange. As we will explain later, this decision will allow us to have a similar size of our sample and will ease the comparative study. If we look at the company's profile in the two countries, they are relatively similar up to a certain point. The financial sector is very well represented in both groups (AXA, BNP Paribas, Crédit Agricole, Société Générale, and EdenRed in France versus Banca Transilvania, BRD, Patria Bank, Fondul Proprietatea, BVB, and SSIF BRK Financial Group in Romania). However, the industrial sector is more heavily weighted in the case of France (Alstom, Airbus, Bouygues, Michelin, Saint-Goben, Schneider, Veolia, Vinci, etc.), while in Romania, the energy sector is the best represented (Petrom, Rompetrol, Electrica, Romgaz, Nuclearelectrica, Transelectrica, Transgaz, etc.). The data that we used related to these companies were collected from different sources: Factiva, Diane database, Bucharest Stock Exchange, Euronext, and the companies' reports. The period of time chosen was only three years because 2017 is the year in which all French publicly listed companies that met certain size criteria were required, by the Copé-Zimmerman law, to meet the 40% quota of women on their boards. This legislation was mainly applicable to CAC 40 companies. Therefore, studying the French companies that have a considerable number of women on their boards offers the possibility of analysing the effect of women directors in a more accurate way than in those cases when fewer women would occupy directorship functions. In addition, it also offers the opportunity to analyse in what way the hard quota instituted by the French government influenced the financial performance of French companies or if it influenced it in any way at all. The study also looks into CAC Next 20 companies. In order to enable a better comparison between the situation in the two countries, all companies listed on the Bucharest Stock Exchange were analysed separately.
Because of the limitations related to the presence in boards imposed by legislation and the choice of the variable ROA, which is not necessarily a proper performance indicator in the financial sector, all financial institution firms were removed from the dataset. This exclusion is in line with previous studies. This resulted in a final sample of 55 firms out of 60 and 165 firm-year observations for France. For Romania, after removing all financial institution firms, the population was reduced to 69 firms. In addition, firms that were going through an insolvency procedure were also removed from the sample, as in their case, the board of directors was replaced by one or two special administrators appointed by a Syndic judge. In the end, the population was reduced to a final sample of 54 firms out of 83 and 162 firm-year observations for Romania. Overall, the two samples are almost identical in size with 55 French firms and 54 Romanian firms.

Descriptive Statistics
Descriptive statistics for French firms can be found in Table 2. The minimum number of women present on French boards is three, while the maximum is nine. French firms have an average number of 5.64 women on their boards as shown by the mean in the table below. The minimum percentage of women on board is 23.53%, while the maximum is 63.64%, with a mean of 43.01%. The minimum value is recorded by a CAC Next 20 company that does not have to comply with the 40% quota. Overall, almost all CAC 40 companies have complied with the 40% quota. Airbus is the CAC 40 company with the smallest number of women on board; however, the company is not registered in France and therefore does not come under the jurisdiction of the Copé-Zimmerman Law. The size of the board of directors ranges between at least 8 members and at most 20 members with a mean of 13.25 members. The return on assets varies from a minimum of −14.8 to a maximum of 29.93% and a mean of 4.21%. The minimum value for Tobin's Q is 0.05, while the maximum value is 7.12, with a mean of 1.10. The minimum value for leverage is 0.28, while the maximum value is 1.20, with a mean of 0.63. Total assets range from a minimum of EUR 1,231,410 to a maximum of EUR 930,695,000, with a mean of EUR 60,398,132. Descriptive statistics for Romanian firms can be found in Table 3. The minimum number of women present on Romanian boards is 0, while the maximum is 4. Romanian firms have an average number of 0.94, a slight increase from the 0.77 mean found by [7]). The minimum percentage of women on board is 0%, while the maximum is 100%, with a mean of 19.39%, which has greatly increased compared to the mean of 0.16 found by Ionascu, Ionascu, Sacarin and Minu [7]. Overall, several Romanian listed companies have no women on board, while some have one or two women on board or at most three women on board, which, in the case of one company, represents 100% of all board members. The size of the boards of directors ranges between at least 3 members and at most 11 members with a mean of 5.09 members. The return on assets varies from a minimum of −0.40 to a maximum of 8.86% and a mean of 0.17%. The minimum value for Tobin's Q is 0.04, while the maximum value is 8.86, with a mean of 0.17. The minimum value for leverage is 0, while the maximum value is 2.03, with a mean of 0.38. Total assets range from a minimum of EUR 1,322,042 (RON 6,266,480) to a maximum of EUR 2,303,500,000 (RON 10,918,590,000), with a mean of EUR 237,318,677 (RON 1,124,890,530).
French and Romanian firms differ in several ways. First of all, French companies have bigger boards. The mean board size of French companies is 13.25 while that for Romanian companies is 5.09. This difference can be explained by the differences that exist between the economic developments of the two countries. The mean for the percentage of women on the boards of French companies is 43.01% while that for Romanian companies is 19.39%. This is due mainly to the difference in regulations concerning female board presence. France has imposed a hard quota, while Romania only offers a recommendation. French companies also have better financial performance measurements. The mean return on assets of French firms is 4.21% while the that for Romanian companies is 0.17%. As far as the mean value for Tobin's Q is concerned, French companies have a mean value of 1.10, while Romanian companies have a mean value for Tobin's Q of 0.47.
Mean comparisons have been conducted for both countries using an unequal variances t-test [41] and can be found in the tables below. After performing the t-test, differences between the means can be observed regarding the board size and the return on assets ratio. Companies that have met the 40% quota tend to have smaller boards with a mean of 12.80 board members, compared to those that did not meet the 40% quota that have a mean of 13.87 board members. In addition, those companies that have complied with the quota tend to have bigger returns on their assets, the mean being 5.40, compared to 2.55 for those companies that have not complied with the legal requirement. There also exists a slight difference in the mean values for Tobin's Q; however, the hypothesis that the two means are not the same could not be invalidated. Companies with 40% or more women on their boards tend to be about the same size according to mean total assets compared to those with less than 40% female directors. (See Tables 4 and 5)  Contrary to Ionascu, Ionascu, Sacarin and Minu [7], Romanian companies with women on board are bigger in terms of assets (RON 1.2 billion compared to RON 900 million) than those that have no female directors. After performing the t-test, no great differences between the two groups of firms can be observed with the exception of the return on assets ratio. Companies that also have female directors on their board have a slightly higher mean for the return on assets ratio at 0.25, compared to 0.03 for those who feature no women on their board.

Regression and Interpretation
The analysis was first performed for the French companies and then for the Romanian companies. In the case of the French companies, pooled OLS regressions were run without including the control variables as a first step. The pooled OLS regressions were run for each of the two dependent variables separately and also for each of the two independent variables used to measure the diversity of the boards of directors: female percentage and the dummy that marks which companies have met the 40% quota. Afterwards, pooled OLS regressions including the control variables were run. Separate regressions with control variables were run for each of the two dependent variables, ROA and Tobin's Q, alternatively, and also for each independent measure of the diversity of the boards of directors. Fixed and random effects regressions were included for each of the regressions. Haussman tests were conducted for each of the pooled OLS regressions with control variables, and their results were used to help choose between the fixed and the random effects models. In cases where the Haussman test rejected, the fixed effects model was chosen. In cases where the results for the Haussman test did not reject, the random effects model was consulted. However, given the fact that fixed effects allow for arbitrary correlation and random effects do not, fixed effects are usually considered to be more convincing than random effects [42]. Therefore, the fixed effects regression was included for further analysis in all cases. Breusch-Pagan tests were also run, their results indicating that the pooled OLS model should be taken into account. In the case of French companies, a simple pooled OLS regression showed significant results only for ROA and not for Tobin's Q. Both measures of gender diversity, female percentage and quota compliance dummy, appear to have a positive effect on the accounting-based measure of firm financial performance ROA. On the other hand, in the case of Tobin's Q, both measures of gender diversity appear to have a positive effect on firm financial performance; however, these results are not significant. Simple pooled OLS cannot be considered sufficient to estimate the relationship between gender diversity on board and firm performance; therefore, control variables were included in new pooled OLS regressions, and a fixed effects model was also analysed.
The results for Tobin's Q remain insignificant in all cases both for the pooled OLS regressions with controls and for the fixed effects models. The relationship between the variables appears to be positive in the case of the pooled OLS regression but becomes negative in the case of the fixed effects model. In the case of ROA, the results remain significant and positive for the pooled OLS regressions with controls. However, when the correlation between both measures of gender diversity and financial performance as estimated through ROA is measured using the fixed effects model, the results show a negative, but not significant, association. The random effects model showed results that were not significant in all of the cases analysed. Therefore, we cannot conclude that gender diversity influences firm financial performance in any way, since the results of our study are not significant and therefore inconclusive. At most, we could conclude that gender diversity has a positive effect on ROA, the accounting-based objective measure of financial performance; however, further analysis in the matter would be needed to fully validate the hypothesis. The results regarding Tobin's Q are contrary to those of Boubaker, Dang and Nguyen [24]. They find that gender diversity has a negative and significant effect on Tobin's Q; however, they use 2SLS regression and control for endogeneity. In regard to the other variables, negative significant results are shown for firm size as measured by total assets and for leverage as measured by the ratio between total debt divided by total assets in relation to both Tobin's Q and ROA. This shows that bigger firms have a weaker firm financial performance compared to smaller firms. In addition, it also shows that firms with more debt also tend to have worse financial performance when compared to their less indebted peers. Overall, these findings are in line with those of multiple studies, including Dale-Olsen, Schøne and Verner [13]; Rose [14]; and Miller and Del Carmen Triana [15], which found no link between boards' gender diversity and companies' financial performance.
One of the arguments used to explain the lack of any observed link between gender diversity of boards and firm financial performance [7] is the fact that a critical mass of at least three female directors [43] has not yet been reached. However, in the case of French companies, this argument cannot be taken into consideration, since all companies in our sample have at least three women on board, and as such, they have all reached the critical mass. Another argument for our findings might be the fact that men and women directors might not differ so much in their behaviour as they are believed to. What is more, not enough time may have passed in order for the higher gender diversity to have any observable effects. After all, only 3 years have passed since most of the French companies in the sample met the 40% quota, and the women who joined their boards may not have yet had the chance to contribute sufficiently to the activity of the board. Last but not least, econometric issues might also be the reason why no link between gender diversity and firm financial performance can be supported. The pooled OLS regressions showed significant and positive results. On the other hand, the results for the fixed effects models were not significant, which might suggest that other unobservable firm characteristics that have not been included as control variables might be the reason for the significant and positive effect found through the pooled OLS regression.

Discussion
In the case of Romanian companies, the same steps as those outlined for the analysis of French companies were followed. Simple pooled OLS regression showed mostly negative and insignificant results in all cases both for ROA and for Tobin's Q. Only female presence has been found to have a positive but insignificant effect on ROA. As in the case of French companies, simple pooled OLS cannot be considered sufficient to estimate the relationship between gender diversity on the board and a firm's performance; therefore, control variables were included in new pooled OLS regressions, and a fixed effects model was also analysed. Pooled OLS regressions with control variables also showed results that are not significant in all cases, except one, and negative. Female presence is shown to have a significant and slightly negative result on firm financial performance as measured by Tobin's Q. The fixed effects models showed results that are not significant except for the case of the effect of female presence on ROA. The results for the fixed effects model are positive but not significant in the case of the effect on ROA of female percentage. They are negative and not significant in the case of female percentage and female presence and Tobin's Q. As previously mentioned, female presence seems to have a positive and significant effect on ROA. The random effects model showed results that are not significant in all of the cases analysed. Therefore, we cannot conclude that gender diversity influences firm financial performance in any way, since the results of our study are not significant and, therefore, inconclusive. At most, we could conclude that female presence has a positive effect on ROA, the accounting-based objective measure of financial performance; however, further analysis in regard to endogeneity would be needed to fully validate this positive effect. (See Tables 6 and 7)  These findings are in line with those of Ionascu, Ionascu, Sacarin and Minu [7] who do not find any significant link between gender diversity and firm financial performance as measured by ROA or Tobin's Q. In regard to the other variables, negative significant results are shown for firm size and for leverage in relationship with both Tobin's Q and ROA, similar to the case of French companies. This shows that bigger firms and firms that have more debt tend to have a weaker firm financial performance compared to smaller firms and those with less debt. Overall, as in the case of French companies, these findings are in line with those of multiple studies, including Dale-Olsen, Schøne and Verner [13]; Rose [14]; and Miller and Del Carmen Triana [15], which found no link between boards' gender diversity and companies' financial performance.  As previously stated, one of the arguments used to explain the lack of any observed link between gender diversity of boards and firm financial performance [7] is the fact that a critical mass of at least three female directors [43] has not yet been reached. In the case of Romanian companies, this might be the case, as most of them have few to no women on their boards. Another argument for our findings, similar to the case of French firms, might be the fact that men and women directors might not differ so much in their behaviour as they are believed to. Last but not least, econometric issues might also be the reason why no link between gender diversity and firm financial performance can be supported. As in the case of French firms, other unobservable firm characteristics that have not been included as control variables might be the reason for the significant and negative effect found through the pooled OLS regression for the effect of female presence on Tobin's Q and the negative but not significant result of the fixed model. (See Tables 8 and 9)

Conclusions
This study analyses the relationship between gender diversity and firm financial performance. Its aim is to try to contribute to the business case argument for gender diversity on boards of directors by analysing the existing situation in two European Union member countries: France and Romania. The two countries offer an interesting ground for comparison as they are examples of opposite ends of the spectrum in terms of female board representation as well as in terms of legislative approach in the matter. The case of France is of particular interest as the country is the only one to have met a 40% quota at a European Union level. The main objective of this study was to examine how the gender diversity of boards of directors influences firm financial performance in each of the two countries and to compare the existing situations in the two countries. The main hypothesis was that there is a positive association in both countries between the gender diversity of the boards of directors and the financial performance of the listed companies. Panel data analysis is conducted on a sample of 54 Romanian listed companies and 55 CAC 40 and CAC Next 20 listed French companies between the years of 2017 and 2019. The study uses both an accounting-based measure of financial performance (ROA) and a financial-based measure of financial performance (Tobin's Q). In addition, it uses two measures of female representation (female proportion and female presence for Romania, and female presence at least equal to the imposed quota for France).
The analysis conducted fails to provide any evidence of a link between boards' gender diversity and companies' financial performance. Results are not significant, regardless of the measures used. These findings are in line with those of multiple studies including Dale-Olsen, Schøne, and Verner (2013); Rose (2007); and Miller and Del Carmen Triana (2009). While the analysis has failed to find a positive link between female presence and firm financial performance, it has not outlined a negative one. This could still be considered as a contribution to the business case argument for female representation on the board of directors, as female presence does not have a negative impact on the bottom-line goal of profit making for the companies which they are a part of. The fact that no evidence regarding a negative effect of female presence on the boards of directors has been found can be used to at least defend, if not support, the case for the introduction of a European directive and as an argument for other European Union countries to willingly support the increased presence of women on their companies' boards of directors.
Several limitations of the study can be considered. Firstly, the study does not address endogeneity concerns. In addition, it uses a limited data sample that only covers the previous three years. A three-year period is a relatively short period of time, which could imply that not enough time has passed before any significant effects of female presence on boards of directors could be observed. In addition, the study does not use lagged data given that the period of time between 2017, the deadline of the 40% quota for CAC 40 companies, and now is too short to make the use of lagged data possible. Lagged data would allow for a better observation of the impact of female presence on firm financial performance as it would take into account the fact that some time has to pass before any significant effect can be observed. In the case of Romanian companies, limitations might include the small number of women present on boards or lack of their presence altogether, which implies that a minimum threshold might not have been met that would make any potential effects observable. The study also included only a limited number of control variables, which means that the relationship between board diversity and firm performance was not analysed in accordance with all factors that might also influence it, and therefore results might not be as accurate as possible.
In terms of future research proposals, further analysis could be conducted to compare the situation of French companies in the years previous to the deadline of the 40% quota and the years after companies met the required quota. In addition, analysis regarding the effect of the 40% quota introduced through the Copé-Zimmerman law could be repeated at a later time after more years have passed since all CAC 40 companies complied with the quota. This would allow for a better observation of the impact of gender diversity on firm financial performance. In the case of Romania, a future investigation could be conducted to compare the former state-owned companies with the new entrepreneurial companies that were listed on the Bucharest Stock Exchange, in order to see if there is an improvement of the gender diversity on the boards and its impact on financial performance. Moreover, it would be interesting to investigate the relation between the size of the boards, which can vary significantly, and the financial performance of the companies. In the case of both Romanian and French companies, further research could focus on analysing whether the presence of women as chairs of the board has a significant effect on firm financial performance, for example. In addition, further research could also include more control variables, so as to better estimate the relationship between board gender diversity and firm financial performance and to better account for differences in the results between the pooled OLS regression and the fixed model. In the case of French companies, as suggested by [22,24], further research could focus not only on publicly listed companies but also on smaller ones, which represent the majority of French companies. The same research can also be done in the case of Romania, or at least for the smaller companies that are listed on the AeRo, which is the alternative market of the Bucharest Stock Exchange, with more than 270 listed firms, and which was designed especially for SMEs and start-up companies.