The effect of the sex diversity of the board of commissioners on firm performance and the role of the ethnic background of president commissioners

Abstract

This study aims to obtain empirical evidence on the effect of the sex diversity of the board of commissioners (SDBOC) on firm performance (FP). In addition, this study examines the role of the ethnic background of the president commissioner (ETC) as a moderating variable that strengthens the influence of the sex diversity of the board of commissioners (SDBOC) on firm performance (FP). The sample consists of manufacturing companies, registered as public companies in Indonesia between 2017 and 2021, resulting in 783 observations. The data analysis includes several multivariate regression techniques consisting of ordinary least squares (OLS), fixed effects (FE), and the Heckman selection model for data robustness and hypothesis testing. The results showed that the sex diversity of the board of commissioners increased firm performance, as measured by Tobin’s Q (TQ), but it had no significant effect on the market to book value (MBV). Furthermore, the ethnic background of the president commissioner strengthened the influence of the sex diversity of the board of commissioners on firm performance, as measured by TQ and MBV. The results support the theories used, namely agency theory, resource dependence theory, and upper echelon theory. Shareholders can use the result of this study to highlight certain aspects during the appointment of the board of commissioners, in terms of their gender diversity and the ethnic background of the president commissioner. Meanwhile, the required number of women on the board of commissioners, to improve board decisions, equates to two. This is supported by the critical mass theory.

Keywords:

• Sex diversity of board of commissioners
• firm performance
• ethnic background of president commissioner
• Tobin’s Q
• market to book value

REVIEWING EDITOR:

• Nor Shaipah Abdul Wahab
• Taylor’s University – Lakeside Campus
• Malaysia

SUBJECTS:

• Business, Management and Accounting; Economics; Industry & Industrial Studies

1. Introduction

Gender diversity is a popular issue today, and this issue is in accordance with the Sustainable Development Goals (SDGs) of the United Nations, issued in 2015, which are aimed at gender equality. Gender equality focuses on empowering women in all fields of work. The issue of gender diversity in work, according to SDGs, outlines that men and women have equal rights to be involved in the leadership of companies. One of the leadership positions in Indonesian companies is a seat on the board of commissioners, which adheres to the two-tier system. This coincides with the agency theory, in which the conflict of interests between shareholders and management can be minimized by having a board of commissioners to act as a supervisory party (Akram et al., 2020; Assenga et al., 2018; Jensen & Meckling, 1976; Pucheta-Martínez & Gallego-Álvarez, 2020; Roudaki, 2018). The oversight structure will be stronger if there is sex diversity on the board. Sex diversity is the diversity of sexs on the board of commissioners (sex diversity of the board of commissioners/SDBOC) (Ahmadi et al., 2018; Bae & Skaggs, 2019; Osahon & Hassan, 2021; Simionescu et al., 2021). The agency theory states that SDBOC can strengthen the supervisory structures by bringing a wide range of perspectives, experiences, and approaches, which in turn can increase the effectiveness of the board and create a more inclusive supervisory environment (Hosny & Elgharbawy, 2022).

Financial Services Authority Regulation Number 33/POJK.04/2014 does not regulate the diversity of the board of commissioners. In fact, women in Indonesia who have become members of a board of commissioners constitute 14.9%, which shows the low proportion of women who serve in public companies (International Finance Corporation, 2019). In other words, sex diversity in public companies in Indonesia is low, even though the board of commissioners has an important position in companies, owing to its supervisory and management duties. Notably, the presence of a diverse board of commissioners can enrich the perspective and experience of the board of commissioners, thus enabling better and more innovative decision-making that increases firm performance (Bin Khidmat et al., 2020; Charles et al., 2018; Duppati et al., 2020; Dwaikat et al., 2021). Therefore, the importance of SDBOC for the progress of companies is evident. According to Nguyen and Huynh (2023), the resource dependence theory (RDT) outlines that companies with high corporate performance must have sufficient resources to develop their businesses. Therefore, a board of commissioners must have the ability to oversee its company’s operations, to develop its business. The supervision process will run well if there is sex diversity (Arvanitis et al., 2022; Ben Slama et al., 2019; Saha & Maji, 2022).

Empirical evidence suggests that the sex diversity of the board of commissioners positively influences firm performance (Abdullah et al., 2022; Arvanitis et al., 2022; Garcia-Blandon et al., 2019; Hosny & Elgharbawy, 2022; Joecks et al., 2023; Nguyen & Huynh, 2023; Nur & Darvin, 2023; Safiullah et al., 2022; Sahoo et al., 2023; Soare et al., 2022). However, other evidence states that the presence of women on the board of commissioners reduces the profitability of companies (Adams & Ferreira, 2009; Bogdan et al., 2023; Ghafoor et al., 2022; Mehrotra et al., 2023; Nigam et al., 2022), while other studies have shown that SDBOC has no effect on FP (Almarayeh, 2021; Alshirah et al., 2022; Arora, 2022; Wang, 2023).

Previous research into the effect of SDBOC on FP has found inconsistent evidence. In this regard, to mitigate the influence of SDBOC on FP, the role of the ethnic background of the president commissioner, as a moderator, may help achieve the alignment of interests between the agents and principals. Indonesia is a very ethnically diverse country. As a matter of fact, Indonesia has approximately 1,340 ethnic groups, according to the 2010 BPS census (Informasi, 2017). Therefore, it is important to examine the ethnicity of the president commissioner, as a moderating variable in the relationship between the sex diversity of the board of commissioners and firm performance in Indonesia, in order to provide a clearer view of how diversity can affect company performance in the country. Furthermore, in line with Frazier et al. (2004), introducing and including moderating variables into research models is important when there are inconclusive research findings. Previous theoretical findings by Mohammad and Wasiuzzaman (2020) found evidence that the ethnicity of the board moderates the influence of family ownership on earnings management. Other findings suggest that the ethnic diversity of the board moderates the relationship of board characteristics with company performance (Karim et al., 2022). Findings by Mohammad and Wasiuzzaman (2020) and Karim et al. (2022) can be used as a reference to show that the ethnic background of president commissioners is a moderating variable in this study. This research is important because the role of the president commissioner is that of the main actor, who determines an effective system of supervision. The president commissioner plays a role in coordinating the entire board in a company, in accordance with the upper echelon theory (UET).

The upper echelon theory states that the business, knowledge, and experience of a leader are the factors influencing the success of a company (Akram et al., 2020; Bin Khidmat et al., 2020; Neffe et al., 2022; Prosvirkina & Wolfs, 2021). A leader serves as the main actor in the company, and the main actor in question has the role of the president commissioner and leads the supervisory team. The ethnic background of the president commissioner allows the company to obtain superior resources, to implement efficient corporate governance in order to improve firm performance (Chuah & Hooy, 2018; Darmadi, 2013; Hogan & Huerta, 2019; Karim et al., 2022). Moreover, it can enrich the perspective and experience of the board, in order to make the best decisions and consider the interests of all the stakeholders (Oldford et al., 2021).

The contribution of this study, first and foremost, entails proposing the ethnic background of the president commissioner as a moderating variable in the research model. So, the novelty of this study is to develop and empirically test the role of the ethnic background of the president commissioner in moderating the influence of SDBOC on FP, using multiple theories, namely the agency theory, the resource dependence theory, and the upper echelon theory. Second, this study uses the measurement of the ethnic background of the president commissioner in Indonesia based on lineage, namely a background of matrilineal traditions (Minangkabau, Dayak, Nias, Sumba, and Toraja) or patrilineal traditions (Javanese, Sundanese, Balinese, Batak, and Madurese) (Badan Pusat Statistik [BPS], 2010). According to Riani et al. (2021), ethnicity in Indonesia can be demonstrated through matrilineal and patrilineal kinship systems. Previous research has also measured ethnicity, including Malay, Chinese, and Indian ethnicities (Karim et al., 2022). Third, this research contributes to the analysis of SDBOC, ETC and FP using multiple theories, namely the agency theory, the resource dependence theory, and the upper echelon theory. The results of this study support the agency theory, where a company must have an effective supervision system as part of the role of the board of commissioners. The resource dependence theory confirms this study, in that SDBOC can affect FP. The upper echelon theory supports this study, by showing that the ethnic background of the president commissioner is important because it is a corporate resource that determines the effective supervision of the company, thus strengthening the influence of SDBOC on FP. Fourth, this study contributes to companies that go public. It helps them pay attention when selecting candidates for the board of commissioners, by considering sex diversity and ethnic diversity. Finally, this study can provide and encourage proposals to start including a minimum number of women on the boards of commissioners of companies going public. The required number of women members of boards of commissioners, who can improve the boards’ decisions, is two members, in accordance with the critical mass theory.

2. Theoretical development and hypotheses

2.1. Theoretical development

2.1.1. Agency theory

The agency theory outlines the conflict of interest between shareholders and managers (Cieślak, 2018; Nguyen & Huynh, 2023; Panda & Leepsa, 2017; Taherinia et al., 2017). However, the conflict of interest can be mitigated through the board of commissioners, which serves as the right hand of the shareholders, whose duty it is to represent the company’s shareholders. In addition, the board of commissioners is tasked with overseeing the company’s activities and providing advice to the board of directors, in order to increase firm performance (Abdullah et al., 2022). In carrying out its supervisory duties, the board of commissioners will work efficiently when it has diverse board memers (Nigam et al., 2022). One of the diversities of the board of commissioners is sex diversity, namely having both male and female members. In this regard, the sex diversity of the board of commissioners can increase firm performance (Galbreath, 2018; Okoyeuzu et al., 2021; Song et al., 2020).

2.1.2. Resource dependence theory

The resource dependence theory (RDT) states that businesses always want to achieve sustainable strategic goals through the resources that they have, in order to create a competitive advantage (Hatane et al., 2022; Khan et al., 2019; Sahoo et al., 2023). Accordingly, RDT suggests that the sex diversity of the board of commissioners will provide various resources to the company, in order to improve the company’s performance (Shehata et al., 2017). In addition, the sex diversity of the board of commissioners outlines that such diversity will result in an efficient surveillance system. Women are seen as more careful when working; thus, they will work more effectively when paired with men. Therefore, based on RDT, the sex diversity of the board of commissioners can increase the effectiveness of its monitoring (Dissanayake et al., 2021; Hosny & Elgharbawy, 2022; Nguyen & Huynh, 2023).

2.1.3. Upper echelon theory

The upper echelon theory (UET) states that leaders in a company are important actors who influence their subordinates (Kolev & McNamara, 2020; Neffe et al., 2022; Prosvirkina & Wolfs, 2021). In fact, the upper echelon theory also describes the art of leadership, where the leader has the ability to control and coordinate the team. While UET assumes that the diversity of the board of commissioners is viewed from the point of view of the members’ gender (Mahsina & Agustia, 2023), the theory also states that the leader’s effort, knowledge, and experience are the success factors of the company (Akram et al., 2020; Bin Khidmat et al., 2020; Neffe et al., 2022; Prosvirkina & Wolfs, 2021). A leader serves as the main actor in the company, and the main actor has the role of the president commissioner and leads the supervisory team. The upper echelon theory in this study also presents the variable of the ethnic background of the president commissioner.

2.1.4. Ethnic leadership styles in Indonesia

Indonesia is a country with a rich ethnic and cultural diversity. There are more than 1,300 ethnic groups spread throughout Indonesia (Badan Pusat Statistik (BPS), 2010). Some of the largest ethnic groups in Indonesia include the Javanese, Sundanese, Batak, Minangkabau, Madurese, Balinese, Bugis, and Dayak. According to Riani et al. (2021), ethnicity in Indonesia can be shown through matrilineal and patrilineal kinship systems. For instance, the Javanese leadership style is a type of leadership that is authoritative and intimate with its group members (Endraswara, 2013). Javanese tribal leaders are known by the term ‘father’ or ‘mother’, which is often used to refer to older and respected people. In addition, in Javanese leadership, values such as politeness, wisdom, and modesty are highly valued. Meanwhile, the leadership style of the Sundanese tribe tends to be more democratic and participatory (Wati & Hum, 2005). Sundanese tribal leaders seek to involve all members of the group in the decision-making process and strive to reach mutual agreement on issues. Values such as hard work, honesty, and brotherhood are also highly valued in Sundanese culture. Next, the Batak leadership style entails strong and authoritative leadership (Sidabutar & Mas’ud, 2016). The leader of the Batak tribe is considered to be a figure who must be able to provide protection and security for the members of the group. Furthermore, in the leadership of the Batak tribe, values such as honesty, discipline, and simplicity are highly valued. On the other hand, the Minangkabau leadership style is known as a matriarchy, or a leadership system dominated by women (Elfira, 2013). In this system, leadership is held by women called ‘grandfathers’. The grandfathers’ leadership style is usually participatory and democratic. A Datuk is also expected to solve problems in the group fairly and wisely. Finally, the Dayak leadership style tends to be more egalitarian and participatory (Nawawi et al., 1986). In the Dayak leadership system, important decisions are taken through consensus, and each member of the group is expected to contribute to the decision-making process. Values such as honesty, courage, and simplicity are also highly valued in Dayak culture.

Leadership styles based on ethnicity have been examined by Passakonjaras and Hartijasti (2020), who found evidence of a complete leadership theory that would be suitable in the context of Indonesia with its various ethnicities. The full range leadership theory states that leaders combine transformational, transactional, and laissez-faire leadership according to the conditions, the employees and the work at hand. Passakonjaras and Hartijasti (2020) also found evidence that Padang’s ethnic leadership uses a combination of transformational and transactional leadership.

2.2. Hypotheses

2.2.1. Sex diversity of the board of commissioners and firm performance

The sex diversity of the board of commissioners has an impact on the effectiveness of the board in carrying out its monitoring activities. The board of commissioners consists of a balanced composition of men and women who, if they carry out their duties well, will ultimately improve their firm’s performance (Hosny & Elgharbawy, 2022). Based on RDT, firm performance increases when the company has competent resources. One of the company’s resources involves sex diversity in the monitoring process. According to Sahoo et al. (2023), the presence of women enhances the interests of a business’s stakeholders, as they are good at solving business problems; thus, monitoring becomes more effective, and the firm’s performance increases.

RDT confirms the empirical findings that the sex diversity of the board of commissioners entails diverse resources, in terms of the members’ opinions and knowledge, to ensure that the monitoring process runs well (Alshirah et al., 2022; Arvanitis et al., 2022). As such, this is important for the continuity of the business. Other findings have also shown that the sex diversity of the board of commissioners ensures that the men and women work in a team that has diverse experience, varied educational backgrounds, and various skills that allow it to carry out its supervisory activities well, in a way that increases its firm’s performance (Zhu et al., 2022). The presence of women on the board of commissioners may increase the monitoring’s efficiency, thus increasing management’s productivity and the firm’s performance (Abdullah et al., 2022). Other empirical evidence also suggests that the sex diversity of the board of commissioners positively affects firm performance (Ghafoor et al., 2022; Mahsina & Agustia, 2023; Nguyen & Huynh, 2023; Safiullah et al., 2022; Wang, 2023). Based on previous research and RDT, the first hypothesis in this study is as follows:

H1: The sex diversity of the board of commissioners influences firm performance

2.2.2. Sex diversity of the board of commissioners, the ethnic background of the president commissioner and firm performance

According to RDT, firm performance is influenced by the sex diversity of the board of commissioners (Hosny & Elgharbawy, 2022; Joecks et al., 2023; Safiullah et al., 2022). However, other studies have shown that sex diversity negatively affects firm performance (Adams & Ferreira, 2009; Ghafoor et al., 2022), or that the sex diversity of the board of commissioners has no effect on firm performance (Alshirah et al., 2022; Bae & Skaggs, 2019). Notably, the inconsistency in these results implies an opportunity to conduct further research.

Inconsistent research results can be mitigated by including the variable of the ethnic background of the president commissioner. Evidently, the ethnic background of the board of commissioners can increase a company’s competitive advantage (Karim et al., 2022). In this vein, the ethnic background of the president commissioner can be showcased through the leadership style adopted in managing his/her subordinates. Ethnicities in Indonesia are diverse, with different characters involved (Riani et al., 2021). Thus, the ethnic background of the president commissioner can be a resource for the company, allowing the supervisory system to be run more effectively.

Studies conducted in Malaysia have divided the ethnicity of the board of commissioners into three main groups: Malay, Chinese, and Indian (Karim et al., 2022). The ethnic diversity of the board of commissioners is very useful for supervising its company effectively and increasing firm performance, owing to the different points of view and knowledge of certain ethnic groups. Similarly, the ethnic background of the president commissioner may also impact the leadership with respect to how the members of the board of commissioners are organized to carry out their duties as a supervisory board. According to UET, the leader is the main actor who determines the success of the team (Mahsina & Agustia, 2023). Thus, the effectiveness of the supervisory board, due to the ethnic background of the president commissioner, will improve firm performance. Other research has also shown that the board’s home country encourages improved performance (Garcia-Blandon et al., 2019). As such, ethnic diversity in Indonesia is an attractive factor for the leadership in companies, and the ethnic background of the president commissioner can have an impact on the relationship between the sex diversity of the board of commissioners and firm performance. Based on previous research and UET, the following hypothesis is developed:

H2: The ethnic background of the president commissioner moderates the influence of the sex diversity of the board of commissioners on firm performance

Correspondingly, based on a theoretical review, empirical evidence, and the hypothesis development above, the research model can be described in Figure 1.

Figure 1. Research model. Note: FP denotes firm performance; SDBOC denotes the sex diversity of the board of commissioners; ETC denotes the ethnic background of the president commissioner; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; NBOD denotes the size of the board of directors; NWBOD denotes the number of women on the board of directors; NBODI denotes the size of the independent board of directors; PRONBOID denotes the proportion of independent directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity.

3. Research design

3.1. Sample selection and data sources

The population in this study comprised companies that were going public in the Indonesian manufacturing sector from 2017 to 2021. Manufacturing companies were chosen as the research sample because manufacturing companies have diverse sectors (for example: food, automotive, chemical and others), therefore the diversity of the manufacturing sectors would allow for a broader and more comprehensive analysis (Soebyakto et al., 2018). The data sample obtained included a total of 207 companies, giving 783 observations using unbalanced panel data. However, the sample identified was limited, due to the unavailability of data about the ethnic background of the president commissioners, which was publicly submitted through the companies’ annual reports. Data for this study were collected archivally from the annual reports, company websites, and the Indonesian capital market data were obtained from www.idx.co.id. Here is the sample acquisition table (Table 1).

Table 1. Operational definitions of variables.

No	Criterion	2017	2018	2019	2020	2021	Total
1.	Manufacturing company listed on the Indonesia Stock Exchange	223	225	235	245	255	1,183
2.	Companies that did not provide the data needed in the study	(66)	(67)	(75)	(99)	(93)	(400)
Total sample		157	158	160	146	162	783

3.2. Variable measurements

The dependent variable in this study was firm performance (FP), which was proxied through a market-based ratio, namely Tobin’s Q (TQ) (Ngo et al., 2019; Saha, 2023; Yıldız et al., 2019) and the market to book value (MBV) (Kao et al., 2019; Liew & Devi, 2021; Zarefar et al., 2022). Meanwhile, the independent variable in this study included the sex diversity of the board of commissioners (SDBOC), using the Blau index measurement (Đặng et al., 2020; Maji & Saha, 2021; Singh et al., 2019). In addition, the sex diversity of the board of commissioners was measured using measurements in accordance with the previous research, as follows: percentage of women on the board of commissioners (PRONWBOC) (Ahmadi et al., 2018; Mastella et al., 2021); variable dummy for the presence of women on the board of commissioners (1) or none (0) (WBOCDUM) (Dwaikat et al., 2021; Leyva-Townsend et al., 2021), and the number of women on the board of commissioners (NWBOC) (Hatane et al., 2022; Mohsni et al., 2021). Finally, the moderating variable in this research was the ethnic background of the president commissioner, using patriarchal and matriarchal-based categories (Riani et al., 2021). The control variables that affect FP were also included in the research model to overcome omitted variable bias (Gujarati, 2003). According to Whited et al. (2022), the control variable in the research model is important for controlling the research model. This was done by selecting previous research variables that were used to test the influence of the independent variable on the dependent variable, by paying attention to the validity and the appropriate theoretical considerations.

The control variables used were divided into two, namely governance mechanisms (Shukeri & D Alfordy, 2022) and firm characteristics (Cid et al., 2021; Khan et al., 2019). Variable governance mechanisms are widely recognized as significant predictors of firm performance (Maji & Saha, 2021). The governance mechanisms consisted of the age of the president commissioner (AGE), the education level of the president commissioner (GRADE), the educational background of the president commissioner (MAJORITY), the size of the board of commissioners (NBOC), the size of the board of directors (NBOD), the size of the independent board of directors (NBODI), the proportion of independent directors (PRONWBOD), the number of women on the board of directors (NWBOD), the size of the audit committee (NAC), and the proportion of woman on the audit committee (PRONWAC).

An older president commissioner should have a deeper knowledge of market changes and business trends, thus improving FP (Osei Bonsu et al., 2023). A president commissioner with a high education level should have a strong ability to make rational decisions, thus improving FP (Osei Bonsu et al., 2023). The educational background of the president commissioner is important for companies, because they have expertise in the business field so that they can carry out supervisory activities properly, increasing FP (Molinero-Díez et al., 2022). A larger board of commissioners could allow for more effective oversight mechanisms, thus increasing FP (Kusumastati et al., 2022). A larger independent board of directors can help companies better manage risk, and thus improve FP (Maji & Saha, 2021). The presence of women on the board of directors brings a variety of perspectives, ideas and experiences, thus increasing FP (Pidani et al., 2020). A larger audit committee could provide better oversight of the financial reporting, thereby increasing FP (Meah et al., 2021). A larger proportion of women on the audit committee could play an important role in ensuring transparency, accountability, and integrity in the decision-making processes, thereby improving firm performance (Omotoye et al., 2021).

On the other hand, firm characteristics consisted of firm age (FIRMAGE), company size (FIRMSIZE), leverage (LEV), and liquidity (CR). FIRMAGE suggests companies that have been in business for longer may be doing well, so FP increases (Sahoo et al., 2023). FIRMSIZE was expected to have a positive influence on company performance, as larger companies enjoy better competitiveness and economies of scale (Maji & Saha, 2021). LEV was used as a monitoring method, as managers take decisions to go into debt, thereby increasing FP (Maji & Saha, 2021). Companies with well-managed liquidity show the company’s operational efficiency, so FP increases (Boshnak et al., 2023).

The following table presents the operational definitions of the dependent, independent, moderating, and control variables (Table 2).

Table 2. Operational definitions of variables.

Variables	Code	Measurement	Reference
Panel A: Dependent variable
Tobin’s Q	TQ	(Market value equity + Total debt)/Total assets	(Hosny & Elgharbawy, 2022; Zarefar & Narsa, 2023)
Market to book value	MBV	Total market capitalization/Total equity	(Yıldız et al., 2019; Zarefar & Narsa, 2023)
Panel B: Independent variables
Sex diversity of the board of commissioners	SDBOC	BI = 1 − ${\sum }_{i=1}^k{\mathit{\text{Pi}}}^2$ Information : BI = Blau index k = Number of members of the board of commissioners in each category i = Category P = Proportion of members of the board of commissioners to each member	(Leyva-Townsend et al., 2021; Limbasiya & Shukla, 2019; Hosny & Elgharbawy, 2022)
Number of women on board of commissioners	NWBOC	Total number of women on the board of commissioners	(Yıldız et al., 2019)
Proportion of women on board of commissioners	PRONWBOC	Total number of women on the board of commissioners divided by the total number of members of the board of commissioners	(Yıldız et al., 2019)
Presence of women on board of commissioners	WBOCDUM	1 If it has one or more women on the board of commissioners; 0 if otherwise	(Yıldız et al., 2019; Zarefar & Narsa, 2023)
Panel C: moderating variable
Ethnic background of president commissioner	ETC	1 = President commissioner with matrilineal tradition backgrounds (Minangkabau, Dayak, Nias, Sumba, and Toraja) 2 = President commissioner with patrilineal tradition backgrounds (Javanese, Sundanese, Balinese, Batak, and Madurese)	(Riani et al., 2021)
Panel D: Control variables (governance mechanisms)
Age of president commissioner	AGE	Age of president commissioner	(Osei Bonsu et al., 2023)
Education level of president commissioner	GRADE	1 = Senior high school 2 = Bachelor’s 3 = Master’s 4 = Doctorate 5 = Professor	(Osei Bonsu et al., 2023)
Educational background of president commissioner	MAJORITY	1 = Economics 2 = Technical 3 = Other	(Molinero-Díez et al., 2022)
Size of the board of commissioners	NBOC	Total number of members of the board of commissioners	(Kusumastati et al., 2022)
Size of the board of directors	NBOD	Total number of members of the board of directors	(Kusumastati et al., 2022)
Size of the independent board of directors	NBODI	Total number of members of the independent board of directors	(Maji & Saha, 2021)
Proportion of independent directors	PRONBOID	Total number of members of the independent board of directors/total number of members of the board of directors	(Maji & Saha, 2021)
Size of women board of directors	NWBOD	Total number of women on the board of directors	(Pidani et al., 2020)
Size of the audit committee	NAC	Total number of members of the audit committee	(Meah et al., 2021; Zarefar & Narsa, 2023)
Proportion of women audit committee	PRONWAC	Total number of women on the audit committees/total number of members of the audit committees	(Omotoye et al., 2021)
Panel E: Control variables (firm characteristics)
Firm age	FIRMAGE	Number of years since the company was founded	(Sahoo et al., 2023; Yıldız et al., 2019; Zarefar & Narsa, 2023)
Company size	FIRMSIZE	Ln total assets	(Maji & Saha, 2021; Sahoo et al., 2023; Yıldız et al., 2019)
Leverage	LEV	Total debt divided by total assets	(Maji & Saha, 2021; Sahoo et al., 2023; Zarefar & Narsa, 2023)
Liquidity	CR	Current assets divided by current liabilities	(Boshnak et al., 2023)

3.2.1. Model specification and empirical methods

Data analysis were carried out using a quantitative approach, and data testing were carried out using the base ordinary least square (OLS) regression method. The following equation presents the equation 1(1) $${\mathrm{\text{FP}}}_{\mathrm{i},\mathrm{t}}={\mathrm{\beta }}_0+{\mathrm{\beta }}_1{\mathrm{\text{SDBOC}}}_{\mathrm{i},\mathrm{t}}+\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{\beta }}_{\mathrm{n}}{\mathrm{\text{CV}}}_{\mathrm{\text{it}}}+{\mathrm{\epsilon }}_{\mathrm{\text{it}}}$$(1) model to test Hypothesis 1: (1) $${\mathrm{\text{FP}}}_{\mathrm{i},\mathrm{t}}={\mathrm{\beta }}_0+{\mathrm{\beta }}_1{\mathrm{\text{SDBOC}}}_{\mathrm{i},\mathrm{t}}+\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{\beta }}_{\mathrm{n}}{\mathrm{\text{CV}}}_{\mathrm{\text{it}}}+{\mathrm{\epsilon }}_{\mathrm{\text{it}}}$$(1)

FP was the dependent variable measured by TQ and MBV, SDBOC (the sex diversity of the board of commissioners) was the independent variable measured using the Blau index, and CV was the control variable (governance mechanisms and firm characteristics).

The second hypothesis aimed to answer the research question of whether the president commissioner’s ethnic background (ETC) could moderate the relationship between SDBOC and FP. The dependent variables in this study consisted of TQ and MBV, while the independent variable was SDBOC. The control variables used in this study were the same as those in model 1. The following is the equation 2(2) $${\mathrm{\text{FP}}}_{\mathrm{i},\mathrm{t}}={\mathrm{\beta }}_0+{\mathrm{\beta }}_1{\mathrm{\text{SDBOC}}}_{\mathrm{i},\mathrm{t}}+{\mathrm{\beta }}_2\text{ETC}+{\mathrm{\beta }}_3\text{SDBOC*ETC}+\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{\beta }}_{\mathrm{n}}{\mathrm{\text{CV}}}_{\mathrm{\text{it}}}+{\mathrm{\epsilon }}_{\mathrm{\text{it}}}$$(2) model used to test the second hypothesis: (2) $${\mathrm{\text{FP}}}_{\mathrm{i},\mathrm{t}}={\mathrm{\beta }}_0+{\mathrm{\beta }}_1{\mathrm{\text{SDBOC}}}_{\mathrm{i},\mathrm{t}}+{\mathrm{\beta }}_2\text{ETC}+{\mathrm{\beta }}_3\text{SDBOC*ETC}+\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{\beta }}_{\mathrm{n}}{\mathrm{\text{CV}}}_{\mathrm{\text{it}}}+{\mathrm{\epsilon }}_{\mathrm{\text{it}}}$$(2)

Remarks: FP = TQ and MBV; SDBOC = Sex diversity of the board of commissioners; ETC = ethnic background of president commissioner; CV = control variable.

4. Results and discussion

4.1. Descriptive statistics

A summary of the descriptive statistics is shown in Tables 3–6, consisting of the dependent, independent, and control variables with a five-year observation period from 2017 to 2021. The descriptive statistical tables show the wide distribution of data for all the variables in this study. TQ had a value ranging from 0.12 to 7.76, with an average value of 1.43, whereas MBV had a fairly distant value ranging from −8.26 to 70.57, with an average value of 2.17. Meanwhile, the average for sex diversity (SDBOC) was 0.15, with a range from 0.00 to 0.50. The maximum SDBOC value, which was far from one, showed that the level of sex diversity of the boards of commissioners in Indonesia was still low. Based on Table 4, 64.11% of companies in Indonesia did not have sex diversity; thus, 64.11% of the boards of commissioners were represented by men. Nonetheless, boards of commissioners with one female member constituted 28.86%, while 5.87% had two women members, and 1.16% of boards had three women members. Hence, the maximum number of women on the board of commissioners was three. Meanwhile, Table 5 shows that 92.46% of the presidents’ ethnic backgrounds were dominated by patriarchal ethnicities, which was indicated by code 2, while code 1 showed that 7.54% of the president commissioners’ ethnic backgrounds were matrilineal. Table 6 shows that 71.01% of the presidents’ ethnic backgrounds were dominated by indigenous Indonesians, as shown by code 1. The remaining backgrounds of the president commissioners, who come from foreign ethnicities, was 28.99%, as indicated in code 2. The remaining variables, NBOC, PRONBOID, NBOD, NWBOD, NBODI, NAC, PRONWAC, FIRMAGE, FIRMSIZE, LEV, and CR all showed wide variations, thereby implying adequate variations in the research variables.

Table 3. Descriptive statistics.

Variable	Obs	Mean	Std. Dev	Min	Max
TQ	783	1.43	1.11	0.12	7.76
MBV	783	2.17	5.73	−8.26	70.57
SDBOC	783	0.15	0.20	0.00	0.50
ETC	783	1.93	0.26	1.00	2.00
AGE	783	61.41	11.18	36.00	89.00
MAJORITY	783	1.72	0.86	1.00	3.00
GRADE	783	2.24	0.78	1.00	5.00
NBOC	783	3.96	1.77	2.00	10.00
PRONBOID	783	0.18	0.16	0.00	0.67
NBOD	783	4.61	2.06	2.00	12.00
NWBOD	783	0.58	0.88	0.00	6.00
NBODI	783	0.69	0.52	0.00	3.00
NAC	783	3.00	0.35	1.00	5.00
PRONWAC	783	0.20	0.25	0.00	1.00
FIRMAGE	783	39.91	17.69	4.00	104.00
FIRMSIZE	783	21.58	1.65	17.72	25.92
LEV	783	0.49	0.27	−0.91	0.94
CR	783	0.01	0.02	0.00	0.10

Note: TQ denotes Tobin’s Q; MBV denotes the market to book value, SDBOC denotes the sex diversity of the board of commissioners; ETC denotes the ethnic background of the president commissioner; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; NBOD denotes the size of the board of directors; NWBOD denotes the number of women on the board of directors; NBODI denotes the size of the independent board of directors; PRONBOID denotes the proportion of independent directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity.

Table 4. Frequency of NWBOC.

Number of women	Number of observations	Frequency (%)
0	502	64.11
1	226	28.86
2	46	5.87
3	9	1.16
Total	783	100

Table 5. Frequency of ETC.

Code	Ethnic background of president commissioner
	Frequency	Percentage (%)
1	59	7.54
2	724	92.46
Total	783	100

Table 6. Frequency of native and non-native Indonesian ethnicities.

Code	Ethnic background of president commissioner
	Frequency	Percentage (%)
1	556	71.01
2	227	28.99
Total	783	100

4.2. Correlation analysis

The correlation analysis results are presented in Table 7 for all the dependent, independent, and control variables in the research model. Correlation analysis in the regression analysis was used to ensure that the research model was free from multicollinearity issues. The correlation analysis was also based on pairwise correlation coefficients. The pairwise correlation coefficients in Table 6 showed that the degree of correlation among all the variables was rather low, indicating no serious multicollinearity problems among the variables included in our study. In addition, according to Gujarati (2003), the maximum limit of the correlation coefficient is 0.70. Based on the results in Table 6, all the coefficients were below 0.70; thus, the pairwise correlation coefficients showed that the research model was free from any potential multicollinearity.

Table 7. Pairwise correlations.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
(1) TQ	1.000
(2) MBV	0.257*	1.000
(3) BI	0.027	−0.008	1.000
(4) ETC	0.027	0.028	0.062	1.000
(5) AGE	−0.078	−0.040	0.075	−0.123*	1.000
(6) MAJORITY	0.017	0.039	0.014	−0.032	0.340*	1.000
(7) GRADE	0.057	0.063	−0.129*	0.056	−0.281*	−0.487*	1.000
(8) NBOC	−0.030	0.006	−0.078	−0.023	0.268*	0.060	0.103*	1.000
(9) NBOD	−0.014	0.101*	−0.104*	0.010	0.201*	0.056	0.045	0.586*	1.000
(10) NWBOD	0.025	0.114*	0.091	0.005	0.046	−0.014	0.039	0.080	0.328*
(11) NBODI	0.035	0.011	−0.051	−0.040	−0.100*	−0.016	0.012	0.013	0.102*
(12) PRONBOID	0.037	−0.016	0.032	−0.017	−0.196*	−0.047	−0.012	−0.247*	−0.357*
(13) NAC	0.063	−0.018	−0.115*	−0.011	−0.031	0.025	−0.037	0.130*	0.151*
(14) PRONWAC	−0.087	−0.026	0.246*	−0.023	0.037	0.010	−0.015	−0.081	−0.065
(15) FIRMAGE	0.030	0.094*	0.047	−0.006	0.140*	0.036	0.096*	0.196*	0.173*
(16) FIRMSIZE	−0.063	0.019	−0.143*	−0.030	0.218*	0.099*	0.045	0.578*	0.628*
(17) LEV	−0.011	0.003	−0.015	0.017	−0.067	−0.063	0.122*	−0.009	−0.009
(18) CR	0.029	−0.039	−0.077	0.016	−0.046	−0.108*	0.106*	−0.046	−0.036

Table

Variable	(10)	(11)	(12)	(13)	(14)	(15)	(16)	(17)	(18)
(10) NWBOD	1.000
(11) NBODI	0.044	1.000
(12) PRONBOID	−0.089	0.794*	1.000
(13) NAC	0.018	0.035	−0.042	1.000
(14) PRONWAC	0.052	−0.068	−0.009	−0.031	1.000
(15) FIRMAGE	0.113*	−0.037	−0.099*	0.038	0.032	1.000
(16) FIRMSIZE	0.139*	0.020	−0.251*	0.179*	−0.126*	0.250*	1.000
(17) LEV	0.044	0.037	0.047	−0.023	0.090	−0.051	−0.054	1.000
(18) CR	0.044	0.004	0.009	0.002	0.018	−0.007	−0.075	0.497*	1.000

Note: TQ denotes Tobin’s Q; MBV denotes the market to book value; SDBOC denotes the sex diversity of the board of commissioners; ETC denotes the ethnic background of the president commissioner; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; NBOD denotes the size of the board of directors; NWBOD denotes the number of women on the board of directors; NBODI denotes the size of the independent board of directors; PRONBOID denotes the proportion of independent directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity. *p < 0.1

4.3. Multivariate analysis

4.3.1. Empirical findings on the effect of the sex diversity of the board of commissioners on firm performance

The first research hypothesis outlines that the sex diversity of the board of commissioners (SDBOC) affects firm performance (FP). Table 8 presents the empirical findings for the effect of SDBOC on FP, as measured using market-based TQ and MBV. Models 1 and 4 were tested with the independent variable only, models 2 and 5 were tested using the independent and control variables, and models 3 and 6 were tested using the independent variable, the control variables, and firm-fixed effects (year effect and industry effect). Models 1, 2, and 3 in Table 8 show that SDBOC had a positive effect on TQ. These findings support H1, in that the greater sex diversity of the board of commissioners does increase FP. Besides, SDBOC’s positive influence on TQ is consistent with the prediction that greater SDBOC can provide better supervision of management, because of the board’s diverse ideas, knowledge, and experience in carrying out tasks, thus improving firm performance (Nur & Darvin, 2023; Sahoo et al., 2023; Soare et al., 2022). Hence, SDBOC’s contribution can increase the success of a company (Safiullah et al., 2022). The presence of women on the board of commissioners also enhances the supervisory function, and makes executives more cautious and vigilant when making investment decisions (Sahoo et al., 2023). The current results support the research results reported by previous studies (Abdullah et al., 2022; Hosny & Elgharbawy, 2022; Joecks et al., 2023; Nguyen & Huynh, 2023; Nur & Darvin, 2023; Soare et al., 2022).

Table 8. Effect of SDBOC on FP.

	(1)	(2)	(3)	(4)	(5)	(6)
	TQ	TQ	TQ	MBV	MBV	MBV
Intercept	1.451***	1.924***	1.694***	2.197***	7.913***	6.241***
	(39.49)	(7.95)	(10.16)	(8.68)	(2.01)	(1.58)
SDBOC	0.499**	0.766*	0.757*	2.208	3.097	3.276
	(1.95)	(1.74)	(1.74)	(0.80)	(1.10)	(1.17)
AGE		−0.005	−0.004		−0.033	−0.025
		(−1.29)	(−1.47)		(−1.62)	(−1.23)
MAJORITY		0.129***	0.114***		0.668 **	0.560**
		(2.71)	(2.68)		(2.40)	(2.01)
GRADE		0.132**	0.137***		0.669**	0.694**
		(2.40)	(2.37)		(2.16)	(2.25)
NBOC		−0.031	−0.033		0.679***	0.671***
		(−0.91)	(−1.02)		(3.48)	(3.46)
PRONBOID		0.407	0.251		2.561	1.554
		(1.17)	(0.58)		(0.82)	(0.50)
NBOD		0.077	0.076		0.409**	0.402**
		(1.54)	(1.44)		(2.12)	(2.09)
NWBOD		0.347***	0.367***		1.178**	1.074**
		(5.14)	(5.22)		(2.11)	(1.93)
NBODI		−0.069	−0.035		−0.792	−0.580
		(−1.07)	(−0.49)		(−0.91)	(−0.67)
NAC		0.221**	0.225***		−0.422	−0.386
		(2.99)	(3.35)		(−0.71)	(−0.65)
PRONWAC		0.423***	0.421***		−0.916	−0.884
		(4.37)	(−2.210)		(−1.09)	(−1.06)
FIRMAGE		0.003**	0.003**		0.023**	0.020*
		(2.35)	(2.37)		(1.93)	(1.69)
FIRMSIZE		0.083**	0.084***		0.341**	0.345**
		(2.83)	(3.09)		(1.99)	(2.02)
LEV		−0.172**	−0.217**		0.768	0.453
		(−2.44)	(−2.21)		(0.87)	(0.51)
CR		2.541	2.573		−18.144	−17.451
		(1.19)	(1.08)		(−1.24)	(−1.20)
Year effect			Yes			Yes
Industry effect			Yes			Yes
Adj. R^2	0.19	0.22	0.37	0.00	0.05	0.12
N	783	783	783	783	783	783
F-stat	0.887***	0.936***	0.909***	0.447***	0.304***	0.339***

Note: TQ denotes Tobin’s Q; MBV denotes the market to book value; SDBOC denotes the sex diversity of the board of commissioners; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; NBOD denotes the size of the board of directors; NWBOD denotes the number of women members of the board of directors; NBODI denotes the size of the independent board of directors; PRONBOID denotes the proportion of independent directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity. t-Statistics in parentheses; *p < 0.10; **p < 0.05; ***p < 0.01.

These empirical findings support the two theories used in the present study, namely the agency theory and RDT. The agency theory and RDT support the idea that female members of the board of commissioners carry out supervisory duties more effectively, in a way that increases the company’s legitimacy and improves firm performance (Arvanitis et al., 2022; Hosny & Elgharbawy, 2022). Altering the gender composition of Malaysian companies’ boards, by appointing women directors to the boards, and adding more directors, reportedly led to higher financial performance (Abdullah et al., 2022). Hosny and Elgharbawy (2022) found evidence that women’s councils can improve board effectiveness and help companies access finance and other resources. The oversight process, with women included in the composition of the board of commissioners, can also improve market-based financial performance (TQ) (Joecks et al., 2023). Similarly, Nguyen and Huynh (2023), in their study, found that women members of the boards of commissioners had a positive influence on the financial performance of real estate enterprises listed on the Vietnam Stock Exchange. Thus, greater sex diversity on the board of commissioners increases company profitability (Nur & Darvin, 2023). However, the results of this study do not support those of other past studies (Almarayeh, 2021; Alshirah et al., 2022; Arora, 2022; Wang, 2023) in which women members of the boards of commissioners did not affect market-based (TQ) financial performance.

Models 4, 5, and 6 in Table 8 show that SDBOC had no effect on FP proxied with MBV; hence, these empirical findings do not support H1. Since SDBOC had no effect on MBV, this result was consistent with a very low MBV score, at an average of 2.14. Table 3 shows that the average MBV value of 2.14 was far from the maximum MBV value of 70.57. MBV also had a higher fluctuation, because its standard deviation was greater than its average value of 5.73. This finding is in line with the findings reported by Liew and Devi (2021). In addition, since the minimum MBV value reached −8.28, this showed a negative equity book value. MBV research has shown that the market values a company lower than its equity book value. If it is associated with SDBOC, then the sex diversity of the board of commissioners is not able to affect the value of the company through the value of equity. The low MBV was due to several industries in this study, which tended to be volatile or of a high-risk nature (e.g. clothing and textile manufacturing, the electronics industry, automotive, the chemical and pharmaceutical industry, the oil and gas industry, computer manufacturing and information technology). Generally, industries that have low MBVs are those that tend to rely on significant physical assets and have thinner profit margins. In the observation year 2020–2021, the whole world experienced the COVID-19 pandemic, resulting in panic in the stock markets and falling MBV. Clothing and textile manufacturing companies, and those in the electronics industry, automotive, and oil and gas industries in Indonesia also experienced a decline in their share prices.

Additionally, the current results did not support RDT, which found empirical evidence that the sex diversity of the board of commissioners was associated with diverse resources, in terms of the members’ opinions and knowledge, to ensure that the monitoring process ran well and FP increased (Alshirah et al., 2022; Arvanitis et al., 2022). Companies that had sex diversity on their boards of commissioners tended to have more effective monitoring systems, but when the type of industry was high risk or fluctuated, investors were not interested in buying shares. Such conditions caused the book value of equity to be lower; as a result, MBV became low. The current results are also in agreement with Fujianti (2018), who reported that gender diversity had no effect on MBV.

4.3.2. Empirical findings on the moderating effect of the ethnic background of the president commissioner on the relationship between the sex diversity of the board of commissioners and firm performance

The second research hypothesis outlines that the ethnic background of the president commissioner (ETC) moderates the influence of the sex diversity of the board of commissioners (SDBOC) on firm performance (FP). Table 9 presents the empirical findings on the moderating effects of ETC on the relationship between SDBOC and FP, as measured by TQ and MBV. Models 1 and 3 were tested with the independent variable, the moderating variable and the control variables, models 2 and 4 were tested using the independent variable, the moderating variable, the control variables, and firm-fixed effects (year effect and industry effect).

Table 9. Moderating role of ETC.

	(1)	(2)	(3)	(4)
	TQ	TQ	MBV	MBV
Intercept	1.798***	1.560***	8.105**	6.531**
	(7.12)	(5.30)	(2.27)	(2.10)
SDBOC	1.648**	1.604*	3.412	3.125
	(1.94)	(1.73)	(1.24)	(1.20)
ETC	0.087*	0.089	0.008	0.021
	(1.67)	(1.42)	(0.04)	(0.09)
SDBOC*ETC	0.750**	0.760**	1.600*	1.666**
	(2.22)	(2.08)	(1.82)	(2.05)
AGE	0.005*	−0.004	−0.035	−0.028
	(1.76)	(−1.35)	(−1.43)	(−1.49)
MAJORITY	0.128***	0.112**	0.614*	0.507*
	(2.68)	(2.46)	(1.77)	(1.72)
GRADE	0.133***	0.136**	0.667*	0.687**
	(2.60)	(2.43)	(1.83)	(1.90)
NBOC	−0.031	−0.029	0.666***	0.655***
	(−0.97)	(−0.85)	(7.12)	(−6.28)
PRONBOID	0.328	0.162	2.214	1.119
	(0.84)	(0.36)	(0.77)	(0.35)
NBOD	0.074	0.072	0.381	0.369
	(1.60)	(1.47)	(1.47)	(1.36)
NWBOD	0.341***	0.356***	1.228**	1.130**
	(5.08)	(4.68)	(2.18)	(2.03)
NBODI	−0.081	−0.052	−0.734	−0.540
	(−1.28)	(−0.72)	(−1.08)	(−0.81)
NAC	0.226***	0.232***	−0.415	−0.377
	(3.48)	(3.86)	(−0.99)	(−0.93)
PRONWAC	0.420***	0.415***	−0.952	−0.919**
	(4.73)	(4.07)	(−1.44)	(−1.44)
FIRMAGE	0.003**	0.003**	0.024	0.021
	(2.47)	(2.59)	(1.27)	(1.12)
FIRMSIZE	0.084***	0.085***	0.341*	0.345*
	(3.21)	(3.23)	(1.73)	(1.81)
LEV	−0.173**	−0.220**	0.800	0.491
	(−2.31)	(−2.15)	(1.26)	(1.03)
CR	2.466	2.535	20.171*	19.716**
	(1.21)	(1.13)	(1.79)	(2.10)
Year effect		Yes		Yes
Industry effect		Yes		Yes
Adj.R^2	0.060	0.067	0.069	0.080
N	783	783	783	783
F-stat	0.928***	0.902***	0.581***	0.574***

Note: TQ denotes Tobin’s Q; MBV denotes the market to book value; SDBOC denotes the sex diversity of the board of commissioners; ETC denotes the ethnic background of the president commissioner; SDBOC*ETC denotes the interaction variable between the sex diversity of the board of commissioners and the ethnic background of the president commissioner; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; PRONBOID denotes the proportion of independent directors; NBOD denotes the size of the board of directors; NWBOD denotes the number of women on the board of directors; PRONWBOD denotes the proportion of women on the board of directors; NBODI denotes the size of the independent board of directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity. t-Statistics in parentheses; *p < 0.10; **p < 0.05; ***p < 0.01.

Models 1 and 3 in Table 9 were tested using the SDBOC*ETC interaction with TQ and the control variables, while models 2 and 4 were tested using the interaction of SDBOC* ETC with TQ, the control variables, and fixed effects (year effect and industry effect). Models 1 and 3 showed that the coefficient of SDBOC*ETC was positive and significant; this suggested that the relationship between SDBOC and TQ had improved significantly. For example, based on the relationship between SDBOC and TQ in Table 8, model 3 had a coefficient value of 0.757, while SDBOC in Table 9 showed that model 2 had a coefficient value of 1.604. These results provide support for H2, which meant that the president commissioner’s ethnic background moderated the relationship between SDBOC and TQ. Likewise, models 2 and 4 showed that the value of the coefficient of SDBOC*ETC was positive and significant, and this showed that the relationship between SDBOC and MBV had also improved significantly. For example, based on SDBOC’s relationship in Table 8, model 6 had a coefficient value of 3.276, while SDBOC in Table 9 showed that model 3 had a coefficient value of 3.125. These results support H2, which meant that the president commissioner’s ethnic background moderated the relationship between SDBOC and MBV.

Overall, the analysis results support UET, which states that the success of a company depends on the main actor (Neffe et al., 2022). The main actor in this study referred to the president commissioner in charge of supervising a company, and the supervision system of the company would be effective if the president commissioner of the company had competent resources. Based on UET, these competent resources can be seen from the knowledge and experience of the leaders who determine the success of the company (Prosvirkina & Wolfs, 2021). Leadership experience can also be determined by the ethnic background of the president commissioner.

In this study, the ethnic background of the president commissioner was grouped into two categories: ethnicity adhering to a matriarchal system, and ethnicity adhering to a patriarchal system. The ethnic backgrounds of most of the president commissioners in manufacturing companies in Indonesia stem from ethnicities that adhere to patriarchal beliefs. Table 4 shows that 92.46% of the president commissioners came from ethnic backgrounds that adhere to the patriarchal system. Based on the field data, a total of 724 observations demonstrated ethnicities that adhered to the patriarchal system, which consisted of Javanese with 45%, Sundanese with 30%, Batak with 5%, and other tribes with 20%. Leaders who came from the Javanese group showed a style of leadership that was more polite, wise, and respectful of others (Endraswara, 2013). When a president commissioner from a Javanese ethnic background had leadership spirit, the company valued other people more, thus strengthening the influence the sex diversity of the board of commissioners had on firm performance. Meanwhile, a president commissioner of Sundanese ethnicity showed the type of leadership that tended to be democratic and participatory (Wati & Hum, 2005). Therefore, the president commissioner with a Sundanese ethnic background had a democratic, participatory leadership spirit that strengthened the influence of the sex diversity of the board of commissioners on firm performance. Finally, a president commissioner who was of Batak ethnicity, had strong and authoritative leadership (Sidabutar & Mas’ud, 2016), and this likewise strengthened the influence the sex diversity of the board of commissioners had on firm performance.

The results of this analysis are consistent with the empirical findings reported by Karim et al. (2022). Specifically, Karim et al. (2022) found that the ethnic background of the board of commissioners served as a company resource, which created a competitive advantage that improved firm performance. Other empirical evidence has also shown that the ethnic background of a president commissioner, who has more effective supervisory capabilities, can reduce agency costs, thus increasing firm performance (Darmadi, 2013). Ethnic commissioners who have wise, democratic, participatory, respectful, authoritative, and protective leadership styles do improve the monitoring system, and such leaders further support SDBOC in improving FP.

Nonetheless, the current results do not support the research undertaken by Guest (2019), which found no evidence of ethnic diversity, but the boards of commissioners do strengthen their monitoring to increase firm performance. The results also contradict the findings of Hassan et al. (2015) in which ethnic diversity had no impact on the market value, but was inversely correlated. Although Gul et al. (2016) supported the results of this study, in their study, the relationship between the ethnic diversity of the boards of commissioners and corporate performance was non-linear. At low levels of ethnic diversity, there were positive associations; these associations were negative at higher levels of ethnic diversity.

4.4. Additional analysis

4.4.1. Indigenous and non-native Indonesian ethnicities

The ethnic backgrounds of president commissioners in Indonesia comprise indigenous Indonesians and non-native Indonesians (people from foreign countries). A value of one was given to president commissioners with an indigenous Indonesian ethnic background, whereas a value of two was given to those with non-native Indonesian ethnic backgrounds. Table 6 shows that 71.01% of the presidents’ ethnic backgrounds encompassed indigenous Indonesians, which meant that the presidents’ ethnic backgrounds were dominated by native Indonesians, while the remaining 28.99% of the president commissioners’ ethnic backgrounds encompassed foreign countries. The ethnic grouping for the presidents of the board of commissioners was based on Hassan et al. (2015) study, which grouped ethnicities into Bumiputra and non-Bumiputra (Chinese and/or Indian) categories. Based on the condition that the ethnic backgrounds of president commissioners from foreign countries encompass diverse ethnicities, the research sample was divided into indigenous Indonesian and non-native Indonesian ethnic groups.

Table 10 presents the empirical findings regarding the influence of SDBOC on FP, with the samples divided into indigenous Indonesian and non-native Indonesian ethnic groups. Models 1 and 6 were tested with the independent variable, the control variables and firm-fixed effects (year effect and industry effect), models 2 and 5 were tested using the independent variable and the control variables, Models 3 and 4 were tested with the independent variable, the moderating variable, the control variables and firm-fixed effects (year effect and industry effect). Models 3 and 4 in Table 10 examined the moderating effect of SDBOC*ETC on FP, as measured by TQ and MBV. Model 3 showed that the coefficient value of SDBOC against TQ was positive and significant, and this result was consistent with the values of the coefficient of SDBOC-TQ in Table 10, where models 1, 2, and 3 increased from −0.235, −0.571, and 1.351 respectively. The analysis results were consistent with those in Table 9; therefore, it could be concluded that the ethnic background of the president commissioner moderated the relationship between SDBOC and TQ. The results from testing model 4 also provided the same empirical evidence, such that the value of the coefficient of SDBOC against MBV was positive and significant. Similarly, it can also be concluded that the ethnic background of the president commissioner moderated the relationship between SDBOC and TQ and MBV. Based on Table 6, it can be concluded that 71.01% of indigenous Indonesians moderated the influence of SDBOC on FP. While the remaining 28.99% of foreign ethnicities moderated the influence of SDBOC on FP.

Table 10. Moderating role of indigenous and non-native Indonesian ethnicities.

	(1)	(2)	(3)	(4)	(5)	(6)
	TQ	TQ	TQ	MBV	MBV	MBV
Intercept	0.663	3.725	0.672	4.135	8.014**	3.725
	(0.73)	(0.55)	(0.736)	(0.594)	(2.026)	(0.557)
SDBOC	−0.235	−0.571	1.351**	2.742	−0.051	−0.571
	(−1.12)	(−0.70)	(2.325)	(1.246)	(−0.048)	(−0.707)
ETC			0.058	−0.194
			(0.517)	(−0.335)
SDBOC*ETC			1.310***	2.808*
			(3.065)	(1.875)
AGE	0.005	0.024	0.004	0.022	−0.034	0.024
	(1.22)	(1.32)	(1.053)	(1.211)	(−1.646)	(1.326)
MAJORITY	0.121**	0.621*	0.105*	0.558*	0.627**	0.621*
	(2.04)	(1.81)	(1.796)	(1.683)	(2.257)	(1.814)
GRADE	0.122**	0.741	0.120**	0.737	0.631**	0.741
	(2.19)	(1.54)	(2.131)	(1.531)	(2.036)	(1.542)
NBOC	−0.006	−0.284*	−0.006	−0.286*	−0.656***	−0.284*
	(−0.16)	(−1.70)	(−0.161)	(−1.732)	(−3.373)	(−1.703)
PRONBOID	−0.028	−0.882	0.028	−0.663	2.212	−0.882
	(−0.04)	(−0.32)	(0.047)	(−0.245)	(0.716)	(−0.323)
NBOD	0.009	0.320	0.016	0.340	0.390**	0.320
	(0.29)	(1.18)	(0.508)	(1.252)	(2.028)	(1.180)
NWBOD	−0.426***	−1.336**	−0.424***	−1.315**	1.158**	−1.336**
	(−3.41)	(−2.03)	(−3.488)	(−1.975)	(2.078)	(−2.032)
NBODI	−0.041	−0.554	−0.047	−0.588	−0.752	−0.554
	(−0.26)	(−0.75)	(−0.314)	(−0.800)	(−0.864)	(−0.755)
NAC	0.114	−0.246	0.098	−0.310	−0.428	−0.246
	(0.93)	(−0.58)	(0.790)	(−0.712)	(−0.715)	(−0.580)
PRONWAC	−0.263	−1.279**	−0.247	−1.281**	−0.909	−1.279**
	(−1.56)	(−2.11)	(−1.462)	(−2.124)	(−1.080)	(−2.114)
FIRMAGE	0.002	0.010	0.002	0.010	0.024*	0.010
	(0.70)	(0.40)	(0.731)	(0.434)	(1.955)	(0.401)
FIRMSIZE	−0.024	−0.222	−0.025	−0.224	−0.337*	−0.222
	(−0.56)	(−0.65)	(−0.600)	(−0.664)	(−1.954)	(−0.655)
LEV	−0.070	1.887	−0.036	2.072	0.603	1.887
	(−0.36)	(1.25)	(−0.187)	(1.346)	(0.682)	(1.259)
CR	0.597	−21.939*	0.600	−23.368*	−14.571	−21.939*
	(0.23)	(−1.83)	(0.231)	(−1.822)	(−0.966)	(−1.835)
Year effect	Yes		Yes	Yes		Yes
Industry effect	Yes		Yes	Yes		Yes
Adj.R^2	0.25	0.21	0.25	0.04	0.02	0.03
N	783	783	783	783	783	783
F-stat	0.909***	0.935***	0.890***	0.371***	0.304***	0.339***

Note: TQ denotes Tobin’s Q; MBV denotes the market to book value; SDBOC denotes the sex diversity of the board of commissioners; ETC denotes the ethnic background of the president commissioner; SDBOC*ETC denotes the interaction variable between the sex diversity of the board of commissioners and the ethnic background of the president commissioner; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; PRONBOID denotes the proportion of independent directors; NBOD denotes the size of the board of directors; NWBOD denotes the number of women on the board of directors; NBODI denotes the size of the independent board of directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity. t-Statistics in parentheses; *p < 0.10; **p < 0.05; ***p < 0.01.

Essentially, the analysis results are consistent with UET, which states that a company becomes successful owing to the role of the main actor (Neffe et al., 2022). In this study, the president commissioner was the main actor, who acts as the leader supervising the company. The ethnic backgrounds of the original Indonesian president commissioners, which included Javanese (having a wise character and respecting others), Sundanese (having a democratic, participative leadership spirit), and Batak (having a strong, authoritative leadership spirit), were postulated to be able to create harmonious gender diversity on the board of commissioners, which ultimately increases FP. These findings are consistent with those of Karim et al. (2022), who showed that the board of commissioners’ ethnic backgrounds are a company asset that can increase FP. However, the current results do not support the study by Hassan et al. (2015), which found that ethnic diversity had no impact on a company’s market value, thus showing an inverse correlation.

4.4.2. Effects of WBOCDUM, PRONWBOC, and NWBOC on FP

Additional analysis was also performed by replacing the proxy of the independent variable with WBOCDUM, PRONWBOC, and NWBOC. Table 11 presents the empirical findings on the influence of WBOCDUM, PRONWBOC, NWBOC, and FP. Models 1 and 4 were tested with the independent variable (WBOCDUM), the moderating variable, the control variables and firm-fixed effects (year effect and industry effect), models 2 and 5 were tested using the independent variable (PRONWBOC) the moderating variable, the control variables and firm-fixed effects (year effect and industry effect), while models 3 and 6 were tested with the independent variable (NWBOC), the moderating variable, the control variables and firm-fixed effects (year effect and industry effect).

Table 11. Effects of PRONWBOC, NWBOC, and WBOCDUM on FP.

	(1)	(2)	(3)	(4)	(5)	(6)
	TQ	TQ	TQ	MBV	MBV	MBV
Intercept	2.280***	2.291***	2.267***	−39.291	−42.164	−42.043
	(10.78)	(12.58)	(9.27)	(−0.84)	(−0.90)	(−0.83)
WBOCDUM	1.154***			12.199
	(4.41)			(1.20)
WBOCDUM*ETC	0.430***			2.559
	(2.92)			(0.70)
PRONWBOC		4.859***			6.314
		(4.73)			(0.21)
PRONWBOC*ETC		2.030***			9.037
		(3.92)			(0.73)
NWBOC			0.885***			11.954
			(3.09)			(1.27)
NWBOC*ETC			0.384***			2.537
			(3.86)			(0.79)
AGE	0.005*	0.006*	−0.005	−0.015	0.003	0.005
	(1.72)	(1.83)	(−1.497)	(−0.19)	(0.04)	(0.05)
MAJORITY	0.129***	0.131***	0.126***	0.725	0.685	0.671
	(3.24)	(3.09)	(3.12)	(1.09)	(0.93)	(0.82)
GRADE	0.128**	0.133**	0.129**	−0.839	−0.845	−0.860
	(2.52)	(2.34)	(2.31)	(−0.63)	(−0.65)	(−0.64)
NBOC	−0.015	−0.030	−0.020	3.843*	−3.578	−3.584*
	(−0.44)	(−0.93)	(−0.60)	(1.65)	(−1.92)	(−1.79)
PRONBOID	0.242	0.385	0.265	20.606	21.635	21.457
	(0.80)	(1.33)	(0.80)	(0.72)	(1.13)	(0.98)
NBOD	0.068	0.077	0.072	5.277	5.312	5.309
	(1.49)	(1.55)	(1.55)	(1.14)	(1.24)	(1.17)
NWBOD	0.337***	0.342***	0.331***	−10.961	−10.324	−10.357
	(5.12)	(4.93)	(4.57)	(−0.89)	(−0.97)	(−0.92)
NBODI	−0.055	−0.071	−0.056	0.126	−0.212	−0.146
	(−1.09)	(−1.22)	(−0.92)	(0.02)	(−0.04)	(−0.03)
NAC	0.210***	0.198***	0.205***	−4.134	−4.136	−4.123
	(3.75)	(3.55)	(3.38)	(−1.18)	(−1.28)	(−1.21)
PRONWAC	0.404***	0.394***	0.391***	−6.599	−6.127	−6.112
	(5.01)	(4.41)	(4.20)	(−1.31)	(−1.53)	(−1.43)
FIRMAGE	0.004***	0.004**	0.004***	−0.027	−0.025	−0.024
	(2.71)	(2.89)	(2.72)	(−0.50)	(−0.50)	(−0.49)
FIRMSIZE	0.098***	0.097***	0.099***	1.751	1.765	1.754
	(4.19)	(3.73)	(3.78)	(0.91)	(0.97)	(0.91)
LEV	−0.083	−0.065	−0.073	2.948	3.033	3.035
	(−1.33)	(−0.90)	(−1.00)	(0.79)	(0.90)	(0.89)
CR	0.242	0.221	0.170	−3.473	−18.144	−1.800
	(0.33)	(0.24)	(0.20)	(−0.03)	(−1.24)	(−0.02)
Year effect	Yes	Yes	Yes	Yes	Yes	Yes
Industry effect	Yes	Yes	Yes	Yes	Yes	Yes
Adj.R^2	0.12	0.15	0.12	0.01	0.01	0.01
N	783	783	783	783	783	783
F-stat	0.934***	0.935***	0.9341***	0.305***	0.304***	0.306***

Note: TQ denotes Tobin’s Q; MBV denotes the market to book value; WBOCDUM*ETC denotes the interaction variable between the presence of women on the board of commissioners and the ethnic background of the president commissioner; PRONWBOC denotes the proportion of women on the board of commissioners; PRONWBOC*ETC denotes the interaction variable between the proportion of women on the board of commissioners and the ethnic background of the president commissioner; NWBOC denotes the number of women on the board of commissioners; NWBOC*ETC denotes the interaction variable between the number of women on the board of commissioners and the ethnic background of the president commissioner; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; PRONBOID denotes the proportion of independent directors; NBOD denotes the size of the board of directors; NWBOD denotes the number of women on the board of directors; NBODI denotes the size of the independent board of directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity. t-statistics in parentheses; *p < 0.10; **p < 0.05; ***p < 0.01.

Models 1, 2, and 3 showed that WBOCDUM, PRONWBOC, and NWBOC positively affected TQ. However, the test results for models 4, 5, and 6 showed that WBOCDUM, PRONWBOC, and NWBOC had no effect on MBV. With the presence of women members of the board of commissioners measured using a dummy, the percentage and number of female board of commissioners’ members increased TQ (Hosny & Elgharbawy, 2022; Joecks et al., 2023), but it did not affect MBV (Fujianti, 2018). The presence of women on the board of commissioners showed sex diversity in the board of commissioners, as they brought diverse knowledge and experience, which increased the monitoring and TQ (Sahoo et al., 2023). The results were also consistent with RDT, which states that the presence of women members on the board of commissioners can create greater efficiency in the company’s supervision and increase FP (Abdullah et al., 2022). However, since many companies had low MBV rates, the presence of women members of the board of commissioners as supervisors had no influence on it. Despite the sex diversity in the board of commissioners, if the type of industry is very high-risk, then investors will be reluctant to invest and this decreases the equity value. High-risk industries (e.g. clothing and textile manufacturing, the electronics industry, automotive, the chemical and pharmaceutical industry, the oil and gas industry, computer manufacturing and information technology) may create uncertainty and reduce investor interest, which may further affect MBV.

The results for WBOCDUM, NWBOC, and PRONWBOC testing against TQ in models 1, 2, and 3, in Table 11, were consistent with those of Table 8 for model 1 to model 3, and the results could support H1. However, the results for WBOCDUM, NWBOC, and PRONWBOC against MBV in models 4, 5, and 6, in Table 11, were consistent with those of Table 8 for model 4 to model 6. Hence, the results did not support H1.

The empirical findings in Table 11 also showed that the interactions of WBOCDUM*ETC, PRONWBOC*ETC, and NWBOC*ETC with TQ had positive and significant coefficients. These results showed that the ethnic background of the president commissioner could strengthen the influence of sex diversity, as measured through dummy variables, namely the presence of women board of commissioners’ members (WBOCDUM), the percentage of women board of commissioners’ members (PRONWBOC), and the number of women board of commissioners’ members (NWBOC), on TQ. These results supported UET, which states that the leader is the main actor and determines the success of a team (Mahsina & Agustia, 2023). The main actor in this study involved the role of the president commissioner from the viewpoint of his/her ethnic background, in order to strengthen the influence of sex diversity on TQ. The test results for models 1, 2, and 3 in Table 11 are consistent with those for models 1 and 2 in Table 9.

The interactions of WBOCDUM*ETC, PRONWBOC*ETC, and NWBOC*ETC with MBV showed positive and insignificant coefficients. Therefore, the results showed that the ethnic background of the president commissioner could not strengthen the influence of sex diversity, as measured through dummy variables, namely the presence of women board of commissioners’ members (WBOCDUM), the percentage of women board of commissioners’ members (PRONWBOC), and the number of women board of commissioners’ members (NWBOC), on MBV. These results did not support UET, as outlined by Mahsina and Agustia (2023), as the condition of the data in the field was found to demonstrate low MBV values and some values were also negative.

4.4.3. Effects NWBOC1, NWBOC2, and NWBOC3 on FP (critical mass theory)

One of the points of the Sustainable Development Goals (SDG) is gender equality. Gender equality can be practiced in the corporate world, and it includes the presence of women on the boards of commissioners, which function as the management’s supervisors. According to Safiullah et al. (2022), the role of sex diversity on the board of commissioners is crucial for the success of a business. However, Elmagrhi et al. (2018) highlighted that no consistent evidence exists about how many women commissioners are needed to efficiently improve FP. To answer this, we adopted the critical mass theory, which describes the concept that there is a minimum number of individuals from one particular group (e.g. women), required in an environment or organization, for significant changes in gender dynamics to occur (Garanina & Muravyev, 2021; Tleubayev et al., 2020).

This study follows Brahma et al. (2021), who divided three dummy variables to represent the level of sex diversity. Based on the descriptive statistics in Table 4, the maximum number of women on the board of commissioners in one company is three. Therefore, we arranged sex diversity into three categories, namely the presence of one female member on the board of commissioners (WBOC1), the presence of two women members on the board of commissioners (WBOC2), and the presence of three women members on the board of commissioners (WBOC3).

Models 1, 2, and 3 in Table 12 show that increasing the number of women board members to a certain level had a significant effect on TQ. Model 1 was tested with the independent variable (board of commissioners consisting of one woman/NWBOC1), the control variables and firm-fixed effects (year effect and industry effect), model 2 was tested using the independent variable (board of commissioners consisting of two women/NWBOC2), the control variables and firm-fixed effects (year effect and industry effect), and model 3 was tested with the independent variable (board of commissioners consisting of three women/NWBOC3), the control variables and firm-fixed effects (year effect and industry effect).

Table 12. Effects of NWBOC1, NWBOC2, and NWBOC3 on FP.

	(1)	(2)	(3)
	TQ	TQ	TQ
Intercept	1.012	0.728	0.954
	(1.108)	(0.803)	(1.042)
NWBOC1	−0.136
	(−0.949)
NWBOC2		0.279*
		(1.669)
NWBOC3			−0.236
			(−0.466)
AGE	0.003	0.005	0.004
	(0.881)	(1.298)	(0.933)
MAJORITY	0.118*	0.115*	0.119*
	(1.926)	(1.937)	(1.958)
GRADE	0.115**	0.114**	0.117**
	(2.041)	(2.046)	(2.079)
NBOC	−0.006	−0.009	−0.009
	(−0.203)	(−0.228)	(−0.294)
PRONBOID	0.165	0.292	0.185
	(0.255)	(0.471)	(0.285)
NBOD	−0.032	0.027	−0.028
	(−0.954)	(0.815)	(−0.833)
NWBOD	−0.016	−0.490***	−0.013
	(−0.301)	(−3.769)	(−0.250)
NBODI	−0.025	−0.095	−0.033
	(−0.158)	(−0.617)	(−0.211)
NAC	0.143	0.145	0.136
	(1.185)	(1.177)	(1.129)
PRONWAC	−0.233	−0.235	−0.212
	(−1.310)	(−1.403)	(−1.196)
FIRMAGE	0.001	0.002	0.001
	(0.405)	(0.648)	(0.364)
FIRMSIZE	−0.032	−0.032	−0.029
	(−0.739)	(−0.772)	(−0.674)
LEV	−0.072	−0.059	−0.063
	(−0.368)	(−0.309)	(−0.325)
CR	1.437	0.655	1.258
	(0.561)	(0.258)	(0.489)
Year effect	Yes	Yes	Yes
Industry effect	Yes	Yes	Yes
Adj.R^2	0.23	0.25	0.23
N	783	783	783
F-stat	0.940***	0.928***	0.936***

Note: TQ denotes Tobin’s Q; MBV denotes the market to book value; NWBOC1 has a value of 1 if it has one female member on the board of commissioners and 0 otherwise; NWBOC2 has a value of 1 if it has two women members on the board of commissioners and 0 otherwise; NWBOC3 has a value of 1 if it has three women members on the board of commissioners and 0 otherwise; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; PRONBOID denotes the proportion of independent directors; NBOD denotes the size of the board of directors; NWBOD denotes the number of women on the board of directors; NBODI denotes the size of the independent board of directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity. t-Statistics in parentheses; *p < 0.10; **p < 0.05; ***p < 0.01.

Specifically, model 2 showed that the presence of two female board of commissioners’ members (NWBOC2) increased TQ; however, the presence of three female members on the board of commissioners led to a decrease in FP. Nonetheless, NWBOC2 had a positive and significant coefficient of 0.279, and when NWBOC3 was included in the model, the value decreased to a negative and insignificant one, of -0.236. These results suggested that increasing the number of women on a board of commissioners up to a certain number (two members) could improve board decisions. Typically, companies that have three female board of commissioners’ members find conflicts among the board members increase, which negatively impacts TQ. Likewise, with the presence of one woman member on a board of commissioners (NWBOC1), the negative and insignificant coefficient was -0.136. This showed that one female member on a board of commissioners increased FP ineffectively, as measured by TQ, and these test results were consistent with the critical mass theory. However, in relation to the effect of SDBOC on MBV in Table 7, the insignificant results found in the tests using the critical mass theory were not tested.

4.5. Robustness analysis

Robustness analysis is a statistical analysis that aims to create a research model that is strong or resistant to variations or disturbances in the data. The robustness analysis for the effect of SDBOC on FP, as measured by TQ and MBV in this study, could be carried out in two ways:

1. Heckman selection model.

2. Fixed-effects.

The sample for this study was limited to the manufacturing companies listed on the Indonesia Stock Exchange from 2017 to 2021. Limiting research samples to just manufacturing companies, with five years of data, typically creates a bias in the sample’s selection process. However, the presence of bias in the sample’s selection process, when tested using the regression analysis, would negatively affect the reliability of the findings. According to Đặng et al. (2020), the Heckman selection model can be used to overcome the potential negatives in the reliability of the findings. The Heckman selection model estimates the selection model (the probability model for entry into the sample) and the master regression model (the model that relates the dependent variable to the independent variable) (Heckman, 1979). By integrating these two models, the Heckman test overcomes selection bias and produces more consistent estimates.

The Heckman selection model in this study consisted of two stages. The first stage integrated the independent variable of SDBOC into a dependent variable measured by a dummy variable (SDBOCDUM). Specifically, SDBOCDUM was used to analyze the decision to appoint (1) or not appoint (0) a female member to the board of commissioners and have it continue to act as a supervisory board in the company. Accordingly, this study identified the independent variables that could influence a company’s decision to appoint or not appoint a female member to the board of commissioners. Based on the previous studies by Saha (2023), an instrument variable that can affect SDBOCCUM was employed, namely the lagged value of the percentage of women board of commissioners’ members (LAGPRONWBOC). In the first stage, the dependent variable was SDBOCDUM, while the independent variable affecting SDBOCDUM was LAGPRONWBOC. Notably, the control variables were also incorporated into the research model.

In the second stage of the Heckman selection model, we calculated IMR1 and integrated it into a regression model to control self-selection bias. The results for the Heckman selection model are presented in Table 13 for models 2 and 3. As can be seen in Table 13, model 2 shows that IMR1 had a positive and significant coefficient, and it was also found that SDBOC had a positive effect on TQ. These test results were consistent with those in Table 8 for models 1, 2, and 3. Subsequently, the test results showed that ETC moderated the effect of SDBOC-TQ, and these results were likewise consistent with those in Table 8 for models 1 and 2.

Table 13. Heckman two-stage regression and fixed effects.

	Heckman two-stage regression			Fixed effects
	(1)	(2)	(3)	(4)	(5)
	SDBOCDUM	TQ	MBV	TQ	MBV
Intercept	0.710	0.741	−89.858	1.965***	51.293
	(0.44)	(0.789)	(−0.988)	(9.57)	(−0.92)
SDBOC		0.682**	−2.436	2.225***	17.392
		(2.375)	(−0.301)	(2.72)	(0.68)
SDBOC*ETC		0.346***	0.586	0.606**	1.519
		(3.217)	(0.187)	(2.56)	(0.43)
LAGPRONWBOC	5.975***	–	–	–	–
	(13.27)	–	–	–	–
IMR1		0.067**	−0.361	–	–
		(0.599)	(−0.276)	–	–
AGE	0.016**	0.004	0.141	−0.004	0.039
	(1.97)	(0.979)	(1.098)	(−1.63)	(0.35)
MAJORITY	−0.267**	0.121*	−0.338	0.119***	0.020
	(−2.55)	(1.903)	(−0.267)	(3.30)	(0.02)
GRADE	−0.434***	0.129**	−0.930	0.139***	−0.667
	(−3.48)	(2.143)	(−0.472)	(2.64)	(−0.54)
NBOC	0.242***	0.001	−4.791	−0.007	3.596*
	(3.38)	(0.024)	(−1.011)	(−0.20)	(1.66)
PRONBOID	−1.445	0.113	21.126	0.121	13.805
	(−1.40)	(0.195)	(0.774)	(0.33)	(0.66)
NBOD	−0.010	0.025	5.781	0.069	5.219
	(−0.15)	(0.756)	(0.964)	(1.58)	(1.11)
NWBOD	−0.495**	−0.446***	−15.612	0.351***	−11.678
	(−2.18)	(−3.613)	(−1.049)	(5.18)	(-0.88)
NBODI	−0.188	−0.081	−1.013	−0.044	1.178
	(−0.67)	(−0.566)	(−0.314)	(−0.81)	(0.20)
NAC	0.018	0.149	−1.909	0.227***	−3.788
	(0.08)	(1.196)	(-0.691)	(3.70)	(−1.10)
PRONWAC	0.62*	−0.278*	−9.006	0.417***	−6.549
	(1.90)	(−1.675)	(−1.250)	(4.76)	(−1.34)
FIRMAGE	0.001	0.003	0.070	0.003***	−0.047
	(0.19)	(0.731)	(0.872)	(2.67)	(−0.74)
FIRMSIZE	−0.151**	−0.036	3.302	0.098***	1.722
	(−2.25)	(−0.848)	(0.957)	(4.12)	(0.88)
LEV	−0.144	−0.006	2.808	−0.137	0.658
	(−0.41)	(−0.036)	(0.425)	(−1.42)	(0.16)
CR	−13.753***	−0.162	−19.136	0.349	1.272
	(−2.81)	(−0.249)	(−1.098)	(0.51)	(0.01)
Year effect	Yes	Yes	Yes	Yes	Yes
Industry effect	Yes	Yes	Yes	Yes	Yes
Adj.R^2	0.11	0.24	0.07	0.12	0.02
N	684	783	783	783	783
F-stat	0.233***	0.118***	0.211***	0.871***	0.351***

Note: TQ denotes Tobin’s Q; MBV denotes the market to book value; SDBOC denotes the sex diversity of the board of commissioners; ETC denotes the ethnic background of the president commissioner; SDBOC*ETC denotes the interaction variable between the sex diversity of the board of commissioners and the ethnic background of the president commissioner; LAGPRONWBOC denotes the lagged value of the percentage of women on the board of commissioners; IMR1 denotes the effect of self-selection bias; AGE denotes the age of the president commissioner; MAJORITY denotes the educational background of the president commissioner; GRADE denotes the education level of the president commissioner; NBOC denotes the size of the board of commissioners; PRONBOID denotes the proportion of independent directors; NBOD denotes the size of the board of directors; NWBOD denotes the number of women on the board of directors; NBODI denotes the size of the independent board of directors; NAC denotes the size of the audit committee; PRONWAC denotes the proportion of women on the audit committee; FIRMAGE denotes the firm’s age; FIRMSIZE denotes the company’s size; LEV denotes leverage; CR denotes liquidity. t-Statistics in parentheses; *p < 0.10; **p < 0.05; ***p < 0.01.

Additionally, Table 13 shows that model 3 demonstrated negative and insignificant coefficients for IMR1. Moreover, the value of the variable of SDBOC’s coefficient against MBV was positive and insignificant. These test results were consistent with those in Table 8 for models 4, 5 and 6. Based on the Heckman testing, the selection model for this study was, therefore, robust.

A robustness analysis can also be carried out using fixed effects to overcome endogeneity problems (Al Farooque et al., 2020; Bin Khidmat et al., 2020). Table 13 for models 4 and 5 presents the test results using fixed effects. Overall, model 4 showed that SDBOC positively affected TQ, and these results were consistent with those in Table 8 for models 1, 2, and 3. Besides, the interaction of SDBOC*ETC with TQ was positive and significant, which was consistent with the results in Table 9 for models 1 and 2. Meanwhile, model 5 showed a positive and insignificant SDBOC-MBV coefficient. These test results were consistent with those in Table 8 for models 4, 5, and 6. Based on the fixed effect model, the results of this study were robust.

5. Conclusion, implications, limitations, and future research directions

The purpose of this study is to test whether SDBOC influences FP and to ascertain whether the chief commissioner’s ethnic background can moderate the relationship between SDBOC and FP. We studied companies for five years, from 2017 to 2021. The first contribution of this study enriches the literature on the relationship of SDBOC’s effect on FP when measured with TQ, but there is no effect when measured with MBV. The second contribution of the study is the ethnic background of the chief commissioner (ETC) does moderate the relationship between SDBOC and FP, as measured by TQ and MBV. Furthermore, the third contribution is that this study examines whether ETCs grouped into indigenous and non-indigenous Indonesian ethnic groups moderate the relationship between SDBOC and FP. Remarkably, the results showed that ETC does have the ability to moderate the relationship between SDBOC and FP, as measured by TQ and MBV. The fourth research contribution involves the presence of female members on the boards of commissioners, as measured using the influence of a dummy (WBOCDUM), a percentage (PRONWBOC), and the number of female members on a board of commissioners (NWBOC) on FP. The results show that WBOCDUM, PRONWBOC, and NWBOC have a positive effect on TQ; however, the study found that WBOCDUM, PRONWBOC, and NWBOC have no effect on MBV. The fifth research contribution is whether ETC moderates the influence of WBOCDUM, PRONWBOC, and NWBOC on FP. Interestingly, the results show that ETC is able to moderate the effects of WBOCDUM, PRONWBOC, and NWBOC on FP, measured by TQ, but not MBV. The sixth contribution of this study is that increasing the number of female members on a board of commissioners to a certain level, namely two female board members, has a significant effect on TQ. The seventh research contribution involves the adoption of a multi-theoretical framework that helps explain the various reasons that define FP, namely the agency theory, the resource dependence theory, and the upper echelon theory, to explain the impact of SDBOC on FP.

The practical contribution of this study is that companies need to pay attention to the composition of their boards of commissioners by encouraging gender diversity when conducting a monitoring process to improve FP, considering that Indonesian culture tends to follow patriarchal beliefs. In addition, the findings of this study can provide input for shareholders to pay attention to the ethnic background of the president commissioner, who will oversee the operations of the company. The ethnic background of the chief commissioner is important, considering that Indonesia’s large number of ethnic groups have differing leadership styles, which can influence supervision. Thus, this research makes a theoretical contribution, and can be applied to future studies, so it can be replicated, and its practical contributions can be used as a reference for companies in the election of their boards of commissioners.

One of the limitations of this study is that it only focuses on manufacturing companies in Indonesia, which is a developing country. In this regard, future research can compare FP in developed and developing countries, and also across various sectors such as the service sector, the mining sector, and the trade sector in these countries. Next, since FP’s variable measurement is only contingent on the market-based concept, further research can compare FP using market-based and accounting-based concepts (return on assets; return on equity). With regard to the expansion of the theory, the results of this study also suggest that future studies can improve its theoretical foundations by using other theories, such as the feminist theory (Teng et al., 2022), the human capital theory (Martinez-Jimenez et al., 2020), and the socio-psychological theory (Pandey et al., 2022) when examining the impact of SDBOC on FP. Finally, although the current results are strong for the estimates and alternative models, they limit the analysis for sex diversity to the board of commissioners. Hence, further research may consider the impact of other factors, e.g. country-level factors (inflation rate, macroeconomic conditions), and further enhance the readers’ understanding by conducting in-depth interviews and qualitative analyses, to gain further insights into the relationships between ETC, SDBOC, and FP.

Author contributions

Conceptualization, N.E. and N.S.; methodology, N.E. and N.S.; data curation N.E., and N.S.; validation, N.E.; formal analysis, N.E. and N.S.; writing—original draft preparation, N.E.; writing—review and editing, N.S.; supervision, N.S.; project administration, N.S.; All authors have read and agreed to the published version of the manuscript.

Acknowledgment

The authors received no direct funding for this research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data used in the present study will be available on request, from the corresponding author.

Additional information

Notes on contributors

Nanik Ermawati

Nanik Ermawati is a PhD student in Accounting, Faculty of Economics and Business, Airlangga University. He is also active as a lecturer at the Bachelor of Accounting Faculty of Economics and Business, Muria Kudus University. Her research focus is management accounting, corporate governance.

Noorlailie Soewarno

Noorlailie Soewarno SE., MBA., Ak., is an Professor of accounting at Faculty of Economics and Business, Universitas Airlangga, Indonesia. She is an active and professional researcher and has published numerous articles in reputable journals. Her focus includes management accounting, strategic management, corporate social responsibility, and green accounting.