“Is gender diversity good for the quality of accruals in Indonesia?”

There is no regulation in emerging markets, i.e. Indonesia imposing the gender diversi- ty quota. Therefore, based on the different characteristics and attributes between developed and emerging markets, this study aims to analyze the impact of the boardroom’s gender diversity on the quality of accruals, i.e. earnings in big Indonesian firms. This study used a sample of 100 big manufacturing firms in Indonesia. Moreover, this paper utilized a widely used proxy for the quality of earnings and the boardroom’s gender diversity. Data panel with fixed effect method was used to compute the quality of ac- cruals while path analysis was utilized for hypotheses testing. Furthermore, the results showed evidence that the boardroom’s gender diversity indirectly influenced the qual- ity of accruals through cash flow variability. However, the presence of female directors is not a significant inducement for managers to report better quality of earnings. The managerial implication of this study is that the boardroom’s gender diversity is good for the emerging markets since it supports better accounting accruals indirectly. Future studies should incorporate many industries and consider the potential of endogeneity, which has not been tackled in this paper. in big firms. The results also support the idea that


INTRODUCTION
Business nowadays facing the rapid dynamics, these dynamics coming from both internal and external factors. Along the complexity of the business environment and technology advancement, there is an increased awareness of the boardroom's gender diversity (Jizi & Nehme, 2017). Boardroom's gender diversity is seen as a governance tool that can improve the quality of governance, transparency, monitoring, and protection of the rights of shareholders. Women's representation is seen as a signal of governance quality. Gender diversity in the board of directors can promote new talents and provide access to important parties from the external environment. the boardroom's gender diversity improves transparency and monitoring of management activities, therefore the quality of decision-making is also increasing (Herring, 2017;Isidro & Sobral, 2015;Triana & Asri, 2017). Hence, it is logic if the boardroom's gender diversity has positive relationship with earnings quality. Some studies in developed market found the positive effect between earnings quality and the boardroom's gender diversity (Jadiyappa et  However, different setting occurs in Indonesia. There are no regulated quotas in Indonesia regarding the boardroom's gender diversity. In other words, there is no corporate governance recommendation for appointing female executives to companies listed on the Indonesia Stock Exchange (IDX). It is highly likely that the implementation of IFRS, high-quality accounting standards, promotes cross-country investment (Krismiaji et al., 2019). Concisely, there is no regulation in Indonesia imposing the boardroom's gender diversity quota under the Indonesia Stock Exchange's (IDX) rules and the Financial Services Authority. So it is important to scrutinize this issue in Indonesia' setting. Saona et al. (2019) suggested that the number of female directors has increased especially in European nations. Their results confirmed the advantages of having a balanced gender in directors' positions.

LITERATURE REVIEW, AIM, AND HYPOTHESES
In fact, the presence of female directors can reduce earnings management. In addition, it was also shown that the advantages of having female directors, such as better firm performance and better earnings management, are more evident in countries that impose quota restrictions on the composition of companies' boards. Similarly, Pasaribu (2017) showed the importance of female directors on firm performance especially for smaller firms in the UK. According to resources dependence theory (RDT), which is proposed by Pfeffer and Salancik (1978), the board of directors is functioned to communicate to external parties especially in dealing with resource dependence; hence, the relationship of boardroom's diversity and firm performance is to make minimal environmental dependences. Hillman et al. (2009) stated that the board's diversity shows commitment to minorities and it may add legitimacy to companies. It sends a signal that companies support equal opportunities. Another related theory is agency theory, popularized by Jensen and Meckling (1976

Sample and data
This study followed the previous study in the sample selection process ( Table 1.
Further, this study implemented a purposive sampling method. The first requirement of the purposive sampling was public companies listed in Indonesia Stock Exchange from 2012 until at least 2017. Secondly, the sample was manufacturing companies. The main reason for using the manufacturing firms was that there was a different approach to obtaining net income across different industries. Therefore, only one specific industry was taken. Consequently, the results could be compared among firms.
In addition, the data were hand-collected from the Indonesian Capital Market Directory (ICMD) and the firms' official websites. In addition, the reason for using the year 2012 as a starting point was that Indonesia has fully started to implement IFRS convergence in accounting practice since 2012. According to Juniarti et al. (2018), the IFRS convergence in Indonesia started with the adoption phase of 2008-2010, the preparation in 2011, and the implementation phase in 2012. Based on those characteristics, the data used in this study is panel data and it will be analyzed by using the panel regression method.

Models and measurements
The paper used Eviews for the calculation of earnings quality; meanwhile, LISREL was used for structural equation modeling (SEM) for the path analysis method. According to Hair et al. (2007), the sample required to test hypotheses in SEM is between 100 and 200. In addition, the sample selection is also in line with Teh et al. (2017). Hence, the sample in this paper is in line with the requirement.
The study implemented accrual-based earnings management proxies for quality of earnings because previous research has mostly employed accruals modification as a proxy for the earnings management ( where TCA a,b is ∆CA (the difference of current assets) -∆CL (the difference of current liabilities) -∆Cash (the difference of cash) + ∆STDEBT (the difference of debt) -DEP (company's depreciation); CFO a,b is the cash flow from operations of the firm a in year b; ∆REV a,b is the change in revenue; PPE a,b is property, plant, and equipment. All values are divided by total assets. ε a,b is residual. The quality of earnings is the standard deviation on the estimated residual of the total current accrual of firm a in year b as in equation (1). It indicated that a larger variability of the residuals represents a lower quality of earnings. However, according to Ismail et al. (2015), the quality of earnings computation demands sufficient data. At least eight years of data are demanded to compute the quality of earnings. However, it was stated that some emerging markets such as Malaysia were not able to provide the data required. Therefore, the study followed Ismail et al. (2015) and used six years of annual data for quality of earnings computation. The residuals of equation (1) (1). There was 500 firm-years observation for quality of earnings (AQF) measurement.
As for boardroom diversity (BOARD1), the study followed Tarigan et al. (2018) and Kılıç and Kuzey (2016). This paper used Blau Index for BOARD1 variable. The formula for the Blau Index is: where P i 2 is the portion of male and female board members and n is two. 0 is the minimum value while the maximum is 0.5 (Kılıç & Kuzey, 2016). This paper used the proportion of female directors from 2012 to 2017. For example, if there was a difference in female directors' proportion from 2012 to 2017, then the latest condition was used.
Moreover, low quality of earnings could result from high business uncertainties. These uncertainties could lead to the high variability of cash flows from operations. The presence of females in the boards could send a signal to various stakeholders such as investors and customers that the company supported gender equality and it can bring stability in cash flow since many female directors are appointed as chief financial officers. Moreover, this study used Ismail et al. (2015) and Nugroho et al. (2019) for cash flow stability measurement (CFOVA R). Hence, the formula for cash flow stability is: where CFO a was the cash flow from operating and a CFO was the average of cash flow from operating. In this study, a standard deviation of cash flow from five years of data was used.
Further, the structural equation model was constructed. This model was processed with LISREL statistical program. The models are: where AQF is quality of earnings as previously defined; CFOVA R is the cash flow variability as previously described; BOARD1 is the boardroom's gender diversity variable computed with Blau Index; SHAR is a control variable, it is dummy variable (1 is for a sharia-compliant firm) else 0, the researcher used Indonesia Sharia Stock Index (ISSI) as a reference whether a company is a sharia-compliant or not; BIG4 is also a control variable, it is dummy variable (1 is for big4 auditing firm) else 0; SIZE is a control variable, it is the natural logarithm of companies' total assets; S A L E SVA R is a control variable, it is the standard deviation of sales revenue from 2013 to 2017; LOSS is a control variable, it is dummy variable (1 is for the companies that have a loss from 2013 to 2017) else 0; LEV is also a control variable, it is total debt divided by total assets.
where TCA a,b is the total current accrual of the company a in year b (as previously described in equation (1); CFO a,b is the cash flow from operations of firm a in year b. All values are divided by total assets. ε a,b is residual.
According to the model (AQD), the quality of accruals is the standard deviation of residuals. In other words, the larger the variation of residuals in equation (7), the lower the quality of earnings. Further, the study followed Saona where BOARD2 is the boardroom's diversity variable computed with Shannon Index; the rest of the variables are previously mentioned. Figure 1 is the research framework.

RESULTS
Based on the mean values ( quality of earnings (AQD and AQF), significantly better cash flow (CFOVAR), and significantly better LEV. Based on the mean difference statistical test, manufacturing companies with gender diversification are statistically different in earnings management (AQF and AQD), cash flow variability (CFOVAR), and LEV compared to the ones without gender diversification. Table 3 shows the descriptive statistics for dummy values. The number of sharia-compliant big manufacturing companies (SHAR) with female directors was marginally less than the companies without gender diversification. Similarly, the number of manufacturing companies with gender diversi-fication in boards that used BIG4 auditing firms was marginally less than the companies without gender diversification. In addition, the number of manufacturing companies with gender diversification in boards was more profitable compared to the companies without gender diversification and this result was in line with Dwiharti and Adhariani (2018).
This paper used path analysis to test the hypotheses using LISREL. Therefore, it was required to compute the goodness of fit. Table 4 shows the goodness of fit from structural equations (4), (5), and (6). According to the goodness of fit values, it shows that the equation modeling was in line  with the threshold values except for RFI and AGFI, which were still acceptable. In other words, the equation could be used to test the hypotheses. In addition, Table 5 shows the direct effect of variables in equations (4), (5), and (6). This study also observed the indirect effect. In this case, there was only one indirect effect studied, variable BOARD1 on AQF through CFOVAR. Table 6 shows the indirect effect of this particular variable. The meth-od used for analyzing the mediation effect of CFOVAR variable was the Sobel test.
In Table 5, variable BOARD1, which is boardroom's gender diversity, has a negative relationship and is insignificant with quality of earnings (AQF).
Based on the Sobel test (Table 6), it is confirmed that cash flow variability is the mediator for the  boardroom's gender diversity and earnings quality in Indonesian manufacturing firms.
Further, Table 7 shows the results of a sensitivity check. Similar to Table 6, the method used for analyzing the mediation effect of CFOVAR variable was the Sobel test. Moreover, consistent with the previous result, Table 7 also shows the goodness of fit. The results of Table 7 confirm the result of Table  5. Moreover, an indirect effect of BOARD2 on AQD through CFOVAR is in Table 8 and the results are  in line with Table 6. Hence, this study found robust evidence that cash flow variability is the mediator of the boardroom's gender diversity and quality of accruals for Indonesian big manufacturing firms.

DISCUSSION
Manufacturing companies with female directors were not statistically different in SIZE and sales variability (SALESVAR) compared to the ones without gender diversification. Moreover, BOARD1 (Blau Index) and BOARD2 (Shannon Index) show that male directors were still dominating in the manufacturing companies' board of directors. Tarigan

CONCLUSION
This study aims to examine the boardroom's gender diversity and quality of accruals i.e. earnings in Indonesia by incorporating a mediator that indirectly connects the boardroom's gender diversity to the quality of earnings in Indonesia. By using the panel regression this study found that the boardroom's gender diversity does not have a significant effect on the quality of earnings, while the involvement of women in the board of directors would improve corporate governance and they become controlling agents. The results revealed that cash flow variability is the mediator for the boardroom's gender diversity and earnings quality in Indonesian big manufacturing firms. The results also support the idea that female boards avoid risk-taking action directors rather than commissioners. This study revealed that, unlike many European countries, there is no corporate governance recommendation for appointing female executives to companies listed in the Indonesia Stock Exchange (IDX). In addition, female and male directors are supported by sharia-compliant status and it can reduce earnings management practice. So Indonesian companies should consider some females on the board, also the Financial Service Authority should consider issuing some regulations regarding these matters.
There are some limitations. First, the sample was limited to manufacturing firms. There was a different approach to obtaining net income across different industries. Therefore, only one specific industry was taken. Consequently, the results could be compared among firms. It is recommended that further studies can incorporate many industries. Second, the results could be influenced by endogeneity, which has not been tackled in this paper.
The managerial implication is that although the boardroom's gender diversity does not directly affect earnings management, it is encouraged that the boardroom's gender diversity should be maintained.
The results also revealed that female directors are able to stabilize cash flow variability.