The influence of multinational board characteristics on earnings management in foreign subsidiaries

The influence of multinational board characteristics on earnings

management in foreign subsidiaries

Master thesis in Accountancy and Controlling Author: Rachèl Voppen

Student number: S2944758 Supervisor: dr. Simona Rusanescu Date of submission: 18-01-2020 Word count: 11.973 words University of Groningen

2 ABSTRACT

Monitoring the quality of the financial reports is one of the main responsibilities of the Board. This study examines how parent Board independence and gender diversity might influence the level of earnings management in the foreign subsidiary and will bring new literature insights. Extant literature documents that Board independence and gender diversity are negatively associated with the level of earnings management in standalone firms. This study shows if the parent Board composition is able to influence the earnings management decisions of the subsidiaries and might increase the financial reporting quality. The data is gathered from multinationals which parent firm is in the United States and subsidiaries in Europe. The results indicate that parent Board independence has no association with the subsidiaries’ level of earnings management. The absence of proximity makes it difficult for the parent Board to supervise the decisions. Also, the lack of information independent members face, might decrease Board effectiveness. Therefore, parent CEO do not have to engage earnings management abroad. Parent Board gender diversity has an increasing effect on the level of earnings management in the subsidiary. Suggesting that the parent CEO is not able to engage earnings management at parent level and is taking advantage of the multinationals’ corporate structure by doing this abroad.

Key words: Multinationals, Board of Directors, earnings management, gender diversity, independence

3 TABLE OF CONTENT

INTRODUCTION ... 4

THEORY ... 9

Parent Board independence and subsidiaries financial reporting quality ... 11

Parent Board Gender diversity and subsidiaries financial reporting quality ... 14

METHODOLOGY ... 17 Variables ... 17 Empirical model ... 19 RESULTS ... 22 Correlations ... 23 Regression ... 25 DISCUSSION ... 28 Implications ... 32 Future research ... 33 REFERENCES ... 35 APPENDIX A ... 40 APPENDIX B ... 41

4 INTRODUCTION

According to the Financial Times: ‘96% of the S&P 500 firms in the US are using at least one non-Generally Accepted Accounting Principles (GAAP) method in their financial statements’ (Financial Times, 2019). These S&P 500 firms mostly consist of multinational corporations (MNC) which have subsidiaries in different foreign countries (Beurs FD, 2020). The

International Accounting Standards Board (IASB), an independent organisation that develops and approves International Financial Reporting Standards, found similar results when

analyzing different variations of operating profit (Financial Times, 2019; IASPlus, 2020). This phenomenon, in which companies use different methods to alter their financial reporting to mislead stakeholders about the performance, is known as earnings management (Healy & Wahlen, 1999). Over the years, various researchers have studied how earnings management affects the company and its financial reporting quality (e.g., Beuselinck, Cascino, Deloof, & Vanstraelen, 2019; Healy & Wahlen, 1999; Burgstahler & Dichev, 1997; Beyer, Guttman, & Marinovic, 2019). According to the Conceptual Framework of the IASB, the users of

financial reports e.g., existing and potential investors, lenders, creditors and staff, must rely on financial reports for much of the financial information they need. Therefore, the

information must give a reliable and relevant representation (IFRS, 2018). When companies manage earnings to mislead these stakeholders about their performance, this could give an incorrect view of the organization and users may suffer wealth losses (Healy & Wahlen, 1999; Wilson, 2011). This is why understanding factors which are likely to motivate and constrain firms’ earnings management activities, is of interest to researchers (Wilson, 2011).

The majority of studies in the earnings management literature examine factors which play a role in how pervasive earnings management is. Managers have different types of incentives to engage in earnings management, e.g., external contract incentives, management compensation contract incentives, regulatory motivations and capital market motivations (Healy & Wahlen, 1999; Noronha, Zeng, & Vinten, 2008). Prior studies have suggested that internal corporate governance mechanisms play a crucial role in deterring earnings

management (e.g., Dechow, Ge, & Schrand, 2010; Bekiris & Doukakis, 2011; García-Meca & Sánchez-Ballesta, 2009). The internal corporate governance mechanisms consist of the Board of Directors (Board) and ownership structures (Denis & McConnell, 2003; García-Meca & Sánchez-Ballesta, 2009). Monitoring the quality of the financial reports is one of the main responsibilities of the Board (Board & Administrators, 2018). One way of assessing this financial reporting quality, is examining to what extent earnings are managed (Van Tendeloo & Vanstraelen, 2011; Herath & Albarqi, 2017). Since Boards want to fulfill their

5 responsibilities and achieve quality, they can play a role in the level of earnings management (Orazalin, 2019; Barja & Cadez, 2018; Lo, Wong, & Firth, 2010; Dechow et al., 2010). The Board will in most cases try to reduce the managed earnings in order to enhance the quality of the financial reports.

In MNCs the aim of the Board is to improve the financial reporting quality of the MNC. The financial statements are consolidated in MNCs, which means that the financial statements of the parent firm and subsidiaries are reported together. Therefore, the MNC Board is likely to be concerned about the financial reporting quality of the subsidiaries and the parent firm (Beuselinck et al., 2019; Dechow et al., 2010). Several studies suggest that MNCs have substantial power over their subsidiaries and influence their decisions. This power could extend to financial reporting decisions of the subsidiaries as well (Egbe, Adegbite, & Yekini, 2018; Robinson & Stocken, 2013; Beuselinck et al., 2019). Financial reporting quality of the MNC is important for the Board, it might be likely that the parent Board aims to reduce the extent of earnings management at the subsidiary level. On the other hand, the parent firm can contribute to an increase in the subsidiaries’ level of earnings management. For the parent CEO it might be difficult to engage earnings management at the parent level, because the Board can oversee this. In MNCs most of the subsidiaries are abroad, so the parent CEO can take advantage of the absence of proximity. When the monitoring of the parent Board is effective, the CEO will not be able to manage earnings in the parent’s individual financial statement. Instead, the parent CEO may choose to engage earnings management in the foreign subsidiary, because the parent Board oversight is unlikely to be that effective (Warren et al., 2011; Al-khabash & Al-Thuneibat; 2008; Breton & Taffler, 1995). In this case, the parent Board might not deter parent CEOs from managing earnings abroad.

The studies about earnings management and Boards are focussed on the consolidated statements. As a result, the influences of the parent Board on the level of earnings

management of the subsidiaries is not being considered (Beuselinck et al., 2019). If the parent Board is not able to reduce earnings management in the subsidiaries, this could reduce the MNCs’ financial reporting quality. Therefore, it is important to study the financial statements of the subsidiaries, instead of the consolidated statements, to examine whether the parent Board is able to curb the level of earnings management of foreign subsidiaries (Bonacchi, Cipollini, & Zarowin, 2018; Beuselinck et al., 2019).

The aim of this study is to provide evidence on the extent to which a parent Board influences the level of earnings management in the foreign subsidiaries. The results can show

6 if the parent Board composition has an influence on the earnings management decisions of the foreign subsidiaries and how the financial reporting quality of the MNC might be increased.

Most of the MNCs have foreign subsidiaries which have their own Board (Beurs FD, 2020). Several studies suggest that the parent CEO has substantial power over the

subsidiaries’ Boards and influence their decisions (Robinson & Stocken, 2013; Egbe et al., 2018; Beuselinck et al., 2019). The role of the parent Board is to advise and monitor the parent CEO. Another main responsibility of the parent Board is to improve the financial reporting quality of the MNCs consolidated statements. When the parent Board wants to improve the financial reporting quality, it might use the influence on the subsidiaries (Beuselinck et al., 2019; Dechow et al., 2010). So, the parent Board could play a role in the level of earnings management of the subsidiary. It is also possible that the parent Board might not be effective in deterring earnings management in the foreign subsidiaries. For example, the absence of proximity makes it difficult for the parent Board to supervise the decisions of the foreign subsidiaries’ CEO and Board (Watson-O’Donnell, 2000).

Several studies found evidence that characteristics of a firms’ Board are likely to influence the level of earnings management (Barja & Cadez, 2018; Lo et al., 2010; Dechow et al., 2010; Davidson, Goodwin-Stewart, & Kent, 2005). The literature on the effects of the Board on the level of earnings management examine different characteristics of the Board (e.g., independence, size, tenure, age and gender diversity) (Barja & Cadez, 2018). The most studied Board characteristics will be addressed in this study: gender diversity and

independence. In the literature there is a conflicting view on the relationship between Board independence and earnings management (Klein, 2002; Vafeas, 2005; Agrawal & Chadha, 2005; Chen, Chang, & Wang, 2015). Several studies found that independent Board members in standalone firms are better monitors. Independent members are less subjected to influences from managers and are better in monitoring, which is likely to result in higher earnings quality (Klein, 2002; Osma, 2008; Vafeas, 2005). Literature widely documents that

independent Boards are better able to constrain earnings management. Accordingly, in order to enhance the MNCs’ financial reporting quality in MNCs, independent Boards can

discourage earnings management in the foreign subsidiaries. Therefore, the expectation is that higher Board independence is associated with a decrease in earnings management in the foreign subsidiaries. The independent directors are more objective and less subjected to the influence of managers and might therefore be better in monitoring and discouraging earnings management. However, if independent parent Boards deter earnings management at the parent level, it is likely that the parent CEO takes advantage of the corporate structure of the MNC.

7 Specifically, MNCs control many subsidiaries located in different countries around the world, which might enable the parent CEO to manipulate the earnings abroad, far away from the scrutiny of the parent Board (Watson-O’Donnell, 2000). This could result in an increase. A reason for a non-existent association between earnings management and parent Board independence is the lack of information independent Board members face. The independent Board members are outsiders of the firm and have less access to information than the insiders of the firm.

Gender diversity is in most studies associated with better Board effectiveness.

However there are also some studies which report the opposite (e.g., Orazalin, 2019; Terjesen, Sealy, & Singh, 2009; Damak, 2018; Albaum & Peterson, 2006). Prior studies on gender diversity provide evidence that a more diverse Board, so with the presence of male and female members, improves the financial reporting quality. Female directors are in comparison to male directors, more ethical in their behavior and professional judgment. Also, female members improve the communication within the Board. Better communication increase the monitoring ability and can result in less earnings management (García-Sánchez, Martinez-Ferrero & García-Meca, 2017; Fan, Jiang, Zhang, & Zhou, 2019; Orazalin, 2019; Damak, 2018). In MNCs the better communication and monitoring ability of the gender diverse parent Board can discourage earnings management in the foreign subsidiaries, which can lead to a decrease. On the other hand, the gender diverse parent Board can better prevent earnings management at the parent level. So the parent CEO might manage the earnings in the foreign subsidiaries. However, more communication of the gender diverse Board can be time

consuming and less effective (Smith, Smith, & Verner, 2006). Also, in gender diverse Boards there are more opinions and conflicts. These inefficiencies can result in less influence and overview of the parent Board over the subsidiaries. So the parent Board might not be effective in deterring earnings management in the subsidiaries. Given the arguments presented earlier, it is expected that gender diversity in the parent Board has a positive association with the level of earnings management in the foreign subsidiary.

In line with the above reasoning and research, the research question addressed in this study is: To what extent does the MNC Board independence and gender diversity affect the

level of earnings management of the foreign subsidiary?

The before mentioned studies about Board independence and gender diversity are in standalone firms. It is important to study this, to see which characteristics could play a role in the relationship between parent Board and subsidiaries’ financial reporting quality. Given the crucial monitoring role of the parent Board within MNCs, investigating whether two different

8 Board characteristics may deter foreign subsidiaries from engaging in earnings management provides new insights on Board effectiveness. This study aims to fill this gap.

This research focuses on a sample of MNCs with parent firms in the United States (US) and subsidiaries in Europe (EU) in the period of 2012 till 2017. The final sample

consists of 5298 observations of 883 unique subsidiaries. Discretionary accruals are used as a measure for earnings management. In order to measure the discretionary accruals, the Jones (1991) model is used. The main independent variables in this study are parent Board

independence and gender diversity. Board independence is measured as the percentage of independent directors divided by the total number of Board members. Board gender diversity is measured as the percentage of female directors divided by the total number of Board members.

The results suggest that parent Board independence has no association with the level of earnings management in the subsidiary. This indicate that the independent parent Board is not able to oversee the reporting decisions in the subsidiary. In addition, this study finds that parent Board gender diversity has a positive effect on earnings management in the subsidiary. This suggests that the parent CEO is taking advantage of the absence of proximity of the Board and engage earnings management in the foreign subsidiaries.

This study adds to the existent body of literature in the following ways. First, this study contributes to the literature on the relation between Board characteristics and earnings management. The existent studies on earnings management are focused on the consolidated financial statements and do not research the effect of the parent firm on subsidiaries (Dechow et al., 2010; Bonacchi et al., 2018). The studies in standalone firms show a conflicting view. Some studies state that Board independence has an effect on earnings management (e.g., Klein, 2002; Idris, Siam and Nassar, 2018), some state that there is no effect (e.g., Agrawal & Chadha, 2005). Also, the literature about Board gender diversity in standalone firms have conflicting results. There are studies which state that there is an association (e.g., Orazalin, 2019) and studies which state there is no relationship (e.g., Smith et al., 2006). However, this study investigates the relationship between earnings management in the context of MNCs and not in standalone firms. This provides a more thorough understanding of earnings

management and whether the parent Board limits this in the subsidiaries. This study creates insights on whether the parent Board is able to oversee the decisions of the subsidiaries and the effect of Board composition on earnings management. Next, this study allows to

distinguish between the effect of the Board on earnings management, since it considers two distinct Board characteristics. In the current body of literature these elements have conflicting

9 associations with earnings management (e.g., Klein, 2002; Vafeas, 2005; Chen et al., 2015; Agrawal & Chadha, 2005). Looking at these different characteristics of the parent Board bring a more nuanced perspective on how parent Board influences the financial reporting quality of the MNC, in particular the financial reporting quality of subsidiaries. The results of this study can be linked to the results of studies in standalone firms. Some studies in

standalone firms found a discouraging effect of Board independence and gender diversity, this effect is not seen in MNCs (Vafeas, 2005; Chen et al., 2015; Damak, 2018). This can indicate that the Board is less effective in monitoring financial reporting decisions as it comes to subsidiaries. This is a new insight for the literature.

The structure of the remainder of this paper is as follows: the following section describes the theoretical background of the paper and develops the hypotheses; the third section describes the methodology used in the empirical analysis and the fourth section contains the results of this study. Finally, the last section entails the discussion and conclusion.

THEORY

Financial reporting quality is an important topic in the accounting literature. Financial reporting quality refers to the extent to which financial statements provide reliable and relevant information about the underlying economic position and performance of an organisation (IFRS, 2018). Most studies in financial reporting quality literature are about measuring the quality (e.g., Beest, Braam, & Boelens, 2009; Cheung, Evans, & Wright, 2010; Gordon & Gallery, 2006) and what affects the financial reporting quality (see Herath & Albarqi (2017) for an overview). One way of assessing the quality of financial reports is examining to what extent earnings are managed, this phenomenon is known as earnings management (Van Tendeloo & Vanstraelen, 2011, Herath & Albarqi, 2017). Earnings management is perceived as a challenging issue due to its negative impact on the quality of the financial reports (Barghathi, Collinson, & Crawford, 2018; Ascioglu et al., 2012;

Habbash, Sindezingue, & Salama, 2013). According to Healy and Wahlen (1999): ‘Earnings management occurs when managers use judgment in financial reporting and in structuring transactions to alter the financial reports to mislead stakeholders about the underlying performance of the company’.

Discretionary accruals are used as a proxy to measure earnings management (Dechow et al., 2010; Bartov, Gul, & Tsui, 2001). Accruals are the difference between the reported earnings and cash flows from operations. The total accruals can be decomposed into

10 discretionary and nondiscretionary accruals (Dechow & Dichev, 2002). Nondiscretionary accruals are the expected level of adjustments to the cash flows. They arise from doing ‘usual business’ as a company (Drake & Fabozzi, 2012). Whereas discretionary accruals are

adjustment self-selected by the manager or CEO. The discretionary accruals enable to opportunistically change the financial numbers, due to the generally accepted discretion they get. E.g., flexibility in the decision of the depreciation method (Ahmed, 2018; Healy, 1985). This study uses discretionary accruals as a proxy for earnings management as well, since it is widely used in literature (Dechow et al., 2010).

Most of the studies in the earnings management literature analyze factors that play a role in the earnings management decision and the incentives and motivations for managers (e.g., Healy & Wahlen, 1999; Francis, Huang, Rajgopal, & Zang, 2008; Dechow et al., 2010; Capalbo, Frino, Mollica, & Palumbo, 2014). According to Healy and Wahlen (1999) there are different types of incentives for managers that encourage or discourage earnings management (e.g., external contract incentives, management compensation contract incentives, regulatory motivations and capital market motivation). Prior research suggest that internal corporate governance mechanisms play a crucial role in deterring earnings management (Dechow et al., 2010; García-Meca & Sánchez-Ballesta, 2009; Bekiris & Doukakis, 2011). These internal corporate governance mechanisms can be classified in: Boards and ownership structure (Denis & McConnell, 2003; García-Meca & Sánchez-Ballesta, 2009). Boards can play a crucial role in earnings management because monitoring the quality of the financial reports is one of their main responsibilities (Board & Administrator, 2018; Barja & Cadez, 2018; Lo et al., 2010). Since earnings management is an indicator for the financial reporting quality, it is likely that the Board aims to discourage this (Van Tendeloo & Vanstraelen, 2011; Herath & Albarqi, 2017; Davidson et al., 2005).

Most of the MNCs have foreign subsidiaries which have their own Board (Beurs FD, 2020). Different studies suggest that MNCs have substantial power over the subsidiaries and exert substantial influence on the financing, operating and investment decisions of the subsidiary. It is reasonable that this power could extend to financial reporting decisions as well (Egbe et al., 2018; Robinson & Stocken, 2013; Beuselinck et al., 2019). As mentioned before, one of the main responsibilities of the parent Board is to improve the financial reporting quality of the MNC. The parent firm uses the individual financial statements of the subsidiaries to draw up the consolidated report of the MNC. Therefore, it is presumable that the parent Board uses the substantial influence to achieve improvement of the financial

11 Board might evaluate the subsidiaries’ management and monitor the behavior in interest of the shareholders (Du, Deloof & Jorissen, 2015). However, the parent Board might not have enough influence to discourage earnings management in the foreign subsidiary. The absence of proximity can make it difficult for the parent Board to supervise the decisions of the CEO and management of the foreign subsidiaries (Watson-O’Donnell, 2000). Also, the parent CEO can take advantage of the absence of proximity. When the parent Board is likely to monitor effectively on parent level (and not on subsidiary level), the parent CEO may choose to engage earnings management in the foreign subsidiary (Warren et al., 2011; Al-khabash & Al-Thuneibat; 2008; Breton & Taffler, 1995).

Several studies found that Board effectiveness is likely to influence the level of earnings management in standalone firms. The different Board characteristics increase or decrease the Board oversight (Barja & Cadez, 2018; Lo et al., 2010; Dechow et al., 2010; Davidson et al., 2005). However, in the current body of literature there is no study on parent Boards characteristics and level of earnings management in the financial statements of the subsidiaries (Beuselinck et al., 2019; Dechow et al., 2010). It is important to do research to see if the Board characteristics also, or do not play a role when it comes to the influence on the subsidiary. This can create new insights in the literature on financial reporting quality in subsidiaries in particular and how parent Boards influence the quality of the MNCs’

consolidated statements. Also, the results for MNCs can differ from the results in standalone firms, because of the corporate structure of MNCs. Next, this research focuses on two different parent Board characteristics which gives a better understanding how earnings management is influenced. Furthermore, this research shows if the parent Board composition can play a role in the financial reporting quality.

In sum, it is not clear if parent Boards could have an influence on the level of earnings management in the subsidiary. The insights of this study create a more thorough

understanding of earnings management and how a MNC can increase their financial reporting quality. Another insight of this study is, whether the parent Board composition plays an important role in enhancing the quality. This study is aiming to fill this gap and shows if parent Boards can curb earnings management of foreign subsidiaries and if their effectiveness depends on Board composition.

Parent Board independence and subsidiaries financial reporting quality

The main role of the Board is to provide effective and unbiased monitoring. This task can be carried out the best if the one who is monitoring is independent of the one who is supervised

12 (Nolan, 2005). Board independency generally refers to the proportion of independent

directors within the Board (Tribbitt & Yang, 2017).

Independent directors are said to prevent conflicts of interests. Furthermore, they should monitor the management to detect and prevent managerial fraud, e.g., managerial mismanagement and inquire corporate affairs (Beleya & Ramendren, 2012). Independent directors are seen as more valuable than insiders, since they depend less on the management (Ringe, 2013). This can lower for example agency problems. Therefore, independent directors can enhance the Board effectiveness (Yermack, 1996). Independent directors have access to external resources that complement the skills of the insiders (Hilmann & Daziel, 2003; Faleye, 2015). Several studies present evidence suggesting that firm performance and

effective governance increase with Board independence (e.g., Brickley, Coles & Terry 1999; Dechow, Sloan & Sweeney, 1996; Klein, 2002). A higher percentage of Board independence could be related with more effective monitoring, because they are less subjected to the influence of managers and shareholders and therefore more objective (Vafeas, 2005; Chen et al., 2015). Effective monitoring could play a role in deterring earnings management. Effective monitors have more control over the financial reporting decisions. The studies of for example Klein (2002), Idris et al. (2018) and Jaggi, Leung and Gul (2009) found a negative association between Board independence and earnings management in standalone firms. This indicates that a higher number of independent directors in the Board reduce earnings management and thereby increase the financial reporting quality. Also, independent Boards are likely to provide stricter monitoring of managerial behavior with respect to earnings management, which could have a discouraging effect.

Nevertheless, the exclusion of inside directors can also harm the Board effectiveness. For example, it can negatively impact the formulation and execution of corporate strategies and weaken the effectiveness of the monitoring. Since, outsiders of the firm feel less involved in the firm’s strategy. Also, there is more distance between the independent Board members and the executives of the firm. This decrease the Boards’ access to firm specific information, because such information is costly to transmit through others (Fama & Jensen, 1983; Faleye, 2015). There are studies which find that Board independence in standalone firms is not or positively related with earnings management. Agrawal and Chadha (2005) find that the probability of a company managing earnings, is unrelated with the independency of the Board. They state that the Board has a variety of other issues on its agenda besides overseeing the financial reporting of the firm (e.g., hiring, compensation and overseeing the business strategy). So it is possible that even a competent and independent Board may fail to detect

13 accounting problems (Agrawal & Chadha, 2005). Also other researchers found, that the increase of Board independence itself does not reduce earnings management in standalone firms. Chen et al. (2015) argue that outside directors’ monitoring role might be hindered by their lack of information. When the effectiveness of monitoring of the financial reports is hindered by the poor information access this could result in less control over earnings management decisions.

In MNCs the financial reporting quality is important for the parent Board. Independent directors in the parent Board could have an influence on financial reporting decisions of the subsidiaries (Egbe et al., 2018; Beuselinck et al., 2019). In standalone firms Board

independence could result in stricter monitoring and managerial behavior, this can also happen in MNCs (Jaggi et al., 2009). The independent parent Board members could use the stricter monitoring to discourage earnings management in the subsidiaries. The parent Board can achieve this by their controlling role to evaluate the management of the subsidiary (Du et al., 2015). Also, the parent Board can set up rules and targets for the subsidiaries regarding discouraging earnings management.

However, the parent CEO can takes advantage of the corporate structure of MNCs, when the independent parent Board is more effective and deter earnings management at the parent level. Specifically, MNCs control many subsidiaries located in different countries around the world. This might enable the parent CEO to manipulate the earnings abroad, far away from the scrutiny of the parent Board (Warren et al., 2011; Al-khabash & Al-Thuneibat; 2008; Breton & Taffler, 1995; Watson-O’Donnell, 2000). This could result in an increase in the foreign subsidiaries.

On the other hand, the lack of information independent Board members face, could result in less effective monitoring of the financial reports of the subsidiaries (Chen et al., 2015). The independent Board members are outsiders of the firm, therefore they can have less information access than the insiders of the firm. When the parent Board is less effective in monitoring, it is hard to supervise the decisions of the foreign subsidiaries’ CEO and

management. This might allow the parent CEO to manage earnings at the parent level and this lower the incentives to manage the earnings abroad.

In sum, independent Boards are less subject to managers’ influence and are stricter in monitoring managerial behavior regarding earnings management. This can also be the case in MNCs. Financial reporting quality is an important responsibility for the Board. In MNCs quality of the financial statements of the parent firm and its subsidiaries is important, because the consolidated financial statements are drawn up on this. In order to enhance the MNCs’

14 financial reporting quality, the parent Board might uses stricter monitoring to discourage the level of earnings management in the foreign subsidiary (Chen et al., 2015; Vafeas, 2005; Jaggie et al., 2009; Klein, 2002). The parent Board can evaluate the management and CEO of the subsidiary and setting up rules and targets regarding earnings management. On the other hand, when the parent Board is effective enough to discourage earnings management at the parent level, the parent CEO has less opportunities to manage the earnings. The parent CEO can try to manipulate the earnings in the foreign subsidiaries, where the parent Board oversight is less likely to be that effective. Nevertheless, the lack of information the

independent parent Board could face, could result in less effective monitoring of the financial reporting of the subsidiaries. The parent Board cannot succeed in discouraging earnings management, when they are unable to oversee the decisions in the foreign subsidiaries (Chen et al., 2015). When there is a less effective parent Board, the parent CEO can manage the earnings at the parent level (and there is no need to do this abroad). The conflicting arguments lead to the following first hypothesis:

Hypothesis 1: There is an association between parent Board independence and level of earnings management in the subsidiary.

Parent Board Gender diversity and subsidiaries financial reporting quality

As mentioned before, the parent Board composition may have an influence on the level of earnings management of the subsidiary. An aspect of the Board composition has received growing attention, which is gender diversity (Damak, 2018). There are differences between the Board members in leadership styles, communications skills, risk aversion and

conservatism due to personal characteristics. Males and females have different cognitive frames, which affect their human behavior. Researchers and firms want to know how this affect the firm. (Qi & Tian, 2012; Post & Bryon, 2015; Damak, 2018).

Board gender diversity generally refers to the proportion of female directors within the Board. The better distributed the Board is, the more effective the Board is seen. The perfect distribution will be 50/50, here all the personal characteristics of males and females are represented. Besides, gender diversity is considered as a key attribute of good corporate governance (Smith et al., 2006; Damak, 2018; Carter, Simkins, & Simpson, 2003).Female members are more efficient in obtaining information, which can reduce information

asymmetry between Board and managers (Arun, Almahrog & Aribi, 2015). Obtaining more information can improve the communication and Board discussions. The improvement in communication and discussion might improve the monitoring ability of the Board (Terjesen et

15 al., 2009; García-Sánchez et al., 2017; Srinidhi, Gul, & Tsui, 2011; Fan et al., 2019). Also, females are less likely than males to take risks, especially in financial aspects, and to engage in unethical behavior (Pan & Sparks, 2012; Gul, Fung & Jaggi, 2009; Damak, 2018).

However, there are studies that state the opposite. Female directors provide more and faster their opinions and critical questions than male directors. As a result, there are conflict and tensions within the board, this might be time consuming. This can contribute to less effective monitoring and slow decision making (Smith et al., 2006; Hambrick, Cho, & Chen, 1996).

Prior literature suggest that gender diversity in Boards improve the financial reporting quality of the standalone firm (Srinidhi et al., 2011; Orazalin, 2019; Alshaer & Zaman, 2016). These studies showed that firms with higher level of female representation on the Board are less likely to manipulate earnings. Executives who are monitored by diverse Board are less likely to engage in earnings management. Due the diversity in opinions and backgrounds they are better in monitoring and controlling activities (Albaum & Peterson, 2006; Forte, 2004; Damak, 2018; Post, Rahman, & Rubow, 2011). Also, the higher risk aversion of female directors and their ethical behavior could result in discouraging earnings management (Gul et al., 2009; Pan & Sparks, 2012). So, gender diversity can play a role in the level of earnings management.

As mentioned before, parent Boards can have an influence in the financial reporting decisions of the subsidiaries (Egbe et al., 2018; Beuselinck et al., 2019). When the parent Board is more gender diverse, this could influence the subsidiaries’ decisions. The parent Board aiming to improve the financial reporting quality of the MNC. It is important to have a high quality in the parents’ and subsidiaries’ reports. A gender diverse Board can better control and monitor the level of earnings management in the subsidiaries (Terjesen et al., 2009; Srinidhi et al., 2011). In the gender diverse parent Board there is less information asymmetry and better communication, this can result in better monitoring and controlling activities (Arun et al., 2015). The parent Board can use this when evaluating the management of the subsidiary regarding earnings management. The parent Board can set up rules and targets and monitor this strictly in the subsidiary by controlling if these are achieved. As a result of higher risk aversion and more ethical behaviour of female members, the parent Board can be more motivated to discourage earnings management in the subsidiaries (Gul et al., 2009; Pan & Sparks, 2012; Damak, 2018). The performance of the parent Board is, among other things, judged on the quality of the MNCs’ consolidated statements.

Nevertheless, when the parent Board is more effective this can increase the level of earnings management of the subsidiaries. When it is hard for the parent CEO to engage

16 earnings management at the parent level, the parent CEO can take advantage of the corporate structure of the MNC. Specifically, the many foreign subsidiaries might enable the parent CEO to manipulate the earnings abroad, this is far away from the scrutiny of the parent Board (Warren et al., 2011; Al-khabash & Al-Thuneibat; 2008; Breton & Taffler, 1995; Watson-O’Donnell, 2000). On the other hand, the increase in conflicts and tensions can result in less effective monitoring of the financial reporting decisions of the subsidiaries (Smith et al., 2006; Hambrick et al., 1996). When there is less effective monitoring it is hard to supervise correctly the parent firm. Therefore, it is more easy for the parent CEO to engage earnings management at the parent level and the parent CEO might not need to do this abroad.

In sum, a more gender diverse Board might be better in controlling and monitoring the financial reporting quality, because of the better communication and risk aversive behavior of female directors (Damak, 2018). This can also be the case in MNCs. The parent Board can use the better monitoring and controlling when evaluating the management of the subsidiary regarding earnings management. The parent Board can set up rules and targets and monitor this strictly in the subsidiary. Also, the parent Board can be more motivated to discourage earnings management in the subsidiaries, as a result of the higher risk aversion and more ethical behavior of the female members (Gul et al., 2009; Pan & Sparks, 2012; Damak, 2018). However, when the parent Board is more effective it is difficult for the parent CEO to manage earnings at the parent level. The parent CEO will try to engage earnings management in the foreign subsidiary, far away from the scrutiny of the parent Board. This could result in an association between parent Board gender diversity and level of earnings management in the foreign subsidiary. Also, in gender diverse parent Boards there are more opinions and conflicts, which could result in less effective monitoring (Smith et al., 2006). When there is less effective monitoring it might be easier for the parent CEO to manage the earnings at the parent level, instead in the foreign subsidiary. The conflicting arguments lead to the following second hypothesis:

Hypothesis 2: There is an association between parent Board gender diversity and level of earnings management in the subsidiary.

17 Figure 1: Conceptual model

METHODOLOGY

To test the hypotheses, first all listed parent firms from the US are identified by using Compustat. The firms in the banking- and financial sector are discarded, because of the different regulation and capital structure requirements for these firms (Aggarwal, Jindal, & Seth, 2019). Also, firms with missing Board related information are excluded. Information about the parent Boards is gathered from the MSCI database. The IDs of the parent firms are used to obtain data about Board gender diversity and independence. The MSCI Board information is obtained for the period 2012 till 2017. For these listed non-financial parent firms, the names and jurisdiction of all material subsidiaries, included in exhibit 21.1 or 21 of 10-K filling, are hand collected from Edgar. The period for the hand collection is 2012 till 2017. Hereafter, the names and countries of the material subsidiaries are matched with those of the firms covered by the Orbis database. Given the coverage of Orbis database, only subsidiaries from 23 EU countries are selected. This is done for the period 2012 till 2017. Thereafter, subsidiaries operating in the financial- and banking sector are deleted because of their different regulation and capital structure requirements (Aggarwal et al., 2019). Finally, the financial information of the subsidiaries is collected from the Orbis database. After discarding subsidiaries with unavailable financial information, the final sample consists of 5298 observations of 883 unique firms.

Variables

The dependent variable in this study is the level of earnings management in the subsidiary. Following previous literature, discretionary accruals is used as proxy for the level of earnings management (Dechow et al., 2010; Bartov et al., 2001). The Jones (1991) model is employed for the purpose of estimating the discretionary accruals (DACC_SUB), this model is widely

18 used in the literature (Dechow et al., 2010). The Jones model proposes a model that basically splits total accruals in discretionary and nondiscretionary accruals. The model attempts to control for the effect of changes in a firm’s economic circumstances on nondiscretionary accruals (Dechow, Sloan, & Sweeney, 1995). The Jones (1991) model is presented in equation (1), where total accruals are regressed (TACC_SUB) cross-sectional. The total accruals are regressed on the change in sales of the subsidiaries (SALES_SUB) and the level of property, plant and equipment (PPE_SUB). Then, all the variables are scaled by total assets (TA_SUB). In model: 𝑇𝐴𝐶𝐶_𝑆𝑈𝐵𝑖,𝑡 𝑇𝐴_𝑆𝑈𝐵𝑖,𝑡−1 = 𝛽0 1 𝑇𝐴_𝑆𝑈𝐵𝑖,𝑡−1+ 𝛽1 ∆𝑆𝐴𝐿𝐸𝑆_𝑆𝑈𝐵𝑖,𝑡 𝑇𝐴_𝑆𝑈𝐵𝑖,𝑡−1 + 𝛽2 𝑃𝑃𝐸_𝑆𝑈𝐵𝑖,𝑡 𝑇𝐴_𝑆𝑈𝐵𝑖,𝑡−1+ 𝜀𝑖,𝑡 (1) Where:

TACC_SUB = Total accruals of the subsidiary

TA_SUB = Total assets of the subsidiary in year t-1 for company i

ΔSALES_SUB = Change in sales of the subsidiary in year t for company i

PPE_SUB = Property, plant and equipment of the subsidiary in year t for company i

ε = Error term (or residual)

In order to compute the TACC_SUB, equation (2) is used. To calculate the total accruals, the change in current liabilities (CL_SUB) and the depreciation expenses (DEP_SUB) need to be deducted of change in non-cash current assets (NCA_SUB). So, to be able to estimate the total accruals the following equitation is used:

𝑇𝐴𝐶𝐶_𝑆𝑈𝐵𝑖,𝑡 = ∆𝑁𝐶𝐴_𝑆𝑈𝐵𝑖,𝑡− ∆𝐶𝐿_𝑆𝑈𝐵𝑖,𝑡− 𝐷𝐸𝑃_𝑆𝑈𝐵𝑖,𝑡 (2)

Where:

ΔNCA_SUB = Change in non-cash current assets of the subsidiary in year t for company i

ΔCL_SUB = Change in current liabilities of the subsidiary in year t for company i

19 The Jones model is estimated for each year and industry combination. DACC_SUB is the absolute value of the existence of the association. Where, the higher the value, the greater the manipulation of the accruals.

The independent variables in this study are the parent Board independence (B_INDEP_PAR) and Board gender diversity (B_GENDER_PAR). Consistent with prior literature, the directors are classified as insiders or outsiders of the firm (Klein, 2002; Jaggi et al., 2009). Insiders are current employees of the firm or past employees, relatives of the CEO or other managers, or have significant business relationship with the firm (Klein, 2002). Outsiders have no ties to the firm beyond being a Board member. The B_INDEP_PAR is the ratio of the number of outside directors to the total numbers of parent Board. To determine the gender diversity, Board members can be classified as male and female members. So, the B_GENDER_PAR is the ratio of the number of female directors to the total number of parent Board directors (Damak, 2018; Orazalin, 2019).

Empirical model

To investigate whether parent Board independence and gender diversity influence subsidiaries’ earnings management, the following empirical model is used:

𝐷𝐴𝐶𝐶_𝑆𝑈𝐵_𝑖,𝑡 = 𝛽₀+ 𝛽₁𝐵_𝐼𝑁𝐷𝐸𝑃_𝑃𝐴𝑅_𝑖,𝑡+ 𝛽₂𝐵_𝐺𝐸𝑁𝐷𝐸𝑅_𝑃𝐴𝑅_𝑖,𝑡 + 𝛽₃𝑆𝐼𝑍𝐸_𝑆𝑈𝐵_𝑖,𝑡+

𝛽₄𝐿𝐸𝑉_𝑆𝑈𝐵_𝑖,𝑡+ 𝛽₅𝑅𝑂𝐴_𝑆𝑈𝐵_𝑖,𝑡+ 𝛽₆ 𝐶𝐹_𝑆𝑈𝐵_𝑖,𝑡+ 𝛽₇𝐿𝑂𝑆𝑆_𝑆𝑈𝐵_𝑖,𝑡+

𝛽₈𝑆𝐴𝐿𝐸𝑆_𝐺𝑅𝑂𝑊_𝑆𝑈𝐵_𝑖,𝑡+ 𝛽₉𝐴𝑈𝐷𝐼𝑇_𝑃𝐴𝑅_𝑖,𝑡+ 𝛽₁₀𝐵𝑂𝐴𝑅𝐷_𝑆𝐼𝑍𝐸_𝑃𝐴𝑅_𝑖,𝑡 +

𝛽11𝐿𝐸𝑉_𝑃𝐴𝑅𝑖,𝑡+ 𝛽12𝑅𝑂𝐴_𝑃𝐴𝑅𝑖,𝑡 + 𝛽13𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌_𝑆𝑈𝐵𝑖,𝑡 + 𝛽14𝐶𝑂𝑈𝑁𝑇𝑅𝑌_𝑆𝑈𝐵𝑖,𝑡 +

𝛽15𝑌𝐸𝐴𝑅𝑖,𝑡+ 𝜀𝑖,𝑡 (3)

This study includes control variables in order to account for the potentially confounding effects of specific firm characteristics of the subsidiary that may affect earnings management (Orazalin, 2019). First, size of the subsidiary (SIZE_SUB), measured as the natural logarithm of the total assets of the subsidiary, is included as control variable. SIZE_SUB is included as control variable, since firm size may decrease earnings management (Damak, 2018; Kim, Liu, & Rhee, 2003; Francis, Maydew, & Sparks, 1999). This is because large-sized firms are usually audited by more experienced accounting firms that could help prevent earnings misrepresentation (Francis et al., 1999). Also, larger firms could take the reputation costs into account when engaging in earnings management. Second, financial leverage of the subsidiary (LEV_SUB), this is calculated as total debt divided by total assets. LEV_SUB is inserted in

20 order to control for the leverage incentives for earnings management (Kim, Park, & Wier, 2012). Firms with high leverage could engage in earnings management for different

motivations, e.g., improve the conditions at which they secure funding, avoid debt covenant violations and raise additional debt under favorable contracting terms (Anagnostopoulou & Tsekrekos, 2017; Rodríquez-Pérez & van Hemmen, 2010; Sweeney, 1994). Also, return on assets of the subsidiary (ROA_SUB) is included to control for firm performance. Firm performance may affect managers’ incentive to engage in earnings management. Firms that performed well in the past, may choose to smooth their earnings by engaging in income increasing earnings and then reversing it in the future. Therefore ROA_SUB can be associated with the discretionary accruals (Du & Shen, 2018; Kothari, Leone, & Wasley, 2005).

ROA_SUB is calculated as net income divided by the total assets. Another variable that is added to control for firm performance, is cash flow of the subsidiary (CF_SUB). For the same reasons as for the ROA_SUB. CF_SUB is measured as cash flow divided by total assets (Beuselinck et al., 2019). Another control variable is loss of the subsidiary (LOSS_SUB). LOSS_SUB is an indicator variable equal to 1 when the subsidiary reports a loss and 0 when there is no loss. Prior research suggest that firms that have losses use large negative accruals to take a ‘big bath’ in the current period in order to report higher future earnings (Lee & Son, 2009). Therefore, a loss could have an association with earnings management and is included. Finally, there is a control variable for sales growth of the subsidiary (SALES_GROW_SUB). This is measured as the current period sales minus the prior period sales divided by the prior period sales (Jaiswall & Raman, 2019). Studies suggest that companies with higher growth rates in sales are more prone to presenting higher levels of discretionary accrual, because they present a higher margin for earnings management (Fama et al., 2016).

There are also control variables included for additional parent Board characteristics that might influence the subsidiary financial reporting quality. The first control variable is audit quality of the parent firm (AUDIT_PAR). Measured by an indicator variable equal 1 if the company is audited by a Big 4 accounting firm (EY, Deloitte, KPMG and PwC) and 0 when it is not (Orazalin, 2019). Big 4 accounting firms might provide higher quality audits. This could result in less earnings management opportunities in the parent firm (Francis & Yu, 2009). Next, the size of the Board can play a role in earnings management. Studies state that larger Boards are unmanageable and fail to function effectively (Kapoor & Goel, 2017; Yermack, 1996). The Board becomes less effective, because the coordination and process problems overwhelm the advantage from having more people to draw on (Jensen, 1993). Therefore, a control variable for Board size of the parent (BOARD_SIZE_PAR) is included.

21 Further, the leverage of the parent company is included (LEV_PAR). To control for the same reasons as LEV_SUB. LEV_PAR is calculated as the parents total debt divided by total assets. Finally, the model controls for the return on assets of the parent company (ROA_PAR). As mentioned before, this is included to control for firm performance, which may affect the incentive to engage in earnings management. ROA_PAR is measured as net income divided by the total assets. Prior studies have suggested that these specific firm characteristics are useful as control variables (e.g., Chen et al., 2015; Kim et al., 2012; Damak, 2018; Klein, 2002; Prencipe, Markarian, & Pozza, 2008).

There are also variables included to control for the different industries, countries and years in the empirical model. For the different industry of the subsidiaries (INDUSTRY_SUB) is added. To control for the different countries where the subsidiaries are located

(COUNTRY_SUB) is included. Finally year effects (YEAR) is included for the time trend. Since the dependent variable DACC_SUB is continuous, the method to estimate the model is the Ordinary Least Squared (OLS). Table 1 presents the definitions of all the variables included in the empirical model. The definitions of the variables INDUSTRY_SUB, COUNTRY_SUB and YEAR are given in Appendix A.

Table 1: Variable definitions

Variable Definition

DACC_SUB Absolute value of the discretionary accruals of the subsidiaries, measured by the Jones (1991) model.

B_INDEP_PAR Board independence of the parent Board, measured by the number of outsiders divided by the total members of the Board.

B_GENDER_PAR Board gender diversity of the parent Board, measured by the number of female directors divided by the total members of the Board.

SIZE_SUB Size of the subsidiary, determined by the natural logarithm of total assets.

LEV_SUB Leverage of the subsidiary, measured as total debt divided by total assets.

ROA_SUB Return on assets of the subsidiary, determined by net income divided by total assets.

22 Table 1: Variable definitions (continued)

Variable Definition

CF_SUB Cash flow of the subsidiary, measured by cash flow divided by total assets.

LOSS_SUB Indicator variable if the subsidiary reports a loss, equal to 1 if the subsidiary reports a loss and 0 when there is no loss.

SALES_GROW_SUB Sales growth of the subsidiary, measured by the current period of sales minus the previous period sales divided by the previous period sales.

AUDIT_PAR Audit quality of the parent firms, defined as 1 if the company is audited by a Big 4 firm and 0 when this is no Big 4 firm. BOARD_SIZE_PAR Board size of the parent firm, measured as the number of Board

members.

LEV_PAR Leverage of the parent firm, measured as total debts divided by total assets.

ROA_PAR Return on assets of the parent firm, determined by net income divided by total assets.

RESULTS

The descriptive statistics of all variables used in this research are shown in Table 2. The mean of the absolute value of discretionary accruals is 14% of the total earnings. On average, three quarters of the directors of the parent Boards are independent, while 12% are female directors. The average parent Board is composed of 24 members. Also, almost all parent firms are audited by a Big 4 company. The leverage of the parent firm is on average one quarter and the parent firm has a return on assets of 7%. When taking a look at the subsidiaries, the size of the logarithm of the assets is around 10. Next, the leverage of the subsidiaries is around 43% and the subsidiaries have a return on assets which is on average the same as the return on assets as the parent firm. Furthermore, around 10% of the subsidiaries report a loss. The table shows that the subsidiaries’ cash flow relative to the total assets is 9%. Finally, the sales growth of the subsidiaries is on average 0.

23 Table 2: Descriptive statistics

Variable N Mean Median Std. Dev. Min Max

DACC_SUB 5298 0.14 0.08 0.17 0 1.03 B_INDEP_PAR 5298 0.73 0.73 0.12 0.27 1 B_GENDER_PAR 5298 0.12 0.12 0.08 0 0.44 SIZE_SUB 5298 10.01 10.01 1.72 3.18 13.86 LEV_SUB 5298 0.43 0.40 0.26 0 1.08 ROA_SUB 5298 0.07 0.05 0.09 -0.19 0.97 LOSS_SUB 5298 0.10 0 0.30 0 1 CF_SUB 5298 0.09 0.08 0.09 -0.13 0.98 SALES_GROW_SUB 5298 0 0 0.41 -1 2.13 AUDIT_PAR 5298 0.97 1 0.18 0 1 BOARD_SIZE_PAR 5298 23.65 23 7.86 3 51 LEV_PAR 5298 0.26 0.25 0.16 0 0.98 ROA_PAR 5298 0.07 0.06 0.06 -0.76 0.42

Notes: This table reports the descriptive statistics for the final sample. The definitions of the variables are given in Table 1.

Correlations

The correlations among the variables are shown in Table 3. The coefficient of the correlation between subsidiary discretionary accruals and parent Board independence is not statistically significant. Suggesting that there is no relationship between the discretionary accruals of the subsidiaries and the proportion of outside directors in the parent Board. This is not in line with Hypothesis 1. Hypothesis 1 predicted an association between subsidiaries’ discretionary accruals and parent Board independence. The absence of correlation might indicate that there is no association.

Next, the coefficient of the correlation between subsidiary discretionary accruals and parent Board gender diversity is positively statistical significant. So, there could be a positive relationship between the level of discretionary accruals of the subsidiary and how gender diverse the parent Board is. This means a positive relationship: the more gender diverse the parent Board, the more earnings management in the subsidiary. This outcome is in line with Hypothesis 2, since the positive correlation indicates that there is an association between discretionary accruals and Board gender diversity.

24 Table 3: Correlation matrix

1 2 3 4 5 6 7 8 9 10 11 12 13 1 DACC_SUB 1 2 B_INDEP_PAR 0.02 1 3 B_GENDER_PAR 0.04 0.21 1 4 SIZE_SUB -0.14 0.07 0.13 1 5 LEV_SUB 0.31 -0.03 0.02 -0.09 1 6 ROA_SUB 0.00 0.03 -0.03 -0.09 -0.18 1 7 LOSS_SUB 0.36 -0.05 -0.05 0.00 0.15 -0.41 1 8 CF_SUB 0.18 0.03 -0.01 -0.06 -0.18 0.93 -0.36 1 9 SALES_GROW_SUB 0.01 0.07 0.05 0.10 0.05 0.07 -0.05 0.08 1 10 AUDIT_PAR -0.04 0.07 0.03 0.08 -0.01 0.02 -0.05 0.02 0.02 1 11 BOARD_SIZE_PAR -0.04 0.02 0.19 0.24 0.01 -0.01 -0.05 -0.02 -0.04 0.21 1 12 LEV_PAR -0.03 0.21 0.05 0.10 -0.06 0.04 -0.01 0.06 -0.03 0.07 0.10 1 13 ROA_PAR -0.03 -0.03 0.06 0.08 0.00 0.11 -0.07 0.10 0.02 0.13 0.03 -0.10 1 Notes: This table reports the correlation coefficients for the variables used in this study. The correlation coefficients with a significant level of 5% are shown in boldface. The definitions of the variables can be found in Table 1.

25 Also, the coefficients of the correlations indicate that the absolute value of

subsidiaries’ discretionary accruals is positively statistical significant with leverage of the subsidiary and the likelihood of reporting a loss in the subsidiary. This is indicating that the absolute value of discretionary accruals is higher in subsidiaries that are more leveraged and are more likely to report a loss. Furthermore, the absolute value of discretionary accruals is negatively statistical significant with the size of the assets of the subsidiary, audit quality of the parent firm, Board size of the parent and leverage of the parent firm. Suggesting that the absolute value of discretionary accruals is lower in subsidiaries that are larger in size of the assets, have a higher parent firm audit quality, have a higher parent Board size and have more leveraged parent firms. However, the coefficients of the correlation between subsidiary discretionary accruals and return on assets of the subsidiary and parent firm, cash flow of the subsidiary and sales growth of the subsidiary are not statistically significant. This suggests that there is no relationship between these variables and discretionary accruals of the subsidiary.

All of correlation coefficients for subsidiary discretionary accruals are relatively small. Since they are relatively small, it is unlikely to face multicollinearity issues in the analyses.

Regression

The regression analysis for model (3) is shown in Table 4. In all the columns the dependent variable is DACC_SUB. The first column (1) displays the results for only the control variables. The second column (2) displays the results for the control variables and

B_INDEP_PAR. Next, the third column (3) shows the results for the control variables and B_GENDER_PAR. Finally, the fourth column (4) displays the results for the complete model. The results of the full model (column 4) suggest that the Board independence of the parent firm is not significantly associated with discretionary accruals of the subsidiaries (p>0.10). Therefore, Hypothesis 1 could not be supported by this result. So there is no relationship between parent Board independence and discretionary accruals of the subsidiary. This could indicate that independent parent Boards might not restrict earnings management in MNCs, since they seem unable to curb earnings management in foreign subsidiaries. The reason for this might be that the absence of proximity can make it difficult for the parent Board to supervise the decisions of the CEO and management in the subsidiaries (Watson-O’Donnell, 2000). Another explanation could be that, independent directors are outsiders of the firm, their monitoring role could therefore be hindered by lack of information. When there is a lack of information, the parent Board might not know exactly what decisions the subsidiaries make.

26 Also, the other issues on the parent Board agendas, besides overseeing the financial reporting quality, can result in less effective monitoring. These reasons could result in less effective monitoring of the subsidiaries’ financial reports and an absence of the relationship.

However, gender diversity of the parent Board is positively and significant associated with discretionary accruals of the subsidiaries (p<0.01). In terms of economic significance, the magnitude of the coefficient of 0.09 suggests that every one percentage point increase in the gender diversity, increases the discretionary accruals with 0.09. This result provides support to Hypothesis 2 ,which predicted that there is an association between parent Board gender diversity and discretionary accruals in the subsidiaries. The positive association is not consistent with the general view on gender diversity. Most studies state that there is a negative association (e.g., Srinidhi et al., 2011; Orazalin, 2019; Alshaer & Zaman, 2016). This can indicate that the parent CEO is trying to engage earnings management in the foreign subsidiaries, because the gender diverse parent Board is more effectively in monitoring the managed earnings at the parent level.

Based on the results for the control variables it can be observed that the leverage of the subsidiary and the subsidiaries’ cash flow are positively and significantly associated with the discretionary accruals of the subsidiary. This indicates that the more leveraged the subsidiary, the more earnings management there is. Also, the more profitable the subsidiary, measured as cash flow, the more earnings management is shown. A significant and negative association is found for the size of the assets of the subsidiary, the return on assets of the subsidiary and the audit quality of the parent firm with discretionary accruals of the subsidiary. Meaning that the larger the size of the assets of the subsidiary, the lower the level of earnings management. Also, when the return on assets of the subsidiary is increasing, the level of earnings

management is decreasing. Finally, when the audit quality of the firm is higher, the level of earnings management is lower.

The coefficients of the full model (column 4) are similar to the results in column 1, column 2 and column 3. To see if these results also hold with other measures of the variables a robustness test is conducted in Appendix B.

27 Table 4: Regression analysis

(1) (2) (3) (4) B_INDEP_PAR - 0.00 (0.774) - 0.01 (0.676) B_GENDER_PAR - - 0.09*** (0.004) 0.09*** (0.004) SIZE_SUB 0.00*** (0.000) 0.00*** (0.000) -0.01*** (0.000) -0.01*** (0.000) LEV_SUB 0.21*** (0.000) 0.21*** (0.000) 0.21*** (0.000) 0.21*** (0.000) ROA_SUB -0.21*** (0.002) -0.22*** (0.002) -0.21*** (0.003) -0.21*** (0.003) LOSS_SUB 0.00 (0.935) 0.00 (0.937) 0.00 (0.883) 0.00 (0.884) CF_SUB 0.31*** (0.000) 0.31*** (0.000) 0.30*** (0.000) 0.32*** (0.000) SALES_GROW_SUB 0.00 (0.497) 0.00 (0.493) 0.00 (0.468) 0.00 (0.463) AUDIT_PAR -0.02* (0.062) -0.02* (0.060) -0.02* (0.075) -0.02* (0.073) BOARD_SIZE_PAR 0.00 (0.967) 0.00 (0.979) 0.00 (0.537) 0.00 (0.484) LEV_PAR 0.00 (0.676) -0.01 (0.657) 0.00 (0.805) 0.00 (0.775) ROA_PAR -0.02 (0.616) -0.02 (0.602) -0.03 (0.464) -0.03 (0.445) Observations 5298 5298 5298 5298 R-squared 0.13 0.13 0.14 0.14

Notes: This table reports the results of the estimation of model (3). The definitions of the variables are given in Table 1. In (1), (2), (3) and (4) the dependent variable is DACC_SUB. The first row for each variable is the coefficient and the T-statistics are reported between parentheses. Year, industry and country fixed effects are included but not reported. The *, **, ***, denote statistically

28 DISCUSSION

Due to the influence of earnings management on the financial reports of the firms, it has been a popular topic of research (Wilson, 2011). The information in the financial reports must give a reliable and relevant representation, so the users can rely on the information in order to make decisions (IFRS, 2018). When companies manage earnings to mislead the users about the performance, this could give an incorrect view of the organization and users may suffer wealth losses (Healy & Wahlen, 1999). Understanding the factors likely to motivate and constrain firms’ earnings management activities is therefore of interest to researchers (Wilson, 2011). Monitoring the quality of the financial reports is one of the main responsibilities of the Board, therefore Boards can play a crucial role in earnings management (Board &

Administrator, 2018; Baria & Cadez, 2018; Lo et al., 2010). It is likely, that in order to enhance the quality, the Board tries to discourage earnings management (Van Tendeloo & Vanstraelen, 2011; Herath & Albarqi, 2017; Davidson et al., 2005). Several studies found evidence that Board effectiveness play a role in the level of earnings management (Barja & Cadez, 2018; Lo et al., 2010; Dechow et al., 2010; Davidson, Goodwin-Stewart, & Kent, 2005). Board effectiveness is determined by Board composition, in which characteristics as Board independence and gender diversity are considered most relevant.

In the earnings management literature there is a conflicting view between Board independence and earnings management (Klein, 2002; Vafeas, 2005; Agrawal & Chadha, 2005; Chen, Chang, & Wang, 2015). Studies in standalone firms found that independent Board members are better monitors. They are less subjected to influences from the managers and can make more independent decisions (Klein, 2002; Osma, 2008; Vafeas, 2005). In MNCs the parent Board might want to discourage earnings management in all the reports that are used for the consolidated statements. The independent Board can use its superior

monitoring skills, to discourage earnings management at subsidiary level. However, better monitoring of the independent parent Board can also lead to an increase in the subsidiaries’ level of earnings management. If independent parent Boards deter earnings management at the parent level, it is likely that the parent CEO takes advantage of the corporate structure and choose to manage the earnings in a foreign subsidiary.

In the earnings management literature, there is also a conflicting view between Board gender diversity and earnings management (e.g., Orazalin, 2019; Terjesen, Sealy, & Singh, 2009; Damak, 2018; Albaum & Peterson, 2006). The female members in a gender diverse Board are more ethical in their behavior and professional judgement. The female members also improve the communication and monitoring ability within the Board. This can result in

29 less earnings management (García-Sánchez, Martinez-Ferrero & García-Meca, 2017; Fan, Jiang, Zhang, & Zhou, 2019; Orazalin, 2019; Damak, 2018). When the parent Board is better in communication and monitoring they can set up rules and targets regarding earnings

management in the subsidiaries. On the other hand, when they monitor more effective at the parent level, the parent CEO might manage the earnings in the foreign subsidiary.

Seeing that there is a conflicting view in the literature to these two characteristics, it is important to study the association with earnings management. When the parent Board wants to improve the financial reporting quality, they need to know how the characteristics and composition can play a role in benefitting the quality. Therefore, the research question of this study is: To what extent does the MNC Board independence and gender diversity affect the level of earnings management of the foreign subsidiary?

This study focus on MNCs which have foreign subsidiaries (Beurs FD, 2020). Different studies suggest that MNCs have substantial power over the subsidiaries and influence their (financial) decisions (Egbe et al., 2018; Robinson & Stocken, 2013;

Beuselinck et al., 2019). The main responsibility of the parent Board is to improve the quality of the financial reports of the MNC. It is presumable that the parent Board uses the influence to achieve improvement of the financial reporting quality of the subsidiaries. However, the parent CEOs may use the foreign subsidiary for managing earnings. Most of the MNCs have a lot of subsidiaries, so it is not possible for the parent Board to monitor each individual

subsidiary. Also, the absence of proximity makes it difficult for the parent Board to supervise the decisions of the CEO and management in the foreign subsidiary. Also, the parent CEO might engage earnings management at the subsidiary level. It is unlikely for the parent Board oversight to be that effective (Warren et al., 2011; Al-khabash & Al-Thuneibat; 2008; Breton & Taffler, 1995). When the parent Board is more focused on reducing earnings management in the foreign subsidiary, there will be less monitoring of and control over the earnings management at the parent firm. Because the parent Board is more focused on the subsidiaries than the parent firm. In this case the parent CEO will engage earnings management at parent level. So, the parent Board may not restrict earnings management in the consolidated

statement by reducing earnings management in the foreign subsidiaries. In addition, foreign subsidiaries may have their own incentives regarding earnings management (Du et al., 2011).

In the current body of literature there is no study on parent Board characteristics and level of earnings management in the financial statements of the subsidiaries (Dechow et al., 2010). The goal of the parent Board is to improve the financial reporting quality and monitor this. It is useful for the parent Board to know if Board independence and gender diversity can

30 play a role in enhancing the quality. This research investigates if parent Board independence and gender diversity could have an influence on the level of earnings management of the subsidiaries.

In order to accomplish their responsibilities, it is important for a Board to monitor the executive management effective and unbiased. This can be accomplished the best, if the director is independent of the company (Nolan, 2005). Studies show that independent members are less subjected to the influence of managers and are therefore likely to provide stricter monitoring of managerial behavior. This could result in more control over the financial reporting decisions ( Vafeas, 2005; Chen et al., 2015; Jaggi et al., 2009). As mentioned before, the parent Board will try to enhance the financial reporting quality of the MNC and discourage earnings management. In order to achieve this, the parent Board can strictly evaluate and monitor the subsidiaries’ management decisions and set up rules and targets. However, the parent Board has a variety of other responsibilities besides financial reporting quality of the MNC. Therefore, they can lose oversight and can provide less

effective monitoring to the foreign subsidiaries. Especially the absence of proximity makes it difficult for the parent Board to supervise the decisions of the foreign subsidiaries (Watson-O’Donnell, 2000). Also, the monitoring role of independent directors might be hindered by their lack of information because they are outsiders of the firm. When the effectiveness of monitoring is hindered by the poor information access, this can also result in less control over the financial reporting decisions. Furthermore, the parent CEO can use the absence of

proximity to engage earnings management in the foreign subsidiary. The expectation was that parent Board independence is associated with the level of earnings management in the foreign subsidiary. However, this study found that there is no association between parent Board independence and the level of earnings management in the foreign subsidiary. This suggests that the parent’s independent directors might not limit earnings management practices in consolidated financial statement. Their presence does not enhance financial reporting quality of foreign subsidiaries. This might be because the absence of proximity can result in less effective monitoring and has no effect on the subsidiary. Also the lack of information independent members face, could result in less effective monitoring. Therefore, the parent CEO can engage earnings management at parent level and will not do this abroad. Another possible reason could be that the composition of the subsidiaries’ Board (if there is one) maybe play a bigger role than the composition of the parent Board. The parent Board can be less involved than the subsidiaries’ Board in this case.