The Effect of Family Ownership and Control on Dividend Policy of Publicly Listed Firms in Indonesia and Malaysia

Several recent studies show that family firms is dominating public corporations around the world, especially in emerging market (Shleifer & Vishny, 1986), (Porta, Lopez-de-Silanes, Shleifer, & Vishny, 1997), (Claessens, Djankov, Fan, & Lang, 2002). In most emerging markets, large and growing companies mostly concentrated owned by families (Porta, Lopez-deSilanes, Shleifer, & Vishny, 1997). Claessens, Djankov, & Lang (2000) defined family firms as a company with a shareholder who owns 20 per cent voting right and cash flows (Claessens, Djankov, & Lang, 2000). Prior studies proved that governance mechanisms such as dividends, debt, and board structure are significant in controlling agency problem type II which is relatable to free cash flow theory that applied in family firms (Shleifer & Vishny, 1986),(La et al., 2009). Family firms with high growth rates will have a preference using free cash flow (FCF) to expand the company and rely on external leverage. However, if the company has a surplus of free cash flow, then the free cash flow will be allocated for payment to shareholders as a dividend. Thus, this can be a positive signal to the shareholder that the family company will expand the company using the free cash flow reserve fund. Considering related regulation, only family firms with good corporate governance have a preference for generating sufficient free cash flow to pay dividends (Bhattacharya, 1979). Several reasons that support a positive influence of family ownership on dividend policy come from corporate governance literature which suggests that internal control mechanisms may need to be complementary, especially in countries that have shareholder’s weak legal protection (Al-Malkawi, Bhatti, & Magableh, 2014), (Pindado et al., 2011). The board Indonesian Capital Market Review 12 (2020) 1-11

structure with differentiating responsibilities between the executive and independent directors is called the two-tier board system (Ponnu, 2008) and applicable in Indonesia. Companies that embrace this system will be more transparent on responsibilities and rights incorporate strategies, so there is no overlapping role. Another case is that Malaysia applying one-tier board system, which is the role of the executive and independent directors is unity and replaces each other so that the executive director also has the responsibility as an independent director (Hidayat & Utama, 2003). Furthermore, Andres (2008) suggests that family-controlled firms usually generate earnings higher than those of non-family firms, therefore pay a higher dividend, especially in Europe and the US (Faccio et al. 2001).
Moreover, the association between expropriation and dividend payout is stronger in Hong Kong family firms (How et al. 2008). The reason might be that they want to ensure the minority shareholders are well-protected from the expropriation practices. Second, previous studies suggested that public companies in Southeast Asia generally owned by one shareholder or a group of shareholders (Djankov, & Lang, 2000). Therefore, these largest shareholders will have a significant role in setting dividend policy since they can put pressure on management to adopt a particular policy (Truong and Heaney, 2007).
Indonesia and Malaysia have relatively similar family business characteristics. For example, family businesses in Indonesia grew from 42 per cent in 2016 to 65 per cent in 2018. This growth is also faster than the global family business growth, which is 39 per cent compared with 16%. Meanwhile, 87% of family businesses in Indonesia and 90% of family businesses in Malaysia expected to grow in the next two years. Another similarity is that family businesses in Indonesia and Malaysia are generally owned by the second generation and operate the most in the manufacturing sector (PwC, 2018). However, the findings of the study on the effect of family ownership on dividend policy are different between Indonesia and Malaysia. For example, Benjamin, S. J. et al. (2016) found that firms with a higher level of family ownership experienced higher dividend payout, vice versa when examining the relationship between family share ownership and dividend payout in Malaysia. In contrast, studies on the effect of ownership structure on dividend policy of the listed Indonesian firms concluded that familycontrolled firms pay less dividend than those of foreign and government-controlled firms (Setiawan et al. 2015 andDuygun et al. 2018). studies inline with Yousaf et al. (2019) found that family firms in Pakistan pay lower dividends to shareholders. This finding consistent with Villalonga and Amit, 2006), Hu et al. (2007), andDe Cesari (2009), where most of the family firms expropriate minority shareholders for their-owned benefits.
Third, the placement of family members in strategic positions generally occurs in family firms (Villalonga & Amit, 2004). Family members who placed in the CEO position can minimize the company's additional costs, such as placing a professional manager. However, once the company led by its descendants, it can produce a lower stock price than founder that can bring value to the company (Anderson et al., 2003). A study on listed companies in the US suggests that the founders showed bias toward other family members if occupied as CEO positions, resulting in less optimal investment and lower profitability (Singell and Thornton, 1997).
Fourth, from 2011 to 2015, Indonesia and Malaysia are still experiencing a transitional period after the global financial crisis of 2008-2010. The growth pattern of fluctuating GDP has adjusted to the global financial condition in tackling the crisis by conducting an expansion of economic policies in each country. At the time of the financial crisis in 2009, Indonesia has a positive GDP rate despite a declining growth pattern. On the other hands, Malaysia has declining and negative growth pattern (-0.69%) but still that value was small and does not have a significant impact on the country.
Despite the contradicting findings on the impact of family ownership on dividend policy, and the different board system between Indonesia and Malaysia is attractive to examine. Be-sides, studies that directly compare those effects on dividend policy in two countries rarely conducted. Therefore, this study aims to investigate the influence of family ownership and family control companies, with the characteristics of family ownership classification following Chen, Cheung, Stouraitis, & Wong, (2005) on the dividend policy distributed to shareholders. Besides, this study is also investigating the impact of economic conditions, especially in the period of global financial crisis 2008-2010. The remaining of the paper is as follows. Section 2 consists of the relevant literature review and hypotheses development, follows by research methodology in Section 3. Section 4 describes and analyses findings from the study and concludes the study in Section 5.

Literature Review and Hypothesis Development
The agency conflicts between majority and minority shareholders found in firms with concentrated ownership, where the ultimate owners are the family members (Claessens et al., 2002) (Shleifer & Vishny, 1986) (Luo & Liu, 2014. This condition occurs because the majority of shareholders controlled the firm so that it would indicate as expropriation (Porta et al., 1997). In this case, the company's wealth given to the majority shareholders without considering the interests of others (Ronald C Anderson et al., 2018). Jensen (1986) suggests that agency problem between family as controlling shareholders and minority shareholders resulted in lower dividend policy.
Prior studies distinguished different conclusions of the influence of family ownership and control on dividend policy. Anderson et al. (2018) suggest that family firms have fewer agency problems than non-family firms. Thus, they pay dividends higher than other firms (Pindado, Requejo, & De, 2011;Yoshikawa&Rasheed, 2010;Atmaja, Tanewski&Skully,2009). On the other hand, Subramaniam & Shaiban, (2011), Wei, Wu, Li, & Chen ( 2011), Atmaja, (2017 suggest that family firms tend to be associated with more agency problems than non-family firms. Furthermore, they have more significant control mechanisms (La et al., 2009;Shleifer & Vishny,1986;Porta et al., 1997). This implies that firms with high family control pay fewer dividends than other firms, especially in countries with weak legal shareholder protection policy (Facco&Lang, 2007). Based on the above discussion, this study proposes the following hypotheses.

H1:
Family ownership and control has significant impact on dividend policy.
Driffield (2007) suggests that family-controlled firms will minimize agency conflict and therefore maximize a firm's value through capital structure decision. A study by Haron (2018) found that concentrated family ownership in emerging markets has a significant and positive impact on corporate debt financing. Furthermore, Jensen & Meckling (1976) propose that the majority of shareholders have to escalate incentive and their ability to oversee managerial performance, about leverage. Therefore, the level of debt used as the control mechanism for managers, and that concentrated family ownership in emerging markets have a significant and positive impact on corporate debt financing (Haron, 2018). An empirical study by Haron (2018) found that family firms in Indonesia and Malaysia are still dependent on debt financing. Also, Thanatawee (2012) found that a firm's leverage has a positive and significant impact on dividend policy since higher leverage will minimize agency costs. Based on the above discussion, this study proposes the following hypothesis.

H2:
The level of debt of family firms has a positive effect on dividend policy.
A study in developing countries found that the board of director's system in family firms is an essential tool to monitor management decision and lowering agency costs (Ponnu, 2008). Two board systems usually use by firms, i.e., two-tier and one-tier board system. The executive director has a direct responsibility in the management. In contrast, the independent di-rectors were responsible for monitoring and controlling the implementation of corporate practices by applicable rules and provide independent consideration to overcome misleading between directors, firm performance, and utilizing resources. In the two-tier board system, there are different responsibilities between the executive and the independent directors (Ponnu, 2008). Indonesia and Malaysia applied the different board system. The application of the two-tier board system in Indonesia will receive the benefit of more distinct responsibilities, rights, and obligations in implementation of firm strategies, so it may increase the quality of corporate governance since there are no overlapping roles (Hidayat & Utama, 2003).
In contrast, in one-tier board system which applied in Malaysia, the role of the executive and independent board is unity and replaceable to each other so that the executive director also has the responsibility as the independent director (Ponnu, 2008). Porta et al., (1997) suggest that corporate governance has the significant influence on dividend policy, where majority shareholders prefer to utilize firm's cash flows to their interest than pay dividend to minority shareholders. Therefore, we proposed that the different board system will have a different dividend policy. Thus, the hypothesis is as follows.

H3: The board system of directors of a family
firm has a significant influence on dividend policy.
The placement of family members in strategic positions such as CEO position generally occurs in family firms to minimize the additional costs incurred by the company. (Belén Villalonga & Amit, 2004). However, the expropriation for free cash flow in the family firms affects dividend policy. Firms which majority shareholders and led by its founder create additional value for all shareholders by paying the dividend. However, when a family company led by its descendants tend to have a lower value and mostly does not pay the dividend (Anderson et al., 2003). A study by Shleifer & Vishny (1986) suggests that when the founders of family firms have a role as CEOs, they will be more innovative and will enhance value for the company. Therefore, the hypothesis is: H4: There is a positive influence of founder CEOs in family firms on dividend policy.
Family members usually are the ultimate owners of the family firms and tend to have more control over management. Therefore, family firms tend to have less diversify ownership (Anderson et al., 2003 andFaccio &lang, 2007). Furthermore, they may face unexpected external shock such as the global financial crisis. Several studies show that family as majority shareholders likely to allocate more free cash flows on investment (Attig et al., 2015;and Lins et al., 2013). The hypothesis is as follows.
H5: There is a negative influence of global financial crisis on the dividend policy

Data and Sample Selection
This study uses 37 Indonesian families listed firms and 67 Malaysian families listed firmsthe data obtained from the Indonesian Stock Exchange and Bursa Malaysia. The sample firms are the company with family ownership marked by the existence block holder (control rights) with vulnerable value 0% -5% and 5% -33% (Chen et al., 2005), (Benjamin et al., 2016). They must also have a positive value of net income. It assumes that if the company has a negative net income, then the company will not pay dividends (zero dividends).
The amount of family ownership of the listed firms obtained from the annual report of each firm, while the information of the board system  Anderson et al. (2003) collected from the firm's annual report and the Bloomberg database. Besides, data on leverage, cash ratio, ROA, profitability lag, firm size, and growth, dividend payout ratio, net income, total assets are collected manually from DataStream / Thomson Reuters / Annual Report. The data of listed companies both in Indonesia and Malaysia collected from the official website of IDX and Bursa Malaysia.

Definition of Variables
The dependent variable in this study is dividend policy, measured by the dividend payout ratio. The independent variables are family ownership, the board of director system, and leverage level. This study also considers the effect of control variables, namely cash ratio, return on assets (ROA), lag profitability, firm size, and growth. Dummy variables used to take into account the crisis year of 2008 -2010 and family control.
During the global financial crisis, the stateowned enterprises (SOEs) significantly reduced their dividend payouts compared to those of private firms. However, private firms have to adjust their expenses to realize a similar dividend payout. Besides, profitability is of essen-tial determinants of dividend payout (Fama and French, 2001;Ankudinov and Lebedev, 2016). Furthermore, A study by Floyd et al. (2015) found that companies prefer to lessen their stock repurchase rather than to reduce dividend payment. Baker et al. (2001) suggested that previous dividend pattern, earning stability, and the level of current and predicted future earnings are relevant factors of firms' dividend policy. Therefore, we include the interaction of return on asset (ROA) and crisis year to check the impact of ROA on dividend policy in the crisis periods. Table 1 presents the variables used in this study, as well as their measures and the sources of citation.

Model Specification
This study conducted using Generalized Least Square (GLS) estimation. The first model is used to test the influence of family ownership and family control on the dividend payout 5

Empirical Results
This Section consists of the description of data employed in this study and empirical analysis of the results from the estimation models. Table 3 presents the statistics of the data used in this study. It consists of the mean value, standard deviation and minimum and maximum values.

Descriptive Statistics
According to Table 3, it shows that the mean value of the dividend payout ratio is 0.0189, with a standard deviation of 0.0348. There are several family companies in Indonesia and Malaysia, paying dividends despite having nega-tive earnings. This phenomenon indicates that the family firm has a constant dividend payout rate to preserve corporate image or reputation. However, some firms did not pay the dividend in Malaysia. (Sapura Industrial BHD, Brem Holdings BHD, Hup Seng Industries BHD, Salcon BHD, Sapura Resource BHD. It happened in 2005 due to Malaysia's newly imposed freefloating system on ringgit and an evaluation of the capital market policy by imposing shortselling bans that could create various sentiments from shareholders. This causes the family companies listed on Bursa Malaysia to be reluctant to pay dividends (Source: Financial Times: "Malaysia Relaxed Short-Selling Ban").  DPR i,j,t = α+β 1 FamOwn i,j,t -β 2 Lev i,j,t +β 3 Size i,j,t +β 4 Profit i,j,t +β 5 Cash i,j,t +β 6 Growth TA i,j,t +β 7 Dcontrol5 i,j,t +β 8 Dcontrol533 i,j,t +β 9 Dtier i,j,t +β 10 Dfounder i,j,t +β 11 lagprofitability i,j,(t-1)   conducted -the level of leverage measured by a debt ratio. The leverage has a mean value of 1.and standard deviation value of 2.6711 indicates that family firms in Indonesia and Malaysia still prioritize external funding of companies derived from debt so that the debt level still lacks the criteria as a tool for a family firm to take expropriation measures that caused agency problem. If the family wants to expand its firm, the firm will use free cash flow (FCF) to repay the debt to external parties and can lower the dividend rate distributed to the shareholders. Table 4 presents the empirical results of the two models using the generalized least squares method. Results in Table 4 show that both models generate similar indications. It shows that there is a positive and significant influence of family ownership on the dividend payout ratio. This finding is consistent with (Setia- Atmaja et al., 2009, andYoshikawa &Rasheed, 2010), (Setianto &Sari, 2017, andPindado et al., 2011). According to agency theory Vishny, 1986, andClaessens et al., 2002), the presence of family ownership as shareholders has a high degree of exponential indication in exploiting the company's free cash flow (FCF) for personal purposes and prioritizing share-holder interests held by the family. However, this study finds that there is a positive relationship between family ownership with the level of dividend payout ratio, which is contrary to the evidence of agency theory. This result indicates that the higher the percentage of family ownership in Indonesia and Malaysia as majority shareholders, the higher the degree of firm' disclosure and a lower degree of expropriation by the majority shareholders.

Results
Generally, investors believe that a bird in the hand is worth two birds in the bush (Setia- Atmaja et al., 2009). It indicates that investors prefer cash dividend payments rather than capital gains for the sake of dividend taxes. The minority shareholder in the family firms more appreciates dividend than capital gain in the future. They also assume that retained earnings invested in new projects always contain risks and can be misused or invested in speculative projects which may create negative returns. Another reason is that dividend payout regularly disciplines the management so that dividend is used as a policy to improve the firm's corporate governance (Ismail, Haron, and Idayu, 2016).
The family firms in Indonesia and Malaysia have a preference to allocate free cash flow (FCF) as debt payments if the family company is in an expansionary phase so that firms would reduce dividend payouts. This case is 7  Mulyani et al., (2016), Arifin, (2003, and Anderson et al., (2018). Implementing the two-tier system would be better for reducing agency problems and expropriation action issues by majority shareholders in family firms with separation of ownership and control in firms (Gillmore Doppia, 2016). However, the results of this paper show that the implementation of two-tier in Indonesia will hurt the level of dividends paid. This finding indicates that the two-tier system that applied in Indonesia has not effectively reduced the agency problem in the family firms so that the index of corporate governance companies in Indonesia is still below the average compared to other companies in ASEAN countries.
This study also finds that the 2008-2010 global financial crisis in Indonesia and Malaysia has no significant impact on the stability of dividends paid. This result is with a study conducted in Oman (Al-Malkawi, et al., 2014). As reported in www.cnnindonesia.com, the 2008-2010 global financial crisis did not have a high volatility shock even in 2008 when the Fed decided to raise interest rates by 0.25 points indicating capital outflows from Indonesia and Malaysia to the US. Furthermore, although the firms increased the dividend payment during crisis periods, the effect is not significant. This indicates that family firms choose to pay more dividend to minimize agency problems with minority shareholders.

Robustness Tests
In order to check the consistency of the model, we conducted robustness test by changing the measurement of dependent variable, i.e: dividend payout ratio using cash dividend divided by net income before extra items (La Porta, 2009), (Mulyani et al., 2016). We also used generalized least square models and found somewhat weaker but very similar results on the influence of family ownership, family control, and leverage on dividend payout ratio. Results in Table 5 exhibit a similar effect of family ownership, family control, and leverage on dividend policy of Indonesian and Malaysian firms. Hence, our findings were consistent with the basic model (See Table 4).

Conclusions and Limitation
This study aims to investigate the influence of family ownership and family control on a firm's dividend policy in Indonesia and Malaysia. It found that the higher the percentage of family ownership the greater the dividend payout ratio. In addition, the same effect applies to the percentage of family control. In contrast, we found that firms with higher debt levels will re-8  (Attig, Boubakri, El, & Guedhami, 2015). The findings have two essential implications. For policymakers, the indications of potential expropriation could provide consideration to encourage more transparency, investor protection, and overall good governance practices. For shareholders, the findings indicate that investments in family firms that distribute positive dividends or family firms that seem to mitigate agency conflicts through the use of debt or dividend are reasonable This study included classification percentage of family ownership at level 0% -5%, 5%-33%, and >33% in Indonesia and Malaysia's companies. Dividend payment (dividend payout ratio) is a way to reduce agency problems between shareholder and management level. The percentage of family ownership at 5%-33% is more concentrated so that the minority shareholders will giving much concern into uses of free cash flow using dividend policy to minimize agency problem and enhancing corporate's reputation. However, due to the ambiguity of the beneficiaries in the emerging countries, the effect of ownership structure might not be concluded as shown in the results. The unclear structure may cause obscure conclusion and be a limitation of this study 9 39 (2013)