The effectiveness of the internal corporate governance mechanism and the ownership of the gov- ernment and agencies

Article history: Received: February 27, 2021 Received in revised format: April 29 2021 Accepted: May 6, 2021 Available online: May 6, 2021 This paper examines the impact of the government and its agencies’ ownership on the effectiveness of one the main internal governance mechanisms, namely; board of directors, for a sample of 140 energy and petrochemical Saudi listed firms over 2012-2019. The Saudi Arabia provides an interesting context due to the domination of government-linked corporations’ ownership. This setting arranges for the impact of such ownership on the board of directors’ monitoring and advisory roles. The board of directors’ effectiveness is measured as an interaction term of the board size and meetings of the board of directors. The study finds that government-linked energy and petrochemical corporations’ ownerships are inversely related to the board of directors’ effectiveness. This result is sensitive to the measurement of the board of directors’ effectiveness as each variable consisting of the board of directors’ effectiveness was examined individually. The study also finds that government-linked corporations’ ownership had a strong negative impact on the board size. In contrast, the proposed model does not provide any evidence supporting the relationship of the governmentlinked corporations’ ownerships with board meetings. Overall, the evidence supports the substitution hypothesis on the relationship of government-linked corporations and board of directors’ effectiveness. © 2021 by the authors; licensee Growing Science, Canada

This paper examines the impact of the government and its agencies' ownership on the effectiveness of one the main internal governance mechanisms, namely; board of directors, for a sample of 140 energy and petrochemical Saudi listed firms over 2012-2019. The Saudi Arabia provides an interesting context due to the domination of government-linked corporations' ownership. This setting arranges for the impact of such ownership on the board of directors' monitoring and advisory roles. The board of directors' effectiveness is measured as an interaction term of the board size and meetings of the board of directors. The study finds that government-linked energy and petrochemical corporations' ownerships are inversely related to the board of directors' effectiveness. This result is sensitive to the measurement of the board of directors' effectiveness as each variable consisting of the board of directors' effectiveness was examined individually. The study also finds that government-linked corporations' ownership had a strong negative impact on the board size. In contrast, the proposed model does not provide any evidence supporting the relationship of the governmentlinked corporations' ownerships with board meetings. Overall, the evidence supports the substitution hypothesis on the relationship of government-linked corporations and board of directors' effectiveness.

Introduction
Ownership structures are an essential element of governance, particularly when a well-designed legal framework is not present. Ownership structures are becoming increasingly popular, with numerous organizations finding them a viable alternative (Eng & Mak, 2003;Barako et al., 2006;Ntim & Soobaroyen, 2013). Dominant internal shareholders frequently exploit the advantages of private control, which gives them insider access to crucial data (Shleifer & Vishny, 1986;Dyck & Zingales, 2004). Internal shareholders have the capacity of controlling and influencing directors to defend their own interest over those of other parties (Coffee, 1991;Shleifer & Vishny, 1986). Dominant controlling shareholders are not incentivized to promote board control or audit committees, since the more intensely the environment is monitored, the more expensive it will be for them to leverage the situation for private gain (Leftwich et al., 1981;Shleifer & Vishny, 1997). Owing to this, these shareholders become a controlling force in the company and employ their influence over directors for the advancement of their own interest (Mendez & Garcia, 2007). Greco (2010) suggests that ownership in the hand of insiders has a negative influence on the frequency of board and/or audit committee meetings. Specifically, ownership structures are generally regarded as being a central influence on the ways directors behave. The ownership structure of the companies headquartered in the Gulf Corporation Council (GCC) area has been alleged to be a primary source of corporate governance difficulties in the region. Niemi (2005) proposes that there should be a holistic debate regarding ownership, suggesting that a focus on concentration ownership with no examination of alternatives could cause erroneous assumptions to be made about the part various owners play.
In Saudi Arabia, concentrated ownership of companies is quite common as the majority of listed companies are owned by the government (Al-Janadi, Rahman & Alazzani, 2016). Baydoun and Willett (2006) state that firms based in the GCC are generally owned by the government. The government of Saudi Arabia owns a huge majority of the shares in the publicly listed firms in the country. This massive government share in public companies is attributable to the fact that the government tends to privatize companies. Additionally, the Treasury of the Saudi Arabian government frequently invests in listed companies, increasing government shareholdings in such companies further. (Al-Janadi et al., 2016;Baydoun and Willett, 2006). The majority of companies in government ownership and control are generally sensitive politically because their operations are subject to public scrutiny. It is highly expected that these organizations should be conscientious in their operations (Ghazali, 2007). Some research findings report negative issues regarding government-owned companies in terms of performance (Megginson and Netter, 2001;Shleifer, 1998), poor reporting of conservative earnings (Bushman & Piotroski, 2006), a lack of proper transparency (Bushman et al., 2004), pay back the voters for their support, for contributing politically, and for their bribery (Rajan & Zingales, 2003;Shleifer & Vishny, 1993, 1994La Porta et al., 2002), voluntary disclosure (Al-Janadi et al., 2016), the focus on implementing governmental objectives at the expense of shareholders' objectives (Firth & Rui, 2007), disclose earnings with lower quality in their report (Chaney, Faccio & Parsley, 2011), acquire the property of minority investors for public use (La Porta et al., 2002;Shleifer & Vishny, 1997) and offer financial insurance to outside shareholders in exchange for political and financial support (Wang et al., 2008), and less modified opinions reported (Chen et al., 2000).
Government control over a board of directors is regarded as a strategy employed by the state for protection of its own interests and property values. The majority of the board members that represent the government receive recompense from local government dependent on their administrative rank rather than the quality of their performance (Xu & Wang, 1999;Zhou & Wang, 2000). They may not have sufficient management skills to be effective supervisors of a variety of management behaviors. In addition, the priorities of local government are not necessarily the same as that of the national government. It should be noted that government influence on the board of directors frequently involves interference in management structures. The government may at times demand a move away from the pursuit of maximum profit, focusing instead on developing infrastructure and maximizing social benefits. Empirically, some prior studies investigated individually the characteristics of board meeting in different settings Aljaaidi, Bagai & AlAbdullatif, 2021;Greco, 2010;Vafeas, 1999;Hahn & Lasfer, 2007;Baccouche and Omri, 2014;Hahn & Lasfer, 2016;Lin et al., 2014). Regarding the characteristics of board of directors and the ownership, Greco (2010) looked at the correlations between insider ownership/ownership concentrations and board meeting frequency. This research revealed a negative correlation between insider ownership and board meeting frequency. Al-Janadi, Rahman & Alazzani (2016) looked at the moderating influence of the ownership of corporations linked with government and the correlation of voluntary disclosure and corporate governance. The research showed that the ownership of corporations linked with the government has a negative influence in the moderation of the correlation between corporate government elements, i.e., voluntary disclosure, board size, and non-executive directors. This indicates that corporations linked to the government are negatively impacted in terms of effective corporate governance. It's important to note that existing research regarding board of directors' discipline does not take into account the degree to which government ownership of a corporation limits the power of its board of directors. This research represents an attempt to make good this lack in previous research. Additionally, this research measures a company board of directors' power employing composite measures of frequency of board meetings and size of board. The outcomes demonstrate that measuring the overall impact of a combination of factors shows stronger influences than measuring individual influences (Cai et al., 2009;Ward et al., 2009;Agrawal & Knoeber, 1996;O'Sullivan et al., 2008). To the best of our awareness, a study examining the impact of the government-linked corporations on the effectiveness of the board of directors does not exist, giving a contribution to the extant literature.
The paper proceeds as follows: Section 2 describes the research design and methodology. Section 3 highlights the empirical results and discussions. Section 4 provides the conclusions and implications.

Sample and Data
This study utilized a number of 140 firm-year observations for energy and petrochemical companies listed in Saudi Stock Exchange that their corporate governance and financial information are publicly available for the period 2012-2019. The data was collected by hand from the financial statements. The final sample is depicted in the following Table 1:

Empirical Results and Discussions
The descriptive statistics are illustrated in 2. These include the mean, standard deviation, minimum and maximum of each variable in the sample data set.  Table 2 shows that differences in values exist among the variables included in the model of this study. It is shown that the range of BDEFF is from 14 to 90 with an average of 42.5 and a standard deviation of 12.77067. The mean of BDSIZE is 9 with a maximum of 11 and a minimum of 6 and a standard deviation of 1.15096. The mean of BDMEET is 5 with a maximum of 10 and a minimum of 2 and a standard deviation of 1.44760. The range of GLCO is from .00 to 0.79 with an average of .14 and a standard deviation of . Pooled Ordinary-Least Square (OLS) was utilized to test the level of effect of government-linked corporations' ownership with the power of the board of directors' effectiveness. Table 3 reports that the adjusted R 2 is 0.37, indicating that the model has explained 37% of the variance in the board of directors' effectiveness. The F-value for the model is statistically significant at the 1% level, indicating that the overall model can be interpreted. This also gives an indication that the board of directors' model is a good fit. Table 3 displays that GLCO has a negative relationship with the board of directors' effectiveness BDEFF (t = -3.611, p-value < .001, one-tailed significance). This result gives support to the previous empirical studies such as (Megginson and Netter, 2001;Bushman & Piotroski, 2006;Bushman et al., 2004;Rajan & Zingales, 2003;La Porta et al., 2002;Al-Janadi et al., 2016;Firth & Rui. 2007;Chaney, Faccio & Parsley, 2011;La Porta et al., 2002;Wang et al., 2008;Chen et al., 2000). The results of this study also indicated that family ownership FOWN is negatively associated with board of directors' effectiveness BDEFF (t = -2.455, p-value < 0.020, one-tailed significance). This result is consistent with Brunninge and Nordqvist (2004). The results of this study showed that there is a positive relationship between domestic corporate ownership DOMOWN and board of directors' effectiveness BDEFF (t = 2.761, p-value < 0.010, one-tailed significance). This result is in line with several previous studies (Chibber & Majumdar, 1999;Djankov & Hoekman, 2000;Khanna & Palepu, 2000;Barbosa and Louri, 2005;Claessens et al., 2000;Dharwadkar, George & Brandes, 2000;Douma et al., 2006). The results of this study also depicted that firm size FSIZE is marginally and positively associated with board of directors' effectiveness BDEFF (t = 1.509, p-value < 0.142, one-tailed significance). This result is consistent with , Greco (2010), Hahn (2007), Brick & Chidambaran (2010), Lin, Yeh, and Yang (2014), Baccouche and Omri (2014). The results of this study also illustrated that leverage LEV has a positive association with the board of directors' effectiveness BDEFF (t = 1.938, p-value < 0.062, onetailed significance). This result is consistent with the extant research (Greco, 2010;Baccouche and Omri, 2014;Hahn and Lasfer, 2016). The results of this study reported a positive association of firm performance ROA with board of directors' effectiveness BDEFF (t = 1.846, p-value < 0.075, one-tailed significance). This results in inconsistent with the previous studies' results (Aljaaidi, 2021;Aljaaidi, Alothman, Sharam and Bagais, 2021;Vafeas, 1999;Brick and Chidambaran, 2010;Hahn, 2007;Baccouche and Omri, 2014;Hahn and Lasfer, 2016;Al-Najjar, 2012;Adams & Ferreira, 2007).

Additional Analysis
Additional analysis has been carried out to examine the extent to which our results are sensitive to the individual variables consisting the board of directors' effectiveness which are board size BDSIZE and board meetings BDMEET.  Table 4 shows that the results of the board meeting model are identical with those of the original model of the board of directors' effectiveness. This gives evidence that the results of the board meeting as an individual dependent variable are not sensitive to the results produced when using the board meeting as one of the individual variables consisting the score of board of directors' effectiveness. Table 5 depicts that the results of the board size model are different to some extent with those of the original model of the board of directors' effectiveness. This gives evidence that the results of the board size as an individual dependent variable are sensitive to some extent compared with those produced when using the board size as one of the individual variables consisting the score of board of directors' effectiveness. The difference in the results were reported for the government ownership as its relationship was negative and significant in the original model, board of directors' effectiveness. While in the board size model the result was insignificant (t = -1.181, p-value < 0.247, one-tailed significance).

Conclusions and Implications
Using a sample of 140 firm-year observations from energy and petrochemical industries, this study has examined the impact of the government-linked corporations' ownership on the power of their board of directors in Saudi Arabia. The setting of Saudi Arabia provides an interesting institutional setting for this issue due to the domination of the government and its agencies on companies. Therefore, the negative impact of the government-linked local corporations is likely to reflect the malfunction of their board of directors' advisory and monitoring roles. This might give an indication that the government focuses on the economic, social and other purposes rather than the profitability purpose. Our key conclusion is that the ownership of the government-linked local corporations impacts negatively on its board of directors. This result is sensitive to the measurement of the board of directors used in the study which is the interaction term of the board size and meetings. As the result indicated a negative association between the ownership of the government-linked local corporations and board of directors' meetings and this association has not been found for the board size. Our results are consistent with the agency theory prediction and the substitution hypothesis. The results reported by this study are somewhat inconclusive on the measurement of the board of directors' effectiveness. This study used an interaction term of the most common board of directors' characteristics, board size and meetings, as the measurement for the board of directors' effectiveness. Future line of research may use a composite measurement for the board of directors' effectiveness of several board of directors' characteristics such as financial expertise, CEO duality, and multiple directorships. In specific, this study contributes to the extant research by providing an empirical evidence of the association between the government-linked local corporations and the effectiveness of their board of directors in the context of Saudi Arabia. Future studies may extend this model to the other GCC countries to examine its validity. The findings of this study have theoretical contribution to the agency theory and practical implications for Saudi market, companies' management, and external auditors.