Institutional Ownership Types and ESG Reporting: The Case of Saudi Listed Firms

Abstract

The main aim of this study is to investigate the influence of institutional investors’ ownership (INOW) on firms’ environmental, social, and governance (ESG) reporting in Saudi Arabia. Using data on ESG reporting from the Bloomberg database for 206 Saudi-listed firms spanning the period from 2010 to 2019 and employing ordinary least squares regression (OLS), the results show a significant and positive association between INOW and ESG reporting. When institutional investors are classified into government and privately managed institutions, the research findings clearly show that only government-managed institutional investors (Govt_IO) are linked to ESG reporting in a positive and significant way, whereas there is no significant association between privately managed institutions (Prvt_IO) and ESG reporting. In addition, when the ESG score is disaggregated by individual pillars, we find Govt_IO is positively associated with environmental score and social score. These results suggest that the association between INOW and ESG varies depending on the types of INOW, as well as the ESG components. Even after several additional analyses, including tests for endogeneity, the main results of this study still hold.

1. Introduction

Institutional investors representing pension funds, insurance companies, banks, sovereign wealth funds, and other institutions such as hedge and mutual funds, endowment funds, and private equity firms that invest money on behalf of beneficiaries are dominant players in the global financial markets. As explained by the authors in [1], the growth in institutional investors’ ownership (INOW) has caused an increase in ownership concentration, with just a very few institutional investors now holding a significant proportion of the shares in the majority of publicly listed companies. Thus, by holding sufficiently sizeable positions in corporations, institutional shareholders, or “blockholders,” have a non-negligible impact on the shareholder vote outcomes. There is ample evidence that INOW affects how companies are run and, in turn, affects organisational outcomes such as executive compensation [2], firm innovation [3,4], and earnings management [5].

In recent years, one significant area where the institutional investor’s role has the potential to make a difference has been in encouraging investee companies to actively participate in various environmental, social, and governance (ESG) practices. The United Nations Principles for Responsible Investing (PRI), which establishes six principles of responsible investing, reports that some 3750 PRI signatory institutional investors, managing about USD 120 trillion, have made the following commitment: “As institutional investors, we have a duty to act in the best long-term interests of our beneficiaries. In this fiduciary role, we believe that environmental, social, and corporate governance (ESG) issues can affect the performance of investment portfolios” [6] p. 6. With this declaration, the PRI signatory investors acknowledge that ESG is an integral factor in their investment decisions, and that their investee companies are required to make appropriate ESG disclosures. So, by embracing the PRI, the so-called “socially responsible institutional investors” help to change the ESG landscape of the companies in which they have invested and push for a more sustainable global financial system.

The impact of INOW on ESG reporting is well discussed in the literature. Most recently, the authors in [7] show that INOW elicits greater corporate transparency by championing firms to voluntarily disclose information on climate change risk. Furthermore, study [8] indicates that there is a positive causal association between INOW and ESG performance using data from 41 countries. Nevertheless, study [9] analyses empirical research that examines the impact of INOW on ESG or corporate social responsibility (CSR) reporting globally and gathers inconsistent results. INOW has a positive impact on ESG disclosure in China [10], Indonesia [11], Malaysia [12], Pakistan [13], Palestine [14], and Spain [15]. However, several studies reveal a negative association between institutional shareholdings and ESG disclosure in Jordan [16], Malaysia [17], South Africa [18], and Poland [19]. Meanwhile, a sizeable number of studies produce insignificant results, which suggests institutional investors are heterogeneous. These include studies in Bangladesh [20], China [21], India [22], Qatar [23], Saudi Arabia [24,25], and the United States [26].

As noted by study [27], p. 242, the conflicting findings imply that the “direction of causality” is unclear and that there may be other factors at play in the relationship between INOW and ESG reporting. The same observation is echoed by study [9], p. 283: “Regarding institutional investors’ nature, long-term versus short-term, active versus passive, and financial versus sustainable institutional investors can be found in practice. Sustainable (non-financial) and long-term oriented institutional investors will include ESG issues in their decision-making in line with financial goals.” Overcoming the limitation of many prior studies that analyse the role of institutional investors as a homogeneous group without considering their heterogeneity, this study addresses the gap in the literature and examines the influence of INOW on ESG transparency among Saudi publicly traded firms.

We are guided by prior studies that show institutional investors are differentiated according to their resources and monitoring abilities [28], investing preferences and horizons [29,30,31], and ownership (state or private owned) [32], which affect their attitudes regarding corporate sustainability and, therefore, their interests in ESG reporting [33,34,35] and the pressure they exert on firms to disclose such reports [36]. Consequently, the pattern of the link between INOW and ESG reporting will vary depending on the types of INOW [37]. In this study, institutional investors are classified as either government-owned (public) or non-government-owned (private).

In addition, studies [38,39] and other scholars contend that using overall ESG score may not yield the same results, compared to when the composite ESG score is disaggregated into the three individual pillars, namely E, S, and G. For example, study [40] shows that, although the Brazilian stock market positively valued firms with higher ESG performance during 2010–2015, the environmental pillar has stronger value relevance, as compared to governance pillar (mildly significant) and social pillar (no significant relationship with stock prices). Mindful that there may be significant variation in the institutional investor-ESG disclosures nexus across different ESG pillars, we are motivated to examine whether institutional investor types influence the disclosure of the three ESG components differently.

To investigate the influence of institutional investor heterogeneity on ESG reporting, we employ a sample of Saudi publicly listed firms for many reasons, as follows. First, as a significant country-level sustainable development initiative, Saudi Arabia launched a comprehensive and extraordinary vision for the year 2030 (the Saudi Vision 2030) in 2016. The Saudi Vision 2030 addressed some requirements associated with ESG issues [41,42]. The vision states that maintaining environmental sustainability is critical for the well-being of present and future generations as well as for the comfort of the people [43]. In addition, in line with the Saudi Vision 2030 objectives, the Saudi Capital Market Authority (CMA) revised the Saudi Corporate Governance Code (SCGC) in 2017 [44]. The articles 87 and 88 of SCGC, which are specifically connected with ESG, were included in the updated SCGC, along with other major ESG-related articles. Based on the revised SCGC, Saudi companies are required to report on their ESG activities. Furthermore, CMA became a partner exchange supporting the UN “Sustainable Stock Exchanges Initiative” in 2018 [45]. In line with this, in October 2021, CMA issued “ESG Disclosure Guidelines, 2022” to help Saudi publicly listed companies navigate ESG issues. According to study [45] p. 7, “The guidance document provides an overview of ESG for issuers, the key questions to consider, and it introduces a selection of the reporting options available for companies to quantify their ESG practices and embark on their own ESG journey.” Study [46] also mentions that Saudi Arabia came in second for effectively incorporating ESG concepts within the Middle East and North Africa region (MENA region).

Second, unlike other countries in the MENA region, Saudi Arabia is a substantial “G-20” economy, ranked as the world’s largest oil producer and hosting some of the world’s largest international corporations [47,48]. For instance, Saudi Arabia accounted for 44% of overall Arab market capitalisation and 25% of overall Arab GDP, respectively [46,49,50,51]. As a result, it is an attractive market for scientific research, especially in areas related to corporate governance and sustainability. Third, the Saudi capital market supports the growth of asset management and promotes institutional investment. The CMA Strategic Plan 2021–2023 outlines that “CMA attaches utmost importance to the asset management industry as one of the main securities businesses regulated by the Authority and one of the most important sources of institutional investing… Markets with a large number of institutional investors are characterized by a low level of volatility and a high level of efficiency, transparency, and governance practices.” [52], p. 21.

The current study employs a set of 36 Saudi publicly listed companies (206 company-year observations) spanning the period from 2010 to 2019. We found a statistically significant positive association between INOW and overall ESG reporting, environmental reporting, and social reporting. When an institutional investor is classified into government (Govt_IO) and privately managed institutions (Prvt_IO), we find that only Govt_IO is positively and significantly associated with overall ESG reporting, environmental reporting, and social reporting. There is no significant association between (i) INOW and governance disclosure, (ii) Govt_IO and governance disclosure, and (iii) Prvt_IO and overall and disaggregated ESG reporting. These results suggest that the association between INOW and ESG disclosures is contingent on the types of institutional investor, as well as the components of ESG disclosures. Even after conducting additional analyses and tests for endogeneity, the main results of this study still hold.

By examining the relationship between types of institutional investor and ESG reporting, the current study adds to the literature as follows. First, the majority of previous studies that investigate the determinants of ESG reporting in the MENA region have focused on various corporate governance aspects such as board characteristics, CEO characteristics, and ownership structure [24,25,41,46,50,53,54,55,56]. The current study expands the literature by investigating the association between institutional investor heterogeneity and ESG reporting, a topic that receives scant attention, particularly in the MENA region. Second, prior research has classified INOW as local versus foreign [57], pressure-resistant or pressure-sensitive [58], independent against grey [59], or transient versus dedicated [60]. Our study is one of a few studies that categorise INOW into government versus privately managed, encouraged by studies [32,61]. Third, we examine the determinants of both overall ESG disclosure, as well as individual components of ESG pillars.

The rest of this article is structured as follows. Section 2 and Section 3 discuss the theoretical frameworks that underpin this study, related literature, and the development of the hypotheses. Research methods and data collection are discussed in Section 4. The study results and discussion are discussed in Section 5. In Section 6, the results of further investigations are shown, and in Section 7, the conclusion, implications, and limitations of the study are discussed.

2. The Theoretical Framework

The theoretical framework of this study is covered by agency and stakeholder theories to explain the relationship between ownership structure and ESG reporting [14,55,62,63,64]. According to study [65], agency theory can be considered as “a theory of the ownership structure of the firm.” Study [66] claims that agency theory is mainly focused on solving the problems that could appear in the agency relationship. An agency problem may occur when a conflict of interest exists between the goals of agents (managers) and principals (shareholders) [14,65], as both of them have different preferences and objectives [67]. Thus, managers may try to maximise their own interests rather than shareholders’ wealth [63]. In this regard, study [68] recommends that where shares are widely held, the agency conflict is likely to be more severe than when they are held in a few hands. Accordingly, managers may freely disclose information in order to resolve agency problems with owners and demonstrate that they are acting in the best interests of the shareholders [69]. Study [14] claims that companies with large blockholders can successfully mitigate the agency conflict between managers and shareholders as they have a significant motive to engage in monitoring activities. Given their influential role, institutional investors have the power to impact the firms’ ESG activities by pressuring management to release more ESG information to meet their non-profit objectives and enhance their own and the firms’ reputation [8,12,14,34,35,70].

According to stakeholder theory, companies must consider their stakeholders’ interests in addition to maximising their shareholders’ value [62]. A stakeholder can be described as “any group or individual who can affect or is affected by the achievement of the organisation’s objectives” [71] p. 46. Stakeholder theory claims that it is indeed essential for firms to maintain good relationships with all of the key stakeholders in order to preserve and strengthen corporate legitimacy [62,64]. In this regard, ESG management and reporting are critical for meeting the prospects of stakeholders [72]. Based on agency and stakeholder theories, we contend that the higher INOW is, the more likely it is that a company will decide to disclose its ESG activities.

3. Literature Review and Hypothesis Development

According to [35], over time, investors’ interest in the ESG performance of their investee firms has grown dramatically [73]. Socially Responsible Investment (SRI), despite its 18th-century religious origins [74], is now an “investment philosophy” that is widely accepted by the investment community, especially institutional investors [75], who incorporate ESG aspects into their investing choices [76]. The main challenge faced by institutional investors when integrating ESG information in investment decision process is the lack of comparability of ESG information across firms and rating agencies [76,77]. Institutional investors leverage on their considerable INOW on the international capital stock markets to exert influence over these markets and, in particular, on companies’ ESG strategies and decisions [8,35,78]. As noted by [35], among the channels used by institutional investors in exercising their influence are via their voting rights to influence the management “voice” [79,80], or their existence on the firms’ board of directors [36,81], or indirectly, or their purchases or threats to sell their stock “voting with their feet” [8,80].

In addition, institutional investors’ influence stretches to ESG policies aimed at improving firms’ ESG performance [8,12,34]. They have a higher probability of supporting voluntary disclosure of ESG-related activities because it is seen as good management practice that can provide them with further information about the company, allowing them to make a more objective assessment of the company and, therefore, avoid possible risk in the long run [10]. Study [82] claims that ESG disclosure is an effective tool for institutional investors to demonstrate their feeling of responsibility to outsiders. Furthermore, study [83] contends that institutional investors play an essential part as a monitor by pressuring managers to publish more ESG information. In a similar vein, study [8] claims that institutional investors may be driven by social responsibility, especially if they live in countries where society believes in the importance of ESG issues.

In the Saudi context, even though there is a lower INOW among Saudi firms [84], the Saudi government proactively pursued institutional investors by gradually allowing foreign ownership on its Tadawul stock market. In particular, the CMA gives more priority to foreign institutional investors with long-term investment objectives [85]. In addition, the SCGC encourages institutional investors to actively make an effort to enhance Saudi firms’ governance and disclosure policies [49]. Previous Saudi studies found mixed results related to the influence of INOW on the firms’ outcomes. For example, study [49] documents a positive correlation between INOW and corporate governance disclosure among Saudi companies. On the other hand, studies [24,25] reported that in the Saudi context, INOW has no noticeable effect on ESG reporting.

Even though the relationship between INOW and ESG reporting is increasingly being investigated, empirical findings are inconsistent [62,86]. Previous studies, for instance, found no significant association between INOW and ESG reporting [22,23,24,26,87,88]. Furthermore, some empirical research finds a negative relationship between INOW and ESG reporting [16,17,18,19]. This suggests that institutional investor does not always use their ownership to support companies’ ESG pledges [89]. On the other hand, some previous studies have found that INOW has a positive effect on the disclosure of ESG information, which leads to better ESG reporting [7,8,10,11,12,13,14,15,62,90,91]. Even though previous research on the relationship between INOW and ESG reporting has shown mixed results, we propose the following:

Hypothesis 1 (H1).

There is a positive relationship between institutional investors’ ownership and ESG reporting.

Ref. [38] explores the impact of the firm’s access to finance of the three components of ESG individually and find that it is driven by both social and environmental pillars, suggesting that both social and environmental disclosures can reduce a firm’s vulnerability to capital constraints, which is crucial for its survival and growth. Likewise, in analysing the value relevance of ESG information, study [39] argues that professional investors are likely to focus on the performance in each of the ESG pillars rather than the composite or aggregate ESG score which may conceal important information. Accordingly, mindful that the investors may value the individual ESG pillars differently, and, likewise, institutional investors may accord different priorities to the individual components of the ESG, we create three sub hypotheses for each of the ESG pillars as follows:

Hypothesis 1a (H1a).

There is a positive relationship between institutional investors’ ownership and environmental reporting.

Hypothesis 1b (H1b).

There is a positive relationship between institutional investors’ ownership and social reporting.

Hypothesis 1c (H1c).

There is a positive relationship between institutional investors’ ownership and governance reporting.

Institutional investors have various governance practices and investment horizons, which influence firm performance [34,92] and their attitude toward corporate sustainability and, as a result, their interests in ESG information [33,34,35,36]. Study [93] argues that institutional investors are different in terms of their nature, so it is important to explore the different groups of institutional investors and assess which group will have the biggest effect on corporate sustainability.

Given the heterogeneity of institutional investors, prior studies have shown that the association between INOW and ESG reporting is different according to the different types of institutional investor. For instance, earlier studies document that institutional investors with a long-term investment horizon is correlated with higher ESG performance compared to those with a short-term investment horizon [34,37]. In a similar way, prior studies find that pressure-resistant institutional investors (i.e., those that have little interaction with their investee firms) typically exert active monitoring among these firms [29] and strongly push their managers to disclose more information related to ESG aspects [36]. In contrast, pressure-sensitive institutions (i.e., those that have strong interaction with their investee firms) normally embrace a passive attitude [29]. Consequently, they are demotivated to impact such firms’ ESG practices [36].

In addition, institutional investors can also be classified according to the types of owners: government-owned (public) or non-government owned (private). Earlier research shows that government engagement in business has a positive impact on corporate sustainability [24,25,27,94,95,96]. According to study [14], government-owned corporations are assumed to be more ESG-oriented since their operations are scrutinised by the public. As a result, they must pay more attention to the public’s interests [94]. Study [97] claims that government-owned corporations are likely to face pressure to publish more sustainability information in order to be visible and transparent to a diverse set of stakeholders. As a result, the government-linked institutional investors may place a greater emphasis on ESG goals than on shareholder value [95].

Additionally, government investment in investee firm is often long-term and therefore is not driven by short-term profit motives alone [98], implying that government would consider ESG issues when making investment decisions [27]. Consequently, it is reasonable to expect that a government owner would support socially responsible activities among their investee firms in their role as institutional investors [35,95]. In contrast, private (as opposed to public) institutional investors do not usually contribute to stakeholder engagement due to their shorter investment horizons and preference for stock liquidity and maximising shareholder wealth [99]. In line with this notion, previous studies document negative association between ownership by private institutional investors and ESG reporting [19]. Therefore, the following research hypotheses are formulated:

Hypothesis 2 (H2).

There is a positive relationship between ownership by government managed institutional investors and ESG reporting.

Hypothesis 2a (H2a).

There is a positive relationship between ownership by government managed institutional investors and environmental reporting.

Hypothesis 2b (H2b).

There is a positive relationship between ownership by government managed institutional investors and social reporting.

Hypothesis 2c (H2c).

There is a positive relationship between ownership by government managed institutional investors and governance reporting.

Hypothesis 3 (H3).

There is a negative relationship between ownership by privately managed institutional investors and ESG reporting.

Hypothesis 3a (H3a).

There is a negative relationship between ownership by privately managed institutional investors and environmental reporting.

Hypothesis 3b (H3b).

There is a negative relationship between ownership by privately managed institutional investors and social reporting.

Hypothesis 3c (H3c).

There is a negative relationship between ownership by privately managed institutional investors and governance reporting.

4. Research Methodology and Data

4.1. Sample and Data

The current study focuses on examining the association between INOW and ESG reporting among Saudi-listed companies. To examine the study hypotheses, the ESG data was gathered using the Bloomberg database. Only Saudi firms with a yearly ESG score from Bloomberg and that have the relevant information for the sample period spanning from 2010 to 2019 were selected for the analysis. As a consequence of this, our final sample includes a total of 206 firm-year observations (used in the main model of the study), encompassing 34 unique firms from 2010 to 2019.

Table 1. Tests of differences in mean between Saudi firms covered and not covered by Bloomberg database for year 2019.

	Saudi Firms Included in Sample (n = 34)	Saudi Firms Excluded from Sample (n = 180)	t-Test
Operating Cash Flow	18,362,425	182,015	3.180 ***
Total Assets	149,887,760	2,450,351	7.472 ***
Revenues	52,295,938	1,021,685	3.368 ***
Market Capitalisation	251,666,388	2,373,854	2.607 ***
Net Income	11,530,119	89,205	2.797 ***

*** p < 0.01.

displays the differences in means between Saudi companies having ESG score in the Bloomberg database and other Saudi companies listed in Saudi stock exchange in 2019 but without ESG score in the Bloomberg database, based on some financial characteristics such as operating cash flow, total assets, revenues, market capitalisation, and net income. The results in Table 1 indicate a significant mean difference in all financial characteristics between sampled firms and non-sampled firms, which suggests that Bloomberg tends to favour ESG coverage for larger companies.

We chose 2010 as the beginning point for our research since the SCGC came to be compulsory for Saudi firms in 2010 [100]. Further, we extended the study period until 2019, two years following the issuance of the current SCGC in 2017, and three years following the announcement of Saudi Vision 2030, which prioritises environmental and social concerns [50]. Data on institutional investor ownership and other control variables are gathered by hand and retrieved from annual reports of Saudi firms on the Tadawul stock market website. The data for other control variables was collected using Refinitiv (formerly Thomson Reuters) DataStream.

4.2. Variables’ Measurements

4.2.1. Dependent Variable: ESG Reporting

Most previous studies evaluated ESG disclosures from corporate websites, and other reporting channels using content analysis based on predetermined criteria [101,102]. This study uses Bloomberg’s ESG disclosure score as a proxy for ESG reporting because of its widespread use. More specifically, Bloomberg ESG scores with a 0–100 scale are used to assess our dependent variables (i.e., overall ESG reporting and the three individual pillars), and companies that disclose a minimal level of ESG data receive a score of 0.1, while those who publish every data item receive a score of 100. Bloomberg’s ESG database employs the most wide-ranging method to assess companies’ ESG activities and performance [103]. This method is based on the requirements of the Global Reporting Initiative (GRI) for corporate sustainability reporting [104].

4.2.2. Independent Variable: Institutional Investors’ Ownership

INOW is our main independent variable, and we follow previous research [14,19,24,62,105] to measure INOW as the proportion of shares held by institutional investors to the total shares outstanding. We further categorise institutional investors into Govt_IO and Prvt_IO. Govt_IO comprises government-managed pension funds, unit trust funds, and other institutions managed by the government, while Prvt_IO includes banks, insurance companies, privately managed mutual funds, and unit trusts [19,32,35].

4.2.3. Control Variables

This study model integrates various firm-level control variables to help prevent model misspecification and to capture additional factors that may have an influence on ESG reporting. Due to this, the size of the board of directors (BSIZE) as well as other aspects of corporate governance were taken into consideration in our research. A larger board size provides more diversity in terms of nationality, professional backgrounds, and expertise, which may benefit a company’s involvement in ESG activities [14]. Thus, we expect that BSIZE is positively associated with ESG reporting. We measure BSIZE as the total number of directors on the firm’s board [14,106,107,108,109].

We also control for board independence (BIND), as more board independence will help firm management incorporate broader ways of forming sustainable collaboration with the community, in addition to higher levels of ESG reporting [110]. Thus, we assume that BIND will be positively associated with ESG reporting. BIND is measured as a proportion of the board’s independent directors [35,108,111,112,113,114,115]. Further, we control for board of director meetings (BMET). More frequent board meetings give more possibilities to convert board experience, information, and abilities into boosting a firm’s results, including ESG activities [56,116,117]. Hence, we assume a positive correlation between BMET and ESG reporting. BMET is calculated as the board meeting number [41,56,108].

In addition, we also control for firm age (AGE), as older firms have a tendency to disclose more ESG activities [14]. AGE is measured as the natural logarithm of the number of years from the company’s establishment [14,24,32,56,112]. According to study [118], companies of a large size are anticipated to encounter stricter regulatory restrictions and are likely to be inspected by the public. As a result, they are under increased pressure to publish more information regarding ESG initiatives in order to rationalize their ESG practices [119]. SIZE is computed as the natural logarithm of the company’s market value of equity [32,120,121].

Furthermore, we control for firms’ profitability (ROA). Previous research indicates that profitable firms are more likely to participate in ESG initiatives. ROA is measured as a firm’s net profit scaled by total assets [14,24]. We control for firm leverage (LEVGE), which may influence ESG reporting [122,123], as firms with more debt are expected to have lower investment in ESG activities [124]. LEVGE is calculated as total debt divided by total assets [14,62,123,125,126]. Finally, we include firms’ dividends (LNDVD) as a control variable based on study [127], who reported that firms that pay high dividends also report more ESG activities, implying that they have sufficient resources to ensure that the needs of shareholders and any other stakeholders are met in a satisfactory manner. LNDVD is calculated as the natural logarithm of annual dividends.

4.3. Model Specification

In order to investigate whether or not there is an association between the ownership of institutional investors and ESG reporting, ordinary least squares (OLS) regression models with Huber–White robust standard errors were utilised to avoid heteroscedasticity and serial correlation issues. The following models are utilised in this study in order to investigate the hypotheses that have been posed:

$$\begin{array}{ll}ES{G}_{it}=& {\beta }_0+{\beta }_{1}INO{W}_{it}+{\beta }_{2}BSIZ{E}_{it}+{\beta }_{3}BIN{D}_{it}+{\beta }_{4}BME{T}_{it}+\\ & {\beta }_{5}LNAG{E}_{it}+{\beta }_{6}LNSIZ{E}_{it}+{\beta }_{7}RO{A}_{it}+{\beta }_{8}LEVG{E}_{it}+{\beta }_{9}LNDV{D}_{it}+\\ & \mathrm{Year} \mathrm{dummies}+{\epsilon }_{it}\end{array}$$

(1)

$$\begin{array}{ll}EN{V}_{it}=& {\beta }_0+{\beta }_{1}INO{W}_{it}+{\beta }_{2}BSIZ{E}_{it}+{\beta }_{3}BIN{D}_{it}+{\beta }_{4}BME{T}_{it}+\\ & {\beta }_{5}LNAG{E}_{it}+{\beta }_{6}LNSIZ{E}_{it}+{\beta }_{7}RO{A}_{it}+{\beta }_{8}LEVG{E}_{it}+{\beta }_{9}LNDV{D}_{it}+\\ & \mathrm{Year} \mathrm{dummies}+{\epsilon }_{it}\end{array}$$

(2)

$$\begin{array}{ll}SOCIA{L}_{it}=& {\beta }_0+{\beta }_{1}INO{W}_{it}+{\beta }_{2}BSIZ{E}_{it}+{\beta }_{3}BIN{D}_{it}+{\beta }_{4}BME{T}_{it}+\\ & {\beta }_{5}LNAG{E}_{it}+{\beta }_{6}LNSIZ{E}_{it}+{\beta }_{7}RO{A}_{it}+{\beta }_{8}LEVG{E}_{it}+{\beta }_{9}LNDV{D}_{it}+\\ & \mathrm{Year} \mathrm{dummies}+{\epsilon }_{it}\end{array}$$

(3)

$$\begin{array}{ll}GO{V}_{it}=& {\beta }_0+{\beta }_{1}INO{W}_{it}+{\beta }_{2}BSIZ{E}_{it}+{\beta }_{3}BIN{D}_{it}+{\beta }_{4}BME{T}_{it}+{\beta }_{5}LNAG{E}_{it}+\\ & {\beta }_{6}LNSIZ{E}_{it}+{\beta }_{7}RO{A}_{it}+{\beta }_{8}LEVG{E}_{it}+{\beta }_{9}LNDVD+\\ & \mathrm{Year} \mathrm{dummies}+{\epsilon }_{it}\end{array}$$

(4)

$$\begin{array}{ll}ES{G}_{it}=& {\beta }_0+{\beta }_{1}Govt\_I{O}_{it}+{\beta }_{2}Prvt\_I{O}_{it}+{\beta }_{3}BSIZ{E}_{it}+{\beta }_{4}BIN{D}_{it}+\\ & {\beta }_{5}BME{T}_{it}+{\beta }_{6}LNAG{E}_{it}+{\beta }_{7}LNSIZ{E}_{it}+{\beta }_{8}RO{A}_{it}+{\beta }_{9}LEVG{E}_{it}+\\ & {\beta }_{10}LNDV{D}_{it}+\mathrm{Year} \mathrm{dummies}+{\epsilon }_{it}\end{array}$$

(5)

$$\begin{array}{ll}EN{V}_{it}=& {\beta }_0+{\beta }_{1}Govt\_I{O}_{it}+{\beta }_{2}Prvt\_I{O}_{it}+{\beta }_{3}BSIZ{E}_{it}+{\beta }_{4}BIN{D}_{it}+\\ & {\beta }_{5}BME{T}_{it}+{\beta }_{6}LNAG{E}_{it}+{\beta }_{7}LNSIZ{E}_{it}+{\beta }_{8}RO{A}_{it}+{\beta }_{9}LEVG{E}_{it}+\\ & {\beta }_{10}LNDV{D}_{it}+\mathrm{Year} \mathrm{dummies}+{\epsilon }_{it}\end{array}$$

(6)

$$\begin{array}{ll}SOCIA{L}_{it}=& {\beta }_0+{\beta }_{1}Govt\_I{O}_{it}+{\beta }_{2}Prvt\_I{O}_{it}+{\beta }_{3}BSIZ{E}_{it}+{\beta }_{4}BIN{D}_{it}+\\ & {\beta }_{5}BME{T}_{it}+{\beta }_{6}LNAG{E}_{it}+{\beta }_{7}LNSIZ{E}_{it}+{\beta }_{8}RO{A}_{it}+{\beta }_{9}LEVG{E}_{it}+\\ & {\beta }_{10}LNDV{D}_{it}+\mathrm{Year} \mathrm{dummies}+{\epsilon }_{it}\end{array}$$

(7)

$$\begin{array}{ll}GO{V}_{it}=& {\beta }_0+{\beta }_{1}Govt\_I{O}_{it}+{\beta }_{2}Prvt\_I{O}_{it}+{\beta }_{3}BSIZ{E}_{it}+{\beta }_{4}BIN{D}_{it}+\\ & {\beta }_{5}BME{T}_{it}+{\beta }_{6}LNAG{E}_{it}+{\beta }_{7}LNSIZ{E}_{it}+{\beta }_{8}RO{A}_{it}+{\beta }_{9}LEVG{E}_{it}+\\ & {\beta }_{10}LNDV{D}_{it}+\mathrm{Year} \mathrm{dummies}+{\epsilon }_{it}\end{array}$$

(8)

Winsorizing was applied to each and every continuous variable to mitigate the problems that were brought about by outliers.

Table 2. Variable definitions.

Variables	Abbreviation	Explanations
Dependent variable
Environmental, Social, and Governance	ESG	The Bloomberg score is based on the level of ESG disclosures made by a firm
Environmental	ENV	The level of environmental disclosures made by a firm
Social	SOCIAL	The level of social disclosures made by a firm
Governance	GOV	The level of governance disclosures made by a firm
Independent variables
Institutional investors’ ownership	INOW	The percentage of shares held by institutional investors
Government-managed institutional investors	Govt_IO	The percentage of shares held by government institutional investors
Privately managed institutional investors	Prvt_IO	The percentage of shares held by privately managed institutional investors
Control variables
Board size	BSIZE	Total number of directors on the board of the firm
Board independence	BIND	The percentage of independent directors to all board members
Board meeting	BMET	Calculated as the number of board of directors’ meetings.
Company age	AGE	The number of years a company has been operating.
Age	LNAGE	The logarithm of firm age
Company size	SIZE	Market capitalisation
Size	LNSIZE	The logarithm of market capitalisation
Return on assets	ROA	Net profit scaled by total assets
Leverage	LEVGE	Ratio of total debt to total assets
Total dividend	DVD	Firms’ annual dividends
Dividend	LNDVD	The logarithm of annual dividends
Variables utilized in the further analysis
Institutional investors’ ownership	INOWDMY	Dummy variable coded 1 if INOW is above the sample median, 0 otherwise
Government-managed institutional investors	Govt_IODMY	Dummy variable coded 1 if Govt_IO is above the sample median, 0 otherwise
Privately managed institutional investors	Prvt_IODMY	Dummy variable coded 1 if Prvt_IO is above the sample median, 0 otherwise
Inverse Mills ratio	IMR	The inverse Mills ratio was calculated using the ESG probit model.
Market to book ratio	MTB	The percentage of market value of equity to book value of equity
Systematic risk	BETA	Market risk defined as the link between the volatility of the stock and the volatility of the market.

shows the measurements of all the variables that were used in this study:

5. Empirical Results

5.1. Descriptive Statistics

Table 3. Descriptive statistics.

Variables	Maximum	Minimum	Mean	SD
ESG	48.347	2.194	14.621	9.685
ENV	44.961	1.550	11.729	12.917
SOCIAL	61.667	3.333	17.814	14.221
GOV	64.286	3.571	40.751	11.733
INOW%	98.180	0.000	28.904	26.802
Govt_IO%	98.180	0.000	23.991	26.966
Prvt_IO%	40.000	0.000	4.913	11.376
BSIZE	12.000	7.000	9.660	0.890
BIND%	100.000	20.000	42.229	11.920
BMET	16.000	2.000	6.049	2.195
AGE	86.000	2.000	27.820	16.512
LNAge	4.454	0.693	3.109	0.717
SIZE	7,050,000,000	2,485,332	77,600,000	491,000,000
LNSIZE	22.676	14.726	17.065	1.090
ROA	38.200	−3.420	4.861	5.957
LEVGE	74.430	0.000	23.576	18.323
DVD	274,000,000	0.000	2,825,449	19,200,000
LNDVD	19.430	0.000	10.255	6.121

displays the descriptive statistics, which shows that there is a wide array of ESG reporting levels across the sample companies, with a mean of 14.621, while the ESG minimum score is 2.193 and the maximum is 48.347. This result implies that ESG scores among Saudi-listed firms were low throughout the study period, which in itself is comparable with previous scientific research findings [41,46,53,128,129]. Among the three pillars, governance score is the highest with an average of 40.8%, followed by social disclosure (17.8%) and environmental disclosure (11.7%). The average of INOW is 28.9%, ranging from nil to 98.2%. Furthermore, the average percentage of Govt_IO and Prvt_IO are 24.0% and 4.9%, respectively. Regarding the board characteristics variables, the findings in Table 3 show that the average scores of BSIZE, BIND, and BMET are 9.66, 42.2%, and 6.05, respectively. Regarding other control variables, the mean of AGE, SIZE, ROA, LEVGE, and annual dividends are 27.82, SAR 77.6 billion, 4.86, 23.58, and SAR 2.8 billion, respectively.

5.2. Correlation Analysis

Table 4. Correlation matrix.

Variables	ESG	INOW	Govt_IO	Prvt_IO	BSIZE	BIND	BMET	LNAGE	LNSIZE	ROA	LEVGE	LNDVD
ESG	1.000
INOW	0.209 ***	1.000
Govt_IO	0.221 ***	0.909 ***	1.000
Prvt_IO	−0.031	0.200 ***	−0.226 ***	1.000
BSIZE	0.038	−0.372 ***	−0.398 ***	0.067	1.000
BIND	−0.066	−0.086	−0.095	0.022	0.185 ***	1.000
BMET	−0.086	0.381 ***	0.449 ***	−0.166 **	−0.034	0.087	1.000
LNAGE	0.255 ***	0.174 **	0.170 **	0.008	0.213 ***	−0.033	0.004	1.000
LNSIZE	0.237 ***	0.564 ***	0.572 ***	−0.026	−0.066	−0.266 ***	0.269 ***	0.294 ***	1.000
ROA	−0.024	−0.247 ***	−0.188 ***	−0.136 *	0.089	−0.016	−0.187 ***	0.143 **	−0.037	1.000
LEVGE	0.042	−0.069	0.071	−0.327 ***	−0.315 ***	−0.269 ***	−0.026	−0.250 ***	−0.156 **	−0.143 **	1.000
LNDVD	0.059	0.179 ***	0.158 **	0.048	0.114 *	−0.085	0.103	0.585 ***	0.457 ***	0.264 ***	−0.375 ***	1.000

*** p < 0.01, ** p < 0.05, * p < 0.1.

displays the correlation matrix among all variables. A Spearman correlation matrix is used to examine the direction and degree of a linear correlation between the variables, which assists in determining the possibility of multicollinearity among the variables. Study [130] asserts that when the correlation coefficient value between the different variables is significantly larger than 0.8, multicollinearity is anticipated to be “harmful”. The findings in Table 4 indicate that the maximum correlation coefficient is 0.58 between LNAGE and LNDVD, followed by between firm size and institutional investor ownership (0.56–0.57). Therefore, multicollinearity was not considered an issue in our investigation because these values did not exceed 0.8. The findings in Table 4 also indicate that INOW and Govt_IO are positively and significantly correlated with ESG reporting at the 1 percent significance level, implying that firms with higher ownership by INOW and Govt_IO engage in more ESG activities. In addition, we used the variance inflation factor (VIF) test, which can also be used to determine whether or not multicollinearity is present. The results for VIF (unreported) for all variables are less than 3.0, which is lower than 10 [131].

5.3. Regression Results and Discussion

Table 5. OLS Regression Results (Total Institutional Investors’ Ownership).

Independent Variables	(1)	(2)	(3)	(4)
	ESG	ENV	SOCIAL	GOV
INOW	0.096 ***	0.175 **	0.188 **	0.040
	(0.034)	(0.079)	(0.084)	(0.044)
BSIZE	1.468	1.662	1.188	2.715 ***
	(1.143)	(2.512)	(2.527)	(0.952)
BIND	0.027	−0.235	−0.104	0.133 **
	(0.042)	(0.160)	(0.168)	(0.065)
BMET	−0.851 ***	−2.376 ***	−1.423 *	0.192
	(0.266)	(0.761)	(0.748)	(0.447)
LNAGE	2.430 **	4.034 *	3.935	2.755 *
	(0.971)	(2.200)	(2.654)	(1.446)
LNSIZE	1.384	−1.267	−1.354	−0.443
	(0.880)	(2.410)	(2.240)	(0.874)
ROA	0.099	0.811 *	0.550 ***	−0.165
	(0.087)	(0.416)	(0.186)	(0.166)
LEVEG	0.083 **	0.193 *	0.285 ***	0.016
	(0.039)	(0.110)	(0.103)	(0.040)
LNDVD	−0.186	−0.405	−0.202	0.036
	(0.164)	(0.393)	(0.399)	(0.211)
Constant	−32.366 **	30.348	19.472	2.999
	(13.056)	(40.484)	(36.139)	(18.731)
Observations	206	88	110	206
R-squared	0.316	0.341	0.249	0.210
Time Dummies	Yes	Yes	Yes	Yes

*** p < 0.01, ** p < 0.05, and * p < 0.

shows the main findings to investigate the potential impact of INOW on ESG reporting, using OLS estimates with robust standard error. We include the overall ESG score in model (1), while the individual ENV, SOCIAL, and GOV pillars are exhibited in models (2), (3), and (4), respectively. The results, which can be seen in Table 5, demonstrate that all models have a very high level of significance (p-value < 0.001) and R² of 0.316, 0.341, 0.249, and 0.210 for models 1, 2, 3, and 4, respectively, which suggests that the models are well-fitted and that several of the variables we use to explain ESG reporting have a significant effect.

In line with the study prediction, the findings in shown in Table 5, model (1), indicate that there is a positive and significant connection between total INOW and ESG reporting (Coef. = 0.096, p-value = 0.005). According to this result, companies that seem to have a higher degree of INOW have quite a greater propensity to participate in and continue to invest in ESG efforts. This enables us to acknowledge that our primary hypothesis was correct (H₁). The current study’s findings are contradictory to those of previous studies which found insignificant [22,23,24,26,87,88] or negative [16,17,18,19] empirical results for the link between INOW and ESG reporting.

On the other hand, our research findings are in agreement with the conclusions that were reported by the vast majority of earlier research [7,8,10,11,12,13,14,15,62,90,91], which documents a positive relationship between INOW and ESG reporting. These results support that institutional investors in the Saudi stock market are active monitors and help to facilitate a company’s transparency by disclosing more ESG activities. This is in tandem with the argument put forward by [69] that an institutional investor is strongly motivated to monitor a company’s disclosure and boost its information transparency. Very few firms have an environmental pillar score (n = 88) and social pillar score (n = 110), as compared to governance pillar score (n = 206). Models (2), (3), and (4) show that institutional investors are significantly associated with an environmental and social score but have limited influence on the governance pillar score. These results suggest that institutional investors in Saudi are more concerned with environmental and social issues which have long-term consequences, as compared to governance issues. Thus H₁, H_1a, and H_1b are supported, but not H_1c.

Table 6. OLS Regression Results (Types of Institutional Investors’ Ownership).

Independent Variables	(1)	(2)	(3)	(4)
	ESG	ENV	SOCIAL	GOV
Govt_IO	0.109 **	0.226 **	0.243 **	0.063
	(0.042)	(0.095)	(0.104)	(0.051)
Prvt_IO	0.044	0.001	0.009	−0.049
	(0.039)	(0.081)	(0.085)	(0.060)
BSIZE	1.607	1.859	1.786	2.955 ***
	(1.180)	(2.462)	(2.473)	(0.983)
BIND	0.018	−0.278 *	−0.134	0.117 *
	(0.044)	(0.165)	(0.166)	(0.065)
BMET	−0.961 ***	−2.939 ***	−1.987 **	0.001
	(0.298)	(0.842)	(0.842)	(0.459)
LNAGE	2.282 **	3.047	2.822	2.498 *
	(0.985)	(2.165)	(2.610)	(1.460)
LNSIZE	1.178	−2.179	−2.172	−0.800
	(0.956)	(2.653)	(2.353)	(0.917)
ROA	0.078	0.838 *	0.445 **	−0.201
	(0.090)	(0.449)	(0.193)	(0.165)
LEVEG	0.068 *	0.097	0.204 *	−0.010
	(0.040)	(0.121)	(0.112)	(0.044)
LNDVD	−0.175	−0.413	−0.189	0.055
	(0.163)	(0.378)	(0.383)	(0.211)
Constant	−28.199 **	63.357	47.013	10.213
	(13.802)	(47.347)	(39.870)	(19.151)
Observations	206	88	110	206
R-squared	0.320	0.370	0.275	0.218
Time Dummies	Yes	Yes	Yes	Yes

*** p < 0.01, ** p < 0.05, and * p < 0.

illustrates the findings of the second and third hypotheses, which are related to the heterogeneity of institutional investors (government owned and privately managed). It can be observed that the coefficient of Govt_IO is positive (Coef. = 0.109), and significant at the 5 percent level (p-value = 0.010). This finding supports our second hypothesis (H₂), which argues that Govt_IO is positively associated with ESG reporting. This finding is in line with those from earlier research which documents a positive association between government ownership and ESG reporting [24,25,27,94,95,96]. This finding also shows that the government, in its role as an institutional investor, may use force to get the companies in which it has invested to share more ESG information. Similar to the first hypothesis, when we disaggregate the ESG score by individual pillars, we find that government institutional investors are more inclined to boost the environmental and social score, but not the governance score, supporting our sub hypotheses H_2a and H_2b, but not H_2c.

Furthermore, Table 6 demonstrates no statistically significant link between Prvt_IO and ESG reporting, contrary to our expectations (no support for our third hypothesis (H₃)). This insignificant effect of Prvt_IO on ESG reporting supports the argument that, due to their shorter investment horizons, Prvt_IO does not often engage in public shareholder activism [99]. The finding is also at odds with [19], who found a significant negative correlation between private IO and ESG reporting.

In terms of control variables, board meetings (BMET) were negatively and significantly associated with ESG reporting. It implies that more frequent meetings by board members are negatively associated with ESG reporting. This result contradicts that of [132], who reports a positive and significant link between BMET and the quality of integrated reporting. However, this result is in line with [41] who document a negative link between BMET and ESG reporting. Firm age (LNAGE) has a positive and statistically significant relationship with ESG, indicating that older companies have a greater propensity to participate in ESG reporting activities. This finding is in tandem with [24], which states that Saudi companies that are older are more likely to disclose their CSR practices. Leverage has a positive and significant connection with ESG reporting, supporting the results of [14]. With regard to other control variables, no significant link was found between BSIZE, BINDP, LNSIZE, ROA, and LNDVD, with ESG reporting.

6. Further Investigation

To test the sensitivity or robustness of the association between institutional investors’ ownership and ESG reporting, additional analyses are carried out as reported in the following sections.

6.1. Alternative Measurement of Institutional Investors’ Ownership

In order to support the conclusions drawn from this research on the connection between INOW and ESG reporting, we use a dummy variable as another measurement of institutional investors’ ownership. Following previous studies [5,105,133], a dummy variable of ‘1’ is given if the levels of institutional investors’ ownership for different categories (INOWDMY, Govt_IODMY, and Prvt_IODMY) are more than the median for the entire sample, and ‘0’, otherwise. In general, the findings in

Table 7. OLS Regression Results (Using Dummy Measurement for Institutional Investor Ownership).

Independent Variables	(1)	(2)
INOWDMY	5.200 ***
	(1.350)
Govt_IODMY		2.998 **
		(1.408)
Prvt_IODMY		−0.853
		(1.479)
BSIZE	1.614	0.969
	(1.046)	(1.018)
BIND	−0.014	−0.005
	(0.046)	(0.047)
BMET	−0.726 ***	−0.705 ***
	(0.247)	(0.263)
LNAGE	2.878 ***	2.677 ***
	(0.922)	(0.908)
LNSIZE	1.719 **	2.006 **
	(0.799)	(0.817)
ROA	0.122	0.055
	(0.091)	(0.092)
LEVGE	0.089 **	0.052
	(0.038)	(0.041)
LNDVD	−0.243	−0.232
	(0.162)	(0.164)
Constant	−39.460 ***	−34.934 **
	(12.841)	(13.475)
Observations	206	206
R-squared	0.327	0.301
Time Dummies	Yes	Yes

*** p < 0.01, ** p < 0.05.

, Columns (1) and (2) concur with the main analyses in Table 5 and Table 6.

6.2. Endogeneity Tests

Ref. [134] emphasises that endogeneity is the most notable and common error found in empirical studies of corporate finance. Furthermore, study [14] claims that the unclear results in prior research on the association between ownership structure and ESG reporting are frequently attributed to the endogeneity problem. To address potential endogeneity in our models such as the possibility that responsible companies attract a certain kind of institutional investor, we used the two-stage least squares (2SLS) approach using instrumental variables (IVs), which has been widely used by prior research to alleviate such a problem [18,41,50,135,136]. To find the appropriate IVs, we look for instruments that may possibly affect INOW but not ESG reporting, unless they are influenced indirectly by other exploratory variables. Following earlier research [32,136,137,138,139], the chosen IVs for (INOW, Govt_IO, and Prvt_IO) are market to book ratio (MTB), company size (LNSIZE), and systematic risk (BETA). Study [140] demonstrates that different types of institutional investors have different investment preferences for MTB. Further, study [141] finds that INOW are more likely to be attracted to companies of a large size. Finally, study [137] contends that systematic risk is more likely to influence INOW. Therefore, we utilized BETA as a proxy for market risk.

The results of first-stage regressions are presented in

Table 8. 2SLS Regression Results (First Stage).

First Stage Regression	(1)	(2)	(3)
	INOW	Govt_IO	Prvt_IO
BSIZE	−11.156 ***	−10.947 ***	−0.209
	(1.440)	(1.336)	(0.829)
BIND	0.097	0.208 **	−0.110 *
	(0.103)	(0.093)	(0.064)
BMET	2.438 ***	3.700 ***	−1.263 ***
	(0.658)	(0.623)	(0.403)
LNAGE	5.914 ***	6.564 ***	−0.649
	(2.131)	(2.126)	(1.248)
LNSIZE	−0.012	0.498	−0.510 ***
	(0.333)	(0.341)	(0.153)
ROA	−0.134	0.107	−0.241 ***
	(0.092)	(0.090)	(0.048)
LEVGE	−0.540 *	−0.572 *	0.032
	(0.323)	(0.319)	(0.154)
LNDVD	−5.274 ***	−5.208 ***	−0.067
	(1.814)	(1.772)	(0.768)
LNSIZE	11.910 ***	12.581 ***	−0.671
	(1.206)	(1.143)	(0.695)
BETA	−5.212	−2.359	−2.853
	(5.946)	(5.781)	(2.484)
Constant	−77.046 ***	−124.877 ***	47.830 ***
	(25.787)	(23.729)	(16.076)
Time Dummies	Included	Included	Included
Observations	206	206	206
R2	0.589	0.588	0.128
F-Test (IV)	23.71	22.96	2.00

*** p < 0.01, ** p < 0.05, and * p < 0.

, where INOW, Govt_IO, and Prvt_IO are used as dependent variables and regressed over chosen IVs and control variables. The findings in Panel A show that MTB is significant and negatively linked with INOW and Govt_IO. In the same vein, LNSIZE has a significant and positive association with INOW and Govt_IO. The reported significant F-tests indicate that the IVs are robust, denying the assumption that they can be eliminated from the first-stage regressions. The findings in

Table 9. 2SLS Regression Results (Second Stage).

Second Stage Regression	(1)	(2)	(3)
	ESG	ESG	ESG
INOW	0.169 ***
	(0.050)
Govt_IO		0.166 ***
		(0.047)
Prvt_IO			−2.144
			(1.432)
BSIZE	2.320 **	2.227 **	−0.703
	(0.997)	(0.961)	(1.866)
BIND	0.002	−0.013	−0.217
	(0.038)	(0.038)	(0.180)
BMET	−0.976 ***	−1.174 ***	−2.985
	(0.308)	(0.342)	(1.937)
LNAGE	1.913 *	1.806 *	1.650
	(1.020)	(1.004)	(2.685)
ROA	0.147	0.065	−1.048
	(0.103)	(0.091)	(0.709)
LEVGE	0.093 **	0.049	−0.505
	(0.038)	(0.037)	(0.379)
LNDVD	−0.111	−0.109	−0.061
	(0.148)	(0.146)	(0.305)
Constant	−17.399 *	−11.286	78.500
	(9.843)	(9.124)	(53.759)
Time Dummies	Yes	Yes	Yes
Observations	206	206	206
R2	0.296	0.304	-

*** p < 0.01, ** p < 0.05, and * p < 0.

display the findings from the second-stage regressions, where the ESG reporting is the dependent variable. In tandem with the main study results reported in Table 5 and Table 6 for OLS regression, the 2SLS methodology supports a significant and positive link between INOW, Govt_IO, and ESG reporting. These findings imply that our main findings are not driven by endogeneity.

6.3. Heckman’s (1979) Two-Stage Self-Selection Model

There is a lot of discussion about self-selection bias within the body of scholarly work pertaining to institutional investors [32,142]. Given that our sample might be skewed, we employ Heckman’s (1979) two-step technique as a robustness test to account for this possible issue. Following the technique of Heckman’s analysis, first, we estimate the Inverse Mills Ratio (IMR) from the regression models that predict parameters associated with INOW [143], where in all regression models, the dependent variables are INOWDMY, Govt_IODMY, and Prvt_IODMY, dummy equal to 1 if the percentage of INOW in the firms is above the study sample median and 0 otherwise. The results for first-stage probit regression are presented in

Table 10. Two-stage estimation (Heckman).

Independent Variables	INOWDMY	ESG	Govt_IODMY	ESG	Prvt_IODMY	ESG
	(1)	(2)	(3)	(4)	(5)	(6)
INOW		0.092 ***
		(0.030)
Govt_IO				0.088 **
				(0.041)
Prvt_IO						−0.002
						(0.042)
BSIZE	−1.094 ***	0.554	−1.084 ***	−1.410	−0.376 **	−1.058
	(0.186)	(2.385)	(0.192)	(2.060)	(0.171)	(1.964)
BIND	0.049 ***	0.066	0.065 ***	0.173	−0.030 **	−0.070
	(0.014)	(0.116)	(0.015)	(0.118)	(0.015)	(0.128)
BMET	0.069	−0.781 ***	0.052	−0.829 ***	−0.286 ***	−1.661
	(0.058)	(0.279)	(0.057)	(0.280)	(0.090)	(1.274)
LNAGE	0.037	2.437 **	0.771 ***	4.269 ***	−0.148	2.594 **
	(0.218)	(0.972)	(0.235)	(1.582)	(0.227)	(1.121)
LNSIZE	0.679 ***	1.964	0.694 ***	3.271 *	−0.381 **	1.180
	(0.148)	(1.796)	(0.151)	(1.738)	(0.173)	(1.781)
ROA	−0.092 ***	0.010	−0.079 ***	−0.189	−0.124 ***	−0.437
	(0.027)	(0.295)	(0.028)	(0.205)	(0.033)	(0.555)
LEVGE	−0.020 **	0.066	0.007	0.078 *	−0.058 ***	−0.146
	(0.008)	(0.050)	(0.007)	(0.047)	(0.012)	(0.257)
LNDVD	0.002	−0.186	−0.024	−0.253	−0.013	−0.295 *
	(0.025)	(0.164)	(0.026)	(0.163)	(0.033)	(0.178)
IMR		1.260		3.982		4.421
		(4.028)		(3.037)		(5.129)
Constant	−2.224	−35.837 **	−6.786 **	−50.704 **	14.805 ***	9.485
	(3.100)	(17.747)	(3.044)	(23.999)	(3.663)	(61.352)
Time Dummies	Yes	Yes	Yes	Yes	Yes	Yes
Observations	206	206	206	206	206	206
Pseudo R2	0.404		0.397		0.340
R2		0.316		0.325		0.287

*** p < 0.01, ** p < 0.05, and * p < 0.

columns (1), (3), and (5). The estimated IMR is incorporated as a corrective factor for any sample selectivity bias in the second stage. The findings in Table 10 columns (2), (4), and (6) support those in our main regressions (Table 5 and Table 6).

7. Conclusions

Using agency and stakeholder theories, this study examines the impact of various types of institutional investors on aggregated and disaggregated ESG reporting in an emerging economy. Relevant data from Saudi-listed companies from 2010 to 2019 were analysed using 206 company-year observations and various firm-level attributes. The findings imply that these linkages do actually vary across various categories of institutional investors and the ESG pillars. In particular, we find that total INOW is significantly and positively linked with ESG reporting. Hence, this result supports our first hypothesis which expects a positive association between INOW and ESG reporting. With regards to ESG pillars, the results show a positive and significant association between INOW, ENV, and SOCIAL pillars, but no significant association is found for GOV pillar. In addition, the study findings indicate that Govt_IO is significantly and positively associated with ESG reporting, supporting our second hypothesis. Further, the results show that Govt_IO is positively and significantly associated with ENV and SOCIAL pillars, but no significant association is found with GOV pillar. On the other hand, Prvt_IO appears to have no significant correlation with overall ESG reporting or individual ESG pillars. Therefore, our third hypothesis which expects a negative association between Prvt_IO and ESG reporting is not supported. The positive effect of INOW and Govt_IO on ESG reporting confirms that institutional investors in Saudi Arabia, especially the institutions managed by the Saudi government, are proactive monitors that may help companies be more transparent by disclosing informative ESG narratives. This is consistent with a study by [69], which claims that an institutional investor has significant incentives to scrutinise companies’ disclosures and boost information transparency.

The theoretical implications of our research are as follows. Firstly, this study’s results expand earlier research by acknowledging the institutional investor heterogeneity and providing a more extensive examination of their ESG disclosure preferences. So, our results help explain why previous research on the effect of INOW and ESG reporting [34,35,36,37] is fragmented. Secondly, this study provides insight into ESG reporting for developing economies and countries with a majority Muslim population.

In terms of practical implications, given the relatively limited size of institutional investors in the Saudi capital market, the promotion of institutional ownership in the capital market might be one potential strategy to encourage Saudi companies to enhance corporate transparency. In addition, to legitimise their operations, Saudi companies’ management should integrate sustainable business activities and disclose more ESG information. The management of Saudi companies should also put more focus on the value of ESG initiatives and be aware of how to raise the credibility of non-financial disclosures. Further, Saudi companies should consider the preferences of different institutional investors with respect to corporate sustainability when formulating their policies for voluntary ESG reporting. Our findings also show that institutional investors are essential mechanisms of Saudi firms’ corporate governance. When institutional investors, especially those owned by the Saudi government, own more of Saudi companies’ shares, their ESG reporting improves, especially the environmental and social components.

Our study has a limitation with respect to the sample composition. Only publicly traded large Saudi companies that had relevant ESG data from Bloomberg database were thoroughly investigated because not all of the required data from 2010 to 2019 was readily accessible for the majority of companies. This limits the generalisation of the results because less than 40 publicly traded Saudi companies had Bloomberg ESG relevant information during the study period.