This study aims to investigate the influence of board independence on agency costs in Jordan, an emerging economy, and examine whether managerial ownership moderates this relationship.
This study uses three agency cost measures, the expense ratio, asset utilization ratio and Q*free cash flow, and examines their relationship with board independence and managerial ownership using OLS regression, with two-stage least squares estimation to address potential endogeneity.
Contrary to expectations, board independence in Jordan is associated with significantly higher, not reduced, agency costs on some measures, suggesting it may be counterproductive, not merely ineffective. The interaction with managerial ownership is negative and significant on some measures, indicating it complements board independence in constraining agency costs.
The findings suggest that agency theory’s predictions are conditional on institutional context, calling for theoretical models that integrate institutional, cultural and ownership factors shaping board effectiveness in emerging economies. Future research should examine whether compliance-driven rather than genuine monitoring independence explains the counterproductive effect observed here, using measures of substantive independence and longitudinal data.
The incongruence between board independence and agency cost control challenges the assumption that Western governance models transfer seamlessly to emerging markets, pointing to a need for stronger regulatory enforcement and minority investor protections in Jordan.
Although board independence is well-documented in developed markets, evidence from emerging economies remains limited. This study shows that independent directors in Jordan are counterproductively associated with higher agency costs, going beyond prior null or mixed MENA results. It also demonstrates that managerial ownership complements board independence in reducing these costs, and offers an institutional explanation, grounded in Jordan’s mandatory shareholder−director requirement, for why imported governance codes can raise rather than lower agency costs in relationship-based economies.
1. Introduction
The agency problem refers to the tendency of corporate managers to pursue their own interests at the expense of shareholders. Despite extensive efforts by regulators, shareholders and governance experts to mitigate agency costs through various control mechanisms, agency costs remain stubbornly persistent in corporations around the world, manifesting through diverse channels including managerial entrenchment, perquisite consumption and suboptimal investment decisions, even as governance reforms proliferate (Jensen & Meckling, 1976; Singh & Davidson, 2003).
A common governance mechanism used globally to address agency problems is the appointment of independent outside directors, whose monitoring role is intended to constrain managerial opportunism (Fama & Jensen, 1983; Dalton, Daily, Ellstrand, & Johnson, 1998). However, the effectiveness of this mechanism across different institutional contexts remains debatable.
As independent directors are increasingly viewed as essential for mitigating agency costs (Dalton & Dalton, 2011), their appointment has become standard practice in developed markets. In the USA, for example, the percentage of large public firms with majority-independent boards increased from 20% in 1950 to approximately 75% by 2005 (Gordon, 2006), with this upward trajectory continuing in the following decades.
A series of corporate scandals in the early 2000s, including those of Enron, WorldCom and Tyco in the USA, and HIH in Australia, among others, exposed serious board oversight failures. Investigations revealed that these firms’ boards had failed to detect or prevent accounting irregularities and managerial misconduct, raising fundamental questions about board effectiveness (Bhagat, Bolton, & Romano, 2008). In response, governance reformers in multiple jurisdictions advocated for greater board independence. The Cadbury Report 1992 in the UK, the Sarbanes-Oxley Act 2002 in the USA and several subsequent reports commissioned in the UK (e.g. Smith Report 2003; and Higgs Report 2003) recommended increasing independent director representation to strengthen monitoring capacity.
Such reforms initiated in developed countries also spilled over to developing countries by both push and pull factors. First, they were pushed into developing countries by international bodies such as the World Bank and the International Monetary Fund (IMF) to promote private sectors. Second, such reforms were pulled in by regulators in developing countries who borrowed corporate governance rules and practices from developed countries (Rashid, 2011).
The relationship between board independence and agency costs, well-established in developed markets (McKnight & Weir, 2009), appears far less clear in emerging economies. Prior studies conducted in the context of developing countries have produced divergent results. For example, Rashid (2015, 2018) reported only conditional or limited effectiveness of independent directors in Bangladesh, while Lefort & Urzúa (2008) documented outcomes that vary with institutional conditions in Chile. Evidence from Russia suggests that independent directors may not necessarily improve governance outcomes and can be less effective than insiders in certain settings (Muravyev, Berezinets, & Ilina, 2014). In the Gulf Cooperation Council (GCC) countries, Hamdan (2018) found that board independence influences agency costs in a context-dependent manner. Studies in Malaysia (Germain, Galy, & Lee, 2014) and Pakistan (Yasser, Mamun, & Seamer, 2017) similarly reveal inconsistent or, in some cases, negative associations between board independence and firm performance.
These conflicting findings suggest that institutional context may moderate the effectiveness of independent directors. This paper examines this contextual contingency by analyzing the board independence−agency cost relationship in Jordan, an emerging economy characterized by concentrated ownership and relationship-based governance. Corporate governance research across the MENA region has examined a range of settings, including the GCC countries (Hamdan, 2018), Egypt (Al-Najjar & Clark, 2017), the United Arab Emirates, where board independence has been found to negatively relate to firm performance (Farhan, Obaid, & Azlan, 2017) and Saudi Arabia, where it is associated with improved disclosure outcomes (Chebbi & Ammer, 2022). Yet, despite this growing body of evidence, findings remain fragmented and Jordan has received relatively limited scholarly attention.
This gap is notable because Jordan presents a distinct institutional tension: it combines a mandatory governance code, introduced in 2009 and revised in 2017, with deeply entrenched family ownership structures that create friction between formal governance requirements and informal control mechanisms, a dynamic that remains underexplored in the literature. This tension is reflected in an observable pattern among listed firms: approximately 90% of listed companies exhibit concentrated family ownership with family-based executives holding large stakes (Al-Azzam, Qura’an, & Mohameed, 2015), yet these firms have been increasingly appointing independent directors in response to regulatory directives. This combination of traditional ownership structures and imported governance practices, including widespread chief executive officer (CEO)-duality where powerful family shareholders assume executive roles, creates distinct challenges for independent director effectiveness.
Previous studies examining board independence effects have relied predominantly on traditional financial measures, such as return on assets (Dalton & Dalton, 2011), Tobin’s Q (Rubino, Tenuta, & Cambrea, 2017; Singh, Tabassum, Darwish, & Batsakis, 2018) and market-to-book ratios (Zeitun & Tian, 2007). However, accounting profits are susceptible to manipulation (Pham, Suchard, & Zein, 2011), potentially compromising research validity. To overcome this limitation, this study uses three direct agency cost proxies: asset utilization ratio (AUR), expense ratio (EXR) and Tobin’s Q free cash flow interaction (Q*FCF).
2. Ownership structure and board practices in Jordan
Unlike many developed markets, most of the listed Jordanian firms exhibit a concentration of ownership with high insider representation on the board. Such high concentrations of ownership sometimes act as an internal control mechanism.
In contrast to firms in the West, corporate boards in Jordanian firms represent a one-tier system (EBRD, 2017), where both the chair of the board and the CEO perform their responsibilities together. Sometimes, there is no distinction between the board and the management because the CEO also acts as the board chairperson and vice-versa, a phenomenon we label as CEO duality. This CEO duality may sometimes appear in a different form. For example, a father can serve as the board chairperson while his son is the CEO, or an elder brother can hold the position of the chair of the board while a younger brother is the CEO, thereby keeping all powers within the family. This type of corporate governance structure in Jordan reflects the close-knit family structure of Arab societies. The widespread prevalence of this CEO duality in Jordan often hampers the ability of the boards of many listed Jordanian firms to perform their oversight responsibilities.
Jordanian boards are often strongly dominated by a single family, and the requirement that directors be shareholders sits uneasily with genuine independence (EBRD, 2017); only two of Jordan’s ten largest listed firms disclose non-family directors. Entrenched managerial shareholding (Lasfer, 2006) likely further undermines outside directors’ monitoring power.
A further governance challenge is weak institutional enforcement: the dominance of privileged families often impedes regulators’ ability to compel compliance, exposing the corporate sector to heightened agency conflicts and erosion of minority shareholder rights.
The Jordan Securities Commission (JSC) issued the first, non-mandatory Jordanian corporate governance code in 2005, made it mandatory in 2009 (Circular No. 12 / 1/4659) and revised it in 2017 to require at least one-third of board members be “independent,” meaning free of financial ties to the firm, its management, affiliates or auditors. While outside directors play a pivotal monitoring role in Anglo-Saxon countries, this mechanism may transfer poorly to an Arab country like Jordan, given its distinct culture, organizational dynamics and value systems (Abi-Raad, 2019).
3. Literature review, theory and hypotheses
3.1 Literature review
Research examining the relationship between board independence and agency costs has grown significantly (e.g. Rashid, 2015; Rubino et al., 2017), with existing evidence indicating that four board attributes, board composition, formation processes, structure and characteristics of board members, may directly affect firm efficiency (Dalton & Daily, 1999).
Rashid (2015) found board independence reduces agency costs (via asset utilization), and Hamdan (2018) reported similar conditional effects in the GCC; Al-Najjar & Clark (2017) showed governance structures shape financial policy across MENA. More recently, Sarhan, Ntim, & Al-Najjar (2019), studying five MENA countries including Jordan, documented that firm- and country-level governance quality jointly shape board effectiveness, reinforcing the institutional contingency perspective. Similarly, Jabbouri & Almustafa (2021), drawing on 12 MENA economies, showed that institutional quality determines whether internal governance mechanisms reduce or exacerbate firm-level inefficiencies. This contingency logic extends elsewhere: board effects on agency costs vary with governance dimensions and firm attributes in Iran (Salehi, Sadatifar, & Adibian, 2019) and with ownership structure and economic conditions in the UK (Allam, 2018), while board composition and ownership concentration jointly shape firm value in a country-specific manner (Jentsch, 2019; Singh et al., 2018). Collectively, these findings support an institutional contingency perspective on board independence.
Countering this view, however, a substantial body of evidence documents negative or null effects. Yasser et al. (2017) and Song, Hoof, & Park (2017) found that higher proportions of independent directors are associated with decreased firm value and lower Tobin’s Q, while Rubino et al. (2017) concluded that independent directors provide limited benefit in resolving agency problems within family-owned firms. Singh & Davidson (2003) similarly argued that outside directors are ineffective in constraining agency costs, a conclusion echoed by Rosenstein & Wyatt (1990), who found no clear empirical support for the superiority of independent over inside directors. In emerging economy contexts specifically, Muravyev et al. (2014) found that better governance outcomes in Russia were associated with insider rather than outside directors, and Rashid (2018) reported no positive relationship between board independence and firm performance in Bangladesh. In developed markets, McKnight & Weir (2009) found no impact of outside director proportions on agency costs in the UK, while Bird, Huang, & Lu (2018) documented significant negative associations between independent directors and firm performance.
Finkelstein & Hambrick (1996) attributed these empirical inconsistencies to overlooked contextual factors, including firm strategy, life cycle and board interaction dynamics, while suggesting boards may still affect performance through quality control initiatives.
3.2 Theory and hypotheses
Agency theory conceptualizes the corporation as a nexus of contracts wherein principals (owners) and agents (managers) have inherently conflicting goals. Managers, driven by self-interest, will undertake self-serving activities detrimental to principals’ economic welfare unless constrained by control mechanisms, creating the agency problem (Jensen & Meckling, 1976). Corporate boards serve as essential internal control mechanisms protecting shareholders from managerial opportunism (Dalton et al., 1998). The board’s primary role is exercising the governance function (Tricker, 1994) by incorporating outside directors to reduce agency costs, with board independence emerging from balanced power between outside and inside directors (Rashid, 2015).
Outside directors, because of their experience and qualifications (Fields & Keys, 2003), are often viewed as an effective tool for monitoring the performance of managers and thwarting their opportunistic activities, thus adding value to the firm (Fama & Jensen, 1983). Dalton et al. (1998) argued that such outside directors can indeed ensure the benefits of superior performance to the firm because of their independence from firm management.
It is also important to take into consideration the influence of Jordanian national culture on its corporate culture. Existing literature suggests significant relationships between Hofstede’s (1980) cultural dimensions and corporate governance (Humphries & Whelan, 2017). Over 95% of Jordanian people are Sunni Arabs, displaying the basic tenets of Arab cultural values. The Jordanian society is clan-based, and people owe their loyalty to their tribe and family (Al-Ganideh & Elahee, 2018).
Jordan’s cultural profile intensifies challenges for independent director effectiveness. With a high collectivism score of 70 (Hofstede, 2026), Jordanian society prioritizes family loyalty over institutional commitments, making it difficult for outside directors to challenge management that shows allegiance to family-based inside directors and controlling shareholders. Additionally, Jordan’s high uncertainty avoidance score of 65 reflects managerial discomfort with ambiguity and preference for maintaining status quo. This cultural orientation predisposes Jordanian managers toward traditional practices that satisfy inside directors rather than innovations proposed by independent outsiders. Together, the cultural traits of family primacy and resistance to change suggest independent directors face both relational barriers to acceptance and structural resistance to their governance contributions.
Despite the distinctive institutional and cultural context of Jordanian corporations − particularly the prevalence of family ownership structures and strong collectivist norms − board independence can still play a meaningful role in reducing agency costs. Agency theory suggests that outside directors enhance monitoring effectiveness by limiting managerial opportunism and aligning managerial actions with shareholder interests (Fama & Jensen, 1983; Dalton et al., 1998). While Jordan’s high collectivism and clan-based loyalty systems may weaken informal accountability mechanisms and constrain internal monitoring within family-dominated boards, these same conditions increase the need for structurally independent oversight to counterbalance concentrated control. Independent directors, by virtue of their external status, professional expertise and reduced attachment to controlling families, can introduce objective judgment into board deliberations and improve transparency in decision-making. Even in environments characterized by high uncertainty avoidance and resistance to change, formal governance structures such as board independence become particularly valuable as substitutes for weaker market and institutional controls. Therefore, although cultural and ownership patterns may moderate its effectiveness, board independence is still expected to reduce agency costs in Jordanian corporations by strengthening monitoring and improving alignment between controlling shareholders, managers and minority investors. This discussion leads to the following hypothesis:
Board independence is associated with reduced agency costs in Jordanian firms.
We posit that managerial ownership, particularly its interaction with board independence, may further help alleviate agency costs, consistent with the “convergence of interest” hypothesis (Jensen & Meckling, 1976). This is supported by evidence linking managerial ownership to agency costs and performance: an inverse relationship with agency costs (Ang, Cole, & Lin, 2000); a positive association with AUR (Singh & Davidson, 2003; Rashid, 2016); a strong correlation with agency costs (McKnight & Weir, 2009); and negative relationships with firm performance and Tobin’s Q (McConnell & Servaes, 1990).
In situations where the absence of directors’ shared ownership contributes to elevated agency costs, managerial ownership may have dual implications: while moderate levels can align interests, excessively concentrated ownership may enable managers to exert undue influence on board members and potentially expropriate firm wealth (Jentsch, 2019). Accordingly, we argue that the interaction between board independence and managerial ownership is likely to play a complementary governance role, such that higher board independence, when combined with appropriate levels of managerial ownership, may further strengthen monitoring effectiveness and reduce firm agency costs.
Based on this reasoning, we propose the following hypothesis:
The interaction between board independence and managerial ownership will alleviate firm agency costs.
4. Research method
4.1 Sample selection
There were 171 companies listed on the Amman Stock Exchange (ASE) as of September 30, 2022. Following the pattern in some earlier studies (e.g. Rashid, 2015, 2016), this research was restricted to nonfinancial firms as financial firms (e.g. insurance firms and banks) are subject to different CG disclosure rules and a different regulatory structure. After removing 91 financial firms, we arrived at 80 nonfinancial firms. Based on the published annual reports, we were able to collect data for the period of 2006–2022 for 55 firms, creating a balanced panel of 935 observations (Table 1). We could not find data for 25 firms for the period of 2017–2022. However, we included these firms as well and arrived at an unbalanced panel of 1,210 (55 × 17 + 25 × 11) observations representing 80 companies (Table 3).
Sample construction
| Item | No. of firms |
|---|---|
| Number of listed firms in 2022 | 171 |
| Less: firms in the financial sector | 91 |
| Total nonfinancial firms | 80 |
| Firms with non-availability of data | 25 |
| Number of firms in the sample | 55 |
| Number of observations (balanced) 55x17 years (from 2006 to 2022) | 935 |
Both the balanced and unbalanced panels are composed of a variety of industries (Tables 2 and 3) as per the Global Industry Classification Standard (GICS). The continuous variables were winsorised at the 1st and 99th percentiles to reduce the likelihood of getting misleading findings caused by outliers. This threshold is consistent with established practice in corporate governance panel data research (Rashid, 2015, 2016; McKnight & Weir, 2009) and is warranted given the extreme positive skewness in the raw distributions of EXR and Q*FCF.
4.2 Definition of variables
4.2.1 Dependent variables.
This study uses asset utilization ratio (AUR), expense ratio (EXR) and Q interaction of free cash flows (Q*FCF) as reliable proxies for firms’ agency costs (Ang et al., 2000; Singh & Davidson, 2003; Rashid, 2015). According to Ang et al. (2000, p. 82), AUR is the “proxy for the loss in revenues attributable to inefficient asset utilization” and is calculated as the ratio of annual sales to total assets. Put another way, it measures management’s capacity to use assets efficiently (Singh & Davidson, 2003). Singh & Davidson (2003) further posited that a low AUR indicates management culpability in using assets in “non-cash flow generating and probably value-destroying ventures” (p. 799).
Expenses ratio (EXR) is well known as a reliable direct proxy for agency costs. It is defined as the ratio of operating expenses (company’s selling, general and administrative expenses, such as rents, managerial salaries, advertising and marketing, supplies and utility expenses) to total annual sales and refers to how effectively a firm’s management controls operating costs. Ang et al. (2000) posited that EXR “captures excessive expenses including perk consumption” (p. 82). A low EXR should reflect lower agency costs and vice versa (Singh & Davidson, 2003).
The third measure of the firm’s agency cost used in this study is the “Q free cash flow interaction.” This measure is the interaction of free cash flows and growth opportunities (McKnight & Weir, 2009; Rashid, 2015, 2016). Following Rashid (2015, 2016), the growth opportunities were defined by dummy variables: one (1) if the firm’s Tobin’s Q was less than 1 (pointing out a poorly managed company), or zero (0) otherwise. Following McKnight & Weir (2009) and Rashid (2015), cash flows were calculated by operating income before depreciation minus the sum of taxes plus interest expense and dividends paid divided by total assets. A firm with high (low) growth opportunities was expected to have low (high) agency costs (McKnight & Weir, 2009). Hence, the high value of this measure reflects a high agency cost.
As evidence of the mitigation of a firm’s agency costs, a negative association between board independence and agency cost is indicated by EXR and Q*FCF. Similarly, a positive relationship between board independence and agency cost can be measured using AUR.
4.2.2 Key independent variable.
The independent variable in this study is board independence. Following Rashid (2018), we operationalize board independence as the number of outside directors to total number of directors on the board (BDIND).
4.2.3 Control variables.
Several control variables are included in this study. They are ownership variables, the board size, firm size, firm age, sales growth, liquidity, leverage, research and development expenditure (R&D), years of operation and industry where the firm is operating.
Three ownership control variables are used in this study: Managerial share ownership (MGTOWN), institutional ownership (INSTOWN) and largest blockholding (LBOWN). MGTOWN is measured as the percentage of shares owned by directors; INSTOWN is defined as the percentage of shares held by financial institutions, while LBOWN is measured as the percentage of shares owned by the three largest blockholders.
Board size may influence the capacity of boards to work efficiently. In accordance with Elsayed (2007), in this study, board size (BDSIZE) denotes the total number of members on the board of directors. Firm size (SIZE) in this study is measured as the natural logarithm of the total assets.
A firm’s sales growth is important for its survival (Durnev & Kim, 2005). It may also indicate a firm’s future performance. Sales growth (GROWTH) is measured as the percentage of current year sales minus previous year sales, divided by previous year sales.
Liquidity helps identify firm-specific attributes. Liquidity (LIQ) is calculated as current assets divided by current liabilities. Debt is an essential governance tool because higher debt levels help decrease agency costs. Debt may ameliorate the debtholder’s agency problem and affect performance (Jensen & Meckling, 1976). Leverage (LEV) is defined as the ratio of total debt to total assets.
Firms with high research and development expenses (R&D) are more likely to have high growth and may have a high valuation (Durnev & Kim, 2005). R&D is calculated as a percentage of total research and development expenditure, divided by sales. The definition of all the variables is presented in Table 4.
Definitions of variables
| Variable | Abbreviation | Definition |
|---|---|---|
| Asset utilization ratio | AUR | The sales divided by total assets |
| Return on assets | ROA | ROA is calculated as “earnings before Interest and Taxes” (EBIT) divided by the book value of average total assets |
| Expense ratio | EXR | The operating expenses divided by sales |
| Tobin’s q and free cash flow interaction | Q*FCF | The interaction between Tobin’s q and free-cash flow |
| Independent director | BDIND | The proportion of independent director on the board |
| Managerial ownership | MGTOWN | The proportion of shares held by the board of directors |
| Institutional ownership | INSTOWN | The proportion of shares owned by the institutional investors |
| Blockholder ownership | LBOWN | The proportion of shares held by the three largest block holders |
| Board size | BDSIZE | The number of directors on the board of a firm in a given financial year |
| Firm age | AGE | The number of years the firm is listed on the stock market |
| Sales growth | GROWTH | The percentage increase in sales over the previous financial year |
| Liquidity | LIQ | The current assets divided by the current liabilities |
| Leverage | LEV | The total debt divided by total assets |
| Research and development | R&D | The research and development expenditure divided by sales |
Industry effects may assist in identifying “unobserved heterogeneity at the industry level” (Elsayed, 2007, p. 1207) and performance measures may vary among industries. Industry effects (GICS) are controlled for by dummy variables of (1) if the firm is in the industry, or (0) otherwise. The regression equations are also controlled for a time effect by adding “time dummies” for the year the observations are made.
4.3 Regression model specification
To test H1, the following model is developed:
To test H2, the following model is developed:
where Y is the agency cost measure (AUR, EXR or Q*FCF) and BDIND is board independence; the remaining controls (MGTOWN, LBOWN, INSTOWN, BDSIZE, SIZE, AGE, GROWTH, LIQ, LEV, R&D) are as defined in Section 4.2.3; year and industry are year and industry dummies; and α, β and ɛ are the intercept, coefficients and error term, respectively.
For performing sound statistical analysis, there is a necessity to meet the assumptions of normality and controlling for multicollinearity, heteroskedasticity and endogeneity. Although Coakes & Steed (2001) argued that the violations of normality are of little concern when the sample size is large (greater than 30), we nevertheless conducted tests for normality. The residual test/histogram-normality test of the regression equation produced a “Bell shape” curve conforming to the normality of analyzed data.
The correlation matrix (Table 5) and variance inflation factors (all below 4) indicate no material multicollinearity. The Breusch–Pagan–Godfrey test indicates heteroscedasticity, which we correct for using White’s (1980) correction. We also control for industry (two-digit SIC codes) and year, since the relationship between board independence and agency costs may vary by industry. Following Rashid (2015, 2016), the F-test for the predicted value of board independence was marginally insignificant (F = 1.6393 and relevant p = 0.2008). This finding indicates that there are no signs of potential endogeneity between agency costs and board independence, suggesting that both the OLS estimate, and IVs are consistent. As discussed in Section 5.3, we further address potential endogeneity by estimating two-stage least squares (2SLS) models with an industry-median instrument for board independence.
Correlation matrix of the explanatory variables
| Items | Variable | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | VIF |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | BDIND | 1 | 1.39 | ||||||||||
| 2 | MGTOWN | −0.438*** | 1 | 2.07 | |||||||||
| 3 | LBOWN | −0.358*** | 0.563*** | 1 | 2.08 | ||||||||
| 4 | INSTOWN | 0.233*** | 0.002 | 0.334*** | 1 | 1.39 | |||||||
| 5 | BDSIZE | 0.001 | 0.170*** | −0.204*** | −0.056* | 1 | 1.56 | ||||||
| 6 | SIZE | −0.135*** | 0.161*** | 0.06 | 0.059 | 0.468*** | 1 | 1.42 | |||||
| 7 | AGE | −0.124*** | −0.052 | 0.052 | 0.287** | 0.092*** | 0.092*** | 1 | 1.27 | ||||
| 8 | GROWTH | 0.039 | −0.024 | 0.030 | 0.015 | −0.017 | 0.001 | −0.157 | 1 | 1.04 | |||
| 9 | LIQ | −0.055* | 0.111*** | 0.147*** | −0.054* | −0.120*** | −0.195*** | −0.182*** | 0.028 | 1 | 1.27 | ||
| 10 | LEV | 0.072** | −0.171*** | 0.036 | 0.149*** | −0.015 | 0.198*** | 0.288*** | 0.000 | −0.405*** | 1 | 1.37 | |
| 11 | R&D | 0.016 | 0.057 | 0.008 | −0.129*** | 0.015 | 0.061 | 0.078** | −0.035 | −0.100*** | 0.121*** | 1 | 1.06 |
5. Findings
5.1 Descriptive statistics
Table 6 contains the descriptive statistics of the variables: the mean, median, maximum, minimum and standard deviation. The averages of agency costs in terms of AUR, EXR and Q*FCF are 56.4, 33.8 and 0.3%, respectively. On average, 47.6% of company board members are independent. This number is close to the independent board members in some Anglo-American countries. The number of independent directors range from 0% (no independence) to 90.9% of total directors. Jordanian Corporate Governance Codes (JCGC) recommended at least one-third of board members should be independent. This is comparable to the MENA-region average of 41.2% (Al-Najjar & Clark, 2017), though well below the 71.0% reported for Australia (Krishnamurti & Velayutham, 2017).
Descriptive statistics of the sample
| Variables | N | Mean | Median | Min. | Max. | SD |
|---|---|---|---|---|---|---|
| AUR | 935 | 0.564 | 0.519 | 0.004 | 2.571 | 0.429 |
| EXR | 935 | 0.338 | 0.182 | 0.015 | 4.652 | 0.592 |
| Q*FCF | 935 | 0.003 | 0.002 | −0.627 | 0.541 | 0.084 |
| BDIND | 935 | 0.476 | 0.455 | 0.000 | 0.909 | 0.256 |
| MGTOWN | 935 | 0.452 | 0.418 | 0.001 | 0.974 | 0.261 |
| LBOWN | 935 | 0.517 | 0.487 | 0.090 | 0.969 | 0.219 |
| INSTOWN | 935 | 0.225 | 0.114 | 0.000 | 0.925 | 0.263 |
| BDSIZE | 935 | 8.506 | 9.000 | 5.000 | 13.000 | 2.469 |
| SIZE | 935 | 17.076 | 17.175 | 3.761 | 21.004 | 2.169 |
| AGE | 935 | 2.936 | 3.045 | 0.000 | 3.807 | 0.697 |
| GROWTH | 935 | 0.102 | 0.030 | −0.956 | 4.970 | 0.643 |
| LIQ | 935 | 2.617 | 1.575 | 0.060 | 29.036 | 3.785 |
| LEV | 935 | 0.334 | 0.287 | 0.004 | 0.938 | 0.224 |
| R&D | 935 | 0.015 | 0.000 | 0.000 | 0.229 | 0.043 |
Table 6 shows the mean board size as 8.5 members, comparable to Australia’s 8.63 (Krishnamurti & Velayutham, 2017) and Malaysia’s 7.65 (Germain et al., 2014), though smaller than the Euro Top 100 mean of 16.963 (Green & Homroy, 2018). In Jordan, the average percentages of director ownership, largest blockholders and institutional investors are 45.2, 51.7 and 22.5%, respectively.
5.2 Results
Table 7 presents the relationship between board independence and three agency cost proxies: asset utilization (AUR), expense ratio (EXR) and Q*FCF. Although the results are not identical across the balanced and unbalanced panels, they converge on a single conclusion: H1 is not supported. In the balanced panel, board independence exhibits an insignificant negative relationship with AUR but significant positive relationships with EXR and Q*FCF. In unbalanced panel, only the EXR measure shows the expected negative relationship. The balanced panel results are particularly instructive in this regard: the significant positive coefficients for EXR and Q*FCF suggest that, far from reducing agency costs, board independence may be associated with higher agency costs along these dimensions.
Relationship between board independence and agency cost
| Items | Dependent variables | |||||
|---|---|---|---|---|---|---|
| Balanced panel (935 observations) | Unbalanced panel (1,210 observations) | |||||
| (1) | (2) | (3) | (4) | (5) | (6) | |
| AUR | EXR | Q*FCF | AUR | EXR | Q*FCF | |
| BDIND | −0.066 (−1.32) | 0.218** (2.30) | 13.099* (1.92) | −0.045 (−1.02) | −3.754** (−2.03) | 4.383 (0.82) |
| MGTOWN | −0.245*** (−4.05) | 0.079 (0.69) | 34.304*** (3.34) | −0.223*** (−4.15) | −0.561 (−0.25) | 28.280*** (3.71) |
| LBOWN | −0.015 (−0.21) | −0.039 (−0.30) | 33.642*** (3.21) | 0.013 (0.21) | 6.652*** (2.59) | 34.884*** (4.46) |
| INSTOWN | 0.081 (1.50) | 0.148 (1.47) | −13.532 (−1.25) | 0.080 (1.63) | −1.850 (−0.90) | −9.440 (−1.12) |
| BDSize | 0.036 (0.85) | −0.305*** (−3.76) | 17.224** (2.46) | 0.002 (0.04) | 4.713*** (2.99) | 17.005*** (3.31) |
| SIZE | −0.004 (−0.70) | −0.014 (−1.24) | 12.677*** (7.63) | −0.002 (−0.39) | −0.598*** (−2.64) | 8.809*** (7.61) |
| AGE | 0.042** (2.16) | −0.077** (−2.09) | 4.612* (1.94) | 0.049*** (2.99) | −3.370*** (−4.89) | 3.478* (1.95) |
| GROWTH | 0.062*** (3.86) | −0.093*** (−3.06) | 0.454 (0.21) | 0.062*** (4.64) | −1.483*** (−2.66) | 0.574 (0.38) |
| LIQ | −0.001 (−0.27) | 0.016** (2.43) | 1.278*** (2.69) | −0.003 (−1.01) | 0.016 (0.13) | 0.895** (2.50) |
| LEV | −0.101 (−1.40) | 0.224* (1.66) | −58.255*** (−5.17) | −0.104* (−1.71) | 6.025** (2.36) | −45.613*** (−5.77) |
| RD | −0.662*** (−2.70) | 1.623*** (3.51) | −34.281 (−1.17) | 0.003 (0.34) | 0.074 (0.19) | −2.033** (−2.06) |
| Constant | 0.513*** (3.76) | 1.292*** (5.03) | −281.177*** (−10.47) | 1.013*** (6.87) | 9.217 (1.49) | −257.575*** (−12.27) |
| Industry fixed effect | Yes | Yes | Yes | Yes | Yes | Yes |
| Year fixed effect | Yes | Yes | Yes | Yes | Yes | Yes |
| R-square | 0.4825 | 0.1937 | 0.5091 | 0.4813 | 0.5418 | 0.4615 |
| F | 20.80 | 5.36 | 17.38 | 19.81 | 25.25 | 14.83 |
| Prob > F | 0.0000 | 0.0000 | 0.0000 | 0.0000 | 0.0000 | 0.0000 |
This table presents the summary results of the relationship between board independence, board attributes and agency cost under different agency cost measures; the t-statistics asterisks indicate significance at p < 0.01 (***), p < 0.05 (**) and p < 0.10 (*) levels, respectively
We also notice that the managerial ownership is a source of agency cost as there is a significant negative relationship between managerial ownership and AUR measure of agency cost (Models 1 and 4 in Table 7) and there is a significant positive relationship between managerial ownership and Q*FCF measure of agency cost (Models 3 and 6 in Table 7).
The coefficient of largest blockholder is positive and significant under Q*FCF measure of agency cost (Model 3 of Table 7) for balanced panel (935 observations) and positive and significant under both EXR and Q*FCF measure of agency cost (Models 5 and 6 of Table 7) for unbalanced panel (1,210 observations). This finding implies that blockholders cannot control firm agency cost. The results also suggest that institutional ownership does not have any influence on firm agency cost.
Table 8 reports the interaction effects of board independence and managerial ownership on agency costs. The coefficient of the interaction effects of board independence and managerial ownership are negative and significant under Q*FCF measure of agency cost (Model 3 of Table 8) for balanced panel (935 observations) and negative and significant under both EXR and Q*FCF measure of agency cost (Models 5 and 6 of Table 8) for unbalanced panel (1,210 observations), suggesting board independence and managerial ownership to some extent can jointly control the agency costs.
Interaction effects of board independence and managerial ownership on agency costs
| Items | Dependent variables | |||||
|---|---|---|---|---|---|---|
| Balanced panel (935 observations) | Unbalanced panel (1,210 observations) | |||||
| (1) | (2) | (3) | (4) | (5) | (6) | |
| AUR | EXR | Q*FCF | AUR | EXR | Q*FCF | |
| BDIND | 0.013 (0.15) | 0.016 (0.10) | 47.571*** (3.88) | 0.010 (0.13) | 1.760 (0.55) | 40.906*** (4.27) |
| MGTOWN | −0.175** (−2.02) | −0.098 (−0.60) | 69.657*** (4.77) | −0.172** (−2.16) | 4.568 (1.37) | 67.486*** (5.90) |
| BDIND*MGTOWN | −0.176 (−1.11) | 0.448 (1.50) | −73.074*** (−3.37) | −0.121 (−0.87) | −12.184** (−2.09) | −76.787*** (−4.56) |
| LBOWN | −0.009 (−0.12) | −0.055 (−0.42) | 31.091*** (2.98) | 0.014 (0.23) | 6.812*** (2.66) | 30.189*** (3.87) |
| INSTOWN | 0.077 (1.44) | 0.156 (1.54) | −14.120 (−1.31) | 0.079 (1.60) | −1.977 (−0.96) | −8.017 (−0.96) |
| BDSize | 0.044 (1.01) | −0.324*** (−3.95) | 18.543*** (2.67) | 0.006 (0.16) | 5.180*** (3.26) | 17.701*** (3.48) |
| SIZE | −0.005 (−0.84) | −0.012 (−1.03) | 11.788*** (7.07) | −0.003 (−0.49) | −0.654*** (−2.87) | 8.179*** (7.10) |
| AGE | 0.042** (2.16) | −0.076** (−2.09) | 4.534* (1.92) | 0.049*** (3.01) | −3.344*** (−4.86) | 3.660** (2.07) |
| GROWTH | 0.063*** (3.90) | −0.094*** (−3.11) | 0.734 (0.35) | 0.062*** (4.67) | −1.447*** (−2.60) | 0.739 (0.49) |
| LIQ | −0.001 (−0.26) | 0.016** (2.41) | 1.297*** (2.75) | −0.003 (−1.00) | 0.021 (0.17) | 0.946*** (2.68) |
| LEV | −0.102 (−1.42) | 0.226* (1.67) | −56.290*** (−5.03) | −0.105* (−1.73) | 5.930** (2.33) | −44.044*** (−5.63) |
| RD | −0.646*** (−2.63) | 1.583*** (3.42) | −29.838 (−1.03) | 0.003 (0.32) | 0.058 (0.15) | −2.047** (−2.10) |
| Constant | 0.471*** (3.33) | 1.399*** (5.25) | −288.208*** (−10.79) | 0.963*** (6.07) | 4.119 (0.62) | −282.348*** (−13.17) |
| Industry fixed effect | Yes | Yes | Yes | Yes | Yes | Yes |
| Year fixed effect | Yes | Yes | Yes | Yes | Yes | Yes |
| R-square | 0.4833 | 0.1957 | 0.5187 | 0.4816 | 0.5436 | 0.4746 |
| F | 20.32 | 5.29 | 17.52 | 19.46 | 24.94 | 15.30 |
| Prob > F | 0.0000 | 0.0000 | 0.0000 | 0.0000 | 0.0000 | 0.0000 |
This table presents the summary results of the interaction effects of board independence and managerial ownership on agency costs; the t-statistics asterisks indicate significance at p < 0.01 (***), p < 0.05 (**) and p < 0.10 (*) levels, respectively
The findings support H2 that the interaction between board independence and managerial ownership alleviates agency costs, consistent with the convergence-of-interest hypothesis (Jensen & Meckling, 1976). When managers hold equity stakes, they have stronger incentives to align with shareholder interests, which appears to strengthen the monitoring effectiveness of independent directors. In Jordan, board independence alone does not achieve optimal monitoring outcomes, but functions as a complementary mechanism once managerial ownership is present: independent directors provide oversight while managerial ownership reduces the underlying conflict of interest, together producing lower agency costs than either mechanism alone.
5.3 Robustness test
As additional robustness checks, we re-estimated our models using alternative performance measures (ROA and Tobin’s Q) and separately for the pre- and post-COVID-19 periods (2006–2019 vs 2020–2022). Results, available from the authors, are consistent with those reported above, showing no material change in the board independence−agency cost relationship across the two periods.
As further robustness checks, we re-estimated our models using 5th/95th percentile winsorization and 2SLS, instrumenting board independence with industry-median board independence by two-digit GICS sector and year. Results, available from the authors, are directionally consistent with those reported here.
6. Discussion and conclusion
Decades of governance reform have rested on the premise that independent directors discipline managerial behavior. This study’s findings challenge that premise in the Jordanian context. The findings show that outside independent directors in Jordanian listed firms do not serve as effective safeguards against agency costs. In the balanced panel, in particular, the positive associations between board independence and EXR and Q*FCF suggest that, within Jordan’s institutional setting, the appointment of independent directors may be counterproductive rather than merely ineffective. This pattern is consistent with compliance-driven practices that signal adherence to governance reforms without meaningfully enhancing oversight.
It is plausible that listed companies in Jordan have appointed independent directors in response to regulatory directives rather than to gain from their expertise or networks. In an in-group-oriented country like Jordan, such directors may struggle to introduce a fresh perspective welcomed by dominant family members, or to exert effective monitoring; at worst, their involvement may elicit resistance from management, exacerbating existing problems. We are not challenging agency theory per se; Jensen & Meckling (1976) explicitly recognized that governance effectiveness depends on institutional environments. Rather, we are testing the institutional boundary conditions under which this prediction holds. In Jordan, where concentrated family ownership, mandatory shareholder−director requirements and weak enforcement mechanisms combine to undermine director independence in practice, we hypothesize that the institutional prerequisites for effective independent monitoring are absent.
The findings of this study have several theoretical implications worth noting. Agency theory does not assume that any governance mechanism is universally effective; it predicts that the value of monitoring depends on the severity and source of the agency problem (Jensen & Meckling, 1976). Our findings refine this prediction in two specific ways. First, they show that the dominant agency problem in concentrated-ownership emerging economies is not the classic manager-versus-shareholder conflict but a controlling-family-versus-minority-shareholder conflict, for which externally appointed directors who are required to hold equity stakes provide a structurally inadequate remedy, given that shareholding ties them to the very interests they are meant to oversee. Second, the negative and significant interaction between board independence and managerial ownership (H2) reveals that governance outcomes are shaped not by single mechanisms in isolation but by configurations of complementary mechanisms. This finding reaffirms the contingency perspective advanced by Filatotchev, Jackson, & Nakajima (2013), and is consistent with the cross-country evidence of Sarhan et al. (2019) and Jabbouri & Almustafa (2021), both of whom show that institutional quality is a key determinant of whether internal governance arrangements reduce or amplify firm-level agency conflicts in MENA economies. Relative to prior work in Bangladesh (Rashid, 2015, 2018), Chile (Lefort & Urzúa, 2008) and the GCC (Hamdan, 2018), which report board independence as neutral or only conditionally effective, our contribution lies in identifying the specific institutional mechanism, namely the mandatory shareholder–director requirement combined with concentrated family ownership, that transforms board independence from a neutral device into a governance liability.
Our findings both align with and extend prior evidence from emerging economies. Consistent with Rashid (2015, 2018) in Bangladesh and Lefort & Urzúa (2008) in Chile, we find that independent directors are less effective in environments characterized by high ownership concentration. However, in the balanced panel, the positive association between board independence and certain agency cost measures (EXR, Q*FCF) goes beyond the null or mixed effects reported in prior work. This may reflect particularly acute institutional constraints in Jordan, including the mandatory shareholder–director requirement and the near absence of nonfamily directors in large listed firms (EBRD, 2017). In contrast to Hamdan’s (2018) finding of partial effectiveness in GCC countries, Jordan appears to present a more challenging governance environment.
The findings of this study are novel in two specific respects. Unlike prior studies that report board independence as merely ineffective or neutral in emerging markets, this study documents a counterproductive effect, in that mandatory governance compliance can increase observable agency costs. This distinction carries direct policy relevance. If independence were merely ineffective, regulators could tolerate its absence or rely on other mechanisms. If it is counterproductive, however, regulatory mandates that ignore institutional fit may actively harm minority shareholders. The interaction finding is equally novel in the Jordanian context, as no prior study of listed Jordanian firms has tested whether managerial ownership complements board independence at the agency cost level. The positive finding on at least two of three measures indicates that governance reform in Jordan should focus on incentive alignment rather than structural independence per se. For the broader MENA region, where institutional environments share key structural features (Sarhan et al., 2019), these findings provide empirically grounded caution against the mechanical adoption of Anglo-American board structures without parallel strengthening of investor protection, enforcement capacity and ownership disclosure requirements.
Three institutional features compound this problem in Jordan. First, with concentrated ownership, conflicts arise between controlling and minority shareholders rather than the classic principal–agent problem. Second, controlling owners who also occupy managerial roles produce boards that represent both ownership and management simultaneously, a dual role that limits the board’s ability to address agency problems (Lefort & Urzúa, 2008). Third, outside directors in developing economies may not be truly independent (Rashid, 2015): in Jordan, the shareholder requirement, combined with outside directors typically being selected by insiders from minority shareholders outside the controlling family, fosters close ties conducive to collusion, consistent with a “give-and-take” governance dynamic (Singh et al., 2018).
What distinguishes this study from prior work is the directional finding in the balanced panel: board independence is associated with higher, not merely unchanged, agency costs in Jordan. Unlike Bangladesh (Rashid, 2015, 2018) and Chile (Lefort & Urzúa, 2008), where board independence proves neutral or conditionally effective, our evidence suggests a counterproductive effect under Jordan’s mandatory shareholder–director rule. This finding qualifies agency theory’s standard prescription: formal governance structures can raise agency costs when institutional prerequisites for effective monitoring are absent. The managerial ownership interaction result further refines theory by showing that convergence-of-interest alignment is a necessary complement to structural independence, not a substitute. In developed economies, independent directors operate within ecosystems of complementary monitoring, including regulatory watchdogs, financial analysts, an active press and activist shareholders; Jordan lacks much of this infrastructure. Hence, board independence alone cannot adequately safeguard minority shareholders’ interests. This confirms agency theory’s predictions while revealing the limitations of isolated governance mechanisms, and implies that regulators should both ensure genuine director independence and monitor inside shareholders more closely, given that managerial ownership only partially offsets the agency cost of board independence.
Jordan shares key institutional characteristics with other Arab economies, including concentrated ownership structures, relationship-based business practices and developing legal enforcement mechanisms. These commonalities suggest our findings have broader applicability across the MENA region. Regulators in Arab countries pursuing governance reforms should recognize that mandating board independence without strengthening complementary institutions, such as legal investor protection, transparent disclosure requirements and effective enforcement mechanisms, will prove insufficient to reduce agency costs. Our findings underscore that effective governance reform requires institutional congruence, not merely the mechanical adoption of Western board structures.
Similar to many other studies, this study is subject to some limitations. Our operationalization of board independence follows the Jordanian Securities Commission (JSC) definition of “outside directors” per JCGC criteria, which deviates materially from OECD independence standards. Since Jordanian law requires all directors to be shareholders, even formally designated “independent” directors hold ownership stakes, creating potential alignment with controlling families rather than minority shareholders. This institutional incongruity is not a measurement flaw but a central mechanism explaining the counterproductive findings reported here. Future research should develop measures of substantive rather than formal independence, and examine whether the governance dynamics documented in Jordan extend to other MENA economies where similar ownership structures and regulatory transplants coexist. Notwithstanding these limitations, this study contributes to the corporate governance literature by providing robust empirical evidence that the effectiveness of board independence is contingent on institutional context, and that mandatory governance codes, when introduced without complementary enforcement infrastructure and ownership reforms, may undermine the very objectives they are designed to achieve.
Jebreel Mohammad Al-Msiedeen, PhD, is an Assistant Professor of Accounting in the College of Business at Tafila Technical University, Jordan. He obtained his PhD from the University of Southern Queensland, Australia. Al-Msiedeen’s research focuses on corporate governance, corporate intellectual capital reporting, corporate social responsibility reporting and accounting education. He has published research articles in several international journals including Education and Information Technologies, Data in Brief and Asian Economic and Financial Review.
Afzalur Rashid, PhD, CPA, is a Senior Lecturer of Accounting in the School of Business at the University of Southern Queensland. Rashid has taught at universities across Australia and internationally, bringing extensive pedagogical experience to his role. His research focuses on carbon emissions and disclosure, corporate governance, corporate social and environmental reporting and earnings management. He has published in leading journals including the Asia Pacific Journal of Management, International Review of Economics and Finance, International Review of Finance, International Review of Financial Analysis, Journal of Business Ethics, Journal of Contemporary Accounting and Economics, Journal of Financial Research and Managerial Auditing Journal. Rashid serves as editor and guest editor for several international journals. In 2025, Rashid was recognized among the world’s top 2% scientists in Accounting, Business, and Management in the global ranking developed by Stanford University and Elsevier.
Mohammad Niamat Elahee, PhD is a Professor of International Business at Quinnipiac University, CT, USA. His research interests include artificial intelligence, cross-cultural negotiations, consumer animosity and wine marketing. Elahee has published in several reputed journals including Journal of Consumer Marketing. Review of International Business and Strategy, Journal of Business Strategy, European Business Review, Thunderbird International Business Review to name a few. Elahee was a US Fulbright professor in Jordan. He has also been a reviewer for the US Department of Education as well as several leading business journals and serves as an associate editor of the Journal of Comparative International Management.
Funding
This research received no external funding.
Data availability
The raw data required to reproduce the above findings are available from the corresponding author on request.
