The CEO’s last name matters: family control, independent oversight and tax avoidance

Tax avoidance (TA), the use of legal methods to reduce taxable income, has become a prominent issue in corporate finance and governance, particularly in developing economies where public resources are scarce and fiscal capacity is weak. While TA is not illegal, its ethical legitimacy is increasingly questioned, especially when excessive avoidance undermines tax equity and public finance (Gaaya et al., 2017; Khelil and Khlif, 2023). This concern is particularly acute in countries like Bangladesh, where the tax-to-GDP ratio is among the lowest in South Asia (10.3%) and annual tax losses due to evasion are estimated at USD 703 million (Hossain et al., 2024; Shakila, 2019). Given that sustainable development requires substantial public revenue, understanding firm-level drivers of TA is both practically and academically urgent (Moazzem et al., 2023).

An emerging body of research highlights that CEOs play a central role in shaping a firm’s tax strategies. However, the literature has primarily focused on CEO tenure, compensation, or financial expertise, with limited attention to the role of family CEOs, particularly in contexts characterized by family ownership and control (Bauweraerts et al., 2020; Boubaker et al., 2022; Cirillo et al., 2025; Gaaya et al., 2017; Kuo, 2022). In firms led by a family-member CEO, the potential for Type II agency problems, specifically, the divergence of interests between dominant family shareholders and minority investors, becomes markedly more pronounced (Amin and Naz, 2025; Villalonga and Amit, 2006). Family CEOs may prioritize socioemotional wealth or personal gain over shareholder value, potentially engaging in aggressive TA strategies with minimal resistance, especially in environments with weak regulatory enforcement.

Although some recent studies have begun to explore the relationship between family CEOs and TA (Brune et al., 2019; Rahmadhani and Tjaraka, 2022), they are largely situated in developed economies with strong institutional checks. The findings of these studies may not be directly generalizable to countries like Bangladesh, where regulatory oversight is limited, and family influence over management and boards is more entrenched. Furthermore, these studies often fail to isolate the role of board governance in shaping or mitigating the behavior of family CEOs. Hence, while the prior literature offers valuable foundations, it does not adequately address how board-level mechanisms influence the family CEO–TA relationship in developing economies.

Bangladesh provides a compelling and underexplored setting for this inquiry. The corporate sector in Bangladesh is overwhelmingly dominated by family-owned and family-managed firms, with estimates suggesting that over 70% of listed companies are under family control (Tabassum, 2024; Uddin et al., 2023). In these firms, family members frequently occupy the CEO position, which amplifies concerns about self-serving behavior and resource diversion. The development of corporate governance codes in Bangladesh remains a continuous process. A turning point was created by the Bangladesh Securities and Exchange Commission (BSEC) in 2006 with the setting up of a formal code of governance that has been comprehensively revamped in 2012 and 2018 to make provisions under it stronger. However, compliance often remains symbolic rather than substantive (Muttakin et al., 2018; Rashid, 2018). Independent directors are mandated on boards, but their effectiveness is frequently questioned due to issues of appointment bias, lack of expertise, and close ties to controlling families (Rashid, 2018). Additionally, political connections are pervasive in Bangladeshi firms and have been shown to provide access to favorable financing, regulatory leniency, and protection from enforcement actions (Rashid et al., 2024). In such an environment, family CEOs with political affiliations may face even weaker monitoring, as political protection reduces the disciplining effect of independent directors. These contextual realities make Bangladesh a particularly relevant setting for examining how family CEOs influence TA and how governance mechanisms such as board independence may (or may not) constrain such practices.

The present study seeks to bridge the gap by investigating the relationship between family CEOs and TA in the Bangladeshi manufacturing sector and exploring how board independence moderates this association. Unlike prior studies that primarily focus on family ownership or general governance quality, this study specifically examines the CEO’s family status and offers a more direct measure of leadership influence on corporate behavior. By employing both static and dynamic panel data analyses on a sample of the top 100 listed manufacturing firms from 2019 to 2023, this research contributes new empirical evidence on the role of executive identity in shaping corporate tax behavior in the context of emerging economies. Additionally, the study contributes to the literature by extending the analysis to consider political affiliation as a conditioning factor. In many emerging economies, political ties provide informal protection and regulatory leniency, which may weaken the board’s ability to restrain managerial opportunism. The findings reveal that while board independence generally reduces the likelihood of TA by family CEOs, this moderating effect becomes insignificant when the family CEO holds political connections, highlighting the limitations of formal governance in politically influenced environments.

This contribution of this study to the expanding literature on corporate governance is made in five important ways. First, it advances the literature on corporate governance and TA by focusing on family CEOs, rather than the more commonly examined dimension of family ownership. By isolating the CEO’s family status, the paper provides a nuanced understanding of executives’ identity from a behavioral point of view and its role in shaping corporate tax strategies.

Second, it offers a novel contextual contribution by examining the Bangladeshi corporate environment, which is characterized by concentrated family control, weak enforcement of governance codes, and the widespread influence of political connections. This extends the literature beyond the developed-country focus of most prior studies, offering evidence from an emerging economy where agency conflicts manifest differently.

Third, it contributes to agency theory by showing that family CEOs in Bangladesh are associated with greater engagement in TA, consistent with the Type II agency problem (principal–principal conflict). At the same time, it refines the theoretical debate with socioemotional wealth theory by showing that in weak institutional environments, economic self-interest tends to dominate over reputational concerns.

Fourth, it adds to the literature on monitoring mechanisms by examining the moderating role of board independence. The findings show that while independent directors can constrain the opportunistic tax strategies of family CEOs, their effectiveness is contingent on the CEO’s political affiliation. This highlights the limits of formal governance mechanisms in politically influenced contexts.

Finally, the paper makes a practical contribution by offering implications for regulators and policymakers. Specifically, it suggests strengthening the substantive independence of boards in Bangladesh, enhancing the transparency of director appointments, and improving disclosure practices related to tax governance.

The remainder of the paper is organized as follows: Section 2 discusses the theoretical framework of the study, whereas Section 3 discusses the literature review and develops hypotheses based on the prior studies. Section 4 presents the samples, variables, and research models. Section 5 demonstrates the results of descriptive statistics, bivariate analysis, and multivariate analysis. Section 6 provides an in-depth discussion of the findings, whereas Section 7 provides an additional analysis for getting more insights. Finally, Section 8 draws the conclusion by addressing the implications, limitations, and areas for future research.

Agency theory (Jensen and Meckling, 1976) provides a foundational perspective of the principal-agent relationship, which is fundamentally prone to conflicts of interest between owners and managers. In non-family firms, managers may act opportunistically to maximize their own utility through perks, entrenchment, or opportunistic financial decisions at the expense of shareholder wealth. This creates the classical Type I agency problem, which is a central focus of corporate governance research.

In family-controlled enterprises, governance dynamics are inherently more intricate as a result of highly concentrated ownership. This concentration of equity has the potential to mitigate classical principal–agent problems by harmonizing the objectives of dominant shareholders and management roles frequently occupied by the same family members. This overlap helps mitigate information asymmetry and strengthens monitoring, thereby limiting managerial opportunism (Amin et al., 2025; Anderson and Reeb, 2004; Shleifer and Vishny, 1989). However, the same concentration of power can give rise to principal–principal conflicts, where controlling family members exploit minority shareholders by leveraging privileged access to information and decision-making authority (Hasan et al., 2014; Villalonga and Amit, 2006). Such conflicts are especially pronounced in emerging markets characterized by weak governance mechanisms and limited investor protection.

From this perspective, family CEOs may use their discretion to adopt aggressive TA strategies that enrich the controlling family at the expense of transparency and minority shareholder value (Khelil and Khlif, 2023). TA, in this case, becomes a channel for expropriation, allowing family leaders to extract private benefits while preserving their dominance in the firm’s leadership. Moreover, as recent studies suggest, the quality of corporate governance plays a decisive role: stronger governance systems may restrain opportunistic TA, while weaker oversight may enable family CEOs to pursue more aggressive strategies (Itan et al., 2024). This underscores the ambivalent role of family control capable of both reducing traditional agency costs and intensifying new forms of conflict when assessing tax behavior.

Socioemotional wealth theory (Berrone et al., 2012; Gomez-Mejia et al., 2014) offers an alternative explanation for family firm behavior. Unlike agency theory, which focuses narrowly on economic incentives, Socioemotional wealth accounts for the emotional attachment, identity, and reputational concerns of family owners. The socioemotional wealth model is rooted in behavioral agency theory and developed specifically for family business contexts, making it especially relevant for analyzing the multiple and shifting goals that characterize family-controlled firms (Kuo, 2022). A central dimension of socioemotional wealth is the strong identification between the family and the firm. Stakeholders often perceive the firm as an extension of the family (Berrone et al., 2010; Dyer and Whetten, 2006). This intertwining of identities means that reputational risks borne by the firm directly affect the family’s social standing. Thus, practices such as aggressive TA may be avoided because they expose the family to reputational damage or accusations of greed, particularly in times of heightened public scrutiny (Austin and Wilson, 2017; Dyreng et al., 2016).

Moreover, socioemotional wealth emphasizes loss aversion: family owners are more sensitive to potential socioemotional losses than to equivalent financial gains (Gomez-Mejia et al., 2014). From this perspective, family firms may view TA as a risky strategy with uncertain long-term consequences, despite its short-term financial benefits (Brune et al., 2019). Preserving family reputation, intergenerational control, and community legitimacy often outweighs the appeal of immediate tax savings. Finally, socioemotional wealth theory acknowledges heterogeneity among family firms. Strategic choices such as TA may vary depending on the leadership role of the CEO, whether founder-led or successor-led, and the extent to which family values shape decision-making (Brune et al., 2019). These nuances make socioemotional wealth particularly useful for explaining why some family firms adopt conservative tax practices while others behave more aggressively.

The relationship between family CEOs and corporate TA is complex and shaped by competing theoretical perspectives. From an agency theory standpoint, family CEOs may use their dual role as owners and managers to expropriate minority shareholders through aggressive TA strategies (Amin et al., 2025; Khelil and Khlif, 2023; Villalonga and Amit, 2006). Ownership concentration reduces traditional principal–agent conflict but intensifies principal–principal conflict, allowing family leaders to extract private benefits via opaque tax practices (Amin et al., 2025; Anderson and Reeb, 2004; Hasan et al., 2014). Weak governance environments exacerbate this tendency, as seen in emerging economies such as Bangladesh and Indonesia (Itan et al., 2024).

On the other hand, socioemotional wealth theory emphasizes the non-economic goals of family firms such as reputation, identity, and legacy, which may discourage TA. Recent evidence supports this view: Amin et al. (2025) and Amin and Naz (2025) show that family founders and family ownership are generally associated with lower TA, consistent with family owners’ desire to protect their long-term reputation and maintain socioemotional value. However, these effects are not uniform. Cirillo et al. (2025) reveal a counterintuitive dynamic whereby family CEOs with high transgenerational responsibility can have a greater propensity for TA. This activity appears most vividly under two situations: when members of the next generation are directly involved in the business operations of the firm, or during extreme financial hardship. In these circumstances, the utilitarian motivation to preserve and transfer wealth from one generation to the next can dominate reputation concerns, leading to more confrontational tax planning.

The heterogeneity of family leaders is further highlighted when personal expertise and power are considered. Amin and Naz (2025) find that CEOs with financial expertise tend to pursue more aggressive TA, thereby weakening the SEW-driven conservatism of family ownership. Conversely, Oussii and Klibi (2024) report that greater CEO power reduces TA, especially when institutional investors provide external monitoring. Their results suggest that powerful CEOs may prioritize long-term stability and legitimacy, opting for less risky tax strategies when constrained by institutional oversight.

Taken together, the literature illustrates a nuanced and sometimes contradictory relationship between family CEOs and TA. Some evidence highlights opportunistic motives driven by agency conflicts (Cirillo et al., 2025; Khelil and Khlif, 2023), while other studies emphasize socioeconomic wealth-driven conservatism (Amin et al., 2025; Amin and Naz, 2025). Additional contingencies, such as next-generation involvement, financial expertise, CEO power, and governance quality, appear to critically shape tax strategies in family firms. These contrasting findings reflect contextual and institutional differences. In environments with weak regulatory enforcement, such as many developing countries, the agency-driven logic may dominate. The discretionary power enjoyed by family CEOs in such settings, especially in the absence of effective oversight, can lead to higher levels of TA. Given this theoretical and empirical ambiguity, the following non-directional hypothesis is drawn:

Previous studies have documented the role of independent directors in enhancing corporate governance and reducing agency costs (Bhagat and Black, 2002; Fama and Jensen, 1983). Independent directors are presumed to enhance oversight efficacy owing to their detachment from familial affiliations, thereby mitigating managerial opportunism such as aggressive TA strategies. Consequently, the moderating function of board independence in the relationship between family CEOs and TA gains particular significance.

Empirical evidence on the moderating effect of board independence in the context of family firms and TA is limited, especially in developing economies. Some studies suggest that independent boards can act as a check on family CEOs, reducing the likelihood of TA. For instance, Chen et al. (2010) found that firms with more independent boards engage in less TA. This is because independent directors are more likely to challenge the family CEO’s decisions and prioritize long-term corporate sustainability over short-term tax savings. Armstrong et al. (2015) also documented that financial complexity and a lower degree of tax evasion are positively correlated with board member independence. Butt and Ahmed (2021), Lanis and Richardson (2011), and Salihu and Kawi (2021) further confirm the negative association between board independence and TA.

The broader corporate governance literature reinforces this monitoring role of independent boards in different domains. For example, Saleh et al. (2025) show that board diversity and independence strengthen the effectiveness of ESG practices in reducing corporate risk, emphasizing the role of independent boards in fostering long-term sustainability. Similarly, Mansour et al. (2025) highlight that board attributes such as independence and diversity significantly influence environmental disclosure across ASEAN-5 firms, underlining the importance of board monitoring for corporate transparency. In another stream, Shubita et al. (2025) find that advertising and sales promotion improve firm performance, suggesting that oversight mechanisms such as board independence may be critical to ensure that such investments align with shareholder value rather than managerial opportunism. Moreover, Khalaf et al. (2024) emphasize the importance of system trust and integrity in organizational governance, indirectly supporting the argument that independent boards enhance accountability by building trust in organizational processes and outcomes.

Conversely, some evidence shows that the presence of independent directors does not always translate into effective governance, particularly in family firms. McClure et al. (2018) and Richardson et al. (2013) argue that independent directors may sometimes facilitate opportunistic practices by endorsing management’s rent extraction. Similarly, Anderson and Reeb (2004) and Villalonga and Amit (2006) suggest that independent directors in family firms may be reluctant to challenge entrenched family interests.

Overall, the literature suggests that board independence plays a crucial, albeit context-dependent, role in moderating family CEOs’ tax strategies. Based on these arguments, the following hypothesis is developed:

This study utilizes a panel dataset comprising the top 100 manufacturing firms listed on the Dhaka Stock Exchange (DSE), selected based on market capitalization, over the five-year period from 2019 to 2023, resulting in an initial sample of 500 firm-year observations. The starting year of 2019 was chosen to align with the implementation of the revised Corporate Governance Code (CGC) issued by the Bangladesh Securities and Exchange Commission (BSEC) in August 2018. As firms began complying with the new governance guidelines from the financial year 2018–2019, 2019 represents the earliest full year of mandatory compliance, making it a suitable baseline. The reason behind selecting the top 100 firms is that they represent more than two-thirds of the market capitalization of the listed manufacturing companies. However, due to the unavailability of annual reports, the final sample was limited to 445 firm-years. Data have been collected from annual reports and company websites. Table 1 shows that the sample comprises 9 different manufacturing industries, where the highest number of observations is from the pharmaceuticals and chemicals industry, followed by the textile industry and the engineering industry.

Three different proxies have been used to measure TA, as not all tax-aggressive conduct is likely to be captured by a single metric (Lin et al., 2014). As a result, several company TA strategies have been developed in recent literature (Dyreng et al., 2008, 2010; Frank et al., 2009).

Effective tax rate (ETR) has been used as the first proxy of TA. ETR is measured by using the ratio of total tax expense and pre-tax income. Since a lower effective tax rate indicates a higher level of corporate TA, ETR is inversely correlated with TA (Frank et al., 2009). It has been used by prior studies, as it can encompass any form of tax reduction achieved through the utilization of tax shelters and loopholes within the tax legislation (Lanis and Richardson, 2011; Dyreng et al., 2017; Gaaya et al., 2017).

The second proxy of TA considered in the study is cash flow effective tax rate (CFETR) which has been measured by using the ratio of total tax expense and operating cash flows. The data used to calculate this metric comes from cash flow statements, which may not account for the effects of earnings management (Richardson et al., 2013; Gaaya et al., 2017).

The third and final proxy of TA used in the study is book-tax difference (BTD), which has been widely used in prior studies (Lin et al., 2014; Gaaya et al., 2017). BTD is measured by using the difference between pre-tax book income and estimated taxable income and scaling it by total assets. Estimated taxable income is measured by dividing the current tax expense by the statutory tax rate.

To measure the independent variable, Family CEO (CEO_FAM), a dummy variable is constructed, taking the value 1 if the firm’s Chief Executive Officer is either the founder or a close relative of the founder, and 0 otherwise. This measurement approach is consistent with prior research on family influence in executive leadership (Brune et al., 2019; Kuo, 2022; Lee and Bose, 2021). The moderating variable, Board Independence (BIND), is measured as the proportion of independent directors on the board relative to the total number of board members, reflecting the degree of autonomy and external oversight in corporate governance structures.

In addition to the primary variables of interest, the study controls for both board- and firm-level characteristics. The board-specific control variables include: (1) board size (BSIZE), measured as the natural logarithm of the total number of directors; (2) board gender diversity (BGD), represented by the proportion of female directors; and (3) board diligence (BDIL), proxied by the number of board meetings held in a fiscal year. The firm-specific control variables include: (1) profitability, measured by return on assets (ROA); (2) leverage (LEV), calculated as the ratio of total liabilities to total assets; (3) firm size (FSIZE), measured as the natural logarithm of total assets; and (4) firm age (FAGE), defined as the number of years since incorporation. These controls account for heterogeneity in governance structures and firm characteristics that may influence TA behavior.

To test the hypotheses, the following models have been developed:

The definition of variables is shown in Table 2.

Table 3 shows the descriptive statistics of the variables used in the study. The average value of ETR is 21.79% with a minimum value of −55.09% and a maximum value of 100%. CFETR has a slightly higher mean of 30.24% with a minimum value of −72.47% and a maximum value of 100%. The average value of BTD is 43.36% with a minimum value of −76.96% and a maximum value of 84.37%. The degree of corporate TA practices differs greatly throughout the companies in the sample. CEO_FAM has a mean value of 71.69% which indicates that more than two-thirds of the sample companies are run by the founders or their relatives. Average BIND is 26.28% which is consistent with the minimum requirement of 20% by the Corporate Governance Code. Average BISZE is almost 8, ranging from 4 to 19. BGD in the sample firms is found to be very low, with an average value of 19.73% only. Average BDIL is almost 11, ranging from a minimum of 3 board meetings a year to a maximum of 19 board meetings a year.

Table 4 shows the Pearson pairwise correlation matrix and variance inflation factor (VIF) results of the model. The study has found a positive and significant correlation between ETR and CFETR. Another proxy for TA, BTD, is negatively and significantly correlated to both ETR and CFETR. This indicates a strong correlation among different TA measures, which is consistent with prior studies. CEO_FAM is found to be positively correlated with ETR and CFETR and negatively correlated with BTD at 1% significance level. This gives a primary indication that firms run by family CEOs tend to get engaged in TA. As none of the correlation values of the independent and control variables exceeded 0.7 in the matrix, multicollinearity is assumed to be absent in the model (Gujarati, 2003). To further assure the absence of multicollinearity, the VIF test has also been performed. As the mean VIF (1.35) is in between 1 and 10, it provides further evidence of the absence of multicollinearity problem (Wooldridge, 2016).

Table 5 shows the regression results based on the three proxies of TA: ETR (Model 1), CFETR (Model 2), and BTD (Model 3) using a static panel data analysis. As the Breusch-Pagan test showed the presence of heteroskedasticity in the models, the GLS method has been used to estimate the regression equation. The results of model 1 and model 2 show that CEO_FAM is negatively associated with ETR and CFETR at 5% significance level. Both ETR and CFETR are inverse functions of TA, which indicates that CEO_FAM is positively associated with TA. This also holds true for model 3, where the study found a positive association between CEO_FAM and BTD at 1% significance level. All the results indicate that TA is more prevalent in companies that are operated by family CEOs. Thus, Hypothesis 1 can be accepted.

To investigate the moderating effect of board independence on the association between family CEO and TA, CEO_FAM × BIND has been generated. The findings show that CEO_FAM × BIND is positively associated with ETR and CFTR at 1% significance level and negatively associated with BTD at 5% significance level. Such a reversal in relationships indicates that board independence can prohibit family CEO’s engagement in TA practice. Thus, Hypothesis 2 can also be accepted.

Among the board-specific control variables, only the BSIZE is found to have a negative association with TA in all three models. Larger boards can provide better monitoring and control, which can prevent TA (Minnick and Noga, 2010). Among the firm-specific control variables, the study found a positive association between ROA and TA in the first two models. Highly profitable firms engage in TA to reduce tax payments (Lanis and Richardson, 2012). The study also found a positive association between LEV and TA in all three models. Higher leverage allows firms to benefit from tax-deductible interest payments, which can reduce taxable income and thus lower overall tax liabilities (Richardson et al., 2015).

Although the results of models 1–3 show a positive and significant association between CEP_FAM and TA and the existence of a moderating effect of BIND on such a relationship, the GLS model does not address the endogeneity problem of the variables. In order to address this issue, GMM, a dynamic panel data estimator, has been used (Table 6). The study has selected the two-step system GMM over the difference GMM, primarily due to its utilization of optimal weighting matrices. The findings are consistent with the findings of the GLS method in Table 5. This confirms the engagement of family CEOs in TA practice. Regarding the moderating effect of BIND, the results are also found to be consistent with the GLS method, which confirms that board independence decreases the tendency of family CEO’s engagement in corporate TA. The reported diagnostic tests confirm the robustness and validity of the estimated models. The significance of the F-statistic indicates that the models are jointly well-specified. The rejection of the null hypothesis for first-order autocorrelation (AR1) coupled with the failure to reject the null for second-order autocorrelation (AR2) supports the presence of first-order serial correlation while affirming the absence of second-order correlation, a necessary condition for consistent estimation. Furthermore, the Hansen test’s insignificant result validates the instruments used, confirming they are exogenous and correctly excluded from the main equation.

The findings reveal that the presence of a family CEO is positively associated with TA, indicating that family-controlled firms in Bangladesh are more likely to engage in strategies that minimize their tax liabilities. This is consistent with existing literature that suggests family firms often prioritize the preservation and accumulation of family wealth and engage in TA (Cheng, 2014; Gaaya et al., 2017; Ibrahim et al., 2021; Kovermann and Wendt, 2019; Niu et al., 2024). Family CEOs, who typically hold significant ownership stakes and personal ties to the firm, may have both the incentive and the ability to adopt aggressive TA strategies that reduce the firm’s tax burden, thereby maximizing the financial benefits to the family. The propensity of family CEOs to engage in TA may be particularly noticeable in Bangladesh, where family ownership is widespread and corporate governance institutions remain less mature than in more developed markets. From the perspective of agency theory, this behavior reflects a Type II agency problem in which controlling families may prioritize their own interests over those of minority shareholders. TA thus emerges as a mechanism through which family CEOs transfer benefits to the family, potentially at the expense of other stakeholders, by exposing the firm to legal, financial, and reputational risks. Moreover, the socio-cultural context of Bangladesh, characterized by strong family loyalty and patronage networks, further reinforces this behavior. In such an environment, ensuring the family’s financial security often outweighs the pursuit of long-term corporate sustainability, especially in the presence of weak enforcement and opaque governance practices.

The study also finds that board independence significantly moderates the relationship between family CEOs and TA. This suggests that independent directors play an important role in mitigating the opportunistic behavior of family CEOs, thereby constraining aggressive tax practices. This finding aligns with prior evidence that stronger board independence enhances monitoring and reduces TA (Armstrong et al., 2015; Butt and Ahmed, 2021; Salihu and Kawi, 2021). In the Bangladeshi context, where governance mechanisms are often undermined by family dominance, the presence of genuinely independent directors becomes particularly valuable for ensuring transparency and accountability. Independent directors, being less entangled with family interests, are more inclined to safeguard the rights of all shareholders rather than favoring controlling families.

Recent research in related governance domains provides further support for the broader monitoring role of board independence. Saleh et al. (2025) demonstrate that board diversity and independence strengthen the role of ESG practices in reducing corporate risk, while Mansour et al. (2025) show that independent boards improve environmental disclosure across ASEAN-5 firms. These findings highlight that independent directors contribute to corporate accountability and sustainability well beyond financial oversight. Similarly, Shubita et al. (2025) emphasize the importance of monitoring investments such as advertising and promotion to ensure they contribute to shareholder value rather than managerial rent-seeking, reinforcing the relevance of independent boards in constraining opportunism. In addition, Khalaf et al. (2024) stress the role of trust and integrity in effective governance, indirectly supporting the argument that independent boards build legitimacy and stakeholder confidence by restraining managerial excesses.

Taken together, the findings underscore the importance of board independence in family-dominated business environments such as Bangladesh. While family CEOs may be inclined toward TA to preserve family wealth, the presence of independent directors can provide an institutional safeguard, ensuring that corporate decisions align with the long-term interests of the firm and its stakeholders rather than short-term family gains.

In order to have deeper insights into the relationships among the variables, the study has also examined the association between politically connected family CEOs and TA. Similar to the primary independent variable, the study created a dummy variable CEO_POL, where 1 denotes a family CEO who is a current or former member of the parliament or top-level government official. Table 7 shows the regression result, where it can be seen that politically connected family CEOs also engage in TA. The coefficients of the models in Table 7 are higher than those of Table 6 which implies that TA practice is even higher in firms led by politically affiliated family CEOs. However, the study found an insignificant role of board independence in moderating the relationship between these variables. This indicates that while independent directors can effectively limit TA practices by family CEOs, their influence diminishes significantly when the family CEOs have political affiliations. In the context of Bangladesh, where political ties often translate into substantial power and influence, family CEOs with such connections may wield considerable control over corporate governance structures. This control can undermine the independence and effectiveness of independent directors, who are tasked with safeguarding the interests of shareholders and ensuring compliance with ethical standards.

The study investigates the association between family CEOs and TA and how board independence can moderate this association. Based on a sample of 100 DSE-listed manufacturing companies of Bangladesh over the 2019–2023 period, the study has found that Bangladeshi companies led by family CEOs tend to avoid taxes more aggressively to reduce their tax liabilities and maximize their personal gains. The existence of type II agency problems due to a lack of minority investors’ protection and a poor governance structure provides opportunities to the family CEOs to avoid taxes and extract private benefits. However, the study also found that a greater level of board independence can prevent family CEOs in their engagement in corporate TA.

This study makes several important contributions to theory. First, examining the relationship between family CEOs and TA in Bangladesh offers new insights into the intersection of agency theory and socioemotional wealth theory in a weak institutional setting. While socioemotional wealth perspectives generally argue that family leaders act conservatively to preserve reputation and legacy (Amin et al., 2025; Amin and Naz, 2025), our findings support the agency-driven view that family CEOs in Bangladesh engage in TA to maximize private benefits. This contrast highlights how the predictive power of SEW theory may be context-contingent, with its relevance weakened in environments where regulatory enforcement is limited and reputational sanctions are less binding. Second, the study provides empirical evidence on the Type II agency problem (principal–principal conflict), showing that controlling families may expropriate minority shareholders through TA strategies. This extends prior agency literature by demonstrating that TA can function not only as a managerial opportunism tool but also as a family opportunism mechanism in emerging economies. Third, the moderating role of board independence empirically validates the monitoring hypothesis of agency theory, underscoring that effective governance mechanisms can constrain opportunism even in settings where formal institutions are weak. Taken together, these findings refine existing theories by emphasizing that family firm behavior is shaped not solely by socioemotional concerns but also by institutional conditions, thereby extending the generalizability of agency and socioemotional wealth perspectives to emerging-market contexts.

From a practical standpoint, the findings carry significant implications for policymakers, practitioners, and investors. For regulators and policymakers, the results emphasize the importance of strengthening board independence in substance, not just in numbers. This requires reforms in the governance codes of Bangladesh, such as more stringent qualification criteria, transparent appointment processes, and safeguards against long tenures that may erode independence. Furthermore, our findings support the introduction of tax governance disclosures, where firms would be required to report on board-level oversight of tax planning and risk management. Such disclosures would increase transparency and reduce information asymmetry between controlling families and outside investors. For corporate practitioners, particularly board members and executives, the study highlights the need to integrate tax risk into broader governance and compliance frameworks. Independent directors should actively question aggressive tax practices that may generate short-term gains but expose the firm to reputational or legal risks. For investors and institutional stakeholders, the findings provide a clear signal that family CEO-led firms may require closer monitoring. Incorporating governance quality, particularly board independence and CEO identity, into investment screening and engagement strategies could mitigate risks associated with opportunistic TA. Finally, for international development partners and governance reform advocates, the study underscores the role of institutional context: strengthening enforcement mechanisms and investor protection frameworks can shift the balance from opportunistic family behavior toward reputationally driven conservatism, ultimately improving tax compliance and corporate accountability.

This study has several limitations that offer directions for future research. Firstly, the sample is limited to manufacturing firms, restricting its scope. Including financial institutions in future studies could provide more comprehensive results. Secondly, the focus on Bangladesh may affect the generalizability of the findings. Replicating this study in other institutional settings would help assess the universality of the relationships between family CEOs, board independence, and TA. Finally, the study only considered board independence as a moderator. Future research should explore other governance mechanisms, such as audit committee quality or shareholder activism, to gain a fuller understanding of how to mitigate tax avoidance in family-controlled firms.