Integrated Reporting Quality, Tax Avoidance, and Sustainable Development: Evidence from South Africa

Abstract

This study examines the association between Integrated Reporting (IR) quality and tax avoidance among South African listed firms from 2012 to 2021, and whether this relationship differs across the highest and lowest levels of IR quality. The extent to which the adoption of a Combined Assurance (CA) model strengthens the IR monitoring role in reducing tax avoidance, as well as the IR quality link with ESG-related implications of tax avoidance, are also explored. IR quality is directly derived from the EY Excellence in Integrated Reporting Awards ranking. This ranking evaluates firms’ adherence to the IR framework and is thus employed as a comprehensive proxy for IR quality. Tax avoidance is captured through multiple proxies. The main findings reveal no significant overall association between IR quality and tax avoidance, suggesting a decoupling between IR and tax behavior. However, when examining firms at the highest and lowest levels of IR quality, a significant negative relationship emerges only for the top performers (highest IR quality), indicating that IR constrains tax avoidance only when supported by a strong ethical corporate culture. Firms adopting CA exhibit higher tax avoidance, suggesting that IR and CA may be constrained by underlying corporate culture and used symbolically. Higher IR quality is also associated with lower tax avoidance relative to GDP and reduced potential revenue losses relative to government expenditures on education, health, and environmental protection. These findings contribute to the literature on IR, corporate governance, and tax avoidance, while also informing policymakers and regulators on the need to strengthen IR and CA frameworks through enhanced tax transparency requirements, thereby supporting equitable resource mobilization, institutional trust, and long-term sustainable development.

1. Introduction

In the current economy, corporate tax avoidance has become a critical issue for shareholders and various stakeholders (Salhi et al., 2019). Governments have become increasingly alert about tax avoidance, especially with taxes representing the world’s largest source of revenue for governments (Wahab et al., 2022). Such practices have significant implications, as tax revenue is one primary source of financial support for economic, social, and environmental development (Hossain et al., 2024). This is particularly important in developing nations, which account for a considerable share of global tax revenue losses (Cobham & Janský, 2018). In these countries, tax revenues play a crucial role in addressing high levels of poverty, and financing investments in critical areas such as infrastructure, education, and healthcare (Gunn et al., 2020).

From a theoretical perspective, corporate tax avoidance can be understood through the lens of Agency Theory and information asymmetry (Jensen & Meckling, 1976; Jensen, 1986). The separation between ownership and control creates incentives for managers to engage in opportunistic behaviors that may not align with broader stakeholder interests (Jensen & Meckling, 1976). In this context, information asymmetry between managers and stakeholders allows such practices to remain opaque, limiting external monitoring and increasing agency costs (Jensen, 1986). Tax avoidance, in particular, is often characterized by complexity and low transparency, making it difficult for stakeholders to assess firms’ tax strategies and their broader societal implications.

Therefore, engaging in tax avoidance practices may be seen as creating a conflict with a firm’s social responsibility (Wang et al., 2020). This opportunistic behavior undermines and exploits the implicit social contract between corporations and society, thereby causing harm to society (Hoi et al., 2013). This breach of social responsibility leads to public outrage, boycotts, and the reputational damage of tax shaming, as seen in high-profile cases involving Amazon, Starbucks, and Google (Barford & Holt, 2013). In response to these pressures, firms have incorporated social and environmental (ESG) information into their business strategies and operations (Eccles et al., 2014).

Improved transparency, facilitated by reporting regulations and frameworks, can produce significant outcomes by motivating firms to alter and enhance their managerial practices (Cao et al., 2024). Building on this, the present study focuses on Integrated Reporting (IR), an emergent corporate reporting model that combines financial and non-financial information (Eccles & Krzus, 2010). This new reporting paradigm aims to promote greater corporate transparency by providing insights into value creation for all stakeholders over the short, medium, and long term (De Villiers et al., 2017; Wu & Zhou, 2022).

With improved transparency and reduced information asymmetry, IR has the potential to strengthen managerial accountability (IFRS, 2021), as broader transparency exposes managerial actions to greater stakeholder scrutiny and may force alignment with the diverse interests of stakeholders while fostering value creation. Prior research has invoked and provided some evidence about the IR effectiveness in enhancing performance, information quality, and monitoring managerial opportunistic behavior (e.g., Barth et al., 2017; Lee & Yeo, 2016; Muttakin et al., 2020; Obeng et al., 2020; Wu & Zhou, 2022; Zhou et al., 2017). Accordingly, it can be expected that IR may discourage aggressive tax avoidance practices.

However, the extent to which IR can effectively influence corporate tax behavior depends on whether its stakeholder-oriented accountability framework can outweigh the prevailing shareholder- and manager-centric view of tax strategy. Tax avoidance remains appealing to managers because it is often framed as a value-maximizing strategy for shareholders and themselves, particularly when compensation plans designed to incentivize aggressive managerial behavior that aligns managers’ interests with those of shareholders are in place (Armstrong et al., 2012; Hanlon & Slemrod, 2009; Rego & Wilson, 2012; Wang et al., 2020). At the same time, IR’s effectiveness is questioned by critics who claim it prioritizes capital providers over broader societal interests (Abhayawansa et al., 2019; De Villiers & Sharma, 2020; Flower, 2015). It is further argued that IR may be used symbolically, as an impression management tool rather than a substantive change (Ahmed Haji & Anifowose, 2016; Melloni et al., 2017).

Thus, while IR may improve transparency, its ability to curb tax avoidance depends on how firms balance stakeholder pressures against financial incentives. In practice, this tension may lead firms to use IR symbolically to mitigate reputational damage from tax avoidance (Ahmed Haji & Anifowose, 2016). It may also result in organizational decoupling, whereby firms adopt IR symbolically to satisfy institutional expectations and regulatory pressures while continuing to implement tax strategies independently of integrated thinking. Such decoupling reflects a separation between formal reporting structures and operational practices (Meyer & Rowan, 1977), ultimately limiting IR’s capacity to constrain tax avoidance in any substantive way.

This uncertainty highlights the need to investigate the relationship between IR quality and corporate tax avoidance and to assess the broader sustainable development implications of such practices, particularly in contexts where transparency and accountability pressures are heightened through mandatory IR adoption.

In 2011, South Africa, under the King Report on Corporate Governance (King III) became the first country to make IR a requirement for firms listed in the Johannesburg Stock Exchange (JSE), through an initial “apply or explain” approach that later became a mandatory requirement (i.e., an “apply and explain” basis under the King IV, in 2017) (Barth et al., 2017; Zhou et al., 2017; Wang et al., 2020). Although this requirement preceded the publication of the International Integrated Reporting Council (IIRC) Framework in 2013, it was broadly aligned with the emerging principles of IR, which were later formalized by the IIRC.1 South Africa provides a unique setting to study IR, offering a long-term perspective on how IR and managerial decision-making have evolved over time. The country has established itself as a global leader in corporate governance and reporting standards, further reinforcing the relevance of this setting for studying IR adoption and effectiveness (T. G. Hoang et al., 2020; Muttakin et al., 2020).

In this context, this study investigates whether IR quality is associated with tax avoidance among South African listed firms. IR quality is operationalized using a variable directly derived from the Ernst & Young (EY) Excellence in Integrated Reporting Awards ranking (EY, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021), while tax avoidance is measured using three proxies: the effective tax rate, the cash effective tax rate, and the long-run effective tax rate.

Although IR quality is inherently multidimensional, this study conceptualizes it primarily as a proxy for transparency, accountability, and integrated thinking. The EY Excellence in Integrated Reporting Awards ranking captures firms’ overall adherence to the IIRC Framework, including its guiding principles, content elements, and fundamental concepts such as value creation and the use of multiple capitals (EY, 2015, 2016, 2017, 2018, 2019, 2020, 2021). This proxy therefore reflects the extent to which firms provide an integrated view of value creation across short-, medium-, and long-term horizons. By increasing the visibility of corporate activities, higher IR quality is expected to strengthen managerial accountability through greater stakeholder scrutiny. These transparency and accountability attributes are highly pertinent in the context of tax avoidance.

This analysis is particularly relevant given the increasing prevalence of tax minimization strategies in the global corporate landscape (Lanis & Richardson, 2015). The study responds to the call by De Villiers et al. (2017) for further research into the association between IR and tax avoidance. This relationship is also explored across firms with the highest and lowest levels of IR, as the implementation of IR varies significantly among firms (McNally et al., 2017; Lemma et al., 2019), affecting the framework’s ability to achieve transparency and align with stakeholders’ needs. Furthermore, the study examines whether the adoption of a Combined Assurance (CA) model strengthens IR as a monitoring mechanism to mitigate tax avoidance. CA was introduced in South Africa under King III as a tool to enhance transparency alongside IR (IODSA, 2016), potentially strengthening oversight and accountability within corporate reporting practices.

In additional analyses, the study further explores the ESG-related implications of corporate tax avoidance in relation to IR quality, shedding light on the role of transparency and accountability in a mandatory IR setting. This is particularly important because tax revenues are essential for financing sustainable development, especially in developing countries facing substantial revenue losses (Cobham & Janský, 2018; Wahab et al., 2022).

Overall, the findings reveal no significant association between IR quality and corporate tax avoidance among South African listed firms, suggesting potential decoupling between IR practices and tax behavior. However, when examining firms at the extreme levels of IR quality, a significant negative association emerges for those with the highest IR quality, while no such relationship is observed among firms with the lowest IR quality. These results suggest that IR may constrain tax avoidance only when implemented at high quality. Further analysis suggests that firms adopting a CA model may still engage in higher tax avoidance, consistent with the view that governance mechanisms such as IR and CA may be limited by underlying corporate culture (Bird & Davis-Nozemack, 2018).

Extending the analysis, we conclude that higher IR quality is also associated with a lower economic magnitude of tax avoidance relative to GDP and with reduced potential revenue losses relative to government expenditures on education, health, and environmental protection, indicating that high-quality IR may indirectly support sustainable development.

This study contributes to the literature in several ways. First, it adds to the growing body of accounting research on IR by providing evidence that advances the understanding of the implications of IR, as prior studies primarily focused on the adoption and drivers of IR, as well as the relevance of IR to capital markets (e.g., Zhou et al., 2017; Leukhardt et al., 2022; Rossignoli et al., 2022). Second, few studies have explored the relationship between IR and opportunistic managerial decisions, particularly in the context of corporate tax avoidance. While prior research has examined the link between IR and earnings management (e.g., T. G. Hoang et al., 2020; Obeng et al., 2020; Shirabe & Nakano, 2022; Wu & Zhou, 2022), studies directly investigating the IR–tax avoidance relationship are very scarce. To the best of our knowledge, Donkor et al. (2022) and Fourati et al. (2025) represent notable exceptions and provide inconsistent evidence. This study, therefore, helps to fill an important gap in the literature. Third, the study also complements the growing literature on CA as a credibility-enhancing mechanism (e.g., Boujelben & Medhioub, 2025; Decaux & Sarens, 2015; H. Hoang & Phang, 2021; Donkor et al., 2021). Fourth, it advances the literature by examining IR quality (rather than mere adoption), by using the independently evaluated EY Excellence in Integrated Reporting Awards ranking. This externally validated measure reduces subjectivity, enhances comparability across firms, and enables more robust statistical inferences (Barth et al., 2017; De Villiers et al., 2017; Wang et al., 2020; Wu & Zhou, 2022). Fourth, exploring the South African setting is also important to the literature: (1) prior research mainly focuses on the role of disclosure regulations in developed markets (Leuz & Wysocki, 2016); (2) evaluating the IR-tax avoidance relationship is relevant in developing countries, where a substantial portion of global tax revenue losses occur (Cobham & Janský, 2018); (3) South Africa is the only country that has mandated the implementation of IR for listed firms for more than a decade. Donkor et al. (2022) and Fourati et al. (2025) explored the IR-tax avoidance relationship with data on South African listed firms from 2011 to 2017 and 2016 to 2021, respectively. In contrast, our dataset spans a full decade (2012–2021), allowing us to capture variations in IR quality over time and providing a long-term perspective on the IR and tax avoidance relationship in South Africa. Finally, the study extends the literature by examining the link between IR quality and the broader ESG-related impact of tax avoidance.

The findings of this study challenge the common assumption that higher IR quality necessarily improves managerial decision-making and promotes responsible governance. Instead, they support the decoupling hypothesis by showing that IR quality and tax avoidance can remain disconnected, while also revealing that IR constrains tax avoidance only when its core principles are embedded within a strong ethical corporate culture—namely, at the highest levels of IR quality. By examining the broader implications of tax avoidance in terms of the potential erosion of public resources relative to GDP and government expenditures on education, health, and environmental protection, the analysis provides novel evidence that IR may indirectly support sustainable development by encouraging more responsible tax behavior. These findings offer important insights for policymakers, regulators, and practitioners by highlighting the need to more explicitly incorporate stronger tax transparency requirements into the IR and CA frameworks. Doing so would enable corporate reporting practices and governance mechanisms to better support equitable resource mobilization, strengthen institutional trust, and contribute to long-term sustainable development.

The remainder of this study is structured as follows. Section 2 reflects the existing literature, focusing on tax avoidance, IR, and their relationship. Section 3 describes the sample and research methodology. Section 4 presents and discusses the empirical results, including sensitivity and additional analysis. Section 5 concludes the study.

2. Literature Review and Hypotheses Development

2.1. Tax Avoidance: Definitions, Risks, and Contradictory Outcomes

Previous literature provides a comprehensive definition of tax avoidance, describing it as the downward management of taxable income through various tax planning activities, which may include legal strategies, those operating in a gray area, and potentially illegal actions (Abdelfattah & Aboud, 2020; Lanis & Richardson, 2018; Watson, 2015).

Although there is no unanimous definition of tax avoidance, it is generally viewed as a continuum of tax planning strategies. At one end, there are legal, low-risk strategies that reduce taxes while maintaining compliance; as one moves along the continuum, the strategies become progressively more aggressive, approaching illegality and tax evasion (Hanlon & Heitzman, 2010; Lisowsky et al., 2013). Wang et al. (2020) explain that tax planning involves reducing the explicit tax burden in compliance with tax laws by strategically designing investment and business activities, while tax evasion occurs when firms minimize their tax burden by violating tax laws.

This study follows Dyreng et al. (2008) in defining tax avoidance as any practice that reduces the tax burden, given that empirical studies cannot clearly distinguish between tax avoidance and tax evasion practices. The primary difference between these practices lies in legal technicalities, related to tax systems, and loopholes; both have similar effects on the taxes collected by tax authorities (Dyreng et al., 2008).

Tax avoidance typically involves a firm retaining cash within the firm instead of being remitted to the government (Wang et al., 2020). This practice can undermine the role of tax payments, as they are considered a basic social obligation of taxpayers and are often used to improve social welfare (Sikka, 2010). Consequently, the public expects firms to pay their fair share of taxes to support the community (Lanis & Richardson, 2018).

Yet when firms fail to meet these expectations, the evidence shows that news of firms engaging in tax avoidance practices often leads to a decline in their market value (Hanlon & Slemrod, 2009). The public backlash against such practices, when exposed, can also bring substantial reputational and financial damages to the firms, as demonstrated by the high-profile Panama Paper scandal, where the implicated firms lost around $174 billion in market value (O’Donovan et al., 2019).

Therefore, firms carefully consider the risks associated with tax avoidance when determining their tax strategies (Austin & Wilson, 2017; Graham et al., 2014), as they are concerned about potential reputational damage. The survey conducted by Graham et al. (2014) indicates that 69 percent of executives considered reputation as a critical variable when deciding which tax planning strategies to avoid. Furthermore, Austin and Wilson (2017) provide evidence showing that firms with valuable brands tend to be more cautious about their tax planning and maintain higher effective tax rates compared to firms with less valuable brands.

While firms carefully weigh the reputational risks of tax avoidance when designing their tax strategies (Austin & Wilson, 2017; Graham et al., 2014), empirical evidence suggests that these concerns do not always materialize into tangible adverse outcomes. For instance, Gallemore et al. (2014) examine the potential negative consequences for the firm’s managers. They find no evidence of increased chief executive officer (CEO), chief financial officer (CFO), or auditor turnover in the three years following the tax avoidance revelation relative to the turnover rates for matched control firms. Their findings show that tax avoidance also does not cause significant effects on loss of sales, increased advertising costs, or heightened media scrutiny. Additionally, Chyz and Gaertner (2018) found that CEOs are more likely to be dismissed when their firms incur relatively high tax burdens compared to their peers. These contradictory outcomes can be attributed to the complex nature of tax avoidance, whereby the successful reduction in tax liabilities directly lowers operational costs and enhances shareholder value (Huseynov & Klamm, 2012; R. Chen et al., 2022), which can offset potential negative repercussions for both the firm and its leadership.

This perspective is further supported by the fact that the primary benefit of tax avoidance is the maximization of firm value through reduced tax payments (Hanlon & Slemrod, 2009). Shareholders might view efforts by management to minimize taxes as essential, particularly since governments typically claim a substantial share of a firm’s pre-tax profits—around one-third—thus providing shareholders with a reasonable justification for the use of tax avoidance strategies (S. Chen et al., 2010).

In this context, firms may tie management’s compensation to aggressive tax planning, directly linking managerial incentives and tax minimization efforts. Thus, compensation plans are designed to incentivize aggressive managerial behavior that aligns managers’ and shareholders’ interests (Armstrong et al., 2012; Gaertner, 2014; Rego & Wilson, 2012).

Rego and Wilson (2012) argue that firms employ equity-based compensation to align managers’ interests with those of shareholders and that equity risk incentives encourage managers to pursue risky yet value-enhancing tax-saving strategies. Using a sample of S&P 500 and S&P 400 firms, they concluded that equity risk incentives significantly drive corporate tax aggressiveness, leading to increased tax minimization efforts.

Nevertheless, this approach may drive managers towards more aggressive tax strategies that surpass shareholder preferences, thus creating a conflict of interest between shareholders and managers (Armstrong et al., 2015; Chyz & Gaertner, 2018). Armstrong et al. (2015) empirically substantiate this concern, demonstrating that risk-taking equity incentives are positively associated with tax avoidance, particularly at the upper end of the tax avoidance spectrum. This evidence suggests that high-risk equity incentives can drive managers to adopt extreme tax strategies beyond the shareholders’ tolerance, suggesting that compensation structures meant to solve agency problems may inadvertently create new ones. This managerial overreach emerges from organizational culture dynamics. Bird and Davis-Nozemack (2018) argue that persistent tax avoidance practices may foster a culture of diminished respect for rules, potentially influencing decision-making processes across the organization. This cultural shift helps explain why tax avoidance, once initiated, often escalates beyond optimal thresholds. In line with this, Abdelfattah and Aboud (2020) note that such behavior may arise from either the desire to maximize shareholder value or managerial opportunism. The latter is particularly concerning, as the complex and opaque nature of tax avoidance transactions enables managers to pursue personal gains with limited detection or accountability (Hossain et al., 2024; Minnick & Noga, 2010).

Although tax avoidance does not automatically mean the existence of an agency problem, as managers are usually expected to prioritize shareholder value maximization (Hanlon & Heitzman, 2010), there is increasing concern about how managers may leverage tax avoidance strategies to serve their interests (Armstrong et al., 2015; Crocker & Slemrod, 2005; Desai & Dharmapala, 2006, 2009; Desai et al., 2007). Specifically, the opaque and complex nature of many tax avoidance arrangements can create opportunities for managers to obscure their actions and divert corporate resources for personal benefit (Desai, 2005; Desai & Dharmapala, 2009). Crocker and Slemrod (2005) further emphasize this by noting that tax decisions are made by managers rather than shareholders and how shareholders lack the information needed to assess the financial and legal implications of tax avoidance strategies.

Such strategies increase information asymmetry, making it more difficult and costly for shareholders and external monitors to oversee managerial behavior effectively (Wang et al., 2020). Empirical evidence by Kim et al. (2011) demonstrates that aggressive tax planning is associated with a greater tendency to withhold or delay the disclosure of bad news, thereby increasing the firm’s exposure to stock price crashes, supporting the contention that aggressive tax planning provides managers with a means to conceal negative performance and increase information asymmetry. This occurs when the accumulation of bad news passes a threshold, and the information is revealed to the market at once, leading to a large negative drop in stock price (Habib & Hasan, 2017).

Considering these risks, shareholders may prefer more conservative tax planning strategies. Supporting this view, Desai et al. (2007) provide evidence from Russian oil firms, where shareholders seem to favor limiting tax avoidance. Their findings show that firms, subject to increased tax enforcement, experience higher market valuations, suggesting that investors respond positively to efforts to curb tax avoidance. Chyz and Gaertner (2018) extend this logic to executive accountability in the U.S. context, revealing that excessive tax avoidance, as evidenced by low tax payments, increases the risk of forced CEO turnover. Furthermore, Desai and Dharmapala (2006) provide evidence that proper incentive alignment can reinforce this preference for tax planning conservatism. Their findings indicate that high-powered executive compensation is associated with lower levels of tax avoidance, suggesting that shareholders may intentionally structure executive compensation to discourage aggressive tax avoidance practices.

The tax avoidance literature presents a fundamental tension: while tax avoidance can enhance shareholder value through reduced tax burdens (Hanlon & Slemrod, 2009; R. Chen et al., 2022), it simultaneously introduces substantial risks and damages (Hanlon & Slemrod, 2009; O’Donovan et al., 2019). One key risk lies in its tendency to increase information asymmetry (Wang et al., 2020; Kim et al., 2011), making it more difficult for investors to assess managerial actions. This opacity can exacerbate agency conflicts, whereby managers exploit tax avoidance to pursue personal interests at the expense of shareholders (Desai & Dharmapala, 2006; Armstrong et al., 2015). These concerns underscore the importance of robust corporate governance mechanisms, including the effectiveness of monitoring practices by an independent board of directors, internal audit departments, and external auditors, to align the interests of managers, shareholders, and society at large (Eugster et al., 2024; Kovermann & Velte, 2019).

2.2. The Association Between Integrated Reporting and Tax Avoidance

The governance challenges surrounding tax avoidance, particularly the risks of information asymmetry and agency conflicts, highlight the critical need for mechanisms that enhance transparency and accountability. In this context, enhanced transparency, introduced by financial and non-financial disclosure regulations, is expected to produce significant outcomes by encouraging firms to adjust their managerial practices and providing external parties with the ability to more effectively monitor corporate performance (Cao et al., 2024). Leuz and Wysocki (2016) discuss how disclosure regulations have real effects on firm behavior. This effect is closely tied to the idea that a more robust information environment reduces information asymmetry between managers and stakeholders, thereby improving external monitoring (Kanodia & Sapra, 2016; Leuz & Wysocki, 2016).

This study focuses on IR, an emerging reporting framework that (arguably) enhances transparency and accountability (IIRC, 2013). IR represents an evolution in corporate disclosure practices, with the potential to bring about significant changes to the current reporting norms (Higgins et al., 2014). Companies are increasingly adopting the IR framework, as it offers deeper insights that go beyond mere financial statements (Cohen & Karatzimas, 2015). Its uniqueness lies in its ability to integrate both financial and non-financial disclosures into a new holistic report (Churet & Eccles, 2014).

This shift represents a significant advancement in corporate reporting, responding to the demands of various stakeholders. The IIRC (2013) (now part of the IFRS foundation) describes IR as a process based on comprehensive thinking that concisely presents an organization’s strategy, governance, performance, and prospects. Its design aims to enhance the information available to external stakeholders while simultaneously improving managerial decision-making processes (Eccles & Krzus, 2010).

IR adopts a principles-based approach focused on explaining how organizations create value over time through the integration of financial and non-financial information. In contrast to more recent ESG regulatory frameworks, IR emphasizes connectivity, strategic orientation, and integrated thinking rather than standardized metrics. By comparison, the European Union’s Corporate Sustainability Reporting Directive (CSRD) and the associated European Sustainability Reporting Standards (ESRS) introduce a more prescriptive model requiring detailed, comparable disclosures and the application of double materiality, thereby broadening the focus beyond enterprise value to include societal and environmental impacts (Ruohonen & Kullas, 2024). These approaches can be viewed as complementary, with IR providing an integrative and narrative perspective on value creation, while CSRD and ESRS enhance comparability and accountability. This distinction is particularly relevant when interpreting IR quality measures, which capture integrative and narrative aspects of reporting that may not be fully reflected in standardized ESG metrics.

IR catalyzes organizations to adopt a holistic thinking perspective. According to Eccles and Krzus (2010), this approach becomes more apparent when considering the concerns of all stakeholders. As IR embodies the concept of integrated thinking, the deeper an organization embraces this mindset, the more seamlessly relevant information from various aspects of the business will flow into its reporting, analysis, and decision-making processes, ultimately contributing to corporate value creation (IFRS, 2021). In this context, embedding integrated thinking into corporate practices ensures that stakeholder needs are naturally incorporated into business operations (IIRC, 2013).

Previous studies, mainly from South Africa, have linked IR to various financial indicators. The main findings indicate that firm value, stock liquidity, cost of capital, and expected future cash flow tend to improve with IR quality (e.g., Barth et al., 2017; S. Chouaibi et al., 2022; Lee & Yeo, 2016; Muttakin et al., 2020; Raimo et al., 2022; Zhou et al., 2017). IR can also enhance managerial decision-making and help control earnings management (Obeng et al., 2020; Shirabe & Nakano, 2022; Wu & Zhou, 2022). Moreover, Zhou et al. (2017) find that better alignment of integrated reports reduces analyst forecast error and mitigates information risk, suggesting that IR can enhance the reporting environment of firms.

Overall, prior research suggests that IR functions as a mechanism that promotes transparent reporting. Fung (2013) describes transparency as a direct solution to the challenges organizations face by providing information that external stakeholders need to evaluate performance and ensure accountability. Consistent with the core ideas of Cao et al. (2024) and Leuz and Wysocki (2016) on how disclosure regulations can significantly alter the relationship between organizations and stakeholders, IR appears to enhance corporate transparency.

In this context, the improved transparency fostered by IR adoption may discourage companies from engaging in tax avoidance strategies, making such practices more easily detectable by stakeholders. According to Balakrishnan et al. (2019) and Zhang et al. (2022), tax avoidance tends to increase opacity and undermine corporate transparency. Drawing on IR assumptions, greater transparency should limit managers’ ability to pursue aggressive tax strategies for personal or short-term gains, as stakeholders gain better access to the information necessary to scrutinize such behaviors. Research by Venter et al. (2017) found that IR is positively associated with the transparency of tax disclosures.

Also, Segal et al. (2017) point out how taxation is perceived as a significant factor that should be integrated into an organization’s core principles and included in IR by South African firms. Under this reasoning, companies with higher levels of IR quality should exhibit significantly lower levels of tax avoidance.

The studies of Donkor et al. (2022) and Fourati et al. (2025) examine the relationship between IR and tax avoidance in South Africa, reporting mixed findings. Donkor et al. (2022) used content analysis to evaluate the IR quality of South African listed companies and captured tax avoidance practices by comparing the statutory tax rate to the effective tax rate of those firms from 2011 to 2017. They conclude that higher IR quality is associated with lower levels of corporate tax avoidance among JSE-listed firms. However, they also note that while increased transparency due to higher IR quality reduces tax avoidance, firm complexity weakens the significant negative relationship between IR quality and corporate tax avoidance practices. Fourati et al. (2025) investigate the IR quality and tax avoidance link, and the moderating effect of board gender diversity on the IR quality tax avoidance nexus. They used data from South African listed firms from 2016 to 2021 and the effective tax rate as a measure of tax avoidance. IR quality is computed through the rankings issued in the annual EY Excellence in Integrated Reporting Awards. The empirical results indicate a positive and non-significant impact of IR quality on tax avoidance, while also revealing a negative and statistically significant effect of IR quality on tax avoidance in the presence of a high percentage of women on the board of directors.

Notwithstanding the proclaimed benefits of IR, namely as a mechanism for transparency and accountability, researchers are skeptical of the usefulness of IR (e.g., Ahmed Haji & Anifowose, 2016; Brown & Dillard, 2014; Flower, 2015; McNally et al., 2017; Lemma et al., 2019). Ahmed Haji and Anifowose (2016) and McNally et al. (2017) argue that IR may have a limited effect on decision-making and reporting practices. Despite the holistic nature of IR, researchers suggest that managers remain primarily driven by short-term goals to maximize shareholder returns and managerial self-interest (Higgins et al., 2014; Stubbs & Higgins, 2014; Flower, 2015). Furthermore, integrated thinking, as defined in the IR framework, links governance, performance, and strategy to drive organizational change, and its implementation is expected to deepen process understanding, promote cross-departmental dialogue, and shape organizational culture by breaking down silos (Dumay et al., 2017; Dumay & Dai, 2017). However, embedding integrated thinking is challenging in organizations with deeply entrenched cultures, and, so, the effectiveness of IR depends on whether existing cultural norms allow for behavioral change; if IR fails to permeate day-to-day operations, it risks becoming counterproductive to building a truly transparent and well-managed business (Dumay & Dai, 2017). Therefore, IR can be used at a symbolic level to conceal managerial opportunistic actions.

The decisions regarding both IR and tax avoidance may also be unrelated. In this context, the true effectiveness of IR may not be realized, as it may not be reflected in the firm’s actual practices, but rather be driven by pressures from regulatory bodies and the capital markets. Maclean and Behnam (2010) emphasize that firms frequently face challenges in complying with externally imposed regulatory requirements, especially when these contradict profit-maximizing activities. Tax management serves as a good example of this tension.

Moreover, adopting an integrated approach to taxation may compel managers to assess various trade-offs in their decision-making processes (Venter et al., 2017). This also adds to the tension, given how managers need to plan the firm’s taxes properly to respond to IR principles or pursue maximizing shareholder value. This conflict may result in a form of organizational decoupling, where formal reporting frameworks do not lead to substantive behavioral change. Decoupling can be an approach that allows managers to mitigate stakeholder conflicts when firms face conflicting stakeholder demands (George et al., 2006).

The decoupling hypothesis suggests that organizations respond to institutional pressures by adopting outward-facing policies and practices that align with societal expectations, even if these initiatives do not fully reflect their internal operations, ultimately aiming to preserve legitimacy (Meyer & Rowan, 1977). Bromley and Powell (2012) further point out that decoupling happens when policies in organizations are implemented, yet they are disconnected from the core operations. They also explain that this type of decoupling is significant when organizations are trying to fulfill different objectives. This form of decoupling is particularly relevant in the context of IR, which promotes a holistic approach requiring firms to address multiple, and sometimes conflicting, stakeholder interests.

Bromley et al. (2012) also emphasize that the decoupling approach is used by firms to respond to institutional pressures, reflecting variations in their capacity, willingness, or alignment with imposed policies. This phenomenon is particularly pronounced in the context of IR in South Africa, where, in response to regulatory pressures to mandatorily adopt IR, listed firms often present vague integrated reports, suggesting a kind of symbolic compliance and also that IR may often be used as a tool to enhance corporate legitimacy (Ahmed Haji & Anifowose, 2016; Brown & Dillard, 2014; Stubbs & Higgins, 2014).

Ahmed Haji and Anifowose (2016) conducted a comprehensive analysis of 246 integrated reports from South African firms from 2011 and 2013. Their findings indicate that IR is used at a symbolic level and primarily for legitimacy purposes. Brown and Dillard (2014) and Stubbs and Higgins (2014) further express concerns about the inconsistency of information that companies disclose in integrated reports. The main issue lies in the significant ambiguity surrounding the implementation of IR, as managers interpret and apply the framework according to their perceptions (Lai et al., 2018). Given IR’s principle-based approach (IIRC, 2013), its effectiveness depends not only on applying the guidelines but also on how thoroughly companies embrace the underlying principles and integrate “integrated thinking” into corporate decision-making and operations (De Villiers & Sharma, 2020). Zhou et al. (2017) highlight that the “apply or explain” basis of IR (only later transformed into an “apply and explain” basis under the King IV, in 2017) means full alignment with the framework could take several years, leading to considerable variations in the IR quality among South African firms (McNally et al., 2017; Steenkamp & Roberts, 2022).

Decoupling scenarios have been observed in corporate social responsibility (CSR) practices, where companies may fail to integrate tax policies into their CSR practices, leading to a lack of correlation between the two areas (Mayberry & Watson, 2021). Firms seem to separate their tax avoidance and CSR efforts, aiming to maximize firm value in each area independently. By doing so, firms seek to reduce tax payments while maintaining the perception that they are paying their fair share of taxes (Mayberry & Watson, 2021).

In sum, empirical evidence directly examining the relationship between IR quality and corporate tax avoidance remains limited and inconclusive. Existing studies, primarily focused on South Africa, report mixed findings: some document a negative association that is weakened by firm complexity (e.g., Donkor et al., 2022), while others find no significant overall relationship but identify a negative effect only under specific governance conditions, such as higher board gender diversity (e.g., Fourati et al., 2025). Given the inconclusive empirical evidence and competing theoretical arguments supporting both a positive and a negative association between IR and tax avoidance, as well as perspectives suggesting no significant relationship between the two constructs, we formulate the following hypothesis under the null form:

H1. 

Tax avoidance is not associated with IR quality.

The relationship between IR and tax avoidance may exist in specific settings. Mayberry and Watson (2021) argue that the absence of evidence for a relationship in a sample does not necessarily imply its nonexistence, suggesting that such a relationship may emerge within relevant subsamples. This suggestion is confirmed by Marques et al. (2024), who find evidence of a relationship between CSR and tax avoidance in cases characterized by extreme CSR engagement and aggressive tax behavior. Similarly, Hoi et al. (2013) and Davis et al. (2016) report significant associations between tax avoidance and CSR, specifically in settings of pronounced CSR activity.

A plausible explanation for these findings lies in the role of organizational culture. As Hoi et al. (2013) suggest, firms with a culture of unethical behavior tend to display consistent patterns of misconduct across different domains. In such firms, unethical decision-making is not isolated but rather embedded within the organization’s strategic orientation and managerial values. Within this context, tax avoidance may foster an organizational environment characterized by a weaker commitment to rules and regulatory compliance. As noted by Bird and Davis-Nozemack (2018), tax avoidance can shape corporate culture by eroding norms of integrity. This effect may not be confined to tax-related practices but can extend to other areas of the firm’s operations (Hoi et al., 2013), suggesting a reinforcing cycle between corporate culture and unethical behavior. Therefore, a culture of non-compliance may foster a non-substantive adoption of IR, weakening its intended governance role in enhancing transparency and accountability.

A lack of substantive implementation of IR as a truly holistic framework implies that it may be formally adopted while remaining disconnected from core organizational practices (Bromley & Powell, 2012; De Villiers & Sharma, 2020). As discussed earlier, this implementation gap aligns with broader concerns regarding the symbolic adoption of IR, as prior evidence suggests that many firms adopt IR primarily for legitimacy purposes rather than to drive meaningful organizational change (e.g., Ahmed Haji & Anifowose, 2016; McNally et al., 2017; Segal et al., 2017; Stubbs & Higgins, 2014). Such patterns are often associated with managerial decisions driven by short-term objectives and opportunistic behaviors (Flower, 2015; Higgins et al., 2014; Stubbs & Higgins, 2014).

In contrast, in firms with a mindset that embraces transparency and accountability, ethical behaviors are expected to spread across all areas of the firm. In this sense, De Villiers and Sharma (2020) highlight that IR effectiveness relies not just on adherence to its formal guidelines, but also on the extent to which firms internalize its core principles and embed “integrated thinking” throughout their strategic and operational decision-making, that is, on the way IR is embedded in the company’s culture.

If substantially adopted, IR links governance, strategy, and performance, fostering integrated thinking, to support organizational transformation through a more comprehensive understanding of processes, enhanced cross-functional interaction, and improved decision-making (Dumay et al., 2017; Dumay & Dai, 2017). It promotes a holistic representation of how organizations create value over time (De Villiers et al., 2017; Eccles & Krzus, 2010; Wu & Zhou, 2022).

By enhancing transparency and reducing information asymmetry, IR can reinforce managerial accountability by increasing stakeholder oversight and encouraging alignment with broader interests (IFRS, 2021). Prior evidence suggests that IR contributes to improved performance, higher information quality, and greater oversight of managerial opportunism (e.g., Barth et al., 2017; Lee & Yeo, 2016; Muttakin et al., 2020; Obeng et al., 2020; Wu & Zhou, 2022; Zhou et al., 2017). Therefore, the more deeply IR is embedded within corporate culture, the more effective it is expected to be in constraining opportunistic behaviors.

In sum, the substantive adoption of IR may be constrained in organizations with deeply entrenched cultures, as its effectiveness depends on the extent to which existing norms enable genuine behavioral change. Otherwise, IR may remain largely symbolic and fall short of improving transparency and managerial practices (Dumay & Dai, 2017). Therefore, firms with low IR quality are expected to exhibit higher levels of managerial opportunism, including tax avoidance practices, whereas firms with high IR quality are more likely to exhibit lower levels of tax avoidance.

We focus on the comparison between the lowest and highest levels of IR quality to capture potential non-linear effects on tax avoidance. This approach is consistent with Mayberry and Watson (2021), who argue that heterogeneity in the full sample can obscure meaningful relationships that become more visible within more homogeneous subsamples. Importantly, this prediction does not contradict the absence of an association in the full sample, but rather suggests that such an association may be conditional and observable only at extreme levels of IR quality. Accordingly, we examine whether the relationship between IR quality and tax avoidance differs across firms with the highest and lowest levels of IR quality and formulate the following hypothesis in its alternative (non-directional) form:

H2. 

The association between IR quality and tax avoidance differs between firms with the highest and lowest levels of IR quality.

2.3. IR Quality and Tax Avoidance: The Role of Combined Assurance

A substantial body of literature has examined how corporate governance mechanisms—such as board and audit committee characteristics, external auditing, and ownership structure—affect the quality of IR (e.g., Y. Chouaibi et al., 2022; Dragomir & Dumitru, 2023; Erin & Adegboye, 2022; Permatasari & Tjahjadi, 2023). However, limited attention has been devoted to the role of CA in shaping IR quality.

The emergence of CA is related to IR, and its prominence has expanded in parallel with the growing adoption of IR practices (H. Hoang & Phang, 2021). Since its formal introduction in the South African King III governance report in 2009 (IODSA, 2009), CA has been promoted alongside IR as a complementary mechanism for enhancing corporate transparency and accountability. CA is based on the implementation of three core lines of defense (IIA, 2022; Boujelben & Medhioub, 2025; H. Hoang & Phang, 2021). The first line of defense bears primary responsibility for identifying and mitigating risks related to daily operational processes (IIA, 2020). The second line of defense is responsible for risk management oversight, ensuring compliance with regulatory requirements while also encompassing critical functions such as human resources, cybersecurity, data privacy, and ESG reporting (IIA, 2020, 2022). Moving to the third line of defense, the internal audit function, through its independent review, provides assurance and advice on the appropriateness and efficacy of governance structures and risk management frameworks (IIA, 2020). Additionally, external auditors are also viewed as an external line of defense (Boujelben & Medhioub, 2025; H. Hoang & Phang, 2021).

By coordinating the three lines of defense, CA delivers holistic assurance to the board on the effectiveness of risk management and internal control systems, thereby mitigating the inefficiencies and gaps associated with fragmented assurance practices (Decaux & Sarens, 2015). In this sense, CA represents a comprehensive approach that may solve problems associated with the normal assurance of the IR framework adoption. Consistent with this rationale, the King III report recommends that companies adopt CA to strengthen the control environment and enhance the credibility and transparency of corporate disclosures, and therefore the quality of IR (IODSA, 2016).

Previous studies on CA provide evidence that it can improve corporate governance and compliance and reporting quality, and reduce analysts’ forecast errors (Donkor et al., 2021; Zhou et al., 2017). Using a sample of South African listed firms, Donkor et al. (2021) conclude that CA quality is significantly related to both IR quality and sustainability reporting quality. Moreover, in the context of tax avoidance, Boujelben and Medhioub (2025) reported a negative association between CA and tax avoidance levels. Thus, CA seems to increase transparency and reduce agency problems.

However, in terms of financial reporting quality, Donkor et al. (2021) show that while CA quality is significantly and negatively related to real earnings management, it is not significantly associated with accrual-based earnings management. Conversely, the presence of CA demonstrates a significant negative relationship with accrual-based earnings management but no significant effect on REM. Furthermore, Prinsloo and Maroun (2021) argue that CA models are often developed conservatively and tend to be guided by a limited range of assurance methodologies, rather than employing a pluralistic verification framework for integrated and sustainability reporting. As a result, the scope of CA is frequently constrained, and empirical evidence on its broader effectiveness remains limited.

From an agency perspective, CA can be viewed as a governance mechanism that enhances monitoring and reduces information asymmetry by coordinating multiple assurance providers across the three lines of defense. By strengthening internal controls, risk oversight, and the credibility of disclosures, CA is expected to improve the effectiveness of IR in enhancing transparency and accountability. In this sense, CA may reinforce IR’s ability to constrain managerial opportunistic behavior, including corporate tax avoidance.

Prior evidence provides some support for this expectation, showing that CA is associated with improved reporting quality and governance outcomes (e.g., Donkor et al., 2021; Zhou et al., 2017), as well as lower levels of tax avoidance (Boujelben & Medhioub, 2025). However, the effectiveness of CA appears to be context-dependent, with mixed findings regarding its impact on different dimensions of financial reporting quality (Donkor et al., 2021), and concerns that its implementation may be limited in scope or largely symbolic (Prinsloo & Maroun, 2021).

Taken together, these arguments suggest that CA may not directly constrain tax avoidance in all cases, but rather enhance the effectiveness of IR as a transparency and accountability mechanism. Accordingly, we formulate the following hypothesis in its alternative form:

H3. 

The presence of CA strengthens the negative association between IR quality and tax avoidance.

Overall, the emerging literature on IR quality and tax avoidance remains underdeveloped and provides limited and inconclusive evidence. Prior studies offer limited insight into how IR quality varies across firms, as well as how extreme levels of IR quality relate to tax avoidance behavior. Moreover, the role of additional governance mechanisms, such as CA, in shaping this relationship remains largely unexplored.

This study addresses these gaps by providing a more granular examination of the association between IR quality and tax avoidance. Specifically, it captures variation in IR quality across firms using a multi-level measure, considers the role of extreme levels of IR quality, and incorporates CA as a governance mechanism that may influence the effectiveness of IR in constraining tax avoidance.

3. Research Methodology

The study aims to examine the association between IR quality and tax avoidance practices among JSE-listed firms. The following subsections outline the empirical strategy employed to address this research objective.

3.1. Regression Model

The research hypotheses are examined by using the following model:

TaxAv_i,t = α + β₁ IRQ_i,t + β₂ SIZE _i,t + β₃ ROA _i,t + β₄ MTB _i,t + β₅ LVG _i,t + β₆ IIN _i,t
+ β₇ CIN _i,t + β₈ NOL _i,t + β₉ IndM _i,t +β₁₀ AEM _i,t + β₁₁ Big4 _i,t + Industry FE + Year FE + ε_i,t

where TaxAv, represents a proxy for tax avoidance; IRQ is the IR quality; and SIZE, ROA, MTB, LVG, IIN, CIN, NOL, IndM, Big4, and AEM are a set of control variables that, according to prior studies, are determinants of tax avoidance.

3.1.1. Measurement of Tax Avoidance

The prior literature presents several proxies to measure tax avoidance, most of which capture nonconforming tax avoidance. This type of avoidance involves strategies that reduce tax liabilities without directly lowering book income. Although nonconforming measures have limitations, as they do not capture tax planning strategies that simultaneously decrease both tax expenses and book income, we adopt them in the present study for several reasons. First, nonconforming tax avoidance is a significant, if not predominant, tax planning strategy (Badertscher et al., 2019). Second, we are still far from reaching a consensus on how to measure conforming tax avoidance, making it difficult to rely on a single, standardized approach. Finally, firms engaging in nonconforming tax avoidance are also likely to pursue conforming tax avoidance, making the former a reasonable proxy for overall tax avoidance by firms.

Accordingly, the dependent variable for tax avoidance is measured using the effective tax rate (ETR) and the cash effective tax rate (CETR). The ETR captures both current tax expenses and deferred taxes, while the CETR reflects the amount of tax paid (Hanlon & Heitzman, 2010). Additionally, a long-term measure of cash effective tax rate (LETR) is also employed to reduce the year-to-year volatility associated with the annual measures (Hanlon & Heitzman, 2010).

As noted by Lanis and Richardson (2013), the ETR reflects a firm’s ability to reduce tax payments relative to pre-tax income, with lower ETR commonly interpreted as evidence of tax avoidance.

The basic form of annual ETR is commonly employed to measure tax avoidance and is computed based on the information disclosed by firms in their financial statements (Austin & Wilson, 2017; Dyreng et al., 2008). The ETR, as shown in expression (1) below, is calculated as the ratio of total tax expense to pre-tax book income. The tax expense reflected in the ETR represents the total tax liability for the current year, regardless of the timing of the actual payment (Wang et al., 2020). Thus, the ETR is influenced by both tax avoidance practices and financial accounting regulations. The formula of the ETR for firm i at time t is:

$${ETR}_{it}= {\displaystyle \frac{{Total tax expense}_{it}}{{Pre\text{-}tax income}_{it}}}$$

(1)

Building on previous research, this study adopts the cash-effective tax rate (CETR) as a key measure of tax avoidance, recognizing its ability to address the limitations of the ETR highlighted in the literature (Austin & Wilson, 2017; Dyreng et al., 2008). As the ETR includes current and deferred taxes, Dyreng et al. (2008) note that strategies such as accelerating deductions and deferring income for tax purposes constitute tax avoidance practices that the ETR might overlook. Also, using CETR as a tax avoidance measure can help mitigate the effects of accrual earnings management, which may influence the ETR. Yet, the CETR has limitations: cash taxes paid can be inconsistent due to timing differences and be related to a previous period, while the pre-tax income must be related to the current year. Nevertheless, Watson (2015) suggests that the CETR is widely regarded in the accounting literature as a reliable proxy for measuring tax avoidance. The CETR2 for firm i at time t is calculated as follows:

$${CETR}_{it}= {\displaystyle \frac{{Cash tax paid}_{it}}{{Pre\text{-}tax income}_{it} }}$$

(2)

A long-term measure is additionally employed to capture tax avoidance, considering an average of three consecutive years of cash ETR, consistent with Austin and Wilson (2017). Hanlon and Heitzman (2010) argue that this approach mitigates the year-to-year volatility associated with the annual ETR measures. By providing a broader view of a firm’s tax expenses over an extended period, it reduces potential noise associated with annual fluctuations. Long-run cash ETR (LETR) proposed by Dyreng et al. (2008) is calculated by summing a firm’s total cash tax paid over three years and dividing this by the total pre-tax income over the same period (excluding the effects of special items).3

In line with previous studies, all tax avoidance proxies are truncated to the 0–1 range (e.g., Abdelfattah & Aboud, 2020; Austin & Wilson, 2017; Dyreng et al., 2008; Lanis & Richardson, 2013). As larger effective tax rates values originally reflect lower corporate tax avoidance, to facilitate interpretation, the ETR proxies are multiplied by −1, so that higher values now indicate higher levels of corporate tax avoidance.

3.1.2. Integrated Reporting Quality

The explanatory variable is the IR quality (IRQ), which is directly operationalized with data from the annual EY’s Excellence in Integrated Reporting Awards rankings from 2012 to 2021 of the top 100 JSE-listed companies (EY, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021).

Starting from 2010, all JSE-listed firms were required to issue an integrated report, following the King III guidelines. As a result, EY began publishing an annual excellence report in 2011, in which they evaluated the top 100 firms listed on the JSE in terms of IR quality. These companies collectively account for approximately 90% of the total market value of the JSE, as highlighted by Barth et al. (2017).

EY’s assessment of IR quality is based on a structured mark plan and an independent adjudication process. The mark plan is developed by three academic adjudicators affiliated with the University of Cape Town, in collaboration with EY’s Professional Practice Group, and is explicitly aligned with the principles of the IIRC Framework (EY, 2015, 2016, 2017, 2018, 2019, 2020, 2021).

According to the EY reports (EY, 2015, 2016, 2017, 2018, 2019, 2020, 2021), the mark plan evaluates integrated reports along two main dimensions: the IIRC Framework’s guiding principles and content elements. Evaluation and scoring are carried out independently by the academic adjudicators. A standardized scale is used to assign scores to seven guiding principles (strategic focus and future orientation, connectivity of information, stakeholder relationships, materiality, conciseness, reliability and completeness, and consistency and comparability) as well as to eight content elements (organizational overview and external environment, governance, business model, risks and opportunities, strategy and resource allocation, performance, outlook, and basis of preparation and presentation). The evaluation also considers the IIRC Framework’s fundamental concepts, particularly value creation and the use of multiple capitals (EY, 2015, 2016, 2017, 2018, 2019, 2020, 2021). As Bondar et al. (2024) recognize, the assessment of IR quality is a complex process.

Following the individual assessments, scores assigned to the guiding principles, content elements, and fundamental concepts are aggregated and combined with adjudicators’ overall impressions and recommended rankings. The final ranking is determined through a process that integrates quantitative scores with qualitative judgment and deliberation among adjudicators, ensuring consistency and reducing subjectivity (EY, 2015, 2016, 2017, 2018, 2019, 2020, 2021). Based on the outcome of this process (i.e., the overall score), each company is assigned to one of five categories, “Top 10”, “Excellent”, “Good”, “Average”, and “Progress to be Made”, which reflect decreasing levels of adherence to the principles of IR (EY, 2015, 2016, 2017, 2018, 2019, 2020, 2021).

The scores assigned to guiding principles, content elements, and the fundamental concepts of the IIRC Framework are not publicly available; only the IR quality categories are available. However, Barth et al. (2017) had proprietary access to the underlying EY score sheets used by the adjudicators and found them to be consistent with the IIRC Framework. Accordingly, the EY Excellence in Integrated Reporting Awards scoring system provides a granular and informative assessment of IR and is considered a comprehensive and credible proxy for IR quality (Barth et al., 2017; De Villiers et al., 2017).

In this study, the EY Excellence in Integrated Reporting Awards categories are converted into a categorical variable ranging from 1 to 5 (IRQ), where higher values indicate higher IR quality. In short, for companies classified as: “Top 10”, IRQ = 5; “Excellent”, IRQ = 4; “Good”, IRQ = 3; “Average”, IRQ = 2; “Progress to be Made”, IRQ = 1. This approach was also used by Wu and Zhou (2022) and Eloff and Steenkamp (2022).

3.1.3. Control Variables

In addition to the variables of interest, we include a comprehensive set of control variables consistent with the established literature (e.g., Davis et al., 2016; Donkor et al., 2021; Hoi et al., 2013; Lanis & Richardson, 2012, 2015, 2018). These controls account for both firm-specific characteristics and governance factors that may influence the effective tax rate.

SIZE is measured by the natural logarithm of total assets. While larger firms have greater opportunities for tax avoidance, they may also face significant reputational damage from engaging in such practices (Rego, 2003; Minnick & Noga, 2010). ROA is included to account for profitability and computed as pre-tax income before extraordinary items divided by lagged total assets, given its significant influence on tax avoidance (Davis et al., 2016; Donkor et al., 2022; Lanis & Richardson, 2018; Minnick & Noga, 2010). Following Davis et al. (2016) and Lanis and Richardson (2012), market-to-book ratio (MTB) is added, as firms experiencing rapid growth are more likely to make investments in tax-favored assets (S. Chen et al., 2010). Leverage (LVG) is calculated as total debt divided by total assets, and captures the potential tax benefit associated with debt financing, commonly referred to as debt tax shields (Lanis & Richardson, 2018; Zeng, 2019). Furthermore, we control for asset structure through two key intensity measures: intangible intensity (IIN) and capital intensity (CIN). These variables capture distinct tax planning opportunities, with IIN calculated as intangibles divided by lagged total assets (Donkor et al., 2022; Hoi et al., 2013; Lanis & Richardson, 2012), while CIN is measured using net property, plant and equipment (PPE) divided by lagged total assets (Davis et al., 2016; Hoi et al., 2013). Both measures are relevant as firms with higher asset intensity typically possess greater flexibility in implementing tax reduction strategies (e.g., accelerated depreciations, tax incentives). Additionally, we control for firms with net operating losses (NOL) carryforwards, as these allow firms to offset future taxable income, decreasing the effective tax rate (Austin & Wilson, 2017; Lanis & Richardson, 2018; Mayberry & Watson, 2021). This variable is included as a dummy equal to one if a firm reports a negative net income in the previous year, and zero otherwise.

To account for the effects of corporate governance on tax avoidance, independent board members (IndM) are taken into consideration and calculated as a percentage of independent board members relative to the total number of board members on the firm’s board. The board of directors is the highest internal control mechanism for monitoring the actions of top management and is expected to influence tax-related strategies (Lanis & Richardson, 2012). Furthermore, being audited by a Big Four auditor (Big4) is considered, as it may help to reduce tax avoidance via enhanced monitoring and higher quality audit (Lanis & Richardson, 2015).

To ensure that the observed relationship between IR quality and tax avoidance is not driven by earnings management, we control for accrual-based earnings management (AEM). AEM refers to the use of discretionary accounting accruals to influence reported earnings without affecting actual cash flows and is widely recognized as being closely linked to tax avoidance (Kim et al., 2011; Hoi et al., 2013; Lanis & Richardson, 2018). It is measured using the residuals from the modified Jones model (Dechow et al., 1995), where larger residuals indicate higher levels of AEM.

Finally, industry- and year-fixed effects are included in the model to control for differences in effective tax rates across industries and time-specific shocks.

All variables are defined in Appendix A.

3.2. Sample and Data

This study investigates the association between tax avoidance and IRQ using data from JSE-listed firms from 2012 to 2021. The sample is drawn from the EY Excellence in IR Report, a ranking of the top 100 JSE-listed firms based on their IRQ, provided by Ernst and Young.

As a start, firms’ rankings were collected from the EY’s Excellence in Integrated Reporting Awards (EY, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021), yielding an initial dataset of 1000 observations. In line with prior research, only non-financial firms were retained, resulting in the exclusion of 345 observations pertaining to banks and financial institutions. Financial data and additional relevant information for the remaining 655 firms were gathered from the Refinitiv database. Firms lacking complete financial data were subsequently excluded from the sample. The final sample size varies between 400 and 411 firm-year observations, depending on the specific measure of tax avoidance used.

Table 1 reports descriptive statistics for the sample variables. All the continuous independent variables are winsorized at the 1st and 99th percentiles to mitigate the influence of outliers. The IRQ score has an average of 2.886 and a standard deviation of 1.177, reflecting the variation in IR quality within the sample. Firms exhibit a mean of ETR and CETR of 0.316 and 0.299, respectively, along with an average LETR of 0.315. Further exploration of the sample characteristics reveals that the firms have an average of 156 million in their total assets. The profitability of these firms, as reflected by the mean of ROA, is 12% which suggests a solid level of financial performance for firms in South Africa, along with an average MTB of 2.852. The sample has an average leverage of 27%, with an intangible asset ratio and a capital intensity ratio of 17% and 40%, respectively. Additionally, a minority of firms reported NOL during the previous year, amounting to 3%, and the average AEM in the firms is 0.04. Finally, regarding the governance indicators, on average, 61% of board members are independent, contributing to a balanced and accountable governance framework, and 96% of the firms are audited by Big 4 auditors.

Table 1. Descriptive statistics.

Variable	N	Mean	SD	Min	Max
ETR	406	0.316	0.132	0.001	0.958
CETR	400	0.299	0.143	0.001	0.964
LETR	411	0.315	0.145	0.001	0.927
IRQ	411	2.886	1.177	1.000	5.000
Size	411	17.541	1.399	14.220	21.317
ROA	411	0.124	0.109	−0.093	0.688
MTB	411	2.852	2.222	−1.826	10.918
LVG	411	0.271	0.185	0.003	1.119
IIN	411	0.173	0.198	0.000	0.940
CINT	411	0.403	0.245	0.016	1.437
NOL	411	0.034	0.181	0.000	1.000
IndBM	411	0.618	0.153	0.250	0.923
AEM	411	0.043	0.041	0.000	0.218
Big4	411	0.968	0.175	0.000	1.000

The sample period is 2012–2021. N is the number of firm-year observations. SD is the standard deviation. Descriptive statistics for dependent and independent variables are presented. Descriptive statistics of independent variables are based on the largest sample (LETR). All continuous variables are winsorized at the 1st and 99th percentiles except for ETR, CETR and LETR, which are restricted to fall in the interval of 0 and 1. All variables are described in Appendix A.

Table 2 presents the Pearson correlations between all the variables. Pairwise correlations among the independent variables are below 0.5, suggesting that multicollinearity is unlikely to affect the main model. The tax avoidance measures (ETR, CETR, and LETR) are positively and significantly correlated, indicating consistency across the tax avoidance metrics. Yet, the correlations between IRQ and the tax proxies are weak and statistically insignificant, suggesting no straightforward bivariate relationship between IR quality and tax avoidance.

Table 2. Correlation matrix.

	ETR	CETR	LETR	IRQ	SIZE	ROA	MTB	LVG	IIN	CINT	NOL	IndBM	AEM	Big4
ETR	1.0000
CETR	0.4830 ***	1.0000
LETR	0.3570 ***	0.5765 ***	1.0000
IRQ	0.0195	−0.0126	0.0088	1.0000
SIZE	−0.1358 ***	−0.0425	−0.0639	0.1611 ***	1.0000
ROA	0.2453 ***	0.2097 ***	0.2694 ***	0.0600	−0.2167 ***	1.0000
MTB	0.0623	−0.0157	0.0565	0.0476	0.2165 ***	0.3870 ***	1.0000
LVG	0.1076 **	0.0746	0.0669	0.2144 ***	0.1968 ***	0.0362	0.0324	1.0000
IINT	0.1551 **	0.0944 *	0.1013 **	−0.0325	0.1088 **	0.1400 ***	−0.0168	0.4497 ***	1.0000
CINT	−0.0433	0.1390 ***	0.1046 **	0.2314 ***	0.2108 ***	0.0213	−0.1430 ***	0.2876 ***	−0.3308 ***	1.0000
NOL	−0.0112	−0.0347	−0.0340	−0.0502	0.0496	−0.0421 **	−0.0440	−0.0936 *	−0.1105 **	−0.1651 ***	1.0000
IndBM	−0.1356 ***	−0.1230 **	−0.1724 ***	0.0033	0.3112 ***	−0.1410 ***	−0.2298 ***	−0.0172	−0.1117 **	0.1454 ***	0.0121	1.0000
AEM	−0.0536	0.0234	0.0120	0.0678	−0.1627 ***	0.1038 **	0.0824 *	−0.0379	−0.0788	−0.0661	0.0083	−0.1300 **	1.0000
Big4	0.0123	0.0109	0.0158	0.1953 ***	0.2108 ***	−0.0897 *	−0.1271 **	0.0297	−0.1193 **	0.1602 ***	0.0339	0.1959 ***	0.0502	1.000

Table 2 Pearson correlation matrix. This table presents pairwise correlations for the study’s variables. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. All variables are described in Appendix A. All continuous variables are winsorized at both 1% and 99%. The figures are based on the LETR sample. (The coefficients of the correlations are quantitively similar across the models when the tax avoidance measures are changed between ETR, CETR and LETR).

4. Empirical Results and Discussion

4.1. Baseline Results

The results in this section are derived from Ordinary Least Squares (OLS) regressions with robust standard errors to account for heteroskedasticity. The estimations incorporate industry- and time-fixed effects to control for common shocks which could cause cross-sectional correlation in the errors.

Table 3 presents the outcomes of regression analysis conducted to examine the relationship between tax avoidance and the quality of IR. Different dependent variables are employed as proxies for tax avoidance: ETR, CETR, and LETR.

Table 3. Regression of IR quality on tax avoidance.

	Dependent Variable
	(1)		(2)		(3)
	ETR		CETR		LETR
	Coef.	t-Stat.	Coef.	t-Stat.	Coef.	t-Stat.
Explan. variable
IRQ	−0.001	−0.11	−0.011 *	−1.73	−0.001	−0.12
Control variables
Size	−0.007	−1.09	−0.009	−1.24	−0.005	−0.74
ROA	0.319 ***	5.07	0.322 ***	3.94	0.376 ***	5.23
MTB	−0.008 ***	−2.72	−0.006 *	−1.71	−0.004	−1.17
LVG	0.050	1.01	−0.015	−0.29	−0.026	−0.55
IIN	0.008	0.17	0.119 **	2.09	0.147 ***	2.67
CINT	−0.034	−1.05	0.090 **	2.07	0.085 **	2.12
NOL	0.018	0.46	0.028	0.62	0.042	0.78
IndBM	−0.045	−1.20	−0.107 **	−2.27	−0.083 *	−1.64
AEM	−0.164	−1.00	0.101	0.66	0.011	0.07
Big4	−0.007	−0.20	0.022	0.55	0.027	0.78
Industry fixed-effects	Yes		Yes		Yes
Time fixed-effects	Yes		Yes		Yes
R2	0.215		0.163		0.200
N	406		400		411

*, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. All p-values are two-tailed. The OLS regressions are estimated using panel data for the JSE-listed firms over the period 2012–2021. All continuous variables are winsorized at the 1 percent and 99 percent levels. Year and industry dummies are included in each specification and for the sake of brevity, the results for these dummies are not reported. All variables are defined in Appendix A.

As suggested by the correlation analysis, multicollinearity statistics do not reveal distortions of results due to correlation among independent variables. The Variance inflation factor (VIF) for column (1) is 1.99, for column (2) 2.05, and for column (3) 2.02, indicating that multicollinearity is not a significant problem (Studenmund, 2014).

In the first and third specifications, with the ETR and LETR as proxies for tax avoidance, the IRQ coefficient is negative but not statistically significant. Yet, in the second specification, using the CETR, the IRQ coefficient is negative and statistically significant at the 10% level. Overall, this evidence suggests a limited relationship, indicating that IR quality does not substantially influence managerial tax avoidance behavior among South African listed companies. These findings contradict those of Donkor et al. (2022), who conclude that higher IR quality reduces opportunistic tax avoidance among South African listed firms. However, they are consistent with Fourati et al. (2025), who report a positive but statistically insignificant association between IR quality and tax avoidance in the same context.

Donkor et al. (2022) examine the IR-tax avoidance relationship using data on South African listed firms from 2011 to 2017, while Fourati et al. (2025) analyze the period 2016–2021. By contrast, our dataset spans a full decade (2012–2021) following the adoption of IR in South Africa, enabling a more comprehensive assessment of variations in IR quality over time and a longer-term perspective on its relationship with tax avoidance. This extended timeframe may help explain the divergence between our findings and those of Donkor et al. (2022).

The findings support H1, and reveal that IR quality and tax avoidance are, overall, unrelated among JSE-listed firms. The decoupling phenomenon between IR quality and tax behavior seems to prevail among these firms. This observed separation aligns with Meyer and Rowan’s (1977) argument that organizations often separate formal structures from operational practices to manage conflicting institutional demands. Such decoupling may reflect firms’ efforts to optimize IR quality and tax avoidance independently. It also underscores a gap in implementing IR as a truly holistic framework—IR is adopted formally but remains disconnected from core operations (Bromley & Powell, 2012). If South African listed firms fully embraced IR and its underlying principle of integrated thinking, greater transparency would be encouraged, and IR could act as a constraint on aggressive tax avoidance practices.

The observed decoupling between IR and tax practices can be understood as an organizational response to conflicting stakeholder demands, as theorized by Meyer and Rowan (1977). In the context of tax avoidance, this decoupling may stem from the tension between shareholder value maximization and broader societal expectations for corporate tax contributions (Abdelfattah & Aboud, 2020). Faced with these contradictory pressures, firms adopt decoupling strategies. This pattern has been noted in the context of IR, which has itself been criticized for prioritizing the interests of capital providers while marginalizing broader societal concerns (Brown & Dillard, 2014; Flower, 2015).

The findings highlight the lack of substantive alignment between the disclosed IR principles and actual tax strategies. This aligns with Segal et al. (2017), who observed that tax risk management is often underdeveloped in IR disclosures, with most firms treating it as a compliance exercise rather than a strategic opportunity. Such an implementation gap reflects broader concerns about the symbolic nature of IR adoption, supported by evidence that many firms pursue IR primarily for legitimacy purposes rather than to drive meaningful organizational change (Ahmed Haji & Anifowose, 2016; McNally et al., 2017; Stubbs & Higgins, 2014).

Regarding control variables, we only find significant coefficients for ROA, IIN, CIN, MTB, and IndBM. Consistent with prior research, the analysis reveals a positive relationship between ROA and tax avoidance in the three specifications, indicating that profitable companies are more likely to engage in tax avoidance (e.g., Dyreng et al., 2008; Lanis & Richardson, 2012; Zeng, 2019). Likewise, both IIN and CIN show positive associations with the CETR and LETR. The significance of IIN comes in line with the findings of Zeng (2019), highlighting how firms with more intangibles engage more in tax-aggressive behavior. In line with S. Chen et al.’s (2010) arguments, CIN enables tax avoidance, partly because of depreciation policies which create discretionary opportunities for firms to manage taxes. In the opposite way, MTB shows a significantly negative coefficient in the first and second specifications, indicating that firms with high growth opportunities avoid aggressive tax planning. Finally, the negative association of independent board members (IndBM) emphasizes the positive influence of independent board oversight in mitigating tax avoidance practices (Lanis & Richardson, 2012).

To address the second research hypothesis, we analyze firms with the highest and lowest IRQ levels. This comparison allows us to examine whether the association between IR quality and tax avoidance varies according to the extent of a firm’s commitment to high-quality IR practices. In other words, it provides deeper insights into whether high-quality IR has a more pronounced impact on mitigating tax avoidance behavior, and whether low-quality IR fails to generate such outcomes. Dummy variables are used to identify those firms in the highest and lowest levels of IRQ: IRQ_highest is a dummy equal to 1 if a firm is in the highest level of IRQ (IRQ = 5), and 0 otherwise; IRQ_lowest is a dummy equal to 1 if a firm is in the lowest level of IRQ (IRQ = 1), and 0 otherwise.

As presented in Table 4, the coefficients for IRQ_highest are negative and statistically significant in both the CETR and LETR specifications, while remaining insignificant for the ETR specification. The results indicate that superior IR quality is associated with reduced tax avoidance. Conversely, the consistently insignificant coefficients for IRQ_lowest across all specifications indicate that low-quality IR is not associated with tax behavior. These results support H2.

Table 4. Analysis of extreme cases of IRQ.

	Dependent Variable
	(1)		(2)		(3)
	ETR		CETR		LETR
	Coef.	t-Stat.	Coef.	t-Stat.	Coef.	t-Stat.
Explan. Variables
IRQhighest	−0.027	−1.29	−0.071 ***	−2.79	−0.042 *	−1.90
IRQlowest	−0.003	−0.21	0.011	0.70	−0.012	−0.70
Control variables
Size	−0.004	−0.65	−0.003	−0.41	−0.002	−0.24
ROA	0.410 ***	5.28	0.406 ***	4.51	0.413 ***	5.54
MTB	−0.012 ***	−2.73	−0.004	−1.28	−0.004	−1.00
LVG	0.044	0.87	−0.070	−1.27	−0.060	−1.11
IIN	−0.006	−0.12	0.083	1.24	0.131 **	2.26
CINT	−0.058 *	−1.62 *	0.069	1.56	0.051	1.21
NOL	−0.027	−0.57	0.014	0.30	0.028	0.47
IndBM	−0.074 *	−1.80 *	−0.114 **	−2.25	−0.161 ***	−3.17
AEM	−0.194	−1.27	−0.041	−0.23	−0.213	−1.44
Big4	−0.006	−0.22	0.052 *	1.65	0.067 *	1.92
Industry fixed-effects	Yes		Yes		Yes
Time fixed-effects	Yes		Yes		Yes
R2	0.207		0.163		0.209
N	406		400		411

*, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. All p-values are two-tailed. The OLS regressions are estimated using panel data for the JSE-listed firms over the period 2012–2021. All continuous variables are winsorized at the 1 percent and 99 percent levels. Year and industry dummies are included in each specification and for the sake of brevity, the results for these dummies are not reported. All variables are defined in Appendix A.

We interpret this evidence as suggesting that IR constrains tax avoidance only when the framework is adopted and implemented within firms characterized by a strong ethical organizational culture, thereby fostering holistic decision-making and enhancing corporate transparency (Dumay et al., 2017; Dumay & Dai, 2017). Consistent with this view, it has been argued that because IR reflects the principle of integrated thinking, the deeper an organization adopts this approach, the more effectively information from diverse areas of the business is incorporated into reporting, analysis, and decision-making, ultimately contributing to corporate value creation (IFRS, 2021). Embedding integrated thinking into organizational processes facilitates the natural consideration of stakeholder interests within everyday business activities (IIRC, 2013). As De Villiers and Sharma (2020) emphasize, the effectiveness of IR depends not only on formal adherence to its guidelines but also on the extent to which firms internalize its principles and embed integrated thinking into strategic and operational decision-making. When firms achieve this, IR’s stakeholder-oriented accountability framework can help shift tax strategies away from shareholder- and manager-centric views (Armstrong et al., 2012; Rego & Wilson, 2012; Wang et al., 2020).

The lack of a full-sample effect of IR in constraining tax avoidance may not be evidence of IR’s failure, but is instead consistent with the known challenges in terms of full IR adoption (Dumay et al., 2017; Higgins et al., 2014). According to Zhou et al. (2017), IR was mandated in South Africa in 2011 on an “apply or explain” basis and only in 2017 became a mandatory requirement (i.e., with an “apply and explain” basis), and therefore, full alignment with the framework may take several years, leading to significant variation in IR quality across firms. This suggests the existence of a threshold level of IR quality of reporting quality—what we term deep, substantive adoption—above which IR is able to constrain tax avoidance and fulfill its objectives of promoting transparency and accountability. Methodologically, this finding supports Mayberry and Watson’s (2021) argument that significant relationships can be obscured in full-sample analyses and may only emerge when examined within subsamples.

Regarding the third hypothesis, we explore whether the adoption of CA can strengthen the IR quality’s ability to mitigate tax avoidance.

We split the sample into adaptors and non-adaptors of CA, and we re-run the baseline models reported in Table 3. Consistent with Boujelben and Medhioub (2025), the identification of CA adoption followed a two-step procedure. First, firm disclosures were searched within integrated reports for relevant keywords, including “combined assurance,” “three lines of defense,” “lines of defense model,” and “three lines of control.” Second, a close textual analysis of the accompanying narratives was conducted to verify the presence of the three key components of a CA model. If the three lines of defense were detected, CA models were identified and the firm was considered to have adopted the CA model.

Following this classification, the baseline regression models (reported in Table 3) were re-estimated separately for the adopters and non-adopters of CA. This split-sample approach enables a direct comparison of whether the relationship between IR quality and tax avoidance differs depending on the presence of a CA model. For brevity, the results for non-adopters are omitted, as they show no significant relationship between IR quality and any measure of tax avoidance (ETR, CETR, LETR), a finding which robustly supports the decoupling hypothesis of the main findings.

As reported in Table 5, the results for the sample of CA adopters reveal a significant and positive association between IRQ and the three proxies of tax avoidance, leading, therefore, to the rejection of H3. This suggests that CA may be strategically leveraged to enhance the appearance of transparency and accountability, while enabling the continuation of opportunistic practices of tax avoidance. The combined use of IR and CA may reflect a legitimacy-seeking strategy: firms project the adoption of governance and assurance mechanisms to external stakeholders while simultaneously exploiting this credibility shield to conceal aggressive tax behavior. This dynamic highlights tensions between stakeholder-oriented and shareholder- and manager-driven incentives that frame tax avoidance as value-maximizing activity (Armstrong et al., 2012; Rego & Wilson, 2012; Wang et al., 2020).

Table 5. IR quality in firms with CA.

	Dependent Variable
	(1)		(2)		(3)
	ETR		CETR		LETR
	Coef.	t-Stat.	Coef.	t-Stat.	Coef.	t-Stat.
Explan. Variable
IRQ	0.023 **	2.43	0.016 *	1.64	0.018 *	1.87
Control variables	Yes		Yes		Yes
Industry fixed-effects	Yes		Yes		Yes
Time fixed-effects	Yes		Yes		Yes
R2	0.309		0.191		0.296
N	174		171		185

* and ** indicate statistical significance at the 10% and 5%, respectively. All p-values are two-tailed. The OLS regressions are estimated using panel data for the JSE-listed firms over the period 2012–2021. All continuous variables are winsorized at the 1 percent and 99 percent levels. Year and industry dummies are included in each specification; for brevity, the results for these dummies are not reported. All variables are defined in Appendix A.

The evidence supports a skeptical view of CA’s effectiveness. As Prinsloo and Maroun (2021) argue, the technically conservative and narrow scope design of prevailing CA models may limit their ability to detect or prevent complex, opportunistic behaviors such as tax avoidance. This interpretation is consistent with Donkor et al.’s (2021) findings, which report conflicting evidence regarding the effectiveness of CA quality and CA presence in curbing earnings management. Their study suggests that while CA may influence certain dimensions of reporting quality, it has limited capacity to address the complexity of earnings management practices.

The ineffectiveness of CA as a tool for corporate governance might also be attributed to corporate culture, as firms engaging in tax avoidance might have developed a culture of disrespecting the rules and regulations. As Bird and Davis-Nozemack (2018) highlight, tax avoidance affects the firm’s culture, ultimately diminishing the culture of integrity. This consequently extends beyond tax avoidance and influences other areas across the firm. Given that corporate culture is an important factor affecting many corporate policies (Hoi et al., 2013), a non-compliance culture can make the CA lose its intended governance role.

The symbolic use of CA and IR may be particularly pronounced in South Africa, where the absence of endorsements of international agreements promoting transparent tax disclosure up until 2023, could increase opportunities for tax avoidance (Barth et al., 2017; Venter et al., 2017).4

4.2. Robustness Analysis

Alternative Tax Avoidance Measures

We check whether the baseline results hold when alternative proxies for tax avoidance are employed. The alternative proxies are computed as the difference between the national statutory tax rate (STR) and both measures of the effective tax rate (ETR and CETR), corresponding to the ETR differential (Hanlon & Heitzman, 2010). The new variables are D_ETR and D_CETR. All things being equal, the wider the spread between the STR and ETR, the greater the tax avoidance. This approach allows a more direct comparison with the findings of Donkor et al. (2022), as discrepancies in results may be attributed to differences in proxies of tax avoidance. Table 6 presents the results using the D_ETR (column 1) and D_CETR (column 2) as the dependent variables.

Table 6. Regression of the ETR Differential on IRQ.

	Dependent Variables
	(1)		(2)
	D_ETR		D_CETR
	Coef.	t-Stat.	Coef.	t-Stat.
Explan. Variable
IRQ	0.008	0.79	0.003	0.30
Control variables
Size	0.035 *	1.65	0.039	1.28
ROA	0.471 ***	4.71	0.602 ***	3.91
MTB	−0.013 **	−2.31	0.001	0.17
LVG	−0.108	−1.48	−0.112 *	−1.68
IIN	−0.055	−0.77	0.069	0.78
CINT	−0.020	−0.22	−0.061	−0.77
NOL	−0.025	−0.81	−0.025	−1.19
SGrth	0.045	1.26	−0.003	−0.05
IndBM	0.033	0.28	−0.029	−0.34
AEM	−0.147	−0.99	0.019	0.13
Industry fixed-effects	Yes		Yes
Time fixed-effects	Yes		Yes
R2	0.156		0.150
N	411		405

*, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. All p-values are two-tailed. The OLS regressions are estimated using panel data for the JSE-listed firms over the period 2012–2021. All continuous variables are winsorized at the 1 percent and 99 percent levels. Year and industry dummies are included in each specification, and for the sake of brevity, the results for these dummies are not reported. All variables are defined in Appendix A.

The estimation shows that IR quality remains statistically insignificant in both specifications. These results are in line with the main findings, indicating that IR is not significantly associated with tax avoidance and suggesting that firms decouple IR from actual tax practices. However, they contradict Donkor et al. (2022), who find a significant negative relationship between IR quality and tax avoidance, though they also note that firm complexity weakens this link.

4.3. Endogeneity

This study acknowledges potential endogeneity threats that could influence the interpretation of the results. However, it argues that the research design substantially mitigates concerns of selection bias. The analysis leverages South Africa’s mandatory IR adoption setting, which inherently limits firms’ ability to self-select into high-quality reporting based on private information (Barth et al., 2017; Wu & Zhou, 2022). Furthermore, the use of an external, third-party quality ranking (EY’s Excellence in Integrated Reporting Awards) provides an objective metric that mitigates subjective measurement error from self-constructed indices, helping reduce concerns about selection bias (Barth et al., 2017).

To address reverse causality concerns, following previous research, we adopt a lead–lag specification by lagging all independent variables by one year (Barth et al., 2017; Dhaliwal et al., 2011; Zhou et al., 2017). We then re-estimate the baseline regressions of IR quality and tax avoidance reported in Section 4.2.

For brevity, in Table 7, we display only the coefficients for IRQ across the three tax avoidance measures. In panel A, the coefficients of IRQ are statistically insignificant when ETR and LETR are used as dependent variables, and a slightly significant negative relationship emerges when CETR is the dependent variable. These findings align with the main findings reported in Table 3. Panel B of Table 7 shows that IRQ_highest has a significantly negative association only with the CETR and LETR, and that IRQ_lowest is not significantly associated with any proxy for tax avoidance. This evidence is consistent with the baseline results reported in Table 4.

Table 7. Endogeneity analysis.

	Dependent Variables
	(1)		(2)		(3)
	ETR		CETR		LETR
	Coef.	t-Stat.	Coef.	t-Stat.	Coef.	t-Stat.
Panel A
Explan. Variable
IRQ	−0.003	−0.57	−0.013 *	−1.65	−0.001	−0.15
Control variables	Yes		Yes		Yes
Industry fixed-effects	Yes		Yes		Yes
Time fixed-effects	Yes		Yes		Yes
R2	0.197		0.122		0.203
N	357		354		365
Panel B
Explan. Variables
IRQhighest	−0.008	−0.33	−0.067 *	−1.79	−0.035 *	−1.42
IRQlowest	−0.010	−0.47	0.010	0.47	−0.020	−0.92
Control variables	Yes		Yes		Yes
Industry fixed-effects	Yes		Yes		Yes
Time fixed-effects	Yes		Yes		Yes
R2	0.199		0.136		0.216
N	357		354		365

* indicates statistical significance at the 10% level. All p-values are two-tailed. The OLS regressions are estimated using panel data for the JSE-listed firms over the period 2012–2021. All continuous variables are winsorized at the 1 percent and 99 percent levels. Year and industry dummies are included in each specification, and for the sake of brevity, the results for these dummies and control variables are not reported. All the variables are those described in Appendix A.

Similar to prior studies (e.g., Barth et al., 2017; Wang et al., 2020; Wu & Zhou, 2022), we note that our methodological choices, while designed to reduce bias, may not fully account for all endogeneity concerns.

4.4. Additional Analysis

IR Quality and ESG Implications of Corporate Tax Avoidance in South Africa

Tax revenue represents one of the primary sources of financial support for economic and social development. Governments rely on tax revenues to finance public goods and services and to support broader economic, social, and environmental development (Hossain et al., 2024). Consequently, corporate tax avoidance practices may have significant societal implications by reducing the fiscal capacity of governments to fund these public investments. From a corporate responsibility perspective, engaging in aggressive tax avoidance may therefore create a conflict with firms’ broader social responsibilities and commitments to sustainable development (Wang et al., 2020).

Building on this perspective, we extend our analysis to examine the relationship between IR quality and the economic, social, and environmental implications of corporate tax avoidance in South Africa.

To capture the macroeconomic relevance of tax avoidance, the tax gap (D_CETR as defined earlier in this study) is measured as a percentage of gross domestic product (GDP), thereby reflecting the potential economic impact of foregone tax revenues at the national level for each firm-year observation.5 The new dependent variable is TAGDP. The societal implications of tax avoidance are captured by expressing the tax gap as a percentage of government expenditure on education and health. The resulting dependent variables, TAGEducation and TAGHealth, measure foregone tax revenues at the national level for each firm-year observation relative to government investment in education and health. Finally, to capture the environmental implications of tax avoidance, the tax gap is expressed as a percentage of government environmental protection expenditure. The resulting dependent variable, TAGEnvironment, measures foregone tax revenues at the national level for each firm-year observation relative to public investment in environmental protection. Data on South Africa’s GDP, education expenditure, and health expenditure are retrieved from the World Bank database, while data on public investment in environmental protection are obtained from the International Monetary Fund (IMF) Climate Change Dashboard.6

IR quality is measured by using the IRQ and the highest and lowest IRQ levels (IRQ_highest and IRQ_lowest).

The empirical results are reported in Table 8. In Panel A, a negative and statistically significant relationship between IRQ and the tax gap expressed as a share of GDP (TAGDP) is observed over the period 2012–2021. This finding suggests that firms with higher-quality IR are associated with lower levels of tax avoidance, thereby reducing the potential economic impact of these practices. Specifically, a one-unit increase in IRQ is associated with a reduction of approximately 47.3 billion Rands in the economic magnitude of tax avoidance (measured relative to GDP), holding other factors constant.

Table 8. IR quality and the ESG implications of corporate tax avoidance.

	Dependent Variables
	(1)		(2)		(3)		(4)
	TAGDP		TAGEducation		TAGHealth		TAGEnvironment
	Coef.	t-Stat.	Coef.	t-Stat.	Coef.	t-Stat.	Coef.	t-Stat.
Panel A
Explan. Variable
IRQ	−4.73 × 1010 *	−1.74	−2.11 × 108 *	−1.64	−3.76 × 109 *	−1.68	−1.34 × 1010 *	−1.69
Control variables	Yes		Yes		Yes		Yes
Industry fixed-effects	Yes		Yes		Yes		Yes
Time fixed-effects	Yes		Yes		Yes		Yes
R2	0.163		0.165		0.164		0.164
N	405		405		405		405
Panel B
Explan. Variables
IRQhighest	−3.35 × 1011 ***	−2.80	−1.60 × 109 ***	−2.77	−2.77 × 1010 ***	−12.78	−9.86 × 1010 ***	−2.81
IRQlowest	−3.13 × 109	−0.05	−9.83 × 107	−0.30	−9.45 × 108	−10.17	−3.12 × 109	−0.16
Control variables	Yes		Yes		Yes		Yes
Industry fixed-effects	Yes		Yes		Yes		Yes
Time fixed-effects	Yes		Yes		Yes		Yes
R2	0.180		0.183		0.182		0.182
N	405		405		405		405

* and *** indicate statistical significance at the 10% and 1% levels, respectively. All p-values are two-tailed. The OLS regressions are estimated using panel data for the JSE-listed firms over the period 2012–2021. All continuous variables are winsorized at the 1 percent and 99 percent levels. Year and industry dummies are included in each specification, and for the sake of brevity, the results for these dummies and control variables are not reported. All the variables are those described in Appendix A.

This result also holds when the tax gap is computed as a percentage of government expenditures on education (TAGEducation), health (TAGHealth), and environmental protection (TAGEnvironment). In these specifications, a one-unit increase in IRQ is associated with reductions in the social and environmental impact of tax avoidance, by approximately 211 million Rands in terms of education expenditures, 3.76 billion Rands of health expenditures, and 13.4 billion Rands of environmental protection expenditures, holding other factors constant. In other words, improved IR quality is linked to less revenue loss that could otherwise fund education, health, and environmental protection.

Panel B of Table 8 further indicates a statistically significant negative association between IRQ_highest and TAGDP, TAGEducation, TAGHealth, and TAGEnvironment, while the relationship with IRQ_lowest remains statistically insignificant.

Taken together, these findings indicate that higher levels of IR quality are associated with lower levels of tax avoidance and, consequently, with a reduced erosion of public resources that could otherwise support economic, social, and environmental investments. This effect appears to be driven primarily by firms exhibiting the highest levels of IR quality, suggesting that IR is most effective when adopted and implemented within organizations characterized by a strong ethical culture that promotes holistic decision-making and greater corporate transparency (De Villiers & Sharma, 2020; Dumay & Dai, 2017; Dumay et al., 2017).

From an ESG perspective, these results highlight the role of IR quality in fostering transparency and accountability, thereby promoting more responsible corporate tax behavior within the broader sustainability agenda. Firms with better IR practices may therefore indirectly support sustainable development by limiting the fiscal impact of tax avoidance and preserving public resources for economic investment and key social expenditures, such as education and healthcare, as well as investment in environmental protection. Overall, these findings reinforce the role of corporate governance and transparency in advancing ESG-related objectives.

5. Conclusions

The purpose of this paper is to examine the association between IR quality and tax avoidance among South African listed firms. The study also explores this relationship across firms at the highest and lowest levels of IR quality. Furthermore, it also investigates whether the adoption of a CA model strengthens the role of IR as a monitoring mechanism for mitigating tax avoidance. Tax avoidance is measured using several proxies, while IR quality is captured directly from the EY Excellence in Integrated Reporting Awards ranking. The analysis is based on a sample of South African listed firms over the period 2012–2021. In additional analyses, the study examines the relationship between IR quality and the ESG-related implications of corporate tax avoidance.

Overall, the findings reveal no significant association between IR quality and tax avoidance practices, suggesting that South African listed firms may decouple their IR practices from their tax behavior. This evidence of decoupling remains robust across alternative measures of tax avoidance. However, at a deeper level, the analysis of the highest and lowest levels of IR quality and tax avoidance provides additional insights into their relationship. When examining subgroups of firms with exceptionally high and exceptionally low IR quality, a significant negative association emerges only for firms with the highest IR quality, while no such relationship is observed for those firms with the lowest IR quality. This pattern suggests that IR can act as a mechanism to constrain tax avoidance, but only in firms that implement the IR framework thoroughly, reliably, and with commitment, and operate within a context characterized by a strong ethical corporate culture. Here, integrated thinking is fully adopted and becomes systematically incorporated into reporting and decision-making processes, thereby enhancing transparency and accountability.

Extending the analysis, our findings suggest that firms adopting a CA model may combine it with IR in a symbolic manner. In particular, the observation that firms with higher IR quality also exhibit higher levels of tax avoidance raises the possibility that IR and CA may be used as a legitimacy tool rather than reflecting substantive changes in underlying practices. This aligns with the notion that corporate culture, particularly one influenced by tax avoidance, can limit the effectiveness of governance tools such as IR and CA in curbing aggressive tax practices (Bird & Davis-Nozemack, 2018).

In additional analysis, the results also indicate that higher IR quality is associated with a lower economic magnitude of tax avoidance relative to GDP and with reduced potential revenue losses within the scope of government expenditures on education, health, and environmental protection. These findings suggest that higher-quality IR may indirectly support sustainable development by limiting the fiscal impact of corporate tax avoidance on public resources.

Overall, we conclude that, in South Africa, only above a certain threshold of IR quality—representing deep and substantive adoption—IR effectively constrains tax avoidance and realizes its intended goals of promoting transparency and accountability. Importantly, the findings further suggest that higher-quality IR indirectly supports sustainable development by reducing the corporate tax avoidance impact on public resources. The results also highlight the limited effectiveness of CA as a mechanism for corporate governance and for improving IR. These findings contribute to the literature on IR, corporate governance, and tax avoidance, as well as on the ESG-related impacts of tax avoidance. They also provide crucial insights for policymakers and regulators seeking to strengthen the IR and CA frameworks, highlighting the need to more explicitly incorporate stronger tax transparency requirements. Doing so would enable corporate reporting practices and governance mechanisms to better support equitable resource mobilization, strengthen institutional trust, and contribute to long-term sustainable development.

The results of this study should be analyzed in light of certain limitations. First, while South Africa offers insights into the effects of mandatory IR effects, the results may not be generalizable to other regulatory contexts. Second, the reliance on EY ratings to measure IR quality restricts the sample to the JSE-listed companies. Third, tax avoidance proxies cannot capture its multifaceted nature, as well as all forms of tax evasion. Finally, the relationship between IR and tax avoidance raises endogeneity concerns that cannot be completely ruled out.

Future research should explore IR in diverse global contexts and explore more deeply how corporate culture dynamics impact its effectiveness, as well as that of CA models. The interaction between IR quality, CA, and the ESG-related impacts of tax avoidance deserves further exploration. Future studies should consider employing proxies for tax avoidance in addition to ETR-based measures and using them together for a more comprehensive approach to tax evasion practices.