Audit quality in the mandatory joint audit setting: a comparative analysis between the developed and developing economies

Concerns about audit quality have prompted regulators to consider mandatory joint audits to enhance the reliability of financial statements (Nurunnabi et al., 2020; Ratzinger-Sakel et al., 2013). While mid-tier auditing firms support joint audits to improve quality, Big Four firms oppose them due to higher costs, and empirical evidence on their efficacy is mixed. On the one hand, some studies demonstrate the benefits of joint audits for audit quality (AbuRaya, 2025; Bianchi, 2018; Ittonen and Trønnes, 2015; Zerni et al., 2012). On the other hand, many others find that joint audits deliver no meaningful improvement over single audits (André et al., 2016; El-Dyasty and Elamer, 2022; Velte and Azibi, 2015) – a pattern that persists even after countries abolished mandatory joint audits in Denmark and Kuwait (Holm and Thinggaard, 2018; Van der Zahn and Tebourbi, 2023). Consequently, researchers still cannot reach a definitive conclusion about the actual impact of joint audits on audit quality.

Deng et al. (2014) emphasized the importance of considering the composition of joint audit pairs. Building on this, many studies have examined how joint audit quality varies across pairings, yielding mixed results. For instance, some found higher quality in pairs of two Big Four firms (Francis et al., 2009; Nurunnabi et al., 2020), suggesting that free-riding concerns exist when a Big Four firm is paired with a non-Big Four firm. Conversely, Paugam and Casta (2012) identified higher audit quality in pairs with one Big Four and one non-Big Four firm, suggesting that direct competition can reduce audit quality. In contrast, André et al. (2016) observed no difference in quality. These conflicting results, along with the standard three-category classification (two Big Four, one Big Four with one non-Big Four and two non-Big Four), indicate that further research is needed with a broader range of joint audit pairs.

Moreover, most previous research on joint audit quality has focused on France or compared it with single-audit systems, rarely examining it in developing economies and the differences between developed and developing contexts (Ali Hassan et al., 2026). Similarly, Hassan et al. (2024) highlighted a scarcity of comparative research on auditing quality between European Union (EU) and non-EU countries. To bridge these gaps, this study compares the quality of mandatory joint audits in France (developed/EU) with those in Tunisia and Morocco (emerging/MENA region) by adopting a six-category pair classification. Moreover, prior research has not examined how joint audit quality varies by company size or how industry specialization affects it. Accordingly, this study also examines the effects of firm size and auditor industry specialization.

To address these gaps, this study examines the following research questions. First, are there significant differences in the quality of mandatory joint audits across joint audit pairs within each country? Second, how does mandatory joint audit quality differ between developed (France) and emerging (Morocco and Tunisia) contexts across various auditor pairs? Third, are there significant differences in the quality of mandatory joint audits across joint audit pairs in emerging contexts (Morocco and Tunisia)? Fourth, does the presence of at least one industry-specialized auditor within a joint audit pair affect the overall mandatory joint audit quality level for that pair? Fifth, does the size of the audited firm affect the level of mandatory joint audit quality between different joint audit pairs?

To address these research questions, we use data from 440 nonfinancial listed companies from 2014 to 2023. These companies are from France (359), Morocco (51) and Tunisia (30). We initiate our analysis with ordinary least squares (OLS) regressions in Stata to investigate variations in mandatory joint audit quality across auditor-pair categories, using absolute abnormal working capital accruals (AWCA) as the primary proxy of audit quality (DeFond and Park, 2001). Moreover, our analyses segment firms into small, medium and large groups and further assess the effect of having at least one industry specialist in each joint audit pair. In line with Van der Zahn and Tebourbi (2023) and Hassan et al. (2026a), we use a six-category auditor-pair model: B4B4, consisting of two Big Four firms; B4S1, comprising one Big Four and one non-Big Four international firm; B4S2, including one Big Four and one local non-Big Four firm; S1S1, featuring two non-Big Four international firms; S1S2, comprising one non-Big Four international and one local non-Big Four firm; and S2S2, consisting of two local non-Big Four firms.

The results of this study have both theoretical and practical implications for understanding audit quality in mandatory joint audit environments. This study is among the first to examine and compare mandatory joint audit quality across a developed economy (France) and emerging markets (Morocco and Tunisia), accounting for both the shared mandatory regimes since the 1990s and key implementation differences, such as auditor rotation, joint audit work division rules and International Financial Reporting Standards (IFRS) adoption. Our approach uses a detailed six-category auditor-pair classification, recent data from 2014 to 2023 and analysis of firm size and auditor industry specialization. The findings indicate that, generally, there are no significant differences in audit quality across most joint audit pairs within and across the three countries, with limited exceptions, and suggest that mandatory joint auditing may reduce variation in audit quality across different pairs. These results support the idea that joint auditing improves audit quality through collaboration, shared responsibility and mutual oversight among auditors. They also emphasize how institutional theory explains convergence in audit quality across joint audit pairs, both within and across countries, despite different accounting standards. This reflects a minimum quality standard, driven by mandatory joint audits across all three countries since the 1990s, that has stabilized audit quality across different pairs. The results further show that audit quality tends to converge even more among different pairs for larger firms due to institutional discipline and unified control.

Based on this framework, the results provide practical recommendations for regulators, policymakers and practitioners. Regulators and policymakers should establish clear guidelines for task allocation within pairs to ensure effective collaboration and reduce knowledge concentration. Specialized training for both local and international auditors should be improved to better match their skills and increase audit quality, especially in small and medium-sized businesses. Transparency, disclosure systems and internal controls should be strengthened across all firm sizes to ensure consistent quality and reduce discrepancies between pairs. Harmonizing best professional practices between local and international firms can minimize gaps caused by differing standards and sustainably enhance audit quality. Finally, policymakers should consider the influence of institutional context and firm size when updating standards or control policies. Such measures can help ensure that regulatory actions effectively promote quality without creating gaps in small firms or emerging markets. The paper is organized as follows: Section 2 offers an overview of auditing markets and regulations. Section 3 presents the theoretical framework. Section 4 reviews the literature and states hypotheses. Section 5 describes the research design and data. Section 6 reports the results. Section 7 includes additional analyses. Section 8 discusses findings, limitations and future research.

This section examines the influence of legal frameworks and national audit regulations on the quality of mandatory joint audits. This study compares mandatory joint audit quality in two distinct regions: France (the only EU country that requires joint audits for nonfinancial firms) and Morocco and Tunisia (the only MENA countries with similar mandates). The regulatory and institutional contexts of the three countries add important nuance to this comparison and point out the importance of the present study. In France, joint audits have been required for listed companies since 1966 (Article 223-3) and for consolidated financial statements since 1984 (Article 823-2) under the Commercial Companies Code (Nekhili et al., 2018). Moroccan law also mandates joint audits for listed companies since 1996 and for banks since 1993 (Ayyash, 2025; Ratzinger-Sakel et al., 2012). According to Article 13 of the Tunisian Commercial Code of 2005, listed banks, insurance companies, companies with consolidated accounts and companies with debts exceeding certain thresholds must undergo joint audits (Abdelmoula, 2023).

Consequently, because France, Morocco and Tunisia have implemented mandatory joint audit requirements since the 1990s, and as both countries have adopted the French public accounting system (Khlif et al., 2020), we anticipate convergence in joint audit practices, which may mitigate significant differences in audit quality across these diverse contexts. However, differences in context across these countries could lead to variations in the quality of mandatory joint audits. For example, unlike France, neither Morocco nor Tunisia regulates the division of labor among joint auditors. Moreover, auditor tenure also differs: French auditors serve six-year terms that can be renewed (Nekhili et al., 2018); Moroccan auditors serve three-year terms that can be renewed once for a total of six years; and Tunisian auditors serve three-year terms that can be renewed up to two times, followed by a mandatory three-year break (Lajmi et al., 2021). Finally, most French companies adopt IFRS. In contrast, Moroccan and Tunisian firms primarily use local standards. These contextual similarities and differences, particularly in implementation details and institutional development, could influence audit practices and the resulting quality of mandatory joint audits. A comparison that accounts for both the shared regulatory tradition and these variations, therefore, fills a critical gap in the literature.

Although joint auditing is mandatory in the three countries studied, companies listed in these countries are free to choose the audit firms that comprise the joint audit pair. The theoretical aspects of joint auditing have not been thoroughly explored (Deng et al., 2014; Zerni et al., 2012). Although agency theory, reputation protection theory and resource dependence theory offer useful background on auditor incentives and resource dynamics in general, their relevance is more conditional in mandatory joint audit settings, where such audits are required by law. In this context, institutional theory provides the most appropriate primary lens because long-standing regulatory requirements and shared institutional pressures are expected to promote convergence across different pair types. This study, therefore, relies primarily on institutional theory, while drawing on the other theories as secondary insights, to examine the relationship between the composition of joint audit pairs and audit quality in two economically, culturally and politically distinct regions: the EU (France) and the MENA region (Tunisia and Morocco).

First, the demand for auditing is traditionally linked to agency theory, which holds that separating ownership from management can create agency problems when managers prioritize their own interests over those of shareholders (Elmashtawy et al., 2024). Therefore, external auditors can reduce conflicts among managers, shareholders and bondholders by verifying financial data (Wallace, 2004). Whereas previous research has used agency theory to partially explain auditor selection, our study uses agency theory to explain how audit quality varies across the different types of audit firms selected to form joint audit pairs. In this context, we expect that joint audit pairs that include at least one of the Big Four are associated with higher audit quality than those that do not, especially in emerging economies characterized by significant information disparities and agency challenges.

Moreover, reputation protection theory offers a different view on auditor behavior, proposing that auditors are more cautious with well-known companies to protect their standing and reduce legal risk (Khoo et al., 2020). Notably, when joint audit pairs include the Big Four, auditors tend to be more cautious, conducting extra tests to reduce audit risk and prevent failure. As a result, audit quality increases compared to joint audit pairs that do not include any Big Four firms. Furthermore, Resource Dependence Theory suggests that organizations rely on resources such as expertise, information and advice to achieve goals and improve performance (Sunny and Apsara, 2026). This theory emphasizes that different resources are required for business success (Taldybayeva, 2025). In the context of auditing, this theory explains how a firm’s reliance on auditors affects audit quality (Taldybayeva, 2025). Thus, in this study, audit firms affiliated with international consortia can leverage global expertise and brand reputation to overcome local institutional weaknesses, particularly in emerging markets (Taldybayeva, 2025). Building on this, an audit pair consisting of two international audit firms, especially those from the Big Four, may have sufficient resources to deliver higher audit quality than other pairs. However, this situation may also create an overreliance by one auditor on the other’s resources, potentially reducing quality. Alternatively, a joint audit pair consisting of an international firm, particularly one of the Big Four, and a local firm may integrate the international auditor’s financial resources with the local auditor’s expertise in local financial standards, potentially resulting in higher audit quality. Yet, it is important to note that the “free rider” problem can arise in such pairs due to the smaller auditor’s overreliance on the expertise and resources of the larger audit firm (Deng et al., 2014).

On the other hand, recently, academics have widely used institutional theory. This theory primarily focuses on understanding the functioning and growth of organizations within larger social and cultural systems (Oliver, 1991). According to this theory, the behavior of organizations is determined by the environment in which they operate, in all its cultural, religious, political and social forms (Krane and Eulerich, 2020). This theory also offers the most comprehensive framework for analyzing disparities in audit quality among nations (Taldybayeva, 2025). Therefore, as this study examines joint audit quality in Tunisia and Morocco – two institutional environments that have not been sufficiently studied – institutional theory may help us understand how joint audit quality varies across different pairs in these emerging institutional economies, potentially leading to different outcomes.

Moreover, according to this theory, organizations are embedded in broader social and regulatory systems that impose pressures to conform to established norms, values and expectations (Taldybayeva, 2025). In response to these external pressures, organizations tend to adopt similar structures and practices, leading to a trend toward organizational similarity (Torres et al., 2020), which is called “isomorphism” (DiMaggio and Powell, 1983). Consequently, we anticipate that, under mandatory joint auditing, various joint audit pairs may adopt analogous structures and practices, potentially leading to a trend toward organizational uniformity in audit practices and, subsequently, a reduction in audit quality disparities among these pairs.

Audit quality underpins stakeholder trust in financial statements (Church et al., 2015). Despite numerous efforts by practitioners and researchers to define audit quality, a universally accepted definition and framework remain elusive (Velte, 2017). This is likely due to the complexity and variability of audit quality (DeFond and Zhang, 2014). From a practitioner’s perspective, audit quality is typically characterized by two fundamental determinants: offering reasonable assurance to users regarding the accuracy and integrity of financial statements and adherence to relevant auditing standards and regulations. From a researcher’s perspective, audit quality is often defined as the likelihood of identifying and disclosing significant misstatements (DeAngelo, 1981).

In 2010, the European Commission (2010), through its Green Paper, promoted joint audits to enhance audit quality, arguing that two auditors increase error detection, consistent with the four-eyes principle. Researchers have since investigated joint audit quality. Several studies have compared the quality of joint audits with single audits, yielding mixed results. For instance, Zerni et al. (2012), Ittonen and Trønnes (2015), AbuRaya (2025) and Bianchi (2018) found a positive association between joint audits and audit quality. However, El-Dyasty and Elamer (2022) found no relationship among the listed Egyptian companies. André et al. (2016), Garcia-Blandon et al. (2021) and Velte and Azibi (2015) similarly found no distinction in quality between joint and single audits. Furthermore, after Denmark and Kuwait abolished mandatory joint audits, studies by Lesage et al. (2017), Holm and Thinggaard (2018) and Van der Zahn and Tebourbi (2023) found no significant change in audit quality. This reveals a discrepancy in the results of previous studies concerning the quality of joint auditing. Moreover, most previous research focuses on France, with limited comparison in diverse or developing contexts. Therefore, there is a need for comparative studies of joint audit quality, especially between developed and developing economies. Our analysis of audit quality in mandatory joint-audit settings across France, Tunisia and Morocco addresses these gaps.

In addition, Deng et al. (2014) highlighted the need to consider joint audit composition and auditors’ differences in technological proficiency, as audit pairs may vary in quality (Lobo et al., 2017). Studies indicate that the composition of audit firm pairs affects audit quality. For instance, Nurunnabi et al. (2020) observed that joint audits by two Big Four firms may deliver the highest audit quality. Likewise, Francis et al. (2009) found that greater Big Four representation is associated with better outcomes, especially when information asymmetry is high. Conversely, Lobo et al. (2017) and Paugam and Casta (2012) demonstrated that pairing a Big Four firm with a non-Big Four firm in France yields the highest quality. André et al. (2016) noted, however, that there is no significant difference in quality between pairs with or without Big Four firms. Moreover, Biehl et al. (2021) reported higher joint audit quality when both auditors have similar high-level experience, whereas free-riding is more likely when experience differs. In a similar vein, Ittonen and Trønnes (2015) found that audit quality rises when joint auditors share characteristics. Conversely, Al-Hadi et al. (2017) demonstrated that including a less technologically proficient firm lowers audit quality in GCC countries. Furthermore, Lobo et al. (2017) and Paugam and Casta (2012) observed that similar firms may compete rather than collaborate.

These studies provide inconsistent evidence on the relationship between the composition of different joint audit pairs and audit quality. In particular, the traditional three-category classification (two Big Four, one Big Four with one non-Big Four and two non-Big Four) appears too coarse to capture meaningful distinctions between international and local non-Big Four firms, especially in emerging markets where local auditors often play a dominant role. To directly address this gap, our study distinguishes non-Big Four firms as international or local, consistent with Van der Zahn and Tebourbi (2023). Furthermore, prior research has not examined whether the presence of at least one industry-specialized auditor or company size affects joint audit quality. In response, we categorize companies as small, medium or large and examine how mandatory joint audit quality varies depending on the presence or absence of at least one industry-specialized auditor in the pair.

DeAngelo (1981) defined audit quality as the probability of detecting and revealing accounting errors. Therefore, it is logical to suggest that this ability may improve when two auditors collaborate to identify and report accounting errors (the four-eyes principle). However, it may vary depending on the type and size of audit firms that make up the joint audit pair, as some research uses them as proxies of audit quality (DeAngelo, 1981). Salehi et al. (2019) and DeFond and Zhang (2014) noted that the auditing literature categorizes external audit firms into two groups: larger firms (the Big Four) and smaller firms (non-Big Four). Sutaryo and Lase (2015) contended that an auditor’s attributes – whether associated with a Big Four, international or local firm – generally influence the audit process. Accordingly, our study categorizes audit firms into Big Four (B4) firms, international firms outside the Big Four (S1) and domestic firms unaffiliated with international networks or the Big Four (S2) (Hassan et al., 2026a; Van der Zahn and Tebourbi, 2023).

There is ongoing debate over whether Big Four audit firms, also known as the “Big N,” deliver higher-quality audits than non-Big Four firms. DeAngelo (1981) hypothesized that Big N firms provide better audit quality due to greater independence, resources, experience, specialized staff, superior training, advanced processes and reputational concerns. Reputation protection theory suggests that Big Four firms conduct more thorough audits to safeguard their reputations, leading to higher-quality audits (Khoo et al., 2020). However, despite these arguments and extensive research, empirical findings on audit quality differences between Big Four and non-Big Four audit firms remain mixed (Eshleman and Guo, 2014; Khurana et al., 2021). In addition, the collapse of Arthur Andersen, a former Big Four firm, challenges the assumption that Big Four audit firms have consistently higher audit quality. These inconsistent results, often based on coarser three-category classifications and single-country designs (primarily France), underscore the limitations of prior work and the value of broader classifications and multicountry comparisons.

Audit quality is often viewed as the combined outcome of auditor competence and independence (DeAngelo, 1981). Deng et al. (2014) suggested that joint audit composition should be considered in joint audit studies. They argue that a large audit firm has a technological advantage over a small one and hypothesize that joint audits pairing a large and a small firm may harm audit quality due to a free-riding problem: a less experienced auditor may rely on the more experienced one, which bears the full reputational loss. Deng et al.’s (2014) model does not anticipate a free-riding problem in a joint audit with two Big Four firms, as both firms would bear reputational losses equally, given their strong reputation and financial resources. However, according to resource dependence theory, Big Four auditors may rely too heavily on one another, and neither may ultimately report the breach, leading to poor overall audit quality.

Moreover, since France, Morocco and Tunisia have a long history of implementing mandatory joint audit practices since the 1990s, this, according to institutional isomorphism, could lead to similarities in joint audit practices among different pairs within each country. This may reduce differences in audit quality across these pairs. Therefore, given the conflicting predictions from agency, reputation protection, resource dependence and institutional isomorphism theories, along with mixed empirical evidence from previous studies, we adopt the following open (undirected) hypothesis:

Furthermore, this research examines and compares joint audit quality in the EU (France) and the MENA region (Tunisia and Morocco). According to institutional theory and documented institutional disparities across nations, including differences in governance standards, disclosure practices, auditor tenure and joint audit work division regulation, such variations may influence the effects of joint auditors on audit quality. For example, in emerging economies such as Tunisia and Morocco, weak governance standards, limited continuous disclosure and ineffective regulatory enforcement heighten the importance of audit quality for enhancing the credibility of reported earnings compared to developed economies such as France. However, these economies often face disparities in staffing, funding and technology among audit firms (Van der Zahn and Tebourbi, 2023). Consequently, large audit firms attract more qualified professionals, which may impact the quality of services offered by the Big Four, international and local audit firms. For instance, Al-Hiyari et al. (2022), analyzing 10 MENA countries from 2009 to 2019, discovered that large audit firms do not demonstrate a significant correlation with audit quality. Moreover, Choi et al. (2008) asserted that a nation’s legal liability framework significantly influences audit quality. Furthermore, Diallo (2021), using a sample of Eastern European and Middle Eastern countries, argued that differences in values and cultural dimensions across countries may lead to reduced or increased self-interested behavior, which in turn increases or decreases agency costs. We thus base our research on the following open (undirected) hypotheses:

Moreover, institutional isomorphism posits that organizations adapt to environmental pressures, including laws and societal expectations (Taldybayeva, 2025). Consequently, they frequently establish analogous structures and practices, resulting in organizational homogeneity (Torres et al., 2020). As Morocco and Tunisia are both emerging economies in the MENA region and share broadly similar institutional features, including long-standing mandatory joint audit requirements, joint audit pairs may develop analogous structures and practices. This leads to organizational uniformity in audit practices and to a reduction in disparities in audit quality. Therefore, our study also proposes the following hypothesis:

This study draws on a multicountry sample of nonfinancial listed firms from France, Morocco and Tunisia over the 2014–2023 period – a timeframe that encompasses the 2014 EU auditing reforms and the COVID-19 pandemic – to examine audit quality under mandatory joint audit regimes. These countries were selected for their long-standing joint auditing tradition: France is the only EU country that mandates joint audits for nonfinancial firms, while Tunisia and Morocco are the only MENA countries with similar requirements. Although France is well studied, research on Morocco and Tunisia is limited. Our study addresses this gap by examining audit quality in these emerging markets. We started with 820 companies before excluding financial firms, single-auditor firms and those lacking financial data. The final sample comprises 440 nonfinancial companies, yielding 4,400 observations: 3,590 from France, 510 from Morocco and 300 from Tunisia. Small sample sizes in Morocco and Tunisia may limit generalizability. Table 1 presents the results of the sample selection process.

We collected data from three primary sources. First, financial data, including total assets, total revenues, net income, total debt, current assets, current liabilities, auditor opinion, company age, closely held shares, fiscal year-end, market-to-book value and number of subsidiaries, were obtained from the Thomson Reuters Datastream database. Second, missing financial data not available from the database were manually collected from the annual reports of French, Moroccan and Tunisian companies for 2014–2023, sourced from their official stock exchange websites. Third, additional details, such as auditor name and type (Big Four, international or local) and whether the company changed at least one auditor during the year (auditor change), were also collected from the annual reports of French, Moroccan and Tunisian companies for 2014–2023 from their official stock exchange websites. Finally, national governance data were collected from the World Bank’s Worldwide Governance Indicators, compiled by Kaufmann et al. (2010), which rate countries from 0 to 100 based on six governance factors: voice and accountability, political stability and absence of violence/terrorism, government effectiveness, regulatory quality, rule of law and control of corruption.

In our study, the primary dependent variable is AuditQuality, which measures audit quality for firm i in fiscal year t. In our study, we treat audit quality – the primary dependent variable – as unobservable directly. While prior studies often rely on proxies such as discretionary accruals, we deliberately select AWCA from the DeFond and Park (2001) model as our primary proxy of audit quality. Francis and Wang (2008) strongly argue that this model suits contexts such as Morocco and Tunisia particularly well, as limited industry observations severely undermine the traditional Jones (1991) model. Accordingly, we implement the DeFond and Park (2001) model as follows:

In equation (1), AWCA represents abnormal working capital accruals in year t; WC_t denotes noncash working capital in year t, calculated as ((current assets – cash and near-cash assets) – (current liabilities – short-term debt)); WC_{t – 1} indicates noncash working capital in the prior year; Sales_t refers to sales in year t; and Sales_{t – 1} signifies sales in the prior year. Following El-Dyasty and Elamer (2022), we scaled AWCA in equation (2) by lagged total assets. As noted by André et al. (2016), absolute discretionary accruals are the primary indicator of audit quality in our study. To minimize the influence of extreme observations, the dependent variable values are winsorized at the 1st and 99th percentiles.

Because our study primarily examines and compares the extent to which the quality of mandatory joint audits varies across joint audit pairs, the main independent or test variables are the composition of joint audit pairs. Sutaryo and Lase (2015) contended that an auditor’s attributes, whether associated with a Big Four, international or local firm, generally influence the audit process. Accordingly, consistent with Van der Zahn and Tebourbi (2023) and Hassan et al. (2026a), we categorize joint audit pair affiliations into six distinct pairs for each of the three countries (France, Morocco and Tunisia). These are defined as follows: B4B4FR/MO/TU_i,t, equals 1 when two Big Four auditors audit a firm and 0 otherwise; B4S1FR/MO/TU_i,t,equals 1 when a firm is audited by one Big Four auditor and one non-Big Four international auditor and 0 otherwise; B4S2FR/MO/TU_i,t,equals 1 when a firm is audited by one Big Four auditor and one local non-Big Four auditor and 0 otherwise; S1S1FR/MO/TU_i,t, equals 1 when two non-Big Four international auditors audit a firm and 0 otherwise; S1S2FR/MO/TU_i,t, equals 1 when a firm is audited by one non-Big Four international auditor and one local non-Big Four auditor and 0 otherwise; and S2S2FR/MO/TU_i,t, equals 1 when two local non-Big Four auditors audit a firm and 0 otherwise. For notational convenience in the country-specific analyses, we append the country identifiers FR, MO or TU to each variable where appropriate.

Control variables were selected based on prior research (e.g. El-Dyasty and Elamer, 2022; Francis et al., 2009; Garcia-Blandon et al., 2021; Haak et al., 2018; Holm and Thinggaard, 2018; Van der Zahn and Tebourbi, 2023; Velte and Azibi, 2015). Table 2 summarizes variable definitions and measurements. Also, we controlled for institutional differences using the National Governance Index (NGI), which averages the six World Bank indicators (Kaufmann et al., 2010); definitions are provided in Table 2.

First, because firms select auditors based on their preferences, this potentially introduces endogeneity through self-selection bias. To address this issue, we apply Heckman’s (1979) two-stage procedure. In the first stage, we use a probit regression with JOINT_AUDIT (ranked 1–6) as the dependent variable to model the determinants of selecting one of the six joint audit pairs. This stage includes control variables such as client size, debt, losses, closely held shares, return on assets and number of subsidiaries. The functional structure is as follows:

The outcomes of this first-stage regression are then used to compute the inverse Mills ratios (IMR1–IMR6) for each of the six joint audit pair choices.

In the second stage, we incorporate the self-selection correction parameters (IMR1–IMR6) as additional control variables in the main regression analysis. This adjustment accounts for endogenous factors specific to each country and enables consistent comparisons both within and across countries.

To test our first hypothesis, we examine how audit quality varies across the six joint audit pairs within each country (France, Morocco and Tunisia) using the following model:

This model is estimated separately for each country by repeating the regression six times, each time using a different pair as the reference category (B4B4 for Model 2a, B4S1 for Model 2b, B4S2 for Model 2c, S1S1 for Model 2d, S1S2 for Model 2e and S2S2 for Model 2f). This reference-pair rotation allows for complete pairwise comparisons of audit quality across all six combinations.

To test our second hypothesis, we examine how mandatory joint audit quality varies across joint audit pairs between the developed economy (France) and the developing markets (Morocco and Tunisia) using the following model:

This model is applied in two separate analyses: one comparing French-Moroccan pairs and another comparing French-Tunisian pairs. In each analysis, the model is run six times, each time using a different French pair as the reference (Model 3a: B4B4FR; Model 3b: B4S1FR; Model 3c: B4S2FR; Model 3d: S1S1FR; Model 3e: S1S2FR; and Model 3f: S2S2FR). The dependent and control variables remain the same as in Model 2.

Finally, to test our third hypothesis, we examine differences in audit quality across joint audit pairs between the two developing economies (Morocco and Tunisia) using the following model:

This model is estimated six times, each time using a different Moroccan pair as the reference (Model 4a: B4B4MO; Model 4b: B4S1MO; Model 4c: B4S2MO; Model 4d: S1S1MO; Model 4e: S1S2MO; and Model 4f: S2S2MO). Again, the dependent and control variables are retained from the previous models. Table 2 summarizes all variable definitions and measurements.

We adopted a quantitative approach to examine how different joint audit-pair combinations influenced audit quality. Accordingly, we used descriptive statistics and panel OLS regression with industry- and time-fixed effects, consistent with prior studies, to precisely isolate the impact of the audit pairs. In line with Bianchi (2018) and Van der Zahn and Tebourbi (2023), we winsorized the dependent variable and all continuous controls at the 1st and 99th percentiles to reduce the influence of outliers. Finally, we excluded from the primary analysis any companies with missing data in Thomson Reuters Datastream or annual reports.

Figure 1 shows that 74.82% of the French sample was audited by pairs with at least one Big Four auditor, compared with 25.18% by pairs without a Big Four auditor. This indicates a strong Big Four influence through collaboration or operation with other firms. In Morocco and Tunisia, the shares audited by pairs with at least one Big Four auditor were 48.04% and 59%, respectively, suggesting a more balanced split in these markets. B4S1 was the top pair in France (30.42%), followed by B4S2 (24.48%) and B4B4 (19.92%). In Morocco, S1S2 led (21.96%), followed by B4S2 (20.98%) and S2S2 (18.82%). In Tunisia, B4S2 was most prevalent (35%), followed by S2S2 (29.67%) and B4S1 (17%). Therefore, these results underscore the significant role of local auditing firms in Morocco and Tunisia.

Table 3 presents descriptive statistics for the primary and control variables, grouped into three panels: French firms (Panel A), Moroccan firms (Panel B) and Tunisian firms (Panel C). The most notable finding is that French firms exhibit lower absolute AWCA compared to Moroccan and Tunisian firms. In contrast, Tunisian firms rank higher than Moroccan firms, suggesting that French firms have higher audit quality than their Moroccan and Tunisian counterparts. However, directly comparing the average AWCA is challenging due to the distinct features of the organizations.

Table 4 reports regression results comparing audit quality (AWCA) across the six joint audit pairs in France, Morocco and Tunisia. For each country, six models are estimated, each using a different reference pair (Model 1: B4B4; Model 2: B4S1; Model 3: B4S2; Model 4: S1S1; Model 5: S1S2; and Model 6: S2S2) to enable complete pairwise comparisons. The regression results show that audit quality generally does not differ significantly across most joint audit pairs in French, Moroccan and Tunisian companies. These results do not support H1, but they are consistent with André et al. (2016), who noted only minor variations among ONEBIG4FR, TWOBIG4FR and NONBIG4FR pairs. This broad convergence can be explained by the long-standing mandatory joint audit systems in place since the 1990s across all three countries, which appear to have fostered standardized professional practices and rigorous oversight. In line with institutional isomorphism theory, such regulatory pressures have encouraged various joint audit pairs to adopt analogous structures and practices, thereby reducing quality disparities despite differences in auditor type or resources. Auditor collaboration and the exchange of expertise under mandatory regimes (Bianchi, 2018) further contribute to this outcome, as does the competitive environment that enables smaller firms to gain greater market share (Ratzinger-Sakel et al., 2013). However, we interpret these results cautiously due to the small sample sizes in Morocco and Tunisia, which may limit generalizability.

Key exceptions nevertheless emerge. In France, S1S2FR outperforms B4B4FR (p = 0.00765**) and B4S1FR (p = 0.00636**) at the 5% level. This partially supports H1 and contradicts Deng et al. (2014), who argued that pairing large and small firms may reduce audit quality. Resource dependence theory offers a plausible explanation: this pairing integrates the global expertise of the international non-Big Four firm with the local regulatory knowledge of the domestic firm, potentially enhancing overall audit quality in the French context, where labor division is regulated, and IFRS is widely adopted. Turning to Morocco, B4B4 audits differ significantly from B4S1 (p = 0.0209**), again partially supporting H1. This result aligns with reputation protection theory, which posits that Big Four firms maintain high audit standards to protect their reputation, especially in emerging markets with weaker governance structures. In Tunisia, S1S1TU shows clear advantages over the other five Tunisian joint audit pairs, partially supporting H1. This finding likely reflects strong competition among international non-Big Four firms, which collaborate to improve their reputations and market positions by delivering higher audit quality. Overall, while institutional isomorphism drives broad convergence under the mandatory regimes in place since the 1990s, specific pair compositions can still yield modest differences in quality in certain institutional settings. However, we interpret these results cautiously due to the small sample sizes in Morocco and Tunisia, which may limit generalizability.

Table 5 compares audit quality levels (AWCA) across joint audit pairs for pairwise country comparisons. In Models 1–6, B4B4FR, B4S1FR, B4S2FR, S1S1FR, S1S2FR and S2S2FR serve as the benchmark for French versus Moroccan firms and French versus Tunisian firms, while B4B4MO, B4S1MO, B4S2MO, S1S1MO, S1S2MO and S2S2MO serve as the benchmark for Moroccan versus Tunisian firms. The results in Table 5 show no significant differences in audit quality between the six French joint audit pairs and the six Moroccan or Tunisian joint audit pairs. Thus, these findings do not support H2 or its sub-hypotheses. Similarly, Table 5 indicates no significant differences in audit quality between the six Moroccan and six Tunisian joint audit pairs, which supports H3.

These patterns of convergence can be explained by institutional isomorphism, which posits that organizations adjust to environmental pressures, including legal frameworks and social norms, by adopting analogous structures and practices, thereby resulting in organizational homogeneity (Torres et al., 2020). Given that France, Morocco and Tunisia have all implemented mandatory joint audit requirements since the 1990s, and that Morocco and Tunisia have adopted elements of the French public accounting system (Khlif et al., 2020), such long-standing regulatory traditions appear to have fostered standardized professional practices across different joint audit pairs – whether local or affiliated with international networks. This institutional pressure toward uniformity reduces the likelihood of substantial quality differences across pairs, even in the presence of implementation variations such as the lack of regulated labor division in Morocco and Tunisia or differences in IFRS versus local standards adoption. The findings therefore illustrate how mandatory joint audit regimes can promote convergence in audit quality both within and across developed and emerging contexts, consistent with the theoretical framework that emphasizes isomorphism under mandatory settings. However, we interpret these results cautiously due to the small sample sizes in Morocco and Tunisia, which may limit generalizability.

In this section, we tested audited firm-size effects by dividing the sample into small, medium and large terciles based on total assets (Al-Hadi et al., 2017). We then reapplied the within-country regressions from subsection 6.2.1. These steps enabled us to identify differences and similarities in audit quality across the six joint audit pairs at different firm sizes. For brevity, Table 6 reports only firm-size comparisons of audit quality across joint audit pairs by country; full tables are available upon request.

Table 6 results indicate that there are no significant differences in audit quality generally across joint audit pairs for small, medium and large firms in the French, Moroccan and Tunisian samples. This consistency is more pronounced in larger firms, where variations in audit quality are notably smaller across pairs. The possible explanation is that larger firms exhibit a greater information gap between management and shareholders, thereby increasing agency risk and prompting all auditors —Big Four, international or local – to exert greater effort, thereby minimizing differences in AWCA levels across joint audit pairs. This pattern is also consistent with institutional isomorphism under the long-standing mandatory joint audit regimes in place since the 1990s, which appear to have fostered standardized practices and unified control, leading to even stronger convergence in audit quality among different pair types for larger firms. However, the results indicate that audit quality is higher for the S1S1TU pair than for the other five Tunisian joint audit pairs in large Tunisian firms. Notably, the S1S1TU pair is observed only among large Tunisian firms, as this audit pair did not audit any small or medium-sized Tunisian firms during the sampled period. These findings may reflect heightened competition in large firms, driving non-Big Four international firms to mobilize additional resources, such as larger audit teams or specialized expertise, to narrow audit quality gaps. Therefore, these results align with and support those in Table 4, thereby increasing their reliability. However, we interpret these results cautiously due to the small sample sizes in Morocco and Tunisia, which may limit generalizability.

To the best of our knowledge, prior research on joint audit quality has not examined the effect of including at least one auditor with industry specialization on audit quality. This additional analysis aims to address this gap by examining the quality of each of the six joint audit pairs and comparing cases in which at least one auditor has industry expertise with those in which neither does.

Industry specialization is not directly observable; therefore, prior studies use indicators such as market share, portfolio share or weighted share to assess it (Audousset-Coulier et al., 2016). Consequently, audit firms with dominant market shares in their sectors are viewed as industry experts (Balsam et al., 2003). Building on this definition, this analysis uses the market-share approach to examine how industry specialization affects audit quality. Audousset-Coulier et al. (2016) highlighted five main proxies for market size: audit fees, assets, sales, number of clients and the square root of assets or sales. However, because audit fee disclosure in Tunisia is incomplete (Hassan et al., 2026b), we use total assets as our key proxy. Following Audousset-Coulier et al. (2016), we identify the firm with the largest market share as the specialist. According to Hassan et al. (2026a), we then calculate an audit firm’s market share as follows:

where:

MarketShare denotes the market share of auditor i in industry k, where i = 1, 2, … indexes audit firms, j = 1, 2, … indexes client firms and k = 1, 2, … indexes client industries. ik denotes the number of audit firms i operating in industry k; ijk denotes the number of clients served by audit firm i in industry k; and A_ijk denotes the total assets of client firm j in industry k audited by auditor i. To assess how industry specialization influences joint audit quality, we used the model outlined below:

The dependent and control variables in this model align with those used in prior models. The model was executed six times independently in each country (France, Morocco and Tunisia), using a different joint audit pair as the test variable each time. The test variables are defined as follows:

B4B4_SPEC_i,t, equals 1 when a firm is audited by two Big Four auditors and at least one is an industry specialist (otherwise 0); B4S1_SPEC_i,t, equals 1 if the firm is audited by one Big Four and one non-Big Four international auditor with at least one industry specialist (otherwise 0); B4S2_SPEC_i,t, equals 1 when a firm is audited by one Big Four and one local non-Big Four auditor with at least one industry specialist (otherwise 0); S1S1_SPEC_i,t, equals 1 if two non-Big Four international auditors are used with at least one industry specialist (otherwise 0); S1S2_SPEC_i,t, equals 1 when a non-Big Four international and a local non-Big Four auditor are used with at least one industry specialist (otherwise 0); and S2S2_SPEC_i,t, equals 1 if two local non-Big Four auditors are used with at least one industry specialist (otherwise 0).

Table 7 indicates that, in most cases across France, Tunisia and Morocco, the presence of at least one industry-specialized auditor within a joint audit pair does not significantly affect overall audit quality. One possible explanation is that joint audits involve shared knowledge and procedural review. This sharing reduces reliance on a single auditor’s expertise, thereby limiting the impact of specialization on overall audit quality under mandatory regimes that emphasize collaboration. However, the study reveals important exceptions. In France, two Big Four firms without industry-specialized auditors (B4B4_NON-SPEC) produced significantly higher audit quality than similar pairs with industry-specialized auditors (coefficient = 0.0336, significant at the 1% level). This may occur because overreliance on the specialized auditor reduces cross-review and balanced effort, particularly in a regulated, mandatory joint audit environment. In contrast, when a French joint audit pair includes a Big Four auditor and a local firm, the presence of an industry-specialized auditor is associated with higher audit quality, though this effect is less significant. Resource dependence theory may explain this result: industry specialization bridges technical gaps between Big Four and local firms, integrating resources and enhancing audit quality. In Morocco, S2S2MO, B4S1MO and B4B4MO pairs without industry-specialized auditors also delivered higher audit quality. These differences are significant at the 5% level (coefficients = 0.435, 0.0439 and 0.433, respectively). These outcomes run counter to prior studies that emphasize the benefits of auditor industry specialization (Balsam et al., 2003; Hegazy and Ebrahim, 2022). Such patterns may reflect the dynamics of emerging markets, where institutional pressures toward uniformity under mandatory joint audit regimes can sometimes lead industry specialists to carry a disproportionate workload. This can limit the partner auditor’s role and reduce the effectiveness of cross-review in joint audits. Taken together, these findings show that the presumed audit quality benefits of industry specialization do not always materialize in joint audit systems, especially when professional responsibility is unbalanced under long-standing mandatory frameworks. Nonetheless, two limitations must be acknowledged. First, the sample sizes in Morocco and Tunisia are small. Second, the outcome may depend on the method used to measure industry specialization (Audousset-Coulier et al., 2016). These factors may constrain the broader application of these findings.