Do common auditors mitigate classification shifting in the family business groups? Evidence from emerging economy

This research study examined earnings management (hereafter, EM) via classification shifting in the family business group (henceforth, FBG) affiliated firms and focusing on the role of common auditors ^[1] in curbing classification shifting in the FBG-affiliated firms. The classification shifting is based on the misclassification of the income statement items/elements but without altering the net income (Malikov and Gaia, 2022). Similarly, McVay (2006) defined classification shifting as the deliberate vertical shifting of income statement items. Bansal (2021) reported that classification shifting is the vertical movement of the income statement items for the reason of reporting favorable operating performance without altering bottom-line income, like accruals earnings management (hereafter, AEM) and real activity manipulation (hereafter, RAM). The practice of classification shifting increases the line items of the income statement, which affects stock valuation and carries importance for investors (Anagnostopoulou and Malikov, 2024). Therefore, the classification shifting practice is a non-bottom-line income manipulation EM technique, while in AEM and RAM, the bottom-line profits/incomes are manipulated (Anagnostopoulou et al., 2021).

In the extant literature, EM is investigated in the FBG-affiliated firms via AEM, RAM or a combination of both (Khan et al., 2025; Khan and Kamal, 2022; Muttakin et al., 2017). However, the evidence of classification shifting in the business group-affiliated firms remains unexplored in EM literature. Apart from the AEM and RAM, FBG-affiliated firms may commit EM via classification shifting. For instance, FBG-affiliated firms may be involved in classification shifting in order to beat or meet analysts' forecasts because investors value core earnings more than non-core earnings and may be involved in classification shifting to reduce losses or influence taxes. Similarly, FBG-affiliated firms may strategically prefer classification shifting over other EM choices, as this practice does not impair firm value and has no effect on future firm performance (Anagnostopoulou et al., 2021) as compared to AEM and RAM. The AEM and RAM reduce future cash flows, impact firm performance and have huge economic costs. Additionally, FBG-affiliated firms may consider classification shifting over other forms of EM due to its less falsification nature, harmless impact on firms and unlikely to damage or harm FBG-affiliated firms' reputations (Malikov and Gaia, 2022; McVay, 2006).

In Pakistan, the dominance of FBGs and family firms, lower investor protection, poor cum family-oriented corporate governance and pervasive EM in Pakistani firms (Khan et al., 2022; Khan and Kamal, 2024a) may stimulate FBG-affiliated firms to be involved in classification shifting. In the related literature, the studies of Beuselinck and Deloof (2014), Khan (2022) and Muttakin et al. (2017) reported that FBG-affiliated firms are involved in EM via AEM. The reason for such involvement is due to the ownership structure of FBG-affiliated firms, which offer incentives for the controlling shareholders to engage in EM, in order to expropriate the minority shareholders' rights. The management involvement in various forms of EM in the FBG-affiliated firms creates the principal (majority) and principal (minority) agency problems (Khan et al., 2025). Similarly, in Pakistan, the studies of Khan et al. (2025) and Khan and Kamal (2022) reported that FBG-affiliated firms in Pakistan are involved in RAM.

On the contrary, the studies of Choi and Kim (2012), Khan et al. (2025), Kouwenberg and Thontirawong (2016) documented that FBG-affiliated firms are not involved in EM due to the socio-economic wealth (SEW) of the family firms and the value of their long-term reputation. Hence, the aforementioned discussion of EM practice in the FBG-affiliated firms reported mixed results, while taking AEM and RAM proxies. Therefore, it is not clear whether FBG-affiliated firms will use classification shifting, as this method of EM is different compared to AEM and RAM. Therefore, the first objective of this research study is to investigate whether FBG-affiliated firms are involved in EM via classification shifting.

Now, a question will arise whether common auditors engaged in a FBG curb EM via classification shifting? Since prior literature suggests that classification shifting is less likely to attract the attention of external auditors (Haw et al., 2011; McVay, 2006; Zalata and Roberts, 2016, 2017; Zalata et al., 2019; Zalata and Abdelfattah, 2021). Likewise, Desai and Nagar (2016) posited that auditors are not effective in detecting classification shifting. Consistent with this argument, Nelson et al. (2002) added that auditors are more likely to waive EM attempts committed by large clients. Likewise, Pakistani FBGs could be large clients for auditors, as they have many affiliated firms (Khan, 2023).

Hence, it is not theoretically clear whether the findings of Desai and Nagar (2016) and Nelson et al. (2002) are applicable to Pakistani FBG-affiliated firms. On the one hand, based on Abid et al. (2018), external auditors may be less effective in the Pakistani capital market due to the lack of incentives to deliver high-quality audits. In line with, Sun et al. (2020) reported that auditors affiliated with business groups (BG) compromise independence due to affiliated firms' size and economic significance. On the other hand, Khan and Kamal (2024b) added that the same network auditors have an effective role in curbing EM in the FBG-affiliated firms. Similarly, Nagar et al. (2021) and Mulchandani and Mulchandani (2022) reported that BIG-4 auditors act as a superior deterrent to mitigate EM and play a vital role in curbing classification shifting.

In addition, the findings of Desai and Nagar (2016) and Nelson et al. (2002) are not about family and FBG-affiliated firms. The extant literature reports that hiring or having common auditors, including, those from BIG-4, may be instrumental in curbing EM (AEM and RAM) in the FBG-affiliated firms due to their knowledge sharing between group affiliates (Chang and Chen, 2008; Fu and Kim, 2024) and their efficacy in preventing misreporting (Khan and Kamal, 2024b). Therefore, to find the answer to the aforementioned study's question, the second objective of this research study is to examine the role of common auditors along with BIG-4 common auditors, in controlling classification shifting in the FBG-affiliated firms in the Pakistani context.

Additionally, the key motivation of this research study is the context of Pakistan belonging to “Majority World” (Uddin, 2025). Pakistan's capital market provides an excellent setting for testing the classification shifting phenomena in Pakistan's FBGs. First, listed firms in Pakistan have a high level of concentration of ownership, mostly in the hands of controlling families (often these firms are affiliated with FBGs). In family firms, agency problems existed between majority and minority shareholders, termed principal-principal agency problems (Ferramosca and Allegrini, 2018; Sun et al., 2017). The presence of the principal–principal agency problem causes information asymmetry, which provides an incentive for the majority shareholders to extract private benefits by engaging in classification shifting, as the controlling family owners have control over companies' accounting policies. Hence, examining classification shifting in Pakistan, an emerging economy, merits proper empirical investigation. Second, the listed firms of the Pakistan Stock Exchange (PSX) are mostly controlled and managed by families' directors/members. Moreover, the board of directors in PSX-listed firms is dominated by family members/directors, having long-tenured independent directors and family loyalists and close friends (Khan, 2022). The strong influence of families on PSX-listed firms' boards may discourage independent directors from working independently, and they may not fulfill their duties out of family proximity. Lastly, Abid et al. (2018) added that external auditors are less effective in the Pakistani capital market due to the non-availability of any incentives to deliver high-quality audits. As a result, Pakistan's specific institutional and governance characteristics make its settings distinct and worthy, which necessitates a systematic effort to check the role of common auditors in limiting EM via classification shifting in the FBG-affiliated firms.

To achieve the study's objective and to explore the relationship of classification shifting in the business groups of Pakistan, a data set of 286 non-financial firms of PSX from 2010 to 2019 has been used. This study's dependent variable is unexpected core earnings estimated under the (McVay, 2006) model, and we assume that its relationship with income-decreasing special items (SPI) will be positive, based on the methodology of (McVay, 2006). The results of the study indicate that there is a positive and highly significant relationship between SPI and unexpected core earnings in the non-financial listed firms, indicating the existence of classification shifting. Thus, the findings of this research study provide evidence that PSX-listed firms are involved in classification shifting. However, the FBG-affiliated firms are not involved in classification shifting, when the SPI interacts with the FBG dummy. Moreover, when we checked the role of common auditors and BIG-4 common auditors in the FBG-affiliated firms, the results show that both common and BIG-4 common auditors reduce classification shifting in the FBG-affiliated firms. Furthermore, the results of this study are robust with alternative measures of classification shifting, such as with working capital accruals, dropping current accruals components from the classification shifting model, and robust analysis.

The empirical evidence of this research study contributed to the EM, FBG and audit literature in several key aspects, especially from “Majority World” (Uddin, 2025) country. First, this research study's findings added to the prior classification-shifting literature by reporting that common auditors and BIG-4 common auditors are new EM determinants and deterrents in the FBG-affiliated firms. Although the studies of Boahen and Mamatzakis (2020), Mulchandani and Mulchandani (2022) and Nagar et al. (2021) had linked BIG-4 auditors with classification shifting. However, these studies investigated BIG-4 auditors' role in non-FBG-affiliated firms. We have provided evidence from the FBG perspective and reported that not only common auditors, but BIG-4 common auditors are instrumental in controlling classification shifting in the FBG-affiliated firms. Second, this research study contributed to the growing EM literature by highlighting the distinct role of common auditors, along with BIG-4 common auditors in FBG-affiliated firms in curbing classification shifting. To the best of the authors' knowledge, this research study is the first study to provide evidence that FBG, or business group firms, are not involved in classification shifting in the emerging market. In addition, this research study is the first to investigate the role of common auditors in curbing classification shifting in the FBG-affiliated firms. Although Khan and Kamal (2024b) had linked common auditors (same network auditors) with AEM and Fu and Kim (2024) had linked them with audit quality. However, in existing literature, the role of common auditors in mitigating EM via classification shifting was not systematically examined. Through this research study, we provide insights into how common auditors, along with BIG-4 common auditors, play an effective role as an external corporate governance mechanism in curbing EM via classification shifting in FBG-affiliated firms.

Third, through this research study, a Pakistani context has been exploited, famous for family-dominated firms with well-known FBGs, which provide incentives for managers to be involved in EM and fill the research gap established in earlier paragraphs regarding FBGs and classification shifting along with common auditors role in curbing classification shiting in FBG-affiliated firms. Finally, previous studies linked FBG-affiliated firms with AEM and RAM (Khan, 2022; Khan et al., 2025; Khan and Kamal, 2022; Muttakin et al., 2017). However, this research study extends EM research with another EM technique or form of classification shifting and links it with FBG-affiliated firms providing the first-ever evidence of classification shifting in FBG-affiliated firms.

The remainder of the study is divided into four parts as follows: Section 2 presents and reports related literature, theoretical framework and the hypotheses of the study. Section 3 provides the details of data, sample, research methodology and research equations, along with a description of the variables. The results and discussion are documented in Section 4. This research study is concluded in Section 5.

The majority of extant literature is based on two theoretical perspectives on the EM-family firms' relationship: the long-term perspective and the managerial opportunism perspective (Gavana et al., 2025). The long-term perspective has been investigated under stewardship theory, stakeholder theory, legitimacy theory and SEW theory, while on the other hand, the managerial opportunism perspective has been examined under institutional theory, entrenchment vs alignment effect theory and agency theory (Gavana et al., 2025). Hence, this research study's hypothesis regarding FBG-affiliated firms and classification shifting will be built under agency and SEW perspectives. According to agency and SEW perspectives, FBG-affiliated firms are not involved in EM due to family interests and preservation of family reputation (Khan et al., 2025; Khan and Suplata, 2025; Kim, 2019).

In addition, in the audit and business group literature concerning the audit quality of the group affiliates, there are three hypotheses: the “agency theory hypothesis,” the “influential power hypothesis” and the “knowledge transfer hypothesis” (Chang and Chen, 2008; Fu and Kim, 2024; Khan and Kamal, 2024b). In FBG-affiliated firms, the traditional agency problem between principal and agent is resolved by family ownership and control (Khan, 2022). Hence, in the FBG-affiliated firms, there exists the agency type II problem between majority (FBG-controlled ownership) and minority shareholders (Khan et al., 2025), which is also called the principal and principal problem (Khan, 2022; Khan et al., 2022; Khan and Kamal, 2022). Therefore, in order to resolve agency type II conflict, agency theory demands independent auditors to play an external corporate governance role to resolve this conflict and protect the interests of minority shareholders (Cuomo et al., 2016). Hence, independent auditors alleviate the agency type II problem by engaging well-reputed audit firms like BIG-4 auditors (Khan and Kamal, 2024b). Hence, independent auditors are paramount in securing minority shareholders’ rights against the expropriation of controlling shareholders in FBG-affiliated firms and in resolving the principal–principal agency conflict (Khan and Suplata, 2025). The independent auditor/firm is such an external monitoring function that it improves the financial reporting quality, aligns the interests of principals and agents (Abid et al., 2018), assuages the small investors’ (minority shareholders’) concerns of majority shareholder expropriation (Khan, 2022) and lessens agency conflicts (Chang and Chen, 2008).

In accounting research, EM is one of the contemporary issues. Accounting academicians are constantly exploring EM with related variables in different contexts. In the EM literature, mostly three EM techniques have been identified and used (Khan and Suplata, 2025). These are the AEM, RAM and classification shifting. The former two are the most researched and widely reported EM techniques in various contexts (Haw et al., 2011; He et al., 2024). However, the latter one, classification shifting, is the least researched EM technique, especially in the FBG-affiliated firms. Zalata and Roberts (2017) reported that some managers are involved in classification shifting to expand core earnings by shifting recurring expenses to nonrecurring expenses without altering the bottom-line income. Moreover, Abernathy et al. (2014) documented in their study that classification shifting has gotten attention internationally because, under the International Financial Reporting Standards (IFRS), firms’ managers have the discretion over revenue and expense misclassification. The classification shifting has two forms: expense and revenue misclassification or shifting (Bansal, 2021). According to McVay (2006), firm management is involved in expense classification shifting by shifting operating expenses to non-operating expenses to overstate operating profit. Similarly, revenue classification shifting involves shifting non-operating revenues to operating revenues to exaggerate operating profit and revenue (Malikov et al., 2018).

In the current literature on classification shifting, there are various motives that stimulate companies' managers to engage in classification shifting. Hence, in the following lines, we are going to discuss various factors that motivate firms and firms' management to engage in classification shifting. Bansal (2021) reported that classification shifting is the vertical movement of the income statement items for the reason of reporting favorable operating performance without altering bottom-line income; like in AEM and RAM, bottom-line income is manipulated. Likewise, classification shifting is an EM technique practiced by the company's management to misclassify income statement items while keeping intact the net profit. The practice of classification shifting increases the line items of the income statement, which affects stock valuation and carries importance for investors (Anagnostopoulou and Malikov, 2024). Similarly, Boahen and Mamatzakis (2024) added that company management is involved in classification shifting due to the fact that investors give high valuation and relevance to the companies' core earnings during their assessment of a business. Nagar et al. (2021) added that companies engage in classification shifting since financial market participants (investors) target core earnings as compared to bottom-line earnings.

Moreover, companies’ management perceive classification shifting as less unethical compared to other EM techniques (Farnsel and Ha, 2024). In the same way, the exclusion of operating expenses for the current period of core earnings enhances core earnings, but in future periods, when such a practice is revealed, it will erode investors’ confidence, investment (Alfonso et al., 2015) and mislead investors from predicting accurate future performance (Anagnostopoulou and Malikov, 2024). Additionally, Bansal (2023) added that firms prefer classification shifting over AEM and RAM due to the low-cost element. In the same vein, Malikov and Zalata (2025) added that when companies' managers' abilities are constrained by managing accruals, then they engage in classification shifting. Furthermore, classification shifting is an inexpensive method of EM in which core operating income has been increased to achieve the benefits of meeting or beating analyst forecasts (Haw et al., 2011).

Having discussed the factors that motivate classification shifting, this practice of EM has substantial hitches for companies, such as creating doubts, skepticism and undermining financial statement quality, misleading investors, financial analysts and carrying considerable implications for companies' well-being and stock valuation (Hessian et al., 2024). Boahen and Mamatzakis (2020) argued that classification shifting manipulation may increase revenues in a short period and hide the costs of business, so classification shifting is unethical behavior on the part of managers by depicting cherry-picked information. In the same vein, Alfonso et al. (2015) reported that involvement in classification shifting significantly increases inefficiencies in resource allocation, particularly mispricing of earnings and causes possible negative market consequences.

In different countries of the world, serious concerns have been raised over classification shifting practices, which cause changes in accounting standards. For example, in Australia, accounting standards were changed relating to the reporting of income to limit classification shifting (Seve and Wilson, 2019) ^[2] and in the USA, the Securities and Exchange Commission raises high concern over classification shifting and charges various companies with misclassification of income statement items (Alfonso et al., 2015). However, regulators' scrutiny is negligible to classification shifting as compared to other forms of EM due to its lack of effect on companies' net income (Farnsel and Ha, 2024; Liu et al., 2022).

In the extant literature, there are various deterrents that curb classification shifting. Such as BIG-4 auditors are famous for their audit quality and reputation. Hence, BIG-4 auditors are more likely to limit classification shifting compared to non-BIG-4 auditors (Mulchandani and Mulchandani, 2022; Nagar et al., 2021). Similarly, according to Liu et al. (2022), CFO officers with longer tenure effectively mitigate classification shifting in the Chinese A-share listed companies. Moreover, Boahen and Mamatzakis (2024) added that the legal environment in the developed and emerging economies plays a restrictive role in classification shifting. In the same way, Haw et al. (2011) added that well-functioning legal institutions and independent external auditors mitigate the misclassification practice. According to Boahen and Mamatzakis (2020), religion through ethical channeling is an effective tool to deter managerial unethical behavior of classification shifting in US firms for the period of 2000–2015. Hence, religiosity effectively reduces both upward and downward unethical managerial behavior of EM (classification shifting). Moreover, internal corporate governance, such as board independence (Mulchandani and Mulchandani, 2022; Zalata and Roberts, 2016) and board diversity (Usman et al., 2022), has inhibited the classification shifting in various research settings.

According to Khan (2022), a business group is a cluster of firms, listed or non-listed, which are connected from the top by a family and control the firms through links such as common CEOs, cross-directorship/management and cross-ownership. Khan et al. (2025) argued that, as compared to stand-alone firms, FBG-affiliated firms have more advantages and incentives to engage in EM. Similarly, Sriniasan and Ajay (2022) posited that the EM phenomenon is more severe in business groups due to institutional voids and a lack of formal institutions in developing economies, which results in poor reporting quality of the firms. The literature of business groups, or FBG and EM, documented contradictory and variegated empirical findings. The studies of Beuselinck and Deloof (2014), Kim and Yi (2006) and Muttakin et al. (2017) reported that business groups or FBG-affiliated firms are positively related to EM. Similarly, the studies of Khan (2022) and Sarkar et al. (2013) also documented that FBG firms are involved in EM. Moreover, Sarkar et al. (2013) reported that the insiders in the FBG-owned business manage and manipulate earnings to get or conceal private benefits. Moreover, they reported the famous EM scam, also called the “Enron of India,” the “Satyam Computer Services,” manipulated its earnings for several years, later to report default. Additionally, Satyam Computer Services was the affiliated firm of Satyam Business Group. In the same vein, FBG-affiliated firms may be involved in classification shifting to report lower core earnings to avail tax-related benefits (Chung et al., 2021). Moreover, FBG-affiliated firms may exploit their minority shareholders via engaging in classification shifting, as it is considered less misrepresentative relative to other forms of EM (Farnsel and Ha, 2024).

In addition, the studies of Khan et al. (2025) and Khan and Kamal (2022) reported that FBG-affiliated firms are not involved in EM as compared to stand-alone or non-affiliated firms and the magnitude of EM is lower compared to stand-alone firms. Moreover, this hypothesis can be linked with the “alignment hypothesis,” and “SEW preservation,” which asserts that the interests of family insider management and outside-dispersed shareholders are aligned and preserve family status. Both parties want to increase the value of a firm and avoid EM, which can harm the reputation of the company and family (Khan, 2022). In addition, group-affiliated firms have more access to finance than stand-alone firms by using group reputation, which results in lower EM in the group-affiliated firms (Kouwenberg and Thontirawong, 2016). The access to external financing by group affiliates does not need to give positive signals to the outsiders through EM. Moreover, keeping in view the classical agency theory explanation, the management is involved in EM for personal benefits. However, management in FBGs is affiliated with family members whose interests are aligned with the firm values and reputations and may not be involved in EM.

In addition, these managerial shareholders in FBGs also lead to reducing the classical principal–agent conflicts. In line with this argument, Lin et al. (2021) find that management of group-affiliated firms may refrain from opportunistic reporting due to their reputation (SEW preservation), high visibility and increased monitoring by the regulator and media. Hence, this discussion concludes that FBG-affiliated firms are less engaged in EM (Kim, 2019). Moreover, the SEW theory suggests that EM practices may by differ between FBG-affiliated firms and stand-alone firms. According to the SEW theory, family firms may not be involved in EM due to long-term family reputation.

As depicted in the aforementioned paragraphs, there is extensive research available on FBG and EM; however, there is little to no prior evidence on whether FBG-affiliated firms are involved in classification shifting or avoid classification shifting. Therefore, due to a lack of literature available on classification shifting and FBG, this research study is going to find their relationship via the following undirectional hypothesis:

In FBG-affiliated firms, there exists the agency type II problem between the majority (FBG-controlled ownership) and minority shareholders (Khan et al., 2025). Therefore, in order to resolve agency type II conflict, agency theory demands independent auditors to play an external corporate governance role to resolve this conflict and protect the interests of minority shareholders (Khan, 2022). Moreover, Chang et al. (2008) posited that having a common auditor (same auditors) increases the auditors’ knowledge, and as a result, increases the audit quality and reduces the management's opportunistic behavior. Hence, external independent auditors under the agency theory perspective increased the credibility and transparency in the companies’ financial statements, so common auditors may do the job of external corporate governance roles in mitigating classification shifting in the business group affiliates.

Another hypothesis of common auditors in the business groups is the “influential power hypothesis,” or “economic dependence hypothesis.” According to this view, common auditors become economically dependent on a large client (business group) having many affiliates. Therefore, the economic dependency of an auditor impairs audit quality due to the fear of the potential loss of losing a large and rich client. According to Khan and Kamal (2024b), the same network auditors on one side increase the auditors' knowledge and understanding of a particular business group, but on the other side, they also increase the economic dependence on a business group, which compromises auditors’ independence and impairs audit quality. Similarly, Fu and Kim (2024) added that audit fee dependence on common auditors in a business group jeopardizes auditors’ independence, and an auditor of a business group may surrender to the group's managers regarding audit opinion. Hence, keeping in view the aforementioned arguments, common auditors may harm audit quality in the affiliates of a business group and may not decrease EM in the form of classification shifting.

According to the knowledge transfer view, engaging the same network auditors or common auditors for the same business group improves audit quality and, in return, limits the company's management opportunistic behavior. Likewise, Fu and Kim (2024) reported that common auditors improve audit quality, especially in those environments and contexts where information flow is opaque. In the same vein, the study of Yang et al. (2016) stated that common auditors improve audit quality, reduce costs in the shape of auditors’ communication and develop the common auditors’ understanding of intra-party transactions of group affiliates. Moreover, having the same network auditor (common auditors) for the various affiliates of the FBG reduces EM, and the same network auditor has a negative relationship with EM in the FBG-affiliated firms in Pakistan (Khan and Kamal, 2024b). In addition, Fu and Kim (2024) reported that internationally common auditors for the affiliated group firms improve audit quality and support the knowledge transfer view.

Hence, keeping in view the aforementioned discussion and logic as well as the limited to unavailable prior literature on the common auditors and classification shifting the EM literature, the following hypothesis of this research study is proposed as follows:

Independent auditors are an important external corporate governance factor who ensure financial transparency in the companies' external reporting. External auditors maintain companies' financial statements free of material misstatements and make sure that these statements depict a true and fair financial position of the companies. Similarly, Haw et al. (2011) added that effective auditing in the companies decreases the information asymmetry between the controlling and non-controlling shareholders and allows all the shareholders to verify the authenticity of the company's financial statements. The audit and EM literature suggest that BIG-4 auditors are more effective in deterring EM compared to non-BIG-4 auditors (Chang et al., 2019; Khan and Kamal, 2024b; Nagar et al., 2021; Viana et al., 2022). According to Fang et al. (2017), in China, group-affiliated firms prefer to appoint the top 10 audit firms in order to improve financial reporting quality for external monitoring. Moreover, they added that in China, top audit firms translate into higher audit, disclosure and earnings quality; strong implications for related-party transactions; and many more benefits associated with them compared to non-top auditors.

Therefore, they (BIG-4 auditors) reduce and curb EM via classification shifting in those settings where law enforcement is weak (Nagar et al., 2021). Keeping the argument in mind, it seems that BIG-4 auditors did not compromise their independence while auditing a rich and influential client, such as a very big and diversified business group. Along the same line, Chang et al. (2019) posited that compared to BIG-4 audit firms, the non-BIG-4 audit firms compromise their audit independence while auditing a large, economically important client having many affiliates. Moreover, Khan and Kamal (2024b) investigated the role of BIG-4 same-network auditors (common auditors) on AEM in the FBG-affiliated firms in Pakistan and reported that BIG-4 auditors have a strong and negative relationship with AEM, which indicates that BIG-4 auditors have an effective role in controlling management opportunistic behavior. In the same vein, Viana et al. (2022) posited that BIG-4 auditors could play an effective role in limiting EM in emerging economies featuring poor regulatory environments, higher uncertainty, poor investor protection and an unstable economy.

Similarly, Nagar et al. (2021) reported that BIG-4 auditors act as a superior deterrent to mitigate EM via classification shifting compared to non-BIG-4 auditors. They further argued that BIG-4 auditors charge higher audit fees, have global operations and fame, have strong incentives and aim to maintain their persistent and uniform reputation universally. Moreover, Mulchandani and Mulchandani (2022) added that in Indian firms, the BIG-4 auditors play a vital role in curbing classification shifting. Hence, we propose the following hypothesis regarding BIG-4 common auditors and their role in mitigating classification shifting in FBG-affiliated firms:

The total population of the study is 524 listed firms on the PSX. The sample selected for the study is 286 firms for the period of 2010–2019, after applying various exclusion criteria as discussed in the following Table 1. The period of 10 years is selected due to the COVID-19 pandemic, the world banking crisis in 2008–09, and the availability of annual reports of the companies listed on the PSX. The sample consists of 163 firms affiliated with FBG, and 123 firms are stand-alone. The data were manually collected from various sources; as mentioned in the study of Khan et al. (2025) related to FBG affiliation, financial data have been downloaded from the SBP sources. The data were unbalanced after estimation of classification shifting and other EM models due to the lagged values and delta variables in these variable estimations. Hence, the data were made balanced by dropping the years 2010 and 2011 due to lag and delta variables. Therefore, the effective period of study is from 2012.

In order to examine the role of common auditors in influencing EM via classification shifting in FBG-affiliated firms. In this research study, the McVay (2006) model has been used and adopted, which focuses on the distribution of firms' expenses between special items and core expenses. It is a two-step process; in the first stage, core earnings have been divided into expected and unexpected components by estimating the following Equation (1). According to McVay's (2006) model, Unexpected_Core earnings are calculated by misclassifying operating expenses as non-operating expenses, which results in inflated unexpected core earnings (hereafter, UE_CES). Moreover, the misclassification of operating expenses to non-operating expenses results in favorable UE_CES. Moreover, it is assumed that when companies' managers misclassify expenses, then the relationship between core earnings and SPI will be positive and significant. Thus, following this proposition, we expect that the relationship between UE_CES and SPI will be positive as evidence for classification shifting in PSX-listed non-financial firms. The UE_CES are estimated in the following equation.

In the above Equation (1), CE is the firm's core earnings calculated as the difference between the sales minus cost of goods sold and selling, general and administrative expenses. The ATO is the ratio of asset turnover. The ACC is the accrual estimated as net income minus cash flow from operations. The ΔSales is the change in revenue and $NEGΔSALESi,$ as a dummy variable, they take the value of one if the change in revenue is negative, otherwise zero. Lastly, all the aforementioned variables are scaled by total sales. The residual from the above Equation (1) will be the proxy of UE_CES.

The second stage of the McVay (2006) model is to regress UE_CES with the absolute value of SPI. In regression, a positive relationship between UE_CES and SPI would show that FBG-affiliated firms in Pakistan boost their companies' core earnings via classification shifting.

We measure the common auditor in this paper as an indicator variable taking the value of one if a business group or FBG has one audit firm and that audit firm is the auditor of at least two affiliated firms of a business group or FBG. Moreover, this definition is in line with the study by Fu and Kim (2024).

FBG-affiliation is measured through a dummy variable taking the value of one if a PSX-listed nonfinancial firm is affiliated with FBG and zero otherwise (firms are non-affiliated with FBG).

Following the classification shifting literature, this research study has included various firm characteristics as control variables because they may influence the classification shifting. According to Malikov and Gaia (2022), to control for profitability and scale effects in classification shifting, the following control variables may be used: firm size (FSIZE), leverage (LEVE) and return on assets (ROA). Similarly, BIG-4 auditors are also included in this research study as a control variable, because the presence of BIG-4 auditors may decrease classification shifting. Moreover, many studies reported that board and audit committee independence are the most important determinants of EM (Faisal et al., 2021; Khan and Kamal, 2021, 2024a). In addition to this, in the studies of Fan et al. (2019) and Malikov and Zalata (2025) used some key control variables related to classification shifting, such as board size, institutional ownership, profitability and tangibility are also adopted and used in this research study. Furthermore, we use the industry dummies and year dummies in regression models to control for fixed effects of industry and year.

The panel data regression techniques have been used in order to test the study's hypotheses. Before estimation and analysis, to overcome the issues of outliers, winsorization has been used with 1%–99% for the whole sample, as it was adopted as a standard in the related studies. Moreover, to test the study's hypothesis regarding the relationship between FBG-affiliated firms and classification shifting (H1), Equation (2) has been used:

In Equation (2), we expect statistically significant coefficients for β₃, which will indicate the presence of classification shifting in the FBG-affiliated firms.

While testing study's hypotheses H2 and H3, Equation (3a), (b) have been used. Additionally, in Equation (3a), there must be negative and statistically significant coefficients for β₇ to provide evidence whether common auditors are effective in curbing classification shifting in the FBG-affiliated firms. Lastly, in Equation (3b), there must be negative and statistically significant coefficients for β₇ to provide evidence that BIG-4 common auditors are effective in mitigating EM via classification shifting in FBG-affiliated firms.

The variable measurement is given in  Appendix.

The descriptive statistics of the study variables are given in Table 2. In Table 2 and in Panel A, the mean, standard deviation (SD), minimum (min), maximum (max) and various quartiles are reported. All the continuous variables are winsorized (1st and 99th percentiles) in order to overcome the outlier effect in the analysis. The most important variable in Table 2 is CE_UES stands for unexpected core earnings for PSX-listed non-financial firms. The mean and SD values of UE-CES are −0.003 and 0.054, along with a maximum value of 0.19. These negative values of UE_UES suggest that FBG-affiliated firms report lower core earnings than anticipated, as reported by Bansal (2021). Moreover, our study's findings are the same as those reported in the study of Khan and Suplata (2025). They reported that the negative values of core earnings in the PSX-listed firms indicate that classification shifting is not prevalent, and these firms did not increase core earnings to achieve various earnings-related motives.

Additionally, the special items' values, also known as non-operating expenses (SPI), are also negative, the same as UE_UES. The mean value of SPI is −0.059, the SD is 0.11 and the minimum and maximum values are −0.72 and 0.39. Moreover, the FBG-affiliation dummy (FBG_AFF) shows that 57% of firms are FBG-affiliated firms in the total sample of the study. Additionally, the common auditor's mean value is 0.39, while the minimum and maximum values are zero and one. The summary statistics of the board of directors' variables, such as board independence (BIND) and audit committee independence (ACIND), show that, on average, there is 14% board independence in the Pakistani firms, while the audit committee independence is 23%. Similarly, on average, there are 10% institutional shareholders in PSX-listed firms, while the maximum institutional shareholders are 70%. Lastly, Panel A reported that 54% of PSX-listed firms are audited by the BIG-4 auditors.

Additionally, in Table 2 and in panel B, the mean and median test of the difference between FBG-affiliated firms and stand-alone (non-affiliated) firms. The main variable of interest in this research study is CE_UES; the results in Panel B depicted that non-affiliated firms compared to FBG-affiliated firms have, on average, more unexpected core earnings, and this result is statistically significant. Similarly, the classification shifting proxy estimated with working capital accruals (CE WCA) is opposite from unexpected core earnings; it is less in the stand-alone firms compared to FBG-affiliated firms, and it is statistically significant. However, the value of SPI, board independence, audit committee independence and asset tangibility (TANG) is similar in both FBG-affiliated firms and stand-alone firms. Moreover, the BIG-4 auditors, leverage (LEVE), return of assets (ROA), loan size (LOANSIZE), profitability (PROFIT) and firm size (FSIZE) are more in the FBG-affiliated firms compared to stand-alone firms.

Table 3 shows the cross-tabulation of the FBG-affiliated firms and the common auditors. The main purpose of the insertion of this table is to report what the percentage of common auditors in the FBG-affiliated firms is. The results reported that almost 40% of FBG-affiliated firms have common auditors, as the cross-tabulation table reports 904 firm-year observations.

Moreover, Table 4 reports the tabulation of the study's dummy variables, as there are too many dummy variables; therefore, we thought it necessary to include it and summarize it. Table 4 shows that 57% of the sample size firms are FBG-affiliated firms, 39.86% affiliated firms with FBG have common auditors, the BIG-4 auditors in the PSX-listed firms are 54%, while BIG-4 common auditors are almost 24%.

Table 5 reports the study's variables correlation matrix. In Panel A, common auditors (COM-AUD) along with the study's variables are reported, while in Panel B, BIG-4 common auditors (BIG-4COM-AUD) along with the study's variables are reported. The correlation between CE_UES and income-decreasing special (SPI) items is positive and highly significant. These results indicate that there is a presence of classification shifting in Pakistani firms. In addition to this, the FBG dummy correlations with CE_UES are positive and statistically significant; however, the correlation between FBG firms and income-decreasing special items are not significant but positive. Moreover, Panel B shows that the results of BIG-4 common auditors' correlation with income-decreasing special items (SPI) are negative and highly significant, indicating that BIG-4 common auditors reduce classification shifting. However, its relationship with unexpected core earnings is positive and statistically significant, which shows that BIG-4 common auditors are not reducing unexpected core earnings. Additionally, the results reported in Tables 5 show that there is no issue of multicollinearity in the data or variables, as all the values of variables come under the cut-off threshold value of 0.80.

In order to test the study's hypothesis H1, stating that “Family business groups affiliated firms are positively (negatively) associated in classification shifting compared to stand-alone companies listed on the Pakistan Stock Exchange.” This research study employed Equation (2) to investigate whether FBG-affiliated firms are involved in classification shifting or not. However, before coming to the study's hypothesis. First, this research study investigated whether PSX-listed firms are involved in classification shifting. Table 6 reported that PSX-listed non-financial firms are involved in classification shifting due to the highly significant positive coefficient of SPI and UE_UES, including FBG-affiliated firms. Likewise, these results provide evidence that PSX firms' managers misclassify core expenses are special items in order to increase core earnings. These results of this research study are consistent with the studies of Bansal and Bashir (2023); Boahen and Mamatzakis (2024); and McVay (2006). Moreover, to check the relationship of FBG-affiliated firms, another regression was carried out by interacting with the FBG dummy with SPI.

In Table 7, the interaction of FBG dummy with SPI is regressed with UE_UES. The results of Table 7 show that the interaction term of the FBG dummy and SPI is negative and statistically significant in GLS with the cluster () command −0.03** (−1.95). These results reported in Table 7 show that FBG-affiliated firms are not involved in classification shifting and do not decrease SPI to increase core earnings compared to stand-alone firms. These results may be attributed to the family's SEW and reputation as reported in the study of Khan et al. (2025). Similarly, Gavana et al. (2025) added that various dimensions of the SEW perspective, such as the family's identification with the firm, family succession aspirations, and the family's image and reputation, reduce EM. Additionally, the findings of Table 7 show that FBG-affiliated firms, compared to stand-alone firms, did not shift core expenses to income-decreasing items (non-core expenses) to boost companies' core earnings.

Furthermore, family-controlled owners use various accounting and EM choices (EM types) to strategically convey the financial reporting information. Khan (2022) reported that FBG-affiliated firms in Pakistan are involved in RAM, while they are not involved in AEM. However, the EM behavior of management damages the future value of firms and family firms' socio-emotional endowment. Therefore, in PSX-listed firms, the FBG-affiliated firms' controlling shareholders may not be involved in the classification due to the preservation of the SEW problem. Furthermore, Ferramosca and Allegrini (2018) added that higher stakes of a family in a firm increased the family's concerns about the family's reputation, and they avoided being involved in an accounting scandal to ruin the family SEW.

In the same line, Gomez-Mejia et al. (2016) posited that in family firms, major managerial choices are driven to enhance and preserve the controlling family's SEW, distant from economic efficiency. However, the EM behavior of management damages the future value of firms and family firms' socio-emotional endowment. Therefore, in PSX-listed firms, the FBG-affiliated firms' controlling shareholders may not be involved in the classification due to the preservation of the SEW problem.

In order to test the study's hypothesis H2, stating that “The common auditors in the family business groups affiliated firms in Pakistan mitigate earnings management via classification shifting.” This research study employed equation 3a to investigate whether common auditors in FBG-affiliated firms mitigate the classification shifting or not. In Table 8, the common auditor's dummy variable is interacted with SPI (income-decreasing special items) and FBG (family business group dummy) and regressed with UE_UES. The results of Table 8 show that the interaction term of common auditors (COM-AUD) with the FBG dummy and SPI is negative and statistically significant at 1% in the GLS Robust −0.37*** (−3.23) with the cluster () command to overcome the heteroscedasticity, multicollinearity and serial correlation. These results reported in Table 8 show that in FBG-affiliated firms, common auditors are effectively controlling classification shifting. Hence, H2 of the study is accepted, which states that common auditors in the FBG-affiliated firms mitigate EM via classification shifting. The effectiveness of common auditors in the FBG-affiliated firms is due to their knowledge in a particular business group and its affiliates; as a result, they increase the audit quality and reduce the management's opportunistic behavior like EM (Fu and Kim, 2024; Khan and Kamal, 2024b). On similar grounds, Fan et al. (2022) added that common auditors build and transfer auditors’ expertise, increase auditors’ connections and act as knowledge transfer pipes between group affiliates, which leads to better financial reporting quality.

Moreover, the results of our study supported the agency theory and knowledge transfer hypothesis. According to our study's findings, common auditors in the FBG-affiliated firms curb classification shifting, which may act as pipes of information and knowledge among various affiliates of FBGs through which financial and audit information is transferred. Hence, through common auditors' information asymmetry is reduced, financial reporting quality is improved, and classification shifting in the FBG-affiliated firms is reduced.

The results of Table 9 show the testing of the H3 of the study, which states that “The BIG-4 common auditor(s) for a business group affiliate will mitigate earnings management via classification shifting in Pakistan Stock Exchange-listed firms.”

Hence, to test the third hypothesis, the BIG-4 common auditor's dummy variable (BIG-4-COM-AUD) interacted with SPI (income-decreasing special items) and FBG and regressed at UE_UES. The results in regression 1, depicted in Table 9, show that the interaction term of FBG*SPI*BIG-4-COM-AUD is negative and highly significant. These results indicate that BIG-4 common auditors mitigate EM via classification shifting in the FBG-affiliated firms in Pakistan. Thus, the results reported in Table 9 support the H3 of the study, that BIG-4 common auditors in FBG-affiliated firms mitigate classification shifting. Moreover, these results may be aligned with the study of Khan and Kamal (2024b). Khan and Kamal (2024b) reported that the same network auditors in the FBG-affiliated firms are effective in controlling AEM. Therefore, in Pakistan, the BIG-4 common auditors effectively deter classification shifting in the FBG-affiliated firms and perform the role of independent external auditors as mandated by the agency theory. Hence, to secure and prevent minority shareholders' rights from the expropriation of controlling shareholders’ in FBG-affiliated firms, as well as to resolve the principal-principal agency conflict, independent auditors are paramount in the shape of BIG-4 common auditors.

These results are consistent with the study of Viana et al. (2022). According to Viana et al. (2022), BIG-4 auditors are associated with higher audit quality and limit EM due to their market incentives, linked with their market-based reputation and legal concerns. In addition, the results of this research are in line with the study of Mulchandani and Mulchandani (2022), who reported that BIG-4 audit firms protect their brand image and have a global business presence, and they will not damage their image if their audited report finds manipulation. Moreover, the results reported in Table 9 are the same as those reported in the study of Khan and Kamal (2024b), which states that in Pakistan, BIG-4 audit firms reduce EM via accrual-based due to their brand image, goodwill and maintain their independence irrespective of how important and large their client (business group).

Moreover, the results reported in Table 9 are in line with the study of Fan and Wong (2005), who reported that in emerging markets, BIG-4 auditors play an external corporate governance role while taking into consideration the agency conflict. Moreover, FBG-affiliated firms may engage and appoint the BIG-4 auditors in order to improve the audit quality, attract potential investors by signaling transparency and credibility in the financial reporting and reduce auditing costs associated with coordination, along with audit efficiency (Yang et al., 2016). Similarly, Nagar et al. (2021) added that BIG-4 auditors act as a superior EM deterrent, and due to their global operations, strong reputation and high incentive for superior audit quality, they prevent classification shifting.

According to Boahen and Mamatzakis (2024), during the calculation or estimation process of classification shifting proxy, researchers rely on the figures reported in the financial statements of the companies. These figures reported in the financial statements may be subject to measurement errors. Therefore, to counter these measurement errors, it is necessary to make a robust analysis by replacing or defining key variables in the classification shifting model (Boahen and Mamatzakis, 2024). Hence, in the following sections, we replace total accruals with working capital accruals and drop the current period accruals in the classification shifting model.

In this research study, an alternative proxy of classification shifting has been used by replacing total accruals with working capital accruals (WCA). Following Bansal (2023), in order to nullify the impact of accruals such as depreciation and other non-recurring accruals in the UE-CES model, in the working capital accruals (UE_CE_WCA) model, we replace total accruals with WCA. In addition, the WCA considers the impact of accounts receivable and stock write-offs (Bansal et al., 2021). Table 10 illustrates the results in GLS robust regression 1, which are reported in order to report results without any biases of panel data regressions. The reported results in Table 10 regression 1 show that the triple interaction term of FBG, SPI and common auditors (FBG_AFFI # COM-AUD # SPI) is negative and highly significant, indicating that common auditors in the FBG affiliates are effective at curbing classification shifting (alternative proxy). Additionally, these results are similar to the main results reported in Section 4.3 and Table 8.

In addition, the BIG-4 common auditors are also linked with the classification-shifting model estimated via WCA in Table 11. The results of regression 1 in Table 11 show that BIG-4 common auditors (BIG-4 COM-AUD) interacting with SPI and FBG dummy (FBG_AFFI # BIG-4 COM-AUD # SPI) is negative and statistically significant in GLS regressions. These results show that BIG-4 common auditors in FBG-affiliated firms are effective in controlling classification shifting, which is calculated with working capital accruals.

To validate McVay's 2006 model, we have dropped the current accruals from the McVay (2006) model. Moreover, McVay (2006) added current accruals in the classification model to control the firm's performance. However, when she dropped it from the model, the relationship between income-decreasing special items (SPI) and unexpected core earnings became inverse and she added that the evidence in the core earnings expectation model of the classification shifting in the US firms might be attributed to the current accruals.

Besides, according to Fan et al. (2010), McVay's (2006) model might prompt a mechanical relationship between UE_CES and SPI, because both the income statement items contained the current accruals. Hence, keeping in view the aforementioned points, we dropped the current accruals component from the classification model and re-ran it. The results of Tables 10–12 reported the classification shifting model without containing the current accruals component (CE_WOCA). Moreover, the findings in Table 12 in regression 1 and 2 show that the results were the same as reported in Section 4.2, which reported that all types of PSX-listed firms are involved in the classification shifting. Similarly, FBG-affiliated firms' results were also reported in Table 12 in regressions 3 and 4, indicating that FBG-affiliated firms are not involved in classification shifting when we dropped in current accruals and contained working capital accruals.

Similarly, in order to check whether the results reported in Section 4.3 are the same and valid with the classification shifting model after dropping the current accruals components from the original model of McVay's (2006) model. The findings in Table 10 in regression 2 depicted that common auditors in the FBG-affiliated firms are effective in curbing EM, as the coefficient and t-statistics in regression 2 of common auditors in the FBG with the classification shifting proxy without current period accruals (CE_WOCA) are negative and statistically significant.

Moreover, in Table 11, the BIG-4 common auditors (BIG-4 COM-AUD) are linked with the classification shifting model without containing current accruals (CE_WOCA), and the results in regression 2 report that it is negative and statistically significant. These results support the study's hypothesis that FBG affiliates have common auditors and BIG-4 common auditors are not involved in classification shifting and that BIG-4 auditors, as an external corporate governance mechanism, play an important role in mitigating EM via classification shifting.

All the aforementioned reported results supported the study's hypotheses, i.e. FBG-affiliated firms are not increasing core earnings via classification shifting and common auditors, and BIG-4 common auditors are effective in curbing EM via classification shifting in FBG-affiliated firms. Although the above results are robust with alternative definitions of classification shifting along with robust panel data analysis. However, this study's results may be biased due to the endogeneity issue. Hence, the instrumental variables (2SLS) regression has been used. Moreover, in the instrumental variable (2SLS) two-step process was implemented. In the first step, FBG-affiliated firms are regressed for the purpose of obtaining the predicated value (P_FBG). In the second step, the fitted values of P_FBG are employed in the structural equation for tackling endogeneity. The results of instrumental variables (2SLS regression) are reported in Table 13. The results of instrumental variables (2SLS regression) reported that in the FBG-affiliated firms of PSX classification shifting is not prevailing, and BIG-4 common auditors, along with common auditors, have effective tools of external governance and have a vital role in curbing classification shifting after controlling for the endogeneity issue.

The FBG-affiliated firms' managers may employ classification shifting along with AEM or RAM simultaneously. Additionally, the methodology adopted in this research primarily focuses on classification shifting, potentially leading to a correlated omitted variable problem. Managers may employ multiple EM techniques simultaneously. Without controlling AEM and RAM, the study's findings on classification shifting could be biased. Therefore, in order to tackle this issue, in the following Table 14, AEM estimated via (Kothari et al., 2005) and RAM proxies estimated via (Roychowdhury, 2006) are taken as control variables. The results are the same as reported in the main analysis after controlling for AEM and RAM.