本文总结了我过去在教学与研究中遇到的困难与困惑的思考。这些思考尚不成熟且浅薄，但因它们 与我交流过的许多会计学界同行的困惑相呼应，故而值得分享，希望能引起更多开放式讨论，推动有意 义的变革。更重要的是，我坚信这些困惑背后的问题，正是导致会计学科吸引力下降、人们对会计作为 学术领域或职业道路丧失信心的根本原因。

In China, the Confucian value system with benevolence and justice as the core and the governance proposition based on moral governance had been basically finalized in the pre-Qin period, providing ideological support for traditional Chinese moral governance (Li Feng, 2018), and also receiving high attention in modern Chinese society (Xiang Yuqiao, 2016; Wu Chanxin, 2019). In modern Chinese society, moral governance is an indispensable governance means and driving force to promote the modernization of the national governance system and governance capacity (Long Jingyun, 2015). It is worth noting that Chinese society, influenced by the governance thought of “virtue is the principal, punishment is supplemented, and rites and laws are combined” in traditional Chinese culture, not only advocates the promotion of people’s inner virtue through education, self cultivation and moral practice, but also emphasizes the selection of virtuous and capable people to participate in governance. However, the agency theory (Jensen and Meckling, 1976), which lays the foundation for the study of corporate finance and corporate governance, is based on the assumption of “economic man” and holds that the agent’s self-interest after the separation of ownership and control rights will lead to agency conflicts and agency costs. Although Smith (1776; 1759) has long pointed out that egoism and altruism are the two basic motivations for human behavior, because it is difficult to change human nature, agency theory only focuses on the role of institutional structure, contracts and informal arrangements in reducing agency conflicts and increasing the interests of both parties to the transaction (Jensen, 1994). However, as for the role of personal morality, some studies have pointed out that personal morality in the exchange link of economic activities is conducive to correcting the excessive profit-seeking motive of exchange, reducing the opportunistic behavior caused by information asymmetry in the exchange process, and reducing transaction costs (Wang Xiaoxi and Yang Wenbing, 2002; Hua Guihong and Wang Xiaoxi, 2004. As a self-motivation and restraint mechanism, personal morality can provide the internal incentive and restraint of individual behavior and make up for the deficiency of external system (Wesson, 2002; Liu Liangbi and Liu Minggui, 2001; Wang Xiaoxi and Yang Wenbing, 2002; Hua Guihong and Wang Xiaoxi, 2004; Li Yining, 2010; Yu Dahuai, 2013). In addition, relevant research in the field of corporate governance have also pointed out that individual managers will become “moral people” with a certain level of morality in the process of their growth and socialization under the influence of their social culture, and the self-motivation and restraint of managers’ ethics enable them to play an important role in reducing agency conflicts and improving corporate governance efficiency (Chen Donghua et al., 2017; Zhu Juan and Chen Donghua, 2024). However, due to the difficulty in quantifying managers’ morality, the existing research on corporate finance and corporate governance, which is mainly based on the empirical research paradigm, has not directly tested the role of managers’ personal characteristics such as managers’ morality. This provides an opportunity for further empirical research on the governance function of managers’ morality. Accounting information is the key basis for internal and external information users of a company to make decisions. Its quality will exert an important influence on the decision-making efficiency of information users, especially external information users of a company, and thus exert an important influence on the normal operation and healthy development of the capital market and market economy. Therefore, research on the quality of corporate accounting information is an important issue in the research on corporate finance and corporate governance (Lu Dong et al., 2012; Ruan Rui et al., 2021; Liu Guangqiang and Wang Di, 2021). Most existing studies believe that the behavior of managers who control the operation of a company in accounting information disclosure is an important factor affecting the quality of accounting information, and managers who aim to maximize their own interests will hide the real accounting information in accounting information disclosure and reduce the quality of accounting information of the company (Healy, 1985; Sweeney, 1994; DeFond and Jiambalvo, 1994; Guidry et al., 1999; Jin Ming, 2000; Cheng and Warfield, 2005; Bergstresser and Philippon, 2006; Li Yanxi et al., 2007; McAnally et al., 2008; Su Dongwei and Lin Dapang, 2010; Jayaraman and Milbourn, 2015; Hsieh et al., 2018). These studies regard managers as “economic men” who seek to maximize their own interests, ignoring their personal morality and its possible role. Deng Deqiang et al. (2014) and Chung and Hsu (2017) studied the restraining effect of individual moral cognition on individual behavior from the perspective of individual moral cognition, and found through experimental research, individual moral cognition can significantly promote honest behavior in budget reporting and management reporting so as to restrain budget slack and reduce the damage to corporate interests. However, the level of individual moral cognition is not completely equivalent to the level of individual morality. Based on the existing relevant studies, this paper intends to directly study the influence of managers’ morality on the quality of corporate accounting information disclosure by using experimental research methods based on the Chinese sociocultural context. Therefore, the first question facing this paper is whether a measurement tool can be found in the experimental research to effectively measure the morality of managers in the social and cultural background of China. As for the measurement of personal morality, among the existing relevant measurement methods, the Scale of Moral Behavior and the Scale of Controversial Moral Behavior can be used to measure the moral status of the subjects to a certain extent (Retting and Pasamanick, 1959; Harding and Phillips, 1986). In addition, the Defining Issues Test (DIT), proposed by Rest (1979) and revised by Thoma et al. (1999), can be used to infer the stage of moral cognitive development of test subjects according to their answers. The Moral Identity Measure (MIM) compiled by Aquino and Reed (2002) sets measurement items from two dimensions, implicit and explicit, to measure the moral identity of the subjects. Among them, implicit identity refers to the importance of moral traits in the self-scope. Explicit identification reflects the extent to which individuals display these traits in social behavior through personal behaviors such as dress, hobbies, and behavior. However, as Yang Zhongfang put it in the preface to his translation of Personality and Psychosocial Measurement (Robinson et al., 1997), “The conception and production of the scale have their specific historical, cultural and social backgrounds and limitations; the questions in the scale reflect the common concerns of everyone in a certain era and a certain society and culture. For people in the context of Chinese social history and culture, these tests may cause difficulties in understanding, and may not be able to measure the concept that they really want to measure because the subject thinks that the content of the test is irrelevant.” In order to effectively assess adult morality under Chinese social and cultural background, based on the connotation of personal morality in Chinese social and cultural context, this paper develops the Personal Morality Scale to measure the moral level of subjects who simulate the role of managers in the experiment. On the basis of the above, this paper studies the influence of the moral level of the subjects in the role of simulated managers on the possibility of hiding real accounting information by simulating the situation of managers’ accounting information disclosure decisions with conflicts between managers and shareholder agents. The research finds that, under the control of other conditions, the higher the moral level of managers, the less likely they are to hide the true accounting information, so the higher the quality of corporate accounting information disclosure. Further, the study examines the influence of managers’ morality on external governance mechanisms. It is found that the higher the managers’ moral level, the smaller the role of external supervision mechanism in reducing the possibility of managers hiding true accounting information, and the smaller the role of managers’ shareholding in reducing the possibility of managers hiding true accounting information, that is, managers’ morality will significantly replace the role of external governance mechanism. In addition, in order to explore whether managers’ morality can also affect their decision-making behaviors other than accounting information disclosure, so as to further clarify the role that managers who control the company’s operation can play in corporate decision-making related to the interests of external investors, this paper also studies the impact of managers’ morality on their investment decisions and thus on the investment efficiency of the company. It is found that under other conditions, the higher the managers’ moral level is, the less likely they are to over-invest and under-invest, and the higher the investment efficiency of the company. The possible research contributions of this paper are as follows: (1) Focusing on the quality of corporate accounting information disclosure, this paper provides direct empirical evidence for the role of managers’morality in corporate governance by constraining their behaviors and alleviating agency conflicts, and forms an important supplement to existing relevant studies (Chen Donghua et al., 2017; Zhu Juan and Chen Donghua, 2024). At the same time, it also provides direct empirical evidence that managers’ morality can affect the role of external governance mechanism, which indicates that managers’ morality can partially replace external governance mechanisms and become an important internal governance mechanism to restrain managers’ behavior and alleviate agency conflicts. This research conclusion contributes to a better understanding of agency conflict and corporate governance in China’s social and cultural context. At the same time, it partly explains why North (1990) said that in different countries, legal systems play different roles in restricting people’s behavior, reducing transaction costs and promoting economic performance. (2) This paper directly discusses the moral level of managers and their role in governance, which is an important echo of relevant studies on butler theory. According to the butler theory, managers are not always opportunists pursuing the maximization of their own interests, but in some cases are dutiful “stewards”, whose behavior is driven by a sense of accomplishment and mission, and aims to pursue the maximization of the interests of the client (Miller and Le Breton-Miller, 2006; Miller et al., 2008; Dominguez-Escrig et al., 2019), therefore, governance strategies guided by the butler theory are different from those guided by the agent theory. According to the butler theory, the key to corporate governance is not that owners supervise and control managers or give managers material incentives, but that owners trust managers and encourage managers to maximize their enthusiasm and creativity through authorization, coordination or spiritual incentives to create more value for the company or owners (Donaldson, 1990; Donaldson and Davis, 1991; Davis et al., 1997; Muth and Donaldson, 1998; Sundaramurthy and Lewis, 2003). This paper has formed an important echo with these studies of butler theory, and further combined with the social and cultural background of China, from the perspective of managers’ moral level, the governance logic of managers’ morality is systematically discussed and empirically tested. (3) By developing the Personal Morality Scale to measure the moral level of the subjects who simulated the role of managers in the experiment, this paper makes it possible to directly test the governance role of managers’ morality under experimental conditions, and provides reference for subsequent relevant studies. At the same time, the use of experimental research methods can control the influence of non research concerns on the research conclusions more accurately by controlling the experimental situation, making the research conclusions more credible and providing reference ideas and directions for future related research.

Corporate misconduct represents a persistent challenge in capital markets worldwide, undermining investor confidence and market integrity. However, despite intensified regulatory efforts by China’s securities regulators, violations by listed companies remain endemic. In 2023 alone, regulatory authorities imposed 2,555 penalties on listed companies, with aggregate fines exceeding 12 billion yuan. This troubling pattern highlights the urgent need to identify novel approaches to constrain corporate malfeasance. The academic literature has extensively documented various determinants of corporate violations, examining both internal organizational factors such as corporate strategy, board composition, executive compensation structures, and ownership configurations, as well as external environmental factors including media scrutiny, analyst coverage, regulatory enforcement, audit quality, and market conditions. However, a critical limitation of this research stream is its predominant focus on monitoring mechanisms per se, with insufficient attention devoted to the practical challenges of implementing effective oversight. Specifically, geographical distance between companies and their monitors creates significant information asymmetry and substantially elevates monitoring costs, potentially undermining the effectiveness of otherwise well-designed governance mechanisms. This study introduces a novel perspective by examining how major transportation infrastructure development—specifically the expansion of China’s high-speed rail network—affects corporate violations through fundamental improvements in information acquisition and monitoring capabilities. The rapid construction of HSR infrastructure since 2008 provides an ideal quasi-natural experimental setting for investigating how reducing geographical barriers influences corporate behavior. HSR dramatically accelerates inter-city personnel mobility and information transmission, generating pronounced time- space compression effects that fundamentally reshape the information environment surrounding corporations. Through enhanced face-to-face interactions and more efficient information flows, HSR opening potentially strengthens external monitoring while simultaneously increasing the difficulty of concealing corporate misconduct. The theoretical foundation underlying this research integrates insights from information economics, agency theory, and corporate governance literature. Three interconnected mechanisms explain how HSR opening constrains corporate violations.First, the information asymmetry mechanism recognizes that inadequate information represents a primary obstacle to effective corporate monitoring. When monitoring is absent or ineffective, managers face stronger incentives and greater opportunities to engage in value-destroying activities or fraud. The acquisition, verification, and analysis of corporate information entail substantial costs that increase with geographical distance. HSR opening dramatically reduces these information frictions by facilitating more frequent and cost-effective information gathering through multiple channels including on-site investigations, informal communications, and expanded social networks. By lowering information acquisition costs, HSR enables various stakeholders—including institutional investors, financial analysts, media outlets, and regulatory agencies—to more actively monitor corporate behavior. Second, the monitoring efficiency mechanism emphasizes how HSR opening directly enhances the effectiveness of various external governance mechanisms. For financial analysts, reduced travel costs and time enable more frequent company visits and field research, improving forecast accuracy and information production. For external auditors, HSR alleviates geographical constraints, reducing audit delays and enhancing audit quality through more thorough on-site examinations. For independent directors, particularly those residing far from company headquarters, HSR substantially reduces meeting attendance costs, enabling more active participation in board deliberations and oversight activities. For regulatory agencies, HSR facilitates more frequent surprise inspections and investigations, strengthening deterrence effects. Collectively, these improvements in monitoring efficiency raise the expected costs of corporate misconduct. Third, the information transparency mechanism highlights how HSR opening improves the overall corporate information environment by increasing both the quantity and quality of publicly available information. With enhanced information access, external parties can more accurately detect suspicious activities and potential violations, while managers face greater difficulty in concealing unfavorable information. Moreover, improved information transparency amplifies the reputational and stock price consequences of revealed violations, creating stronger incentives for managers to maintain compliance. Building on this theoretical framework, the study advances the central hypothesis that HSR opening significantly reduces corporate violation frequency and severity. Furthermore, two specific mechanisms are proposed: HSR opening enhances corporate transparency by reducing information asymmetry, and HSR opening increases independent director meeting attendance by lowering travel costs, both of which constrain corporate violations. The empirical analysis employs a difference-in-differences identification strategy that exploits the staggered rollout of HSR stations across Chinese cities as a quasi-natural experiment. The research sample comprises all non-financial A-share companies listed on the Shanghai and Shenzhen stock exchanges from 2003 to 2018, yielding 30,431 firm-year observations after excluding financial firms, companies with negative equity, ST companies, and observations with missing data. The sample period begins in 2003, several years before China’s f irst HSR line opened in 2008, ensuring adequate pre-treatment observations for validating the parallel trends assumption. The dependent variables capture corporate violations through two complementary measures. The first is a binary indicator equal to one if the company experienced any regulatory violation during the year, and zero otherwise, analyzed using logit regression models. The second measure counts the total number of violations during the year, analyzed using Tobit regression models to accommodate the large proportion of zero observations. Violation data sourced from the Wind database encompasses all regulatory penalties including information disclosure violations, operational irregularities, and leadership misconduct. The key independent variable represents HSR opening status, constructed as a binary indicator equal to one if the city where the company is registered had opened at least one HSR station by year-end, and zero otherwise. This treatment variable exhibits considerable spatial and temporal variation as HSR construction proceeded gradually across different regions over the sample period.The econometric specifications control for a comprehensive set of firm level characteristics including leverage ratio, firm size, return on assets, independent director proportion, CEO chairman duality, operating cash flow, Big Four auditor engagement, institutional ownership, Tobin’s Q, and loss status. All continuous control variables are winsorized at the 1% and 99% levels to mitigate outlier influence. Year and industry fixed effects are included to absorb aggregate time trends and time-invariant industry-specific factors. Standard errors are clustered at the firm level to account for serial correlation. The baseline regression results provide robust evidence supporting the central hypothesis. HSR opening significantly reduces both the probability and frequency of corporate violations. The coefficient on the HSR opening indicator is negative and statistically significant at the 5% level across all specifications. The economic magnitude is substantial: HSR opening reduces the expected number of violations by 0.0033 times on average. HSR opening reduces corporate violations by approximately three-fold. These effects are not only statistically significant but also economically meaningful, indicating that transportation infrastructure improvements can materially influence corporate governance outcomes. The parallel trends assumption underlying the DID identification strategy is validated through event study specifications that estimate separate coefficients for each year relative to HSR opening. The results show no significant differences between treatment and control groups in the five years preceding HSR opening, with significant negative effects emerging only in the first year after opening and persisting thereafter. This pattern confirms that the observed effects reflect causal impacts of HSR opening rather than pre-existing trends. Multiple robustness checks confirm the reliability of the main findings. Placebo tests using hypothetical HSR opening dates seven years prior to actual opening yield insignificant coefficients, ruling out spurious correlations driven by unobserved regional trends. Controlling for other transportation infrastructure including conventional railways, highways, and air routes does not materially affect the results, indicating that HSR’s impact operates through distinct mechanisms beyond general transportation improvements. Excluding major cities where sample concentration might create estimation bias yields qualitatively identical results. Alternative measures of violations including violation types and severity produce consistent findings. To elucidate the pathways through which HSR opening constrains corporate violations, the study conducts systematic mechanism analysis examining two primary channels. First, HSR opening enhances information transparency. Using discretionary total accruals as a proxy for information opacity, interaction term analysis reveals that HSR opening’s restraining effect on violations is significantly stronger for firms with higher ex ante information asymmetry. This finding confirms that improved information environments constitute a key channel through which HSR reduces violations. Second, HSR opening increases independent director meeting attendance, thereby strengthening board monitoring. Using the ratio of independent director absences to required meetings as a measure of monitoring intensity, the analysis demonstrates that HSR opening significantly reduces absence rates, and this reduction mediates part of HSR’s impact on violations. The interaction term between HSR opening and absence rate is significantly negative, confirming enhanced monitoring as a mechanism. Heterogeneity analysis across three dimensions reveals important conditional effects. First, HSR opening more significantly constrains violations in firms with higher agency costs, measured by administrative expense ratios. Second, the effect is stronger for firms facing tighter financing constraints, measured by the SA index. Third, examining violation types shows that HSR opening most significantly impacts information-related violations while having weaker effects on operational and leadership violations, consistent with information mechanisms playing a central role. Beyond direct effects on corporate violations, HSR opening generates two important spillover benefits. First, it improves regional rule of law environments. As listed companies reduce violations, they contribute to strengthened legal institutions through demonstration effects and reduced enforcement burdens. Using the Fan Gang Index of marketization, the analysis confirms that HSR opening significantly enhances legal environment quality in affected regions.Second, HSR opening strengthens urban economic resilience. Listed companies represent critical components of local economies, and their improved governance enhances economic stability. Constructing a comprehensive economic resilience index across three dimensions—resistance and recovery capacity, adaptation and adjustment capacity, and transformation and development capacity—the study demonstrates HSR opening’s positive impact on urban economic resilience, highlighting broader welfare benefits beyond corporate governance improvements. This research makes three significant contributions to the literature. First, it identifies transportation infrastructure as a novel determinant of corporate violations, extending beyond traditional governance factors. By documenting how geographical accessibility affects corporate behavior through information and monitoring channels, this study bridges infrastructure economics and corporate governance research. Second, methodologically, the study employs a credible identification strategy using quasi-experimental variation from HSR rollout to establish causal effects, overcoming endogeneity concerns that plague much corporate governance research. The staggered DID design, validated through parallel trends tests and robustness checks, provides compelling evidence that HSR opening causes reduced violations.Third, theoretically, the study illuminates specific mechanisms linking infrastructure development to corporate behavior, advancing understanding of how geography shapes governance through distinct information acquisition and monitoring efficiency channels. By documenting both direct effects on monitoring costs and indirect effects through information environment improvements, the research enriches our understanding of how external governance mechanisms operate in practice. The findings yield important policy implications. Continued public investment in transportation infrastructure not only generates traditional economic benefits but also produces significant governance externalities by strengthening corporate oversight and reducing misconduct. Policymakers should consider these governance spillovers when evaluating infrastructure investments. Regulators should implement differentiated supervision strategies focusing monitoring resources on geographically isolated firms where information frictions are most severe. Enhancing mandatory information disclosure requirements can complement infrastructure improvements in constraining violations. Future research could explore other infrastructure types, examine international contexts, and investigate longer-term effects to build on these foundational findings.

The accrual basis is the essence of modern accounting. Under the accrual basis, accounting profits can be divided into accrual profits and cash profits. By recognizing accrual profits based on the realization principle and matching principle in contrast to cash profits, accrual profits incorporate management’s specific information advantage regarding the company’s operating activities, providing incremental information beyond cash flow. However, this also creates opportunities for management to manipulate profits. Understanding accrual items is a key focus of accounting academic research. Among these, earnings management research closely related to accrual surplus has become an important branch of accounting, developing various models to measure earnings management, such as the basic Jones model, modified Jones model, intangible asset Jones model, and revenue matching Jones model. The basic principle of these models is that firms within the same industry or the same f irm over a long period of time share similar business characteristics that drive accrual items. Accrual items that cannot be explained by these characteristics are more likely to be manipulated by management. The differences between these models lie in the corporate characteristics used to calculate normal accruals. These models were mostly developed based on accounting practices in Western markets but have been widely adopted by domestic scholars. This paper argues that accrual items are significantly influenced by a firm’s business model. Interpreting the accrual items of Chinese listed companies requires consideration of their unique business models. This paper will examine how China’s relationship-based business model influences a company’s accrual items. Influenced by China’s institutional environment, Chinese firms commonly adopt relationship-based transaction modes, meaning that transactions between parties are not entirely based on fair market prices but rely on trust built through long-term interactions. This leads to relationship-based transactions exhibiting strong personalization, significantly different from the market-based transaction model which is typical in Western markets. As accounting is the written language of economic activities, the measurement of accrual items may be affected by the particularity of China’s transaction mode. In practice, many firms implement differentiated management of accounts receivable collection periods by offering more favorable credit terms to major supply chain partners within their relational networks to maintain cooperation. Therefore, the existence of relational supplier-customer transactions may lead to behaviors in accrual measurement that cannot be fully explained by the variables in existing accrual models. Consequently, existing accrual models may not be suitable for China’s relationship-based transaction background, thus affecting the accuracy of previous research conclusions. Based on the above analysis, this study empirically investigates the following questions: Does the existence of relationship-based supplier-customer transactions increase a firm’s accrual levels? If it does, what is the specific nature of these incremental accruals? Are they normal accruals or earnings management? Additionally, does the meaning of these incremental accruals differ by firm size? Finally, does the presence of relationship-based supplier-customer transactions weaken the effectiveness of existing accrual models? Through a review of the literature, this study finds that in the field of earnings management, accrual models (represented by the basic Jones model) are widely used in empirical accounting research in China. Existing research consistently indicates that supplier-related relational transactions are positively associated with the level of corporate earnings management (Fang and Zhang, 2016; Xu et al., 2015). However, these studies generally assume that accrual models developed in Western markets remain applicable in the Chinese context, overlooking the impact of China’s relational transaction patterns on the effectiveness of such models. Although relationship based transactions also exist in Western countries, they differ significantly from those in China in terms of transaction frequency, nature, and institutional drivers (Li, 2017). Current accrual models do not account for the distinct characteristics of relationship-based transactions in China or their influence on accounting information, which likely makes these models unsuitable for Chinese samples and may undermine the accuracy of research conclusions. According to the “Institution-Market-Firm” logic, firms’ accounting choices are endogenous to their transaction modes, which are further constrained by the institutional environment. Li (2017) points out that China’s institutional environment, arrangements, and socio-cultural factors impose high costs on market transactions, causing Chinese firms to more often choose relational transactions. The impact of relationship based transactions on accrual items may come from two aspects. First, relationship-based transactions may lead firms to conduct more credit-based transactions with legitimate business reasons, thereby increasing normal accrual levels. In relationship-based transactions, relationship-specific assets formed through long-term frequent transactions have high investment irreversibility and specificity, greatly increasing default costs for both parties, which exerts a governance role and fostering mutual trust. Based on this trust and mutual understanding, firms may be more willing to conduct credit-based transactions with network members. Additionally, due to the long term and frequent nature of relationship-based transactions, relying solely on cash payments would incur high transaction costs and reduce supply chain efficiency. Hence, firms prefer credit-based transactions for cost reasons (Petersen et al., 1997). Second, relationship-based transactions may also encourage firms to engage in earnings management, thereby affecting accrual levels. Firms involved in relationship-based transactions often communicate privately rather than rely on public information (Ball, 2000). Thus, firms may lack incentives to provide high-quality accounting information. Moreover, relationship-based transactions tightly bind the interests of parties, and firms may collude with suppliers and customers to manipulate accounting information to maintain stable relationships and relationship-specific assets (Gong et al., 2022). Therefore, relationship-based transactions may influence accruals through both normal accruals and earnings management. Furthermore, firms vary in their reliance on relational transactions, and the relational networks they form are unique, reducing the comparability of accrual items between firms. Hence, the “homogeneity among peers” assumption underpinning existing accrual models may no longer hold, affecting these models’ applicability in China. Based on the above analysis, this study explores the specific meaning of accruals under China’s relational transaction background and the applicability of existing accrual models according to the following research approach. First, the study tests whether relationship-based supplier-customer transactions change firms’ accrual measurement behavior—that is, such transactions may increase firms’ total accrual levels. Second, it examines whether the accruals caused by relational transactions represent earnings management. Drawing on the instrumental variable approach, total accruals are regressed on supplier-customer concentration and other control variables, decomposing total accruals into parts related to and unrelated to relationship-based transactions. By testing the correlations of these two parts with corporate violations or audit opinions, the specific nature of incremental accruals caused by relationship-based transactions is further clarified. Additionally, by distinguishing firm size, the heterogeneity of the specific meanings in large and small firms is examined. Finally, samples are grouped by supplier-customer concentration to test whether the correlation between discretionary accruals calculated by existing models and corporate violations changes at different levels of relationship-based transactions. A declining correlation with increasing relationship-based transaction level indicates that existing accrual models’ effectiveness is weakened in firms with more relationship-based transactions. This paper examines the intrinsic relationship between supplier-customer relationship-based transactions and corporate accounting accruals using a sample of Chinese A-share listed companies from 2011 to 2021. The study finds that supplier-customer relationship-based transactions increase a company’s total accrual level. Furthermore, the incremental accruals resulting from supplier-customer relationship-based transactions are not significantly correlated with corporate fraud, but these accruals increase the likelihood of non-standard audit opinions. This suggests that incremental accruals are more likely to be normal accruals but are sufficient to attract auditors’ attention. Furthermore, it is found that incremental accruals in large companies are normal accruals and are unlikely to attract auditors’ attention, while in small companies they indicate earnings management and are likely to raise auditors’ concerns. Economic consequence tests show that incremental accruals resulting from relationship-based transactions do not increase the risk of stock price crashes in future periods. Finally, this study finds that supplier-customer relationship-based transactions weaken the predictive ability of existing accrual models for corporate misconduct. 2025-1 Vol 3 This study contributes in three ways. First, it clarifies the specific meaning of accruals in the relationship based transaction context. Xu et al. (2015) found that as the proportion of supplier-customer relationship based transactions increases, firms engage in more positive earnings management. Similarly, Fang and Zhang (2016) found that supplier-customer concentration partly explains earnings management levels. However, these studies assume existing accrual models still accurately measure discretionary accruals even when firms engage in many relationship-based transactions. This paper relaxes this assumption and uses corporate violations as a proxy for earnings management. The insignificant correlation between incremental accruals from relationship based transactions and violations suggests these accruals are more likely normal accruals, thus filling gaps in the existing literature. Second, the paper investigates the applicability of existing accrual models in relationship-based transaction contexts. Existing models, based on Western accounting practices, assume homogeneity among peers and interpret the unexplained part of accruals as discretionary accruals. However, due to relationship-based transactions, peer firms may apply very different accounting treatments to accrual items in China, violating this assumption and impacting model applicability. The study finds increasing relationship-based transactions weaken the correlation between discretionary accruals and violations, indicating reduced model effectiveness and offering new directions for model improvement. Third, the findings help regulators better understand corporate accounting information and improve regulatory efficiency. Enforcement announcements and inquiry letters show that earnings management has long been a regulatory focus. The results indicate whether accruals imply earnings management depends on firms’ transaction modes. Also, accruals from relationship-based transactions in large firms are more likely normal accruals, while in small firms more likely earnings management. This can help regulators clarify enforcement priorities and focus. Future research should understand the complex relationship between business activities and accounting figures in China’s market environment. Ball (2013) and Ball (2024) noted that accounting researchers often assume earnings management is pervasive, possibly due to their strong ethical biases, limited understanding of accrual determinants, and overlooking omitted variables for significant findings, while in fact actual earnings management may not be that widespread. Accruals may stem from fundamentals or earnings management, but existing Chinese earnings management research lacks consideration of relationship-based transactions, which may overestimate earnings management. Lastly, there still remain some limitations. First, this study does not propose a new accrual model, which lays a foundation for future improvements. Second, it only considers supplier-customer relationships, ignoring other stakeholder ties that may affect accruals. Third, the conclusion that relationship-based transactions increase accruals may be endogenous. Future research may explore suitable instrumental variables to confirm the robustness.

This study investigates how relational-contract-based governance reshapes both the distribution and the aggregate magnitude of the costs of real earnings management in China’s listed-firm economy. Prior research shows that long-term ties can make real earnings management easier to execute. Yet much less is known about whether, in settings where relational governance is pervasive, the economic burden of such actions is shared across firms connected by those ties. Focusing on Chinese “Xi Zu” groups (business groups organized around a common ultimate controller), we ask whether lineage-based coordination lowers the initiator’s direct cost of manipulating real activities, whether it transfers costs to affiliated firms through internal transactions and resource reallocation, and what the net consequences are at the listed-company level once both direct effects and spillovers are considered. Our motivation rests on two observations. First, in Western market environments built around complete contracting and arm’s-length pricing, firms that engage in real earnings management typically bear the cost themselves. To induce unrelated counterparties to cooperate in non-routine trades—accelerated shipments, unusually large orders, temporary price concessions, or discretionary spending cuts—the initiating firm must compensate these counterparties and incur search and negotiation frictions. Decades of evidence therefore links real earnings management to value destruction for the initiator through distorted operating choices that depress future performance. Second, Chinese markets feature a different organizational logic. Alongside market transactions, many exchanges are governed by relational contracts. Within capital lineages, common controllers, shared information, and long-standing cooperation create a governance structure in which affiliated firms can be mobilized to support short-term reporting goals. In such contexts, a single transaction need not be profitable for the cooperating affiliate if reciprocity and controller coordination promise future benefits, if communication costs are low, and if the reputation embedded in the relationship enforces compliance. These features can reduce the explicit concessions and frictions borne by the initiator. But they do not eliminate the underlying economic distortion. Instead, they can relocate it: affiliates that help to realize the initiator’s target may face capacity misallocation, accept non-market pricing, postpone their own projects, or realign resources in ways that erode their investment and future profitability. In short, relational governance may transform the incidence of real earnings management costs from a firm-level burden to a network-level redistribution. Relational contractual arrangements are widespread in China's institutional environment, manifesting themselves in the form of a one-share shareholding structure, centralized supplier-customer relationships, more prevalent political-business affiliations, and numerous capital groups (“Xi Zu” groups). Among them, with the development of China’s capital market, the “Xi Zu” capital lineage families have emerged and exerted an increasingly important influence. In “Xi Zu” groups, common controllers, information sharing and long-term cooperative relationships constitute a typical relational contractual governance structure. When lineage firms have REM incentives, they are more likely to mobilize “insiders” (e.g., sibling firms of the same lineage) to cooperate with the transaction with lower communication and coordination costs; at the same time, lineage sibling firms may also participate in the transaction out of relationship preservation or in response to the coordination of the controlling shareholders even if they deviate from their own optimal decision-making. Such REM behaviors under incomplete marketization reduce the direct costs of the initiator and may induce resource mismatch and value erosion of the partner through the relationship contract, which is reflected in the spillover and transfer effect of REM costs. To examine these issues, we assemble a panel of Chinese non-financial A-share firms from 2003 to 2018 and identify capital lineages using common ultimate controllers. We compare “Xi Zu” and non-“Xi Zu” firms to assess whether lineage affiliation weakens the link between a firm’s own real earnings management and its contemporaneous valuation, which would indicate lower direct costs for the initiator. We then study spillovers by asking whether real earnings management conducted by other members of the same lineage is associated with lower value, reduced investment, and weaker subsequent performance at the focal firm. Finally, we bring direct and spillover effects together to contrast the aggregate cost borne by lineage firms against that borne by stand alone firms. Throughout, we probe heterogeneity that clarifies when costs are more likely to be internalized or externalized. We examine internal governance by considering settings with high controller funds occupation and weak ownership checks versus those with stronger internal constraints. We examine external oversight by comparing firms with higher market capitalization or operating in more marketized regions to those with lower visibility or weaker market institutions. We also open the black box of the lineage by constructing a simple hierarchy to distinguish “core” members—large in assets and tightly held by the controller—from “edge” members with smaller size and weaker control intensity. Three findings anchor the study. First, lineage affiliation reduces the initiating firm’s direct value penalty from real earnings management. Relative to stand-alone firms, lineage members experience a weaker negative association between their own real earnings management and contemporaneous valuation. This pattern is consistent with relational coordination lowering the explicit price concessions and the communication and bargaining frictions that would otherwise be borne by the initiator in a pure market context. Second, real earnings management generates negative spillovers within lineages. When other members in the same lineage engage in earnings-increasing real activities, the focal firm’s value is lower. More importantly, this lineage-wide manipulation suppresses the focal firm’s investment in the next period and weakens its operating performance in subsequent periods. These real and financial consequences indicate that accommodating affiliates absorb part of the economic burden through distortions in their own operations. Third, after aggregating direct and spillover components at the listed-firm level, lineage firms still exhibit a lower total cost of real earnings management than non-lineage firms on average. Organizational savings achieved through lower frictions and concessions for initiators outweigh, in the aggregate, the harm inflicted on affiliates. This does not imply that lineages are socially efficient; it means that the burden shifts: initiators internalize less, while cooperating affiliates bear more. Opening the lineage structure provides an additional layer of insight. Using a simple, transparent measure based on assets and controller ownership to characterize a member’s importance and control intensity, we separate core and edge firms within each lineage-year. Two patterns emerge. Among core members, the direct value penalty of their own real earnings management is smaller, reflecting privileged access to coordination that lowers initiator-side costs. At the same time, core members are comparatively insulated from lineage-wide manipulation; the negative association between other members’ actions and their own outcomes is weak or absent. Among edge members, the picture reverses. The direct penalty of their own manipulation is larger, and lineage-wide manipulation has a pronounced negative association with their value, investment, and subsequent profitability. In other words, core members lighten their own burden, while edge members bear the brunt of the externalized cost. This asymmetric burden-sharing is a hallmark of relational governance inside hierarchically organized networks and would be obscured if one treated all lineage members as homogeneous. The analysis further reveals how internal governance, external oversight, and intra-lineage hierarchy shape cost incidence. On internal governance, where controller funds occupation is high and ownership checks are weak, spillovers are stronger, consistent with a greater ability of the controller to direct burdens toward other members. Where funds occupation is low and ownership checks are stronger, initiator-side cost reductions remain, but spillovers are muted. On external oversight, firms with larger market capitalizations and those located in more marketized regions display stronger buffering—initiators still benefit from lower frictions, yet negative spillovers to affiliates are contained. The mirror image appears in low-visibility settings: when external monitoring is weak, costs more easily flow from members operating under stronger constraints to those in weaker environments within the same lineage. This mapping from institutional strength to institutional weakness constitutes a concrete pathway by which real earnings management costs travel inside relational networks. Several steps are taken to strengthen the credibility of these inferences. To further mitigate concerns that lineage-wide manipulation and firm outcomes may be jointly driven by unobserved shocks, the analysis complements fixed-effects estimations with instrumental-variable regressions. Specifically, we use the previous year’s average media attention and short-selling activity of other firms within the same lineage—excluding the focal firm—as instruments that shift the monitoring environment and external disciplinary pressure faced by sibling firms. The focal firm’s own media attention and short-selling exposure are directly controlled for in both stages, and all instruments enter the first stage with one-year lags to alleviate simultaneity. Conventional identification diagnostics confirm instrument strength, and the estimated intra-lineage transfer and spillover patterns of REM costs remain robust under these alternative identification designs. The study’s contribution is threefold. First, it extends the analysis of real earnings management costs from a firm-centered perspective to a network perspective. In relationally governed environments, the cost of real earnings management does not rest solely with the initiator; it is spread across affiliated firms, with negative consequences for value, investment, and subsequent performance at the receiving end. Recognizing this externality changes how researchers and practitioners should think about incidence and welfare. Second, it opens the black box of the lineage by documenting a stark hierarchy of burden-sharing. Core members, large and tightly held, are positioned to secure the benefits of coordination while escaping the fallout. Edge members, smaller and more weakly held, absorb the spillovers, especially where internal governance is poor and external oversight is weak. This pattern has direct implications for investors’ portfolio risk assessment and for regulators’ surveillance of networked behavior. Third, it clarifies the boundary conditions under which costs are internalized or externalized. Stronger internal governance and stronger external monitoring induce more internalization by initiators and less damage to affiliates; weaker governance and weaker monitoring do the opposite, enabling costs to flow from nodes with stronger protections to nodes with weaker ones. This conditional view reconciles seemingly conflicting narratives about business groups in emerging markets: relational coordination can both economize on frictions and facilitate opportunistic redistribution, depending on the surrounding institutions. For investors and analysts, the results suggest a network-aware approach to evaluating the risk of real earnings management in China. The exposure of a given holding depends not only on the firm’s own behavior but also on the behavior of other firms connected through common control, especially when the holding is an edge member or operates in a low-visibility environment. For information intermediaries and regulators, the findings recommend surveillance that looks across organizational boundaries: unusual capacity shuffling, synchronized revenue timing, or non-market pricing among related entities can indicate cost transfers rather than genuine efficiency. For controllers and boards, the evidence cautions against strategies that repeatedly offload distortions onto peripheral units. Although such coordination can make real earnings management “cheap” from the initiator’s perspective and, on average, at the listed-company level, the persistent suppression of investment and erosion of future performance at affiliate firms threaten the longer-run health of the network. In sum, relational-contract governance in China alters the incidence and visibility of real earnings management costs. Lineage affiliation lowers the direct cost for initiators by leveraging long-term ties, reputation, and controller coordination, but the burden does not vanish; it is reallocated within the group, with edge members disproportionately bearing the spillovers. Governance and monitoring conditions determine whether lineages act as buffers that absorb frictions or as channels that externalize costs. By embedding China’s capital lineage structure into the analysis of real earnings management, the study provides new evidence oncost spillovers and transfers under relational governance, enriches the framework for understanding earnings management in transition economies, and offers practical guidance for investors, intermediaries, and regulators seeking to identify and restrain opportunistic behavior in complex corporate networks.