Sustainable Development of Chinese Family Firms: A Perspective from Downward Earnings Management before Successions

Abstract

Most initial entrepreneurs of Chinese family firms are approaching retirement, and intra-family successions have occurred frequently in recent years. However, both founders and the public have concerns about successors’ ability of assuming the responsibility of long-term sustainable development of their family business. This paper aims to explore whether and how founders manage earnings before within-family successions for the purpose of smooth successions. Based on a sample of Chinese family firms from 2004 to 2018 and using the ordinary least square (OLS) regression method, this study finds that Chinese family firm founders tend to manage earnings downward in the year before they transfer business to their heirs. Furthermore, founders with poverty experiences and political connections have more incentives to manipulate earnings downward. Firms’ institutional investors and the legal environment inhibit founders’ earnings management behavior before within-family successions, but independent directors do not have any significant impact on the association between successions and earnings management. Finally, it is documented that the downward earnings management leads to more favorable market reactions in the short term but do harm to firms’ long-term sustainable development. The results are robust to the Heckman two-stage analysis and using the alternative measures of earnings management. This study contributes to research on family firm succession and earnings management, and is also informative to policy makers, market participants, and family firm founders.

1. Introduction

The objective of this study is to examine the sustainable development of Chinese family firms from the perspective of earnings management before successions, that is, whether founders manage earnings downward before within-family successions in the hope of helping their children successfully inherit family business. It also investigates whether the downward earnings manipulation by founders is beneficial for firms’ short-term market reactions and the long-term sustainable development of family business.

Some researchers study family firms’ accounting choices from the socioemotional view, and they find family firms are less likely to manage earnings in order to preserve their socioemotional wealth by projecting a positive family image and maintaining a good reputation [1,2]. However, Ansari et al. [3] argue there is a breaking point where the incentives of family firm founders change significantly, and founder CEO reappointment represents such a point where they have incentives to manage earnings upward to maintain their positions in companies. This paper examines such a breaking point where Chinese family firm founders transfer business to their descendants and the incentives of earnings management exceed the costs.

The establishment of private enterprises has been allowed by the Chinese government from 1980s, so most initial entrepreneurs are approaching retirement. As a result, intra-family management successions have frequently occurred in recent years. To guarantee families’ control over the business, most founders choose their children or other blood-related descendants as top managers after they retire. Due to the prevalent administrative approval in market operations, weak investor protection regime, and relationship culture in China, some specialized assets, such as trust from other family members and connections with other stakeholders, is critical to the profitability of family firms [4,5,6]. However, it is difficult for founders to totally transfer these specialized assets to their heirs [6]. Moreover, the one-child policy imposes human capital constraints for within-family successions [7]. Taken together, lack of specialized assets and capabilities present difficulties for successors when inheriting their family business.

Altruism, which is defined as a moral value motivating individuals to take actions that benefit others without any expecting reward, is prevalent in family firms [8,9]. Confucian familism is deeply rooted in Chinese people’s moral education. Hence, family members, especially parents and children, are more closely connected relative to those in Western countries. Moreover, due to the one-child policy effectively implemented in China since 1979, most family firm founders have only one child, who is the center of the whole household but is probably less competitive in work [10,11]. Therefore, it is expected that in order to help their heirs establish a reputation for competency, Chinese family firm founders sacrifice their own reputation and firms’ financial performance by managing earnings downward before intra-family successions in the hope of improving future earnings when their descendants take family business.

This paper examines three research questions. The first is the downward earnings management behavior of founders before intra-family successions. The second is the moderating effects of founders’ characteristics, firms’ corporate governance, and the legal environment on the relationship between within-family successions and earnings management. The last is the impact of the downward earnings management on the short-term market reaction during successions and firms’ long-term financial performance.

From the empirical analysis, it is found that Chinese family firm founders tend to manage earnings downward in the year before they transfer business to their heirs. The result is more pronounced for firms with founders having poverty experiences and political connections. Firms’ institutional investors and the legal environment inhibit founders’ earnings management behavior, but independent directors do not have any significant impact on the association between successions and earnings management. Finally, it is documented that the market reaction is more favorable for firms with larger downward earnings management, but firms’ long-term financial performance is negatively affected by the downward earnings manipulation.

This study is informative to the government, who has been encouraging private entrepreneurs to assume social responsibility by cultivating competent successors for the sustainable development of socialist economy. The results of downward earnings management suggest more supervision of accounting behavior is needed around management successions. The results of this study should also be useful to market participants by helping them be aware of the potential earnings manipulation around family firm successions and make correct investment decisions. It is also meaningful to private entrepreneurs, as the negative impact of earnings management on firms’ long-term financial performance suggests founders should adopt more appropriate methods to help their descendants smoothly inherit family business, instead of managing earnings downward.

This study contributes to two research streams. First, it enriches the literature of family firm succession, most of which focuses on the consequences of management successions. However, to the best of our knowledge, this is the first study on the antecedents of management successions, that is, whether family firm founders manage earnings downward to help their descendants successfully take on family business. Consistent with the expectation, it finds that founders tend to manage earnings before within-family successions.

Second, the study adds to research on earnings management around CEO turnover, most of which independently examines earnings management behavior of the outgoing or incoming CEOs. This paper links the outgoing management with the incoming ones, and for the first time, it suggests it is necessary to consider the relationship between the outgoing management and successors when studying earnings management around management turnover. Consistent with the hypothesis, it finds that relative to family firms with outside successors, earnings manipulation is more likely to happen in enterprises with internal successions.

The rest of the paper is organized as follows. The second section reviews prior literature and develops several hypotheses. The third section describes research design. The fourth section presents empirical results, and the last section sets forth the conclusion, implications, and limitations.

2. Literature Review and Hypothesis Development

2.1. Family Firm and Earnings Management

Research on family firms is mainly based on agency theory and steward theory. According to agency theory, in family firms, the traditional owner–manager agency conflict is mitigated due to the close relationship between management and family controllers [12]. However, a new conflict, the one between controlling family and minority shareholders, arises in family firms. Controlling shareholders may sacrifice the interests of minority shareholders to maximize their wealth [13,14]. In contrast to agency theory, steward theory posits family managers act as stewards rather than agents. According to steward theory, family managers work with altruism for the benefits of the entire organization and its stakeholders as they have a deep emotional investment in companies [15,16]. In terms of earnings management of family firms, Wang [17] provides two competing views, that is, the entrenchment effect and alignment effect. Consistent with the agency-theoretic view, the entrenchment effect posits concentrated ownership creates incentives for controlling shareholders to opportunistically manage earnings for their private interests. However, the alignment effect, stemmed from steward theory, suggests family firms are less likely to behave opportunistically when reporting accounting earnings due to the alignment of the interests of controlling and minority shareholders. Additionally, the socioemotional wealth view, which supplements the steward theory, argues noneconomic factors play a vital role in family firms [18], so they are less likely to manage earnings in order to preserve socioemotional wealth by projecting a positive family image and maintaining a good reputation [1,2].

Many researchers examine earning management behavior of family firms in developed countries and find family firms are less likely to manage earnings, supporting the steward theory and socioemotional wealth view [3,12,17]. However, Fan and Wong [19] argue that research findings in developed countries, such as the US, are not readily applicable to other countries, and the institutional environment should be considered when studying one country’s corporate reporting. Hence, Fan and Wong find in east Asia, where the ownership of listed companies is concentrated in the hands of large shareholders, controlling owners are subject to less governance and are perceived to have strong incentives to withhold or manipulate earnings information for private benefits. Therefore, they find that in east Asian economies, the informativeness of reported earnings is weakened by the high ownership concentration and the large separation of ownership and control.

The legal protection of investors is poor in China, and so it is difficult for investors to win lawsuits once companies violate information disclosure laws or regulations. Additionally, the information disclosure environment in China is not very transparent, and large shareholders tend to obtain information through private channels. Hence, Chinese family founders have opportunities to manipulate earnings.

There are difficulties for successors to inherit family business successfully in China. First, because of the prevalent administrative approval in Chinese market, poor investor protection regime, and the relationship culture in China, the association with stakeholders is a specialized asset that plays a vital role in the development of family firms. The specialized asset is usually highly personalized and nontransferable [6]. Moreover, successors’ relationship management skills are influenced by their emotional intelligence [20]. Therefore, it is impossible for successors to totally inherit and capitalize the specialized assets even when they have inherited the controlling positions in firms, leading to difficulties of the sustainable development of family firms. Second, due to the one-child policy successfully implemented in China from 1979, most Chinese family firm founders have only one child. Hence, talented heir availability for family firm successions is constrained by the policy [7,21]. Therefore, the skepticism of successors’ competence is prevalent among stakeholders, resulting in the decline of firms’ stock prices around within-family successions.

Researchers (Eshel et al., 1998; Holtz-Eakin 1993) [8,9] argue that altruism exists in family firms, that is, family members may take actions that benefit others without any expected rewards. Moreover, Confucian familism, which is characterized by supremacy, unity, sacrifice, emotional attachment, and trust in families, is deeply rooted in Chinese education [22]. According to familism, each individual must make contributions to families, obey family rules, and meanwhile individuals enjoy protection from their families [23]. Hence, influenced by familism, the relationship among Chinese family members is closer relative to other countries. Additionally, the one-child policy results in overprotection and excessive indulgence of the only child from their parents [24]. Therefore, family firm founders may sacrifice the interests of themselves and other outsiders around management successions in the hope of maintaining family control over the business and protecting their descendants. Taken together, the difficulties of inheriting family business successfully for descendants and the close relationship between Chinese parents and children provide incentives for founders’ earnings management behavior before within-family successions.

Therefore, it is expected that Chinese family firm founders tend to manage earnings downward before they decide to pass business to their descendants. The following hypothesis is proposed:

Hypothesis 1:

Family firm founders tend to manage earnings downward in the year prior to within-family successions.

2.2. Moderating Factors

2.2.1. Founders’ Poverty Experience

Psychological researchers find people’s characteristics and behavior are influenced by their childhood experience [25,26]. Applying psychological research results to economic studies, Graham and Narasimhan [27] report that managers with depression experience become more conservative by using relatively little debt. Similarly, Bernile et al. [28] find CEOs who experienced extreme fatal disasters and witnessed the downside become more sensitive to firm risks and behave more cautiously. The first generation of Chinese family firms was born around 1950 and has been raised in tumultuous years. The hardship and famine experiences make Chinese family founders more conservative and cautious when managing business. Therefore, it is conjectured that family firm founders who experienced poverty in childhood are more concerned with the litigation and reputation risks arising from earnings management. Hence, the following hypothesis is stated:

Hypothesis 2a:

The association between within-family successions and earnings management is less pronounced for firms with founders having poverty experience in childhood.

2.2.2. Founders’ Political Connection

Chinese researchers argue political connections bring Chinese firm CEOs privileges, and they find politically connected CEOs are less likely to be fired when firm performance is poor [29,30]. Consistently, Xing et al. [31] suggest politically connected board secretaries are less likely to disclose high-quality earnings forecasts. In the transition economy with a weak legal system to protect family firms, political connections provide private entrepreneurs security, and so it is expected that they are more likely to oversee accounting regulations when pursing their private benefits. Hence, the following hypothesis is proposed:

Hypothesis 2b:

The association between within-family successions and earnings management is more pronounced for firms with politically connected founders.

2.2.3. Independent Director

Outside directors have incentives to effectively monitor management to maintain their reputational capitals [32,33]. Roe (1991) [34] argues that effective monitoring of corporate boards prevents the abuse of power by mangers and thus protect the interests of shareholders. Anderson and Reeb (2004) [15] further find in family firms where the principle–principle conflict is more significant relative to the owner–manager conflict, independent directors can protect the interests of minority shareholders against the exploitation of controlling shareholders.

With respect to the impact of outside directors on earnings management, researchers [35,36,37] find companies with a larger proportion of outside directors are less likely to engage in earnings management. Bammens et al. [38] show although outside directors in family firms are mostly included for their advice role, they are also valuable in constraining earnings management. Moreover, Chi et al. [13] document that Taiwanese family firms are more likely to manage earnings, but independent directors effectively suppress the association between the degree of family control and earnings management. Therefore, it is expected that independent directors in Chinese family firms would negatively moderate the association between within-family successions and earnings management. Hence, the following hypothesis is stated:

Hypothesis 2c:

The association between within-family successions and earnings management is less pronounced for firms with a larger proportion of independent directors.

2.2.4. Institutional Investor

Institutional investors, whose equity investment in firms is usually high, have resources, abilities, and incentives to monitor and influence management decisions [39,40]. Furthermore, Bushee [41] suggests that institutional investors are sophisticated investors searching for long-term firm success rather than short-term profits. Consistently, researchers find companies with a larger proportion of institutional shareholdings are less likely to manage earnings towards managers’ desired level of profits [41,42]. Hence, it is expected that institutional investors would reduce management incentives for earnings manipulation before within-family successions. Therefore, the following hypothesis is proposed:

Hypothesis 2d:

The association between within-family successions and earnings management is less pronounced for firms with a larger percentage of institutional ownership.

2.2.5. Legal Environment

Prior studies document that firms’ information disclosure is influenced by legal regime [43,44]. Lower earnings quality is less likely to occur in countries with a stricter legal environment, as there are greater consequences for management in terms of civil and criminal liability and other punishment and sanctions imposed by regulatory agencies [45]. Consistently, researchers document that financial disclosure is more transparent in countries with stronger investor protection regime [46,47], and in these countries, earnings are less managed and more value-relevant [45,48]. Moreover, Du et al. [49] document that the positive association between writer–auditor relationship and pre-IPO earnings management is more pronounced in Chinese listed firms located in provinces with weak legal environment. Hence, it is conjectured that the association between family firm successions and earnings management would vary in provinces with different legal environment. This leads to the following hypothesis:

Hypothesis 2e:

The association between within-family successions and earnings management is less pronounced for firms located in provinces with stricter legal environment.

2.3. The Impact of Downward Earnings Management

Many researchers study the impact of within-family successions on family firm performance, and their findings are inconclusive [50,51,52,53,54,55]. The aim of downward earnings management for founders is to boost earnings when heirs take on family business, exhibiting successors’ competence and establishing their reputations. Therefore, it is expected that around successions, the market would react more favorably to succession firms with larger prior downward earnings management. However, parents’ altruism in the form of downward earnings manipulation may spoil their children, leading to less responsible and hard-working descendants, which is harmful for the long-term sustainable development of family business [56]. Furthermore, it is difficult to detect earnings management in the short term, but in the long run, firm value is negatively affected by earnings management through deteriorations of firm performance and stock price crashes [57]. Therefore, it is expected that the downward earnings management would do harm to family firms’ long-term performance. Hence, the following hypotheses are proposed:

Hypothesis 3a:

The market reaction around within-family successions is more favorable for family firms with larger downward earnings management before successions.

Hypothesis 3b:

The long-term financial performance is worse for family firms with larger downward earnings management prior to successions.

Based on the above analysis, the following theoretical framework is proposed and presented in Figure 1.

3. Data and Research Design

3.1. Data

The Chinese listed family firm data were collected from China Stock Market and Accounting Research (CSMAR) from 2004 to 2018. In this study, succession was identified by tracking turnovers of chairmen, who are typically the key decision makers in Asian family firms. A succession takes place when a family member (usually the son, daughter, son-in-law, daughter-in-law, niece, or nephew of the founder) or an outside professional is appointed to the position of chairman.

Public disclosures, including company prospectuses and annual reports were used to identify successions. The relationship between founders and successors was obtained through a special research database on family firms in CSMAR. If the relationship could not be clearly determined, the information was looked up in Baidu, China’s most popular research engine. Other financial data and corporate governance information were obtained from CSMAR databases with respect to financial information quality, management characteristics, institutional investors, stock market, etc.

Excluding firms with incomplete data, the final sample consisted of 189 family firms that continued to be managed by family successors and 335 family firms in which the management was passed onto unrelated professionals.

3.2. Research Design

3.2.1. Measures of Earnings Management

Two proxies—accrual and real earnings management—were used to capture earnings manipulation. Specifically, the following models were estimated cross-sectionally by each industry-year with at least 10 observations. The industry was identified using the China Securities Regulatory Commission (CSRC) 2012 industrial classification code. The modified Jones’s model was used to calculate discretionary accruals.

$$\frac{T{A}_{i,t}}{A_{i,t-1}}={\alpha }_1\frac{1}{A_{i,t-1}}+{\alpha }_1\frac{\Delta RE{V}_{i,t}}{A_{i,t-1}}+{\alpha }_2\left(\frac{PP{E}_{i,t}}{A_{i,t-1}}\right)+{\epsilon }_{i,t}$$

(1)

where TA_i,t is total accruals in year t, calculated as operating income minus operating cash flow. A_i,t−1 presents total assets in year t − 1, ΔREV_i,t is the change in sales revenue in year t, and PPE_i,t is the net value of fixed assets in year t.

The coefficient estimates from Equation (1) were then used to estimate the normal accruals for sample firms:

$$ND{A}_{i,t}={\alpha }_0\frac{1}{A_{i,t-1}}+{\alpha }_1\frac{\Delta RE{V}_{i,t}-\Delta RE{C}_{i,t}}{A_{i,t-1}}+{\alpha }_2\left(\frac{PP{E}_{i,t}}{A_{i,t-1}}\right)$$

(2)

where ΔREC_i,t is the change in receivables in year t, and NDA_i,t is normal accruals in year t.

$$AE{M}_{i,t}=\frac{T{A}_{i,t}}{A_{i,t-1}}-ND{A}_{i,t}$$

(3)

where AEM_i,t is the proxy of discretionary accruals. This study focuses on downward earnings management. Therefore, following prior research [58,59], accrual earnings management is measured by the signed value of discretionary accruals (AEM) and a variable consists of the absolute value of the negative abnormal accruals and 0 for positive abnormal accruals (ABSAEM).

Following Roychowdhury [60], the abnormal cash flow from operations (ACFO), abnormal production costs (APROD), and abnormal discretionary expenses (ADISEXP) were calculated using the following models:

$$\frac{CF{O}_{i,t}}{A_{i,t-1}}={\beta }_0+{\beta }_1\frac{1}{A_{i,t-1}}+{\beta }_2\frac{RE{V}_{i,t}}{A_{i,t-1}}+{\beta }_3\frac{\Delta RE{V}_{i,t}}{A_{i,t-1}}+{\epsilon }_{i,t}$$

(4)

where CFO_i,t is operating cash flow in year t, and REV_i,t is sales revenue in year t. The abnormal CFO (ACFO) is the residual of Equation (4).

$$\frac{ PRO{D}_{i,t}}{A_{i,t-1}}={\beta }_0+{\beta }_1\frac{1}{A_{i,t-1}}+{\beta }_2\frac{RE{V}_{i,t}}{A_{i,t-1}}+{\beta }_3\frac{\Delta RE{V}_{i,t}}{A_{i,t-1}}+{\beta }_4\frac{\Delta RE{V}_{i,t-1}}{A_{i,t-1}}+{\epsilon }_{i,t}$$

(5)

where PROD_i,t is production costs, measured as the sum of the costs of goods sold and the change in inventory in year t, and ΔREV_i,t−1 is the change in sales revenue in year t−1. The abnormal PROD (APROD) is the residual of Equation (5).

$$\frac{DISEX{P}_{i,t}}{A_{i,t-1}}={\beta }_0+{\beta }_1\frac{1}{A_{i,t-1}}+{\beta }_2\frac{RE{V}_{i,t-1}}{A_{i,t-1}}+{\epsilon }_{i,t}$$

(6)

where DISEXP_i,t is discretionary expenses, measured as the sum of advertising expenses, R&G and SG&A expenses. The abnormal DISEXP (ADISEXP) is the residual of Equation (6).

$$RE{M}_{i,t}=\left(-1\right)ACF{O}_{i,t}+APRO{D}_{i,t}+\left(-1\right)ADISEX{P}_{i,t}$$

(7)

where REM_i,t is the aggregate measure of real earnings management. Similarly, in this study, real earnings management is measured by the signed value of real earnings management (REM) and a variable consists of the negative REM and 0 for positive REM (ABSREM).

3.2.2. Models

To test Hypothesis 1, the association between within-family successions and downward earnings management was investigated. The following ordinary least square (OLS) regression model is used:

$$\begin{array}{l}E{M}_{it}={\propto }_{1}SUCCESSIO{N}_{it}+{\propto }_{2}SIZ{E}_{it}+{\propto }_{3}LE{V}_{it}+{\propto }_{4}RO{A}_{it}+{\propto }_{5}LOSS\\ +{\propto }_{6}MVB{V}_{it}+{\propto }_{7}FIRMAG{E}_{it}+{\propto }_{8}AUDITO{R}_{it}+{\propto }_{9}DUA{L}_{it}\\ +{\alpha }_{10}FS{R}_{it}\end{array}$$

(8)

where EM refers to variables which proxy for accrual earnings management (AEM and ABSAEM) and real earnings management (REM and ABSREM). The variable of interest is SUCCESSION, which equals one if family business is inherited by family members, and 0 if the control is transferred to outside professionals. A vector of control variables, including firm size (SIZE), leverage (LEV), financial performance (ROA), firm loss (LOSS), the market-to-book-ratio (MVBV), firm age (FIRMAGE), auditor (AUDITOR), duality (DUAL), and shares of the largest shareholder (FSR) were included in the model. Industry and year dummies were included to control for any industrial and temporal fixed effects. The detailed definitions of variables are outlined in

Table 1. Definitions of variables.

Variable	Definition
AEM	Accrual earnings management, measured as the signed value of discretionary accruals
ABSAEM	Accrual earnings management, measured by the variable consisting of the absolute value of the negative abnormal discretionary accruals and zero for positive abnormal discretionary accruals
REM	Real earnings management, measured by the sum of the negative value of the abnormal cash flow from operations, the abnormal production costs and the negative value of abnormal discretionary expenses
ABSREM	Real earnings management, measured by the variable consisting of the absolute value of the negative real earnings management and zero for positive REM
CAR	Firms’ cumulative market-adjusted return for the three days centered on the day of the official announcement of chairman appointment
AVEROA	The average value of firms’ ROA from the 3rd to 5th year after family firm successions
SUCCESSION	A dummy variable equaling one if family business is inherited by family successors, and zero if the firm is succeeded by outside professionals
POVERTY	A dummy variable equaling one if the founder has poverty experiences in childhood and zero otherwise. It is deemed that the founder has poverty experience if he/she experienced the Three Years of Natural Disasters in childhood (0 to 14 years old), that is, the founder was born from 1947 to 1961
PC	A dummy variable equaling one if the founder is currently holding or previously held a position in the People’s Congress, Chinese People’s Consultative Conference, has received honors from the government, or is a party member of Chinese Communist Party, and zero otherwise
INDIR	The proportion of independent directors on the board
INST	The percentage of institutional shareholdings
LEGAL	An index of market intermediaries development and institutional environment from Fan et al. (2011), and the index captures the development of market intermediaries, such as lawyers, auditors and various industry associations, the efficiency of the local courts, and the protection of property rights
SIZE	The natural logarithm of total assets
LEV	Total liabilities/total assets
ROA	Net profit/the average value of total assets at the beginning of the fiscal year and total assets at the end of the fiscal year
LOSS	A dummy variable that equals one if ROA is negative and zero otherwise
MVBV	Market value of equity/book value of equity
FIRMAGE	List years, equaling to the number of years a company has been listed
AUDITOR	A dummy variable equaling one if the company is audited by one of the Big-4 auditors and zero otherwise
DUAL	A dummy variable equaling one if the CEO and chairman are the same person and zero otherwise
FSR	The percentage of shares held by the largest shareholder, calculated as the number of shares held by the largest shareholder divided by the number of total shares
HEIRAGE	The age of family firm heirs

.

To test Hypothesis 2, the moderating effect of founders’ characteristics, corporate governance, and legal environment on the relationship between the within-family successions and downward earnings management was examined. The new variables used in Equation (9) included founders’ poverty experience (POVERTY) and political connections (PC), independent directors (INDIR), institutional investors (INST), and legal environment (LEGAL). Definitions of these variables are presented in Table 1. The following OLS regression model was performed:

$$\begin{array}{l}E{M}_{it}={\propto }_{1}SUCCESSIO{N}_{it}+{\alpha }_{2}POVERT{Y}_{it}+{\propto }_{3}SUCCESSIO{N}_{it}\ast POVERT{Y}_{it}\\ +{\alpha }_{4}P{C}_{it}+{\propto }_{5}SUCCESSION\ast P{C}_{it}+{\alpha }_{6}INDI{R}_{it}\\ +{\propto }_{7}SUCCESSIO{N}_{it}\ast INDI{R}_{it}+{\alpha }_{8}INS{T}_{it}\\ +{\propto }_{9}SUCCESSIO{N}_{it}\ast INST+{\alpha }_{10}LEGA{L}_{it}\\ +{\propto }_{11}SUCCESSIO{N}_{it}\ast LEGA{L}_{it}+{\propto }_{12}SIZ{E}_{it}+{\propto }_{13}LE{V}_{it}\\ +{\propto }_{14}RO{A}_{it}+{\propto }_{15}LOS{S}_{it}+{\propto }_{16}MVB{V}_{it}+{\propto }_{17}FIRMAG{E}_{it}\\ +{\propto }_{18}AUDITO{R}_{it}+{\propto }_{19}DUA{L}_{it}+{\propto }_{20}FS{R}_{it}+{\epsilon }_{it}\end{array}$$

(9)

In Equation (9), variables of interest were SUCCESSION * POVERTY, SUCCESSION * PC, SUCCESSION * INDIR, SUCCESSION * INST and SUCCESSSION * LEGAL. Industry and year dummies were also included to control for any industrial and temporal fixed effects.

To test Hypothesis 2a, the association between within-family successions and the downward earnings management, conditional on founders’ poverty experience, was examined. To test Hypothesis 2b, the variable SUCCESSION was interacted with PC to capture the differences in terms of earnings management between firms with politically connected founders and other firms. To test Hypothesis 2c, whether earnings management before within-family successions is smaller for firms with a larger proportion of independent directors was explored. To test Hypothesis 2d, whether the association between within-family successions and earnings management is less pronounced for firms with more institutional investment was examined. To test Hypothesis 2e, the differences concerning the impact of within-family successions on earnings management depending on the local legal environment are studied.

To test Hypothesis 3, the short-term and long-term effect of downward earnings management before family firm inheritance was examined. The new variables used in model (3) were the short-term market reaction (CAR) and the long-term financial performance (AVEROA). The following OLS regression model was used:

$$\begin{array}{l}CA{R}_{it}/AVERO{A}_{it}={\alpha }_{1}E{M}_{it}+{\alpha }_{2}SIZ{E}_{it}+{\alpha }_{3}LE{V}_{it}+{\alpha }_{4}MVB{V}_{it}+{\alpha }_{5}FIRMAG{E}_{it}+{\alpha }_{6}HEIRAG{E}_{it}+{\alpha }_{7}INDI{R}_{it}+{\alpha }_{8}INS{T}_{it}\\ +{\alpha }_{9}FS{R}_{it}+{\epsilon }_{it}\end{array}$$

(10)

where CAR is firms’ cumulative abnormal return for the three days centered on the day of the official announcement of the chairman appointment. AVEROA is the average value of firms’ ROA from the 3rd to 5th year after within-family successions. In Equation (10), the variable of interest is EM, including accrual and real earnings management. Similarly, industry and year dummies were included to control for industrial and temporal fixed effects. The detailed definitions of variables are presented in Table 1.

4. Results

4.1. Univariate Results

Panel A of

Table 2. Descriptive statistics and univariate analysis.

Panel A: Descriptive Statistics
Variables	N		Mean	S.D.	Min		Max
AEM	524		0.0387	0.1419	−0.2598		0.6239
ABSAEM	524		0.0319	0.0585	0		0.2598
REM	524		0.0311	0.2063	−0.4509		0.6264
ABSREM	524		0.0604	0.1106	0		0.4509
SUCCESSION	524		0.3607	0.502	0		1
POVERTY	524		0.55	0.501	0		1
PC	524		0.81	0.397	0		1
INDIR	524		0.3752	0.0638	0.2353		0.5
INST	524		36.6322	24.5746	0.3563		84.2238
LEGAL	524		10.1423	5.508	1.84		20.61
SIZE	524		21.609	0.9665	20.2		24.4
LEV	524		0.412	0.2015	0.0408		0.7631
ROA	524		0.0411	0.0573	−0.1035		0.1947
LOSS	524		0.12	0.322	0		1
MVBV	524		4.5117	3.2095	0.822		15.1385
FIRMAGE	524		14.76	5.583	5		26
AUDITOR	524		0.01	0.113	0		1
DUAL	524		0.17	0.375	0		1
FSR	524		30.7182	11.4895	11.83		56.76
CAR	182		0.0146	0.0387	−0.0481		0.0887
AVEROA	178		0.0453	0.0426	−0.0391		0.1249
HEIRAGE	182		35.81	5.736	27		48
Panel B: Univariate Analysis of Hypothesis 1
Group	Mean of AEM		Mean of ABSAEM		Mean of REM		Mean of ABSREM
	Mean	t-stat	Mean	t-stat	Mean	t-stat	Mean	t-stat
SUCCESSION = 1 (N = 189)	−0.0115	−2.212 ***	0.0588	2.001 ***	−0.0254	−2.298 ***	0.0917	2.391 ***
SUCCESSION = 0 (N = 335)	0.067		0.0167		0.063		0.0427
Panel C: Univariate Analysis of Hypothesis 3
Group	Mean of CAR				Mean of AVEROA
	N	Mean	t-stat		N	Mean	t-stat
Lower AEM	106	0.031	2.591 ***		80	0.0251	−2.386 ***
Higher AEM	76	−0.0083			98	0.007
Higher ABSAEM	47	0.0292	2.014 **		68	0.02	−2.328 ***
Lower ABSAEM	135	0.0095			110	0.0863
Lower REM	106	0.0212	2.348 **		80	0.028	−3.133 ***
Higher REM	76	0.0054			98	0.0665
Higher ABSREM	51	0.0199	2.442 **		67	0.03	−2.99 ***
Lower ABSREM	131	0.001			111	0.0701

Note: This table presents the descriptive statistics for the variables included in the main analyses and univariate analysis for Hypotheses 1 and 3. Panel A provides summary statistics; Panel B provides the differences in variables for family firms inherited by second generations (SUCCESSION = 1) and those transferring control to professional outsiders (SUCCESSION = 0). In Panel C, firms transferring control to family members are divided into two groups according to the average value of AEM, ABSAEM, REM, and ABSREM. The significance of the difference in means is based on two-tailed t-tests. *** and ** denote statistical significance at the 1% and 5% levels, respectively. All continuous variables are winsorized at the 1st and 99th percentiles. Detailed definitions of variables are presented in Table 1.

reports descriptive statistics for all the variables used in the analysis. The mean values of AEM, ABSAEM, REM, and ABSREM were 0.0387, 0.0319, 0.0311 and 0.0604, respectively. The descriptive statistics of other variables report some interesting results. The mean value of POVERTY was 0.55, indicating 55% of founders had poverty experiences in childhood. The mean PC was 0.81, suggesting a large proportion of Chinese family firm founders have political connections. The mean value of INDIR was 0.3752, indicating most Chinese private enterprises only maintain a minimum number of independent directors required by regulations to cater to policy makers. The mean value of AUDITOR was 0.01, suggesting most Chinese family firms are reluctant to employ Big-4 auditors.

Panel B of Table 2 presents univariate analysis of Hypothesis 1. As displayed by Panel A, family firms inherited by second generations exhibit lower values of accrual earnings management (AEM) and real earnings management (REM), and higher absolute values of negative accrual earnings management (ABSAEM) and negative real earnings management (ABSREM) relative to family firms transferring control to professional outsiders, providing preliminary support for Hypothesis 1.

Panel C of Table 2 shows univariate test of Hypothesis 3. Firms with AEM and REM lower than the average value and those with ABSAEM and ABSREM higher than the average value are classified as firms with larger downward earnings management, and others are those with smaller downward earnings management. As suggested by the results, the former firms exhibit higher short-term market prices (CAR) and worse long-term financial performance (AVEROA) relative to the latter, providing preliminary evidence to support Hypothesis 3.

4.2. Multivariate Analysis

Table 3. Within-family successions and earnings management.

	(1) AEM	(2) ABSAEM	(3) REM	(4) ABSREM
SUCCESSION	−0.0278 *** (−2.65)	0.0086 *** (3.49)	−0.0584 *** (−2.97)	0.024 *** (2.84)
SIZE	0.0349 *** (2.5)	−0.0024 ** (−2.25)	0.0354 * (2.08)	−0.0284 * (−1.84)
LEV	−0.1686 *** (−2.45)	0.055 ** (2.15)	−0.2402 ** (−2.46)	0.1266 * (1.73)
ROA	0.0222 ** (2.05)	−0.0372 ** (−2.19)	1.8016 *** (2.71)	−1.5562 *** (−4.95)
LOSS	−0.123 *** (−2.58)	0.0772 *** (2.75)	−0.0927 *** (−2.97)	0.0541 *** (2.92)
MVBV	0.008 * (1.82)	−0.0012 (−1.38)	0.0129 (1.21)	−0.0075 (−1.48)
FIRMAGE	−0.0023(−0.72)	0.0006 (0.47)	−0.004 (−0.88)	0.0014 (0.67)
AUDITOR	0.0502 ** (2.3)	−0.0453 *** (−2.66)	0.275 ** (2.17)	−0.1717 * (1.84)
DUAL	−0.0012 *** (−3.02)	0.0079 *** (2.83)	−0.0169 *** (−2.52)	0.0111 *** (2.43)
FSR	−0.0007 *** (−2.43)	0.0003 *** (3.42)	−0.0012 *** (−2.5)	0.002 *** (2.69)
Intercept	−0.8514 * (−1.78)	0.1099 ** (2.56)	−0.5257 ** (−2.18)	0.4725 * (1.84)
Year fixed effects	Yes	Yes	Yes	Yes
Industry fixed effects	Yes	Yes	Yes	Yes
Adj. R-sq	0.207	0.2713	0.2276	0.2183
N	524	524	524	524

Note: This table presents the OLS regression results examining the association between within-family successions and earnings management. Columns (1) and (3) report the results with signed values of accrual and real earnings management as the dependent variables, respectively. Column (2) and (4) report the results with dependent variables taking the absolute values of the negative AEM and REM, respectively, and zero for positive AEM and REM. All continuous variables are winsorized at the 1st and 99th percentiles. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively (two-tailed). Detailed definitions of variables are presented in Table 1. Equation (8) is estimated.

shows the results of estimating Equation (8), investigating the relationship between within-family successions and earnings management. The coefficient on SUCCESSION is positive and statistically significant at the 1% level in columns (1)–(4), indicating founders are more likely to manage earnings downward before transferring business to heirs. The results in terms of some control variables, including SIZE, LEV, ROA, LOSS, AUDITOR, DUAL, and FSR, suggest that firms that are smaller, more leveraged, with lower return on assets, reporting losses, audited by non-Big 4 auditors, with duality of chairman and CEO, and having more concentrated ownership are more likely to manage earnings downward. Overall, the findings in Table 3 support Hypothesis 1, that is, family firm founders tend to manage earnings downward prior to within-family successions.

Table 4. Within-family successions, earnings management, and moderating factors.

	(1) AEM	(2) ABSAEM	(3) REM	(4) ABSREM
SUCCESSION	−0.0469 *** (−3.16)	0.0917 *** (2.73)	−0.6629 *** (−2.8)	0.2118 ** (2.16)
POVERTY	−0.0259 (−0.46)	0.0033 (1.14)	−0.0129 (−1.18)	0.0216 (0.62)
SUCCESSION * POVERTY	−0.0289 ** (−2.31)	0.0302 ** (2.07)	−0.1279 *** (−3.1)	0.0739 ** (2.28)
PC	0.1456 * (1.9)	−0.0322 * (−1.99)	0.0606 (1.63)	−0.0246 * (−1.72)
SUCCESSION * PC	−0.002 ** (−2.02)	0.001 ** (2.02)	−0.0821 *** (−2.57)	0.0814 ** (2.13)
INDIR	−0.157 (−0.35)	0.0538 (1.28)	−0.0541 (−1.1)	0.0334 (1.12)
SUCCESSION * INDIR	0.1871 (1.29)	−0.2084 (−0.75)	2.3108 (0.83)	−0.9053 (−1.24)
INST	0.0016 (1.03)	−0.0003 * (−1.73)	0.0029 * (1.74)	−0.0001 * (−1.91)
SUCCESSION * INST	0.0005 *** (3.32)	−0.0004 ** (−2.5)	0.0008 ** (2.39)	−0.0007 *** (−2.62)
LEGAL	−0.0025 (−1.36)	0.0024 ** (2.15)	−0.0146 * (−1.68)	0.0004 (1.1)
SUCCESSION * LEGAL	0.0001 ** (2.13)	−0.001 ** (−2.29)	0.0129 *** (2.82)	−0.0019 *** (−2.39)
SIZE	0.0571 * (1.9)	−0.003 ** (−2.23)	0.0429 ** (2.14)	−0.0107 * (−1.75)
LEV	−0.246 * (−1.89)	0.0548 * (1.99)	−0.3252 ** (−1.99)	0.113 * (1.85)
ROA	0.0994 ** (2.18)	−0.0826 * (−1.66)	1.6707 ** (2.44)	−1.52 *** (−2.49)
LOSS	−0.045 *** (−2.53)	0.0951 *** (2.64)	−0.0037 *** (−2.53)	0.0746 *** (2.82)
MVBV	0.0096 (1.14)	−0.0007 (−1.21)	0.0127 (1.21)	−0.0071 (−1.36)
FIRMAGE	−0.0028 (−0.74)	0.0016 (0.97)	−0.004 (−0.85)	0.0023 (0.98)
AUDITOR	0.058 * (1.87)	−0.064 * (−1.84)	0.323 * (1.74)	−0.1315 * (−1.81)
DUAL	−0.0366 *** (−2.65)	0.0086 *** (2.63)	−0.0263 *** (−2.37)	−0.0263 *** (−2.75)
FSR	−0.0033 *** (−2.51)	0.0002 *** (2.52)	−0.0055 *** (−3.03)	−0.0027 *** (−3.05)
Intercept	−1.1999 * (−1.89)	−0.0291 ** (−2.11)	−0.698 * (−1.88)	0.1029 ** (2.26)
Year fixed effects	Yes	Yes	Yes	Yes
Industry fixed effects	Yes	Yes	Yes	Yes
Adj. R-sq	0.2509	0.2091	0.2967	0.2994
N	524	524	524	524

Note: This table presents the OLS regression results examining the impact of within-family successions on founders’ earnings management behavior, conditional on founders’ characteristics, corporate governance and legal environment. Columns (1) and (3) report the results with signed values of accrual and real earnings management as the dependent variables, respectively. Columns (2) and (4) report the results with dependent variable taking the absolute values of the negative AEM and REM, respectively, and zero for positive AEM and REM. All continuous variables are winsorized at the 1st and 99th percentiles. ***, **, and * denote significance at the 1%, 5%, and 10% level, respectively, in two-tailed tests. Detailed variables definitions are presented in Table 1. Equation (9) is estimated.

presents the multivariate tests of the association between within-family successions and earnings management conditional on several moderating factors, including founders’ poverty experience and political connections, independent directors, institutional investors, and local legal environment. As indicated in Table 4, the coefficients on the interaction variable SUCCESSION * POVERTY in columns (1) and (3) were significantly negative at the 5% and 1% levels, respectively, while it was significantly positive at the 5% level in columns (2) and (4), suggesting founders with poverty experiences have more incentives to manage earnings downward before transferring control to their descendants. The results are inconsistent with Hypothesis 2a, and this is probably because founders with poverty experiences are deeply impressed by the hardship during their childhood, and they strongly do not want their children to face such hardship in the future. As a result, contrary to their usual conservative behavior, they are more likely to manage earnings downward before the inheritance in the hope of creating a better financial environment for their second generations.

The coefficient on the second interaction variable, SUCCESSION *PC, was negative and significant at the 5% and 1% levels in columns (1) and (3), and positive and significant at the 5% level in columns (2) and (4), indicating political connections, which provide private entrepreneurs protection and security, incentivize founders to manage earnings downward before transferring business to their heirs.

With regard to corporate governance, inconsistent with Hypothesis 2c, the coefficient of SUCCESSION *INDIR was insignificant in all the equations, suggesting independent directors do not inhibit founders from manipulating earnings before successions. This is probably because independent directors are chosen by controlling shareholders, so they are actually not independent enough and have little power in terms of supervision. Moreover, the information disclosure of Chinese private firms is not very transparent, and most independent directors do not have channels to obtain firms’ information that plays a vital role in supervising management. The fourth interaction variable, SUCCESSION * INST, is significantly and positively correlated with dependent variables at the 1% and 5% levels in columns (1) and (3), respectively, while it is significantly and negatively correlated with dependent variables at the 5% and 1% levels in columns (2) and (4), respectively, indicating institutional investors effectively prevent founders from artificially reporting poor performance before within-family successions.

The coefficient of the last interaction variable, SUCCESSION * LEGAL, was negative and significant at the 5% and 1% levels in columns (1) and (3), respectively, and was positive and significant at the 5% and 1% levels in columns (2) and (4), respectively, suggesting that in provinces with stricter legal regime, founders are less likely to manage earnings before transferring business to second generations. Overall, the findings indicate that founders’ background, corporate governance, and local legal environment influence founders’ decisions to manage earnings downward in the year prior to the transition of control from founders to second generations.

Table 5. Short-term and long-term effects of earnings management.

	CAR				AVEROA
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
AEM	−0.8589 *** (−2.65)				0.4798 *** (3.96)
ABSAEM		0.5188 ** (2.25)				−0.7091 *** (−3.41)
REM			−0.0157 ** (−2.38)				1.0089 *** (2.79)
ABSREM				0.0237 ** (2.43)				−0.1516 ** (−2.25)
SIZE	0.0238 ** (2.18)	0.0222 **(2.14)	0.0232 ** (2.12)	0.0232 ** (2.13)	0.0321 *** (3.8)	0.0336 *** (3.62)	0.0188 * (1.82)	0.0135 ** (2.31)
LEV	−0.0248 (−1.4)	−0.016 (−1.27)	−0.0432 (−0.64)	−0.0403 (−0.63)	−0.2279 *** (−4.04)	−0.2467 *** (−4.04)	−0.1813 ** (−2.44)	−0.1778 ** (−2.52)
MVBV	0.0008 * (1.97)	0.0012 ** (2.26)	0.0009 ** (2.18)	0.001 ** (2.22)	0.0034 ** (2.51)	0.0047 *** (2.66)	0.0002 ** (2.03)	−0.0044 ** (−2.53)
FIRMAGE	0.0008 * (1.79)	0.0011 ** (2.59)	0.0005 ** (2.26)	0.0004 ** (2.23)	0.0016 ** (2.28)	0.0016 * (2.16)	0.0012 * (1.76)	0.0016 * (2.04)
HEIRAGE	0.0034 ** (2.21)	0.0032 ** (2.29)	0.0031 * (2.08)	0.0031 * (2.05)	0.0024 ** (2.33)	0.002 * (1.9)	0.0015 * (1.71)	0.0013 ** (2.15)
INDIR	−0.1317 (−1.1)	−0.1234 (−1.07)	−0.1395 (−1.15)	−0.1418 (−1.16)	0.052 (0.38)	0.0111 (1.08)	−0.0657 (−0.36)	−0.0787 (−0.45)
INST	0.0005 ** (2.11)	0.0005 ** (2.12)	0.0005 ** (2.22)	0.0005 ** (2.23)	0.0007 * (1.84)	0.0007 * (1.83)	0.0002 ** (2.32)	0.0002 ** (2.38)
FSR	−0.0017 * (−1.78)	−0.0019 * (−2.09)	−0.0016 (−1.61)	−0.0015 (−1.58)	0.0002 *** (2.86)	0.0002 *** (2.83)	0.0005 *** (2.95)	0.0005 *** (2.52)
Intercept	0.5659 ** (2.56)	0.5813 ** (2.73)	0.5654 ** (2.53)	0.5691 ** (2.55)	−0.6886 *** (−3.28)	−0.6267 *** (−2.85)	−0.3072 ** (−2.24)	−0.1896 * (−1.77)
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Adj. R-sq	0.1376	0.2055	0.1195	0.1223	0.2705	0.2308	0.1809	0.2525
N	182	182	182	182	178	178	178	178

Note: Table 5 presents OLS regression model relate the short-term and long-terms effects of earnings management before family firm successions. The sample includes firms transferring control to family members. Columns (1) to (4) report the results with CAR as the dependent variable. Column (5) to (8) report the results with AVEROA as the dependent variable. All continuous variables are winsorized at the 1st and 99th percentiles. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively (two-tailed tests). Detailed variables definitions are presented in Table 1. Equation (10) is estimated.

shows the multivariate regression results of the short-term and long-term effects of earnings management before within-family successions. Columns (1) to (4) report the results regarding the impact of earnings management on short-term market reactions (CAR). As indicated by the results, the coefficients associated with AEM and REM were negative and significant at the 1% and 5% levels in columns (1) and (3), respectively. The coefficients of ABSAEM and ABSREM were positive and significant at the 5% level in columns (2) and (4). Regarding control variables, the market reaction was higher for firms that are larger (SIZE) and older (FIRMAGE), with higher growth opportunities (MVBV), older successors (HEIRAGE), more institutional investment (INST), and less concentrated ownership (FSR). Overall, the results support Hypothesis 3a, that is, the market reaction around within-family successions is more favorable for family firms with larger downward earnings management prior to successions.

Columns (5) to (8) present the results of the impact of downward earnings management on firms’ long-term financial performance. The coefficients of AEM and REM were positive and significant at the 1% level in columns (5) and (7), and the coefficients of ABSAEM and ABSREM were negative and significant at the 1% and 5% in columns (6) and (8), respectively. The results suggest the long-term development of family firms is harmed by the downward earnings management before within-family successions, supporting Hypothesis 3b. The results of control variables suggest the financial performance is better for firms that are larger and older, lower leveraged, with higher growth opportunities, older successors, more institutional investment, and more concentrated ownership.

4.3. Additional Analysis

Researchers suggest incoming CEOs have incentives to minimize reported earnings in the initial year of their tenure [61,62,63]. As incoming CEOs are not responsible for firms’ past performance, they tend to manipulate earnings downward in the first financial year of their tenure, thereby lowering expectations of future performance, improving future reported earnings via the reversing nature of accruals, increasing their compensation, and establishing reputations in the labor market. Hence, this study further analyses whether outside professionals are more likely to manage earnings downward in the first year of their appointment relative to second-generation successors.

Table 6. Discretionary accruals and chairman appointment.

	(1) AEM	(2) ABSAEM
SUCCESSION	0.0068 * (1.71)	−0.0008 * (−1.82)
SIZE	0.0136 *** (2.83)	−0.0085 ** (−2.15)
LEV	−0.0233 ** (−2.26)	0.0015 ** (2.04)
ROA	0.6774 ** (2.49)	−0.3713 *** (−3.00)
LOSS	−0.0782 *** (−2.95)	0.0191 *** (2.85)
MVBV	0.0056 (1.23)	−0.0027 (−1.31)
FIRMAGE	0.0001 (1.04)	−0.0001 (−1.08)
AUDITOR	0.0207 ** (2.26)	−0.0234 * (−1.65)
DUAL	−0.0198 *** (−2.74)	0.0176 *** (2.95)
FSR	−0.0013 *** (−3.08)	0.0006 *** (3.17)
STAY	0.0081 (0.29)	−0.0086 (−0.67)
Intercept	0.2754 ** (2.18)	−0.1441 (−0.9)
Year fixed effects	Yes	Yes
Industry fixed effects	Yes	Yes
Adj. R-sq	0.1491	0.1873
N	558	558

Note: This table presents the OLS regression results examining the association between discretionary accruals and chairman appointment in year t. Column (1) presents the results with the signed value of accrual earnings management as the dependent variable. Column (2) reports the results with the dependent variable taking the absolute values of the negative AEM, and zero for positive AEM. All continuous variables are winsorized at the 1st and 99th percentiles. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively, in two-tailed tests. STAY is a dummy variable taking the value of one if the leaving chairman still stays on the board and zero otherwise. Detailed definitions of other variables are outlined in Table 1. Equation (8) is estimated.

presents the association between downward accrual earnings management and chairman appointment in year t. The coefficient on SUCCESSION was positive and significant at the 10% in column (1), while it was negative and significant at the 10% in column (2). The results suggest outside professionals are more likely to manage earnings downward in the year of their appointment relative to within-family successors.

4.4. Robustness Test

4.4.1. Self-Selection Bias

While this study examines the impact of within-family successions on earnings management, it is also possible that founders’ succession decision is driven by other firm-level characteristics, which may also explain founders’ earnings management behavior. This poses a potential endogenerity problem arising from self-selection bias. To alleviate the concern for endogeneity, a robustness test was conducted using Heckman two-stage analysis.

Table 7. Heckman two-stage test on the impact of within-family successions on downward earnings management.

First Stage Probit	DV = SUCCESSION	Second Stage OLS	DV = AEM
FOUNDERAGE	0.2474 *** (3.27)	SUCCESSION	−0.0296 *** (−2.66)
SIZE	−0.7931 * (−1.6)	SIZE	0.0401 * (1.64)
LEV	2.2574 (1.08)	LEV	−0.1835 *** (−2.52)
ROA	14.3024 * (1.73)	ROA	0.1054 ** (2.21)
LOSS	−1.7818 * (−1.93)	LOSS	−0.1486 * (−1.93)
MVBV	−0.6815 *** (−2.65)	MVBV	0.0077 * (1.76)
FIRMAGE	0.0176 (0.31)	FIRMAGE	−0.0024 (−0.73)
AUDITOR	0.4765 (0.72)	AUDITOR	0.0481 ** (2.25)
DUAL	0.6023 *** (2.8)	DUAL	−0.0046 *** (−3.09)
FSR	0.027 *** (2.82)	FSR	−0.0016 *** (−2.85)
		IMR	0.0013 (0.5)
Intercept	−4.0716 ** (−2.01)	0.1099 ** (2.56)	−0.9594 * (−1.92)
Year fixed effects	Yes	Year fixed effects	Yes
Industry fixed effects	Yes	Industry fixed effects	Yes
Pseudo R-sq	0.2661	Adj. R-sq	0.1062
N	450	N	450

Note: This table presents the Heckman two-stage test on the effect of within-family successions on downward earnings management. All continuous variables are winsorized at the 1st and 99th percentiles. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively, in two-tailed tests. FOUNDERAGE is the age of family firm founders. Detailed definitions of other variables are outlined in Table 1.

presents the results of the Heckman two-stage test. In the first stage, the Probit model was used to regress the dummy variable, SUCCESSION, on several firm-level control variables. The results suggest the coefficient on FOUNDERAGE was positive and significant at the 1% level, suggesting older founders tend to transfer family business to second-generations rather than outside professionals. This is probably because older founders are less familiar with professional management market and have less trust in outsiders, and they are more likely to keep the control of the business within their families. Some control variables were also significantly associated with SUCCESSION. The inverse Mills ratio generated from the first stage Probit model was then used in the second stage OLS regression with the dependent variable of AEM, the variable of interest, SUCCESSION, and the remaining control variables. The second stage regression results show the coefficient on SUCCESSION was negative and significant at the 10% level, consistent with the baseline model results in Table 3. The Heckman test for ABSAEM, REM, and ABSREM was repeated, and the results generally resemble each other (untabulated for brevity).

4.4.2. The Alternative Measures of Real Earnings Management

As suggested by Roychowdhury [60], some activities leading to abnormally high production costs also lead to abnormally low CFO. Hence, counting these two measures can result in double counting. Moreover, these three measures capture different types of real earnings management. Therefore, two alternative measures (ALREM₁ and ALREM₂) of real earnings management were employed. Similarly, two variables consisting of the negative ALREM₁ and ALREM₂, and zero for positive ALREM₁ and ALREM₂ (ABSALREM₁ and ABSALREM₂) were also used in the robustness test. The alternative measures of real earnings management were calculated using the following equations:

$$ALRE{M}_1=APRO{D}_{i,t}+\left(-1\right)ADISEX{P}_{i,t}$$

(11)

$$ALRE{M}_2=\left(-1\right)ACF{O}_{i,t}+\left(-1\right)ADISEX{P}_{i,t}$$

(12)

Equations (8)–(10) are re-estimated using the above alternative measures, and the untabulated results were qualitatively similar to the main findings in Table 3, Table 4 and Table 5.

4.4.3. The Alternative Measures of Other Variables

First, POVERTY was measured using the list of national-level poor counties issued by the Chinese government, and it took the value of one if family firm founders were born in one of the above counties and zero otherwise. Second, the long-term financial performance was measured using the average value of ROE from the 3rd to 5th year after family firm successions (AVEROE). Equations (9) and (10) were re-estimated using the above alternative measures and the untabulated results were similar to those reported in Table 4 and Table 5.

5. Conclusions and Discussion

5.1. Conclusions

Chinese family firm founders’ earnings management behavior before management succession is investigated in this study. It is documented that Chinese family firm founders tend to manage earnings downward in the year prior to the transition of control to the second generation. Founders with poverty experiences and political connections have stronger incentives to manage earnings downward. Institutional investors and the legal environment inhibit founders’ earnings management behavior before within-family successions, but independent directors have no significant impact on the correlation between successions and earnings management. With respect to the consequences of earnings management, the results suggest that the strategic downward earnings management avoid the large decline of stock prices around the within-family successions in the short term but is harmful for firms’ long-term financial performance.

5.2. Implications

The study has several theoretical and practical implications. In terms of theoretical implications, it first contributes to literature of family firm succession by focusing on the antecedents of management succession. Most prior succession studies examine whether firms’ financial performance improves or declines after successions, or compare the performance of companies controlled by inside successors and professional outsiders based on agency theory and steward theory. Different from prior research, this study sheds light on the management behavior before successions. Due to the close relationship between Chinese parents and their children largely originating from familism, it is expected that Chinese family firm founders may do more preparations before successions to help their children survive and thrive after successions. Hence, this study focuses on the earnings management behavior of founders before management successions and finds interesting results to complement the current succession research.

Second, this study also adds to earnings management research. Current related research independently investigates the outgoing or incoming CEO’s earnings management around management turnover. However, much research neglects the blood relationship between outgoing and incoming management in family firms and the influence of the relationship on earnings management during successions. This study explores how founders manage earnings before they pass family business to their descendants, and also tests the earnings management behavior of successors after they inherit family firms in the additional analysis. The results of this study provide evidence on the impact of blood relationship on earnings manipulation, which has not been investigated previously.

In terms of practical implications, this study is first informative to policymakers. Smooth succession is crucial to the long-term sustainable development of family firms. Considering the vital role of private enterprises in national economy and frequent successions in recent years, the Chinese government is concerned about the impact of successions on family firms’ development. The results of this study suggest that considering the negative impact of earnings manipulation on firms’ future performance, more supervision from the government is needed around successions. Second, investors make their investment decisions based on firms’ financial information and manipulated information may mislead investors. Therefore, this study’s findings suggest that investors should be aware of the possible earnings management before within-family successions. Last, due to the Confucian familism which is deeply rooted in Chinese people’s moral education, founders tend to sacrifice their economic benefits for the purpose of smooth successions. However, this study suggests founders should adopt more appropriate methods that are not harmful to firms’ future development to help their children successfully inherit family business.

5.3. Limitations and Future Research

This study has several limitations. First, it was undertaken in a Chinese environment, where family firm management is significantly influenced the Confucian familism and the one-child policy, so the research conclusion cannot be applied in other countries with different culture and political policies. Second, a large number of family firms have not begun their successions, and there will be more and more family firms controlled by successors in the future. Therefore, adding the latest data into future related research will further improve the representativeness of the conclusions. Moreover, other management behavior of family firms around management successions, such as corporate social responsibility and green innovation, can be investigated in future research.