Individualist-Collectivist Culture, Ownership Concentration and Earnings Quality: A Comparison of Western Europe and East Asia

Individualist-Collectivist Culture, Ownership Concentration and Earnings Quality: A Comparison of Western Europe and East Asia Abstract This study explores the effects of individualist-collectivist culture, an important dimension of national cultures, on the entrenchment incentives of large shareholders. Specifically, it investigates how individualist-collectivist culture affects the relationship between ownership concentration and earnings quality. Collecting data from thirteen Western Europe and nine East Asia economies between 1995 and 2011, we find that the poor earnings quality induced by ownership concentration is mitigated in individualist societies after controlling for the legal institutions, financial development and economic wealth. We further find that the poor earnings quality induced by ownership concentration is more pronounced in East Asia. This study sheds light on the role that individualist-collectivist culture plays in shaping corporate insiders ethical behavior. The findings in this study help international investors and auditors better evaluate the earnings quality and agency problems in Western Europe and East Asia. Keywords: Corporate Governance; Earnings Quality; Ownership Concentration; Cultures; Individualism; Collectivism

1. Introduction The extant literature documents the prevalence of concentrated ownership in both Western Europe and East Asia, with controlling shareholders voting rights often exceeding their cash flow rights (Haw, Hu, Hwang & Wu, 2004; Claessens, Djankov & Lang, 2000; Faccio & Lang, 2002). When voting and cash flow rights diverge, lower cash flow rights may fail to provide sufficient incentive alignment to curtail the entrenchment (Claessens, Djankov, Fan & Lang, 2002; Fan & Wong, 2002). Prior literature also documents that cultures play a critical role in corporate governance (Stulz & Williamson, 2003; Licht, Goldschmidt & Schwartz, 2005). However, it is not clear whether and how cultures affect the entrenchment incentives of large shareholders. We fill this gap by examining the effect of individualist-collectivist culture, an important dimension of national cultures, on the relationship between earnings quality and ownership concentration. We then compare this effect between Western Europe and East Asia. Although legal institutions play a central role in corporate governance, the accounting scandals incurred in the U.S. and other countries around the world indicate that mere conformity with legal procedures and regulations is insufficient to shape a better corporate governance capable of eliminating accounting scandals. Many influential researchers are making an effort to seek the factors beyond legal institutions, that affect corporate governance (Haw et al., 2004; Stulz & Williamson, 2003; Dyck & Zingales, 2004). Stulz & Williamson (2003) note that cultural differences cannot be ignored when examining why investor protection differs across countries. Licht, Goldschmidt & Schwartz (2005) state that grouping countries by legal families only provides a partial description of the universe of corporate governance regimes. Consistent with these arguments, recent studies find that the social networks among corporate insiders mitigate the monitory efficiency of directors and audit committee members, which results in 2

lower earnings quality (Hwang & Kim, 2009; Hwang & Kim, 2012). We argue that social ties among corporate insiders are closer in collectivist as opposed to individualist cultural societies. Hiring and promotion decisions usually consider employees in-groups in collectivist societies. In contrast, the hiring and promotion decisions in individualist societies are skill- and rule-based (Jackson, 2007). In collectivist societies, the controlling shareholders usually have closer social connections with managers, directors and other large shareholders, which may facilitate collusion among corporate insiders and reduce the monitory incentive of other stakeholders. Such close social connections among corporate insiders induce severe agency problems. We thus predict that the entrenchment effect of large shareholders is more pronounced; hence the poor earnings quality induced by ownership concentration is more severe in a collectivist culture. Although both Western European and East Asian firms are featured with ownership concentration, its entrenchment effects may differ in these two regions. Western Europe and East Asia demonstrate significant differences in social norms and cultures. For example, Western European cultures feature individualism while East Asian cultures are influenced by Confucianism, which features collectivism. The collectivist culture common in East Asia may mitigate the monitory efficiency of board directors, further facilitating collusions among corporate insiders. Thus, we conjecture that the poor earnings quality induced by ownership concentration will be more pronounced in East Asia. This argument is consistent with Faccio, Lang & Young (2001), who find that dividend rates are higher in Europe, but lower in Asia in the presence of multiple large shareholders. This indicates that in Europe, the other large shareholders appear to help constrain the controlling shareholder s expropriation of minority shareholders, whereas in East Asia, they appear to collude in that expropriation. 3

Collecting data from 22 Western European and East Asian countries, we find that earnings quality (measured by discretionary accruals and accounting conservatism) is negatively related to ownership concentration; however, this negative relationship is mitigated in individualist societies after controlling for the legal institutions, financial development and economic wealth. Our further analyses demonstrate that the negative relationship between ownership concentration and earnings quality is more pronounced in East Asia than in Western Europe. This study contributes to the literature in several ways. First, our findings add to the literature addressing the roles that cultures play in corporate governance (e.g., Stulz & Williamson, 2003; Licht, Goldschmidt & Schwartz, 2005). The extant literature confirms the importance of cultures in corporate governance. This study explores the role that individualism-collectivism plays in constraining the entrenchment effect of large shareholders. Second, our work contributes to the recent studies on social networks. Hwang & Kim (2009) and Hwang & Kim (2012) find that the social networks among corporate insiders mitigate the monitory efficiency of directors and audit committee members, reducing the quality of earnings. We argue that the social networks among corporate insiders are closer, and hence the agency problems are more severe in collectivist cultural societies. Third, we compare the entrenchment effects in Western Europe and East Asia and find that the poor earnings quality induced by ownership concentration is more pronounced in East Asia. This finding may help international investors and auditors better evaluate the earnings quality and agency problems in these two regions. The rest of this paper is organized as follows. Section 2 reviews the related literature and develops our hypotheses. Section 3 introduces the research design and Section 4 describes the data collection process. The empirical results are reported in Section 5 and we present our concluding remarks in Section 6. 4

2. Literature Review and Hypotheses Development 2.1 Individualist vs. Collectivist Culture Hofstede (1980, 2001) classifies national cultures into four dimensions: power distance, uncertainty avoidance, individualism and masculinity. However, he claims that the individualism dimension is the most closely linked to a country s level of economic development. Individualism represents the extent to which people in society think of themselves as autonomous individuals who are responsible primarily to themselves and their immediate families. Individualism focuses on rights over duties and concerns for individuals while emphasizing self-fulfillment and personal autonomy (Hofstede, 2001). Americans, Australians and most Western Europeans are individualists. Within an individualist society, individuals are normally autonomous, independent, self-sufficient and respectful of others rights. Their communications are based on rational principles such as equity, non-interference and equality and their rights are protected by legal institutions such as laws, regulations and rules. Therefore, they are more likely to follow legal institutions. Corporations in individualist societies aim to hire people with the highest educational and professional achievements. Individualism lies in contrast with collectivism, the latter of which emphasizes collective purposes over personal goals and group harmony over individual achievement. Collectivism considers duties over rights, exhibits concern for groups and emphasizes one s role within the group, common fate and social order (Hofstede, 2001). Individuals in collectivist societies are usually assigned particular roles within a group to fulfill their duties and obligations, and they lose face if they fail. Their communications rely on their relationships and the role-based conceptions of justice that consist of compromise and concession. Because their welfare is promoted through group harmony, they tend to be loyal to the groups in which they belong. Most 5

Asian and Latin American countries display collectivist societies. Hiring in collectivist culture is based on qualifications and knowing an applicant or his or her family is considered an important qualification. The hiring of family members is common in such societies, and collaborative decisions are as important as individual technical proficiency. Based on these descriptions, we expect that the relationships or social connections among people within a group will be stronger in a collectivist culture than in an individualist culture. 2.2 The Effect of Individualist-collectivist Culture on the Relationship between Ownership Concentration and Earnings Quality Individualist-collectivist culture has corporate governance implications. Jackson (2007) states that people in individualist culture are more likely to adhere to a universal application of rules and laws, whereas those in a collective culture are more likely to apply rules based on the relationships involved. This is consistent with Smith & Hume (2005), who argue that accountants question doubtful behavior less frequently in strong collectivist cultures. In a recent study, Zhang et al. (2012) find that earnings management is more severe in collectivist as opposed to individualist culture. The recent corporate governance literature notes that independent board directors and audit committee members play a monitoring role in constraining controlling shareholders or managers incentives to extract private benefits of control. However, if they share social networks with the controlling shareholders/managers, then the monitory efficiency is mitigated and earnings quality decreases (Hwang & Kim, 2009; Hwang & Kim, 2012). The literature also documents that the emergence of multiple blockholders either mitigates or exacerbates agency problems, depending on whether multiple large owners monitor or collude with the controlling shareholder (Faccio, Lang & Young, 2001; Laeven & Levine, 2008; Maury & Pajuste, 2005). 6

These studies suggest that the social ties among corporate insiders (including large shareholders, managers, directors and audit committee members) play a governance role. The degree of social ties among people is reflected in national cultures. As we discussed previously, corporations in collectivist societies tend to make hiring decisions based on relationships, making the social ties among corporate insiders stronger. These stronger ties facilitate corporate insiders in colluding for private benefits at the expense of outside investors. 1 In addition, Hofstede (2001) states that the individual interests of managers in collectivist societies are dominated by their group interests, which indicates that the alignment of individual interests within a group will consolidate their group interests in collectivist societies. The information asymmetry between in-group members (corporate insiders) and out-group members (outside investors) is exacerbated by a high level of collectivism. Thus, controlling shareholders are more likely to gain the private benefits of control. Hence, the agency conflicts between ingroup members (controlling shareholders) and out-group members (minority shareholders) are exacerbated by a high level of collectivism. Meanwhile, other stakeholders (e.g., large shareholders, directors and committee members) play a monitoring role to prevent the expropriation by controlling shareholders, moderating the agency problems embedded in concentrated ownership in countries with highly individualistic orientations. In contrast, managers are more likely to conform to the legal institutions (e.g., laws and regulations) and social norms (e.g., integrity and honesty) in keeping with the moral judgments of individualistic societies. Because individual interests are supposed to prevail over collective interests, every stakeholder is more likely to play a monitoring role to prevent expropriation by the controlling shareholders. 7

In sum, we argue that individualism could reduce corporate insiders incentive and ability to acquire personal benefits at the expense of minority shareholders benefits, because it s not easy to block information to the public without the alignment of corporate insiders. We thus hypothesize, Hypothesis 1. The poor earnings quality induced by ownership concentration is mitigated in individualist cultures. 2.3 Effects of Ownership Concentration on Earnings Quality in Western Europe and East Asia Although the ownership structures are highly concentrated in both Western Europe and East Asia, the social environments in which corporations operate are quite different. East Asian cultures are deeply influenced by Chinese cultures that feature collectivism. Compared with those in Western Europe, many East Asian firms are controlled by families, and the relationships are highly valued. In fact, the term crony capitalism is prevalently used to describe the close ties between large corporations in East Asia (Ball, Ashok & Wu, 2003). 2 Social scientists assume that individualism is more prevalent in industrialized Western societies than in other societies, especially more traditional societies in developing countries (Oyserman, Coon & Kemmelmeier, 2002). Thus, there are stronger social ties among corporate insiders in East Asia than among those in Western Europe. Therefore, we expect that the agency problems induced by ownership concentration will be more severe in East Asia than in Western Europe. This is consistent with Faccio et al. (2001), who find that group-affiliated Western Europe corporations pay significantly higher dividend rates than those in East Asia. They argue that the other large shareholders appear to help restrict the controlling shareholder s expropriation of minority shareholders in Western Europe, but appear to collude in such expropriation in East Asia. Based on such discussions, we hypothesize, 8

Hypothesis 2: The poor earnings quality induced by ownership concentration is more pronounced in East Asia than in Western Europe. 3. Research Design We measure earnings quality based on two dimensions: accruals quality and accounting conservatism. 3.1 Measurement of Accruals Quality There are several models for estimating the discretionary accruals. However, given the potential misspecifications in tests of discretionary accruals and their effect on inferences, we follow Haw et al. (2004) and estimate discretionary accruals using the Jones model, controlling the performance factors (ROA) as follows: AC 0 1( 1/ LTA) 2Sales 3PPE 4ROA (1) Following Hribar & Collins (2002), who suggest that the balance-sheet approach to estimate accruals is potentially contaminated and that accruals estimated using the cash-flow statement approach produce less error, total accruals (AC) is defined as the difference between net income and net cash flow from operating activities, scaled by lagged total assets. Sales is the change in sales revenues, deflated by lagged total assets. PPE is the net property, plant, and equipment scaled by lagged total assets. LTA is lagged total assets. Return on assets (ROA), included to control firm performance, is measured by net income divided by current total assets. The residuals are estimated cross-sectionally for each year using all of the firm-year observations in the same two-digit Standard Industrial Classification (SIC) code. A minimum of 10 firms in each country-year-industry combination are required. 3.2 The Effect of Individualism on the Relationship between Ownership Concentration and Accruals Quality 9

To test the impact of individualism on the relationship between ownership concentration and accruals quality, we develop the following linear model. ABDA Own IDV Own 0 1 2 3 * Institutio nalcontrol 0 IDV Year i FirmControl 0 Industry i (2) where dependent variable (ABDA) is the absolute value of residuals obtained from Equation (1). Small (big) absolute values of the residuals are associated with higher (lower) accruals quality. Closely held ownership (Worldscope item 08021) is used to proxy for the ownership concentration (Own). It represents the proportion of equity held by corporate officers, directors and immediate family members; by individual shareholder holdings representing more than5%; by other corporations (except shares held in a fiduciary capacity by financial institutions); and by pension/benefit plans and trusts. Although itself has limitations and includes some noises, it has still wildly been used by researchers (e.g., Fernandes & Ferreira, 2008; Armstrong, Barth & Jagolinzer, 2010). 3 Armstrong, Barth & Jagolinzer (2010) use the closely held ownership to measure information asymmetry, suggesting that information asymmetry is higher and agency problems are more severe in firms with more closely held ownership. The individualism index (IDV), extracted from Hofstede (2001), has been brought into a range between 0 and 1. A higher (lower) score indicates a higher level of individualism (collectivism). To explore the joint effect of IDV and Own on accruals quality, we include the interaction between IDV and Own. 3 is our focus and is anticipated to be significantly negative, suggesting that the negative impact of ownership concentration on accruals quality could be restricted by a high level of individualism. 3.3 The Effect of Individualism on the Relationship between Ownership Concentration and Accounting Conservatism 10

We draw on Ball & Shivakumar s (2005) accrual-cash flow model to test the effect of individualism on the relationship between ownership concentration and accounting conservatism. Total accruals are regressed against the interaction of individualism, ownership structure and timely loss recognition, as presented in Equation (3). AC D CFO D* CFO Own Own* D Own* CFO Own* 0 D* CFO IDV IDV * D IDV * CFO IDV * D* CFO IDV * Own IDV * Own* D IDV * Own* CFO IDV * Own* D* CFO FirmControl 0 FirmControl * D* CFO 3 1 13 Control * D D* CFO 2 8 i Institutio nalcontrol * CFO 2 3 9 Year FirmControl * D 1 Industry e i 14 4 10 Institutio nalcontrol 0 5 FirmControl * CFO 2 11 15 Institutio nal Institutio nalcontrol * 3 6 1 12 7 (3) where total accruals (AC), individualism (IDV) and ownership concentration are as previously defined. CFO is net cash flow from operating activities, deflated by lagged total assets. D is a dummy variable equal to 1 if the current-year CFO is negative and 0 otherwise. We expect that high individualism will restrict the negative effect of ownership concentration on accounting conservatism, thus, 15 should be significantly positive. 3.4 Comparison of Western Europe and East Asia The individualism index in Equations (2) and (3) is replaced by a dummy variable, R, which equals 1 if the firm is located in Western Europe, and 0 otherwise, as shown in Equations (4) and (5). We then rerun the regressions to test our second hypothesis. The other variables are maintained as presented in Equations (2) and (3), respectively. ABDA Own R Own 0 1 2 3 * Control Year i R Industry i FirmControl 0 Institutio nal 0 (4) 11

AC D CFO D* CFO Own Own* D Own* CFO Own 0 * D* CFO R R* D R* CFO R* D* CFO R* Own R* Own* D R* Own* CFO R* Own* D* CFO 1 Control * CFO e 1 FirmControl * D Institutio nalcontrol 0 2 14 8 9 FirmControl * CFO 3 Institutio nalcontrol * D Institutio nalcontrol * D* CFO 3 3 2 10 15 1 4 5 11 FirmControl FirmControl * D* CFO 6 0 Year i 12 Institutio nal 2 Industry i 7 13 (5) We anticipate a significant and positive coefficient ( 15 ) on R*Own*D*CFO, suggesting that the negative association between ownership concentration and accounting conservatism is less prominent in Western Europe. 3.5 Control Variables We include a set of firm- and institutional-level control variables that may affect the relationship between ownership structure and earnings quality. Leverage ratio (LEV) is measured by total liabilities divided by total assets, which controls for the importance of debt contracting in determining the equilibrium level of earnings quality, given that debt contracting represents another potential source of demand for quality earnings. Firms market value (MV) is used to control the large firms with larger and more stable accruals (Dechow & Dichev, 2002) and is measured by the logarithm of a firm s market value. Growth rate (Gr) is measured by the change in sales revenues divided by the beginning sales revenues and controls for a corporation s growth opportunities. Creditor rights (CR) and legal origin (LAW) are included to control for the crosscountry differences of legal environments. Creditor rights capture the degree of legal protection for creditors against defaulting debtors in different jurisdictions while the legal origin indicates the degree of shareholders legal protection. Previous literature suggests that civil law countries have weaker shareholder protection than common law countries in terms of both legal regulations and law enforcement (La Porta, Lopez-de-Silanes, Shleifer & Vishny, 1998). Ball et 12

al. (2003) claim that common law countries give exclusive corporate governance rights to shareholders, whereas civil law countries represent the interests of non-shareholder stakeholders in corporate governance. Consistently, Bushman et al. (2004) document that governance transparency is higher in countries with a common legal origin, indicating that information between insiders and outsiders is less asymmetric. However, some of the literature suggests that the revised anti-director rights capture the degree of legal protection of investors. In a robustness check, we use the revised anti-director rights (RADR) as an alternative measure of shareholder protection. 4 Finally, following La Porta, Lopez-de-Silanes, Shleifer & Vishny (2000), we include the natural logarithm of GDP per capita (LnGDP) to control for cross country differences in economic development, and market capitalization to GDP ratio (MCGDP) to control for cross country differences in financial market development. All of the variables are listed and defined in Table 1. ---------------------------------------------------------------- Insert Table 1 about here ---------------------------------------------------------------- 4. Sample and Data Collection Our sample includes listed firms from thirteen Western European economies (Belgium, France, Italy, Germany, Austria, Finland, Ireland, Spain, Norway, Switzerland, Sweden, Portugal and the U.K.) and nine East Asian economies (Korea, Indonesia, Hong Kong, Malaysia, Singapore, Japan, the Philippines, Thailand and Taiwan). All of the accounting and financial data are retrieved for the years between 1995 and 2011 from the Worldscope database. The initial sample contains 178,676 firm year observations representing 10,568 firms in 22 economies. Financial 13

institutions (SIC 6000-69999) and regulated utilities (SIC 4900-4999) are excluded from our sample. In addition, observations with missing data are eliminated. Following previous discretionary accrual studies (e.g., Leuz, Nanda & Wysocki, 2003; Kothari, Leone & Wasley, 2005), we exclude firm-year observations with insufficient data for computing total accruals or the variables needed to estimate the Jones (1991) model. An additional filter is applied in our sample, such that observations with any two-digit SIC code and year combinations with less than ten observations are removed. The final sample contains 89,208 observations. Due to scaling problems and data errors, all of the variables are winsorized at the 1% level to avoid drawing spurious inferences from the outliers. GDP per capita is extracted from the World Bank database, while the creditor rights and anti-director rights are obtained from Djankov, Mcliesh & Shleifer (2007) and Djankov, La Porta & Lopez-de-Silanes (2008), respectively. Panel A of Table 2 shows the number of firm-year observations per country and the descriptive statistics for country-level variables. The number of firm-year observations varies significantly across countries. The observations in the U.K. and Japan account for a large proportion of our sample, thus we perform sensitivity tests without observations from Japan and the U.K. The regional averages for Western Europe and East Asia are also given in the table. East Asian firms have a higher mean of absolute discretionary accruals (0.07) than that of Western European firms (0.06), indicating that on average, East Asian firms are more likely to use their discretionary power to manipulate accruals compared with Western European firms. In addition, East Asian firms have a relatively higher ownership concentration (0.51) than that of Western European firms (0.46). The mean value of IDV for Western European countries is 0.66 and that for East Asian countries it is 0.24, which means that Western Europe is more individualistic than East Asia. However, the mean value of CLV for Western Europe is 4.55, 14

which is relatively lower than that for East Asia (5.55). Panel B of Table 2 presents the descriptive statistics of firm-level variables. It also demonstrates considerable variation in each variable. For example, the minimum insider ownership (Own) is 0.1%, with a maximum amount of 91.2%. Panel B also exhibits large variations of market value (MV), leverage ratio (LEV) and growth rate (Gr). For instance, the minimum leverage ratio (LEV) is 4.9%, whereas the maximum is 134.3%. ---------------------------------------------------------------- Insert Table 2 about here ---------------------------------------------------------------- Table 3 presents the Pearson correlation coefficients for the variables used in our tests. It displays a significant positive relationship between ownership concentration (Own) and absolute discretionary accruals (ABDA), indicating that when ownership is more concentrated, accruals manipulation is more severe. A significant negative correlation is also found between firms market value (MV) and absolute discretionary accruals (ABDA), which is similar to findings in previous studies (e.g., Warfield, Wild & Wild, 1995). Large firms may display less accruals management due to their relatively well-founded corporate governance. This also indicates that a higher growth rate (Gr) significantly correlates with higher absolute discretionary accruals (ABDA). In addition, the correlation between firms market value (MV) and inside ownership (Own) is significantly negative, which is consistent with Haw et al. (2004). In addition, Table 3 demonstrates a significant negative correlation between total accruals (AC) and cash flow from operating activities (CFO), which is consistent with Dechow (1994) and confirms that firms market value (MV) and growth rate (Gr) are positively related to total accruals. Moreover, it shows that individualism (IDV) is significantly and positively related to inside ownership (Own) 15

and absolute discretionary accruals (ABDA) while collectivism (CLV) is significantly and negatively related to them, suggesting that individualist culture is highly related to earnings quality. ---------------------------------------------------------------- Insert Table 3 about here ---------------------------------------------------------------- 5. Empirical Results 5.1 Results from Accruals Quality Multivariate tests are performed in Equations (2) and (3) to empirically test the prediction that earnings quality falls with a high convergence of ownership structure, and that this negative association is weaker in countries with a high level of individualism. We also include a set of control variables, respectively, to ensure that the results are not subject to important factors that have systematic effects on the cross-sectional variations in earnings quality. Table 4 presents the results from Equations (2) and (4). Column (1) illustrates the relationship between discretionary accruals and ownership concentration. The coefficient on Own is significantly positive (0.01), suggesting that earnings quality is poor in firms with ownership concentration. This result is consistent with the entrenchment effect of large shareholders. However, the coefficient on IDV*Own reported in Column (2) is -0.01 (significant at the 1% level), indicating that the positive association between and Own and ABDA is mitigated in individualistic societies, which supports our first hypothesis. ---------------------------------------------------------------- Insert Table 4 about here 16

---------------------------------------------------------------- Column (3) of Table 4 reports on the comparison of Western Europe and East Asia. The coefficient on Own is 0.01, significant at the 1% level, whereas the coefficient on R*Own is - 0.01, indicating that the poor accruals quality induced by ownership concentration is less prominent in Western Europe. This result is consistent with our second hypothesis. The coefficients on firms market value (MV) in all columns are significantly negative, which indicates that large firms engage in less accruals management due to their well-founded corporate governance. In addition, the coefficients for the leverage ratio (LEV) are both significant and positive; that is, a higher debt ratio indicates increased accruals management. The coefficients on growth rate (Gr) are also both significant and positive, implying that higher growth opportunities spur increased accruals management. In addition, the coefficients on LAW are significant and negative across all of the columns, suggesting that increased legal protection for shareholders results in decreased accruals management. All of the adjusted R²s of these three regressions are around 9.5%, indicating that the models are well designed. 5.2 Results from accounting conservatism Table 5 presents our estimates from Equations (3) and (5). Column (1) shows the regression results from the benchmark model. By conditioning the relation between accruals and operating cash flows on the sign of current cash flows, Ball and Shivakumar (2005) claim that there is a positive relation between accruals and operating cash flows when cash flows are negative (i.e., the coefficient of D*CFO is greater than 0). In confirmation, the coefficient on D*CFO (0.62) is significant and positive, indicating that loss recognition occurs in a timely manner, and the coefficient on Own*D*CFO (-0.26) is significantly negative, indicating that the presence of ownership concentration restricts the recognition of losses in a timely manner. In Column (2), the 17

coefficient on IDV*Own*D*CFO is 0.54 (significant at the 5% level), suggesting that the negative effect of ownership concentration on accounting conservatism is mitigated in individualistic societies, which supports our first hypothesis. Column (3) of Table 5 presents the estimates from accounting conservatism. The results reveal that the coefficient on Own*D*CFO (-0.45) is significantly negative. The coefficient on regional difference (R*Own*D*CFO) is 0.43, suggesting that ownership concentrated firms in Western Europe adopt more conservative accounting policies than those in East Asia, further supporting our second hypothesis. ---------------------------------------------------------------- Insert Table 5 about here ---------------------------------------------------------------- The coefficients on MV*D*CFO are significant and negative. Additionally, the coefficient on LEV*D*CFO is insignificant in Column (1), but significantly positive in Column (2). The coefficients on Gr*D*CFO are significantly positive. The coefficient on LAW*D*CFO is significantly positive in Column (1), but insignificant in Column (2). Finally, the coefficients on CR*D*CFO are not significant across all three columns. The adjusted R²s of the three regressions are about 30%, indicating that the models are well designed. 5.3 Robustness Tests To ensure the validity of our results, we conduct a series of robustness tests and report the empirical results in Table 6. 5.3.1 Using Ultimate Ownership Given that closely held shares consist of the sum of all block holdings that may not accurately capture control rights, we alternately retest our hypotheses using ultimate ownership. Our proxy 18

of inside ownership (Own) is replaced by divergence (DVG), which equals 1 if the difference between the voting and cash flow rights of the controlling shareholders is greater than 0, and 0 otherwise. Meanwhile, we control for the voting rights of controlling shareholders (Vote). Because the ultimate ownership is calculated around 1996, we limit our sample to before 2002 to ensure the validity of our results. As Panel A of Table 6 shows, the coefficients on DVG are all significantly positive across Columns (1) to (3). The coefficient on DVG*IDV is -0.01, significant at the 10% level. The coefficient on DVG*R is -0.01, significant at the 5% level. In addition, Panel B shows that the coefficients on D*CFO are all significantly positive and that the coefficients on Own*D*CFO are all significantly negative across Columns (1) to (3). The coefficient on DVG*IDV*D*CFO is 1.80, significant at the 1% level. The coefficients on R*D*CFO and DVG*R*D*CFO are both significantly positive. These findings suggest that our main results are not sensitive to the use of ownership divergence as a proxy for ownership concentration. 5.3.2 Using an Alternative Cultural Index Although Hofstede s individualism index is used very often in the literature, some researchers question its reliability and validity because it was developed in the 1970s. Thus, we use a collectivism index (CLV) obtained from Gelfand et al. (2004). CLV ranges from 0 to 10 with a higher value representing a higher level of collectivism and a lower level of individualism. Column (4) of Panel A reports significantly positive coefficients on Own and CLV. The coefficient on CLV*Own is 0.00 (significant at the 1% level), indicating that collectivism deteriorates the poor accrual quality induced by ownership concentration. Panel B of Column (4) reveals that both of the coefficients on CLV*D*CFO and CLV*Own*D*CFO are significantly negative, which indicates that the less conservative accounting adopted by ownership 19

concentrated firms becomes much less conservative in collectivist cultures. These findings suggest that our main results remain unchanged despite using an alternative measure of individualism. 5.3.3 Pre- vs. Post-IFRS Adoption Since Europe mandatorily adopted IFRS in 2005, its adoption has increased the earnings quality of closely held firms in Europe (Armstrong et al., 2010). Thus, we expect the entrenchment effect of large shareholders to be reduced upon the adoption of IFRS, increasing the differences in earnings quality induced by ownership concentration between Western Europe and East Asia after the adoption of IFRS. We separate our sample into two sub-samples and run the regressions separately. Column (5) of Table 6 reports the regression results for the period before 2005, whereas Column (6) of Table 6 reports the results for the period after (including) 2005. In Panel A, the coefficient on Own*R in Column (6) is -0.01, and -0.01 in Column (5) (both significant at the 1% level). The magnitude of the coefficient in Column (6) is larger than that of the corresponding coefficient in Column (5), indicating that after Western Europe adopts the IFRS, the accruals quality for ownership concentrated firms improve compared with that of East Asian firms that have not adopted IFRS. In Panel B, the coefficient on R*Own*D*CFO in Column (6) is 0.56, and 0.210 in Column (5) (both significant at the 5% level). These results suggest that after the adoption of IFRS in Europe, Western European closely held firms gain more advantages in terms of accounting conservatism compared with their East Asian counterparts. 5.3.4 Excluding Observations from Japan and the U.K. The data from Japan and the U.K. hold a large proportion of our sample, to the extent that our results might have been driven by the data from these two countries. Hence, we exclude these 20

two countries in our sample. Columns (7) through (9) in Panel A show that the coefficients on Own are all significantly positive. The coefficient on IDV is 0.00 (significant at the 5% level), the coefficient on IDV*Own is -0.03 (significant at the 1% level) and the coefficient on R*Own is -0.02 (significant at the 1% level). In contrast, Columns (7) through (9) in Panel B show that the coefficients on D*CFO and Own*D*CFO are all significantly negative. The coefficient on IDV*D*CFO is 0.42 (significant at the 10% level) and the coefficient on IDV*Own*D*CFO is 0.94 (significant at the 5% level). The coefficients on R*D*CFO and R*Own*D*CFO are both significantly positive (significant at the 10% and 5% levels, respectively). Obviously, our results are consistent with our main tests after dropping the observations from Japan and the U.K. ---------------------------------------------------------------- Insert Table 6 about here ---------------------------------------------------------------- 5.3.5 Excluding observations from 1997-1998 and 2008-2009 We exclude observations from 1997-1998 and 2008-2009 to avoid the effects of the financial crisis. The untabulated results demonstrate that our main findings are not sensitive to this exclusion. 6. Conclusions Because the recent series of accounting scandals have attracted worldwide attention, it is important for accounting professionals to understand the consequences and roots of accounting earnings manipulation. This study focuses on individualist culture, an important dimension of national culture, to explore the influence of individualist cultures on the entrenchment effects of large shareholders. We argue that individualism/collectivism reflects the strength of social ties 21

among people within a group (or corporation). As social connections in a collectivist culture are stronger, such connections among corporate insiders mitigate monitory efficiency and facilitate collusion. We predict that the entrenchment effect induced by ownership concentration will be more severe in a collectivist culture. Furthermore, considering the different degrees of individualism in Western Europe and East Asia, we compare the entrenchment effects of large shareholders in these two regions. We test our predictions in the context of earnings quality, measured by accrual quality and accounting conservatism. Consistent with our predictions, our empirical results demonstrate that the poor earnings quality induced by ownership concentration is mitigated in individualist societies after controlling for the legal institutions, financial development and economic wealth. Because Western European cultures feature individualism and East Asian cultures feature collectivism, our further comparisons document that the poor earnings quality induced by ownership concentration is more pronounced in East Asia than in Western Europe. Our study emphasizes the role of individualist culture in affecting large shareholders entrenchment effects. Notes 1 Hofstede (2001) claims that people s attitudes toward others are independent of group members in individualist societies. Furthermore, Steensma, Marino & Weaver (2000) document that people in collectivist societies exhibit a stronger preference for cooperative strategies. 2 Ball et al. (2003) compare the accounting conservatism among four East Asian countries (Hong Kong, Malaysia, Singapore and Thailand) and find that although these four countries are common law oriented, their financial reporting quality is not higher than that of those under code law. 3 Stulz (2005) compares the results of using the two proxies (i.e., closely held ownership extracted from Worldscope vs. family-controlled ownership data proposed by La Porta, Lopez-de-Silanes & Shleifer (1999) and Claessens et al. (2000)), and shows consistent results to explain the expropriation index and the anti-director rights index. He further discusses the strengths and weaknesses of insider ownership as a determinant of agency problems in detail. 4 The untabulated results display a similar outcome to those using revised anti-director rights ( RADR ). 22

Refferences Armstrong, C. S., Barth, M. E., Jagolinzer, A. D., & Riedl, E. J., 2010. Market reaction to the adoption of IFRS in Europe. The Accounting Review 85(1):31-61. Ball, R., Ashok, R. & Wu, J., 2003. Incentives versus standards: Properties of accounting income in four East Asian countries. Journal of Accounting and Economics 36(1-3): 235-270. Ball, R. & Shivakumar, L., 2005. Earning quality in UK private firms: Comparative loss recognition timeliness. Journal of Accounting and Economics 39(1): 83-128. Bushman, R., Piotroski, J. & Smith, A., 2004. What determines corporate transparency? Journal of Accounting Research 42(2): 207-252. Claessens, S., Djankov, S. & Lang, L., 2000. The separation of ownership and control in East Asian corporations. Journal of Financial Economics 58(1-2): 81-112. Claessens, S., Djankov, S., Fan, J. & Lang, L., 2002. Disentangling the incentive and entrenchment effects of large shareholdings. Journal of Finance 57(6): 2741-2772. Hribar, P. & Collins, D., 2002. Errors in estimating accruals: Implications for empirical research. Journal of Accounting Research 40(1): 105-135. Dechow, P.M., 1994. Accounting earnings and cash flows as measures of firm performance: The role of accounting accruals. Journal of Accounting and Economics 18(1): 3-42. Dechow, P.M. & Dichev, I.D., 2002. The quality of accruals and earnings: The role of accruals estimation errors. The Accounting Review 77(4): 35-59. Djankov, S., McLiesh, C. & Shleifer, A., 2007. Private credit in 129 countries. Journal of Financial Economics 84(2): 299-329. Djankov, S., La Porta, R., Lopez-de-Silanes, F. & Shleifer, A., 2008. The law and economics of self-dealing. Journal of Financial Economics 88(3): 430-465. Dyck, A. & Zingales, L., 2004. Private benefits of control: An international comparison. Journal of Finance 59: 537-600. Faccio, M. & Lang, L.H.P., 2002. The ultimate ownership of Western European corporations. Journal of Financial Economics 65(3): 365-395. Faccio, M., Lang, L.H.P, & Young, L., 2001. Dividends and expropriation. American Economic Review 91(1): 54-78. Fan, J. & Wong, T., 2002. Corporate ownership structure and the informativeness of accounting earnings in East Asia. Journal of Accounting and Economics 33(3): 401-425. Fernandes, N. & Ferreira, M.A., 2008. Does international cross-listing improve the information environment? Journal of Financial Economics 88(2): 216-244. Gelfand, M.J., Bhawuk, D.P.S., Nishii, L.H. and Bechtold, D.J. (2004) Individualism and Collectivism. In R.J. House, P.J. Hanges, M. Javidan, P.W. Dorfman, and V. Gupta (Eds.), Culture, Leadership, and Organizations: The GLOBE Study of 62 Societies: 437 512. Thousand Oaks: Sage. Haw, I., Hu, B.B., Hwang, L. & Wu, W., 2004. Ultimate ownership, income management, and legal and extra-legal institutions. Journal of Accounting Research 42(2): 423-461. Hofstede, G., 1980. Culture s consequences: International differences in work-related values. Beverly Hills, CA: Sage Publications. Hofstede, G., 2001. Culture s consequences: Comparing values, behaviors, institutions, and organizations across nations. 2nd Edition. Thousand Oaks, CA: Sage Publications. Hwang, B.H. & Kim, S., 2009. It pays to have friends. Journal of Financial Economics 93: 138-158. Hwang, B.H. & Kim, S., 2012. Social ties and earnings management. Working Paper. Jackson, T., 2007. Cross-culture sensitivities in developing corporate ethical strategies and practices. Corporate Ethics and Corporate Governance, edited by Walther Ch. Zimmerli et al. Springer. Jones, J., 1991. Earnings management during import relief investigations. Journal of Accounting Research 29(2): 193-228. Kothari, S.P., Leone, A.J. & Wasley, C.E., 2005. Performance matched discretionary accrual measures. Journal of Accounting and Economics 39: 163-197. La Porta, R., Lopez-de-Silanes, F. & Shleifer, A., 1999. Corporate ownership around the world. Journal of Finance 54(2): 471-517. La Porta, R., Lopez-de-Silanes, F., Shleifer, A. & Vishny, R.W., 1998. Law and finance. Journal of Political Economy 106(6): 1113-1155. 23

La Porta, R., Lopez-de-Silanes, F., Shleifer, A. & Vishny, R.W., 2000. Agency problems and dividend policies around the world. The Journal of Finance 55(1): 1-33. Laeven, L. & Levine, R., 2008. Complex ownership structures and corproate valuations. Review of Financial Studies 21(2): 579-604. Leuz, C., Nanda, D. & Wysocki, P.D., 2003. Earnings management and investor protection: An international comparison. Journal of Financial Economics 69(3): 505-527. Licht, A., Goldschmidt, C. & Schwartz, S., 2005. Culture, law, and corporate governance. International Review of Law and Economics 25(2): 229 255. Maury, B. & Pajuste, A., 2005. Multiple large shareholders and firm value. Journal of Banking and Finance 29(7): 1813-1834. Oyserman, D, Coon, H.M. & Kemmelmeier, M., 2002. Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin 128(1): 3-72. Smith, A. & Hume, E., 2005. Linking culture and ethics. Journal of Business Ethics 62(3): 202-220. Steensma, H.K., Marino, L. & Weaver, K.M., 2000. Attitudes toward cooperative strategies: A cross-cultural analysis of entrepreneurs. Journal of International Business Studies 31: 591-609. Stulz, R.M. & Williamson, R., 2003. Culture, openness, and finance. Journal of Financial Economics 70(3): 313-350. Warfield, T., Wild, J. & Wild, K., 1995. Managerial ownership, accounting choices, and informativeness of earnings. Journal of Accounting and Economics 20(1): 61-91. Zhang, X., Liang, X. and Sun, H. 2012. Individualism-collectivism, private benefits of control, and earnings quality: a cross-culture comparison. Journal of Business Ethics. Forthcoming. 24

TABLE 1 Variable Definition Variable Description Source Firm-Level Variables AC Total accruals, which is the difference between net income and cash from operations, scaled by lagged total assets. Worldscope ABDA Absolute value of discretionary accruals. Worldscope CFO Cash flow from operating activities, scaled by lagged total assets. Worldscope D Dummy variable, equals 1 if the current year CFO is smaller than 0, and 0 otherwise. Worldscope Own The percentage of closely held block holdings. Worldscope R Dummy variable, equals 1 if the firm is from Western Europe, and 0 otherwise. Worldscope TA Lagged total assets. Worldscope PPE The net property, plant, and equipment of each firm scaled by lagged total assets. Worldscope Sales Changes in sales revenues, scaled by lagged total assets. Worldscope ROA Return on assets, measured by net income and divided by total assets. Worldscope LA Debt ratio, measured by total liabilities and divided by total assets. Worldscope MV The natural logarithm of firms market value (in U.S. dollars) of common equity. Worldscope Gr DVG Growth rate, measured by changes in sales revenues and divided by the beginning sales revenue. Ownership divergence, equals 1 if the difference between the voting and cash flow rights of the controlling shareholders is greater than 0, and 0 otherwise. Worldscope Faccio & Lang (2000), Claessens et al. (2000) Vote Voting rights of controlling shareholders. Faccio & Lang (2000), Claessens et al. (2000) Country-Level Variables IDV Individualism index, ranges from 0 to 1, with a higher score indicating high individualism orientation. Hofstede (2001) CLV Collectivism index, ranges from 0 to 10, with a higher score indicating high collectivism Gelfand et al. (2004) orientation. LAW Legal origin, equals 1 if the country has a common law tradition, and 0 otherwise. Djankov et al. (2008) RADR The revised anti-director rights index, which is an index aggregating six minority Djankov et al. (2008) shareholders legal rights based on laws and regulations applicable to public firms in 2003. It ranges from 0 to 6, with a higher value indicating strong shareholder rights. CR The creditor rights index, which is an index aggregating four creditor rights. It ranges from 0 to 4, with a higher value indicating strong creditor rights. Djankov et al. (2007) LnGDP The natural logarithm of the gross domestic product per capita in U.S. (thousand) dollars. World Bank databases SMGDP Ratio of total market capitalization over GDP. World Bank databases 25

TABLE 2 Descriptive Statistics Panel A presents the descriptive statistics of country-level variables, which are drawn from 21 economies (13 Western European and 9 East Asian) between 1995 and 2011. Panel B shows firm-level variables with data available to estimate our main regressions. All variables are defined in Table 1. Panel A: Country N ABDA Own IDV CLV LAW CR ASD RADR LnGDP MCGDP Hong Kong 7,357 0.08 0.56 0.25 5.32 1 4 0.96 5 3.30 4.14 Indonesia 2,479 0.08 0.66 0.14 5.68 0 4 0.68 4 0.26 0.30 Japan 19,364 0.04 0.43 0.46 4.63 0 2 0.48 3.5 3.58 0.77 Malaysia 6,579 0.07 0.49 0.26 5.51 1 4 0.95 5 1.68 1.44 Philippines 853 0.06 0.68 0.32 6.36 0 0 0.24 3 0.28 0.48 Singapore 4,539 0.07 0.55 0.2 5.64 1 4 1 5 3.37 1.80 South Korea 5,465 0.07 0.38 0.18 5.54 0 3 0.46 3.5 2.73 0.72 Taiwan 4,495 0.06 0.27 0.17 5.59 0 2 0.56 3 2.76 1.29 Thailand 2,942 0.07 0.56 0.2 5.7 1 3 0.85 4 1.02 0.58 East Asia Avg. 6,008 0.07 0.51 0.24 5.55 0.44 2.89 0.69 4.00 2.11 1.28 Austria 514 0.06 0.52 0.55 4.85 0 3 0.21 2.5 3.58 0.27 Belgium 718 0.06 0.52 0.75 --- 0 2 0.54 3 3.49 0.67 Finland 1,206 0.06 0.36 0.63 4.07 0 1 0.46 3.5 3.50 1.12 France 4,823 0.05 0.55 0.71 4.37 0 0 0.38 3 3.43 0.81 Germany 4,285 0.07 0.52 0.67 4.02 0 3 0.28 2.5 3.45 0.48 Ireland 650 0.06 0.33 0.7 5.14 1 1 0.79 4 3.59 0.54 Italy 1,689 0.05 0.52 0.76 4.94 0 2 0.39 2.5 3.35 0.39 Norway 1,301 0.07 0.47 0.69 --- 0 2 0.44 3.5 4.01 0.51 Portugal 389 0.06 0.61 0.27 5.51 0 1 0.49 2.5 2.76 0.40 Spain 553 0.06 0.49 0.51 5.45 0 2 0.37 5 3.40 0.90 Sweden 1,786 0.05 0.36 0.71 3.66 0 2 0.34 3.5 3.61 1.08 Switzerland 2,000 0.05 0.42 0.68 3.97 0 1 0.27 3 3.88 2.39 United Kingdom 15,221 0.06 0.34 0.89 4.08 1 4 0.93 5 3.44 1.40 Western Europe Avg. 2,703 0.06 0.46 0.66 4.55 0.15 1.85 0.45 3.35 3.50 0.84 26