The Boss Behind the Scenes: Nonfamily Leadership and Earnings Management

ABSTRACT

We examine the effect of nonfamily leadership in family firms, whereby family members do not act as either a board chair or a CEO, on corporate earnings management. We find that firms with nonfamily leadership conduct significantly more earnings management than firms with family leadership, although this effect can be alleviated by more effective internal control and a stronger reputation concern. Additional analyses show that the increased related-party transactions can be one mechanism through which the nonfamily leadership increases earnings management. We further find that nonfamily leaders in family firms are on average weaker but receive higher (excess) payment than their peers in nonfamily firms, all consistent with the role of the nonfamily member as a marionette for the family’s misbehaviors. Moreover, earnings management increases with the degree of family noninvolvement and is mainly driven by nonfamily chairs rather than the CEOs. Overall, our paper suggests that not having a family leader aggravates incentives for the controlling family to manage earnings to realize private benefits.

Keywords:

• Family firms
• Nonfamily leadership
• Family involvement
• Earnings management

Acknowledgments

We thank Jeffrey Ng (the Editor) and two anonymous reviewers for their helpful comments and suggestions. All authors contributed equally to this work.

Data Availability Statement

The data that support the findings of this study are available from public sources as described in the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental Data and Research Materials

Supplemental data for this article can be accessed online at https://doi.org/10.1080/09638180.2022.2114514.

Appendix 1. Measures of earnings management

Appendix 2. Other robustness checks

Appendix 3. Determinants of nonfamily leadership

Appendix 4. Degree of family noninvolvement

Notes

1 Please refer to Tabor et al. (2018) for a detailed review of research on nonfamily members in family firms.

2 Previous literature finds that around 20% of votes is usually enough to have effective control of the firm (La Porta et al., 1999). Moreover, while pyramid ownership structure is common in China, this further enlarge the excess difference between family control rights and cash flow rights. This thus worsens the conflict of interest between the controlling shareholder and minority shareholders (Jiang & Kim, 2020).

3 In untabulated test, we find family firms conduct significantly higher levels of earnings management compared with nonfamily firms, suggesting a severe Type II agency problem in information disclosure in family firms.

4 If a firm has experienced financial losses for two consecutive years, the firm is designated as an ST firm. If the firm cannot recover to be profitable in the following year, it is then designated as a PT firm and faces the risk of de-listing. Special requirements on trading exist for ST and PT firms. For example, the price volatility of ST firms must be within ±5% daily, and the increase of price for PT firms cannot be higher than 5% daily while the decrease of price is unlimited.

5 In the CSMAR database, family relationships refer to blood ties or marriage to the ultimate controller, so our category includes the controllers themselves, husbands and wives, parents and children, siblings, cousins, mothers- and fathers-in-law, sons- and daughters-in-law, aunts and uncles, and so on. If both the database and public information show no family relationship between the board chair (CEO) and the ultimate controller, we define this chair (CEO) as a nonfamily chair (CEO).

6 Please see the Online Appendix 1 for detailed descriptions of the earnings management measures.

7 Given these differences in firm characteristics between family-led and nonfamily-led firms, we employ the propensity score matching method in the robustness tests to alleviate such a possible selection bias.

8 Please see the Online Appendix 2 for other robustness tests showing that our results are robust to alternative definitions of family firm, additional controls for firm leaders’ (i.e., chair and CEO’s) personal characteristics, and additional controls for factors affecting family’s incentives to manipulate earnings.

9 To address the concern that chair/CEO turnover is also endogenous regarding earnings quality, we use unforced turnover, defined as the chair or CEO leaving the firm because of health concerns, retirement, litigation, end of tenure, and so on. The results remain similar, indicating that firms with nonfamily leaders conduct more earnings management than firms led by family members.

10 The results remain similar when we use the strength of local clan culture as an instrument, measured as the logarithm of one plus the number of genealogies documenting family histories in each province.

11 Please see the Online Appendix 3 for results of the first-stage predicting model and discussions on the determinants of nonfamily leadership.

12 When we exclude firms that experience changes in family leadership, our results still hold.

13 Related-party transactions are classified into 17 groups by the contents or objects of the deals in the CSMAR database. Categories include commodity transactions (#01), transaction in assets (#02), provided or received services transactions (#03), agencies (#04), cash transactions (#05), loan guarantees (#06), asset leases (#07), custodies (#08), gifts (#09), non-monetary transactions (#10), equity transfers (#13), debt transactions (#15), joint projects (#17), license agreements (#18), R&D products (#19), compensation for critical executives (#20) and others (#21). In the untabulated test, we exclude the categories #17 to #21 when calculating the total amount of related party transactions and find consistent results.

14 Requirement for the mediation effect to hold is (1) the mediator can predict the independent variable in the first regression; (2) the independent variable can affect the dependent variable in the second regression; and (3) the effect of the independent variable on the dependent variable is weaker in the third regression than that in the second regression when the mediator is included in the model.

15 NFlead in the third-stage regression is still significant after controlling for RPT, suggesting the treatment effect of nonfamily leadership on earnings management is partially rather than completely mediated by the variable RPT (Judd & Kenny, 1981). To the extent that complete mediation is rare since it is difficult to identify all possible mediators of the treatment effect, nonfamily leadership could increase corporate earnings management through other indirect ways. However, as we observed in the third stage of the mediation test that RPT has a large effect on earnings management and in Section 6 that the chair position has a more prominent effect than the CEO position, type II agency problem still plays a prominent role that drives the relationship between nonfamily leadership and earnings management.

16 This database covers more than 400 important online media outlets and more than 600 newspapers in China.

17 Compensation is regressed on firm characteristics (i.e., total revenue, ROA, stock return, leverage, and Tobin’s Q), chair characteristics (i.e., chair age, chair gender, chair education, chair tenure, and chair’s foreign experience), and macroeconomic characteristics (i.e., indicators for the region, year, and industry).

18 We examine the effect of the degree of family noninvolvement in the Online Appendix 4. The results show that earnings management increases with the number of leading positions taken by nonfamily members and with the number of nonfamily individuals acting as firm leaders, which are consistent with our previous findings and confirm that nonfamily leadership increasingly creates strong incentives for the family to conduct expropriation and earnings manipulations.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [grant number 71432008, 72102041]; the Fundamental Research Funds for the Central Universities in UIBE [grant number 20QD37, CXTD12-03, CXTD13-03].