Mandatory Resignation of Politician-Directors and Company Violations: Evidence from China

ABSTRACT

This study analyzes whether the loss of political connections in Chinese listed firms affects the incidence of company violations. Using the partially observable bivariate probit model, the results indicate that the mandatory resignation of politician-directors is 1) negatively related to ex ante violation tendency and 2) positively associated with ex post violation detection. Moreover, this effect emerges as more pronounced when the politician-directors in question are senior officials, incumbent officeholders, or those whose administrative jurisdiction includes the listed companies they serve. Combined, these findings provide a new perspective from which management can effectively govern companies.

KEYWORDS:

• Politician-directors
• loss of political connection
• ex ante violation tendency
• ex post violation detection

Disclosure statement

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

Supplementary Material

Supplemental data for this article can be accessed on the publisher’s website.

Notes

1. In line with the existing literature, a politician-director is here defined as an incumbent or former leading cadre of government, or the National People’s Congress, or the CPPCC, or the public security system.

2. The statistics were obtained from “How many politically independent directors are there in 2,532 public firms in China?” in Southern Weekly, http://www.infzm.com/content/102347.

3. Zhou (2004) proposed the notion of the political tournament system. The theory holds that, under China’s centralized political system, higher-level officials mainly evaluate and promote lower level officials based on economic growth.

4. According to the Civil Servant Law of the People’s Republic of China, leadership positions are divided into national level, provincial and ministerial level, department and bureau level, county and department level, and township section level.

5. For statistics of the resignation of independent directors, please refer to Tables 1 and 2 in the supplementary document.

6. We trace back the year in which the firm committed fraud from the announcement of enforcement actions. For example, if an enforcement action in 2018 states that the firm committed fraud in 2016 and 2017, we include the firm-year observations of 2016 and 2017 in the violation sample and the rest of the firm-year observations in the non-violation sample.

7. For the results of the sample selection process, please refer to Table 3 in the supplementary document.

8. For the idea of the partially observable bivariate probit model, please refer to the supplementary document.

9. By setting control variables, we control fixed effects of individual firms and industries. Such as, industry median of Tobin Q value includes industry impact. IndLit represents the ratio of the number of companies violating regulations in the same industry to the total number of companies in the same industry in the current year and also includes the impact of the industry.

10. For the specific definition of variables, please refer to Table 4 in the supplementary document.

11. The Party School of the CPC Central Committee is a national administrative college, mainly training middle and senior civil servants.

12. For the background statistics of independent directors, please refer to Table 5 in the supplementary document.

13. For the statistical results of the company’s violations, please refer to Tables 6–8 in the supplementary document.

14. For the descriptive statistical results of variables, please refer to Table 9 in the supplementary document.

15. When choosing the matching strategy, we try the methods of nearest neighbor matching, caliper matching, kernel matching, and so on. In order to reduce the loss of sample, the one-to-four nearest neighbor matching within a 0.01 caliper is finally selected.

16. For the complete regression results of the main model, please refer to Table 10 in the supplementary document.

17. For the complete regression results, please refer to Table 11 in the supplementary document.

18. For the complete regression results, please refer to Table 12 in the supplementary document.

19. The firms with over-investment tend to violate rules in advance. The reasons are as follows: according to upper echelons theory, Heaton (2002) found that overconfident CEOs are more likely to issue optimistically and biased forecasts. They overestimate their ability to influence financial results (Deshmukh, Goel, and Howe 2013) and underestimate the probability of failure and pursue innovation (Galasso and Simcoe 2011). Once the investment fails, overconfident CEOs will be more inclined to violate the rules to cover up the loss.

20. Richardson (2006) established a model to measure enterprise investment efficiency and found that over-investment is more likely to occur in enterprises with high free cash flow. Compared with underinvestment, firms with over-investment have higher accounting performance and higher marginal return on investment measured by market value. Therefore, when the enterprise has over investment, the probability of violation being detected is low.

21. For the complete regression results, please refer to Table 13 in the supplementary document.

22. For the results of parallel trend test, please refer to Table 14 in the supplementary document.

23. In China, the eastern region includes Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Guangxi and Hainan.

24. According to statistical results, industries with violation rates of higher than 24% are Agriculture, forestry, and fishery; Timber, furniture, printing, and paper making; Petroleum, chemical, and plastics; Construction; Lodging; IT; Leasing; Education; Health; Comprehension.

25. For the results of Placebo control test, please refer to Table 15 in the supplementary document.

26. For the results of two-stage Heckman regression, please refer to Table 16 in the supplementary document.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [71702096].