Academic Independent Director Abnormal Resignations and R&D Investment: Evidence from China

ABSTRACT

We examine the effects of academic independent directors abnormal resignations on a firm’s R&D investment. We manually collect the resignations data of 3,964 academic independent directors (IDAs), and find that the abnormal resignations of IDAs reduce firms’ R&D investment. The results remain significant after employing a series of endogenous checks and robustness tests. Furthermore, we find that the lack of advising and resourcing channels caused by abnormal resignations are the primary mechanisms for the decline of the R&D investment. Our main finding is more pronounced for non-state-owned enterprises, firms with numerous independent directors, and firms with high profitability and low degrees of financial constraints. We contribute to the literature by proving that the abnormal resignations of IDAs not only decrease firms’ R&D investment but also further reduce the innovation quality and firm value over the long run. Overall, this paper supports that IDAs have positive contributions to R&D investment and firms’ value.

KEYWORDS:

• Academic independent directors
• abnormal resignation
• R&D investment
• China

JEL:

• G30
• G34
• G38

Disclosure Statement

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

Notes

1. Previous literature explores the impacts of executives’, directors’, and shareholders’ characteristics on R&D investment, such as CEO personality traits (Hirshleifer, Low, and Teoh 2012; Sunder et al., 2017), inside directors (Shaikh, O’Brien, and Peters 2018), board gender diversity (Almor, Bazel-Shoham, and Lee 2019; Harjoto, Laksmana, and Yang 2018), and institutional investors (Gao et al. 2019). The impact of independent directors on R&D investment remains controversial (Dong and Gou 2010; Hill and Snell 1988).

2. Regulation 11 stipulates that university lecturers who hold certain administrative positions (faculty deputy-dean or above, including faculty dean, university vice-president, university president) are prohibited from holding directorships in listed firms.

3. We manually record the reasons for IDAs’ resignations. For normal resignations, firms usually have made preparations and selected replacements, which makes it difficult to observe the influence of IDAs’ resignations. Therefore, through screening, we choose only the IDAs with abnormal resignations as the research objects. The specific screening method is shown below.

4. Prior literature presents severe endogeneity problems because the directors’ nomination and appointment are endogenously determined (Harris and Raviv 2008; Wintoki, Linck, and Netter 2012). Adopting a dummy variable that indicates whether firms have at least one academy fellow independent director (AFID), Li et al. (2022) find that AFIDs are positively related to corporate innovation only in non-state-owned enterprises. However, Li et al.’s (2022) study does not address the endogeneity problems in the main findings. Wang (2020) design two quasi-natural experiments – independent directors’ six-year tenure expiration and sudden departures to verify the promotion effects of AIDs on innovation. However, Wang’s (2020) study has two shortfalls. First, the independent directors’ six-year tenure expiration is expected, which introduce the endogeneity issue to the findings. Second, the sample for sudden departures of independent directors is a limited sample, which may have sample bias. Compared with the predictability of tenure expiration, the promulgation of Regulation 11 in this paper is unpredictable, exogenous, and more suitable to employ the quasi-natural experiments. In addition, our study has a more comprehensive sample size compared to Wang’s (2020) study.

5. Since listed firms began to publish data regarding their R&D investment and the information regarding independent director resignations formally and continuously from 2009.

6. For normal resignations, firms usually have made preparations and selected replacements, which makes it difficult to observe the role of IDAs.

7. To eliminate the possible influence of scale factors, we divide R&D investment by total assets and then multiply by 100.

8. For example, result in Column (4) shows that an increase in sudden AID resignations from the 25^th percentile (0) to 75^th percentile (1) of its distribution is associated with a 9.70% (−0.194*(1–0)/2) decrease in the firm’s R&D Intensity in the following year.

9. Resign_D equals 1 if there is at least one IDA resignation for firm i in year t, 0 otherwise.

10. A top-tier university is a university that belongs to China’s “National 985 Project” and “National 211 Project”.

11. We also employ a firm’s R&D Expenditure as the dependent variable and replicate all tests in Table 8. Results are qualitatively the same as the results reported in Table 8. We do not report these results due to the space constraints. However, results are available upon request.

12. We also use a firm’s ROE to replace the Tobin’s Q to measure the firm’s long-term performance. Results are quantitatively as same as what we report above. We do not report the result due to the space constraint, and the result is available upon to request.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [72073101].