Page 207 - 臺大管理論叢第32卷第2期

NTU Management Review Vol. 32 No. 2 Aug. 2022

on the Taiwan Stock Exchange (TWSE) and the Taipei Exchange Market (TPEx). Our
empirical evidence helps the public increase the comprehension of whether DOLI can help
catch a firm’s attention or conversely, divert the firm’s attention away from the overall
risks among its stakeholders, thereby providing more insights into the consequences of

DOLI coverage.

2. Design/Methodology/Approach

Based on the 2008~2018 lists of winners of “Top Corporate Citizen” and “CSR
Awards,” two annual CSR evaluation of publicly-traded firms conducted by the Common
Wealth and the Global Views respectively, we establish four discrete CSR performance
measures. In addition, referring to the inclusion criteria of the Shanghai Stock Exchange

Social Responsibility Index, this research computes each firm’s social contribution value,
social returns on assets and social contribution value per share to act as continuous
measures for CSR performance. To quantify the degree of DOLI coverage, we employ
six measures, including a dummy variable of whether a firm has DOLI coverage, the

insurance amount of DOLI, the average insurance amount of DOLI per insured, the ratio
of insurance amount of DOLI to total assets, the ratio of insurance amount of DOLI to
total equity and the ratio of insurance amount of DOLI to total sales.
This study examines the degree of DOLI coverage affects a firm’s CSR performance

by employing descriptive statistics, correlation analysis and multiple regression estimation.
Empirical evidence from baseline regression (pooled OLS) estimation shows that the
degree of DOLI coverage and CSR performance is positively correlated, meaning that
firms attach greater importance on directors and the management litigation risk reduction

tend to engage more in maintenance of stakeholders’ interests, supporting the view of
complementary relationship between DOLI and CSR. This study also conducts a number
of additional tests, mainly including the analysis of the impacts of DOLI coverage on
various aspects of a firm’s CSR performance, the fixed-effect estimation of regression

equations, two-stage least squares instrumental variable estimation to control for the
endogeneity problem (Angrist and Krueger, 2001), two-stage treatment effect estimation
and propensity score matching to correct for sample self-selection bias (Heckman, 1979;

199