Page 233 - AJWEP-22-4

CEPI & ESG greenwashing: Exec. attention view

time, which is coordinated and arranged by the central correlation analysis was conducted using Stata 18
government. Characterized by a high level of authority (StataCorp, United States). A p<0.1 was considered
and strong enforcement, these inspections are less statistically significant.
susceptible to external factors and thus represent a form
of effective policy impact. Therefore, the DID model 5. Empirical result analysis
12
was employed in this study, the control variables were
added, and the time and industry information were fixed Table 2 presents the descriptive statistics for each
to test the proposed hypotheses. To verify Hypothesis variable. The average ESG Gws value of firms was
1a, the equation for the specific model is as follows: −0.444 (standard deviation = 1.231), with the lowest
Gws , i t α = 0 α + 1 Cepi i t , ∑ j α + j Control j ,,i t + t +u i ε + v , i t value being −3.250 and the highest being 2.866. This
suggests that the level of ESG Gws varies greatly
(II) among firms. In addition, the negative average ESG
Given that corporate ESG performance information Gws value indicates a low level of Gws. The descriptive
is generally disclosed in the subsequent year and policy statistics of the other variables were in line with
14
interventions often exhibit lagged effects, this study previous comparable studies. The mean value of the
used Equation Ⅲ to verify Hypothesis 1b, where the variable Cepi was 0.244, based on 1,291 observations
CEPI variable is lagged by one period: in the experimental group and 3,998 observations in the
control group.
α = α +Gws . L Cepi , ∑ α + Control The correlation analysis results revealed that the
, i t 0 1 i t j j j ,,i t
+ t +u i +v ε , i t (III) correlation coefficients among all variables were <0.8,
suggesting the absence of severe multicollinearity issues
between variables. The results are presented in Table 3.
4.3.2. Mechanism regression
Based on the DID benchmark regression models 5.1. Basic regression analysis
(Equations Ⅱ and Ⅲ), the moderating variable and The results of the benchmark regression are presented in
its interaction with the independent variable were Table 4. The regression results for CEPI were significant
introduced to examine the moderating effect of Fs on regardless of the presence of control variables.
the relationship between CEPI and corporate ESG Gws, Specifically, without considering lagged effects, the
thereby testing Hypothesis 2. This model is expressed coefficient was positive and significant, indicating that
as Equations Ⅳ and Ⅴ: the arrival of CEPI in the current year promotes corporate
Gws , i t α = 0 α + 1 , i t α +Cepi 2 Fs it α + 3 Cepi , i t ESG Gws behaviors. With lagged effects considered,
× it + Fs ∑ j α j j ,,i t + t +u i + v ε Control , i t (IV) the coefficient was negative and significant, suggesting
a reduction in ESG Gws behaviors in the subsequent
α = Gws α + . L Cepi α + Fs α + Cepi year. The potential reasons for this phenomenon are as
, i t 0 1 , i t 2 it 3 , i t follows: first, ESG information disclosure is subject
× it + Fs ∑ j α j j ,,i t + t +u i + v ε Control , i t (V) to a time lag. In the year when inspectors arrive, the
where the subscripts i and t represent enterprises and disclosed ESG data largely reflects the previous year’s
performance, which may remain at a relatively high
years; Gws is the dependent variable, representing the level. Second, upon the arrival of inspectors, enterprises
i,t
degree of ESG Gws; Cepi is the independent variable, may adopt coping strategies by exaggerating their
i,t
representing the resident inspector of CEPI in the environmental protection behaviors to reduce the
current year; L.Cepi is the subsequent year of a CEPI likelihood of scrutiny. This can lead to a “discrepancy
i,t
inspection; is the Fs moderating variable, representing between commitments and actions,” thereby resulting
Fs; Cepi × Fs is the interaction term of the moderator in a higher level of Gws. Third, policy interventions
i,t
it
and independent variable; ∑ j α Control j ,,i t is the set often have a lagged effect. When inspectors arrive,
j
of control variables; u is the fixed year; v is the fixed enterprises may initiate green rectification measures,
t
i
industry; and ε is the random disturbance term. but the effects may not be immediately manifested. As
a result, the suppressive impact on corporate ESG Gws
4.4. Correlation analysis tends to be more observable in the following year, when
In this study, Pearson correlation analysis was employed firms begin to implement genuine compliance efforts in
to evaluate the relationships between variables. The recognition of the policy’s intensity.

Volume 22 Issue 4 (2025) 225 doi: 10.36922/AJWEP025280219