Do Auditors Applaud Corporate Environmental Performance? Evidence from China

(1) The Notice of Supervising the Listed Firms in Shanghai Stock Exchange to Disclose the Annual Report of Year 2008 (SHSE 2008a); (2) The Notice of Supervising the Listed Firms in Shenzhen Stock Exchange to Disclose the Annual Report of Year 2008 (SZSE 2008); and (3) The Guide to Environmental Information Disclosure for Listed Firms in Shanghai Stock Exchange (SHSE 2008b).

Chinese CPA Auditing Standards, No.1211: Understanding the Entity and Its Environment and Assessing the Risks of Material Misstatement: http://​kjs.​mof.​gov.​cn/​zhuantilanmu/​kuaijizhuanzeshi​shi/​zyzz/​200806/​t20080618_​46262.​html.

See Chinese CPA Auditing Standards No.1502: Non-standard Auditing Report (http://​kjs.​mof.​gov.​cn/​zhuantilanmu/​kuaijizhuanzeshi​shi/​zyzz/​200806/​t20080618_​46296.​html).

In China, unqualified opinions with explanatory notes are treated as a form of modified audit opinions because auditors often issue this type of audit report in lieu of a qualified opinion (Gul et al. 2013). Chen et al. (2000) document a negative market reaction to unqualified opinions with explanatory notes in Chinese context.

Our sample covers all firms including firms whose environmental performance scores equal 0, which is less likely to bring the sample selection bias. Moreover, we further conduct additional tests (Tables 9 and 10) after (1) differentiating firms with environmental information disclosure in separate CSR reports from those with environmental information disclosure in notes to financial reports and (2) deleting firms whose corporate environmental performance scores equal 0.

Delmas and Cuerel Burbano (2011) identify four cases according to both a firm’s fulfilling environmental responsibility and its disclosure about environmental performance: (1) a firm fulfills environmental responsibility better and fairly discloses its environmental performance (a vocal green firm); (2) a firm fulfills environmental responsibility better but insufficiently reports its environmental performance (a silent green firm); (3) a firm neither fulfills environmental responsibility better nor fairly discloses its environmental performance (a silent brown firm); and (4) a firm does not fulfill environmental responsibility but reports better environmental performance (a greenwashing firm).

Here are the websites: (1) Shanghai stock exchange: http://​www.​sse.​com.​cn/​disclosure/​listedinfo/​announcement and (2) Shenzhen stock exchange: http://​disclosure.​szse.​cn/​m/​drgg.​htm.

The marginal effect is calculated based on the STATA command of “mfx.” In addition, we can obtain qualitatively similar results if we employ the approach in Wooldridge (2009) to calculate marginal effect.

DeFond et al. (2002) investigate the influence of total fees on going concern opinions and report that the marginal effects range from 0.0 % to 0.6 %. We express our greatest thanks to the referee for their suggestion on the marginal effects.

We express our greatest thanks to the referee for his/her constructive comment about this problem.

In Columns (1)–(3), it can be observed that the statistical software does not return the coefficients of BIG4, DUAL, and STATE. The reasons may be that, for firms with environmental disclosure in separate CSR reports: (1) only 11 (1.22 %) firm-years are audited by international big 4 auditors (BIG4); (2) the chairman serves as the CEO simultaneously only in 22 (2.45 %) firm-years; and (3) about 34 (3.79 %) firm-years are state-owned enterprises.

In Column (1) of Table 11, the statistical software does not return the coefficients of MAO_LAG and BIG4 for the following reasons: (1) all 130 firms (100 %) received modified audit opinions in year t − 1 (i.e., MAO_LAG = 1) and (2) only two firms (1.54 %) are audited by BIG4 auditors.

We also conduct t tests to examine the differences in audit opinion shopping across corporate environmental performances. First, we divide the full sample into two subsamples based on the sample mean: (1) the subsample with high environmental performance and (2) the subsample with low environmental performance. Untabulated results show that, compared with the subsample with high environmental performance, the subsample with low environmental performance has a significantly higher likelihood of audit opinion shopping (t = 3.10). Second, we divide the subsample with high environmental performance into two second-tier subsamples based on the subsample mean, but untabulated results show that the difference in the likelihood of audit opinion shopping between the two second-tier subsamples is insignificant (t = 1.17). Third, similarly, we further partition the subsample with low environmental performance into two second-tier subsamples according to the subsample mean, and untabulated results show that the second-tier subsample with low environmental performance has a significantly higher probability of audit opinion shopping than the second-tier subsample with high environmental performance (t = 3.59). Above findings, taken together, suggest that firms with worse environmental performance are more likely to conduct audit opinion shopping.

We also calculate five additional variables to measure earnings management or accrual quality. DA_ΔCF denotes discretionary accruals based on augmented Jones model of Ball and Shivakumar (2006) considering the change of operating cash flow from operation. DA_RET refers to discretionary accruals based on augmented Jones model of Ball and Shivakumar (2006) considering abnormal returns. Following Dechow and Dichev (2002) and Francis et al. (2005), we compute three variables about accruals quality: AQ ₁₀ , AQ ₁₅ , and AQ ₂₀ . Similarly, untabulated results show that the coefficients on DA_ΔCF, DA_RET, AQ ₁₀ , AQ ₁₅ , and AQ ₂₀ are all negative and significant (−0.0007 with t = −1.82, −0.0010 with t = −2.37, −0.0008 with t = −3.00, −0.0008 with t = −3.05, and -0.0008 with t = −3.19, respectively).