The asymmetric relationship between corporate environmental responsibility and earnings management: Evidence from the United States

Abstract

Purpose

The purpose of this paper is to examine the relationships between earnings management (EM) and subsamples of corporate environmental responsibility (CER).

Design/methodology/approach

KLD data are used to generate subsamples of environmental “strengths” and “concerns”. Differences in EM are studied across subsamples, using discretionary accruals to proxy for EM. The samples consist of 2,171 US firms.

Findings

Firms with at least one environmental strength do not exhibit statistically different levels of EM, relative to environmentally neutral firms, while firms with at least one environmental concern do exhibit statistically greater EM (greater income‐increasing discretionary accruals), relative to other sample firms. Further, firms with multiple environmental concerns exhibit greater EM than firms with a single environmental concern. These findings do not support the political cost hypothesis per the CER/EM literature, but they do support the institutional hypothesis (in the case of environmental strengths) and myopia avoidance hypothesis (in the case of environmental concerns) per the broader corporate social responsibility (CSR)/EM literature.

Research limitations/implications

The findings are limited to US firms; results may not be transferable to other countries. KLD data are binary, and thus may not capture the full array of CER.

Practical implications

The findings may aid interested parties in detecting EM.

Originality/value

The paper provides a new testing environment for theoretical frameworks established in the CER/EM and CSR/EM literatures. Additionally, the findings differ across subsamples, suggesting that the relationship between CER and EM is asymmetric.

Citation

Heltzer, W. (2011), "The asymmetric relationship between corporate environmental responsibility and earnings management: Evidence from the United States", Managerial Auditing Journal, Vol. 26 No. 1, pp. 65-88. https://doi.org/10.1108/02686901111090844