Mandatory Corporate Social Responsibility (CSR) Reporting and Financial Reporting Quality: Evidence from a Quasi-Natural Experiment

Abstract

This study examines the impact of mandatory Corporate Social Responsibility (CSR) reporting on firms’ financial reporting quality using a quasi-natural experiment in China that mandates a subset of firms to report their CSR activities starting in 2008. We find that mandatory CSR disclosure firms constrain earnings management after the policy. The result is robust to a battery of sensitivity tests and more prominent for firms with lower analyst coverage. Further analyses reveal that upward earnings management by mandatory disclosure firms is more likely to be caught after the policy. The findings suggest that mandatory CSR disclosure mitigates information asymmetry by improving financial reporting quality.

Appendix: Estimation of the Magnitude of Earnings Management

Performance-Adjusted Discretionary Accrual Measure

The performance-adjusted discretionary accrual measure is derived from the modified Jones (1991) model as follows (Kothari et al. 2005):

$${{TA_{i,t} } \mathord{\left/ {\vphantom {{TA_{i,t} } {A_{i,t - 1} = \phi_{1} \left( {{1 \mathord{\left/ {\vphantom {1 {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right) + \phi_{2} \left[ {{{\left( {\Delta REV_{i,t} -\Delta REC_{i,t} } \right)} \mathord{\left/ {\vphantom {{\left( {\Delta REV_{i,t} -\Delta REC_{i,t} } \right)} {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right] + \phi_{3} \left( {{{PPE_{i,t} } \mathord{\left/ {\vphantom {{PPE_{i,t} } {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right) + \phi_{4} ROA_{i,t} + \varepsilon_{i,t} ,}}} \right. \kern-0pt} {A_{i,t - 1} = \phi_{1} \left( {{1 \mathord{\left/ {\vphantom {1 {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right) + \phi_{2} \left[ {{{\left( {\Delta REV_{i,t} -\Delta REC_{i,t} } \right)} \mathord{\left/ {\vphantom {{\left( {\Delta REV_{i,t} -\Delta REC_{i,t} } \right)} {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right] + \phi_{3} \left( {{{PPE_{i,t} } \mathord{\left/ {\vphantom {{PPE_{i,t} } {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right) + \phi_{4} ROA_{i,t} + \varepsilon_{i,t} ,}}$$

(3)

where TA _i,t denotes the total accruals calculated as earnings less cash flows from operations for company i at year t; A _i,t−1 is the total assets of company i at year t − 1; ∆REV _i,t is the change in sales revenue; ∆REC _i,t is the change in accounting receivables; PPE _i,t represents property, plant, and equipment; ROA _i,t is net income divided by the total assets at year t − 1; and ε _i,t denotes the residual term. The model (A1) is estimated cross-sectionally by each industry-year group. We require each industry-year group to have at least 10 observations to ensure reliable estimation. The fitted values from Model (A1) are the normal accruals (or nondiscretionary accruals) that arise from companies’ normal operating activities, and the residual term (DA1) denotes the performance-adjusted discretionary accruals that are assumed to be opportunistically chosen by management.

Performance- and Growth-Adjusted Discretionary Accrual Measure

We first estimate the following model for each year-industry group (Raman and Shahrur 2008):

$${{TA_{i,t} } \mathord{\left/ {\vphantom {{TA_{i,t} } {A_{i,t - 1} = \alpha_{1} \left( {{1 \mathord{\left/ {\vphantom {1 {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right) + \alpha_{2} {{\Delta REV_{i,t} } \mathord{\left/ {\vphantom {{\Delta REV_{i,t} } {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }} + \alpha_{3} }}} \right. \kern-0pt} {A_{i,t - 1} = \alpha_{1} \left( {{1 \mathord{\left/ {\vphantom {1 {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right) + \alpha_{2} {{\Delta REV_{i,t} } \mathord{\left/ {\vphantom {{\Delta REV_{i,t} } {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }} + \alpha_{3} }}\left( {{{PPE_{i,t} } \mathord{\left/ {\vphantom {{PPE_{i,t} } {A_{i,t - 1} }}} \right. \kern-0pt} {A_{i,t - 1} }}} \right) + \alpha_{4} ROA_{i,t} + \alpha_{5} BM_{i,t} + \varepsilon_{i,t} ,$$

(4)

where BM _i,t is the ratio of book value to market value of equity. Other variables are defined as in Model (A1). The residual term (DA2) denotes the performance- and growth-adjusted discretionary accruals that are assumed to be opportunistically chosen by management.