Corporate Social Responsibility Reporting in China: Symbol or Substance?
City University of Hong Kong
April 26, 2013
Organization Science, Forthcoming
Harvard Business School Research Paper
This study focuses on how and why firms strategically respond to government signals regarding appropriate corporate activity. We integrate institutional theory and research on corporate political strategy to develop a political dependence model that explains (a) how different types of dependency on the government lead firms to issue corporate social responsibility (CSR) reports and (b) how the risk of governmental monitoring affects the extent to which CSR reports are symbolic or substantive. First, we examine how firm characteristics reflecting dependence on the government — including private versus state ownership, executives serving on political councils, political legacy, and financial resources — affect the likelihood of firms issuing CSR reports. Second, we focus on the symbolic nature of CSR reporting and how variance in the risk of government monitoring through channels such as bureaucratic embeddedness and local government institutional development influences the extent to which CSR communications are symbolically decoupled from substantive CSR activities. Our database includes all CSR reports issued by the approximately 1,600 publicly listed Chinese firms between 2006 and 2009. Our hypotheses are generally supported. The political perspective we develop contributes to organizational theory by showing (a) the importance of government signaling as a mechanism of political influence, (b) how different types of dependency on the government expose firms to different types of legitimacy pressures, and (c) that firms face a decoupling risk which leads them to be more likely to enact substantive actions in situations where they are likely to be monitored.
Number of Pages in PDF File: 48
Date posted: February 12, 2013 ; Last revised: May 4, 2013
© 2016 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollobot1 in 2.484 seconds