| Popis: |
Purpose This study aims to examine the extent to which corporate governance structures and ownership types are associated with the level of corporate social responsibility disclosures (CSRD) in a developing country. Design/methodology/approach Multiple regression techniques are used to estimate the effect of corporate governance structures and ownership types on CSRD using a sample of Libyan oil and gas companies between 2009 and 2013. Findings First, the study results suggest that although the level of CSRD in Libya is low in comparison to its western counterparts, ownership factors have a significant positive influence on CSRD. Second, the authors find board meetings to have a positive impact on CSRD. However, the authors fail to find any significant effect of board size and presence of corporate social responsibility (CSR) committees on CSRD. Overall, the results support prior theoretical evidence that pressures exerted by the government and external stakeholders have a considerable influence in promoting firm-level CSRD activities, specifically as a legitimising mechanism in fragile states. Research limitations/implications First, this study is based on the annual reports, and it did not examine any other reports or other mass communication mechanism that companies’ management may use to disclose CSR information. Future studies might consider disclosures in other channels, if any, such as the internet, CSR reports, etc. Additionally, this study adopts the neo-institutional theory perspective. Future studies might integrate multi-theoretical lenses to offer a richer basis for understanding and explaining CSRD determinants. Originality/value This study contributes to the literature by first providing additional evidence for existing studies, which suggest that on average, better-governed companies are more liable to follow a more socially responsible agenda than poorly governed companies as a legitimising mechanism in fragile states. Also, this study overcomes a major weakness in existing Libyan studies, which have mainly used descriptive data. |
