Pressures to mandate corporate social responsibility (CSR) actions grow every day, with stakeholders increasingly demanding that firms act in a responsible manner. However, little is known about how firms may react to such external demands for CSR, and which practices may arise as a result. In our recent paper, we examine these issues, looking at a set of technological disasters with large negative effects on the environment. The disasters are each caused by a single firm, and trigger stakeholders’ attention for CSR at the industry level. We refer to these CSR practices as demand-driven or socially enforced CSR and study their desirability.
We find increased CSR performance among firms in the affected industries after the disaster. However, we also find that this CSR is not substantive. We document that demand-driven increases in CSR performance do not improve environmental performance, nor do they reduce CSR concerns, or entail costly actions for firms. CSR increases in our setting lead to lower quality financial reporting and lower quality narrative disclosure. Overall, our evidence suggests that socially enforced CSR practices are symbolic and may lead to unfavorable outcomes, including impaired transparency. In other words, CSR appears to be used as an impression management tool (i.e., improving aspects of CSR other than the environmental performance, with the aim to gain legitimacy and sympathy from their stakeholders).
We focus our research on two technological disasters—the BP oil spill and the San Bruno Gas Pipeline Explosion—to identify plausibly exogenous increases in socially enforced/demand-driven CSR performance on firm transparency. Both disasters could have been prevented, or at least mitigated, if the affected firms had done more than just meet the minimum formal obligations imposed by law. Notably, the affected firms were sufficiently large to generate negative spillover effects (i.e., affecting all firms in the industry), in terms of reputational damage, and to spurn a call from stakeholders for CSR actions from non-responsible firms in the industry. This is because firms in the affected industry may share relations with the responsible firm (e.g., with suppliers or clients), use the same inputs, or even have a direct relation with the responsible firm. These links might mean that non-responsible firms are seen as sharing the same values that led to the disaster. More generally, external audiences predict organizations’ future behavior based on the current behavior of firms that are seen as belonging to the same category of firms (i.e., firms in the same industry).
To the extent that these spillovers are economically large for non-responsible firms and drive the attention of firm stakeholders, media, and regulators, it follows that affected firms are likely to react by implementing CSR actions. In our baseline analyses, we do observe an improvement in the CSR performance of the non-responsible firms, especially those with low pre-disaster levels of CSR. This means that, on average, firms that were more affected by the shock are those which had not endogenously developed their own CSR practices but react to social pressures for CSR by developing new CSR practices.
Next, we study whether firms react through increasing “substantive” or “symbolic” CSR. Substantive CSR involvescostly decisions that serve to resolve the specific environmental concern, such as improvements in technology or safety of operations. In turn, symbolic CSR includes actions that serve to distract stakeholders’ attention from the critical areas that need improvement (e.g., safety standards, improvements to operational procedures). Clearly, substantive action is more socially desirable than symbolic CSR.
Large technological disasters, among other circumstances that can likely generate a temporary social demand for CSR improvements, are at risk of prompting symbolic CSR. This type of CSR could potentially help distract stakeholders’ concerns without taking root in the organization and changing its ethical profile. This is because good CSR practices usually develop over time, requiring commitment across all firm echelons. Our findings confirm this hypothesis, as we find no evidence of increases in the environmental performance of non-responsible firms, or of decreases in concerns, while we do find evidence of increases in the external and strength dimensions of CSR. Further, we study whether CSR actions lead to increases in the costs of operational activities (i.e., in sales, general and administrative expenses, and research and development). Symbolic CSR should not, on aggregate, result in substantial spending since actions driven by this type of CSR typically do not entail such costs. Consistent with our prediction, we do not find evidence that these CSR improvements lead to increases in operating expenses. The result is in line with the symbolic nature of the increase in CSR that took place after the disasters.
As a final step, we study whether increases in CSR affect firms’ transparency, which we proxy by using financial statement quality. On the one hand, there is a common agreement about the link between CSR performance and corporate ethical standards in the organization, which enhances the social orientation of the firm. For instance, decision makers (e.g., CEOs, CFOs) that are more ethical will not succumb to pressures to manage earnings to conceal poor performance. On the other hand, demand-driven CSR in response to environmental disasters is likely characterized by urgency, rather than by an ethical overhaul of the firm. CSR practices can be used to gain legitimacy and then conceal poor financial reporting. In fact, symbolic CSR can be implemented together with managerial misbehavior, precisely to mask it.
Using discretionary accruals as measures of earnings management, and the Bog Index as our proxy for the quality of the narrative disclosure, we find a negative relation between CSR and transparency in the post-disaster period. This, again, is consistent with the notion that the CSR actions that emerge in response to the disasters are symbolic.
Our results add to the regulatory debate on whether CSR spending should be imposed by law. Some governments seem to be in favor of mandatory state-regulated CSR. For instance, in 2013, the Ministry of Corporate Affairs of India enacted Section 135 of the Companies Act that mandates certain financial thresholds to spend 2% of average net profits on CSR. However, the dark side of such a legal requirement could well be that firms only symbolically increase their CSR. Thereby, our results are also of relevance to investors and other groups of stakeholders, who are willing to distinguish firms with substantive versus symbolic CSR practices.
Overall, our results raise awareness that, at least in some cases, managers may implement symbolic CSR practices. For instance, when they are forced to improve CSR performance over a short period of time, they may implement practices that are not associated with real changes in firms’ ethical or safety standards and, in particular, do not drive an improvement in transparency.
Juan Manuel Garcia Lara is a Professor of Accounting at Universidad Carlos III de Madrid
Beatriz Garcia Osma is a Professor of Accounting at Universidad Carlos III de Madrid
Irina Gazizova is an Assistant Professor of Accounting at Stockholm School of Economics
This post is adapted from their paper, “Demand-driven corporate social responsibility and accounting quality: Evidence on symbolic vs substantive change after large technological disasters” available on SSRN.
