In times of global ecological crisis, the responsibility of large corporations in environmental degradation is increasingly pointed out. As a result, there has been a surge in private organizations' pledges to reduce their environmental impact in recent years. In this paper, we demonstrate that companies with poor environmental responsibility have incentives to take such pledges to maintain their ability to attract high-skilled human capital. Through a case study on a French climate movement which was initiated by elite students who threatened to boycott job offers from polluting employers, we find that environmental pledges can significantly attenuate this selection effect. Using a unique and large survey database on the climate movement participants (n=2307) and machine learning classifiers, we find that individuals who initially intended to refuse a job offer from a polluting employer were, on average, three times less likely to hold such intentions after being exposed to a corporate environmental pledge. This result can be explained by the fact that intentions to refuse to work for polluting companies, and reactions to environmental pledges are driven by different factors. Furthermore, we find substantial heterogeneity in the response to environmental pledges, which is primarily explained by career perspectives, beliefs about the ecological crisis and support for radical political action in the name of ecology.

Sustainable investing is growing fast and investors are increasingly integrating environmental, social, and governance (ESG) criteria. However, ESG ratings are derived using heterogeneous methodologies and can be quite divergent across providers, which suggests the need for a formal statistical procedure to evaluate their accuracy. This paper develops a backtesting procedure that evaluates how well these extra-financial metrics help in predicting a company's idiosyncratic risk. Technically, the inference is based on extending the conditional predictive ability test of Giacomini and White (2006) to a panel data setting. We apply our methodology to the forecasting of stock returns idiosyncratic volatility and compare two ESG rating systems from Sustainalytics and Asset4 across three investment universes (Europe, North America, and the Asia-Pacific region). The results show that the null hypothesis of no informational content in ESG ratings is strongly rejected in Europe, whereas results appear mixed in the other regions. Furthermore, the predictive accuracy gains are higher when considering the environmental dimension of ESG ratings. Importantly, applying the test only to firms over which there is a high degree of consensus between the ESG rating agencies leads to higher predictive accuracy gains for all three universes. Beyond providing insights into the accuracy of each of the ESG rating systems, this last result suggests that information gathered from several ESG rating providers should be cross-checked before ESG is integrated into investment processes.