In the third phase of the EU ETS, sectors that are classified as carbon leakage risky received all allowances for free up to a specific benchmark. In contrast, not at-risk sectors had to purchase an increasing share of allowances by auction. I exploit the allocation rule based on a sharp cutoff value and historical data to estimate whether free allowances successfully addressed carbon leakage for sectors at the margin. Due to a time lag in the information publication and its implementation, it is possible to investigate whether firms reacted in anticipation of the policy and/or after its actual implementation. I find that sectors that marginally received free allowances experienced similar changes in trading patterns to sectors that purchased allowances by auction both after the publication and implementation of the policy. Thus, I do not find evidence in favor of carbon leakage for marginally eligible sectors. Firms increasingly publish information about their sustainability in annual reports and on products. However, it is unclear which information induces pro-environmental behavior most effectively. In an experiment, we compare the effect of carbon display in kilograms, abatement costs, and social costs, respectively, on individuals' purchasing decisions. Contrasting previous literature, we find that the type of display has no significant impact on purchasing decisions. Nevertheless, most participants believe that social cost information leads to the largest carbon reduction by consumers, and many prefer this information. Our results have implications for managers and policymakers interested in setting standards for CO2 information. Expectations are a key element of strategic environments, accordingly in all competitive situations. As it has already been shown that men's performance in current competitions is affected by the strength of expected future opponents, we investigate whether women are overshadowed by future competitors as well. In doing so, we use data from professional tennis, replicate the results for men and compare it with women's behavior. Extending previous research, we focus on individuals that have already entered a competition and investigate the role of uncertainty in this context, particularly, whether behavior differs and is gender-specific when individuals have to form expectations instead of having accurate information about future competitors. Our findings suggest that highly qualified men and women perform worse in current competitions in case that the expected future opponent is stronger. Moreover, we find that the shadow effect is remarkably weaker if uncertainty about future opponent's ability is eliminated, which tends to be even more pronounced for women. Finally, our results suggest that women (men) rather seem to be sensitive to direct (more distant) incentives. These findings might be transferred to information revelation in promotion contests in order to help highly qualified individuals, especially women, climb the career ladder.