Thesis: This doctoral thesis focuses on two distinct themes: decision-making in situations of risk and economic efficiency. These two questions are treated in an empirical framework using unique databases. The first chapter tests the theoretical models of decision under risk on bets on horseraces to identify the theory that allows the best description of individuals’ choices. This chapter shows that behavioral theories of decision under risk (cumulative prospect theory and rank-dependent utility) are better suited to explain the behavior of bettors observed in the data than the expected utility theory. The second chapter estimates the effect of a merger between two large companies in the urban transport sector in France on their costs. The results show that the merger did not result in efficiency gains for the merging parties. This result can be understood in light of the context in which the merger was conducted, in particular the reasons for the choice of the target, the perception of customers and employees of the merger, the differences in culture between the companies and the operational preparation of the merger. The third chapter examines the impact of greater discretion in tendering procedures on the selection of operators, particularly their productivity. The main result is that the procedure that potentially increases the discretion of public purchasers leads to the selection of less efficient companies than a procedure involving little or no discretion. Further analyzes suggest that the result is due to the effect of greater discretion at the screening stage.