21 september: Luc Bridet thesis

• Jacques CREMER, TSE Researcher, University Toulouse Capitole
• Thomas MARIOTTI, TSE Researcher, University Toulouse Capitole
• Jan EECKHOUT, Researcher, University Barcelone Pompeu Fabra
• Paul HEIDHUES, Researcher, ESMT Berlin

Chapter 1 models persuasion by an informed party and show that laissez-faire leads to excessive persuasion expenditures and delays decisions. An uninformed regulator can use pigouvian taxation to reduce the delay in decision-making and the amount of persuasion expenditures incurred by advocates, with no corresponding decrease in the quality of decisions eventually taken. 

Chapter 2, joint with Margaret Leighton, is an empirical study of individual learning and specialisation decision by American college students. We use detailed transcript information in the /Baccalaureate and Beyond/ dataset to construct and explain the timing of specialisation decided by each student. By relating its cross-sectional variation to later job market outcomes, we quantify skill transferability, the cost of career changes and most importantly, the informational benefits of a delayed specialisation. We then use these structural estimates to compare the current college system to one which imposes specialization at college entry. Overall, expected earnings fall by 1.5%. 

Chapter 3, joint with Peter Schwardmann, studies how the marketplace for financial loans interacts with entrepreneurs' cognition. We show that in a market for project financing dominated by informational frictions, market outcomes reward a particular behavioural deviation from standard preferences: a tendency towards optimistic self-deception. This result stands in stark contrast to those obtained under neoclassical competition. We also show that entrepreneurs' beliefs are subject to some market discipline: as long as anticipatory concerns are limited, entrepreneurs are induced to appraise their projects realistically.