Javier GONZALEZ-MORIN will defend his thesis on Monday 15 June 2026 at 14:00 (Auditorium 5, bâtiment TSE and also by zoom)
Title: Essays in the Theory of Incentives
Supervisor: Professor David MARTIMORT
Memberships are:
- David MARTIMORT : Professor in Economics, TSE, University Toulouse Capitole - Supervisor
- Fabrice COLLARD : Senior researcher, CNRS/TSE-R, TSE, University Toulouse Capitole - Examinateur
- Jean-Charles ROCHET : Professor in Economics, TSE, University Toulouse Capitole - Examinateur
- Malin ARVE : Professor in Economics, Norwegian School of Economics - Rapporteure
- Bentley MACLEOD : Professor in Economics, Yale University - Rapporteur
Abstract :
In this PhD thesis, I study dynamic incentives, private information, and commitment in three essays in economic theory. The common thread is that incentives, contracts, and policy promises depend not only on current outcomes, but also on how actions shape future states, future information, and future credibility. The three chapters analyze optimal contracts or policy promises when e!ort has intertemporal e!ects, when ex post signals a!ect screening, and when policymakers cannot fully commit.
In Chapter 1, Exhaustion and Burnout: a Principal-Agent Approach, I study a dynamic moral hazard problem in which e!ort increases current output but also raises the probability of future exhaustion. I model this as a continuous-time principal-agent problem with hidden action, limited liability, and a productivity state that switches between a productive regime and an exhausted one. The key insight is that exhaustion is harmful for future profits, yet informative about past e!ort, since harder work makes it more likely. The principal can therefore use protection during low-productivity spells as an incentive device. When exhaustion is more informative than success, the optimal contract relies more on payments during downturns; when success is more informative, it relies mainly on success-contingent bonuses. I also show that, when the productivity state is privately observed by the agent, the contract can still induce both e!ort and truthful revelation. The chapter thus shows that technologically harmful interruptions can nonetheless strengthen incentives.
In Chapter 2, On the (Relative) Merits of Money Burning for Optimal Contracting, coauthored with David Martimort, we study a screening problem in which an informed agent reports private information and the principal later receives an ex post private signal correlated with that information. A central contribution of the chapter is to incorporate money burning explicitly as an instrument of contractual design, alongside standard transfers. We characterize the set of incentive-feasible allocations and analyze when the first best can be implemented with transfers alone and when money burning becomes necessary. The analysis shows that the role of money burning depends on two key features of the informational environment: the timing at which the signal is learned and its quality. In particular, money burning is useful only when the signal is sufficiently informative, and its desirability di!ers sharply depending on whether the signal is observed before or after production. The chapter therefore shows that wasteful instruments may become part of the optimal contract because informational constraints make them valuable screening devices.
In Chapter 3, Forward Guidance with Limited Commitment and Reputation, coauthored with Emil Mortensen, we study monetary policy at the zero lower bound when the central bank cannot perfectly commit to future policy. In the standard New Keynesian model, forward guidance can have implausibly large e!ects on current output, giving rise to the forward guidance puzzle. We show that this puzzle is mitigated once private agents recognize that future policymakers may not wish to honor previously announced promises. We build a model with limited commitment and endogenous reputation, in which the central banker may be more or less forward-looking. In this environment, forward guidance remains useful, but its e!ects are bounded because longer promises are harder to sustain credibly. The chapter therefore provides a more disciplined theory of forward guidance by making credibility an endogenous equilibrium object.
