Pier-André Bouchard St-Amant

We conduct an optimal taxation analysis on a population of heterogenous individuals whose perceptions of savings' are linked to their contemporaneous disposable income and that of their parents. Individuals coming from poorer families see savings as less important than those from richer families. Perception of savings thus evolve because of cultural transmission through families, but also because of their current productivity. Aside from the standard equity/efficiency trade-off, optimal taxes account for this cultural transmission of perceptions through two additional channels. First, taxing labor decreases income, which decreases the perception of savings through time, generating a behavioural inefficiency. Second, taxation on savings corrects for the (ill)-perceived value of savings and thus savings' and labor decisions. Numerical simulations suggests the second effect is more important. An increase of 600 basis points on labour tax rates and a subsidy on savings of an order of 30% is shown to increase welfare.

We characterize an optimal redistributive pension scheme when individuals face temptation, but can exert costly self-control (as in Gul & Pesendorfer, 2001; 2004). Our results challenge the common wisdom that forced savings tend to reduce individuals' mental cost of self-control. In our model, individuals heterogenous in the intensity of their self-control problem and in productivity. A pension scheme simultaneously has three effects: it redistributes income across retirees, it forces young individuals to save and it affects their mental cost of exerting self-control, which is entailed by voluntary savings. In a scheme with a proportional income tax and benefits with a contributory component, taxation generally increases the mental cost of self-control. This effect partially offsets the forced-saving benefits of public pensions, and call for a more redistributive system. This result contrasts with a first best optimum, where the mental cost of self-control is driven to zero and consumption is smoothed across periods.

Governments tend to fund universities based on the number of registered students and lump sum transfers. Such policy induces universities to compete for recruiting students, diverting money away from funds devoted to teaching. I show this degradation of quality increases with the degree to which universities compete for the same students. I suggest a policy that achieves the social optimum for any given level of funding. This policy shows that a university's funding rules should not be solely based on its enrollment, but also on the enrollment in competing universities.

I develop a theory that answers the two following questions: what is the best word of mouth campaign that a principal can perform on a social network and second, what is the value of such campaign? I answer these questions for a wide range of behaviors of agents on the network, including the case where agents choose to whom they listen to. This last case is shown to reproduce patterns that have been observed in empirical studies. Such theory allows to price companies that own social networks and it also suggest that such companies have incentives to push a network out of an equilibrium where agents agree on the value of a signal. It also yields an optimal solution computable either analytically or in polynomial time for any given type of network.

La diffusion de ce mémoire se fait dans le respect des droits de son auteur, qui a signé le formulaire Autorisation de reproduire et de diffuser un travail de recherche de cycles supérieurs (SDU-522 - Rév.01-2006). Cette autorisation stipule que «conformément à l'article 11 du Règlement ...