Essays in labour and behavioural economics
The entire literature on adverse selection in the labour market spawned by Greenwald (1986, Review of Economic Studies, 63(3)) has been built, somewhat unwittingly, on the assumption that firms forget the type of a worker after the worker quits. In many contexts, this assumption is implausible. The first three chapters of this thesis therefore explore an alternative approach to modelling labour markets with asymmetric information by assuming firms will never forget a worker's type. The first chapter turns the standard Greenwald result on its head by showing that if the worker knows her own type and productivity is unchanging, the possibility of competitive wage offers from fully-informed previous employers means that adverse selection will never persist. Job changing frictions can cause a semi-separating equilibrium where the more productive workers have their type revealed whilst the least productive workers receive a pooling payoff. But even where asymmetric information persists there is no adverse selection because job changing frictions shield potential employers from the winner's curse. The second chapter investigates the robustness of the non-persistence of adverse selection result where previous employers are asymmetrically informed. The result is found to be robust where firms bid for the worker under a closed but not an open auction. The third chapter finds that, if workers are not sure of their exact value to their employer, there will be an adversely selected stream of job changers in equilibrium, even as the probability of a worker quitting for exogenous reasons approaches zero. Less able workers are quickly revealed as such, whilst more able workers have their type revealed gradually. The fourth substantive chapter of this thesis investigates the widely observed paradox that, despite what traditional economics would lead us to believe, there can be such a thing as too much choice. The model provides a formal theoretical explanation for this phenomenon using the regret theory of Loomes and Sugden (1982, Economic Journal, 92(368)). When options are few it is shown that enlarging the choice set improves welfare, but when options are many, a "less is more" phenomenon emerges. In some cases, excess search options can decrease search.