The incentive properties of the Matthew Effect in the academic competition.

This paper is concerned with the incentive properties of the Matthew Effect by which since Merton [1968] one is usually describing the various cumulative advantages that obviously affect academic competition. We introduce a model of sequential contests in which the agents that have initially produced more are the ones that will be further advantaged in that they are benefiting from intrinsically more productive research positions. We principally show that there is an optimal level of the Matthew effect and that this optimal dynamic bias is increasing with the risk of research activity while it is decreasing with the initial inequalities.