Organizing contests for status: The Matthew Effect versus the Mark Effect

What is the best way to design tournaments for status, where individuals labor primarily for the esteem of their peers? We propose a formal model of status-based competition that contrasts two competing alternatives. The first, following Merton (1968), is the "Matthew Effect," according to which a tournament's architect directs excess resources to elite actors, and thus widens the distribution of rewards by favoring cumulative advantage. The second is the "Mark Effect," under which a tournament's designer instead pushes excess resources to marginal actors, and thus tightens the distribution of rewards through engaging in redistribution. Our results indicate that the Mark Effect is better for the social welfare of most tournaments, but that the Matthew Effect is preferable for elite tournaments, such as top R&D labs, that rank highly against other tournaments. Our main contributions are twofold: We find that, counter to the thrust of Merton's (1968) work, cumulative advantage is not invariably optimal for the functioning of status contests; and we identify circumstances in which the production of superstars is likely to make contests for status better off in aggregate. An empirical extension using panel data on levels of selectivity and faculty salary dispersion of colleges and universities in California is presented.