We study a contest with multiple, nonidentical prizes. Participants are privately informed about a parameter (ability) affecting their costs of effort. The contestant with the highest effort wins the first prize, the contestant with the second-highest effort wins the second prize, and so on until all the prizes are allocated. The contest's designer maximizes expected effort. When cost functions are linear or concave in effort, it is optimal to allocate the entire prize sum to a single “first” prize. When cost functions are convex, several positive prizes may be optimal.

We study two-sided markets with a finite number of agents on each side, and with two-sided incomplete information. Agents are matched assortatively on the basis of costly signals. Asymmetries in signalling activity between the two sides of the market can be explained by asymmetries either in size or in heterogeneity. Our main results identify general conditions under which the potential increase in expected output due to assortative matching (relative to random matching) is offset by the costs of signalling. Finally, we examine the limit model with a continuum of agents and point out differences and similarities to the finite version. Technically, the paper is based on the elegant theory about stochastic order relations among differences of order statistics, pioneered by Barlow and Proschan in 1966 in the framework of reliability theory.

We study the optimal design of organizations under the assumption that agents in a contest care about their relative position. A principal determines the number and size of status categories in order to maximize output. We first consider the pure status case without tangible prizes. Our results connect the optimal partition in status categories to properties of the distribution of ability among contestants. The top status category always contains a unique element. For distributions that have an increasing failure rate (IFR), a proliferation of status classes is optimal, whereas the optimal partition involves only two categories if the distribution of abilities is sufficiently concave. Moreover, for IFR distributions, a coarse partition with two status categories achieves at least half of the output obtained in the optimal partition with many categories. Finally, if status is derived solely from monetary rewards, we show that the optimal partition in status classes contains only two categories.