The organization of the labor market for medical interns and residents underwent a number of changes before taking its present form in 1951. The record of these changes and the problems that prompted them provides an unusual opportunity to study the forces at work in markets of this kind. The present paper begins with a brief history and then presents a game-theoretic analysis to explain the orderly operation and longevity of the current market, in contrast to the turmoil that characterized various earlier short-lived attempts to organize the market. An analysis is also given of some contemporary problems facing the market. A subsidiary theme of the paper concerns the history of ideas: the problems encountered in the organization of this market, and some of the solutions arrived at, anticipated the discussion of such issues in the literature of economics and game theory.

Previous authors who have considered partially nonexcludable goods have claimed that an increase in copyright protection will have the following two effects on social welfare. First, it will decrease the social welfare loss due to underproduction. Second, it will increase the social welfare loss due to underutilization. In this paper we investigate these claims in a formal setting by analyzing a model in which consumers vary only in terms of their costs of obtaining a reproduction. Our analysis provides partial support to the first claim of these previous authors while giving little or no support to the second claim.

This paper addresses the question, What happens to the arrest rate when the number of law enforcers increases? One implication of the analysis is that arrest statistics are a poor instrumental variable for judging the quality of law enforcement. Increasing the number of police can increase of decrease the number of arrests. An increased probability of arrest induces fewer criminal acts; hence the ambiguity. Because of this result, we apply the theory in the setting of college basketball. We find a large reduction, 34 percent, in the number of fouls committed during a basketball game when the number of referees increases from two to three. Additional empirical evidence is presented which suggests that this elastic supply of basketball crime is due to more competent officiating and cleaner play.

A great deal of advertising appears to convey no direct credible information about product qualities. Nevertheless, such advertising may indirectly signal quality if there exist market mechanisms that produce a positive relationship between product quality and advertising expenditures. Two models of this phenomenon are presented. In each, advertising signals quality in the short run. The models differ in their treatment of the effect of advertising on long-run sales. In the first, all high-quality firms ultimately establish reputations for high quality whether they advertise or not. This is shown to imply that advertising can signal quality if and only if high-quality production requires investments in specialized assets that increase fixed costs but not marginal costs. In the second model, where nonadvertising firms never acquire a reputation for high quality, advertising might signal quality even if marginal production costs are somewhat lower for low quality. These conclusions closely parallel arguments previously made by Phillip Nelson.

This paper addresses the question, What happens to the arrest rate when the number of law enforcers increases? One implication of the analysis is that arrest statistics are a poor instrumental variable for judging the quality of law enforcement. Increasing the number of police can increase of decrease the number of arrests. An increased probability of arrest induces fewer criminal acts; hence the ambiguity. Because of this result, we apply the theory in the setting of college basketball. We find a large reduction, 34 percent, in the number of fouls committed during a basketball game when the number of referees increases from two to three. Additional empirical evidence is presented which suggests that this elastic supply of basketball crime is due to more competent officiating and cleaner play.

A great deal of advertising appears to convey no direct credible information about product qualities. Nevertheless, such advertising may indirectly signal quality if there exist market mechanisms that produce a positive relationship between product quality and advertising expenditures. Two models of this phenomenon are presented. In each, advertising signals quality in the short run. The models differ in their treatment of the effect of advertising on long-run sales. In the first, all high-quality firms ultimately establish reputations for high quality whether they advertise or not. This is shown to imply that advertising can signal quality if and only if high-quality production requires investments in specialized assets that increase fixed costs but not marginal costs. In the second model, where nonadvertising firms never acquire a reputation for high quality, advertising might signal quality even if marginal production costs are somewhat lower for low quality. These conclusions closely parallel arguments previously made by Phillip Nelson.