We study the selection of an equilibrium for coordination games: symmetric, simultaneous move, complete information games which have multiple, Pareto-ranked Nash equilibria. We design and experiment to explore regularities in the observed outcomes for this class of games. With replication, we find that the Nash equilibrium concept accurately predicts the strategies chosen by players in these games. However, the equilibrium outcome is not always the Pareto-dominant equilibrium so that coordination failures can arise. Moreover, we find that altering the payoffs of a dominated strategy can influence the selection of a Nash equilibrium. Our results are consistent with a modified version of Harsanyi's tracing procedure in which players initially place some positive probability that their opponent is a cooperative player even though the cooperative strategy may be dominated by another strategy.

This paper develops a two-sector model with imperfect competition in order to explore the positive and normative aspects of the share system advanced by Weitzman in his book. In this model, the degree of competitiveness is measured by a parameter /eta = [(F - l)/F], where F is the number of firms in each sector. With fixed-wage contracts, sector-specific shocks generate aggregate fluctuations in employment and output through a multiplier effect. Introducing share contracts will not yield a Pareto-dominant allocation unless the share parameter is exactly equal to /eta. When the share parameter equals /eta, welfare is increased. Thus to reap the benefits of introducing share contracts requires rather exact information on the competitiveness of the economy.

We study an economy in which producers incur resource costs to replace depreciated machines. The process of costly replacement and depreciation creates endogenous fluctuations in productivity, employment, and output of a single producer. We explore the spillover effects of machine replacement on other sectors of the economy and provide conditions for synchronized machine replacement by multiple independent producers. The implications of our model are generally consistent with observed monthly output, employment, and productivity fluctuations in automobile plants. Synchronization of retooling across plants within the auto industry is widespread, so that the fluctuations observed at the plant level have aggregate implications.

This paper contrasts the dynamic properties of an imperfectly competitive economy with a representative agent, real business cycle model. For both economies, inventories are the important dynamic linkage. The predictions of these models with regards to the comovement of employment across sectors may differ. Empirical evidence on the comovement of employment over the business cycle is consistent with the model of imperfect competition with inventory holdings.

This paper investigates two aspects of the macroeconomic consequences of market participation decisions of imperfectly competitive firms. First, can there exist multiple, Pareto-ranked Nash equilibria indexed by the level of market participation? That is, can there exist both thin and thick market equilibria? Second, can the variations in the degree of competition that stem from shocks to preferences and technologies help to understand observed fluctuations in the macroeconomy? This paper is part of an ongoing research program intended to understand the macroeconomic implications of the coordination of economic activities in model economies without complete and/or competitive markets. In such environments, the fundamental theorems of welfare economics do not apply and it is quite possible for the economy to have multiple, Pareto-ranked equilibria. Low-welfare equilibria represent situations in which individual agents, acting noncooperatively, are unable to successfully coordinate their activities and reach a preferred equilibrium-this is termed a coordination failure. In equilibria of this type, there are $100 bills lying on the sidewalk but it takes the effort of more than one individual (i.e., coordination) to reap these gains! Our model relates to two important strands of the literature on coordination failures. First are model economies in which the deviation from the Arrow-Debreu paradigm arises from the market power of sellers (see, for example, Oliver Hart, 1982). In these models, the number of active firms is usually taken to be exogenous. Our contribution is, in part, to allow the number of firms to be determined by the costs and benefits of market participation. This allows us to relate the degree of competition in the economy to variations in fundamentals like technology and preferences in addition to exogenous (but self-fulfilling) variations in expectations. Second, our model represents another example of a participation externality in which the gains to participating in an activity, such as entering a market, depend on the number of other agents participating as well. The papers by Peter Diamond (1982), Chatterjee (1988), and M. Pagano (1987; 1988) explore market participation externalities of a different variety. In Diamond's work, these externalities arise through the matching process, while Chatterjee and Pagano (1987) explore the risk-reducing effects of large markets. Similar externalities are found in the industrial organization literature on networks, as in Michael Katz and Carl Shapiro (1985). The model explored in this paper highlights a participation externality arising from the interaction of imperfectly competitive firms.1 tDiscussants: Robert E. Hall, Stanford University; Peter Diamond, MIT; Olivier J. Blanchard, MIT.