The main objective is to develop and apply methods for describing optima in matching models of money using the tools of mechanism-design analysis. A secondary objective is to apply those methods to models of substantive interest: models of private money and models with imperfect recognizability of money.
There are three methodological features of the project. First, the trades that people make in meetings are part of what the mechanism determines subject, of course, to incentive constraints. (In most work on matching models of money, the way people trade in their pairwise meetings is taken as given in the sense that people trade according to some assumed bargaining protocol.) Second, in computing optima, lotteries are used to linearize objectives and constraints in the variables that determine current returns. (This technique, which, in part, is borrowed from the use of lotteries in computation in other areas, has not been used for numerical analysis of matching models of money.) Third, potentially nonstationary optima are studied in heterogeneous-agent models in which the state of the economy, a distribution of money or wealth, evolves endogenously and affects individual opportunities.
The project will contribute to the development of monetary theory in two ways. It extends the analysis of an attractive class of monetary models, those with decentralized exchange. Such models represent absence-of-double-coincidence difficulties and are useful in positive applications. (One recent application uses such a model to explain coexistence of money and higher-return assets, while another uses such a model to explain why the private note-issue systems in effect in many countries, including the U.S., at the beginning of the twentieth century did not produce so-called elastic currencies.) And, it brings to bear on such models the kind of mechanism-design approach being applied in other areas. This approach is attractive because the essentiality of money must be a result of a mechanism-design analysis and because the approach bypasses the need to make assumptions about bargaining protocols.
Broader impacts. The proposed use of lotteries in computation and the analysis of nonstationary optima should be applicable to dynamic models with heterogeneous agents in other areas of economics.
