Comment on Search, Bargaining, Money, and Prices: Some Recent Results and Policy Implications

Abstract

Trejos and Wright introduce a new dimension to the study of money, or, more precisely, the study of liquidity. Liquidity confers more than just purchasing power: It also confers bargaining power. If only a small fraction of the population possesses liquid assets, it is difficult for a seller to find a potential buyer with sufficient liquidity to support a transaction. It follows that the threat by a potential purchaser to walk away from a bargaining round caITies greater weight than if liquidity is more symmetrically distributed. This suggests a pathway for heterogeneity in money holdings to have real economic effects. Trejos and Wright incorporate this intuition into a model of production and exchange. They examine how prices, output, and social welfare can be affected by changes in the cross-sectional distribution of liquidity. Trejos and Wright's work represents a tour de force of theoretical modeling. However, the ultimate importance of this line of research will be determined by its contribution to our understanding of monetary non-neutrality and optimal monetary policy. In spite of its subtitle, the Trejos-Wright paper has little to say about these issues. The monetary variable in this paper, denoted M, is the fraction of agents holding money. M does not directly correspond to any policy variable at the disposal of a government or central bank. The paper refers to M as the money supply, an accurate label since, by construction, each agent possesses either one unit of money or none. However, to describe an increase in M as an expansion of the money supply is somewhat misleading, since the impact of a change in M on the economy is not due to a change in the total number of monetary units available, but to the change in the cross-sectional distribution of money. To apply the results of this model to actual questions of monetaxy policy requires specifying the linkage between policy instruments of the monetary authonty (such as the total supply of reserves available to the banking sector) and the cross-sectional distribution of liquidity. The substantive implications of the Trejos-Wright model depend on how this linkage is constructed. In this discussion, I first review the key intuition underlying the Trejos-Wright