Resolving Differences over the Optimal Quantity of Money
比较了无限期模型与世代交叠模型对最优货币数量的不同结论，探讨差异源于有限寿命还是货币的交易功能缺失。
OPTIMAL STEADY-STATE MONETARY POLICY in models of infinitely lived agents calls for a deflation of the fiat money stock at a rate that equates the rates of return of fiat money and capital. This prescription for the optimal quantity of money holds whether the demand for money springs from placing money in the utility function (Friedman 1969), a cash-in-advance constraint (Grandmont and Younes 1973), or spatial separation (Townsend 1980). In contrast, monetary models of overlapping generations have counseled the (golden rule) optimality of a fixed stock of fiat money. 1 This model has often been criticized for its lack of a transactions role for money that would distinguish itself from other assets.2 The seeming difference in this model's implications concerning the optimal quantity of money thus raises an obvious question.3 Does the overlapping-generations model obtain different welfare implications because it assumes a finite lifespan or because it lacks a special transactions role for money? Recent work has reexamined the optimal quantity of money in overlappinggenerations models featuring transactions roles for money. Abel (1987), who takes the broadest and clearest look, finds that the golden rule optimal monetary policy,