Edward Ray, in his comment on my 1976 paper, analyzes a slightly different model than the one I presented, and thus reaches different conclusions. His principal conclusions are that: (i) given wage rigidities, a wage subsidy to producers is needed, and this subsidy is equivalent to the optimal static subsidy that ensures full employment in each sector; and (ii) given the forced equilization of wages across sectors, a subsidy to workers is needed to encourage labor transfers between sectors. Thus, Ray finds that full employment is always desirable, whereas I find that some unemployment is (usually) present along the optimum path.

Mundell [20] demonstrated that in order to achieve balance of payments equilibrium and full employment, monetary policy should be paired with external balance and fiscal policy with Internal balance. The fundamental problem posed by Mundell concerned the methods governmental authorities should utilize to insure that both internal and external balance would be achieved when the underlying structural parameters of the economy were unknown. Mundell argued that the assignment of monetary policy to external balance and fiscal policy to Internal balance followed from the Principle of Effective Market Classification [20, p.76j: "Policies should be paired with objectives on which they have the most direct effect."

James Cassing and Jack Ochs' comment is, I believe, a very interesting extension of the analysis of my paper. Their two basic results are: (i) that congestion will occur in the search for jobs; and (ii) that given costly labor mobility, private decisions regarding voluntary quits will yield a socially optimal adjustment path if individuals have perfect foresight and if there is no congestion (externality) in the search process. Thus, they argue that even if factor prices are not rigid, the presence of congestion in the search process implies private decisions will not be socially optimal, and therefore that a subsidy will be needed to support the optimal plan.

One of the central issues in the ongoing debate concerning fixed versus flexible exchange rates is the comparative stabilizing properties of monetary changes as opposed to exchange rate changes. Fixed and flexible exchange rate regimes have been differentiated according to the ability of the private sector of an economy to alter the nominal value of its domestic money supply via the balance of payments. A disturbance which would lead a nation to experience a balance of payments deficit (surplus) in a fixed rate regime would lead to a depreciation (appreciation) of that nation's currency in a flexible exchange rate regime .2 In order to determine whether monetary or exchange rate changes--and, as a consequence, fixed or flexible exchange rates--best stabilize an economy from a given disturbance, it is necessary to consider the source of the disturbance, the nature of the economic system in question, and the variable(s) to be stabilized.

Economists have long debated the relative merits of fixed versus flexible exchange rate systems. One of the major issues in this debate concerns the ability of the exchange rate system to isolate a nation from external disturbances: see for example, papers by Mundell (1960), Stern (1963), Tower and Courtney (1974), and Enders (1977), Fischer , (1977) added another dimension to the controversy by examining the relative stability of real consumption and prices, assuming that the economy is subject to stochastic disturbances. His principle conclusion for the small open economy is that if disturbances are . external, then a flexible exchange rate regime is better in that foreign disturbances have no impact on the domestic economy.

The theory of exchange rate determination has evolved considerably in recent years. Starting from a supposition that exchange rates were determined by Balance of Trade equilibrium, and hence-reflected "real" factors, the resurgence of the Monetary Theory has caused a sharp change in perceptions as to how exchange rates are determined. The Monetary Approach to the Balance of Payments focused attention upon the role of currencies as assets and hence viewed the exchange rate as a relative price of two assets. Thus, those factors that, determine the demand for each currency - as well as the supply - are seen to explain exchange rates. Since, in the context of the Monetary Approach the demands for currencies are generally transactions demands, and since transaction demands for each currency are generally assumed proportional to nominal domestic output (with, perhaps, the rate of interest also affecting transactions demand), the general conclusion emerges that the exchange rate between two currencies will depend upon the ratio of domestic output levels, as well as the ratio of currency supplies.

Mundell [5] has demonstrated that monetary policy should be paired with external balance and fiscal policy with internal balance. This seminal article led to a voluminous literature 1 which at­ tempted to rectify many of the problems inherent in Mundell's flow equilibrium model. Harry John­ son [2] has characterized this extension of Mun­ dell's work as having "... lent itself to almost infinite mathematical product differentiation, with little significant improvement in quality of eco­ nomic product ..." Although we do not fully agree with Harry Johnson-for there are many deficien­ cies in Mundell's model-we do believe that more can be said concerning the "Assignment Problem" in the context of Mundell's model. Specifically, we examine the effects of introducing discrete lags into the Mundell model. Part A of this note presents a generalized discrete time version of Mundell's model and shows that the "Principle of Effective Market Classification" cannot guarantee stability. Part B then examines how the presence of an out­ side lag affects the results of Part A.

The life cycle hypothesis has become the dominant mode used to analyze the effects of a social security system on private saving, the labor/leisure choice, and social welfare. As both Barro and Samuelson indicate, a fully funded Social Security program (in a world of certainty) would drive out an equivalent amount of private saving. If the interest rate is r, the effects of a payment of a dollar into the social security pool while young would just offset* the effects of receiving (1+r) dollars as a transfer when retired. Papers by Diamond, Hi», and Samuelson, among others, have examined the effects of non-fully funded Social Security schemes in a growing economy. A non-fully funded program can be used to alter the private sector's saving rate and, hence, the capital/labor ratio. Social Security, then, can be used as a policy tool for achieving the (or some variant of the) golden rule growth path.

Past research has argued that, in the presence of disturbances, the choice of an exchange regime depend,s upon: i) the source of disturbances and ii) the openness of the economy. In this paper we demonstrate that, when all goods are tradeable, the source of disturbances is unimportant, whereas attitudes towards risk are of fundamental importance in choosing regimes. Further, we show that, when nontraded goods are present, the degree of commodity sustainability is crucial. Somewhat surprisingly, we also demonstrate that the more closed an economy is, the more likely it is that fixed rates are preferable.

The emergence of the portfolio balance approach 1 has led to a reformulation of the causes of balance-of-trade and payments disequilibria. According to this approach, balance-of-trade deficits and surpluses reflect discrepancies between desired and actual wealth holdings; while balance-of­ payments deficits and surpluses reflect discrepancies between desired and actual money holdings. Thus, balance-of-trade and payments disequilibria are viewed as representing disequilibria within the asset markets. Using this framework several authors2 have examined the self-correcting nature of disequilibria within the balance­ of-payments accounts and the ability of a devaluation to reduce the magnitude of a disequilibrium.