In simple versions of the Mundell-Fleming model for small open economies whose financial markets are fully integrated with the world economy, a monetary expansion produces an increase in output and employment, which is reinforced as downward pressure on interest rates induces exchange rate depreciation, which in turn leads to capital outflows and boosts net exports. A bond-financed budgetary expansion has no such effect on employment and output, however, since upward pressure on domestic interest rates leads to exchange rate appreciation, and a reduction in net exports crowds out the initial increase in public spending. In a world of floating exchange rates and perfect capital mobility, however, a monetary expansion is a beggar-thy-neighbour policy, since the foreign country's net exports will fall, while a budgetary expansion drives growth world-wide. Such models underlie much of the literature on the costs and benefits of international policy coordination, but they are built on shaky microeconomic foundations: they are subject to the well-known Lucas critique of econometric policy evaluation and do not lend easily themselves to a thorough welfare analysis.

In Discussion Paper No. 491, Research Fellow Frederick van der Ploeg uses a class of two-country, intertemporal, perfect- foresight models, with floating exchange rates, uncovered interest parity and nominal wage rigidities, to investigate the robustness of the results derived from two-country models of the Mundell- Fleming-Dornbusch type. For his bench-mark model characterized further by unit elasticities of intertemporal and intratemporal substitution in consumption, no initial holdings of foreign assets and infinite lifetimes he shows that monetary disinflation and increased public spending have no spillover effects on foreign consumption and employment and there are no current account dynamics. He then identifies four main channels of international policy transmission.

First, if foreigners hold domestic assets, the exchange rate appreciation associated with monetary disinflation or increased public spending boosts their real value and requires a bigger home-country trade surplus to service the additional debt, so foreign consumption and unemployment both increase. Second, if the elasticity of intratemporal substitution differs from unity and goods are gross substitutes, then a monetary disinflation has negative short-run but positive long-run effects on net foreign assets and corresponding spillover effects on foreign consumption. If goods are gross complements, these effects are reversed. Third, if the elasticity of intertemporal substitution is greater than one, disinflation leads to the accumulation of foreign assets, a rise and subsequent fall in foreign consumption, a temporary increase in foreign unemployment and nominal interest rate overshooting. Again, if this elasticity is less than one, these effects are reversed. Finally, allowing for finite lifetimes without an intergenerational bequest motive destroys Ricardian debt neutrality and permits current account dynamics and non-trivial spillover effects, even with unit elasticities of intertemporal and intratemporal substitution.

Channels of International Policy Transmission
Frederick van der Ploeg

Discussion Paper No. 491, January 1991 (IM)