In this paper, we consider how recent history might have looked if the major OECD countries had coordinated their macroeconomic policies according to a simple set of feedback rules. These rules are based on proposals recently put forward by John Williamson and Marcus Miller, and are based on the idea of using interest rate differentials to keep exchange rates near their long-run equilibrium level. We find that, in terms of output and inflation, both individual countries and the world as a whole would have been significantly better off by adopting these policy rules.
Williamson and Miller propose an international policy regime involving five elements:
(1) Countries together choose a set of real exchange rate targets which are consistent with medium-term current account equilibrium in each country.
(2) Differences between national interest rates are used to keep real exchange rates near these target levels or within a zone surrounding them.
(3) Countries, either independently or jointly, choose targets for the growth of nominal demand. These targets can be adjusted according to the deviation of inflation from some desired level, the level of output compared to trend, and current account disequilibrium.
(4) National fiscal policies are varied to achieve these national targets for nominal demand growth.
(5) The average level of world interest rates is varied to stabilize aggregate world demand around the aggregate of national targets.
The first and second elements embody the idea of target exchange rates or exchange rate zones. Williamson and others have argued for some time that it would be desirable to limit the volatility of real exchange rates which, they argue, has had damaging effects on the efficiency of the international trading systems. Targets or target zones can limit volatility by providing an anchor for market expectations, and by limiting the ability of individual governments to export inflation by raising domestic interest rates and thereby appreciating their own real exchange rate.
The targeting of real rather than nominal exchange rates, however, provides no anchor for inflation. This is where the third element comes in. The nominal demand targets can be influenced by a variety of domestic factors, such as excess inflation, capacity utilization and the current account. Individual countries can attach their own weights to each factor; Germany, for example, might ignore capacity utilization and concentrate on inflation alone, while other countries could trade off lower inflation against higher output. Fiscal policy in each country can vary in order to pursue these nominal demand targets. Interest rate differentials are assigned to the real exchange rate target. This leaves the overall level of world interest rates undetermined, and so in the fifth element of the proposal these are adjusted to control the aggregate level of world demand.
While these five elements provide the general principles for a regime of international policy cooperation, a number of quantitative issues remain. For example, how fast should interest rates adjust to exchange rate disequilibrium? Should we rely on fiscal or monetary policy to control world demand? In an attempt to answer these questions, this paper carries out an optimal control exercise using GEM, the National Institute's world model. The simulations are carried out over the period 1975 to 1984, and involve the United States, Japan and Germany.
The study involves substituting feedback rules for the existing GEM equations explaining government expenditure and interest rates in the US, Germany and Japan. These feedback rules stipulate that changes in each government's expenditure depend on the level of and change in the gap between actual and target growth of nominal GNP in that country. Real interest rate changes also depend on these variables at the world level: global output disequilibrium would, for example, influence the general level of world interest rates. In addition real interest rates depend on deviations of actual real exchange rates from their target values. Finally, the nominal GNP growth target is a function of inflation, capacity utilization and the current account in real terms. The nominal GNP growth target only partially accommodates the current rate of inflation, thereby applying a steady disinflationary force. However, this is offset to some extent if capacity utilization is below normal levels or if the current account is in surplus.
The model is simulated over the period 1975 to 1984 and optimal control methods are used to determine the appropriate parameters of the feedback rules in light of the quantitative importance of the interactions in the model. For simplicity, the parameters of the feedback rules and growth targets are assumed to be the same for all three countries. The simulations confirm the importance of real exchange rates as an intermediate target. Capacity utilization should play an important role in the setting of nominal GNP targets, but the additional influence from the current account is less important. Fiscal policy should be used fairly actively to move nominal GNP growth towards its target. The results suggest, however, that the influence of global GNP disequilibrium on changes in the target level of world interest rates should be fairly small, because in GEM the size and timing of interest rate effects vary widely across countries. Thus an attempt to reduce global aggregate demand by raising the general level of interest rates would affect some countries more than others, leading to damaging shifts in the distribution of activity between countries and causing movements in real exchange rates.
Simulations of GEM suggest that the implementation of optimal feedback rules would have led to a significantly better output and inflation performance in all three countries over this period. World industrial production would have been about 6% higher than the level which was observed, with very little cost in terms of extra inflation. The optimal feedback rules allow the G3 countries to avoid the 1980/1 recession by preventing the large increase in US interest rates around 1980. The results also suggest that a more efficient way to contain inflation over this period would have been an earlier and more gradual contraction in fiscal policy.
The 1980 rise in US interest rates can be seen as a classic example of the dangers of uncoordinated anti-inflationary policy. Providing other countries do not react, a higher exchange rate brings short-term benefits in reduced inflation, while the deflationary effect of higher interest rates and a high real exchange rate can be counteracted by an expansionary fiscal policy. However, this policy brings with it two serious problems both of which have now become evident. First, the current account deficit produced by a high real exchange rate, together with the budget deficit generated by an expansionary fiscal policy and high debt interest payments, may not be sustainable in the longer term. Second, other countries are likely to react, raising their own interest rates, and this may distort the fiscal/monetary mix in the world as a whole, or simply lead to a world recession. The cooperative policy rules examined here are designed precisely to prevent developments of this kind.
The optimal rules simulated in GEM also substantially reduce the variability of real exchange rates, as we would expect. The main feature of the US real exchange rate under the optimal rule (relative to history) is the absence of the large appreciation after 1980, which in turn follows from the absence of the US interest rate hike. As a result, the US current account does not move into substantial deficit, and the US budget also remains fairly stable.