As a consequence of the 'rational expectations revolution' in macroeconomics the conventional theory of economic policy has had to be revised to take account of 'forward-looking' expectations formed in the private sector. For policy-makers in an open economy in particular, expectations formed in the foreign currency markets play a vital role, and this affects the conduct of monetary policy. The foreign exchange rate is the relative price of two currencies, and this implies that policy in one country will be judged in relation to that pursued elsewhere. Recognizing this, national policy-makers will find themselves in a strategic relationship with each other, as well as with private speculators.

In Discussion Paper No. 27, Research Fellow Marcus Miller and Mark Salmon present an analytically tractable framework in which both these aspects of policy can be considered simultaneously. They use concepts from the theory of games to model the strategic relationship between national policy-makers. The results they obtain by considering a variety of possible strategic relationships also prove useful in analysing macroeconomic policy with a private sector which forms 'rational expectations'. Their paper, though technical in nature, is therefore useful in considering a wide range of policy questions.

To analyse strategic behaviour it is necessary to know how each policy-maker thinks others will react to the policies he chooses. The assumption that each decision-maker takes the others' policy actions (now and in the future) as given leads to the symmetric 'Nash' solution. When one player (the leader) is assumed to choose his policy actions knowing how the others will react to his announced plan of action, one obtains instead an asymmetric 'Stackelberg' solution. Miller and Salmon note that in the latter case, optimal policy is not 'time consistent'. In other words, the leader has no incentive to continue with the announced policy as time moves on from the date at which the plan was formulated. Miller and Salmon consider alternative approaches which do yield 'time-consistent' equilibria.

A government, designing optimal policy in an environment where the private sector has rational expectations, is in a position analogous to that of a Stackelberg leader. It is therefore not surprising to find that optimal macroeconomic policy is 'time inconsistent' in this situation as well; Miller and Salmon discuss several time-consistent alternatives appropriate in this context.

Miller and Salmon use these techniques to construct solutions to problems, common in open-economy macroeconomics, involving several national policy-makers and private markets with rational expectations. As an illustration, they discuss the design of fiscal policy in a Common Market, with fixed internal cross-rates but a floating external exchange rate.