The debate on how to avoid unstable exchange rates has changed its focus, Marcus Miller argued at a lunchtime meeting on 8 July. Problems experienced under the Louvre Accord last year meant that the choice for a reformed international monetary system was no longer primarily between fixed rates and target zones, but rather between target zones and a form of `disciplined floating' where fiscal policy is tied to keeping the current account in balance. Marcus Miller is Professor of Economics at the University of Warwick and Co-Director of CEPR's International Macroeconomics programme. He has written widely on floating exchange rates and the EMS, and has published with John Williamson on target zones, including CEPR Discussion Paper No. 266.
Experience with floating rates since 1973 has proved disappointing in a number of respects, Miller noted. In the first place, exchange rates have moved far more in response to fundamentals than seems rational: the US dollar appreciated by roughly 70% in real effective terms over 1980-5, only to fall by an even larger amount subsequently. Second, these `misalignments' have been associated with unsustainably large payments imbalances. Finally, the credibility of national monetary targets (designed to supply a nominal anchor under floating) has been substantially undermined by the speed of financial innovation and deregulation.
Recent developments, notably the Plaza agreement (to drive the dollar down) and the Louvre Accord (to stabilize the dollar and to shift domestic demand in ways that would accommodate an improvement in the US deficit), demonstrated a marked shift in attitudes towards exchange rates and had put the subject of international monetary reform firmly on the agenda. In the Table overleaf, Miller classified international monetary systems (both past and proposed) by two criteria, flexibility of exchange rates and symmetry of operation. In his talk, he focused on the alternative proposals for international monetary reform put forward by Ronald McKinnon, John Williamson and James Boughton.

McKinnon's 1984 plan for a symmetric Paper Gold Standard envisaged fixed exchange rates between the United States, Japan and Germany and a target for the growth rate of the aggregate price level. The G3 economies would set interest differentials so as to stabilize exchange rates and adjust the average level of interest rates to preserve price stability. Fiscal policy is directed to achieving current account balance. McKinnon was right, according to Miller, to argue that some type of symmetric system is more likely to be acceptable (especially to Germany) than a return to US hegemony, and that monetary policy is easier to coordinate than fiscal policy. But his plan, which rests upon explicit cooperation among G3 central banks, assumed more consensus among national policy-makers than is the case now or in the fore- seeable future, Miller argued; at the Toronto summit, monetary coordination had hardly been discussed.
The prolonged deviations from equilibrium (`misalignments') during the 1980s had resulted from the failure of nominal exchange rates to adjust to offset inflationary differentials so as to keep real exchange rates stable, Miller noted. It is this difficulty that Williamson's target zones were designed to remedy: the G7 countries should adopt targets for central exchange rates that were designed to deliver sustainable current account balances at the highest level of output at which inflation is stable, while monetary policy should be used when rates deviate by more than 10% from such targets. Target values for nominal exchange rates would in effect be adjusted to offset inflation, thus becoming effectively target real exchange rates. To counter the objection that there would be insufficient check on inflation, the proposal was extended to include use of average G7 interest rates to stabilize aggregate nominal growth, while individual countries' fiscal policies could be adjusted to influence domestic nominal growth.
The extended target zone system resembles the McKinnon plan in emphasizing multilateral cooperation on stabilizing exchange rates (but involving G7 rather than G3). In addition, Miller observed, despite their different exchange rate targets, both proposals employ similar mechanisms for controlling global inflation, each assigning the average level of interest rates to stabilize a nominal target. Miller rejected the view that the Louvre Accord represented an experiment with a form of target zones. The Louvre bands were not made public, were narrow and bilateral, and no policy response was specified in the event of deviations from target. The extended target zone system was evaluated in CEPR Discussion Paper No. 221 by David Currie and Simon Wren-Lewis, using the National Institute World Economic Model (GEM). Their results suggested that applying such rules in practice would have led to a significant improvement in economic performance during 1975-85.

Recently, however, the policy assignment embodied in the extended target zone system has been challenged by James Boughton, who argues against using monetary policy for exchange rate targets. Instead, he recommends a return to floating rates, but with two major provisions: that monetary policy be guided by targets for nominal income rather than for money supply, and that fiscal policy be `disciplined' so as to secure balance of payments targets. In addition to the difficulties associated with coordinating fiscal policies, however, a further study by Currie and Wren-Lewis had found that the `J-curve' response of current accounts and the weak link between money and nominal income prevented Boughton's policy assignment from working effectively.
The habit of policy coordination has been established and looks set to continue, Miller concluded. There is general agreement on the need to secure and maintain sustainable exchange rates and to manage aggregate demand so as to accommodate shifts in external balance. Fixed exchange rates, however, are off the agenda. Miller also referred to recent reports of a much greater acceptance of exchange rate targets by the US and Japan than by the Bundesbank. Germany was apparently unwilling to fix rates against the dollar because she lacks confidence in the US's willingness to fight inflation. As Germany was also sceptical of a target zone system, which directs monetary policy towards exchange rates rather than inflation, this suggested a more flexible relationship with the dollar <197> possibly along the lines suggested by Boughton.
The best way forward was unclear, however, since the performance of simple rules varies, depending on the nature of the shocks assumed. This was an important issue for research, Miller noted: better to test the performance of alternative regimes on economic models than on the economy.
One member of the audience wondered why a fixed exchange rate regime was so inflexible, given the longevity of Bretton Woods. Miller argued that the major difference now was the lack of agreement over coordinated policies among the G3. It was also pointed out that the workability of such a system had been fundamentally undermined by the removal of exchange controls and of direct controls over banking systems, as well as by the great increase in funds available for speculation held by the corporate sector. Miller was also asked his views on future UK policies. He noted the wide agreement among economists on the need for a tighter fiscal policy which was consistent with all three alternative proposals for international monetary reform.