Exchange Rates
Peso problems

Since the establishment of the Exchange Rate Mechanism in 1979, a number of member countries have experienced substantial and persistent rises in their real exchange rates. Much recent comment has emphasized the ERM's contribution to counter-inflationary discipline; tying a government's hands to a sound external currency specifically the Deutschmark should prevent time-inconsistent monetary policy and ensure the credibility of inflation targets. Since this credibility is imperfect, however, so long as a country can devalue to ease domestic monetary conditions, member governments have stressed the need to prevent parity changes at least until September 1992. While barriers to parity changes exist, it is this possibility that distinguishes the ERM from a monetary union.

In Discussion Paper No. 774, Research Fellow Patrick Minford focuses on the resulting exchange risk for an ERM member country that seeks to restrain inflation and minimize economic distortions. If trading conditions deteriorate drastically, domestic manufacturers will lobby strongly for subsidies and devaluation, but the moment at which such conditions will occur is uncertain, so there is a risk of future devaluation at all times. Wage negotiators therefore account for this potential `extra' inflation in their contracts. This raises wages and prices in the non-traded sector above those in the traded sector, which is directly disciplined by foreign competition. This raises the real exchange rate and the cost of traded goods, which squeezes the traded goods sector; so the very fear that devaluation will help this sector in bad times damages it in normal times. Government subsidies to the traded sector must then be paid from general taxation, and the resulting distortions reduce overall output. Minford concludes that ERM member governments' reactions are optimal given their incentives, but the asymmetry of their reactions favouring devaluation and subsidy in bad times but no parity change or subsidy in good times leads to a `peso problem': average expected devaluation is greater than in fact occurs, which leads in turn to overvaluation in normal times.