Abstract

Exchange Rate System and Philipps Curve New Econometrically Founded Theses to the Actual Discussion on the “Best” Currency System

The point of departure of our study is the econometrical analysis of Carl-Ludwig Holtfrerich in Kredit und Kapital, No. 1/1982. Holtfrerich finds out that for the Federal Republic of Germany fixed exchange rates create better preconditions for a stability-oriented employment policy than flexible exchange rates. His conclusions are based on an econometric analysis which to our opinion includes some inconsistencies. Especially the fixation of the research periods from 1955 - 1969 as period of fixed exchange rates and from 1970 - 1979 as period of floating, is not economically plausible and distorts the interpretation of the datas, because they do not reflect correctly the changes in the international currency system. So it seems better to us to take the period from 1966 to 1973 (Break-down of the Bretton-Woods-System) as representative for the fixed exchange rates and the period from 1974 to 1981 as representative for the flexible exchange rates. During fixed exchange rates we ascertain a market trade-off between inflation and unemployment, whereas during the floating system no significant connection can be proved. Therefore during fixed exchange rates a reduction of unemployment is combined with higher rates of inflation than during flexible exchange rates. Our additional analysis of the “original” Philipps-trade-off shows, that during the Bretton-Woods-System increases in labour costs are joined with smaller pushes of inflation than in the recent years of floating. But this effect is compensated by the extremely strong trade-off between the development of labour costs and unemployment during fixed parities and the perfect absence of such a trade-off during flexible exchange rates. Therefore in contrast to Holtfrerich we conclude, that for the Federal Republic of Germany the conditions for a stability-oriented employment policy are better under a regime of flexible exchange rates than under fixed parities.