Low-Inflation Targeting and Unemployment Persistence

A recent model by Akerlof, Dickens and Perry (2000) (ADP) predicts that low inflation may cause unemployment to persist at high levels. This finding should be of major interest to European countries where inflation is targeted at low levels. We specify a small open economy version of the ADP model and apply it to Swedish data. The results indicate that raising the Swedish inflation target from 2 to 4% would bring long-run unemployment down by two percentage points, to 2.0-2.5%. EMU membership, with inflation at the average of the present 0-2% band, would raise unemployment to around 6%. Membership thus implies a rejection of a national inflation target that could maximize employment. Given that long run unemployment-inflation trade-offs can be found in other countries as well, there is nothing to suggest that these trade-offs are identical across countries. A single inflation rate in the EMU may then cause unemployment to widely exceed the lowest sustainable rate in individual countries. We also extend the ADP model by showing theoretically that the unemployment minimizing inflation rate could lead to too low output. However, empirically we find, both for Sweden and the U.S., that minimum unemployment and maximum output occur at roughly the same rate of inflation.