The Return of the Original Phillips curve? An Assessment of Lars E. O. Svensson's Critique of the Riksbank's Inflation Targeting, 1997-2012

We examine Lars E O Svensson's prominent critique of the monetary policy of the Sveriges Riksbank (the Swedish central bank) from 1995-2012. Our main objection concerns Svensson's conclusion that the original pre-Friedman/Phelps version of the Phillips curve based on constant inflation expectations has returned for Sweden. Based on estimates of this model, Svensson claims that that the Riksbank's policy has contributed to an average of 38 000 more unemployed a year between 1997-2011. This result is based on Svensson's unrealistic as well as unnecessary assumption of constant inflation expectations anchored at the Riksbank's inflation target of 2 per cent. Data show, however, that the public's inflation expectations have varied between 0 and 4 per cent, thus they have not been anchored. The negative employment effect found by Svensson vanishes once actual data on inflation expectations are included in the estimates of the Phillips curve. The long run non-vertical Phillips curve is transformed into a vertical one, in line with the Friedman/Phelps theory. <p> We have additional objections to Svensson's reasoning. First, we show that the Riksbank has on average met its inflation target between 1995 and 2012. Second, we suggest that the original Phillips curve is too simple a model to draw any firm policy conclusions about unemployment and monetary policy in a small open economy such as Sweden. Third, we do not want to overburden Swedish monetary policy by making the Riksbank responsible for three objectives. It has already two objectives: price stability and financial stability. Criticising the Riksbank for employment losses, as Svensson does, gives priority to a third objective, high employment. Finally, Svensson adopts a short-term perspective by focusing on the period 1995-2011. When we compare the Riksbank's inflation targeting regime with previous monetary policy regimes over the past 100 years, inflation targeting in the past fifteen years is clearly one of the most successful.