ABSTRACT
We discuss the evolution of and the theoretical issues surrounding the Phillips curve, which was originally (Phillips, 1958) a negative empirical relationship between the unemployment rate and the inflation rate. It has been the dominant theory of inflation in central banks in developed countries. Policymakers and politicians seem to like the idea that they could influence the economy by exploiting this relationship, that is, in order to create more employment, they inflate the economy. By late 1960s and through the 1970s, more theoretical analyses showed that such exploitation is fruitless because unanticipated aggregate demand shocks, that is, surprises, ignored by Phillips, is the only significant determinant of real output. Anticipated aggregate demand shocks, on the other hand, are ineffective in stimulating the economy. Every time the monetary authority attempts to stimulate demand, it increases real output (reduces unemployment) in the short run, then inflation expectations shift up the Phillips curve, and real output (unemployment) returns to their long-run equilibrium before the policy change. If the monetary authority continues to stimulate the economy, the Phillips curve keeps shifting up via expectations effects, and, in the long-run, the economy ends up with more inflation and no change in real output or unemployment; hence the Phillips curve is vertical in the long run, and no trade-off between inflation and output (unemployment) exists. Our Sample Generalized Variance tests indicate that the New Keynesian specification, change in inflation–output gap, is unstable in most of the 42 countries in our sample.
