Behold the Phillips curve, yet another statistical correlation handled as an everlasting verity that our inflationary period has simply undermined.
From 2007 to 2019, the usual commentary was “The Phillips curve has develop into flat.” Giant modifications in unemployment correspond to little or no change in inflation, or small modifications in inflation correspond to very large modifications in unemployment, relying on which causal (mis) studying of the correlation you select. To the optimist, permitting a tiny little bit of inflation may dramatically scale back unemployment. To the pessimist, it might take immense unemployment to do something about inflation, ought to now we have to.
Then got here the pandemic. Unemployment shot up with no change in inflation, proper on the curve.
Then got here the inflation. The Phillips curve wakened. It is virtually vertical! (The scales of the 2 axes are totally different).
A lot Fed and commentator pondering depends on the Phillips curve. It is the central approach rates of interest have an effect on inflation, in typical pondering. Excessive rates of interest increase actual rates of interest decrease mixture demand trigger unemployment which causes through the Phillips curve, decrease inflation.
Clearly, one thing may be very incorrect right here. Perhaps expectations shift. Perhaps provide shocks do matter in spite of everything. Certainly one ought to begin with a critical dynamic Phillips curve, as most macro literature does. Perhaps the Phillips curve is versatile up however sticky down, and the pure price shifts round. Perhaps costs are sticky till they don’t seem to be. As Bob Lucas confirmed way back, the slope of the Phillips curve relies on the volatility of inflation. International locations with unstable inflation get no output increase from extra inflation. 1000’s of epicycles could be added, and this put up is a little bit of an invite to take action. Or possibly the Phillips curve was only a correlation in spite of everything, hiding a deeper actuality. (My view, however for an additional weblog put up).
Within the meantime, it is one other good warning to not take statistical correlations too critically, and definitely not as causally as we are likely to do. Comparable to inflation will all the time be 2%. Comparable to actual rates of interest are on a everlasting downward development?
This time of inflation will lead us to rewrite an terrible lot of macroeconomics.