are apparently doomed to repeat old theories as new.
I typically save entries on this blog for times when I am able to write an in depth, longer analysis of some issue that has piqued my interest. This entry however will be relatively concise because I see the issues as being rather simple.
An article in today’s New York Times by Neil Irwin summarizes a much longer paper by Jeremy Rudd, an Economist with the Federal Reserve. The argument of the paper is both straightforward and interesting, though it’s originality can be questioned. Rudd’s argument is that the various inflation augmented Phillip’s curve models, resting on general equilibrium and rational expectations foundations, do not explain the actual behavior of inflation. According to Rudd, long term inflation expectations do not actually have that much impact on current rates of inflation. What is curious, is that this is presented as a novel insight, by both Irwin in the New York Times and by Rudd in his longer article.
What strikes me however is that this analysis is actually rather old: specifically, it’s straightforward Old Keynesian, not to be confused of course with original Keynesianism or Post Keynesianism. Once you allow that the economy is driven primarily by fluctuations in aggregate demand and that you have some degree of nominal wage-price stickiness, along with adaptive expectations, you get a more or less normal state of affairs with a steady trade off between inflation and unemployment. Low unemployment generates increasing wages, which in turn drives up prices and the reverse happens when Aggregate Demand declines. There is a straight forward explanation as well for stagflation and the adjustment down from stagflation in the 80’s and that explanation is in supply side shocks, especially the behavior of oil prices.
My point here is not to praise this model. Nor is my point to critique it, at least not presently. That said, there’s a lot to critique about it. But this relatively simple presentation of the behavior of inflation and unemployment was exactly what was rejected by the rational expectations revolution. Yet the rational expectations view, in its strong form, that agents in the aggregate accurately predict the true mean of macroeconomic variables is both ontologically and epistemologically, improbable. Since these models became popular in economics in the 80’s, we’ve been relying on them now for close to 40 years.
If its the case that the older, rather than the newer model does a better job of explaining the data, why are economists and economic reporters acting so shocked when the newer, flawed model fails to perform? And then why not simply acknowledge that the whole shift was never a progressive shift, but in fact, a regressive shift? Perhaps the Old Keynesians are owed an apology.
Of course, we could always build economics on the foundations of uncertainty as Hyman Minsky proposed.
Citizen Rat 