The resurrection of the Lucas supply function at the hands of Keynesians

Wolfgang Munchau has perhaps the clearest statement of the view that an extra dose of current inflation will boost output:

A helicopter drop means that the ECB would print and distribute money to citizens directly. If it were to distribute, say, €3,000bn or about €10,000 per citizen over five years, that would take care of the inflation problem nicely. It would provide an immediate demand boost, and drive up investment as suppliers expanded their capacity to meet this extra demand.

I don’t mean to pick on Munchnau, who is one of the two or three best columnists in the world (thus the clarity), but I view this as incorrect as stated.  And it is symptomatic of a mistake which I see more and more frequently, including from reputable economists.  Given the genesis of the Great Recession, commentators have become obsessed with stimulating demand, but “mere inflation” does not on its own put people back to work, at least not by much.

To be clear, if aggregate demand is on the verge of falling, and expansionary monetary policy maintains aggregate demand, that will indeed prevent a big increase in unemployment.  And typically that is very much worth doing, and most of all what is useful is an ex ante AD maintenance rule from the get-go.  But that does not mean inflation in any particular state of affairs will boost employment significantly.

Let’s go back in time to the 1970s and 1980s.  Bob Lucas developed a monetary misperceptions version of business cycle theory, in which boosts of inflation encouraged people to work more, at least temporarily, and set off a cyclical pattern of boom and bust.  Fortunately the Keynesians stepped in and criticized Lucas in a rather devastating manner.  The measured responsiveness of labor supply, or for that matter investment, to inflation, or for that matter relative price changes, simply wasn’t that large.  That also was a big problem with the core labor market mechanisms of real business cycle theory; for instance read the prescient critique by Larry Summers (pdf).  Those same arguments imply that today more inflation will boost employment by only small amounts.

And consistent with that claim, the Phillips Curve is not exactly stable as of late.

There are also plenty of papers on inflation and investment.  They are hard to summarize, but overall it is easier to argue that more inflation harms investment rather than helping it (pdf).  And at the most general level, it is real cash flow that predicts investment well, not nominal cash flow.  So I am not so optimistic about more inflation today boosting investment by very much, even though I agree that a higher price inflation or ngdp target in the steady state would be very useful for preventing aggregate demand collapses.

Now you might think we are in special circumstances with rates of price inflation at especially low levels.  What harm is there in risking more inflation and having prices rise at 2 percent, 2.5 percent, or even three percent a year?  I agree with this argument.  (I’d be happy to see higher rates of price inflation if only to erode the value of academic tenure.)  Still, if we ask ourselves what is the best point estimate for how much more an extra dose of inflation today will boost investment (and not just the stock market), any strongly positive answer is based more on faith than on clear evidence.  The mere fact that you know “demand hasn’t been high enough” — which is true — doesn’t have to mean current doses of price inflation are going to get us very far.  Yet that is the mistake I see people making again and again.

People, economists have known this for a long time, it’s just that they now are starting to forget it.

One more point: demand could go up through yet another mechanism.  Imagine the economy becomes more productive, wages rise, and stronger consumer demands percolate throughout the broader economy.  That too is an increase in demand, and for that matter supply, and a decline in the risk premium.  It is quite possible the effect of that kind of demand increase on output is stronger than the effects of higher price inflation.  We should not conflate these two scenarios, and I get nervous when I see the word “demand” without further qualifiers or description.

Here are related remarks from Matthew C. Klein.


Comments for this post are closed