Imaginary Squabbles Part 1: Krugman and DeLong on Inflation

Paul Krugman and Berkeley blogger Brad DeLong appear peculiarly agitated about two of the many articles I have written over the past 42 years. The first, from 2009, summarized a longer critique of Krugman’s claim that “liquidity traps” left monetary policy powerless in the U.S. in the 1930s and in Japan since the 1990s. The second, in June 2010, voiced premature optimism about Ireland’s nascent recovery – months before that country wasted a fortune bailing out the banks.

In these cases and others, Krugman and DeLong go to great lengths to put words in my mouth – a proclivity that others have observed.

Krugman writes, “here’s what I find remarkable about Reynolds and people like him: they have a track record. Here’s Reynolds in 2009 ridiculing my claims that we were in a liquidity trap, so that even large increases in the monetary base would not be inflationary. Here he is in 2010 declaring that Ireland’s embrace of harsh spending cuts will produce an economic boom… . And here we are in 2013, with the Fed’s balance sheet up by more than 200 percent and no inflation, with Ireland still mired in a deep slump …”

What I find remarkable about Krugman and DeLong is that they provide links to my articles. That makes it easy to discover they are misquoting me egregiously. Let’s focus first on Krugman’s notion that my skepticism about the notion of a “liquidity trap” is tantamount to predicting that large increases in the monetary base (bank reserves and currency) must be wildly inflationary.

DeLong dredged-up my skepticism about liquidity traps as proof that, “Alan is part of a group [???] that thinks that policy is dangerously close to producing a mammoth outbreak of inflation … And he’s been telling this story for 3 ½ years or so.” Really? If I had been warning of a mammoth outbreak of inflation for years you might think Google could find some record of that.

My mid-2009 article on liquidity traps did not say or even imply that large increases in the monetary base must cause inflation. On the contrary, I explicitly argued that large increases in the monetary base (bank reserves and currency) were necessary in the wake of bank crises – including 2008 when the Fed was much too tight. “In any bank crisis,” I wrote, “the public wants to hold more currency rather than bank deposits, and banks also want to hold excess reserves as insurance against bank runs. Japan’s central bank never adequately accommodated that demand for bank reserves and currency before 2001 (if then) nor did the Fed in 1929-33. But that does not mean (as the liquidity trap implies) that monetary policy was impotent and merely ‘pushing on a string.’”

That short 2009 piece was distilled from a detailed Cato Journal critique of Krugman’s Return of Depression Economics (RDE), which includes data for the U.S. from 1929 to 1940 and for Japan after 1991. “America spent most of the 1930s in a liquidity trap,” says Krugman; “Japan has been in one since the mid-1990s.” My data show otherwise. But Krugman and DeLong would rather quarrel with an imaginary inflation forecast than with my data.

Krugman’s wrote in his 1999 book that “a Friedman-style focus on a broad monetary aggregate gives the false impression that Fed policy [in the 1930s] wasn’t very expansionary. But it was; the problem was that since banks weren’t lending out their reserves and people were keeping cash in mattresses, the Fed couldn’t expand M2.” My figures show that M2 fell dramatically when the Fed let the monetary base fall in 1931–33 and doubled reserve requirements in 1938, but M2 rose by more than 10 percent a year during the rapid recoveries of 1934-36 and 1939-40. Krugman credits the good years to FDR’s undefined fiscal stimulus. But spending rose most rapidly under Hoover (from 3.4 of GDP to 6.9 percent in 1930-1932), and deficits were just 3.6 percent of GDP from 1933 to 1940.

Krugman’s mistaken impression that the Fed “couldn’t expand M2” was a sensible definition of a liquidity trap because it is testable against facts. By that test, the U.S. has clearly not been in a liquidity trap recently, since M2 grew by 6.6 percent a year from 2008 to 2012.

To a monetarist (a proponent of the quantity theory of money) “quantitative easing” is the only kind of easing there is. But it is the quantity of broadly-defined money that matters in this model – not the quantity of monetary base, and certainly not nominal interest rates. Near-zero nominal interest rates in the U.S. in the early thirties and in Japan since the nineties were not evidence of liquidity traps, but symptoms of negative money growth in the U.S., and slow money growth in Japan(2.3 percent a year from 1991 to 2012). The U.S. today is quite different. You don’t have to be a monetarist (I’m just a sympathizer) to suspect that M2 growing by 7.3 percent in 2011 and 8.5 percent in 2012 was cogent evidence against Krugman’s protracted habit of crying wolf about deflation.

There were a wide range of opinions back in 2009 about the net impact of the Fed’s quantitative easing when combined with the opposing squeeze from tighter bank regulation and interest paid on bank reserves. Krugman focused rhetorically on the extremes, as if doubting his predictions of deflation was equivalent to predicting hyperinflation. In reality, very few economists expected much inflation by now, though many expected somewhat less tepid growth of nominal GDP. When it comes to all the U.S. fiscal and monetary tinkering of the past five years, mainstream macroeconomics has much to answer for.

Krugman’s latest call for central banks to pursue higher inflation suggests that he doesn’t believe his own chatter about liquidity traps. “The key … is for the Fed to convince investors that it will allow somewhat higher inflation in the medium term,” Krugman wrote in a 2013 introduction to his book. The obvious problem here is that (1) the Fed can’t raise expected inflation unless it can raise actual inflation, and (2) if the Fed can raise inflation then there is no liquidity trap. Besides, if investors became convinced that inflation will be 3 or 4 percent for a few years, why would they keep holding 5-year Treasury bonds with a yield below 1 percent? Once bond yields moved up with higher expected inflation, it would sink stocks and raise government interest expense.

To summarize: My 2009 articles on about the history of liquidity traps complained that increases in the monetary were not large enough during past bank crises – not that large increases in monetary base were inherently inflationary. Yet DeLong cites this as proof that “Alan … thinks that policy is dangerously close to producing a mammoth outbreak of inflation.” DeLong is “pissed off” that I deny making an inflation forecast that exists only in his imagination. Really?