This Matt Yglesias tweet criticizes a recent Bloomberg article by Noah Smith:

Yglesias is referring to an important 1998 paper by Paul Krugman, where he argued that when in a liquidity trap a central bank should promise to be irresponsible, that is, commit to higher than normal inflation after the economy has recovered and interest rates have risen above zero.

Like J.M. Keynes, Krugman is very good at expressing unconventional ideas in an evocative fashion, but it’s also important to understand exactly what he is saying.  My reading of the paper is that Krugman is not suggesting that central banks promise to be irresponsible, rather that they promise to be “irresponsible”.  The scare quotes indicate that Krugman actually views his policy recommendation as responsible, but recognizes that a period of above normal “catch-up” inflation might be viewed as irresponsible by conservative central bankers, and even the broader public.

For example, suppose a central bank had a 2% inflation target, but was in the midst of a 5-year period of zero inflation.  Assume that because interest rates are zero the central bank is currently unable to stimulate the economy.  In that case, the central bank might want to promise a 5-year period of 4% inflation after exiting the liquidity trap.  If credible (an important consideration) this promise would raise inflation expectations and reduce real interest rates.   (If I’m wrong and it raised nominal interest rates, then it would allow the central bank to stimulate the economy using conventional policy tools.)

The policy may look irresponsible during the subsequent period of 4% inflation, as the economy will have already recovered from the recession.  But this inflation overshoot is helpful because it creates more bullish expectation during the preceding recession, spurring a more rapid recovery.

One can make a pretty good argument that the Fed’s recent adoption of “average inflation targeting” is at least partly motivated by Krugman’s paper, and subsequent research by scholars such as Michael Woodford, Gauti Eggertsson and Ben Bernanke.  The Fed is committing to having inflation average 2% over the long term (say the entire 2020s) even though it is currently running at below 2%.  That means they are committing to an “irresponsible” above 2% inflation rate in the latter 2020s.  But that’s the responsible thing to do!

So who’s right, Yglesias or Smith?  Each side of the debate has a valid point.  For instance, here’s Smith:

Seeing these developments, economists like Summers and Blanchard might worry that a policy regime change is about to follow — or that the public will believe that one has happened, which would amount to much the same thing. So they might be trying to set themselves up as a counterweight to those who dismiss the importance of deficits, in order to reassure the public that once the Covid-19 crisis ends, leaders will go back to worrying about fiscal prudence.

But given their lack of roles within the Biden administration, economists like Summers and Blanchard might not be the best-positioned to do this. Instead, this is probably a job for the Fed. The central bank should make it clear that if deficit spending leads to substantial inflation — say, over 6% — it will raise interest rates to fight it, even if that means hurting the economy.

I would go even further.  The Fed should set rates at the level necessary to hit their target.  Full stop.  If it’s a 2% AIT over the 2020s, then interest rates should be set at a level that produces 2% inflation over the 2020s, on average.  Yes, they have a dual mandate that includes employment, but that’s why they’d be better off targeting NGDP growth.  Unfortunately, they have not taken my advice, and have committed to an average inflation rate of 2%.  So the Fed should what’s necessary to achieve their target, otherwise their promises will not be credible in the future.

In Yglesias’s defense, at the moment it seems like the public is expecting inflation to average a bit less than 2%.  Here’s David Beckworth and Ramesh Ponnuru:

So inflation hawks have highlighted the difference in yields between Treasury bonds that are adjusted for inflation and those that are not. That difference has been rising in a way that seems to imply that the market is forecasting inflation a little above 2 percent.

But the Treasury ties its inflation-adjusted bonds to a different measure, the Consumer Price Index, and it typically runs higher than Personal Consumption Expenditure inflation. More important, the Fed’s own purchases of these bonds has made the market for them more liquid, thus decreasing their yield.

Take account of such factors, and the forecast for the average inflation rate over the next five years is under 1.5 percent, well below the Fed’s target for action. For all of the recent alarms sounded about inflation, expectations are below where they were before the pandemic started. . . .

But catch-up growth in the dollar size of the economy would almost certainly entail going through a period of inflation above 2 percent. The key questions are whether the Fed will seek such growth, or at least tolerate the inflation that comes with it, and whether markets will have confidence that it will.

The financial markets continue to fear that the Fed will be a bit too “responsible”.

To summarize, Smith is right that the Fed needs to be responsible in the sense of being willing to raise rates if necessary to prevent the sort of high rates of inflation that we saw in the 1970s.  And Yglesias is right that the Fed needs to be “irresponsible” in the sense of allowing a bit above target inflation in the future to offset the below target inflation that we are currently experiencing.

Don’t let terms like “irresponsible” do your thinking for you.  Work out exactly what central bankers need to do to hit their target, and judge them on that basis.

PS.  This post is written from a Keynesian perspective.  I’m somewhat more optimistic than most Keynesians about the ability of central banks to do monetary stimulus at the zero bound.  But I do accept the view that an effective monetary stimulus should lead to expectations of at least some catch-up inflation.  The disagreement relates to how aggressively the Fed should be willing to take concrete steps such as large asset purchases while at the zero bound.  I say, “Whatever it takes”.