Krugman: Inflation targeting is the first best solution

So now Krugman says inflation targeting is the first-best solution to our economic problems. This is consistent with everything he wrote about Japan over a decade ago, and also consistent with a wide majority of non-crazy macro economists across the political spectrum.

So why on earth has he not said this a lot more starting a long time ago? (Note that I've asked this question several times before--here are a couple early examples: 1, 2)  Krugman blames stupid economics.  He just sees misguided conventional wisdom against inflation as too much to overcome.

Huh?  Krugman--the man who boldly, clearly, and effectively challenges establishment views on EVERYTHING--gets weak knees when it comes to inflation targeting?

I see and get that inflation is a four-letter word among some crazy economists and much of the media.  But that's exactly why Krugman should be writing about it. Bernanke must have sympathy for this this view.  After all, it is what he spent much of his academic career advocating.  Conservative economists Ken Rogoff and perhaps Greg Mankiw would even back him, and many others on both sides of the political fence.

I'm annoyed because I think Krugman is in a unique position to raise the level of debate.  If he wrote about inflation targeting a lot, others would have to agree with him or not.  And those who didn't eventually would have had to back up their positions.  If Krugman started boldly with this line of attack a year ago I think there's a chance we'd be in a better place right now.

Incidentally,  my guess for what's restraining Bernanke, Geithner and others from more talk of inflation targeting is diplomacy with China, who would probably be unhappy if we were to target 3-5 percent inflation rather than 2 percent for the next five years or so.

Update:  Krugman and others have said that the key challenge is technical feasibility, not political feasibility.  They say the problem is developing credible expectations for future inflation. (See comments.)

I don't get that argument.  I think that if Bernanke laid out a clear 5-year plan for inflation, first increasing and then decreasing target inflation, markets would respond big time. He could further enforce those expectations through purchases of and sales of treasury bills, especially the inflation-indexed variety.

One argument against this is that it takes too many purchases from the Fed to influence long-term rates enough.  But I think those arguments were based on traditional Treasury bills, not inflation indexed bonds.  The inflation indexed bond market is a lot thinner and would be easier to influence. And despite possible claims of "cheap talk", I think a clear announcement by the Fed of targets going forward would further increase the Fed's leverage.

Risky? Maybe. But I think its potential benefits far outweigh its expected costs.  And unlike the other "second best" alternatives, it doesn't need to pass Congress.


  1. I don't understand why people are surprised by Krugman's position, that in theory raising target inflation would be preferable, but he thinks it would be ineffective because the market would not believe the new targets. It's not about whether he can convince policymakers, it's whether the policymakers can convince the market that the new higher target is credible. If the Fed raised the target in bad times, why would it not lower the target in good times? And if the market thinks it would, then the new target is not credible.

    Krugman has been saying this since at least a year ago when we entered the liquidity trap. Here is a blog post from Nov. 15, 2008 where he says pretty much exactly the same thing about targeting as what he said in his recent post that you are commenting on:
    And then a month later:
    Throughout the whole year, he has been making posts like this:

    On a different note, I think it was DeLong(?) who described the Fed's credit easing program as a one-sided bet with the market. If the economy improves then the Fed could sell the assets it has purchased, remove the expansion of m0, and inflation will be as normal; if the economy did not improve, the mortgage market becomes much worse, the assets the Fed bought drop in value, then suddenly it becomes much more difficult for the Fed to remove the expansion and expected inflation goes up - which would then cause the economic improvement we expecting in the first place. Heads or tails, Bernanke wins. Not a complete story, but certainly an interesting way of looking at it.

  2. KTM1: That's a good point. But I disagree a little on the merits of the argument. If inflation targeting is the first-best solution, then it should have been put up front by the scholars that know (or at least strongly suspect) this to be true.

    There's no excuse for the back-burner treatment it has received.

    By way of contrast, consider the single-payer option for health care. Many if not most health economists, and Krugman, think single payer is the first best solution, and they've made that position widely known, while simultaneously acknowledging political realities and practical feasibility. Krugman's position on inflation targeting has, in comparison, been relatively hidden from public view.

    I also disagree on the feasibility front. I think that if Bernanke laid out his 5-year plan for inflation, and spelled out what he was doing and why he was doing it, markets would respond big time. And they could buy and sell treasury bills, especially the inflation-indexed variety, to enforce those expecatations.

    Risky? Maybe. But I think the merits of the idea coupled with the potential benefits far outweigh the expected costs.

  3. I believe there are phases in the grand scheme of rescuing the US' economy from its near collapse. Had inflation-targeting at 5-6% implemented, the creditors, foreign CB's, US banks,... would have absorbed a serious loss. In the case of US banks, it would have been fatal. Perhaps phase I was to restore to health the financial system. A year into >1 trillion purchases of non-agency and long-term Treasuries in effect has transferred questionable assets to stronger hands, has bought sufficient time for the US banks to improve their balance sheets. Phase II, inflation targeting at 5-6&, is now in play as evidenced by the increase number of op-eds, arguments by credible economists on both sides of the political spectrum calling for such policy. Even members of Congress have called for the continuation of Fed's purchase beyond the initial trillion in 2009. However, PIMCO's shift in its portfolio toward Treasuries and recent survey of capital flows, positive for bond funds and negative for stock funds, signal that the market still bets on the ravaging deflationary forces. Reversing this trend and its expectation is the task for the Fed this year and beyond.

  4. Hi Michael, thank you very much fo the interesting entry.

    I agree with your argument that the first best solution "should have been put up front by the scholars that know this to be true". So I do not agree with Prof. Krubman's first point (convincing policymakers). It cannot be a reason why we give up the first best solution.

    I am doubtful, however, if the "first best" solution is really the first best feasible option, because of the dynamic inconsistency KTM1 mentioned. It is optiomal for a central bank to promise a future higher inflation. But after the recession is over (with higher expected inflation), it is no longer optimal to do so - the central bank (CB) happily goes back to be an inflation fighter. Markets rationally expect this and thus nobody believes the CB's promise (believe me, the markets are to some extent rational!).

    As Kydland and Prescott argue, a central bank cannot commit to even the inflation rate where the central bank believes optimal. I believe this idea is now widely accepted, and therefore we have an active discussion about inflation target - if we can simply assume that markets believe what the CB declares, we do not need inflation targets in the first place.

    Well, if we have an inflation target, then the central bank can commit to higher inflation? I am pessimistic. In order to commit to higher inflation than optimal rate that a CB believes, the CB needs to prepare a penalty in the cases it withdraws the promise (higher inflation). It can be a reputation loss as an inflation fighter, or a direct (non) pecuniary penalty such as Walsh (1995, AER) argues. The former option of course does not work here - promising higher inflation contradicts to be a credible inflation fighter. The latter is very difficult, as Walsh himself argues this is a useful fiction.

    It means that a central bank needs a drastic measure to commit to higher inflation. E.g. giving up descretion completely and legislating fixed interest rate (this is slightly similar to Athey 2005, Econometrica). I do not think this is a realistic idea, though.

    And I cannot agree with your purchasing Inflation Indexed Bonds (IIB) idea. Even if a CB monopolizes the IIBs, and raises so-called "measured expected inflation", it has no impact to the expected inflation in real economy. Furthermore, as you said the IIB market is faily illiquid and every market participant knows that - i.e. even as a measure, it is unreliable.

  5. Just a small correction: 5th paragraph line 5: promising higher inflation itself does not contradict to be an inflation fighter. Carrying out the promise after the economy goes back to normal (leaving high inflation) contradicts - and the CB (and possibly the government too, if the promised inflation is too high) has no incentive to do so.This is the reason why the strong enforcement power is needed, and the exogenous enforcement can be inefficient, as Athey argue.


Post a Comment

Popular posts from this blog

Nonlinear Temperature Effects Indicate Severe Damages to U.S. Crop Yields Under Climate Change

Commodity Prices and the Fed