The Fed releases minutes of their meetings with just a few weeks delay. But the full transcript is kept secret for “an extended period” to insure more candor in the discussions. Matt Yglesias recently suggested I look at the recently released FOMC transcript from February 2005, which contains an extended discussion of the pros and cons of inflation targeting. There’s a lot of interesting stuff:
1. There is pretty general agreement about the Fed’s long term inflation objective. They’d like to see about 1% actual inflation. Because they believe the PCE overstates inflation by 0.5% and the CPI overstates inflation by about 1.0%, they’d like to see about 1.5% PCE inflation or 2.0% CPI inflation. I was quite surprised by the uniformity of views on this issue. I don’t recall anyone who put forward a different number, just a few who discussed the advantages of ranges (1% to 3%) rather that point targets.
2. The committee does not favor a hard inflation target. About half the members (led by Bernanke) favor giving the public a specific quasi-official number as a sort of goal, with the understanding that the Fed might have to deviate in the short run due to financial crises and/or supply shocks. Other readers might disagree with this characterization, as they danced around the issue in a very tentative way, knowing it was actually Congress’s prerogative to set any formal policy goal. The FOMC was obviously a bit distrustful of having Congress get heavily involved in monetary policy.
3. I now feel much more confident in asserting that the Fed’s implicit target is 1.5% PCE and 2.0% CPI inflation. Period.
4. They also clearly indicated that a higher than 2% inflation rate might be appropriate under one of two conditions:
a. Financial crisis
b. Adverse supply shock
And they clearly acted on that belief in 2007 and 2008 when they cut the Fed funds rate during the sub-prime crisis, despite above 2% inflation. However, I think they may have tragically misunderstood the implications of their supply shock arguments, and that’s the issue I’d like to focus on. Consider the following quotation from Cathy Minehan.
So my policy preference for a given level or path of inflation would not be identical all the time. It would depend on what is happening in the real economy, just as in the first half of 2004 we tolerated rather rapid price growth on the basis of our calculation of the degree of excess capacity in the real economy and the temporary nature of the energy price increases. I know that over the long run there’s no tradeoff between growth and inflation and that price stability, however defined, is the best contribution monetary policy can make to economic prospects. But in the short run, when supply shocks can dominate, there can be tradeoffs.
She is saying that if unemployment is sort of high, say 7%, and inflation was 3% due to energy price increases, you would not try to immediately bring inflation down to 2% if it meant pushing unemployment up to 9%. And I am pretty sure that everyone at the Fed agrees with that. However the inescapable implication of the argument is that if you have 9% unemployment and 2% inflation, you’d be better off pushing inflation up to 3% if it would reduce unemployment down to 7%. The argument is completely symmetrical.
I saw no signs that the Fed understood this implication in 2005, and I still don’t. Indeed when Brad DeLong asked Bernanke in 2009 why they don’t raise their inflation target to 3% to boost growth, Bernanke said it would be a very bad idea.
The problem here is that either there is a dual mandate, or there isn’t. If there is, then the Fed would want to allow slightly higher than 2% inflation during adverse supply shocks (and vice versa.) And I think they do. If not, they should aim for 2% inflation come hell or high water. And that’s clearly not what they do. But this dual mandate idea also implies that if unemployment is extremely high due to deficient demand, the Fed should try to aim for above 2% inflation to try to bring it down. Yet the Fed seems to vehemently deny any intention of aiming for above 2% inflation during a period of 9.5% unemployment. Why? What sort of social welfare function allows for accommodating supply shocks (or financial crises) but doesn’t allow for an aggressive move against deficient demand?
Of course all this confusion would be eliminated with . . . NGDP targeting.
[William Poole] The one other case is not a U.S. case but Japan. I think the Japanese have also suffered from not being very clear that they do not want deflation, particularly as the situation developed in the early 1990s. The markets were left quite at sea in trying to figure out where Japanese monetary policy was going to go.
CHAIRMAN GREENSPAN. I think they still are.
MR. POOLE. Yes, they probably still are.
And this probably describes September 2008:
A third potential cost could arise if your credibility were seen to be diminished when inflation differed from the stated objective. Finally, a commitment to an explicit price objective could constrain future actions of the FOMC in an unhelpful manner. For example, the Committee might feel inhibited in responding as aggressively as it would like to a financial crisis if inflation were already to the high side of the Committee’s objective. [Wilcox, et al.]
I don’t quite agree with this, but it is interesting:
[Gramlich] There’s an old experiment that I learned about in graduate school from Richard Ruggles, who used to be a professor at Yale: Offer somebody $10,000 and the choice of ordering from a catalog of all goods and services made this year or five years ago, and take a poll on which option they vote for. Try it. You all give talks to Chambers of Commerce and so forth. I’vebeen doing it for years, and people will consistently vote for the current menu. Obviously, this experiment has to be done at a much higher level of scientific rigor. But I think in the utility sense, even core PCE rates as high as 3 percent may be more or less consistent with price stability, given the great difficulty we have in dealing with technological change in price indexes.
The problem is that consumption is a social activity. Do they mean you have the old catalogue and everyone else has the new one, or everyone has the old catalogue and their current nominal income.
And I’m guessing that someone like Bob Murphy won’t agree with the claim by Greenspan that they “lucked out” after easy money was pursued in 2004 despite the fact that the inflation target called for tightening:
Take the spring of 2004, when we were sitting there with a significant acceleration in core inflation. With a targeted range for inflation, conceivably we would have breached one of the limits of the target range. And the question would have automatically arisen, “Well, what are we going to do about this?” My answer would have been—and indeed at the time was—“nothing.” The reason was that I viewed the rise in prices as wholly the consequence of a rise in profit margins. But that rise in profit margins was sufficiently quick to result in a projection of core final goods prices that would be above any reasonable target we’ve been discussing today.
The point is this: That was a particular case where we knew that unit costs were not moving and that the rise in inflation was wholly a mechanical result of a one-shot event, which couldn’t continue unless unit costs started to accelerate. We lucked out.
I think people overstate the role of easy money in the sub-prime fiasco, but I included this quotation because I know that most people disagree with me.
On other topics, Tyler Cowen recently made this observation in response to data showing brisk growth in industrial output:
Yet the labor market is still “eh.” Here is more, but again note it is wrong to reject the AD factor altogether, though it seems to be becoming less relevant over time. Arguably AD and AS are interacting in unusual and presumably deleterious ways.
I have read too many blog posts attacking a caricatured version of either RBC theory or a narrowly defined notion of “structural unemployment” which requires excess demand for labor in significant parts of the economy. As Arnold Kling points out, the labor market shock can be asymmetric in its effects.
From a different direction, here is Scott Sumner criticizing the recalculation argument. I read Scott as establishing the conclusion that both AD and AS must be at work.
I don’t quite agree with the first part, where Tyler discusses the weak labor market. I’ll simply link to my previous post, which explains why. I do agree that our problems are partly real/structural/supply-side. But I still think it is mostly demand-side, and that demand stimulus would even indirectly improve the supply-side (i.e., UI would be reduced below 99 weeks more quickly, which would further normalize the labor market.)
I was flabbergasted to see my first positive link from Paul Krugman. Of course our views on the need for demand stimulus are similar, but we always seem to tangle on the nuances of some issue. Indeed perhaps 1.5 positive cites, as this Krugman post is probably referring to me and Beckworth. I recently published something in the National Review, and Beckworth published in the Wall StreetJournal.
Now I have a sudden fear I will lose all my right-wing friends. I can just imagine what Bob Murphy will do with this. And of course I’m way too right-wing to ever be embraced by the left. I’ll be in limbo, along with Bartlett, Frum, Lindsey, Wilkinson, etc.
Oh well, que sera, sera.
PS. My consolation is that if Milton Friedman were alive, he’d be in the same awkward position.
PPS: I got a Business Week mention.
Voices from the past
Sat, 15 Jan 2011 20:49:15 GMT