February 01, 2006
Should the Fed worry about going too far?
During each of the past three episodes of monetary tightening, the Fed quickly realized that they had gone too far, and were forced to reduce interest rates after only a short time.... Given that changes in the Federal Funds rate are estimated to take 12 to 18 months to have their full impact on the economy, it seems fair to say that in each of these instances Greenspan probably wished that he had stopped raising rates a bit sooner than he actually did.
The growth rate of the economy in 1995 was indeed weak in the context of the surrounding years. In particular, GDP growth in the first half of the year was woeful. But the policy enacted in 1994 and 1995 arguably set the stage for the years that followed, years characterized by better-than-average growth and stable inflation.
So let me ask this question. Suppose that, in retrospect, we find that the FOMC's current round of rate hikes went a little too far. Suppose that we find that fourth-quarter 2005 GDP growth was not an aberration, but the first of several quarters of sub par economic expansion. But suppose further that we find that the extraordinary sequence of energy-price shocks over the current recovery did not bring a persistent increase in the overall inflation trend. And suppose we find that, following two or three quarters of soft economic activity, GDP expanded at rates between 3 and 4 percent for years after.
Would you complain?
The graph above displays Dave's preferred measure of monetary tightness, which is the spread between the yield on a 10-year U.S. Treasury and the fed funds rate. There are a number of other episodes-- namely 1967, 1984, and 1998-- that share the feature that Dave highlights for 1995. In each of these episodes, the spread became about as low as it is now or even negative, but any slowdown in growth proved to be both mild and temporary, and perhaps worth the cost if such monetary tightening succeeded in preventing a resurgence of inflation.
Against these one could alternatively have picked other episodes when again the spread reached levels such as we currently see-- like 1956, 1960, 1968, 1973, 1978, 1980, 1989, and 2000-- when there was not just a slowdown in the growth rate but a fall in output and an economic recession. Presumably Dave would agree that, at least in some of these episodes, the Fed did indeed regret tightening as much as it did.
Now, one of the points that I understood Kash to be making (also here) was that, given the current state of our ability to make economic forecasts, it's hard to know ahead of time which we're talking about-- is the Fed tightening just enough to produce a slowdown in the growth rate, or will we actually see a recession?
I've been urging caution for the Fed some time. This is because I have been more concerned about the potential output effects of the oil price increases and the narrowing yield spread than the Fed has been. The 2005:Q4 figures seem closer to my predictions than the Fed's, though I admit that other data have been more favorable. I honestly don't know how the rest of the year is going to turn out.
I don't know, and neither does the Fed. And that's the point. We do know that, in situations like the present one, sometimes the economy has ended up in a recession, and sometimes it hasn't. That ignorance, in my mind, suggests caution. In my opinion, the Fed should not have attempted to fine-tune the recovery of 2003, stimulating the economy as they did until they saw an employment growth rate they liked. And the Fed should not be assuming that they have the wisdom to reverse course in time if a significant slowdown starts to materialize. If you're not that sure what lies ahead, you don't drive quite so aggressively.
My recommendation is that the Fed not act on the assumption that they know what the economy will be doing 6 months from now. And I believe that, if you act on that assumption, your current policy decision would be to wait for some more concrete indications before moving interest rates up any higher.
So, to rephrase Dave's question-- if the monetary tightening of 2005 turns into the recession of 2006, would I complain? Sure I would-- that's my job.
Posted by James Hamilton at February 1, 2006 08:20 PMdigg this | reddit
Ultimately, I think the problem is merely that the Fed's key control mechanism has effectively three modes: raise rates, lower rates, and do nothing.They also have limited knowledge (though it's better than most economic participants) and their decisions have a lag time measured in months. Given that, they're going to make mistakes.
From my recall of control theory, I think you can usually settle on an optimal control parameter, if you are allowed to overshoot. So it's not clear to me why overshooting is such a problem. Second, it's not clear why a recession is a problem to avoid. It may turn out that recession is precisely where the US economy needs to be to optimize long term economic growth.
Posted by: Karl Hallowell at February 2, 2006 06:25 AM
er, I meant, you can usually settle on a optimal value for your control parameter *quicker* if you're allowed to overshoot.
Posted by: Karl Hallowell at February 2, 2006 06:26 AM
In terms of growth, I fully agree with you: the FOMC (or anyone else) has very limited ability to forecast, thus they should be careful in raising rates when there are signs of weakness. But what about inflation? Wouldn't the argument flip on inflation? Since they can't forecast inflation very well, they should be careful in stopping raising rates when there are sign of higher inflation? (Core CPI was 2.2 percent for 2005, exactly as for 2004, but optimally trimmed mean CPI, which is supposed to be a better measure of underlying inflaitonary pressure, went from 2.4 in 2004 to 2.8 in 2005; similarly, median CPI went from 2.2 in 2004 to 2.4 in 2005. Further, most forecasters are predicting 2006 core CPI will be slightly higher than 2005).
Posted by: pat at February 2, 2006 08:19 AM
How unique is the current situation, with respect to high consumer debt on variable interest rate devices (credit cards and ARMs)? This seems like a particularly risky time for interest rate overshoot ... but maybe there is never a good time.
Posted by: odograph at February 2, 2006 08:27 AM
I hate to rain on people's parade (especially since I like to see the glass half full myself), but you must realize that the fed has just brought the fed funds rate back to a "normal" level.
Most economists and pundits have been acting as if we are going to "high", when actually, we are returning to the "middle of the road", when it comes to short term interest rates, historically.
Due to the many factors that are depressing our economic progress (i.e. debt, tax situation, consumption vs. personal savings, etc..) are making the market sensitive enough so that in time, the fed WILL NOT be able to raise the fed funds rate high enough above normalacy to tighten WITHOUT sending the economy into a recession. This is the real problem. I like that greenspan has taken baby steps to get us here, but I am troubled by the fact that people are saying, "too far, too much, too high", when by historical standards we are approaching "even."
Just something to ponder on.
Posted by: Andy at February 2, 2006 08:40 AM
Andy, what I view as definitely not normal in the current situation is the relation of the fed funds rate to the 10-year rate.
I agree with you that the situation in 2003 was not normal and that rates definitely had to come up from there. My preference would have been never to have let the funds rate fall below 2% in the first place.
If the Fed were doing its job properly, we should not be seeing 400 basis point swings in the short-term interest rate over relatively short periods.
Posted by: JDH at February 2, 2006 10:03 AM
Pat, my view is that both the bouts of inflation and many recessions have a common contributing factor, which is that the Fed puts more faith in its ability to forecast than is warranted. When the economy is weak, the Fed is too confident that they can stimulate it a little more without causing inflation. It turns out they overdo it and bring about the inflation. Then they are too confident that they can tame inflation without causing a serious downturn. They overdo that as well and we end up in a recession, and the cycle repeats.
Currently we're in the "we're confident we can tame inflation without causing too much harm" part of the cycle again. So how do we avoid deja vu all over again?
My answer is, first, the yield curve slope has been a pretty reliable indicator of overdoing it in the past so it shouldn't be ignored now. Second, if I'm wrong and we do need some more rate hikes to get back to Andy's normal real rate, where is that? Surely not more than another 100 basis points up. So what's the rush in getting that extra 100 basis points? Why not hold at 4.5 for a few months, see if longer yields start to catch up a little more, and if so, then you can add on a few more rate hikes later. If you're not that confident exactly where you want to end up, I don't think you'd be in all that big a rush to get there.
Posted by: JDH at February 2, 2006 10:26 AM
"My preference would have been never to have let the funds rate fall below 2% in the first place."
But it did get to that level, and one thing that interests me is whether it's possible to make an orderly recovery from a situation where the economy has been over-stimulated.
Once speculative excesses take hold and credit growth becomes exponential, how do you choke off that excess other than through severe tightening? I'm not sure it's enough to ratchet up a little, then hang around to see how things are going. It's too easy for debt to spiral out of control. Surely there's an argument that some sort of painful shock is a necessity to restore a sense of rationality?
...which I guess is a verbose way of saying: "The damage has been done - you're damned if you do, and you're damned if you don't."
Posted by: FTX at February 2, 2006 11:18 AM
It's like using your hand to balancing a yardstick on its end. You've got to move your hand fast to keep it in balance. You overshoot intentionally. Undershooting causes things to go out of control because of positive feedback. It's much better to overshoot because you can overshoot back the other way. It won't work to consistently undershoot.
I don't know if the economy is really like this, but try it out sometime and you'll see that at least in this example, overshooting is the only way to keep a system in control.
Posted by: Hal at February 2, 2006 11:27 AM
JDH: "my view is that both the bouts of inflation and many recessions have a common contributing factor, which is that the Fed puts more faith in its ability to forecast than is warranted."
I am in full agreement with you here. In fact, the reason I worry so much about inflation is because I worry that the FOMC will overeact if inflatin comes in higher than expected. While there is no argument that Mr Bernanke is an excellent economist, his speeches in 2003 suggest that he had too much faith in forecasts "than is warranted." I hope he has learnt the lesson.
Posted by: pat at February 2, 2006 12:02 PM
in none of the comments does anyone explain the
function of prices in our economy.If the factors
of production i.e.labor,machinery are not fully
utilized then rising prices will serve as an incentive to increase output.Is that what your
commentators call inflation?What about the variation in prices among the components of the
price indices.Some are rising while others are
stable or falling.How can a blunt instrument like
monetary policy control only those prices that
are rising without deflating stable or falling
prices.Itwas only a few years ago thatMr.Greenspan was worried about deflation.
Finally,does anyone really know what changes in
prices are necessary to optimize the full utilization of resources in our economy?
Posted by: wilfred lumer at February 2, 2006 03:27 PM
I like the rebuff to 4.5% is normal.
I don't like the rebuff to 'so what's wrong with a recession' (if it contributes to our long term stability), so much (But I appreciate the light it throws on the emotive side of that word "recession". We could use less fear. We could.)
Then Hal forces me to look at undershooting/overshooting and I can't get ducks out of my head. Sooner or later (ok it is 14 hikes/nibbles later) the undershooting fails and we get our duck. Clunko.
No, it isn't that easy. We might overshoot and not only does the duck not fall, but the sky darkens, an ominous cloud appears, rain, lightning...hard working people are in the streets looking for something to eat...maybe an economist.
No, that isn't it. It's not those wiley ducks, but when to pause. I know it doesn't sound nearly as dramatic as shooting, but JH's suggestion of a pause now seems judicious.
And we won't have to eat him after all.
Posted by: calmo at February 2, 2006 08:01 PM
Confused here. If a recession "might" be needed, but we have no way of knowing that, how do we ever come to the conclusion that a recession is not something to be avoided?
Efforts to identify the point at which real economic damage results from inflation usually find limits well above the 2% upper target the Fed seems to like or the roughly 3%-3.5%% rise that we have experienced in the most extreme consumer price measures. It is hard to argue that the economy "needs" a recession when inflation is below the level at which trouble is thought to start, and inflation expectations are (U Michigan reported today) cooling.
Probably the calculation that Fed guys are making is the risk associated with each kind of error. Is the risk of overshoot great or small? Is the risk of undershoot great or small? Not too long ago, Fed officials were arguing publicly that the risk of overshoot is small. Whatever the right answer is for the ages, the answer Fed officials have said they are biased toward overshoot.
Posted by: kharris at February 3, 2006 12:49 PM
I have no formal training in economics. I read and invest.
A question for the true Students. How can we have wage inflation in a disinflationary world? If China won't do it India will. If India won't, Bangladesh will. Except for favored corporation executives worldwide, how can inflation take hold in wages? Commodity prices may (and are) rising, but this scenario leads to stagflation.
please, any clarifications for me?
Posted by: outsider at February 4, 2006 01:27 AM
Seems they always go too far
Posted by: mynewsbot at February 4, 2006 06:04 AM
The narrative sounds a tad histrionic to me. If the 25bp increases had been applied with a little less constancy, or less track record, I might have been able rouse myself to a little more interest in this overshoot/undershoot issue.
outsider notes the current inflation is not wage driven (and so the charts from 2000 on is structurally different from previous inflationary periods) nor will it be driven by wages in the future unless unions stage a miraculous recovery. So what clarification do you need outsider? You have the picture of deflation, immodest, unregulated executive compensation, and the question of how wage inflation can re-imbed itself strikes me as rhetorical given the Delphi proposal and the emerging asian economies.
Must be nice not to be weighed down with all the formal economics training and just see stuff straight away.
Posted by: calmo at February 6, 2006 01:51 AM