We simultaneously expect too much and too little of monetary policy. It’s better to expect too much.
A few months back I argued that it was mistake to use a falling unemployment rate as evidence that monetary policy has been too tight in recent years. Unemployment was also declining in the late 1960s and again in the late 1970s, and yet monetary policy was certainly not too tight. I do think policy has been a bit too tight in the past few years, but mostly because inflation has undershot the 2% target and more reliable indicators such as NGDP growth have also remained subdued.
Recent posts by Tyler Cowen and Alex Tabarrok make a similar argument. Here’s Tyler:
Slow labor market recovery does not have to mean the core fix is or was nominal in nature, even if the original negative shock was nominal:
A negative nominal shock was associated with a big rise in European unemployment in the 1970s and 1980s, but it turns out that part of the rise was structural, not fixable with monetary stimulus.
Alex adds the following:
The natural rate can change over time, even in a sustained direction, as the structural characteristics of the economy change, as demand, supply, demographics, information and so forth change. Change does not mean disequilibrium. When the production of apples is bigger this year than last year we don’t jump to the conclusion that last year the apple market was out of equilibrium. Similarly, the fact that unemployment was lower this year than last year does not mean that we weren’t at full employment last year.
All this is true. At the same time, I’m actually rather heartened to see so much scrutiny of recent Fed actions. Even if we now expect a bit too much from the Fed, that’s much better than back in 2008 when we expected almost nothing from the Fed, even as its policies were pushing NGDP sharply lower and needlessly pushing millions into unemployment. At the time, only a few of us were pointing to the disastrous Fed actions, such as paying interest on bank reserves, and also refusing to cut interest rates in May, June, July, August and September, errors that are now widely recognized by almost all well-informed observers.
As time goes by it becomes increasingly obvious that the US business cycle is largely monetary in nature, and that a monetary policy that stabilized 12 or 24-month forward NGDP expectations would dramatically moderate the business cycle. Interestingly, I see signs that the Fed itself is becoming increasingly aware of this fact, which bodes well for the future. Still, it’s too soon to declare victory, and the bloggers who obsess over every slight Fed mistake, er, I mean “misunderstanding“, are actually performing a valuable public service. They are helping to keep the Fed on its toes.
In 2008, the Fed ignored market warnings and we had a severe recession when the equilibrium interest rate plunged. In 2019, the Fed paid attention to market warnings and we had no recession when equilibrium interest rates plunged—even though many pundits predicted a recession, and even though we would have had a recession if the Fed had relied on its traditional “Phillips Curve” models.
In 2008, the Fed said nothing about making up any near-term shortfalls in aggregate demand. Today, Fed official frequently talk about the need to make up shortfalls in inflation with future above target inflation.
In 2008, Fed officials obsessed about above target inflation even as plunging NGDP growth was completely ignored. Today, many Fed officials talk about the value of NGDP as a policy indicator, and warn that inflation can be a misleading indicator.
I don’t know about you, but I call that progress. But then I’m someone who has always favored paying attention to market forecasts of aggregate demand, engaging in level targeting, and prioritizing NGDP over inflation.
PS. I expect a lot of the Fed. That’s why I’m on record predicting only one recession in the next 37 years.
READER COMMENTS
Thaomas
Jan 7 2020 at 5:31am
Alex is correct as a casting of doubt (although not in totally refuting) on the idea that there may not still be some “out of equilibrium” unemployment. Nevertheless, the price level is still below its level if the Fed had hit its “inflation” targets since 2008, inflation-adjusted bond yields suggest that market expectations are still below 2% p.a., and the prime age employment ratio is still below its maximum]
The larger picture, however remains the same, that the Fed was very much mistaken in not having provided much more stimulus much earlier and winding down QE too soon, before hitting a 2% p.a. PL target. Notwithstanding movement in that direction, it’s still not certain that the Fed has committed to a PL trend + maximum employment target.
PS, Scott. Something is happening that prevents my comments at “Money Illusion” from appearing. How do I get in touch with the webmaster? Thanks.
Scott Sumner
Jan 7 2020 at 12:24pm
I fixed it.
Thaomas
Jan 8 2020 at 1:51am
Thanks! 🙂
And apologies to Econ Log readers for the interjection.
Mike Sandifer
Jan 7 2020 at 6:34am
I’m curious as to how economists model nominal fluctuations versus adjusting wages. There seems to be widespread consensus that, surely we can’t have very tight money right now, because wages should have adjusted since the Great Recession. This both assumes money can’t be tight on an ongoing basis, and a specific time frame for wage adjustment. Yet, no economist I ask or read ventures even a guess as to how long wages take to adjust. What I do see however, are apparently baseless opinions from economists about natural rates of unemployment, potential real GDP, and the current and past levels of the neutral interest rate. From outside the profession anyway, there seems to be little reason economists actually know much about these topics.
What model or models should convince us that economists can tell the difference between real and nominal factors affecting growth in the longer term? Too many still can’t even seem to tell the difference between real and nominal shocks.
Scott Sumner
Jan 7 2020 at 12:25pm
There’s a lot of empirical evidence for the long run neutrality of money.
Mike Sandifer
Jan 7 2020 at 2:27pm
Yes, I agree there’s a lot of empirical evidence for the long-run neutrality of money, but that doesn’t address my points. How long is the long run, and under what circumstances?
More importantly, to me, is it possible for money demand to outstrip the expected supply indefinitely? For example, if the growth of the money supply is continually expected to remain below that of real GDP growth potential + Inflation, can money be tight indefinitely?
Scott Sumner
Jan 7 2020 at 11:28pm
I don’t know what you mean by money demand outstripping supply. How would we know this?
Mike Sandifer
Jan 8 2020 at 12:33pm
One indicator of money demand outstripping supply is weak or falling money velocity, though no economist seems to take money velocity very seriously anymore.
And though no economist buys it, if I’m right about r* = NGDP growth expectations, in equilibrium, there’s another indicator.
Basically, my impression is, absent shocks, the growth of the money supply should at least equal RGDP potential + some expected sustainable inflation rate (or perhaps even slight deflation, after expectations adjust). Hence, there’s a lower limit, at least in some short-run(perhaps longer than one might expect) on optimal money supply growth, to avoid RGP growth falling below potential.
More simply, explicitly, the idea is that money demand typically equals real GDP potential + the expected inflation rate.
This idea originally came to me from Milton Friedman, who, if I recall correctly, once said in one of his Freedom to Choose episodes that the money supply needs to grow at least as fast as the economy.
So, if money supply growth is consistently below RGDP growth potential + expected inflation, then tight money can persist for years, with very slow, if any adjustment.
Scott Sumner
Jan 10 2020 at 3:53am
I’m not sure money velocity is helpful in determining the stance of policy.
Thaomas
Jan 8 2020 at 1:57am
Isn’t “neutrality of money” inconsistent with favoring one kind of monetary regime over another? Would not real GDP have been greater 2008-2020 with better monetary policy (NGDP or PL+ employment targeting)?
Scott Sumner
Jan 10 2020 at 3:52am
I am referring to long run money neutrality, not short run.
Mike Sandifer
Jan 10 2020 at 12:45pm
There are certainly readily noticeable negative correlations between changes in velocity and changes in the unemployment rate, but obviously causation can’t simply be assumed. A paleo-monetarist would assume causation, I think.
This is a tough issue on the surface, because both tight money and secular stagnation could plausibly explain the state of the variables in question. And I think both apply, but tight money is the larger factor.
I’m working on a method to determine changes in the neutral interest rate. I’m using the equation I developed to translate changes in NGDP into changes in broad stock indexes, and vice versa, to compare implied changes in NGDP expectation via the S&P 500 to changes in short-term nominal T-rates. The idea is that 1 year Treasury rates basically just track the Fed Funds rate, so that changes 2 and 3 year rates reflect any expected changes in the Fed Funds rate. Hence, add the change in the short term T-rates to the implied change in NGDP apparent in the change in the S&P 500, and you have an estimate of the change in the neutral interest rate. That assumes the neutral interest rate changes 1-for-1 with NGDP growth expectations.
I don’t expect anyone else to have confidence in my GDP changes-to-broad stock index change model, because I haven’t presented precise enough evidence to support it. But, I’m making progress on that front, and I’m thinking of new ways to test the hypothesis that r* should equal GDP growth expectations. An obvious implication is that the real interest rate should equal RGDP growth expectations, in equilibrium. So, perhaps analyzing real shocks can be useful, with respect to how real rates react. Real rates did roughly equal RGDP on average during the Great Moderation.
I also point out some seemingly structural relationships between interest rates and GDP that I don’t see anyone else discussing. For example, I’ve yet to see risk free rates be above GDP without a downturn in GDP. That’s consistent, though not uniquely consistent with the hypothesis that r* = NGDP growth expectations in equilibrium, though it doesn’t address whether interest rates can be below GDP in equilibrium.
Anyway, there’s more work to be done. And don’t get me started on the implications for representative agents!
Brian Donohue
Jan 7 2020 at 10:32am
Excellent post. You have enjoyed a lot of vindication over the past decade, there were reasons to be worried about Powell as a newbie, and I do think he stumbled a couple times on forward guidance last year, but, as you say, equilibrium interest rates fell a lot and quickly and the Fed never let itself get too far behind the curve. As someone who spent a lot of 2019 complaining about Jerome, I award him 2.5 cheers for a job pretty well done last year.
Garrett
Jan 7 2020 at 3:49pm
Scott,
What are your thoughts on the shape of the “TIPS spread curve”? Right now the 30y spread is 178 and the 10y spread is 175, but the 5y spread is 168 and the 2y spread is 148. Do you expect it to flatten over the next decade with the front-end increasing?
Scott Sumner
Jan 7 2020 at 11:26pm
Yes.
stoneybatter
Jan 8 2020 at 9:19pm
Scott, is this just because you expect expectations to roll forward? I.e. the 2y inflation breakeven will, in a few years, reflect the already-higher expectations embedded in the 5y breakeven today? Or is this a rare (unprecedented?) case of you thinking the market is mispriced?
Scott Sumner
Jan 10 2020 at 3:51am
I’m not sure that the TIPS spread is precisely equal to the market expectation of inflation. But it’s partly because I expect actual inflation to be close to 2% in the long run. (I think that’s what you mean by roll forward.)
Comments are closed.