Sumner bashes Summers in the FT
Here’s my new op ed in the Financial Times. As you will see, I relied heavily on Christina and David Romer’s recent article entitled “The Most Dangerous Idea in Federal Reserve History: Monetary Policy Doesn’t Matter.” It’s an excellent article that uses some ideas that I’ve also toyed with, but puts them together in a much more powerful and coherent thesis.
HT: Ricardo
PS. I enjoyed this Caroline Baum piece on options for the new Fed chair.
Tags:
7. August 2013 at 22:57
Nice op-ed.
But perhaps it is time to stop referring to QE as “unconventional.”
If the globe’s major central banks engineer a world of ZLB—and they have been getting there for 30 years—then QE is the default norm.
The next stage of monetary policy is how to handle the QE hoards at central banks. Hold to maturity?
Also, do we need a change in our accounting of bonds? Right now, when the Fed holds a bond to maturity, they collect the principle, and then extinguish the cash. The money collected goes into a black hole.
Maybe better they return the money to the Treasury? That would be true monetization of debt.
8. August 2013 at 02:28
Well written.
8. August 2013 at 02:31
Would I rain on anyone’s parade if I pointed out the obvious fact that rates have never actually been at the “zero bound”, making this whole debate about what should be done at the zero bound a moot point?
No lender would agree to lend money at zero percent interest.
0.25% or 0.50% are not zero bound rates. Those are positive rates.
8. August 2013 at 03:04
So far, at 7 am EDT on Thursday, the comments on your op-ed over there are uniformly awful. It is discouraging. But at least they ran the op-ed.
8. August 2013 at 03:44
Geoff:
By zero bound, I mean interest rates low enough that the central bank has little fire power using “conventional” tools, such as lower rates. The Western world is there, or close, now.
8. August 2013 at 04:10
Ben, Yes, QE should not be viewed as unconventional. But that’s the world I live in.
dtoh, Thanks. I thought it was even better before they cut out 100 words.
Jeff, I generally ignore comments outside of my own blog. Most people have no understanding of monetary policy. Indeed most think policy has been easy over the past 5 years.
8. August 2013 at 06:04
I like Caroline Baum’s argument for appointing a dolphin Fed chair.
1- A dolphin has a large brain for filtering acoustical signals (i.e., noise) from the market.
2- A dolphin can swim backwards and clap its flippers together for Congress.
3- A dolphin can balance growth and inflation (i.e., NGDP) on the top of its nose.
8. August 2013 at 07:18
Hey Scott,
Thanks again for all your tireless work. Yes, ‘Most people have no understanding of monetary policy.’ But, the ability to articulate and defend a policy in a way that is at least understandable to a reasonably intelligent and interested layman is both achievable and crucial in a functioning democracy, and I think a lot of what you’ve done on this blog is in pursuit of that laudable goal. The alternative, hiding behind the wizard’s robe, is for lamos. (As someone who works in a highly technical field, I am very familiar with this type.)
The Romer paper is great, although the 70s experience, it seems to me, had less to do with ‘pessimism’ about monetary policy than a dread of the short-term consequences of monetary contraction. I’m sure there weren’t a gaggle of Austrians that policymakers had to hack their way through.
I definitely agree with this from the Romers: “Nonetheless, it seems hard to assign pessimism about the power of monetary policy a large role in the crisis itself.”
And here’s some (calendar-year) data that I think supports this view:
Change in
Year NGDP RGDP Deflator
2003 6.8% 4.6% 2.1%
2004 6.9% 3.5% 3.2%
2005 6.5% 2.9% 3.5%
2006 4.3% 1.4% 2.9%
2007 3.1% 0.4% 2.7%
2008 (2.0%) (4.1%) 2.2%
2009 2.0% 1.6% 0.4%
2010 3.9% 2.0% 1.8%
2011 5.2% 3.2% 2.0%
2012 3.1% 1.2% 1.8%
I’m interested in understanding more particularly your view of the ‘relative impact’ of what I’ll call ‘a period of insane profligacy during much of the 2000s’ and ‘lousy monetary policy’ in contributing to the recent economic woes. In your view, what would the numbers in the above table look like in a world where you ran the zoo?
8. August 2013 at 07:48
‘…the 70s experience, it seems to me, had less to do with ‘pessimism’ about monetary policy than a dread of the short-term consequences of monetary contraction.’
Read some of Arthur Burns’ congressional testimony. It got to be quite comical as he denied he could do anything about inflation.
8. August 2013 at 08:13
“I’m interested in understanding more particularly your view of the ‘relative impact’ of what I’ll call ‘a period of insane profligacy during much of the 2000s’ and ‘lousy monetary policy’ in contributing to the recent economic woes. In your view, what would the numbers in the above table look like in a world where you ran the zoo?”
>> Interesting question. If I could add to this. During this period of “insane profligacy,” the ratio of housing prices to median incomes grew well above trend. Presumably Scott would have pushed for looser monetary policy around ’07/’08 in an effort to maintain a higher NGDP growth rate. I understand Scott doesn’t believe in bubbles. Ok, fine, but I would think most would agree that the higher growth rate in housing prices vs growth in median incomes becomes unsustainable at some point. To what extent, if any, does a looser monetary policy carried out in the name of NGDP targeting account for this structural imbalance? In other words, conceivably if Bernanke didn’t “tighten” and NGDP growth remained around trend growth from 2007 onward, how would the divergence of housing prices away from median incomes (not to mention the continued expansion in the underlying credit markets and securitization via derivatives) still not have manifested in real economic problems at some point in the future?
8. August 2013 at 09:26
Some interesting new suggestions on ZLB monetary policy (I am sympathetic to the switching of tongues): http://www.ft.com/intl/cms/s/0/2521a202-fde2-11e2-a5b1-00144feabdc0.html#axzz2bOatDm7F
(HT Tyler Cowen on Twitter)
8. August 2013 at 09:31
Oh, I see Caroline Baum already did a better one.
8. August 2013 at 10:34
Brian, That data is actually not accurate. But let’s suppose it was. Growth went from 5.5% per year in the first 5 year period, to less than 2.5% per year during the second 5 year period, using your figures. I favor NGDPLT at a stable rate. The actual rate doesn’t matter very much, anywhere from 3% to 6% would be fine. But keep it stable.
And I agree with Patrick, I recall how Keynesians mocked the view that inflation could be controlled by reducing money supply growth. They said an “incomes policy” was needed. Even Nixon went for price controls.
Having said that, concern about employment was also a factor.
Mike, You mischaracterized my view. I don’t argue that money should have been easier in 2007, or that the Fed should keep housing bubbles from popping. The US housing “bubble” mostly popped from January 2006 to April 2008. And that was a necessary adjustment after the sharp fall in expected population growth. Because monetary policy was appropriate, unemployment only rose from 4.7% to 4.9%. So I reject the premise of your question.
8. August 2013 at 10:37
Thanks Saturos.
8. August 2013 at 10:57
Thanks, Scott. My NGDP data came from FRED:
http://research.stlouisfed.org/fred2/series/GDP/downloaddata?cid=106
And I rechecked my ability to divide A by B, which appears sound, so…take it up with FRED, I guess.
And my 70s quibble was just that. I was raised to believe that inflation is a monetary phenomenon, so I find Burns et al astonishing here, and I happily stand corrected.
What I REALLY want to know is: what percent of the (nebulous metric) ‘pain’ that we have suffered over the past five years is due to monetary policy blunders? Because the vibe I’ve gotten from you is that you believe it is close to 100%.
8. August 2013 at 11:55
Brian Donohue,
I don’t know what you’re doing to that data but hit hurts my eyes.
Change in (%)
Year NGDP RGDP Deflator
2003 4.8 2.8 2.0
2004 6.6 3.8 2.7
2005 6.7 3.4 3.2
2006 5.8 2.7 3.1
2007 4.5 1.8 2.7
2008 1.7 (0.3)2.0
2009(2.1)(2.8)0.8
2010 3.7 2.5 1.2
2011 3.8 1.8 2.0
2012 4.6 2.8 1.7
http://research.stlouisfed.org/fred2/graph/?graph_id=132132&category_id=0
8. August 2013 at 12:05
Brian, Don’t blame the Fred for your mistakes.
Mark, You beat me to it. You’d think after 5 years of saying that in 2009 NGDP fell at the fastest rate since 1938, my readers would give me a bit of slack.
8. August 2013 at 12:42
Great column.
The comments section of your article in the FT is comical, in that supposedly knowledgeable FT readers are incredulous to the fact that Scott Sumner didn’t just drop in from Alpha Centauri to teach these Earthlings a thing or two about monetary policy.
Haha changing hearts and minds one puzzled human at a time.
8. August 2013 at 12:45
OK, I chuckled.
‘A bit of slack’? C’mon man, I sucked up to you pretty hard a couple comments upstream.
Mark, your link didn’t work for me. No matter, the story is roughly similar.
Scott, I’ve come a long way toward your worldview here. Not far enough to be considered an acolyte though. Sadly, to you the world appears to be filled only with acolytes and enemies. 😉
8. August 2013 at 13:01
Brian Donohue,
You wrote:
“I’m interested in understanding more particularly your view of the ‘relative impact’ of what I’ll call ‘a period of insane profligacy during much of the 2000s’ and ‘lousy monetary policy’ in contributing to the recent economic woes.”
This wasn’t addressed at me but I remember very well the Job Loss Recovery:
http://research.stlouisfed.org/fred2/graph/?graph_id=121784&category_id=0
From January 2001 to July 2003 a period of 30 months there were only four months where the private sector created jobs. It wasn’t until February 2005 that private sector employment exceeded the previous record set in December 2000. That was the longest period (until now) without an increase in private sector jobs since the Great Depression.
Thanks to it I changed careers and became an economist.
If there was insane profligacy in the 2000s I certainly didn’t witness it.
“Mark, your link didn’t work for me. No matter, the story is roughly similar.”
The link works fine for other people so there must be something wrong with your computer. And if you think those numbers are roughly similar, remind me not to hire you to do any data crunching. (Are you sure you went to the UD?)
8. August 2013 at 13:31
Mark, Not UD, UC! Yeah, lotta brain cells ago. Go Maroons!
I reckon maybe I’m off a year, tho this doesn’t explain all the difference. I’m using raw NGDP numbers from my link above: 14672.9 at 1/1/2008, 14381.2 at 1/1/2009, 14672.5 at 1/1/2010, which I interpret as a 2.0% drop in 2008 and a 2.0% increase in 2009.
Regardless, the broad story IS similar, isn’t it? NGDP growth averaging 5.5% for 2003-2007 but declining in 2006-07, negative growth for one crisis year, and lower growth (about 4% per year) over the past three years.
And my question is: how much better would RGDP growth (and unemployment) look over the past few years if we had a string of, say, 5% growth numbers for NGDP?
8. August 2013 at 14:09
Brian,
I see what you’re doing. You’re finding the year on year (yoy) changes for the first quarter of each year. There’s nothing necessarily wrong with doing that, but you certainly didn’t state that in your original comment.
All other things equal my *rule of thumb* for the US (and the eurozone) is that any change in NGDP roughly breaks 60% for RGDP and 40% for the deflator.
Chicago? You’re *really* stretching my memory. Did I know you when I lived in Henderson House? Or was it later? Who were some people we both knew?
8. August 2013 at 14:54
Brian, It’s not your fault that you don’t know how to read FRED data, I didn’t at first either. All I meant was that I’ve been doing this for a long time, and if I tell someone their NGDP data is wrong, you can be about 99.99% sure they are wrong. Don’t say to me “take it up with Fred.”
BTW, 1/1/2008 has nothing to do with January 1st 2008, it means calendar year 2008 data.
8. August 2013 at 14:55
Thanks Coleton.
8. August 2013 at 15:10
Mark, yes, I was in Henderson. The last time I saw you was prolly 1985. You were rockin’ a Jerry Garcia beard hanging out in Jay Mikalchus’s and Aaron Isherwood’s Den of Vice.
#misspentyouth
8. August 2013 at 15:55
Brian, yes I sort of remember you now.
As for Aaron and Jay’s Den of Vice, if I recall correctly, the mattresses were on the floor, the light bulbs were tinted, there was incense burning, and they had wet towels under the door.
And, as they say, if you can clearly remember your college days, you probably weren’t really there.
P.S. I subscribe to Sirius solely for their 24 hour a day Dead Channel.
8. August 2013 at 16:20
Today’s New York Times reports that Larry Summers is on the Board of Lending Club which lends to consumers at “annual interest of about 14.5 percent, depending on their credit scores, which can lift it to as much as 30 percent, with fees.” With “plans to expand into small-business and student loans, and may pursue credit cards, insurance and mortgages.” But don’t worry, as Fed Chairman he’ll tighten lender regulation and protect consumers – right.
9. August 2013 at 01:11
Well written and hard hitting article – you’re getting better at op eds.
9. August 2013 at 04:21
Mark, thanks for the 60/40 rule of thumb. I was looking for something like this from Scott, but he’s awfully cagey.
Of course, this rule of thumb is not constant over time, otherwise we could NGDP our way to undreamt riches. Maybe we tried this in the 70s, only to see RGDP growth go down.
Here’s a question: Does MM argue that 2009 would have been a particularly ‘potent’ time to increase NGDP, in terms of its impact on RGDP (so, the 60% is 70% or 80%)? Or is it just to stave off a drop in RGDP and associated macroeconomic dislocation, even if it’s only buying 50% or 40% ‘bang for the buck’?
According to your data, NGDP grew 1.7% in 2008 and dropped 2.1% in 2009, while RGDP dropped 0.3% in 2008 and 2.8% in 2009. Based in your 60% rule of thumb, had the Fed communicated and hit a 5% NGDP growth target for these years, RGDP would have grown 1.7% in 2008 and 1.5% in 2009.
RGDP would have been 6.5% higher at the end of 2009 than it was, which translates to x jobs, blah blah blah. Is that it?
9. August 2013 at 04:30
Mark,
As I look at the years leading up to the crisis, it seems to me that monetary policy was getting less ‘bang for the buck’. You can see signs of what some might call (not me, just some crazy Austrians) a ‘real business cycle’ running beneath a ‘placid’ NGDP surface.
Which leads me to think there may have been less marginal ‘bang for the buck’ to juicing up NGDP in 2008 and 2009 than your 60% rule of thumb. Which may by fine, but inclines toward the second, less awesome argument (staving off macroeconomic dislocation to some extent), which arguably makes sense even if there is only, say, a 20% impact on RGDP.
9. August 2013 at 05:55
Brian, I don’t see any evidence that monetary policy was getting less “bang for the buck in the years before 2008. Recall that in the long run RGDP tends to move toward the natural rate, regardless of monetary policy. That has nothing to do with “bang for the buck.”
Thanks Sean.
9. August 2013 at 06:07
Scott, fine. So let’s use Mark’s 60% rule of thumb.
Based on this, I understand you to be saying something like:
“If the Fed had targeted and hit 5% NGDP growth in 2008 and 2009, RGDP at the end of 2009 would have been 6.5% higher than it actually was.”
Honestly, I have no axe to grind here, and I’m not trying to entrap you or something. My only goal is to understand.
9. August 2013 at 09:40
Brian,
Here’s a relatively simple way to frame this.
The following link is to a dynamic AD-AS diagram, and which can be found in “Modern Principles: Macroeconomics” by Tyler Cowen and Alex Tabarrok. You’ll note that the rate of change in the AD curve is equal to the sum of the inflation rate and the rate of change in RGDP, and so is precisely equal to the rate of change in NGDP:
http://1.bp.blogspot.com/_JqNx8yXnFE8/SxlWoq_PI8I/AAAAAAAABCg/7y9VXIleCrs/s1600-h/Tabarrok-Cowen+ADAS.JPG
Note also the short run AS (SRAS) curve and the Solow growth curve, which is essentially the long run AS curve. In the short run wages and prices are sticky causing the SRAS curve to be upwardly sloped. In the long run money is neutral and wages and prices are flexible so the Solow growth curve is vertical. Thus shifts in AD influence the rate of growth of RGDP in the short run, but not in the long run.
My 60% rule of thumb is essentially a rough estimate of the slope of the SRAS curve. The slope of the SRAS curve is determined by by a variety of supply side factors largely outside of the control of monetary policy.
Note also that the way the SRAS curve is depicted it is concave up. Thus it is more horizontal to the left and more vertical to the right. Shifts in AD (NGDP) will break more towards RGDP on the horizontal portion and more towards inflation on the vertical portion. The more horizontal portion corresponds to when unemployment is high and the more vertical portion corresponds to when unemployment is low.
In addition, shifts in SRAS can occur, and I think there may have been a leftward shift in SRAS during 2008-09 due to a variety of factors (e.g. commodity prices). Such a shift would have temporarily raised the inflation rate and lowered the rate of growth in RGDP.
To the extent that the “potency” of monetary policy is measured in terms of the slope of the SRAS curve I would argue that monetary policy is more potent right now, not less. A change in NGDP is likely to break more towards RGDP when unemployment is high.
9. August 2013 at 11:40
Great stuff Mark. Thanks. I gotta go meditate on this now.
10. August 2013 at 03:40
First the Politco piece, now the Financial Times! I’m proud of you Scott. Then there’s me making Delong. Maybe we’re both making it! LOL
10. August 2013 at 04:14
Oh, and Mark, by ‘insane profligacy of the early 2000s’ I’m actually not talking about monetary policy. Apparently, there’s this thing called ‘the rest of the world.’
🙂
Symptoms include overextended consumers swimming naked, the absurd happenings in the real estate/mortgage market enabling same, tax cuts out the wazoo for EVERYONE, trillions of dollars in wars, Medicare Part D, government finances in disarray everywhere as the Baby Boomers Seven Fat Years draw to a close. Yippee!
11. August 2013 at 13:01
Brian, I think RGDP would have been substantially higher, but I can’t give you an exact figure. 6.5% sounds plausible, but so do 5% and 7%. Even 4% and 8%.