There have been lots of books written on the Great Recession. So why do we need a new one at this late date? Because in my view the other books are wrong. And note that Friedman and Schwartz’s book on the Great Contraction came out 30 years after the event, whereas my new book (entitled The Money Illusion) was delayed by a mere decade. Furthermore, the recent Covid recession has few lessons for monetary policy going forward, whereas the Great Recession has lots of important lessons. Future recessions are likely to be more like the (demand side) Great Recession than the (supply-side) Covid recession.
Nonetheless, there’s no question that my new book is poorly timed. I wrote hundreds of pages arguing that American recessions are almost always caused by tight money, right after we experience a once in a hundred year recession that was caused by a real shock—Covid-19. (The book was written years ago, long before Covid. For reasons that I fail to understand there is a long lag in publishing new books.)
My book would have greatly benefited from either of the following two scenarios:
1. Continued stable growth in NGDP after 2020, leading to a soft landing and proving that stable NGDP growth is best.
2. Another recession caused by a sharp drop in NGDP growth (or high inflation caused by excessive NGDP growth), proving that stable NGDP growth is best.
I was 99% likely to get one of those two outcomes, but got neither. Oh well . . .
And yet I prefer to focus on the bright side. Lots of claims that I made in the early drafts look much better today than in the late 2010s. For instance:
1. I argued that bubbles don’t exist, pointing to the fact that what was once viewed as a tech bubble (in 2000) and a housing bubble (in 2006), now look much less like bubbles. Over the past few years both tech stocks and housing prices have skyrocketed, so my bubble skepticism looks far more persuasive today that in 2018. Can anyone with a straight face still claim that house prices were obviously overvalued in 2006? Or the NASDAQ in 2000?
2. I called myself a “supply and demand side economist”, viewing both sides of the AS/AD model as being important. I argued that the Great Recession was primarily caused by a negative demand shock, but bad supply side policies contributed to the slow recovery. I specifically pointed to the fact that the removal of the extended unemployment benefits at the beginning of 2014 spurred rapid growth in employment. Today, the impact of bad labor market policies is even more obvious, as we have both a labor shortage and high unemployment. Both sides of the economy are important.
3. I argued that the Fed should have tried to get NGDP (or prices) back up to the original trend line after the 2008 recession, and that doing so would have made the recession much milder. In 2020, the Fed adopted a watered down version of that idea (average inflation targeting), and did succeed in getting both prices and NGDP back pretty close to the pre-recession trend line. And low and behold, the recession was much milder than the previous recession, despite forecasts in the spring of 2020 that we faced a major depression.
So I still feel pretty good about what I wrote before Covid, despite the abysmal timing of the publication.
READER COMMENTS
robc
Sep 2 2021 at 2:33pm
Yes. Eyeballing this
https://fred.stlouisfed.org/series/CSUSHPINSA
I would say that housing prices were overpriced in 2006, underpriced in 2012, and maybe overpriced today. It would be a lot easier to eyeball on a semilog plot, but that doesn’t seem to be an option on their site.
Its more clear if you look at specific cities AND separate out land price from building price. It wasn’t a housing bubble, it was a land bubble. In select places. In other places, not so much.
Scott Sumner
Sep 2 2021 at 2:52pm
What evidence do you have that prices were clearly overpriced in 2006? That graph merely shows ups and downs in a market price, which is certainly not evidence of an asset price bubble.
robc
Sep 2 2021 at 2:59pm
Shiller refused to call is a bubble too, although in 2005 he said it was unsustainable.
You have disagreed with me in the past on this, but the fact that Shiller said that and John Paulson made a few billion shorting it is evidence of a “bubble” to me.
Its possible this is a “predicted 12 of the last 5 recessions” kind of thing. Or maybe not.
Scott Sumner
Sep 2 2021 at 6:23pm
Paulson’s success is evidence that he was lucky.
robc
Sep 3 2021 at 8:43am
Or prices were clearly overvalued but he was one of the few with clear vision.
Kevin Erdmann
Sep 3 2021 at 1:44am
Shiller wrote in 2004 that there were local price bubbles but that a nationwide bust wouldn’t happen.
robc
Sep 3 2021 at 9:36am
He was pretty much right. In much of middle America there was a housing slowdown and moderate price decreases, but nothing that anyone would call a bust.
Kevin Erdmann
Sep 3 2021 at 11:34am
Here is St. Louis, which I picked at random. Most of the country looks like this. If you had told Shiller in 2004 that St. Louis had a 10%+ price correction and a 60%+ construction correction coming, he would not have considered that a moderate slowdown and he wouldn’t have said you should expect that.
https://fred.stlouisfed.org/graph/?g=GxuQ
robc
Sep 3 2021 at 12:16pm
I lived here at the time.
https://fred.stlouisfed.org/series/ATNHPIUS31140Q
I bought a house in Q4 2007, it felt like prices were already going down, but I guess they werent. No one was getting their asking price, of anywhere close to it, so it felt like a decline already. I think St Louis was too big of a city, you need to look smaller to see what most of middle America was going through. It is still a 6% decline, but not what I would call a “bust”.
I know I got 2 lowball offers on my condo, but I had no need to sell, just wanted to, so I held out for a decent offer (I asked 140k, got 135k). The house I purchased had been listed for 210k in 2006, but was relisted at 190k in 2007 and I got it for 184k.
Bowling Green, KY (where I moved in 2014) looks flat in comparison: https://fred.stlouisfed.org/series/ATNHPIUS14540Q
Lauren Landsburg
Sep 5 2021 at 12:39am
In September 2008, Shiller was comfortable enough about using the word “bubble” in his EconTalk podcast episode:
Robert Shiller on Housing and Bubbles
Shiller uses and also clarifies his use of the term “bubble” in that podcast episode.
My apologies for being late to the discussion here. Myself, I’m not a believer in bubbles. I think there’s always a reason for prices–be they asset prices or goods prices–to rise, fall, or be what they are. Sometimes–particularly in asset markets–price changes happen precipitously, dramatically, and in quick succession. In asset markets in particular, there are many participants weighing the entire discounted present value of future earnings in what they are bidding/asking/accepting to buy and sell. I think majorly mis-priced assets prices happening in rapidly-trending situations such as what are called “bubbles” not only theoretically are unlikely but also empirically and demonstrably have been rare and difficult to document. Sure, you as an individual can get taken for a ride if you leap without looking. That is, if you believe a bubble is the likely explanation for whatever trend you are leaping on, you are more likely than the average person to get taken for a ride.
I think a lot of the usage of the term “bubble” involves people worrying that folks other than themselves–ordinary folks, what we imagine to be average or ordinary people–either are being taken for a ride or are missing out on opportunities in asset markets. I don’t think either happens as often as imagined.
Jon Leonard
Sep 4 2021 at 8:45pm
It may be quibbling about the definition of a bubble, but I’d propose a working definition as a situation where a significant number of people are buying on margin because the price is going up. This was true in 2006, in the sense that more people were taking aggressive mortgages (and credit issuing standards were declining). That pattern is unsustainable; how often it ends with a crash is hard to say, and it doesn’t particularly depend on the asset being overvalued at the time.
Frank
Sep 2 2021 at 3:09pm
https://fred.stlouisfed.org/series/CSUSHPINSA
To go semi-log click on Edit Graph at upper right. On new pop up go down to Units. Choose logarithmic, or anything else.
robc
Sep 2 2021 at 3:29pm
Thanks, I didnt check out “units”.
I downloaded the data and best fit to the log of the CS index.
From 1987 to today, there are only two periods where the log of CS was more than 3% off the regression line.
From April 2004 to Feb 2008 it was high, 7% high from Oct 2005 to Aug 2006.
From Dec 2011 to Mar 2012 it was 4% low.
That is the percent the log is off, so 7% is a big number.
Today it is 2% high, which is fairly normal variation.
To me a bubble is something significantly higher than normal variation without an underlying fundamental reason. I think the 7% (I think that is 96% in unlogged terms, but someone should check my math) qualifies, but a bubble is in the eye of the beholder.
robc
Sep 2 2021 at 3:32pm
I checked my math, the unlogged values peaked at 39% high and 18% low.
Andrew_FL
Sep 2 2021 at 5:09pm
So your idea of a bubble is when prices follow, what, a bell curve path?
Scott Sumner
Sep 2 2021 at 6:24pm
A bubble is when prices are clearly overvalued.
Andrew_FL
Sep 3 2021 at 12:26am
Then I don’t get how the evolution of prices years later is relevant at all.
Scott Sumner
Sep 5 2021 at 1:41am
The burden of proof is on the bubble proponents. They are the ones making the claim that tech stocks were obviously overpriced in 2000. Where is the evidence? In 2002, the evidence they cited was that prices had fallen sharply after 2000. But now that tech stock prices are more than three times higher, even that evidence is gone. So where is the evidence that stocks were overpriced in 2000?
robc
Sep 3 2021 at 9:37am
By my analysis of the Case-Shiller trendline, housing prices were clearly overvalued.
Scott Sumner
Sep 5 2021 at 1:42am
The trend line is irrelevant to the question of whether housing was overvalued.
Alan Goldhammer
Sep 2 2021 at 5:15pm
Housing prices were not the key factor. Whether they were fairly valued or not is irrelevant given the all the NINJA loans that were issued by Countrywide, WaMu, Wachovia, etc. They these wonderful loans were opaquely packaged by Goldman Sachs and others into debt obligations that were highly leveraged and then crashed and burned just like the inflated Hindenberg did in the mid-1930s. All of this is well documented by numerous authors and of course in the entertaining film “The Big Short.”
Foolish companies such as AIG made huge bets on the wrong side of derivatives. All the NGDP targeting in the world wouldn’t solve that problem. I guess we should have just let all these institutions go bankrupt as Lehmann did and let the vultures pick at the carrion that was left.
Good luck with the book. I’ve read more than enough books on this and won’t be one of your loyal readers who picks up this tome.
robc
Sep 2 2021 at 5:21pm
Yes.
Scott Sumner
Sep 2 2021 at 6:28pm
I hope you’ll change your mind and read my book, where I explain why the conventional wisdom is wrong. With NGDP targeting, bank failures would have been dramatically lower. When nominal income falls, people have less resources to repay nominal debt. That’s why falling NGDP triggers financial crises.
BTW, most of the bank failures had nothing to do with mortgage loans.
Alan Goldhammer
Sep 3 2021 at 8:10am
This is an interesting proposition and runs counter to a lot of what Anat Admati has written regarding banking and capital requirements. Bank loans are on the balance sheet and part of those requirements. It leads to excess leverage which cannot be successfully dealt with once mortgages cannot be paid back.
Scott Sumner
Sep 3 2021 at 12:37pm
Most bank failures during the Great Recession were due to bad commercial loans, triggered by the bad economy (falling NGDP.)
DK
Sep 3 2021 at 3:45pm
Very much enjoy your posts, wherever they appear. But this one is a bit misleading. The proximate cause for most bank failures were bad construction loans. So while these may be labeled “commercial” loans under bank reporting rules, they were directly related to the collapse in residential real estate demand/prices, not to businesses failing.
Scott Sumner
Sep 4 2021 at 2:33pm
Construction loans also apply to commercial, industrial, etc. You can always claim that everything is related, but my point is that most bank failures were not caused by mortgage defaults, they were caused by a bad economy. If NGDP had kept growing at 5%/year, then the financial crisis would have been an order of magnitude smaller.
Jerry Brown
Sep 2 2021 at 7:14pm
I will buy your book and read it. Not sure if I will agree with it though. But you are a good writer in general, so how bad could it be.
What most impressed me back then was the apparent divergence in rental prices for property in many areas and prices to purchase the property. As in prices seemed to go way up while rents were much more stable. Do you address that in your book?
Scott Sumner
Sep 3 2021 at 12:36pm
Jerry, You ask:
“As in prices seemed to go way up while rents were much more stable. Do you address that in your book?”
I don’t recall exactly what I said, but I think I mentioned that with permanently lower real interest rates you’d expect permanently higher P/E ratios, Price/rent ratios, etc. I also mentioned how NIMBYism is driving up home prices.
Dale Doback
Sep 2 2021 at 8:10pm
I feel your position on bubbles has probably been an unecessary distraction from your larger points. While the reality of bubbles is maybe helpful to understanding some of your other main points (like the role of expectations and targeting the forecast), I think too many people get hung up on the bubble stuff. People are always saying such and such was obviously a bubble, while not defining what that even means. Bubble is too vague of a term to disprove and so the burden of proof should always be on the bubble people.
Scott Sumner
Sep 3 2021 at 12:38pm
“Bubble is too vague of a term to disprove and so the burden of proof should always be on the bubble people.”
Yup. And they failed to prove their point.
marcus nunes
Sep 2 2021 at 10:34pm
There was certainly a stoppage of supply. Demand was also constrained by the tanking of velocity (surge in money demand). Its interesting to see the “one-legged” monetarists predicting last year that after Covid there would be very high inflation because broad money supply growth was rising to the tune of almost 30% YoY!
If that hadn´t happened the US would have plunged in a deep depression! Now that velocity stopped falling, broad money supply growth is down to 3.9% (July).
https://marcusnunes.substack.com/p/how-the-inflation-story-changes-when
marcus nunes
Sep 2 2021 at 10:44pm
This is a readable primer on Broad (or Divisia) money supply:
https://www.forbes.com/sites/stevehanke/2018/10/29/the-feds-misleading-money-supply-measures/?sh=58e602c64941
Rajat
Sep 3 2021 at 12:22am
Congratulations, Scott! Where does the book end? Do you cover the so-called ‘test’ of MM in 2013?
Scott Sumner
Sep 3 2021 at 12:41pm
Yes, I cover 2013. The book ends on a somewhat philosophical note.
Todd Ramsey
Sep 3 2021 at 9:26am
“the recent Covid recession has few lessons for monetary policy going forward”
“My book would have greatly benefited from either of the following two scenarios…”
“despite the abysmal timing of the publication.”
I strongly disagree. You personally moved the Overton window enough that the Fed knew to act aggressively and immediately in March 2020, an extraordinary vindication of your efforts.
Publication of your book is your leap into the stands… after scoring the winning touchdown in the Super Bowl of March 2020.
I have to disagree with one more statement: “And yet I prefer to focus on the bright side…”
Scott Sumner
Sep 3 2021 at 12:41pm
Thanks.
rr
Sep 3 2021 at 11:56am
will there be an audible version?
Scott Sumner
Sep 3 2021 at 2:47pm
Yes, but I’m not certain when.
Thomas Lee Hutcheson
Sep 3 2021 at 12:13pm
The Covid recession has few lessons if there are few future supply shocks that are large enough not only to reduce but also to induce a massive increase in saving. But who can rule that out?
But in a larger sense, I agree, it does not hold any NEW lessons. Monetary policy should prevent large falls in nominal GDP as occurred in the early months of the pandemic. [“Shortages” instead of price spikes, deflation rather than inflation was completely perverse.] Banks should have been so flush with reserves that they would have been begging mortgage holders to defer rent payments and business to take bridging loans to keep the doors open until patrons returned. With better monetary policy, there should have been no need for the “paycheck protection program.” But we already knew that about monetary policy
And again, we saw the need for a more generous and automatic unemployment insurance system. UI should replace a high percentage but less than 100% of employee compensation (including health insurance premia) and self-employed income based on economic conditions, neither waiting for Congress to get around to passing relief bills nor prolonging it when labor markets have improved.
Scott Sumner
Sep 3 2021 at 12:37pm
“Bubble is too vague of a term to disprove and so the burden of proof should always be on the bubble people.”
Yup. And they failed to prove their point.
robc
Sep 7 2021 at 9:20pm
Paulson proved it. A million might be luck but a billion isnt.
reed e hundt
Sep 3 2021 at 2:45pm
As FCC chair 93-97 and in general technology cheerleader I caught no end of “told you so” and other versions of contumely from 2000 to 2008 concerning the “bubble” you at last show did not exist. Better late than never, so I think for me at least your timing is fine. Good luck with the book. Reed Hundt
Scott Sumner
Sep 4 2021 at 2:29pm
Thanks Reed. Unfortunately, one must often wait a long time to be vindicated.
Benoit Essiambre
Sep 4 2021 at 8:07am
>Continued stable growth in NGDP after 2020, leading to a soft landing and proving that stable NGDP growth is best.
Sure it wasn’t short-run stable, but it seems to be returning to trend pretty quickly
https://fred.stlouisfed.org/series/GDP
Compared to the price level which seems to be maybe overshooting:
https://fred.stlouisfed.org/series/CPALTT01USM661S
It almost looks like the Fed is targeting GDP level more than price level. The recovery looks better than I would have imagined given the situation with lots of opportunities for workers, pretty goldilocks really. This looks like a win for NGDPLT to me and I say that as someone somewhat skeptical of the practicality of the idea.
Scott Sumner
Sep 4 2021 at 2:34pm
Yes, it’s possible that they are secretly targeting NGDP levels.
Comments are closed.