Scott Sumner  

Milton Friedman argued that the Great Depression occurred despite massive QE

Why is House of Cards more pop... Google's Gamble...

The Great Recession has exposed the fact that many economists believe things that just aren't so. They believe that low interest rates suggest that money is easy. They believe that monetary stimulus is ineffective at the zero bound. And recently I've discovered than many people believe that Friedman and Schwartz argued (in their Monetary History of the United States) that the Great Depression happened because the Fed cut the monetary base, or at least refrained from increasing it. Nothing could be further from the truth.

In their famous 1963 study, Friedman and Schwartz show that the base increased almost 4 fold over 12 years:

December 1929: $6,978 million
December 1941: $23,738 million

Indeed it grew by about 20% during the Great Contraction that ended in March 1933.

I think the problem here is that people either read the Monetary History so long ago that they've forgotten the details, or they read a sort of "Cliff's Notes" version of the book. The actual book is very nuanced in its interpretation of the evidence, and is full of brilliant analysis. And I say this despite the fact that I don't buy two of its key assumptions:

1. The claim that M2 growth is the best indicator of the stance of monetary policy (I prefer the gold/base ratio under the gold standard, and NGDP growth expectations under fiat money).

2. The assumption that the Great Depression in the US is best analyzed by looking at US monetary policy (I prefer a global perspective).

Over at TheMoneyIllusion a commenter named Lord Kelvin made the following observation:

As part of my efforts to be a better-informed econ undergrad, I recently read the Monetary History and was surprised by how much more sophisticated the narrative was than I expected.

Friedman and Schwartz were quite clear that the decline in M2 was due mainly to changes in the deposit and reserve ratios. They blamed the Fed for having assured banks that it would be a sound lender of last resort, then doing nearly nothing in the face of the Great Depression; they claim earlier suspension and the Aldrich-Vreeland act would have worked better. So I agree that Delong reads them incorrectly: they blamed the Fed for replacing a superior system; while you may disagree with this assessment, it's certainly reasonable.

Why does any of this matter? Because people like Paul Krugman and Brad DeLong don't just argue that Friedman had a simplistic view of the Great Depression, they also argue that the Great Recession has somehow discredited his views. After all, the Fed did lots of QE. Wasn't that supposed to prevent a depression?

Obviously if Friedman and Schwartz document a lot of QE during the Great Depression, they could not possibly have believed that the Great Depression was caused by a failure to do QE. A better argument would be that the Great Recession occurred despite the fact that (unlike in the early 1930s) the Fed insured that M2 kept growing. However by the last few years of his life Friedman had moved beyond simply money supply targeting, and shifted to inflation targeting. In 1992 he endorsed Robert Hetzel's proposal to peg TIPS spreads. Both TIPS spreads and actual inflation fell sharply during late 2008, and hence money was too tight either way.

Speaking of Robert Hetzel, he has a nice piece at the Richmond Fed on NGDP targeting. He discusses both the pros and cons, but ends on a positive note:

At the same time, articulation of a benchmark path for the level of nominal GDP would be a useful start in formulating and communicating policy as a rule. An explicit rule would in turn highlight the importance of shaping the expectations of markets about the way in which the central bank will behave in the future.

A benchmark path for the level of nominal GDP would encourage the FOMC to articulate a strategy (rule) that it believes will keep its forecasts of nominal GDP aligned with its benchmark path. In recessions, nominal GDP growth declines significantly. During periods of inflation, it increases significantly.

The FOMC would then need to address the source of these deviations. Did they arise as a consequence of powerful external shocks? Alternatively, did they arise as a consequence either of a poor strategy (rule) or from a departure from an optimal rule?

This is quite similar to the points I made in my recent post about Fed accountability.

One small quibble with the article is that Hetzel overlooks the fact that recent NGDP futures targeting proposals are not susceptible to the circularity problem.

Comments and Sharing

COMMENTS (11 to date)
ThomasH writes:

"...they believe that low interest rates suggest that money is easy. They believe that monetary stimulus is ineffective at the zero bound."

I think these are misunderstanding of the state of policy debate.

I believe the "they" believe that policy to lower interest rates suggest that money is being made easier.

I think "they" believe that conventional monetary stimulus (lowering interest rates, see above) is ineffective at the zero bound and they take (too much for my taste) for granted that unconventional policy like exchange rate depreciation, QE, or even announcing the intention to actually have an inflation target instead of an inflation ceiling target are hugely unpopular with VSP/macromedia types.

Scott Sumner writes:

Thomas, Yes, but it's also not correct to assume that falling interest rates mean money is getting easier. Rates fell during 2007-08 and money was getting tighter.

I see many pundits assuming that monetary policy was out of ammo because rates were at zero. Yes, some did distinguish between conventional and unconventional policy, but not all.


What do you mean that the monetary base increased by 20%?

"From the cyclical peak in August 1929 to a cyclical trough in March 1933, the stock of money fell by over a third."
- Milton Friedman and Anna Schwartz (A Monetary History of the United States, 1867–1960)

Scott Sumner writes:

Mark, When they say "stock of money" they mean either M1 or M2, not the monetary base, which rose sharply.

Blair writes:

Scott, what's your reading of this paragraph from Friedman's 1968 AER article on the role of monetary policy:

"The revival of belief in the potency of monetary policy was fostered
also by a re-evaluation of the role money played from 1929 to 1933.
Keynes and most other economists of the time believed that the Great
Contraction in the United States occurred despite aggressive expansionary
policies by the monetary authorities-that they did their best but
their best was not good enough.' Recent studies have demonstrated
that the facts are precisely the reverse: the U.S. monetary authorities
followed highly deflationary policies. The quantity of money in the
United States fell by one-third in the course of the contraction. And it
fell not because there were no willing borrowers-not because the horse
would not drink. It fell because the Federal Reserve System forced or
permitted a sharp reduction in the monetary base
, because it failed to
exercise the responsibilities assigned to it in the Federal Reserve Act to
provide liquidity to the banking system. The Great Contraction is
tragic testimony to the power of monetary policy-not, as Keynes and
so many of his contemporaries believed, evidence of its impotence. "

Andrew_FL writes:
(I prefer the gold/base ratio under the gold standard, and NGDP growth expectations under fiat money

I don't suppose you can explain why the latter isn't better than the former even under a gold standard?

W. Peden writes:

[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Scott Sumner writes:

Blair, That's a pretty bad typo, unless he was referring to a brief period in 1929-30 when the base fell about 8%. But it rose substantially over the 1929-33 period, as Friedman emphasized in the Monetary History.

Blair writes:

Interesting. If there are other instances when "base" and "stock" of money are used interchangeably, perhaps that explains why so many people get confused.

Mike Fellman writes:


I have to call BS on your misguided comments. The monetary base is irrelavent to the stance of monetary policy. Every mainstream monetary theorist agrees on this. Expanding the monetary base does nothing per se. Central banks implement monetary policy via setting short term interest rates, not the quanity of reserves. The central bank supplies said reserves on demand to the bank at the stipulated price. The cost of reserves is a key constraint on bank lending. As costs fall, in competitive credit markets these savings are passed on to customers. When the price of loans falls, customers take more loans. Presto, m2 increases!

When the price of reserves is already zero, adding reserves does nothing per se. Banks cannot do anything they couldn't do before.

QE is not about expanding the monetary base!!!!!!!! [Originally written in all caps so that economics phds who know nothing about modern money and banking and get all their information from the wall street journal understand!] It is about shifting portfolio preferences and possibly expanding long term credit by lowering its cost by pushing down long term treasure yields!!!!!!

[Comment edited with commenter's permission--Econlib Ed.]

Brendan Riske writes:

Monetary policy has been a joke since QE started. All those trillions in reserves and I dont think anyone really knows what the effects of it will be. The USD is part of the world's monetary base in an extreme way now, look at how the potential of an interest rate hike panicked EM. All we have left now are currency devaluation experiments with the different fiat regimes. Who loses first? Europe or Japan, or maybe even China? Inflation, employment, even GDP are all rigged figures now which barely reflect the economic reality.

Comments for this entry have been closed
Return to top