Skidelsky on FDR’s gold-buying program

Robert Skidelsky has written a well-reviewed biography on Keynes.  Here he comments on FDR’s gold-buying program:

The gold-buying policy raised the official gold price from $20.67 an ounce in October 1933 to $35.00 an ounce in January 1934, when the experiment was discontinued. By then, several hundred million dollars had been pumped into the banking system.

The results were disappointing, however. Buying foreign gold did succeed in driving down the dollar’s value in terms of gold. But domestic prices continued falling throughout the three months of the gold-buying spree.

The Fed’s more orthodox efforts at quantitative easing produced equally discouraging results. In John Kenneth Galbraith’s summary: “Either from a shortage of borrowers, an unwillingness to lend, or an overriding desire to be liquid – undoubtedly it was some of all three – the banks accumulated reserves in excess of requirements. Reserves of member banks at Fed were $256 million more than required in 1932; $528 million in 1933, $1.6 billion in 1934, $2.6 billion in 1936.”

What was wrong with the Fed’s policy was the so-called quantity theory of money on which it was based. This theory held that prices depend on the supply of money relative to the quantity of goods and services being sold. But money includes bank deposits, which depend on business confidence. As the saying went, “You can’t push on a string.”

Keynes wrote at the time: “Some people seem to infer…that output and income can be raised by increasing the quantity of money. But this is like trying to get fat by buying a larger belt. In the United States today, the belt is plenty big enough for the belly….It is [not] the quantity of money, [but] the volume of expenditure which is the operative factor.”

I’m afraid that his analysis is both misleading and inaccurate.  The US gradually depreciated the dollar between April 1933 and February 1934.  During that period unemployment was nearly 25% and T-bill yields were close to zero.  Keynes argued that monetary stimulus would not be effective under those circumstances, and Skidelsky seems to accept his interpretation (which was published in the NYT during December 1933.)

[Note that Keynes certainly did believe in the “pushing on a string” theory–I frequently get commenters insisting that Keynes didn’t believe in liquidity traps.]

Unfortunately, Keynes and Skidelsky are wrong.  The US Wholesale Price Index rose by more than 20% between March 1933 and March 1934.  In the Keynesian model that’s not supposed to happen.  The broader “Cost of Living” rose about 10%.  Industrial production rose more than 45%.

And the gold-buying program was not an application of the quantity theory of money.  George Warren was an opponent of the QT, insisting that the quantity of money was not what mattered, that the price level was determined by the price of gold.  His views were much closer to Mundell than Friedman.  Keynes’s views on these issues were inconsistent and borderline incoherent.  Only a few weeks after Keynes argued that it was foolish to believe that currency depreciation could boost prices, FDR finally stopped depreciating the dollar.  How did Keynes react?  He congratulated FDR for rejecting the policy advice of the “extreme inflationists.”  By early 1934 prices were rising again.

The “disappointing” results that Skidelsky mentions come from cherry-picking a few misleading data points.  After the NIRA wage shock of late July, the real economy slumped and commodity prices started falling.  The value of the dollar also leveled off for a few months.  This led FDR to adopt the gold buying program in late October 1933. Despite the name, the actual “gold buying” was not large enough to be important–rather it was essentially a gold price raising program–a signal of future devaluation intentions.  This was well understood by the markets, and commodity prices tended to rise on days when FDR raised the gold price.  The broader WPI was relatively stable between September and December, and then started rising briskly again in early 1934.

It’s a mistake to focus on the tiny declines in the WPI during November and December 1933, which partly reflected sharply falling prices in Europe.  The key point is that the WPI and industrial production rose strongly during the period of dollar depreciation.

Skidelsky is a big fan of Keynes, but needs to read his hero’s writings with a more critical eye.  Other modern Keynesians like Krugman and Eggertsson have argued that FDR’s dollar devaluation program boosted the economy in 1933, and they are right.  They would also be horrified to see a Keynesian criticizing QE2:

Now the US, relying on the same flawed theory, is doing it again. Not surprisingly, China accuses it of deliberately aiming to depreciate the dollar. But the resulting increase in US exports at the expense of Chinese, Japanese, and European producers is precisely the purpose.

The euro will become progressively overvalued, just as the gold bloc was in the 1930’s. Since the eurozone is committed to austerity, its only recourse is protectionism. Meanwhile, China’s policy of slowly letting the renminbi rise against the dollar might well go into reverse, provoking US protectionism.

The failure of the G-20’s Seoul meeting to make any progress towards agreement on exchange rates or future reserve arrangements opens the door to a re-run of the 1930’s. Let’s hope that wisdom prevails before the rise of another Hitler.

I can’t see how depreciating the dollar against the euro would cause China to depreciate its currency against the dollar—I would have thought exactly the reverse.  What am I missing?  (Or was that a typo on Skidelsky’s part?)  And it wasn’t protectionism that led to the rise of Hitler, it was deflationary monetary policies in the US, France, and Germany.

PS.  Keynes is not my hero, George Warren is.  Anyone criticizing Warren in the blogosphere can expect a sharp rebuke from TheMoneyIllusion.

PPS.  His statement about the official price of gold being raised after October is also misleading.  The par value of gold rose all at once in early 1934.  The gold-buying price was already well above $20.67 by October 1933.  The focus should be on the market price of gold, which rose gradually over a period of about 10 months.

PPPS.  There is one strong similarity to late 1933; the conservative outrage over the gold-buying program forced FDR to stop it long before he reached his objective of reflating the price level to pre-Depression levels.  Now there are signs that conservative outrage over QE2 may be making it harder for Bernanke to achieve his inflation objectives, which are to return the inflation rate to pre-recession levels.  Plus la change . . .

HT:  JimP


Tags:

 
 
 

19 Responses to “Skidelsky on FDR’s gold-buying program”

  1. Gravatar of marcus nunes marcus nunes
    23. November 2010 at 16:04

    If Lord RS only bothered to look at the pictures for prices and ind. production…
    That´s why someone said that economic thinking evolves “funeral by funeral”. Living economists never give up their views, even in the face of simple and easily available evidence. In the case of RS it is worse since he is a huge fan and biographer of Keynes…

  2. Gravatar of Mark A. Sadowski Mark A. Sadowski
    23. November 2010 at 16:06

    Scott,
    I was staring at Eichengreen’s now famous graph from his paper on “The Great Slump” (1992) today. Much has been made of it in the left/liberal blogosphere because supposedly it has a perfect relationship between the “end” of the gold standard and the beginning of economic recovery.

    Being the empiricist that I am, I looked at market prices of gold and discovered it was largely true but had one glaring exception among the five nations Eichengreen mentioned: Germany. Germany maintained a more or less constant market price of gold throughout 1929-1937 (no depreciation). Yet they recovered quite smartly after the Nazis took power. How do you explain this inconsistancy? (Excuse me if you’ve already addressed this question in the recesses of this blog’s history.)

  3. Gravatar of Aí já é “sacanagem”… « Historinhas Aí já é “sacanagem”… « Historinhas
    23. November 2010 at 16:37

    […] Sumner tem um bom post (aqui) derrubando os argumentos de Skildesky que, sendo mais Keynesiano do que Krugman, e ao contrário […]

  4. Gravatar of marcus nunes marcus nunes
    23. November 2010 at 17:05

    @Mark
    From 1933 to 1937 Germany´s GDP PC evolved much like Britain´s and both much better than France´s. To Germany maybe getting Hitler in January 1933 was the “equivalent” to leaving the gold standard (much higher G)

  5. Gravatar of Dustin Dustin
    23. November 2010 at 17:38

    Okay, amateur question hour (no laughing please):

    Looking at the annual gdp numbers (percent change from preceding period) from bea.gov, I see large negative numbers for 31,32,33, and then suddenly for 1934 the numbers are positive and huge. Ditto for 35,36.

    Was this due to the gold buying program? Something obviously turned things around in a big way. Yes?

  6. Gravatar of Mark A. Sadowski Mark A. Sadowski
    23. November 2010 at 17:47

    Dustin,
    You’re asking for causality when in truth no one really knows with absolute certainty.

    I look at the data. I see a depreciation in the value of the dollar. Then I observe an increase in real output. And I reach the same conclusion you evidently have.

    No need for laughter.

  7. Gravatar of JimP JimP
    23. November 2010 at 18:00

    Scott

    Any comments on that Krugman Eggertsson paper?

    http://www.princeton.edu/~pkrugman/debt_deleveraging_ge_pk.pdf

    The way I read it, it seems to acknowledge that monetary policy does dominate fiscal policy, without directly saying so.

  8. Gravatar of Doc Merlin Doc Merlin
    23. November 2010 at 18:03

    @Mark
    “Being the empiricist that I am, I looked at market prices of gold and discovered it was largely true but had one glaring exception among the five nations Eichengreen mentioned: Germany. Germany maintained a more or less constant market price of gold throughout 1929-1937 (no depreciation). Yet they recovered quite smartly after the Nazis took power. How do you explain this inconsistancy? (Excuse me if you’ve already addressed this question in the recesses of this blog’s history.)”

    I’m not Scott but I will make my argument anyway.
    1. Government action isn’t necessary for recovery to happen, it can happen all my itself.
    2. Germany stopped paying back the crippling war reparations.

  9. Gravatar of ssumner ssumner
    23. November 2010 at 19:02

    Marcus, Yes, and my funeral is also much closer than 34 years ago when I started grad school.

    Mark, That’s a good question. Germany adopted strong currency controls, which meant the German government could inflate without worrying about a gold outflow. I believe they also avoided the US/French high wage policy, and instead Hitler held wages down.

    I’m not sure how comfortable I feel having Hitler on my side on the wages question. But then he was also a vegetarian.

    Dustin, It was due to the dollar depreciation policy begun in April 1933, of which the gold-buying program was a part. The gold-buying program itself occurred in late 1933. FDR’s other policies actually slowed the recovery. It was all about dollar depreciation against gold. The devaluation was huge, from $20.67 to $35 dollars an ounce. This dramatically raised expected future prices–and current asset values.

    JimP, The post I just did (just posted) addresses the paper. I read them as favoring fiscal stimulus.

    Doc Merlin, They weren’t really paying them anyway, but I agree that removing that huge uncertainty probably boosted business confidence in Germany.

  10. Gravatar of Mark A. Sadowski Mark A. Sadowski
    23. November 2010 at 19:13

    Scott wrote:
    “I’m not sure how comfortable I feel having Hitler on my side on the wages question. But then he was also a vegetarian.”

    As for myself, I usually eat a double Angus burger with a fried egg on top for lunch at the Burger Studio in the Trabant Center at UD. I guess that makes me an anti-Nazi.

    As for your other points they make sense. But in fairness to me, I expected them.

  11. Gravatar of Joe Joe
    23. November 2010 at 19:27

    If Friedman was wrong that velocity and Md are constant, then doesn’t that mean that the QTOM is false?

  12. Gravatar of Full Employment Hawk Full Employment Hawk
    23. November 2010 at 20:22

    “Government action isn’t necessary for recovery to happen, it can happen all my itself.”

    There was a lot of government action in Nazi Germany that led to the recovery. The credit for Germany’s recovery belongs largely to Hjalmar Horace Greeley Schacht, who was head of the German central bank and for a good deal of the NAZI prewar period finance minister. A decade before he had played a leading role in ending the German hyperinflation.

    One of the main factors leading to the recovery was a very strong fiscal stimulus. This included major public works expenditures, like the autobahn. The role played by public works projects is illustated by the saying at that time that Hitler would paint the Black Forest white if it crated jobs. (I got that from my parents who lived in Germany during this time. I was born in NAZI Germany myself.) But the secret rearmament, which was in violation of the Treaty of Versailles also played a major role. To help pay for the resulting deficit Schacht created a money substitute the MEFO bills, which financed a significant part of the government deficit. Since the MEFO bills could be exchanged for Reichmarks on demand they were definitely part of M2 type monetary aggregates. So a significant part of the fiscal stimulus was financed by printing money.

    By 1936 Germany was back at full employment. This explains a lot of Hitler’s popularity. In the English speaking world, these policies, which provide a wealth of empirical evidence about the effects of monetary and fiscal policies, has unfortunately been largely ignored. But the German recovery was definitely not a case of an economy recovering by itself.

  13. Gravatar of Full Employment Hawk Full Employment Hawk
    23. November 2010 at 20:28

    “If Friedman was wrong that velocity and Md are constant”

    Friedman actually argued that velocity was a stable and preditable function of a few variable, although he quite often argued as if velocity itself actualy was stable.

    By the 1970’s M2 velocity became unstable and unpredictable, which did orthox Friedman monetarism in.

  14. Gravatar of Mattias Mattias
    24. November 2010 at 02:00

    Scott

    Talking of Hitler,

    Have you done any post on Germany’s hyperinflation, the war reparations and the rise of Hitler?

    It is often argued or implied that Germany’s tightening influence on ECB:s monetary policy today is a result of their hyperinflation history, and that this lesson is important for other countries to study today. “Ease now, get a Hitler tomorrow!”

    I just get the feeling that this description of the events is a bit too convenient for Germany. It’s almost like it was the Versailles treaty that put Hitler in power, and not the Germans themselves. As I understand it, the war reparations were not to be paid in Marks so the whole idea of monetizing the debt seems a bit strange.

    Also the link between the hyperinflation and Hitler’s rise to power seems also a bit far-fetched. If Hitler had taken power in the mid twenties, directly after the hyperinflation it would have made more sense.

    It would be interesting to hear your opinion on these matters.

  15. Gravatar of Full Employment Hawk Full Employment Hawk
    24. November 2010 at 07:14

    “Ease now, get a Hitler tomorrow!”

    The German hyperinflation did not lead to a fall of the Weimar Republic or a Hitler. The Weimar Republic fell and Hitler took power at a time when Germany was suffering from deflation and, of course, very bad depression. The lesson that should have been learned by historical events (but was not) is:

    Ease now, OR get a Hitler tomorrow!

    But the experience of the hyperinflation has left the German people paranoid about even a little inflation.

  16. Gravatar of scott sumner scott sumner
    24. November 2010 at 19:19

    Joe, Friedman didn’t assume that money demand was constant, just a stable function of a few variables. It would change as the rate of inflation changed.

    Full Employment Hawk, I agree that the Nazis did a lot of fiscal stimulus. I don’t recall how large as a share of GDP. Does anyone know?

    Mattias, It’s a myth that the hyperinflation led to the rise of Hitler. That occurred in the early 1920s. By 1929 Germany had had five decent years of low inflation. And the Nazi party was still tiny. It was the Great Depression that put the Nazi’s in office, I have no doubts about that.

    FEH, Weird. I answer two comments, and then find you have very similar responses. I should check out your comments first–it would save me work!

  17. Gravatar of Sumner, Skidelsky, Keynes and Liquidity Traps | Federal Reservations Sumner, Skidelsky, Keynes and Liquidity Traps | Federal Reservations
    8. November 2011 at 11:28

    […] was searching for some information and I stumbled on a post Scott Sumner wrote last year about Robert Skidelsky's biography of John Maynard Keynes. I haven't read Skidelsky's book, nor do I know Skidelsky, and its been awful long time since I […]

  18. Gravatar of Sumner, Skidelsky, Keynes and Liquidity Traps Sumner, Skidelsky, Keynes and Liquidity Traps
    12. November 2011 at 09:02

    […] was searching for some information and I stumbled on a post Scutt Sumner wrote last year about Robert Skidelsky’s biography of John Mayndard Keynes. I haven’t read Skidelsky’s book, nor do I know Skidelsky, and its been awful long time […]

  19. Gravatar of Bitcoin Faucet Rotator Blog QE-zero Bitcoin Faucet Rotator Blog QE-zero
    10. April 2015 at 06:59

    […] often point to. For instance, see David Glasner here, David Beckworth here, and Scott Sumner here. Scott also has a very interesting paper on the 1933 gold purchasing program (pdf). No, I was […]

Leave a Reply