Misrepresenting the Recovery from the Great Depression

In today’s Wall Street Journal, Harold Cole and Lee Ohanian try to teach us some lessons from the Great Depression.  According to Cole and Ohanian, those of us who believe that increasing aggregate demand had anything to do with recovery from the Great Depression are totally misguided.

[B]oosting aggregate demand did not end the Great Depression.  After the initial stock-market crash of 1929 and subsequent economic plunge, recovery began in the summer of 1932, well before the New Deal.  The Federal Reserve Board’s Index of Industrial Production rose nearly 50% between the Depression’s trough of July 1932 and June 1933.  This was a period of significant deflation.  Inflation began after June 1933, following the demise of the gold standard.  Despite higher aggregate demand, industrial production was flat over the following year.

Though not wrong in every detail, the version of events offered by Cole and Ohanian is still a shocking distortion of what happened before FDR took office in March 1933.  In particular, although Cole and Ohanian are correct that the trough of the Great Depression was reached in July 1932, when the Industrial Production Index stood at 3.67, rising to 4.15 in October, an increase of about 13%, they conveniently leave out the fact that there was a double dip; industrial production was flat in November and started falling in December, the Industrial Production Index dropping to 3.78 in March 1933, barely above its level the previous July.  And their assertion that deflation continued during the recovery is even farther from the truth than their description of what happened to industrial production.  When industrial production started to rise, the Producer Price Index (PPI) increased almost 1% three months in a row, July to September, the only monthly increases since July 1929.  The PPI resumed its downward trend in October, falling about 9% from September 1932 t0 February 1933, at the same time that industrial production peaked and started falling again.

That is why most observers date the trough of the Great Depression in the US not in July 1932, but in March 1933 when FDR took office in the midst of a banking crisis that threatened to drive the US economy even deeper into deflation and depression than it had been in July 1932. So when Cole and Ohanian assert that recovery from the Great Depression started in July 1932, and go on to say that the recovery took place during a period of significant deflation, it is hard to avoid the conclusion that they are twisting the facts to suit their own ideological predilection.

The misrepresentation perpetrated by Cole and Ohanian only gets worse when they describe what happened during the period of true recovery, April through July 1933.  Contrary to their assertion, deflation stopped in February 1933, the PPI hitting its low point of 10.3.  Prices began to rise as soon as FDR suspended the gold standard shortly after taking office in March (not June as Cole and Ohanian mistakenly assert) 1933, the PPI rising to 11.9 in July (an increase of about 14% over February) when industrial production hit a peak of 5.95, 57% above the March low point.

The two attached charts (supplied courtesy of Marcus Nunes) illustrate the sequence of events that I have just described.  The close correlation between the PPI and industrial production is especially striking.

To assert that the rapid price increases from March to July, a proxy for increased aggregate demand, played no role in what was then (and remains) the fastest increase in industrial production in any four-month period in American history is a gross misrepresentation of the facts.  What is perhaps even more shocking is that Cole and Ohanian would misrepresent facts so easily ascertainable.

Nevertheless, not everything Cole and Ohanian say is wrong.  They properly criticize New Deal policies that slowed down the spectacular recovery from April to July 1933 to almost a crawl.  What stopped April to July recovery almost in its tracks?  The answer is almost certainly that FDR forced his misguided National Industrial Recovery Act through Congress in June, and by July its effects were beginning to be felt.  Simultaneously forcing up nominal wages in the face of high unemployment (though unemployment started had falling rapidly when recovery started in April) and cartelizing large swaths of the American economy, the NRA effectively shut down the recovery that was still gaining momentum.  As shown in the chart representing industrial production, industrial production resumed its rapid expansion almost immediately after the NIRA was unanimously struck down by the Supreme Court in May 1935.  The aborted recovery was a tragedy for the American economy and for the world, but the premature end of (or extended pause in) the recovery tells us nothing about whether an economy can recover from a depression with no increase in aggregate demand.

Another point overlooked by Cole and Ohanian, presumably because it doesn’t exactly fit the ideological message that they want to propagate, is that the timing of the recovery — immediately after the monetary stimulus resulting from suspension of the gold standard – shows that monetary policy can be effective with little or no fiscal stimulus.  It is hard to see how any fiscal stimulus could have taken effect by April 1933 when the recovery had already begun.  Moreover, Roosevelt campaigned as a fiscal conservative, so it would not be easy to argue that anticipated fiscal stimulus was being felt in advance of its actual implementation.

The real lesson the Great Depression is that monetary policy works — for good or ill.

HT:  Bill Woolsey, Marcus Nunes, Scott Sumner

About these ads

39 Responses to “Misrepresenting the Recovery from the Great Depression”


  1. 1 Benjamin Cole September 26, 2011 at 3:29 pm

    Thanks goodness for the Internet, a valuable medium to expose simple canards, such as those foisted off on the public by Cole (no relation) and Ohanian (really no relation).

    What is the purpose of these endless assaults on effective monetary policy? What is this peevish zeal to bring measurable inflation to zero, at huge costs to real output? Why this perverted worship of gold?

    At best, we can hope this is simple posturing until we get a GOP president, at which time the WSJ will editorialize it is time to ramp things up again.

    At worst, some right-wingers really believe this demented ranting about gold and tight money, and thus they will be a threat to USA security and prosperity for a long, long time. What if the GOP wins and the gold-fetishists assume policy positions?

    Japan here we come—or maybe Hoovervilles-R-Us.

    And not, I am not a liberal. My favorite economist is Milton Friedman.

  2. 2 João Marcus Marinho Nunes September 26, 2011 at 3:50 pm

    David
    Great “smackdown” on “ideologigal bias”! One correction:”Contrary to their assertion deflation stopped in February 1933 (not 1932).

  3. 3 Nick Rowe September 26, 2011 at 9:16 pm

    I hadn’t realised how strong and immediate the effects of going off gold were, Those charts tell a very compelling story.

    Hmmm. Maybe Scott is right about those long and variable lags being very short indeed.

    And, as I understand it, nothing much at all was happening to interest rates on government bonds after FDR went off gold, right?

  4. 4 Tyler September 27, 2011 at 5:39 am

    I’m disgusted at how people can be so dishonest, but unsurprised that it was people writing for The Wall Street Journal.

  5. 5 srini September 27, 2011 at 6:16 am

    Let me see: no mention of Emergency banking act, the bank holiday–all of which were done before the act of confiscating gold and two months before devaluing the dollar in terms of gold?

    The issues are more complicated than “expand the monetary base” and voila there are no lags, it is instantaneous. Promising a credible lender of last resort, and having deposit insurance is not the same as expanding the base when financial conditions have stabilized.

  6. 6 skeptonomist September 27, 2011 at 8:00 am

    Glassner says “The real lesson [of] the Great Depression is that monetary policy works — for good or ill.”

    Freezing and revaluing gold is not exactly what is understood by modern monetary policy. The evidence permits the hypothesis that closing the banks and freezing/revaluation were critical events in the turnaround, but obviously the circumstances then were very different from anything since; other countries had abandoned gold and there were continual bank runs, which are now practically non-existent because of Glass-Steagall. It was Glass-Steagall and the action on revaluation by Roosevelt and Congress which saved the banks and probably keyed the turnaround, not monetary policy action by the Fed.

    In the real world of Fed monetary policy, the Fed started cuttting rates in August 2007 – did that prevent the financial collapse and recession or cause an instant turnaround? Glassner answers the WSJ nonsense with more nonsense of his own about monetary policy.

  7. 7 Alex Apostolides September 27, 2011 at 8:18 am

    You can not trust the growth rates of an index with different bases, called the Gershenkron effect, where you can not compare indexes with different base years. In addition if you use a modern base in order to see growth in 1931 – the growth rates of dates a long way from your base have higher than possible growth rates (if it a leaspayres index).

  8. 8 mark calahan September 27, 2011 at 9:12 am

    Just more falsehoods from our version of the Taliban. There are objective facts. Deal with them. Do not ignore or lie about them.

  9. 9 ezra abrams September 27, 2011 at 10:02 am

    this blog would be a lot more convincing if you had a simple graph, what cole said, what the graph said…way to many words to carry the poitns

  10. 10 herbert juarez September 27, 2011 at 10:11 am

    Glasner should refrain from commenting publicly about things he does not understand. That goes for the rest of you, too.

  11. 11 Lars Christensen September 27, 2011 at 10:59 am

    David, great post…it is incredible that anybody would put their names on a piece like the one Cole and Ohanian have produced. Your two graphs clearly illustrates how potent monetary policy can be – and how damaging the wrong supply side policies can be.

    One thing which is really puzzling how they can say that the Great Depression is a negative supply shock when we had massive deflation – not exactly textbook…

  12. 12 Lars Christensen September 27, 2011 at 11:29 am

    Nick, of course Scott is right. Monetary policy is extremely potent and works with short or no lags at all. By a look at the market reaction to the devaluation is also extremely impressive – US stock rallies very strongly from the day FDR takes the US off gold…

  13. 13 Lars Christensen September 27, 2011 at 11:41 am

    By the way it should be remembered that a Market Monetarist perspective on the devaluation would be that the devaluation might actually have an positive impact even BEFORE it actually happened. As Scott Sumner like to remind us – monetary policy works with long and variable LEADS.

    The speculation about FDR pursuing anti-deflation policies was clearly on the rise already in 1932.

  14. 14 Dustin September 27, 2011 at 12:57 pm

    Excellent post, David.

    And Marcus’ charts are very compelling, showing the potency of monetary stimulus as well as the harm that NIRA did.

  15. 15 W. Peden September 27, 2011 at 1:32 pm

    skeptonomist,

    Looking at nominal interest rate movements is an extraordinarily naive way of examining monetary policy.

  16. 16 João Marcus Marinho Nunes September 27, 2011 at 4:00 pm

    @Skeptonomics: You wrote:
    “In the real world of Fed monetary policy, the Fed started cuttting rates in August 2007 – did that prevent the financial collapse and recession or cause an instant turnaround? Glassner answers the WSJ nonsense with more nonsense of his own about monetary policy”.
    No, it didn´t do an instant turnaround because, despite the interest rate cuts, MP was still contractionary! Hard to understand? Read this:

    http://thefaintofheart.wordpress.com/2011/09/18/john-taylor-defends-a-troublesome-target/

  17. 17 David Glasner September 27, 2011 at 10:20 pm

    Benjamin, Bad monetary policy had tragic consequences in the 1920s and 1930s, so the stakes are indeed very high, but all the same, watch your blood pressure.

    Marcus, Thanks and thanks again for the graphs.

    Nick, Listen to Lars. I haven’t looked up interest rates, but I think they stayed pretty low for a very long time.

    Tyler, It was Krugman who accused them of writing in bad faith. I don’t know what their motivation was, but they were very sloppy. I am not trying to evaluate their character.

    Srini, You are right that I did not write anything like a complete history of the Great Depression in one blog post.

    Skeptonomist, There are various ways to conduct monetary policy and manipulating interest rates is only one of them. And I certainly agree that it was FDR who took control of monetary policy from the Fed in 1933 to the great benefit of the US and the rest of the world.

    Alex, Thanks for reminding me of (or teaching me about) the Gershenkron effect. However, the correlation between the PPI and industrial production seems so close that it seems hard to imagine that it would be entirely a statistical artifact. How would you go about estimating the magnitude of the effect or possibly correcting for it? Or is that not possible?

    mark, I am no fan of the Wall Street Journal, but calling them taliban is just way over the top.

    ezra, Thanks for your suggestion, but blogging requires pretty quick turnarounds.

    herbert, Thank for that really valuable input.

    Lars and Dustin, Thanks for your kind words, I don’t understand why Cole and Ohanian weren’t more careful.

    Marcus, Good point.

  18. 18 Alex Apostolides September 27, 2011 at 10:55 pm

    Dear David,
    There are several ways to approach the issue, including a fischer rather than a Laspayeres or Paasche index. Statistical offices used chain linked indexes, but as economic historians we find that this creates problems in longrun series.

    The Gershekron effect can be easily corrected if you have the original data.
    First pick a common base year for all your indicators (if you can). Secondly if you have a long series (over 150 years/quarters ect.) you correct the series’ growth rate. You pick benchmarks of 50 years and then use those growth rates on the orignal series levels. In effect you smooth out the proportion of the growth rate that is created by simply being far away for a base (which could be as high if the series is long).

    You can email me if you want – it is something economic historian take into account all the time.

  19. 19 Alex Apostolides September 27, 2011 at 10:56 pm

    Dear David,
    There are several ways to approach the issue, including a fischer rather than a Laspayeres or Paasche index. Statistical offices used chain linked indexes, but as economic historians we find that this creates problems in longrun series.

    The Gershekron effect can be easily corrected if you have the original data.
    First pick a common base year for all your indicators (if you can). Secondly if you have a long series (over 150 years/quarters ect.) you correct the series’ growth rate. You pick benchmarks of 50 years and then use those growth rates on the orignal series levels. In effect you smooth out the proportion of the growth rate that is created by simply being far away for a base (which could be as high if the series is long).

    You can email me if you want – it is something economic historian take into account all the time.
    Thank you for answering

  20. 20 bill woolsey September 28, 2011 at 12:03 pm

    These aren’t growth rates.

    These are levels, right?

    Of course, they are levels of an index, and so are stated relative to a base year.

    Does that make a difference?

    I suppose if the argument here was that the PPI went up 5% in 1932 and the Industrial production index went up 5% too, and that was important, then this might matter.

    But the argument was just about direction.

    Is it possible that because the indexes used different years, the PPI really was going down when the index went up? And that industrial production really went down, when the index of industrial production went up?

    Maybe I am wrong, but I am pretty sure the Great Depression wasn’t about slower producer price inflation and slower growth in industrial production.

    Where are the growth rates?

  21. 21 bill woolsey September 28, 2011 at 12:05 pm

    I would note that the left of center economists (Krugman, Toma, and DeLong) quoting this post leave out:

    “Nevertheless, not everything Cole and Ohanian say is wrong. They properly criticize New Deal policies that slowed down the spectacular recovery from April to July 1933 to almost a crawl. What stopped April to July recovery almost in its tracks? The answer is almost certainly that FDR forced his misguided National Industrial Recovery Act through Congress in June, and by July its effects were beginning to be felt. Simultaneously forcing up nominal wages in the face of high unemployment (though unemployment started had falling rapidly when recovery started in April) and cartelizing large swaths of the American economy, the NRA effectively shut down the recovery that was still gaining momentum. As shown in the chart representing industrial production, industrial production resumed its rapid expansion almost immediately after the NIRA was unanimously struck down by the Supreme Court in May 1935. The aborted recovery was a tragedy for the American economy and for the world, but the premature end of (or extended pause in) the recovery tells us nothing about whether an economy can recover from a depression with no increase in aggregate demand.”

  22. 22 Alex Apostolides September 28, 2011 at 11:34 pm

    Dear Bill,
    Anything that affects growth rates has an effects on levels too – no it will not change the direction, but if the magnitude is important then the further away from the base you are, the least reliable your growth (and hence level) is. Statistical offices correct for this by changing bases often (and correcting as as I explain above) or by chain-linking.
    We take indexes for granted, but really they can create very big problems, while index number theory is never taught in graduate school.

  23. 23 Barry September 29, 2011 at 5:44 am

    First, I would note back to you that this paragraph discusses a few months over several years; perhaps you should have an accurate quantitative perspective.

    Second, ‘I would note that the left of center economists (Krugman, Toma, and DeLong) quoting this post leave out:’

    Since Krugman has spent years cleaning Chicago’s clock, I’ll accept the obvious interpretation that ‘right of center’ means ‘detached from reality’.

  24. 24 Barry September 30, 2011 at 7:20 am

    Herbert, considering that right-wing economics is collapsing into a morass of revisionism and straight-out lying, perhaps it’s time for you guys to realize that you don’t know what you are talking about, and to shut up.

  25. 25 David Glasner October 2, 2011 at 10:33 am

    Alex, Thanks for elaborating on the point for me. My intuition says that for the purposes that I (and I think Bill as well) am interested in these numbers the effect you are mentioning wouldn’t invalidate the inference I want to make. But obviously my intuition doesn’t count for that much. So the specific question which we may have to pursue off line is what as a practical matter would we have to do to the PPI and industrial production data to address the Gershenkron effect. I at least would need a fairly explicit set of guidelines about how to proceed.

    Bill, I think your intuition probably is the same as mine, but to make a solid empirical case, I think Alex’s point would have to be recognized and addressed.

    Barry, I think it is fair to point out that Krugman et al. did not quote my post in its entirety but picked out the part of it that was most congenial to their position. There’s nothing dishonest about that, but there is nothing wrong in making that observation, either.

  26. 26 Alex Apostolides October 2, 2011 at 12:38 pm

    Feel free to contact me by email (alexapostolides at gmail). I think I can see source, and quickly 1) see if it correction to these effects has already been done
    2) see how been the effect can be
    3) correct it or at least walk you though in how it is corrected.

    Note: the effect only changes the amount of growth but not its direction. So i thinks the argument you make is valid

  27. 27 Alan Reynolds October 5, 2011 at 8:04 am

    A piece I wrote for the Fall 2009 Cato Journal includes tables showing measures of fiscal and monetary policy. The data clearly confirm Romer’s 1991 finding that, “Monetary developments were very important and fiscal policy was of little consequence in the mid and late 1930s.” But they also confirm Romer and Romer’s later finding, using postwar data, that higher tax rates are highly contractionary. And that, too, contradicts what Cole and Ohanian wrote in 2007.

    http://www.cato.org/pubs/journal/cj29n3/cj29n3-11.pdf

    I wrote that, “Devaluations in the United States and Britain clearly helped ‘reflate’ the world economy. Most U.S. and foreign tariffs were specific (levied as cents per item or pound), so effective tariff rates were reduced by higher prices. Reflation in Britain also shrank the real value of the dole, boosting work incentives. . . .

    Although short-term interest rates were very low throughout the 1930s, the quantity of bank reserves and currency remained critically important during the deflationary period of 1930–33, the mildly inflationary recovery of 1934–36, the 1937 recession, and the subsequent recovery after June 1938. . . .

    Mono-causal explanations of the Great Depression are neither necessary nor sufficient. Monetary forces dominated the deflationary periods, 1930–33 and 1938. Yet the strong recovery after June 1938 was mildly deflationary, and the aborted recovery of 1934–36 was mildly inflationary. Moreover, such financial matters as the falling value of collateral for bank loans were surely affected by real events, which include effects of changing tax rates (Shimer 2009) and tariffs on the supply and productivity of labor and capital.

    When examining such negative output shocks from higher tax rates in the 1930s, however, Cole and Ohanian (2007: 33) claim, ‘Tax rates on labor and capital changed very little during 1929–33, which implies they were not important for the decline.’ In reality, marginal tax rates were virtually tripled in June 1932, retroactive to January of that year. The lowest marginal tax rate rose from 1.1 percent to 4 percent; the highest from 25 percent to 63 percent. There were also big increases in excise, corporate, and estate taxes (Blakey and Blakey 1932). Because Cole and Ohanian appear unaware of the Revenue Act of 1932, their estimate that ‘higher taxes played a relatively minor role’ surely understates the 1932 fiscal shock.”

  28. 28 David Glasner October 5, 2011 at 11:25 am

    Alan, I agree with most of this. However, I would take exception to a couple of points. The recovery actually began in March/April of 1933 immediately after FDR suspended the gold standard and the dollar price of gold began rising in world markets. A blistering recovery ensued until it was prematurely aborted by the NRA and a forced increase in money wages. Further dollar depreciation and increased money wages kept prices rising for a while, which accounts for the inflationary trend in 1934-35. The recovery resumed after the NRA was scrapped in 1935.but most of the growth in output stemmed from the boom in between the departure from gold and the start of the NRA.

  29. 29 Alan Reynolds October 8, 2011 at 2:05 pm

    I did not mean to suggest that devaluation was not immediately followed by a very strong recovery. I used annual GDP data in the Cato Journal, so I suppose my reference to the “1934-36″ recovery was unintentionally misleading about the timing.

    I was more explicit about the timing of the 1937-38 slump which almost exactly matched the increase and susequent easing of reserve requirements. Much as I would like to blame higher tax rates in 1936 or regulatory changes, both of the big gyrations in growth from 1933 to 1938 appear primarily monetary in origin.

    The modest CPI inflation of 1934-36 was a big improvement over the previous deflation, and I mention it only because it contradicts the Keynesian assumption that slack labor markets make inflation impossible. Obviously not.

    My main point here is to suggest that Cole and Ohanian made yet another mistake (about no tax hikes in 1932) and that Glasner, by contrast, is factually correct to emphasize the dominant role of money and the Fed after 1932, though not necessarily before (when trade wars, regulations and tax policy also involved global supply shocks).

  30. 30 David Glasner October 9, 2011 at 2:35 pm

    Alan, It seemed to me before that we were largely in agreement. Now, happily, it appears we are almost in total agreement. The only point that I would quibble about is that in the 1934-36 episode the increase in wage rates was imposed by government mandate or by a government sponsored enhancement of trade union monopoly power. So the standard Keynesian story about slack labor markets is not necessarily refuted by that special factual situation.

  31. 31 Rich Paul Freeman November 8, 2011 at 11:16 am

    What is important is to have a monetary system not open to manipulation by the people who run it. You cannot reduce the value of gold beyond the marginal cost of production of gold. That is because people want it for other things. Thus you cannot alternately create inflation (and buy real assets) and create deflation (and sell real assets) in order to manipulate the market and eventually manipulate it into leaving you and your friends will (literally) all the property in the world.

    I love Friedman too, but I believe that the Austrians had it right when it comes to fiat money. Friedman’s system would work if government could be trusted. It cannot. The Austrian proposal does not require a trustworthy government. Friedman said if he were appointed to the Federal Reserve board he would set the machines to increase the money supply by an even 3% forever. This would be fine. But how long before some politically connected banker talked them into firing you and putting in another Greenspan?

  32. 32 imtheoz2003@yahoo.com July 14, 2012 at 5:50 pm

    I’m sorry I got to the part of the Cole/Ohanian study where they said we just took the growth factors from the twenties and plugged them into the thirties to come to our conclusion and realized that these were two people who would say anything to protect their theory. They have PhD’s. How can anyone honestly say that they think this is a valid methodology.


  1. 1 TheMoneyIllusion » Glasner demolishes Cole and Ohanian Trackback on September 26, 2011 at 5:06 pm
  2. 2 Bad Faith Economic History - NYTimes.com Trackback on September 27, 2011 at 7:32 am
  3. 3 “Bashing Cole & Ohanian” from another angle | Historinhas Trackback on September 27, 2011 at 9:50 pm
  4. 4 CBO: Stimulus hurts economy in the long run - Page 11 Trackback on November 25, 2011 at 8:57 pm
  5. 5 Our minds are addled, the result of skillful and expensive propaganda « Fabius Maximus Trackback on December 27, 2011 at 4:03 pm
  6. 6 Too Little, Too Late? « Uneasy Money Trackback on December 12, 2012 at 9:16 pm
  7. 7 Skepticlawyer » Bubble trouble: not an easy money problem Trackback on May 14, 2013 at 8:21 pm

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s




About Me

David Glasner
Washington, DC

I am an economist at the Federal Trade Commission. Nothing that you read on this blog necessarily reflects the views of the FTC or the individual commissioners. Although I work at the FTC as an antitrust economist, most of my research and writing has been on monetary economics and policy and the history of monetary theory. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 280 other followers


Follow

Get every new post delivered to your Inbox.

Join 280 other followers

%d bloggers like this: