Another Angry Bear attack
Every few months I get attacked by the Angry Bear blog. Here Mike Kimel questions my interpretation of the 1933-34 inflation, which I attribute to dollar depreciation:
I see a much simpler story.
1.Aggregate demand was very slack when FDR took office.
2.FDR showed up in Washington with a plan to start spending a lot of money and thus boost aggregate demand.
3.The immediate effect was to convince factories they’d be running down their inventories. That boosted producer prices. It had a much smaller effect on consumer prices because everyone knew the gubmint was going to buy a heck of a lot more producer goods than consumer goods. (The government did buy some consumer goods for the various programs, plus there was a spillover effect, but as the graph clearly shows, the action was on the producer side.)
4.After a bit of time, the public realized FDR wasn’t planning just a one-off, but rather a sustained program of purchases of industrial items. That led them to start using some of their idle capacity, which meant not just selling the fixed amount that was in inventory. The rate of price increases thus dropped.
5.GDP increased at the fastest rate in the United States peacetime history since data has been kept. There was a big hiccup, of course, in 1937 when the government cut back on spending for a while.
There are all sorts of the problems with the argument that the inflation of 1933-34 was caused by expectations of fiscal stimulus. First of all, it’s completely at variance with Keynesian theory, which Kimel seems to be trying to defend. Keynesian theory says demand stimulus doesn’t raise prices when there is “slack,” and there has never been more slack in all of American history than in 1933.
But put that problem aside; assume the Keynesian theory did predict inflation. Prices didn’t start rising when FDR came to Washington with spending plans; they started rising when he began depreciating the dollar. Furthermore, the weekly rise in the WPI index was highly correlated with weekly increases in the dollar price of gold (i.e. currency depreciation.) And those changes (in gold prices) were caused by explicit statements and actions by FDR. Not by fiscal stimulus, which would be expected to appreciate the dollar.
In addition, the fiscal stimulus was far too small to produce the more than 20% WPI inflation observed in the 12 months after March 1933. So the fiscal stimulus theory of 1933-34 inflation fails on all counts. Indeed I’m almost certainly that Keynes himself would acknowledge that it was dollar depreciation that boosted prices in 1933 (he favored the policy), despite the fact that it conflicted with his “bottleneck” theory of inflation.
I don’t have time to fully address all his comments, but a few brief points. Kimel says:
So…. his story requires the devaluation of the currency to worsen the trade balance, and rational expectations to cause a one time explosion in industrial prices and a rather smaller recovery in consumer prices. Rational expectations, however, that came an abrupt halt, at roughly the same amount of time one would predict companies might decide that demand will be sustained enough to start producing more rather than just selling off inventory sitting in warehouses. And his story doesn’t explain why growth was so much faster during the New Deal era than any other period of peacetime since the US began keeping data, nor why there was the big hiccup in 1937.
No, my story doesn’t require devaluation to worsen the deficit, but economic theory suggests it can do so via the income effect. Lars Svensson has written on this subject. And my story does explain 1937, but I don’t have time to write a 400 page blog post on the Depression every time I comment on it. I’ve published numerous articles on the topic over the years. The 1937 downturn was caused by a combination of falling prices (due to gold hoarding) and rising nominal wages (due to the post-Wagner Act unionization drives.) I have no idea what the phrase “rational expectations . . . that came to an abrupt halt” means.
The growth in the New Deal period was very uneven. During March to July 1933 industrial production rose 57%. That ended with the NIRA wage shock, and IP was no higher in mid-1935 (when the NIRA was repealed) than in July 1933. If IP had continued rising after July 1933 at the rate of the fast 1921-22 recovery, the Depression would have been essentially over by the end of 1934.
When output falls sharply due to a demand shock, it’s possible to have very fast “catch-up” growth. Since the hole in 1933 was much deeper than at any time in US history, it’s not surprising that the growth rates during the recovery were quite rapid. But as late as mid-1940 the US economy was still deeply depressed. So it’s a glass half full/half empty situation. There were both good and bad aspects of FDR’s policies. Monthly data shows those effects quite clearly.
Tags: New Deal
13. November 2011 at 12:39
Sure, I am biased, but has anyone noticed that the anti-Market Monetarism arguments seem feeble, ambiguous, even peevish?
Market Monetarism has a structure and logic to it.
Crickey, let’s try Market Monetarism and see if it works. The worst result would be some inflation higher than otherwise (and even that is a positive in the current context).
13. November 2011 at 12:53
Ben,
I would love for NGDP targeting to work. It seems elegant and it would improve people’s lives. But I can’t see it working for the same reason targeting the money supply didn’t: the money supply is endogenous. This became clear with the failure of monetarism in multiple countries in the 1980s. Sumner et. al have not convinced me otherwise.
(It also doesn’t help that their arguments often come laden with right wing ideology – probably Sumner the least so, but generally it seems ‘market monetarists’ are willing to agree with anyone who is opposed to Keynes, whether it’s Austrians, monetarists or whatever else.)
13. November 2011 at 13:23
Sure, I am biased, but has anyone noticed that the anti-Market Monetarism arguments seem feeble, ambiguous, even peevish?,
Yes Ben, I feel the same way.
13. November 2011 at 14:29
The AngryBear people seem like nothing more than misguided partisan hucksters.
13. November 2011 at 18:09
Don’t forget about the Wagner Act. It had passed in 35, but was largely ignored. It was basically just the labor portions of the NIRA so everyone expected it to be overturned, but it was upheld in 1937, which kicked off a wave of strikes and union activity that coincides nicely with the downturn.
13. November 2011 at 18:09
Ben, Dustin and Tommy, There may be good arguments against MM out there, but I’m not seeing them.
Turner, You said;
“I would love for NGDP targeting to work. It seems elegant and it would improve people’s lives. But I can’t see it working for the same reason targeting the money supply didn’t: the money supply is endogenous.”
All monetary policy targets are endogenous. The inflation rate is just as endogenous as NGDP and M2.
I don’t know any market monetarists that are particularly close to the Austrians (in monetary theory.) You should read David Glasner and Nick Rowe, they are probably the least right wing.
13. November 2011 at 18:10
Cassander, I didn’t forget it, I mentioned it in the post. 🙂
14. November 2011 at 11:18
It seems to me that to the extent that Market Monetarists are monetarists you would not expect them to be friendly to the Austrians. Seems to me that both Keynesians and Monearists of any stripe are implicitly opposed to the Austrians who are against fiscal or monetary stimulus. Would you describe yourself as being opposed to Keynes, Scott?
14. November 2011 at 14:22
Scott> Whoops. Missed that paragraph somehow.
14. November 2011 at 18:19
Mike, Hayek also favored NGDP targeting, and he was an Austrian.
15. November 2011 at 00:02
“Prices didn’t start rising when FDR came to Washington with spending plans; they started rising when he began depreciating the dollar. ”
Sorry. Doesn’t work. The devaluations of which you speak didn’t begin until October 23, 1933. Go to FRED. Build a graph of the PPI, percent change from a year ago, from Jan 1933 to Oct 1933. (I’ll be putting that and a longer look graph in an upcoming post.) The price had started rocketing half a year before the devaluations, consistent with FDR showing up with plans to spend money, but not with the devaluations.
To give you an idea, the PPI rose 17% from Jan to Oct of 1933. It took until December 1936 to rise another 17%. In plain English, even cherrypicking the dates (going from March to October makes more sense) to make your story look as good as possible, I don’t see any way to argue that prices rose by a comparable amount, much less faster, after the devaluations than before.
The data is consistent with my points 3 and 4 which you cited above. It is also consistent with the rapid growth that followed being accompanied by low inflation. It isn’t consistent with devaluation of the dollar as a cause for growth.
15. November 2011 at 00:02
“1.Aggregate demand was very slack when FDR took office.”
Any time I see an argument starting with “Aggregate demand” I sigh to myself and shake my head. There is no such thing as AD. Demand is just a way of including everyone else’s supply in a partial equilibrium model, and really has no place in a GE model.
15. November 2011 at 05:04
Mike, You said;
“Sorry. Doesn’t work. The devaluations of which you speak didn’t begin until October 23, 1933.”
This is completely false. FDR began devaluing the dollar in mid-April 1933. I’ve spent 20 years researching 1933, I could recite all the macro data in my sleep.
Doc, There is AD, but it can’t be “slack.” Probably Mike meant that there was economic slack.
15. November 2011 at 05:26
[…] Sumner Digs Deeper On November 15, 2011 by Mike KimelScott Sumner Digs DeeperScott Sumner criticizes my most recent post in which I indicate that Keynesian theory explains growth rates during the New […]
15. November 2011 at 07:50
‘Doc, There is AD, but it can’t be “slack.” Probably Mike meant that there was economic slack.’
Humbug.
In a partial equilibrium model, demand represents the supply of the entire rest of the market and their utility preferences wrt the product. Demand is a way of breaking up a market so you can analyze it.
When you move to general equilibrium, all you have is supply and utility functions. Because you are in general equilibrium, you don’t have anything else on the other side of the trade that can be the demand. Then the demand-supply model does not apply.
If you just say that AS represents real goods markets and AD represents NGDP expectations (or some other nominal thing), then AS and AD are endogenous. Here the demand-supply model does not apply because the two curves are not exogenous.
I’m annoyed at people using partial equilibrium concepts as if they applied in general equilibrium when often they do not.
15. November 2011 at 11:54
Scott,
You know, before this I had never heard of any devaluations in 1933. I’ll admit, this isn’t my area of specialty, but I’m generally pretty good at finding data. I wonder how much real effect these devaluations had on any organization’s business. I note that the Fed’s own Banking and Monetary Statistics which show the value of monthly gold bullion set the price of gold for each month up to January 1934 at 20.67 an ounce. (See bottom paragraph, right column, page 522 here: http://fraser.stlouisfed.org/publications/bms/issue/61/download/134/section14.pdf I’ve seen the same thing in a number of other Fed and Treasury docs over the years, which explains why I never heard of this until now.) And if there are any organization that has an incentive to get that the value of gold right, its the Fed and the Treasury.
I’m having a great deal of trouble finding any series that shows something other than what the Fed used as the price of gold. (I.e., if you have access to such a series, please provide a link.)
That said, I’m assuming that, as you (and Jesse Jones) pointed out, FDR did devalue the currency. Let’s further assume that it had an actual effect, as opposed to being an announcement that was completely ignored. Now, assuming that, you actually have the opposite problem…. the effect doesn’t make sense.
Consider… I reread the Jesse Jones piece more carefully. He indicates that when he and Morgenthau had been called, the price of gold was at 29.01.
Now, 20.67 to 29.01 is a change of 40%. The price would go on to increase another 20% (to 35) by Feb 1934.
And yet by this time (i.e., by the time the price was at 29.01) the PPI had increased by 17%. It would increase by another 3% by Feb 1934. In other words, whereas the first set of devaluations were twice as large as the remaining set, the equivalent first set of PPI increases were six times as large as the last set of PPI increases.
If these devaluations had a real effect and were the cause of the changes in the PPI, why is it that the Jesse Jones / Morgenthau group of devaluations had such a small effect on the PPI relative to the earlier devaluations?
15. November 2011 at 12:51
Doc, It’s just a question of what’s convenient. I find it convenient to define AD as a given level of nominal expenditure. I would prefer the term “demand” not be used, because the concept has nothing to do with “demand” in a microeconomic sense.
Mike, The second period effects were smaller because the real value of gold was rising in late 1933, partly due to worries that other countries like France might be forced to devalue. If the public expected currency devaluation, they tended to hold gold (which would see its price increase with devaluation.) This increased demand for gold raised its relative value, and that meant deflation for all countries still on the gold standard. Consider a traded commodity for which PPP holds. If the dollar price of gold rises by 10%, then the price of traded commodities should rise 10% relative to the price in gold stnadard countries. But if prices are falling in gold standard countries, then the absolute increase may be much less than 10%.
One other minor point–the market price of gold it what matters, and it still hadn’t reached $35 in February 1934.
15. November 2011 at 21:37
@Scott:
“Doc, It’s just a question of what’s convenient. I find it convenient to define AD as a given level of nominal expenditure. I would prefer the term “demand” not be used, because the concept has nothing to do with “demand” in a microeconomic sense.”
Then AS and AD are not exogenous.
Which I guess is one reason NGDP targeting is so much better than CPI targeting.
Anyway, here is a thought: prices are more flexible than long term debt financing payments (at-least in one direction). And, we want the fed (if it exists) to target the least flexible target, so the stickiest target will be the least distorted. This suggests that NGDP targeting is a better than CPI targeting right there.
16. November 2011 at 16:40
Doc, Mankiw argued that wages are very sticky, and I also suggested that might be a good target. I view NGDP as the second best choice.
17. November 2011 at 04:35
“Doc, Mankiw argued that wages are very sticky, and I also suggested that might be a good target. I view NGDP as the second best choice.”
Yuck, no way. NGDP is a better target, because it not only takes into account wage stickiness but also the effects of nominal debt and cash flows (which affect riskiness).
Furthermore, wages don’t only rise from nominal effects but also from productivity. This means in a wage targeting regime, the monetary authority will respond to positive supply shocks with monetary loosening and adverse supply shocks with monetary tightening. THIS IS BAD!
NGDP targeting (or free banking) doesn’t have this problem.
17. November 2011 at 04:37
Also, debt payments are at a micro level much sticker than wages in one direction.
17. November 2011 at 19:04
Doc Merlin, The fact that wages rise from productivity is precisely why they are the best target. You keep nominal wages stable and let productivity reduce prices.
28. March 2013 at 13:39
[…] Sumner criticizes my most recent post in which I indicate that Keynesian theory explains growth rates during the New […]