My Role Model, George Warren

One can learn a lot about monetary theory from studying either commodity standards or fiat money regimes.  But perhaps the most illuminating examples come from an in between system, a transition from one commodity regime to another.  Although most of this (very long) post will seem far removed from current issues, in the end I will argue that there are important lessons.  We will look at 1933, the year of transition from one gold standard (1879-1933) to another (1934-68.)

Let’s begin by noting that the price level in the US in 1933 was about the same as in the 1770s (now it is at least 20 times higher.)  Why was that?  Because changes in the price level are the inverse of changes in the value of money, and the value of gold generally doesn’t tend to show dramatic long term changes.  In the short run, however, shocks to gold supply and demand can create a highly unstable price level.

Robert Mundell argued that the undervaluation of gold after WWI was a root cause of the Depression.  (His student Clark Johnson wrote an excellent narrative of gold issues during the interwar period that was informed by Mundell’s insight.)  The fall in European gold demand during WWI pushed price indices far above trend.  Some of this was reversed in the sharp 1920-21 deflation, but in 1929 prices were still well above pre-war levels.  The situation was precarious, as European central banks were rebuilding gold stocks that had been depleted in the war.  More demand for gold drives up its value, i.e. is deflationary.

We all know what happened next (well not exactly, but I’ll explain that in another post), so let’s jump ahead to 1933.  FDR takes office in March, promising to boost wholesale prices back up to pre-Depression levels.  He uses several tools, but the most effective was loosely based on Irving Fisher’s “compensated dollar plan.”  Fisher’s plan was to raise the price of gold one percent each time the price level fell one percent.  An obscure agricultural economist named George Warren was a big fan of Fisher’s idea, and sold it to FDR with all sorts of fancy charts.

And it worked.

Initially it worked better than any other macroeconomic policy in American history.  But at first the policy’s success was mostly accidental, just a matter of talking the dollar down, not enacting Fisher’s specific plan.  Nevertheless, prices immediately began rising sharply.  Industrial production rose 57% between March and July, regaining over half the ground lost in the previous 3 1/2 years.  Then in late July FDR decided to cartelize the economy and sharply raised wages (the NIRA) and industrial output immediately began falling.  By late October FDR was desperate for another dose of inflation, and asked Warren to come up with a plan.  They decided to have the US government buy gold at a price that would be continually increased in order to reflate the price level.

It was widely expected that the US would return to the gold standard at some point.  The pre-1933 gold standard had a gold price of $20.67/oz., but most thought the new price would be in the high $20s.    That’s what Keynes supported.  But Warren convinced FDR that $28 was far short of what was required to regain the 1926 price level.  Warren advocated a much more radical policy of devaluing the dollar until the commodity price indices regained their 1926 level.

It was a very confusing plan, as they never bought enough gold to equate the government buying price with the free market price in London.  I studied this program intensively and eventually concluded that FDR was playing a sort of game with the markets.   The rising gold-buying price sent a signal to the markets about where FDR might re-peg gold.  (He eventually settled on $35/oz.)

Before proceeding we need to consider what sort of monetary policy we are dealing with here.  In my earlier research on the years 1929-32 I had gradually come to the conclusion that what mattered for aggregate demand wasn’t so much changes in the current setting of the monetary instrument (money supply or interest rates) but rather changes in the future expected path of that instrument.  When I began studying 1933 I immediately recognized that this approach was especially well-suited for analyzing FDR’s dollar devaluation program.

I suppose this way of thinking came out of the rational expectations revolution, but don’t just write this off as Chicago macroeconomics; the prominent new Keynesian Michael Woodford also emphasizes the importance of changes in the expected future path of monetary policy.  His student Gauti Eggertsson (Iceland’s greatest economist) applied this approach to the 1933 case and reached similar conclusions, albeit using a far more sophisticated, and more new Keynesian model than mine.  Krugman recently praised Gauti’s Sept. 2008 AER paper (which cites three of my papers), so my analysis hasn’t moved far out of the mainstream . . . yet.

So FDR planned to continually raise the buying price of gold until commodity prices returned to their 1926 level, and then re-peg the dollar to gold at that new level.  Astute readers may have noticed a problem here, if the markets recognized FDR’s game, wouldn’t commodity prices immediately shoot up to 1926 levels on the announcement of the program?  But then how would FDR know what price at which to re-peg gold?  This is the “circularity problem” that Bernanke and Woodford addressed in their 1997 JMCB paper attacking monetary policies that targeted market forecasts.  (Later I will argue that flaws in that paper might help explain our current predicament.)  In any case, FDR was lucky that his policy initially lacked credibility–so it worked better than it should have.  Let’s see how long it took markets to figure out FDR’s game.

Not long.  With each passing day the New York Times analysis became increasingly sophisticated.  It’s almost as if the NYT discovered the theory of rational expectations 40 years before Lucas.  [Note to noneconomists–this is not a slight against Lucas, but rather a subtle form of praise.  Lucas’ theory implies that the public had always been using rational expectations, it was the eggheads in academia that took decades to figure it out.]  Here’s an example of how the NYT saw the program as a signal of future policy intentions:

“The behavior of the foreign exchange market yesterday offered another illustration of the fact that the decline in the dollar is due not so much directly to the daily advances in the RFC’s gold price as to the government’s avowed intention of cheapening the dollar which these rises in the gold quotation indicate. Yesterday the RFC’s gold price was left unchanged from the previous day, but the market ignored the implication that the RFC was content for a moment not to push further the cheapening of the dollar.  Instead it fastened its attention on the news of the changes in the Treasury Department’s administration, which it interpreted as still further reducing the influence of those opposed to drastic dollar depreciation.” (NYT, 11/16/33, p. 37.)

The “changes” they are referring to are a steady succession of resignations in protest over the program, starting with Treasury Secretary Woodin, followed by key figures like Acheson, Sprague and Warburg.  The program was also too radical for Keynes, even too radical for the socialist Debs (although I presume this was because it reduced real wages.)  Meanwhile there were large public protests and counterprotests over the policy.  It was the most politicized monetary policy moment of the 20th century.  Irving Fisher stuck with Warren and FDR, but almost all other respectable opinion in America and Europe thought Warren was a crackpot.  As do almost all modern economic historians.  Except me.  Eventually the political pressure became so great that FDR was forced to abandon the program before his price level objectives had been reached.

BTW:  When I criticize “Bernanke” in this blog, it should not be viewed personally.  As the demise of the gold-buying program illustrates, he probably lacks the political power necessary implement the sort of radical policy change that I think is needed.  He was my first choice to be Fed chairmen in 2006, and I doubt any other economist would have done better.

There is a great deal of evidence that I won’t get into here that suggests the suspension of the gold standard in March 1933, and gradual devaluation between April and February 1934, almost certainly explain most of the increase in goods prices, stock prices, and industrial production during that period.  But why?  Not because it boosted our trade balance, which actually worsened as the rapid recovery pulled in imports.

Both Gauti and I believe that only the rational expectations hypothesis can explain these events.  He focuses on how the regime change led to higher inflation expectations, and thus reduced real interest rates.  I prefer to think in terms of specific policy signals sent as rising gold prices changed the future expected gold price, and hence the future expected money supply.  I don’t see any non-Ratex explanation that can account for the extraordinary rise in prices and output during March-July 1933.  Nominal interest rates didn’t change much, and open market purchases in 1932 (under the constraint of the gold standard) had accomplished little or nothing.

One reason FDR abandoned the program is that many economists told him that the devaluation was already plenty big enough to restore the 1926 price level.  They said that gold would flood into the U.S. after the gold standard was restored, and the monetary base would soar.  They were right about the gold flows and the base, but not about commodity prices.  Only Warren understood that the rational expectations hypothesis implied that expected future growth in the money supply was already embedded in commodity prices, and the markets were telling FDR that more inflation was needed.  Here is a quotation from Warren’s personal notes a few months later (describing FDR’s disappointment when commodity prices stopped rising in 1934):

“The President (a) wanted more inflation and (b) assumed or had been led to believe that there was a long lag in the effect of depreciation.  He did not understand–as many others did not then and do not now–the principle that commodity prices respond immediately to changes in the price of gold.”

I don’t have the data at home, but I believe the WPI rose over 20% during the devaluation, and then only another 5% or so between 1934 and mid-1940.

Today it seems like the Fed is just waiting around to see if their low interest rate policy will work.  And pundits endlessly debate whether the fiscal stimulus package will succeed.  As far as I’m concerned the verdict is already in.  What matters for policy is not whether something “will work,” but rather whether the optimal forecast suggests it is expected to work.  As Lars Svensson (another guy recently praised by Krugman) emphasizes, policy should adopt a stance that it expected to hit the policy goal.  Right now the stock, commodity, and bond markets are signaling that the current policy stance is woefully inadequate.  Indeed, even the Fed’s internal forecast suggests that we will fall short of their own implicit target.

Because I talk a lot about the early October stock market crash, which led me to develop my critique, I anticipate some will accuse me of advocating that we allow “Wall Street” to lead fiscal and monetary policymakers around on a leash.  That is wrong for three reasons:

1.  I don’t favor fiscal countercyclical policies (unless done Singapore-style through payroll tax cuts.)  And we can’t really afford that because unlike Singapore we foolishly run budget deficits.

2.  The stock market can be a highly misleading indicator, as in 1987.  Rather we should at a minimum consider stock, commodity and bond markets as a whole.

3.  Even point 2 is inadequate.  As Robert Shiller and I have emphasized for years, we desperately need a nominal GDP futures market.  It’s time the Fed set up and subsidize trading in such a market.

I started out by suggesting that the 1933-34 transition offered an exceptional opportunity to see how monetary policy affects the economy.  The reason is a bit complicated.  Let’s start with the assumption that the public expected the US to eventually return to the gold standard at a higher peg that would persist for years, and that (due to intertemporal arbitrage) the spot price of gold was a pretty good indicator of the expected future peg, and that the expected future peg was strongly correlated with the expected future money supply and price level.

(If all that was to hard to follow, picture the gold standard price level fluctuating for decades around a fairly horizontal line when gold was at $20.67/oz. and then zigzaging around a much higher horizontal line when gold was reset at $35/oz.  And by the way, the “zig zags” were often large, and very important.  Nor was the line precisely horizontal)

Do we have any evidence that the interwar public thought of things this way?  Paul Einzig understood interwar psychology better than any of us can today:

“The most common argument which was used in attempting to persuade the United States to agree to the currency truce [in June 1933] was that it would not be binding upon her future monetary policy.  In light of practical experience this argument sounds most unconvincing.  It has been found that, whenever an inconvertible currency is kept stable for a while at a given level, it becomes increasing difficult to change that level without provoking strong opposition.”

Einzig was right, the $35/oz peg adopted in 1934 lasted until 1968.

In 1933-34 we can clearly see a monetary policy transmission mechanism that is usually hidden by a thick fog that economists call “the identification problem.”  Woodford and Eggertsson are right that monetary policy shocks are changes in the future expect path of policy.  But normally those shocks are very hard to identify.  In 1933-34 FDR used gold prices as a policy instrument, allowing us to see clearly that which is usually obscure.

To summarize, the lessons I draw from 1933 are:

1.  Policy lags are less of an issue than most monetary economists assume.  I’m not worried about inflation.  But I will have to put off until a future post the specific identification problems that I see with postwar studies of policy lags, as well as the interwar shocks identified by Friedman and Schwartz.

2.  Rational expectations are essential to understanding monetary and fiscal policy.  The models of the smartest left wingers (such as Krugman’s “expectations trap”) and the smartest right wingers, are all structured around the assumption of rational expectations.

3.  In my view the monetary policy transmission mechanism can be explained as follows:  Monetary shocks change the expected future path of the money supply.  This changes the expected future path of nominal GDP through the excess cash balances mechanism (I’ll explain that mechanism later to noneconomists.)  A change in the expected future path of nominal growth (or “aggregate demand”) changes the current level of AD.  This last point is accepted by new Keynesians, although they prefer to think in terms of AD being affected by real interest rates, not excess cash balances.  (And later I’ll argue that this last point is also what Keynes may have meant by the mysterious term “confidence.”)

Is George Warren really my role model?  Actually I’m just being a bit provocative here.  Obviously Friedman and Schwartz are my real role models.  (Someday I’ll do a post on how their Monetary History changed history, and how it offers an implicit rebuke to the “Official Methodology” of modern macro.)  Yes, the “gold and prices” model of Warren and Pearson is crude and simplistic, but consider the following:

1.  It was Warren who understood that more monetary stimulus was needed in late 1933, when Keynes and the official establishment were almost all opposed.

2.  It was Warren who saw through the establishment’s argument that “policy lags” hid an inflationary time bomb that was waiting to explode.

3.  It was Warren who understood the need for forward-looking monetary policy–policy that was guided by expectations embedded in auction-style market prices.

So this obscure economist was battling almost single-handed against the economic establishment.  If you’ve read my heterodox views on current policy in previous posts, perhaps you’ll understand why I find him such an appealing role model.

PS:  I may be so close to this subject that I simply assumed causal connections that are not obvious.  Let me know if any clarification is necessary.  I published a much longer paper on the gold-buying program in 1997–it is actually far more complex and interesting than what I discussed here.  If my 3 year old manuscript on the Depression ever finds a publisher, Chapter 9 will have the definitive account.  Also, since this is more like a full length article than a short post, I should mentioned that I failed to cite numerous important economists who anticipated some of my views on 1933, including Peter Temin and Barry Eichengreen.

PPS:  Yesterday the Boston Herald interviewed Brad Delong.  Brad said he thought I was a right winger, and that he didn’t agree with many of my views.  Nonetheless I appreciate that despite ideological differences Brad called me “thoughtful.”  I thought his paper on monetarism was very good, and cite it frequently.   I am not a big fan of FDR, but given Brad’s remarks about me being right wing it is fun to be able to get to the left of many of his admirers in this particular post.  (Now if only Paul Krugman would use the phrase ‘thoughtful right winger’ more often.)

PPPS:  You don’t know how liberating it is to write an article knowing people will actually read it, and to not have to ruin the article in order to please potential referees.  Thanks for getting to the end of this very long post.

Tags:

This entry was posted on February 28th, 2009 and is filed under Forecasting, Heterodox macro, Monetary History, Monetary Theory, Rational Expectations. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response or Trackback from your own site.