Blame France 3: Dump A Pile of Argle-Bargle On Their Heads
August 7, 2016
We’ve been talking about the notion that France had something to do with the Great Depression — in particular, activities of the Bank of France in 1928-1933. We looked at a lot of background material (including a lot of charts) earlier:
One of the main descriptions of this idea is the 1998 book Gold, France and the Great Depression, 1919-1932, by H. Clark Johnson. Apparently, Johnson was a student of Robert Mundell.
However, we will look at a little more recent, and also more compact, expression of this idea, in this paper:
“The French Gold Sink and the Great Deflation of 1929-1932” (2012), by Douglas Irwin.
I won’t go into the paper line-by-line, but I will try to summarize the main points.
Here is the entire abstract (summary) of the paper:
the tightening of U.S. monetary policy in 1928 is often blamed for having initiated the downturn, France increased its share of world gold reserves from 7 percent to 27 percent between 1927 and 1932, and failed to monetize most of this accumulation. This created an artificial shortage of gold reserves and put other countries under significant deflationary pressure. A simple calculation indicates that the United States and France shared the blame (in a 60/40 split) for the withdrawal of gold from the rest of the world and the onset of worldwide deflation in 1929. Counterfactual simulations indicate that world prices would have been stable during this period, instead of declining calamitously, if the historical relationship between gold reserves and world prices had continued. The deflation could have been avoided if central banks had simplymaintained their 1928 cover ratios.
What does all this mean?
The author begins with the idea that there was a “monetary deflation” at the onset of the Great Depression. This is a popular notion, because, if you are stuck in the Prices Interest Money box, as virtually all economists of the time were and most economists today, then you can’t really explain the Great Depression without relying on Money. Prices and Interest don’t really get you there, as these were reasonably (though perhaps not entirely) flexible at the time. The Keynesians blamed an “autonomous decline in aggregate demand,” which doesn’t really mean anything at all. This was a common notion in the past — in part because people before 1960 or so were a lot more hesitant about blaming Money. However, the “it just happened and we can’t explain why” version has been dissatisfying over time, so people have been more and more enthusiastic about relying on Money for their explanations. This has been enabled by the general deterioration of understanding of monetary matters over the decades.
Maybe you don’t believe me when I say that common Keynesian interpretation was that “it just happened for no good reason.” Don’t these people have PhDs or something? Seriously?
In Golden Fetters: The Gold Standard and the Great Depression, 1919-1939 (1996), an influential book with a Keynesian bent, Barry Eichengreen says: “The initial downturn in the United States enters this tale as something of a deus ex machina.” He continues: “The tightening Federal Reserve policy of 1928-29 seems too modest … Hence the search for other domestic factors that might have contributed to the severity of the downturn, such as structural imbalances in American industry, an autonomous decline in U.S. consumption spending, and the impact of the Wall Street crash on wealth and confidence.”
The search was apparently unsuccessful, as he has little more to say on the topic.
If that is what you have — and it is a pretty thin gruel as even Eichengreen admits — it is not surprising at all if people then latch upon Money as their explanation, especially as Money is all that is left when you are in the Prices Interest Money box.
The basic problem with Money, and the reason why people didn’t blame Money until, decades later, their understanding of Money had deteriorated sufficiently that their theoretical inconsistencies didin’t bother them, is that you really can’t get there with Money.
As I’ve been saying over and over again, in different ways, you have some basic propositions regarding the gold standard system:
2) Base money naturally adjusts — via bullion conversion if by no other route — to maintain the value of the currency at the gold parity. Thus, central banks cannot just go expanding or contracting the monetary base willy nilly, or their currencies’ value would deviate from their gold parities. They would become floating currencies. We looked at this in detail with the Federal Reserve. None of the Federal Reserve’s actions during 1927-1928, or even the gigantic government bond purchase program of 1932, made any difference to base money. Gold conversion canceled it all out. If you think of a gold standard system as something like a “currency board linked to gold,” with all of the automatic characteristics of a currency board, this is perhaps easier to understand.
June 12, 2016: Milton Friedman Blames the Federal Reserve
June 5, 2016: Irving Fisher and “Debt Deflation”
May 22, 2016: The Devaluation of the British Pound, September 21, 1931
May 14, 2016: Credit Expansion And Contraction Of The 1920s and 1930s #2: Paying Off Debt
April 3, 2016: Credit Expansion and Contraction in the 1920s and 1930s
February 7, 2016: Blame Benjamin Strong 2: So Obvious It’s Hard To Believe
January 31, 2016: Blame Benjamin Strong
There was no “contraction in Federal Reserve monetary policy” in 1928. There was a contraction in “Fed Credit”, but this was offset by gold conversion inflows, resulting in no meaningful change in the monetary base. This is exactly what one would expect under a gold standard system.
France did increase its gold reserves, but this doesn’t really matter. The value of the Franc would have been unchanged whether gold reserves were large or small. The value was linked to gold at the gold parity. We saw that much of this bullion accumulation had to do with the overt decision, in 1928, to reduce foreign exchange reserve holdings and to own bullion instead — a return to the regular operating practices of the Bank of France pre-1914. Swapping one asset for another doesn’t change the monetary base, or monetary conditions.
There was no “artificial shortage of gold reserves.” Any central bank could have done just as France did, and swapped some of its lending and bondholdings (foreign or domestic) for bullion. The reason that they didn’t is because central banks wanted to keep some portion of their reserves in interest-bearing debt. The inflows of bullion into France didn’t deplete bullion holdings elsewhere. We saw that, while the 1928-1929 bullion swap was taking place, U.S. gold reserves actually rose by a large amount. Bank of Engalnd gold reserves fell, but by a small margin, and that was most easily explained by the increase in government bond holdings at the time. There is a vague claim that increases in bullion held by France/U.S. caused a reduction in bullion holdings in the rest of the world. How did this happen? The U.S. held no foreign exchange. France’s foreign exchange was held in British pounds and dollars. If the Bank of France redeems British pound base money for gold, why does that make gold fly out of the vaults of the Bank of Poland? Don’t they have locks?
As the Bank of France bullion swap began in June 1928, all central banks in the world except for France and the U.S. held 7,100 tons of bullion. In June 1929, when it ended, they held 7,102 tons. France’s holdings rose from 1,710 tons to 2,160 tons during that time.
“The deflation could have been avoided if central banks had simply maintained their 1928 cover ratios” is a roundabout way of saying: central banks should have expanded their monetary bases. This is the old “excess reserves” argument — supposedly, the idea that having bullion reserves above the regulatory minimum means that central banks have carte blanche to expand at will. This is nonsense, and the Fed proved it in 1932. If you expand base money by purchasing bonds, the only result is that gold conversion cancels it all out so that total base money is unchanged.
Those are the big arguments, as presented in the abstract. Let’s look at some of the details:
This is nonsense. Any central bank could have purchased bullion on the open market, just as central banks had been doing for nearly a century. How do you think they got their gold in the first place? Santa Claus? How would purchases of gold by France cause another central bank — let’s say the Bank of Italy — to be unable to puchase gold, just as France did?
Besides, even if it were true — which it is not — the values of currencies, and consequently their base money supplies, would have been the same. The amount of gold in a vault determines neither.
This is another way of saying “The increase in bullion reserves at the Bank of France had no effect on other central banks’ bullion reserves at all.” Why should it?
By the time of France’s second reserve-asset swap, from October 1931 to June 1932, the Bank of France’s bullion reserves rose from 3,501 tons to 4,843 tons. The rest of the world ex-France and U.S. rose from 6,880 tons to 7,009 tons. But, by this time many central banks had already left the gold standard, so it mattered even less. The only way that Irving can get the negative numbers he wants is to use “proportions of total gold reserves.” It should be obvious that if one central bank’s proportion rises, for whatever reason, then another’s must fall, since there are only one hundred percentages available. The only way you would avoid negative numbers is if the proportions never changed. But we are supposed to believe that these negative numbers caused the Great Depression, because they represent “deflation.”
This massive redistribution of gold reserves might not have been a problem for the world economy if the United States and France had been monetizing the gold inflows. That would have been playing by the rules of the game of the classical gold standard: countries receiving gold inflows would have had a monetary expansion that would have balanced the monetary contraction in countries losing gold. But as already noted, there were no agreed-upon rules of the game in the interwar gold standard. Both France and the United States were effectively neutralizing—or not monetizing—the inflows to ensure that they did not have an expansionary effect. The United States was explicitly sterilizing the gold inflows by conducting open market operations (purchases of Treasury securities) to offset their monetary impact. France was not explicitly sterilizing the gold inflows, but the inflows failed to have much expansionary effect on the country’s monetary stance for reasons that will be discussed below. As a result, this asymmetry—countries receiving gold failed to expand, while countries losing gold had to contract—gave the international monetary system a severe deflationary bias.
There’s a lot of argle-bargle here, which is difficult to untangle. (That’s why these line-by-line kinds of examinations tend to be troublesome.) One of the basic notions — a wrong notion — is that there is some kind of “price-specie flow mechanism” at work, in which gold inflows by one central bank is matched by gold outflows by some other central bank. This is completely wrong. Gold “inflows” happen by two basic mechanisms. One is a reserve asset swap, which the Bank of France was doing. They took their British pound foreign reserves, and bought gold with it. This has no effect on the monetary base. The other mechanism is gold conversion. If the value of the franc is above its gold parity, then the Bank of France becomes the highest bidder in the world market for gold, and everyone sells their gold to the Bank of France. This produces a gold conversion inflow, which is matched by an increase in base money (necessary to fund the purchase). The result is an increase in base money, which tends to depress the value of the currency back to its gold parity value, at which point gold inflows cease.
This process is basically independent of all other central banks. There is no “price-specie flow mechanism.” Thus, there is no “asymmetry,” since the “symmetry” that they are referring to is the idea that gold inflows here must me matched by gold outflows there. Giulio Gallarotti, in Anatomy of an International Monetary Regime: The Classical Gold Standard (1995), effectively demolishes all these old fallacies, and presents a picture very much like the one I am presenting here.
Now, as we saw earlier, the Bank of France actually did do exactly what I described. There was a reserve-asset swap, British pounds for gold, which did not change the monetary base. Then, there was an additional large gold inflow, related to increasing demand for francs and thus a tendency for the franc’s value to rise vs. its gold parity, which was accompanied by equivalent expansion in base money.
Irving says that “The United States was explicity sterilizing the gold inflows by conducting open market operations (purchases of Treasury securities) to offest their monetary impact.” This is a conceptual error — to offset the inflow, you would have to sell bonds. The U.S.’s gold inflow was associated with a decline in bills discounted in 1930. It was actually the decline in “Fed Credit” which was the primary cause of the gold inflow. The reduction of base money inherent in the reduction of bills discounted tended to cause the dollar’s value to rise vs. its gold parity. This produced gold inflows cancelling the decline, with the result that base money was essentially unchanged, from what it would have been if no action had been taken.
The paper has a few vagaries related to the idea that an “artificial shortage of gold” resulted in what amounts to a gigantic rise in the value of gold. I still haven’t dealt with this in detail, but we did see earlier that there was no unusual increase in bullion holdings at central banks throughout the 1920s and 1930s. The supposed mechanism causing this rise didn’t exist.
Finally, the last paragraph:
Here we find what all of this really amounts to: it is a long-winded excuse for central banks to deal with the problems of the Great Depression with a big dose of money-printing — and explicit devaluation. Supposedly, this was all a long argument that France or someone was not really adhering to the principles of the gold standard; but now we find at the end that their mistake was that they were, and should have been expanding/devaluing instead! This prescription is no different than that given by Eichengreen (among Keynesians) or Friedman (among Monetarists). The reason that the arguments don’t really make sense is that they aren’t supposed to make sense. They are just supposed to render you sufficiently battered by nonsensical details that you are ready to accept the conclusion. We also find that Robert Mundell is also basically a devaluationist, stuck in the Prices Interest Money box even though he was instrumental in finally breaking the walls of that box in the 1970s. But, it really took others to pick up that ball and run with it. Mundell himself never really got out of the box, but he did make a hole in the side that others eventually escaped from.
For the most part, this paper amounts to the “throw a lot of spaghetti at the wall and see what sticks” kind of argument. What is the argument exactly? Was it that gold’s value rose? That France “sterilized” gold inflows? That the increasing proportion of world gold reserves held by France had some kind of importance? That there was some kind of “asymmetry” with other central banks? That the fact that the increase in base money in France was not reflected in M2 had some importance? That central banks’ “excess reserves” gave them a free pass to print money willy-nilly, which they didn’t take advantage of? All of those together? Basically, the technique here is to dump enough argle-bargle on people’s heads, combined with some heated rhetoric, that they become docile and cooperative. And, it works for most people.
It is a little difficult (or should be) to make the claim that France was somehow not operating the gold standard system properly, since the goal of a gold standard system is to maintain the value of the currency at the gold parity value. Failure is expressed by a currency that deviates from this parity value; or, in a lesser way, with a parity value that can be maintained only via coercion such as capital controls or “foreign exchange intervention.” Rather, the franc maintained its gold parity value through some very difficult circumstances, 1929-1936, as did the U.S. dollar. If anything, they were examples of extraordinary success. If the Bank of France really wasn’t supplying as much base money as it should have, it would have been expressed by a rise in the franc’s value vs. its gold parity.
Instead, I assert:
2) Currency values were demonstrably at their gold parities. Thus, there was no rise in currency value (vs. gold) causing “monetary deflation.”
3) Base money is basically a residual: it is what it needs to be to maintain the gold parity value.
Let me say again that my version of things was basically the commonly-accepted version of things, until around 1960. As Bank of England chief Montagu Norman said: “There was nothing we could do.” Not “we did nothing,” but “nothing we could do.” Why not? Because they had to maintain their gold parities, and base money was basically a residual of this obligation. There was no avenue for meaningful discretionary action, as the Federal Reserve demonstrated in abundance in 1932. This is true of any currency board today. If you just think about how today’s currency boards work, and simply change the “standard of value” from another currency to gold, you can get a good idea of what was going on in the 1920s and 1930s.