My review of Sumner

This will be coming out in The Journal of the History of Economic Thought, but probably not for another year or so. In the meantime, you get it first:


Sumner, Scott B. The Midas Paradox: Financial Markets, Government Policy Shocks, and the Great Depression. The Independent Institute, 2015. xviii, 507 pp. Hardcover, $37.95.

The Midas Paradox is the result of Scott Sumner’s many years of contemplating the causes of the Great Depression. Sumner adopts a novel “gold market approach” in this work, holding that the demand side shocks largely responsible for the worldwide economic downturn were increases in state and private actors’ desire to hoard gold. The “paradox” of the title arises from the logical fact that, with a relatively fixed amount of gold available, it is impossible for all market participants to increase their gold holdings simultaneously. When they try to do so, the result is instead a decrease in economic activity.

The heart of the book is nine chapters of detailed historical analysis of the macroeconomic history from 1929 through 1938, employing Sumner’s gold market approach. To brutally abbreviate Sumner’s narrative, in 1929, monetary policy mistakes, particularly the failure of France and the United States to print an amount of currency commensurate with their rapidly increasing gold stocks, led to the stock market crash and banking panics. Roosevelt’s decision to inflate by taking the US off the gold standard spurred a recovery that was nipped in the bud by the NIRA’s high wage policies, which delivered a supply side shock. The economy recovered again, albeit less than it might have, due to repeated attempts to raise wages by the Roosevelt administration. Then the final Great Depression setback occurred when private gold hoarding caused the “Roosevelt depression” of 1937.

Sumner is a diligent researcher, so there is no point in this reviewer disputing the facts on the ground as he describes them. Instead, this review will focus on the methods he uses in handling these facts, which sometimes oscillate uneasily between historical and theoretical analysis.

Throughout this work, Sumner is a strong proponent of the efficient markets hypothesis and the “wisdom of crowds.” For instance, he criticizes those who think anyone can reliably outguess financial markets: “A modern example of this conundrum occurred when many pundits blamed the Fed for missing a housing bubble that was also missed by the financial markets” (p. 12).

What should we make of this statement? We can divide bubble theories into three broad categories: collective irrationality theories, partial information theories, and prisoner's dilemma theories. In collective irrationality theories, a “mania” gets going in some market, and market participants buy because they are carried away by their “animal spirits.” Per these theories, someone outside the market (such as the Fed!) might be able to spot the bubble precisely because they are not swayed by the emotions influencing market participants.

In partial information theories, a bubble can occur when a random price movement is mistakenly taken by market participants, who lack perfect knowledge, as a sign that someone else knows more than they do. Therefore they follow this random jiggle, leading others to follow it as well. Theoretically at least, if we have a bubble of this sort, an agency like the Fed might have better information at its disposal than market participants, and so be able to spot the bubble.

Finally, in the prisoner's dilemma theories, market participants may be well aware that a bubble is inflating, but their best strategy is to try to profit from the bubble as long as it is expanding. Since these theories typically posit that a “big player” (e.g., the Fed) is actually causing the bubble, the Fed is also clearly capable of ceasing to cause it at any point in time.

Perhaps what Sumner is thinking is that, although there may have been bubble in the housing market, the market for, say, mortgage-backed securities ought to have been acting on this fact, if it was detectable at all. But it is hard to see how such action in the MBS market wouldn't have prevented the housing bubble itself, since cutting off the flow of mortgage funding would seem likely to have done so.

So, perhaps Sumner just meant, “Bubbles can't exist, because markets won't allow them.” But if so, he phrased this oddly.

It is hard to know what to make of some Sumner’s other claims as well. For example, he writes:
Romer argued that the 1929 stock market crash sharply reduced consumer confidence, and that this was a major factor depressing aggregate demand. But the quite similar stock market crash in 1987 seemed to have no impact at all on economic growth, suggesting that the direct impact of stock prices on real output is certainly very small. (p. 39)
What we have here is an oscillation between an historical mode of explanation and a scientific one. Science deals in generalizations, and has nothing to say about specific events other than citing them as instances of its generalizations. (“The arrow will fall back to earth, as it was shot in the air with less than escape velocity.”) History, on the other hand, deals with particulars, and a resort to scientific generalizations is at most a stopgap in an historical explanation, filling in where we lack actual historical knowledge. (“Although we have no evidence for what happened in that battle, the Egyptians probably won, as armies with such a preponderance of force typically do.”)

But a stopgap generalization like that above cannot be used to defease an historical narrative based on actual evidence: it is silly to write, “It is unlikely that some obscure Macedonian named Alexander conquered the Persian Empire, as armies facing such an imbalance in force are almost always defeated.” We are dealing with extensive evidence that these things did happen, and we can only reasonably dispute that evidence with other historical evidence showing, for instance, a widespread conspiracy to fake the existence of Alexander.

So even if, somehow (based on only two cases!) Sumner has established a scientific generalization that “generally, stock market crashes have very little impact on real output,” that says nothing about the actual historical evidence for whether, in the particular case of the Great Depression, the 1929 crash had a large impact. What starts a panic in one crowd in one place and time may be shrugged off nonchalantly by another crowd in another place and time; events are what they are interpreted to be by human beings, and human interpretation is not constant.

Or consider the following passage:
On June 4, 1928, the New York Times (p. 4) reported 'Credit Curb Hinted by Reserve Board.' The market actually rose slightly on June 5, but then, over the following week, the Dow plunged 7 percent. Policy news ought to be incorporated into securities prices almost immediately, and thus, it is unclear whether the Fed's announcement had any impact on the markets. (47)
This is an odd way to do historical research: one goes in knowing what should happen—policy news ought to be incorporated into securities prices almost immediately—and then creates the facts—“it is unclear whether the Fed's announcement had any impact”—based on the pre-existing theory. There is a sound historical way to determine whether the news about the Fed was what “caused” this drop: detailed examination of the journals, memoirs, letters, newsletters and so on of stock market participants of that time, seeking to determine if this was really what was worrying them.

Similarly, Sumner writes, “in order to understand the October [1929] crash, one needed to explain why it would have been sensible for investors to be highly optimistic in September 1929, and somewhat pessimistic in November 1929” (60-61).

Again, Sumner is introducing his conclusions as a criterion for what facts will be acceptable. Of course, no one embarks on an attempt to explain some historical episode with a “blank slate”: every effort at understanding is an effort to understand better what is to some extent already understood. There is nothing wrong with Sumner starting with the hypothesis that investors were “sensible” in September 1929, and seeing if it holds up. But here Sumner posits their being sensible, not as a conjecture to be explored, but as a given which any explanation must incorporate. And given that is a pre-condition he has placed on any acceptable explanation of what occurred, it is inevitable that the end result of his inquiry will be that they were, indeed, sensible.

While the history Sumner describes is complex, he has done a good job of convincing the reader that it is not needlessly so: there was indeed not a single event, “the Great Depression,” but a whole series of demand- and supply-side shocks that continually battered the economy.

Following his historical narrative, Sumner concludes his book with three chapters drawing broader conclusions from his history. In the first, he contemplates the impact of the Great Depression on twentieth-century macroeconomics. The most important conclusion he reaches here is that the Keynesian elevation of fiscal over monetary policy is based on a misreading of the ineffectiveness of the Fed’s open-market purchases in 1932, which ineffectiveness, he contends, was due to gold outflows, and not a “liquidity trap.”

In this chapter, he makes a strong argument for the importance of history for theory, claiming:
While working on this project I have gradually come to the conclusion that modern macroeconomics, macro history, and the history of thought are a seamless whole; it is impossible to really understand any one filed without also having deep knowledge of the other two. (357)
His next chapter analyzes the causes of the Great Depression. Although Sumner is a self-described ‘libertarian’ his policy conclusion stemming from his study is nonetheless pro-intervention:
If there is a root cause of the Great Depression, it lies somewhere in the painful birth of the modern world, the difficulty that societies had letting go of the “barbarous relic,” and moving to a more mature, and interventionist, monetary policy regime. (393).
Whether this represents admirable flexibility or culpable wishy-washiness, I leave it to the reader to decide.

Sumner also makes a very interesting observation of “the lessons of history” in this chapter:
This leads to another cautionary observation about the “lessons of history.” Either of the two policy counterfactuals [of tighter or looser monetary policy] for 1928-1929 might have led to a smaller 'Great Depression,' but in retrospect the undervaluation of gold made some sort of downturn almost inevitable. Had either alternative strategy been followed, and a modest depression resulted, that alternative would have almost certainly received historical censure. (397)
Sumner offers as an analogy the allied policy towards Germany after World War I. He argues, plausibly I think, that Allied policy was just about perfectly bad. If Allied peace terms had been more forgiving, Germany would have recovered better, and the appeal of the Nazis would have been blunted. But if the policy had been even harsher than it was, then, Nazis or no Nazis, Germany would have been too weak to remilitarize the Rhineland, threaten Czechoslovakia, and so on.

A final, minor gripe I have is that, for those of us not as versed as Sumner in the monetary institutions and history of the late international gold standard, it would have been useful to have a bit more explanation of the technical details of what was occurring. For example, he mentions “gold sterilization” several times, without ever describing what this consists of. The reader can, of course, look this up elsewhere, but since the book is not aimed exclusively at specialists, why not briefly describe the procedure?

In summary, Sumner has written an excellent and thought provoking book that should be read by all students of the Great Depression. His unusual focus on the gold market and the wealth of detail he presents on what actually occurred during the late 1920s and the 1930s certainly will repay careful study.


Popular posts from this blog

Libertarians, My Libertarians!

"Machine Learning"

Misunderstanding dynamical systems