The primary propagating mechanism in the American Depression identified by Friedman and Schwartz in their classic Monetary History of the United States revolved around banking panics. They identified the first of three banking crises in December 1930 with the failure of the Bank of United States. Had the banks responded to panic by restricting payments (a nineteenth-century practice), Friedman and Schwartz claimed, the Depression need never have happened. They argued that restriction in 1893 and 1907 had quickly ended bank suspensions and promoted economic recovery.
The events after the restriction of payments in 1893 and 1907 show that the American economy of the time was very stable. A restriction of payments is defined as a refusal on the part of banks to honor their commitment to exchange deposits for currency at par. When a single bank refused to redeem its obligations at par, it was legally bankrupt. But when banks acted in concert, there was an effective devaluation of deposits against currency.
The price of deposits was determined, like all prices, by the forces of supply and demand. People who were afraid that the price of deposits would decline wanted to sell, driving down the price. People who thought that the price of deposits had already fallen and was due to rise back toward par wanted to buy, driving up the price. The market price was where the supply from the former group just matched the demand from the latter. The currency premium in 1893 and 1907 was never more than 4 percent; it had fallen to almost nothing in a month, even though full resumption came somewhat later. Most people, in other words, expected the banks to resume payments at par speedily. They did not anticipate a major depression or further bank crises. They did not rush to sell discounted deposits.
Friedman and Schwartz therefore adopted an inconsistent position toward the banking crisis of 1930. On the one hand, they said that the economy was unstable, that a small event set off the Great Depression. In fact they traced the cause of the Depression back to the death of Benjamin Strong in 1928, even though their main story starts with the banking crisis in 1930. On the other hand, they implied that the economy was very stable, that a restriction of payments would have resulted in only a tiny change in the price of deposits - like the 2 percent or 3 percent seen in 1893 and 1907 — and that this change would have brought the economy back onto an even keel. They cannot have it both ways. Either there was an impulse more powerful than the death of the head of the New York Fed or the economy was far less stable in 1930 than in 1893 and 1907 (and a suspension of bank payments would have had only limited impact). As noted above, the former position is taken here.
Friedman and Schwartz argued that the banking failures in December 1930 reduced the supply of money by increasing the banks’ demand for reserves and the public’s demand for currency. This in turn depressed spending. If it happened this way the monetary restriction should have affected income through the financial markets. Even if the progress of the Depression eventually led to lowered demand for money and low interest rates, we still should observe a rise in interest rates at the time of the banking crisis — before any effects of the banking failures had run their course. No such credit stringency is observed at the start of 1931.
There was an increase in bank failures in November and December of 1930. But much of the rise of liabilities in failed banks was due to the failure of just two banks. Caldwell and Company failed in Tennessee, and the Bank of United States failed in New York City. Both of these banks had undergone reckless expansion in the late 1920s, and their overblown empires collapsed under the pressure of the emerging Depression.
If the liabilities of these two banks are subtracted from the total liabilities in failed banks in those months, it emerges that the rise in other bank failures was clearly noticeable but not of the same scale as the rise of bank failures in the summer and fall of 1931. The level of bank failures also returned to its earlier level at the end of 1930, where it stayed for four months. There was no reaction in the markets for short-term credit, aside from a temporary rise in rates in Tennessee. There was no fall in the stock of money at the end of 1930. There was no shock to the quantity of money that could have produced a large macroeconomic effect. There was no direct effect of the “first banking crisis.”
Instead, there was the beginning of a movement to increase currency in the hands of the public. This movement was small relative to the other events of the time. The change in the rate of growth of the money supply from the 1930 “banking crisis,” therefore, was swamped by changes from other causes. As a result there was no reason to expect interest rates to react to such a change.
Alternative mechanisms have been proposed for the effects of banking crises. The most popular recent view, due to Ben Bernanke, argues that the effect of banking panics operated through credit rationing. Credit became harder to get for many borrowing firms, which had to shop around for loans or do without. Published interest rates did not reflect this added cost because they were the cost of loans granted, not loans refused.
Any lender had imperfect knowledge of the comparative risks of different firms. Banks specialized in making the best use of the available data. They acquired most of the loan business because they were the low-cost intermediaries. When banks failed, they no longer could extend credit, and other banks switched to more liquid loans to protect themselves. This reduced the supply of the most efficient intermediation services and raised their cost and consequently the cost of loans to borrowers.
This hypothesis typically is tested by time-series regressions explaining the movements of industrial production. A more direct test examines the progress of different industries. Bernanke noted explicitly that the rising cost of credit intermediation hurt households and small firms much more than large firms. Bank failures then should have hurt industries populated by family firms and other small businesses more than those composed of large, well-established firms.
But the presence of large firms is positively related to the fall in production, not negatively as the credit rationing hypothesis predicts. Comparison with 1937—38 reveals that the cross-sectional pattern of industrial decline in the Great Depression was not unusual. Despite the banking crises, the pattern of industrial decline — as opposed to its magnitude and duration — was unexceptional. There is no evidence that the pattern of industrial decline was rendered unusual by the dramatic collapse of the banking system.
We need to take care here not to throw the baby out with the bath water. The American financial system was being battered at the end of the 1920s by the stock market decline, business failures, bank failures, and international events. After the stock market crash, firms shifted their new offerings from stocks to bonds. Net new stock offerings fell by $2.5 billion from 1929 to 1930, while net new bond offerings rose by $1.4 billion. The price of lower-grade industrial bonds then began to decline in late 1930. The increased supply of bonds lowered their price. Business and bank failures decreased the demand for bonds by increasing their perceived risk.
A gap opened up between the cost of bank loans to firms that could borrow at the prime rate (falling steadily in 1929 and 1930) and the cost of industrial bonds for smaller firms. This is the kind of premium that Bernanke was talking about, although market prices reflected this premium rather well. The spread between the prime rate and other interest rates is a good indicator of monetary pressure even without bank failures. In addition, since bonds were being reclassified to show their increased risk at this time, the return on risky bonds was rising for two reasons: bonds of a given riskiness were worth less, and any given bond was becoming more risky. The largest firms had access to credit at costs far lower than smaller firms. The cross-sectional pattern of industrial decline shows, however, that access to credit did not determine which industries declined.
Bank failures undoubtedly accentuated the Depression. International comparisons of countries with and without banking difficulties suggests that banking difficulties in general were harmful. But the mechanism by which bank failures had their effects is not clear. As a result, their importance in the American contraction is still a matter of dispute.
At about the same time as the stock market crash, the prices of raw materials and agricultural goods - which had already been tending slowly downward — began to fall precipitously. Charles Kindleberger identified the fall in commodity prices as one of the primary channels through which deflation spread, from “stock prices to commodity prices to the reduced value of imports.” Although a change in prices only reallocates income, he argued, the effect is asymmetric. The losers found their budgets curtailed and were forced to cut spending; the winners did not correspondingly increase theirs.
The prices of agricultural products and raw materials had been falling in the 1920s as a result of the overexpansion of production during and after the First World War. Various attempts to prop them up through tariffs or purchases had proved ineffective. Inventories accumulated as the production of many raw materials exceeded demand at the market price. The costs of holding these stocks and conducting orderly marketing rose as credit conditions were tightened at the end of the 1920s. In the credit squeeze that always came to the United States in the fall, many owners of these inventories failed in 1929. Further price declines were of course in store as the demand for raw materials contracted.
The effects of the price declines on different groups need to be distinguished. For countries whose agricultural or mineral products were the main source of foreign currency, the fall in price was a disaster. Devaluations were the frequent response. But for importing countries the decline in product prices was a plus. Even if Kindleberger is right and the price decline did not cause spending to rise, it allowed greater monetary ease. (It reduced any inflationary pressure, and it increased the real money supply.) The United States experienced both effects. Farmers suffered, while the rest of the economy gained. The net effect of the initial fall in commodity prices in the United States therefore probably was positive, since there were many more consumers than producers of these commodities in the United States.
The gain was limited, however, as prices in general began to decline in 1930. The more pervasive deflation cannot be attributed to the breakdown of cartels, and it was not closely correlated with the stock market. It was a reflection of the falling aggregate demand that came from the preceding credit stringency. Both the stock market crash and the collapse of taw materials prices were part of the propagating mechanism by which this tightness affected economic activity, but they were only part of a complex picture.
Finally, a recent paper provides a new explanation for the dramatic fall in consumption in 1930. Martha Olney argues that the structure of consumer credit made consumption highly volatile at this moment in history. If a consumer defaulted on an automobile loan, to take the most important form of consumer credit, he or she did not retain any equity in the automobile used as security. Consumers therefore cut back their consumption in an effort to retain their equity in their new cars as their incomes fell in the recession of 1929. A dramatic fall in consumption from 1929 to 1930 was the result.
World History
