Published online by Cambridge University Press: 11 May 2010
This paper focuses on changes in household balance sheets during the Great Depression as transmission mechanisms which were important in the decline of aggregate demand. Theories of consumer expenditure postulate a link between balance-sheet movements and aggregate demand, and applications of these theories indicate that balance-sheet effects can help explain the severity of this economic contraction. In analyzing the business cycle movements of this period, this paper's approach is Keynesian in character in that it emphasizes demand shifts in particular sectors of the economy; yet it has much in common with the monetarist approach in that it views events in financial markets as critical to our understanding of the Great Depression.
1 See Friedman, Milton and Schwartz, Anna J., A Monetary History of the United States (Princeton, 1963)Google Scholar.
2 See Bolch, Ben and Pilgrim, John, “A Reappraisal of Some Factors Associated with Fluctuations in the United States in the Interwar Period,” Southern Economic Journal, 39 (Jan. 1973), 327–44CrossRefGoogle Scholar; Gordon, Robert A., Economic Instability and American Economic Growth: The American Record (New York, 1974)Google Scholar; and Temin, Peter, Did Monetary Forces Cause the Great Depression? (New York, 1976)Google Scholar.
3 See George D. Green, “The Economic Impact of the Stock Market Boom and Crash of 1929,” in Federal Reserve Bank of Boston, Consumer Spending and Monetary Policy: The Linkages, Monetary Conference (June 1971), pp. 189–220; and Kirkwood, John B., “The Great Depression: A Structural Analysis,” Journal of Money, Credit and Banking, 4 (Nov. 1972), 811–37CrossRefGoogle Scholar.
4 Fisher, Both Irving, “The Debt-Deflation Theory of Great Depressions,” Econometrica, 1 (Oct. 1933), 337–57CrossRefGoogle Scholar, and Kindleberger, Charles P., The World in Depression 1929–39 (London, 918 1973)Google Scholar, promote the view that the deflation did affect aggregate demand and that it is one explanation for the seriousness of the contraction.
5 Ando, Albert and Modigliani, Franco, “The ‘Life-Cycle’ Hypothesis of Saving: Aggregate Implications and Tests,” American Economic Review, 53 (Mar. 1963), 55–84Google Scholar, and Mishkin, Frederic S., “Illiquidity, Consumer Durable Expenditure, and Monetary Policy,” American Economic Review, 66 (Sept. 1976), 642–54Google Scholar.
6 Friedman and Schwartz, A Monetary History.
7 The financial asset measure is a gross measure; it does not net out household liabilities. It includes currency plus demand deposits, time and savings deposits, corporate and government bonds and notes, corporate equity, life insurance, and other miscellaneous assets.
8 Models of the demand for household liabilities either are set in a nominal framework [Michael K. Evans and Avram Kisselgoff, “Demand for Consumer Installment Credit and Its Effects on Consumption,” in Duesenberry, James, Fromm, Gary, Klein, Lawrence and Kuh, Edwin, eds., The Brookings Model: Some Further Results (New York, 1969)]Google Scholar or have an adjustment mechanism that views the change in liabilities as a function of real variables [Mishkin, Frederic S., “Household Liabilities and the Generalized Stock-Adjustment Model,” Review of Economics and Statistics, 58 (Nov. 1976), 481–85CrossRefGoogle Scholar and Motley, Brian, “Household Demand for Assets: A Model of Short-Run Adjustments,” Review of Economics and Statistics, 52, (Aug. 1970), 236–41]CrossRefGoogle Scholar. These models have the characteristics that a severe deflation will cause an increase in the real value of household liabilities in the short run, even if the consumers' desired levels of liabilities in real terms have fallen. The desired drop in real liabilities thus may not occur for several years, as was the case during the Great Depression.
9 The impact of household balance-sheet shifts on aggregate demand during the 1973–1975 recession is analyzed in Frederic S. Mishkin, “What Depressed the Consumer? The Household Balance-Sheet in the 1973–75 Recession,” Brookings Papers on Economic Activity, 1977:1, pp. 123–64. Household balance-sheet changes during this period do appear to have been an important factor promoting the severity of the 1973–75 recession, as seems to have been the case during the Great Depression.
10 The rental value of the stock of consumer durables is only a small fraction of consumption. A useful approximation for consumption is the expenditures on non-durable goods and services. The discussion that follows in the next section of the paper will use this definition.
11 Franco Modigliani, “Monetary Policy and Consumption,” in Federal Reserve Bank of Boston, Consumer Spending and Monetary Policy: The Linkages, Monetary Conference (June 1971), pp. 9–84. In this article, Modigliani also cites references which lend further support to the life-cycle hypothesis.
12 The life-cycle models estimated in the current version of the MPS model (see MITPENN-SSRC, Quarterly Econometric Model Equations, January 1975, mimeo) and in Mishkin, “What Depressed. …,” use different sample periods, different definitions of the wealth variable, and, in Mishkin, a different estimation technique. The overall wealth effect is still highly significant and does not deviate more than 10 percent from the .053 value.
13 See Barry Bosworth, “The Stock Market and the Economy,” Brookings Papers on Economic Activity, 1975:2, pp. 257–90 and Mishkin, “What Depressed. …”
14 An important point here is that an increase in indebtedness matched by an increase in holdings of tangible assets that leaves net wealth constant would still lead to a future decline in consumer expenditure. A decrease in the value of financial asset holdings matched by an increase in the value of tangible assets that left net wealth unchanged would also lead to a decline in consumer demand. Hence, the fact that net wealth is unchanged is no guarantee that changes in the composition of the household balance sheet have no effect on the expenditure behavior of households.
15 See Mishkin, “Illiquidity, Consumer Durable Expenditure, …,” for the specifics of the formal analysis from which this discussion is derived. The liquidity-hypothesis model also allows for perceptions of risk-influencing expenditure behavior. This aspect of the hypothesis is not discussed here as it is not crucial to the analysis of this paper.
16 Mishkin, “Illiquidity, Consumer Durable Expenditure,…”; Mishkin, Frederic S., Illiquidity, the Demand for Consumer Durables and Monetary Policy, Ph.D. Thesis, Massachusetts Institute of Technology, 1976Google Scholar, published as Research Report 61, Federal Reserve Bank of Boston, Boston, 1977; and Kearl, J. R. and Mishkin, Frederic S., “Illiquidity, the Demand for Residential Housing, and Monetary Policy,” Journal of Finance, 32 (Dec. 1977), 1571–86CrossRefGoogle Scholar.
17 The user capital cost measure equals the interest rate plus depreciation, multiplied by the relative price of consumer durables or housing.
18 Kearl and Mishkin's expenditure model uses houshold liabilities excluding mortgage debt as the DEBT variable because mortgage debt is not expected to influence the consumer's decision to purchase new housing. See Kearl and Mishkin, “Illiquidity, the Demand for Residential Housing…,” for a discussion of this point.
19 Kearl and Mishkin, “Illiquidity, the Demand for Residential Housing…” and Mishkin, “Illiquidity…,” contain more detailed discussion of the estimated models above. The consumer expenditure model given here does not allow for a distributed lag on stock market asset holdings in the determination of the desired consumer durables stock, a feature of some of the estimated models in Mishkin, “Illiquidity.…” The expenditure model given here is more useful than models with these lags for the calculations in the next section, as it has less complicated dynamic short-run balance-sheet effects.
20 Kearl and Mishkin, “Illiquidity, the Demand for Residential Housing…,” indicate that liquidity considerations might be important to consumers' desires for leased assets, such as multi-family housing, and find significant balance-sheet coefficients in multi-family housing equations. These coefficients, however, are not that robust when credit rationing variables are added as explanatory variables, and post-sample tracking of a multi-family housing equation with balance-sheet effects in Mishkin, “What Depressed…,” casts doubt on the existence of these effects for multi-family housing. Therefore, the multi-family housing equation with balance-sheet effects is not used for the analysis of the following section and is not reported on here.
21 The theory behind the liquidity hypothesis indicates that any consumer obligation that requires a commitment of payments in the future, whether or not it is classified as a household liability, will prove to be a deterrent to tangible asset expenditure. Consumer obligations excluded from the household liabilities measure—lease payments, rent, contractual saving, insurance payments, education expenditures, etc.—should be highly correlated with the debt measure, since the same factors affecting the willingness to enter into these obligations should affect the willingness to incur debt. Hence, the debt variable might well proxy for these other consumer obligations. Therefore, the debt coefficient would reflect their influence as well as the influence from measured household liabilities, and it probably would overstate the actual expenditure impact from measured household liabilities.
22 See Mishkin, “What Depressed …,” and the references in footnote 16.
23 Analyzing the impact of household balance-sheet changes on the business cycle by using the desirable methodology of dynamic simulation experiments with a macroeconomic model of the U.S. economy is possible with postwar, quarterly data. Mishkin, “What Depressed…,” has carried out this exercise.
24 On the other hand, the liquidity models' short-run effects given above, which were estimated on quarterly data, will overstate the direct impact of balance-sheet changes on annual expenditure. These models are flexible-accelerator models in which changes in independent variables have.their largest impact in the initial period. Since this period is a quarter rather than a year in the estimated models, the effect of debt and financial asset changes on expenditures over a year would be less than the estimated short-run quarterly effects given above.
25 Clearly, at least one group of consumers, the farm population, suffered severe financial distress during the Great Depression. J. Munger, A Preliminary Study of Farmer Bankruptcy Experiences in the Dakotas 1928–1952, Agricultural Economics Pamphlet (Mar. 1955), Agricultural Economics Department, South Dakota State College, estimates that 29 percent of all farms in the United States underwent forced sales (foreclosures) in the decade 1930–1939.
26 See Bolch, Ben, Fels, Rendigs and McMahon, Marshall, “Housing Surplus in the 1920's?” Explorations in Economic History, 8 (Spring 1971), 259–83CrossRefGoogle Scholar; Bolch and Pilgrim, “A Reappraisal …”; and Hickman, Bert G., “What Became of the Building Cycle?” in David, Paul and Reder, Melvin, eds., Nations and Households in Economic Growth: Essays in Honor of Moses Abramovitz (New York, 1973)Google Scholar.
27 See R. A. Gordon, Economic Instability, for example.
29 See Temin, Did Monetary Forces.…
30 The work of Brown, E. Cary, “Fiscal Policies in the Thirties: A Reappraisal,” American Economic Review, 46 (Dec. 1956), 857–79Google Scholar, indicates that government fiscal policy was not a primary factor in the collapse of aggregate demand during the Great Depression period, and thus fiscal policy is not cited as an important cause of the Depression.
31 Friedman and Schwartz, A Monetary History.
32 Friedman and Schwartz, A Monetary History, mention the large decline in velocity in 1930, but are unclear as to its cause. There is also a problem of causality in the Friedman and Schwartz argument. See Temin, Did Monetary Forces …, for a discussion of this point.
33 See Burck, Gilbert and Silberman, Charles, “What Caused the Great Depression?” Fortune, 51 (Feb. 1955), 94–99, 204, 206, and 209–11Google Scholar.
34 Kirkwood, “The Great Depression …,” pursues an econometric analysis of the Depression that attributes the decline in aggregate demand to the stock market crash, yet his explanation for the stock market impact is essentially that discussed above.
35 See Temin, Did Monetary Forces. …
36 Galbraith, J. K., The Great Crash, 1929 (London, 1955)Google Scholar is one exception. For a typical reaction of economists to the stock market view of the Depression, see Green, “The Economic Impact …,” and the comments on his article. The discussion so far also has ignored the view that economic events outside the United States were important in causing the Depression. See Kindleberger, The World in Depression, for example. Although it is entirely possible and probable that the collapse of the European economies increased the severity of the downturn in the United States, causality seemed to flow from the U.S. economic collapse to Europe and not the other way around.
37 Fisher, “The Debt-Deflation Theory.” In addition, see Clark, E., ed., The Internal Debts of the United States (New York, 1933)Google Scholar and Minsky, H. P., “Longer Waves in Financial Relations: Financial Factors in the More Severe Depressions,” American Economic Review, 54 (May 1964), 324–35Google Scholar.
38 The 1930 contraction is somewhat different from the 1938 contraction in that demand effects from changes in household liabilities were a larger proportion of balance-sheet effects in 1930 (65 percent) than in 1938 (25 percent). Thus, stock market effects appear to have been more important in 1938 than at the beginning of the Depression. Fluctuations in the valuation of common stocks also might affect aggregate demand by influencing the household's desired level of liabilities. In 1929 the stock market boom might have made households more willing to accumulate debt because of their higher perception of net worth. In this sense, the stock market speculative boom could have set the stage for the Depression.
39 Many economists (Green, “The Economic Impact…,” for example) object to an important stock market role in business cycle fluctuations during the Great Depression because stock market assets were highly concentrated in the hands of a small segment of the U.S. population. The highly skewed distribution of stock market assets does not invalidate the argument contained here which cites the importance of stock market effects. The calculations of balance-sheet effects found in Table 4 use coefficient estimates that already reflect the skewed distribution of stock market wealth. In addition, some recent cross-section work by Friend, Irwin and Lieberman, Charles, “Short-Run Asset Effects on Household Saving and Consumption: The Cross-Section Evidence,” American Economic Review, 65 (September 1975) 624–33Google Scholar, indicates that the skewed distribution of stock market holdings would not greatly diminish the aggregate stock market impact, as has often been thought to be the case. The analysis of the stock market role in the Great Depression presented here is quite different from the expectations view of stock market effects. The liquidity and life-cycle hypotheses imply that even if the consumer's mood did not change as a result of the stock market crash, he still would have cut back on his expenditures because of the deterioration of his balance sheet. Certainly, to the extent that pessimism was bred by the stock market decline, the economy was affected. The approach here also differs from other discussions ofstock market effects on the economy, such as James Tobin, “Monetary Policy in 1974 and Beyond,” Brookings Papers on Economic Activity, 1974:1, 219–32, in that it focuses on the consumer rather than on the firm.
40 See: Tobin, James, “A General Equilibrium Approach to Monetary Theory,” Journal of Money, Credit and Banking, 1 (Feb. 1969), 15–29CrossRefGoogle Scholar; Foley, D. and Sidrauski, M., Monetary and Fiscal Policy in a Growing Economy (New York, 1971)Google Scholar; Modigliani, “Monetary Policy …”; Sprinkel, Beryl, Money and Stock Prices (Homewood, Ill., 1964)Google Scholar; and the references in Cooper, Richard V. L., “Efficient Capital Markets and the Quantity Theory of Money,” Journal of Finance, 29 June 1974), 887–908CrossRefGoogle Scholar.
41 Tobin, “A General Equilibrium Approach …,” and Modigliani, “Monetary Policy. …”
42 The balance-sheet approach found here does not discount important effects on aggregate demand from other sources. Factors such as the surprising drop in non-durable consumption, demographic effects on residential housing markets, international instability, and possible effects on business investment, could all have contributed to the Great Depression collapse.
45 Albert Ando and E. Cary Brown, “Lags in Fiscal Policy,” in Studies for the Commission on Money and Credit, Stabilization Policies (Englewood Cliffs, N. J., 1963) 97–163Google Scholar.
Full text views reflects PDF downloads, PDFs sent to Google Drive, Dropbox and Kindle and HTML full text views.