A public choice theory of the great contraction

Conclusion

The conventional (Chicago) wisdom about the Great Contraction is that it was the result of a massive policy failure. According to this view, the collapse of the economy was brought about mostly by the fact that the monetary policymakers of the day were more concerned with internal power struggles than with supplying reserves to member banks — an explanation which, with apologies to Professor Stigler, must have elicited uproarious laughter at the Bankers Club.

It has been our purpose in this paper to suggest an alternative theory. Briefly stated, the restrictive monetary policy pursued by the Fed from 1929 to 1933 was motivated by the economic interests of member banks. The Fed's policy produced a massive differential failure rate between member and nonmember banks in which the latter were eliminated at a rate at least five times higher than the former. Fed member banks faced substantially less competition in the banking market after the ‘house-cleaning’ than before. Although all banks suffered net losses in the short run, in the long run member bank profitability was enhanced by the reduction in the number of nonmember competitors. At the same time, the Fed itself benefitted from a substantial increase in the proportion of the total banking system within its bureaucratic jurisdiction and from a change in its method of finance.

Putting the Great Contraction in interest-group terms, we suggested that the Fed was acting as the agent of congressmen serving on important oversight committees, who in turn were representing the interests of the member banks in their states. (Due to interstate branching restrictions, the state is the relevant geographic market within which banks compete.) We tested this hypothesis using bank failure rate data across states and found that, holding other things equal, the failure rates of nonmember banks in the early 1930s were significantly higher in states with representation on the House Banking and Currency Committee.

Of course, not even Friedman and Schwartz (1963: 299, 301) argue that Fed policy was the sole direct cause of the Great Contraction, but instead maintain that the Fed acted in such a way as to take a ‘relatively severe’ contraction and turn it into ‘by far the most severe business-cycle contraction ... in the whole of U.S. history.’ But to the extent that the Great Contraction was worsened by Fed policy, and this in turn led to the succession of events called the Great Depression, the latter was a side-effect of economically rational interests of Federal Reserve member banks.

To say the least, this would appear to have been an enormously inefficient method of obtaining the postulated benefits. Certainly, there were other, less costly methods available by which member banks could have reduced the number of their nonmember rivals.

This is perhaps the most difficult problem our theory of the Great Contraction confronts. We can offer two responses. First, although the literature on the Great Depression tends to focus on fluctuations in national macroeconomic aggregates (e.g., the money supply, unemployment, and so on), it is clear that the severity of the Depression differed radically across states. States with relatively large agricultural sectors tended to be especially hard hit, suffering the highest rates of unemployment and the most significant numbers of bank failures. By contrast, states in some other areas (most notably New England and the Southeast) seemed to be less seriously affected. We note that in some of our regressions, the number of nonmember bank failures was negatively and significantly related to state unemployment rates in 1934. This is certainly not conclusive evidence, but it does suggest a pattern of differential costs associated with the Great Contraction. That is, the costs of providing benefits to Federal Reserve member banks in states with representation on the relevant oversight committees may have fallen more heavily on the economies of states not so represented. If so, this would cast a very different light on the actual costs of these policies to the relevant political decision-makers.

Second, the potential dilemma in this case is no worse than that encountered in the application of the economic theory of politics to many other kinds of government behavior. Economists since Adam Smith have ascribed protectionist trade policies to the influence of special-interest groups, who benefit modestly at the cost of relatively huge economic inefficiencies. Many other government actions that might be plausibly explained from an interest-group perspective would similarly seem to involve costs to the economy which grossly outweigh the benefits special interests might reasonably expect. (Presumably, even in such seemingly extreme cases, marginal benefits equal marginal costs to the relevant actors.) Some writers have suggested that the political marketplace in modern democracies is characterized by much larger transactions costs than are normally encountered in ‘ordinary’ markets (see Demsetz, 1982); this might help explain the seemingly huge discrepancies we observe between apparent (concentrated) benefits and (diffused) costs in Depression-era monetary policy. In any event, this problem is not peculiar to our theory, but is shared by many applications of the interest-group model of government.

It is worth emphasizing that the events we have described do fall into the interest-group model's stylized depiction of concentrated benefits versus diffuse costs. Moreover, without having a complete picture of all of the options available to the monetary policymakers, we cannot be sure that any of the alternatives was in fact more efficient. We do know the path that was chosen and that, contrary to the conventional wisdom, this path had a rationally-motivated, interest-group basis.