We set out initially to reconstruct each of the four banking crises of the Great Depression. That task is now completed. What remains to be done is to extract from that account what we have learned, if anything, about the banking crises of the Great Depression. Did these crises differ from pre–1914 banking panics? And, if they did, how did they differ? Did panic–induced failures in the early 1930s differ from the nonpanic–induced suspensions in the twenties? Did panic–induced failures cause output to vary? And, finally, what was the response of the Federal Reserve to accelerated bank closings between 1930 and 1933? We have the knowledge and experience of four separate banking crises with which to attempt to answer these important questions.

Our description and analysis of the first three crises revealed that decomposition by Federal Reserve District disclosed wide geographical disparities in the incidence of bank suspensions and hoarding. So much so that we concluded that the 1930 banking crisis and the April–August 1931 crisis were region specific. These two crises were not nationwide in scope, and the expenditure effects associated with the 1930 crisis were minimal.

Unlike pre–1914 panics, the “eye” of the banking disturbance was not the New York money market. The establishment of the Federal Reserve System in 1914 had altered the response of the New York money market in at least two important ways: the short–term interest rate response was muted, and strong stock market reactions were eliminated. There was in contemporary economic jargon “a change in regime” which had the effect of altering permanently expectations about how the central money market would react to unanticipated financial shocks.