Mason Gaffney asks how far the fall in AD after 1929 could be ascribed
to a contraction of demand deposits.

Lauchlin Currie, _The Supply and Control of Money in the United States_
(Harvard, 1934, Table 1) constructed the first series on the money
supply for the US and showed that demand deposits fell by 33.3 percent
between June 1929 and June 1933. Friedman and Schwartz (1963) put the
decline at 36 percent.

The money supply (cash plus DDs) contracted by about 25 percent in this
period after allowing for the big cash drain that occurred as confidence
in banks fell and the Fed adopted a basically passive stance, eschewing
open-market purchases sufficient to get the banks out of debt.

Mason suggests that the value of real estate collateral was inflated in
the late 1920s by previous deposit expansion. In fact demand deposits
increased only by 8.2 percent between June 1925 and June 1929 (Currie,
ibid.). Savings deposits increased somewhat faster over the same period
(by 22.2 percent; or about 5.5 percent a year), but savings exert a
deflationary not an inflationary influence.

The stock market peaked in October 1929. The Fed had been preoccupied
with asset values rather than consumer and wholesale prices (these were
stable or falling), thanks to the pernicious influence of the real-bills
doctrine, and squeezed the real economy even after its mid-1929 peak.

The demand for money as a proportion of GDP (hence measured income
velocity) was relatively stable, but was highly volatile as a
(declining) proportion of asset values. The latter, incidentally, were
not especially inflated relative to earnings, and the price-earnings
ratio was not as high as at several more recent periods of asset price
booms.

Roger Sandilands