Copyright Guilford Publications, Inc. Fall
2001
In our recent article on "Inflation and Accumulation" (Science &
Society, Vol. 64, No. 3), we claimed that the real price of financial
intermediation in Israel was positively correlated with inflation. This
claim is correct both theoretically and empirically, although the method
we used to illustrate it contains a mistake that needs to be rectified.
Briefly, our argument on pp. 289291 of the article was as follows.
Financial institutions are commonly viewed as intermediaries between
depositors and borrowers. From a neoclassical perspective, the real price
of their intermediation should be independent of the rate of inflation.
Since the principal of both deposits and loans is already indexed to
inflation, in order for the real price of intermediation to remain
unchanged all it takes is a fixed spread between nominal lending and
deposit rates. Strictly speaking, this last statement is wrong: the ratio
between nominal lending and deposit rates should indeed be independent of
inflation, but not their difference.
Symbolically, let P be the price index, LR the nominal lending rate for
the next period, DR the nominal deposit rate for the next period, and RPI
the real price of intermediation. For any period t:


In other words, the expost real price of intermediation is determined
by the nominal interest rate spread set in the previous period, divided by
the ratio between the current price index and the price index prevailing
in the previous period. As defined in the above expression, RPI, is
already corrected for, and should therefore be independent of, price
inflation. However, this is not what happened in Israel. As argued in the
article, and as illustrated in Figures 1 and 2 which correct for our
original misrepresentation, the real price of financial intermediation has
been indeed positively and tightly correlated with inflation (note the
logarithmic scale). Figure 1 shows the comovement of the two variables
based on annual data (now updated until 1999). During times of high
inflation, however, yearly data could conceal significant intrayear
redistribution. This potential bias is somewhat reduced in Figure 2, which
is based on monthly data since 1984, and which, again, corroborates our
basic claim. Overall it seems clear that inflation helped boost the real
price of financial intermediation when it rose, and undermined it when it
fell.1


[Footnote] 
1 This conclusion is of course as good as the
data on which it is based. Indeed, many crucial aspects of
"intermediation" during the inflationary period remain unknown.
These include, among others, the ways in which the banks manipulated
"value days" (on which they charged interest but did not pay it),
the constantly changing structure of interest rates on credit
windows and overdraft, the shifting pattern of commissions on
security trading, and the temporal handling of various government
taxes and levies. During the hyperinflation of the mid1980s, these
features of intermediation were probably more important than the
lendingdeposit spread. The banks, however, together with the
finance ministry and the central bank, remain silent on what
actually transpired. 
[Author Affiliation] 
JONATHAN NITZAN SHIMSON BICHLER

[Author Affiliation] 
Department of Political Science York University
4700 Keele St. Toronto, Ontario M3J 1P3 Canada 
nitzan@yorku. ca POB 4283 
Jerusalem. Israel 91042 tookie@inter. net. il

