Beruflich Dokumente
Kultur Dokumente
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to Journal
of Political Economy.
http://www.jstor.org
On Passive Money
Julio H. G. Olivera
Universityof BuenosAires
I
There are two ways in which the quantity of money may be included in a
general equilibrium system: as a datum, whose value is fixed exogenously,
or as an unknown, whose value is determined by the equilibrium system
itself. Let them be called active money and passive money, respectively.
Current monetary theory uses the first approach. The Ricardo-WalrasKeynes tradition in general equilibrium models is clearly of the activemoney variety. This is also true of the more contemporary versions, such
as Patinkin's well-known system (1965).' Apart from the main stream of
thought, however, a number of analyses exist which imply passive money.
The gold-standard model, as it is usually expounded, belongs to this
class. The price of gold in terms of money is given, while the stock of
money is endogenously determined by the set of equilibrium conditions.
The form in which the quantity of money adapts to its equilibrium level
differs according to the institutional mechanism: Convergence is assured
through monetization and demonetization of gold in the pure goldcurrency system, and through convertibility of notes into gold and vice
versa in the gold-standard system with paper money. In either case, the
money supply is allowed to conform to its equilibrium volume. This kind
of monetary system, in which the money price of a certain product or
commodity is a datum and the quantity of money a dependent magnitude,
may be denoted in general as a commodity standard.
I am indebted to Professors Martin J. Bailey, Geoffrey Maynard, Don Patinkin,
and Henry C. Wallich. and to my colleague Rolf Mantel for interestingcomments.
' Gurley and Shaw, whose analysis also belongs to this class, define expressively
what I call active money: "The nominal stock of money is an adventitious datum
cast into the structureof the model" (1960, p. 35).
805
8o6
JOURNAL OF POLITICALECONOMY
807
PASSIVE MONEY
Active-moneymodelsof generalequilibriumhavereceivedconsiderable
attention.Theirdeterminateness,
comparativestatic behavior,and stability propertieshave been exploredat length. In markedcontrast,passivemoney models of general equilibriumhave not even been formulated
explicitly.We are, at best, only vaguely awareof the bearingand latent
implicationsof passivemoneyfor the economicsystem.If the hypothesis
of passive money is to be properlyassessed, such implicationsshould
be carefullybroughtout. There is no ground to overlook or minimize
their significance.Consequently,the ensuingsectionswill be devoted to
a study of the passive-moneyequilibrium,includingan attemptto comparethe essentialpoints with the doctrineof activemoney.
II
(1)
E(s/p, M/p, i) = 0,
(2)
D(M/p, i) = 0,
(3)
whereT, E, and D mean excess demandin the marketsfor labor, commodities,and money, respectively;s denotes the money-wagerate, i the
interest rate, p the money price of commodities,and M the nominal
quantityof money. Since this systemis a widelyknown analyticaldevice
(Patinkin1965),5 let us pass directlyto the applicationswhichconcernus
here.
4 This is the usual assumption in short-period monetary models; see, however,
Mundell (1965).
5 The model above, however, is given solely in terms of the net demand functions,
whereas Patinkin specifies the properties of the supply and demand functions separately. Thus the model with which we propose to deal is actually more general than
Patinkin's.
8o8
JOURNAL OF POLITICALECONOMY
The case of active money stands out quite obviously from the basic
structure: Given M, the group of equations determines s,p, and i. The labor
standard is derived from the same framework by interchanging the roles of
s and M, so that s becomes a datum and M an unknown. Therefore, given
the value of s, the three equations determinethe levels of M, p, and i. If p,
instead of s, is fixed from outside, we get the commodity standard, with the
system telling the values of M, s, and i. Each of these models-active
money, the labor standard, and the commodity standard-is thus consistent with equilibrium.
This condition is easily seen to fail under the credit-standard model.
Here it is the money price of bonds, and consequently the rate of interest,
which acts as a datum, the unknowns being p, s, and M. But since the
latter come into the equilibrium system solely through their ratios, the
model so generated is inconsistent in general. It contains three independent
relations, equations (1)-(3), to determine two unknowns (Mip and sip).
In the light of this, it is hardly surprising that the credit standard had
always been enveloped by a margin of ambiguity. Tooke and Wicksell,
for example, drew mutually contradictory inferences from much the same
theoretical axioms.
To obtain determinateness, it is necessary to relax any one of the
equilibriumconditions. There are two main possibilities from the economic
standpoint. One is to drop the full-employment condition (which is precisely what Tooke did), which leads to a quasiequilibrium system of
Keynesian style. The other is to transform the money-market equation
into an identity, essentially in the vein of Say's Identity (this, it seems clear,
was Wicksell's procedure).6 Either way, the model ceases to be overdetermined. However only the real system becomes determinate-the
equilibrium values of Mlp and s/p are compatible with an infinite number
of monetary equilibria.
mI
Space forbids a closer examination of these issues, but what has been
said is perhaps sufficient to bring out the central points. Let us now direct
our attention to the effects of contingencies tending to produce change.
Our frame of reference will again be the equilibrium system, equations
(1)-(3), with the appropriate monetary specification. We shall utilize the
method of comparative theory which has been long familiar in analytical
economics. A small variation in one parameter will be supposed to come
about. The corresponding total derivatives of the excess demand functions
must all be zero, for we are relating two equilibrium states with one another.
6 Wicksell's theory implies that the money market keeps continually balanced
(Keynes 1958, p. 198).
809
PASSIVEMONEY
(4)
E, < ? .
(5)
-Da = 0
[ds/da<
01
dp/da <01.
di/da <
(6)
The results carry the opposite sign in the case of a variation of the propensity to invest.
lB. Change in liquiditypreference.-It will be symbolized by a corresponding change of the parameter A. The shift vector is taken to be
TA = ?
Ex = ?
[DE > ]
(7)
8io
JOURNAL OF POLITICALECONOMY
<0
Eds/dA
dp/dA <01.
(8)
di/dA> o
IC. Change in the money supply.-A glance at the equilibrium system
suffices to show that, given a variation of M, equilibrium is reestablished
through an equiproportional change of p and s, with i remaining constant.
This is the well-known derivation of the Quantity Theory as a proposition
in comparative statics.
So much for the case of active money. Let us now focus on the behavior
of the labor standard under the same hypothetical changes.
2A. Change in the propensity to save.-The
response vector is found to be
dpld =
dMIda> 0
(9)
di/da < 0
With a reversal of signs, the foregoing vector is applicable to any variation
of the propensity to invest.
2B. Changein liquiditypreference.-The shift vector is (7). The response
vector is now
dp/d = O
dM/dA
(10)
>0]
di/dA> o
2C. Change in s.-Obviously, equilibrium is restored through an equiproportional adjustment of p and M, with i unaltered. We are thus led
to a kind of converse statement of the Quantity Theory.
Finally, let us envisage the effects of the same types of disturbance in a
commodity-standard system.
3A. Change in the propensity to save.-The shift vector is again (5). The
response vector is
ds/da = 01
dM/da >
di/da < 0
(11)
8ii
PASSIVEMONEY
As in the preceding cases, the negative of this vector describes the response
to a shift in the propensity to invest.
3B. Change in liquiditypreference.-With (7) as shift vector, the effect
on equilibrium is
E
ds/dA =
0]
dM/dA > 0
di/dA >
(12)
oJ
IV
These results are far from anomalous, but still their scope would be narrow
if equilibrium were not stable. Let us therefore turn to the dynamics of the
subject. We shall concentrate attention on the so-called asymptotic stability
the tendency of a perturbed system to return, over time, to the equilibrium state. Only small perturbations will be considered. Proceeding in
the usual fashion, we shall approximate the dynamic system through
adjustment laws which are linear in the deviations from equilibrium.
Our knowledge about the respective coefficients will be limited to their
signs. If such information suffices to indicate stability of the dynamic
process, we shall say that the system is qualitatively stable.8 If stability
cannot be determined without specifying the numerical values of the
parameters, we shall say that the system is qualitatively nonstable. If the
data about the signs preclude the stability of the system, we shall refer to
it as qualitatively unstable.
A word now on the criterion of stability. Since we must operate with
characteristic equations of the third degree, a convenient shortcut to the
investigation of stability is provided by one of the rules of Lienard and
Chipart (Gantmacher 1960, p. 220). Namely, given the characteristic
polynomial X3 + a x2 + a2x + a3, a necessary and sufficient condition
that all its roots have negative real parts is
a,,
a3
(13)
7 The symmetry, however, is only mechanical; the spirit of the genuine Quantity
Theory involves active money (Friedman 1956, esp. pp. 16-17).
8
Sign stable, in the terminology of Quirk and Ruppert (1965).
JOURNAL OF POLITICALECONOMY
8I2
Having set forth these premises, let us briefly sketch their application
to the matter in hand. In the active-money case, the adjustment laws are
normally
hl > 0,
ds/dt = hiTL(slp),
dp/dt = jiEL(s/p, Mo/p, i),
jl > 0,
k, > 0.
(14)
The subscript o signifies that the variable is regarded as exogenous, and the
superscriptL denotes that the function is approximated by the linear part
of its Taylor series in the neighborhood of equilibrium. This dynamic
system fulfills all the conditions in (13) and is, therefore, qualitatively
stable.9
Let us now take up the hypothesis of passive money. Under the commodity standard, the adaptability of the money supply must replace the
flexibility of the price level. The resulting dynamic system is
dsldt = h2TL(sIpO),
h2 > 0,
(15)
0,
k2 > 0,
<
i2
h3 < 0,
(16)
13 > 0,
k3 > 0.
The test for stability exhibits a different pattern in this case. Although a,
and a3 are again positive, the sign of a1a2 - a3 depends on the parameter
values. The system is hence qualitatively nonstable.
This raises the question of how the coefficient of monetary adjustment
affects the stability of equilibrium. Is a high or a low speed of response
of the money supply more likely to promote stability? The result ultimately
turns on the sign of the expression
S-h j3T'
-h3
(k3DMIpEjEi
2M
--i3Es/l
MEs
p2)1 '
(17)
which is the uncertain term in a1a2 - a3, the other terms being necessarily
positive. It is a cumbersome formula, but its import is clear. If the term
(17) is positive, the system is stable, whatever the absolute magnitude of
h3. But if (17) is negative, the chance of stability is greater the smaller h3 is
in absolute value. Hence, on balance, a slowly responsive money supply
8 Incidentally, this proves in a formal manner Patinkin's thesis in chap. 13, secs. 2
and 3 (Patinkin 1965, pp. 316-28).
813
PASSIVEMONEY
Our topic is not one to be dealt with exhaustively in a few pages, and it
would not be difficult to amplify or qualify the foregoing argument. But
we hope to have carried the analysis far enough to permit some broad
conclusions. The passive-money economy we have examined is quite regular in behavior, from both the static and the dynamic standpoints, under
the regime of the commodity standard. It is much less "well behaved"
under the labor standard, for although it yields definite comparative
static results, it offers no assurance of stability. And it proves anarchic
under the credit standard-overdetermined in real terms and indeterminate in monetary values.
Parallels of this kind can be drawn only with reference to a specified
equilibrium structure. Needless to say, alternative models may lead to
conclusions which differ from the ones just summarized. We cannot even
exclude the possibility that in other analytical settings the passive-money
economy may behave more regularly, in the above sense, than the activemoney system. So we must be careful not to overassert the particular
results established in the preceding sections. The chief point is rather that
there exists no necessary congruity, either statical or dynamical, between
the active-money type of model and its passive-money "duals." This fact
must be emphasized, more than any special instance of it.
We must not conclude without saying a word about economic growth.
Some of its consequences on passive money, and vice versa, have been
attended to by the theories of inflation cited at the outset of this paper.
But the problem should be subjected to a closer dissection with the
apparatus of modern growth theory. The recent attempts at integrating
the doctrines of money and growth are based on active-money premises,
since they consider the rate of increase in the money supply as given
exogenously. It would be worth while to undertake a similar integration
from the passive-money point of view. In neither case, however, can we
expect to be able to predict the directions of change of the variables with
the aid of qualitative information alone.10
One more matter may be mentioned before closing these remarks.
Active money is not always separated by a sharp line of demarcation from
passive money. Very often they merge together in actual fact. To the extent
10 That is an additional reason in favor of Harberger'sadvice (and practice) of
bringingthe "inflation and growth" controversydown to rigorous empirical analysis
(Harberger1963, 1964).
814
JOURNAL OF POLITICALECONOMY
that money prices are downward inflexible, the historically given prices fix
a minimum to the monetary stock. Below this lower bound there is no
general equilibrium solution. But above the passive minimum, insofar as
prices are upward flexible, money becomes active. Price increases may well
happen to be structural-the monetary floor may rise because of autonomous factors uplifting them-but there is room for monetary inflation
above the floor. It would thus seem that a mixed model, in which active
and passive money are amalgamated, would come nearest to the real
course of events.
References
Archibald, G. C. "The Qualitative Content of Maximizing Models." J.P.E.
73 (February 1965): 27-36.
Friedman, Milton. "The Quantity Theory of Money-A Restatement." In
Studies in the Quantity Theory of Money, edited by Milton Friedman.
Chicago: Univ. Chicago Press, 1956.
Gantmacher, F. R. The Theory of Matrices. Translated by K. A. Hirsch.
Vol. 2. New York: Chelsea Publishing Co., 1960.
Gurley, John G., and Shaw, Edward S. Money in a Theory of Finance. Washington: Brookings Institution, 1960.
Harberger, Arnold C. "Dynamics of Inflation in Chile." In Measurement in
Economics: Studies in Mathematical Economics and Econometrics in Memory
of Yehuda Grunfeld, edited by Carl F. Christ et al. Stanford, Calif.: Stanford
Univ. Press, 1963.
. "Some Notes on Inflation." In Inflation and Growth in Latin America,
edited by W. Baer and I. Kerstenetzsky. Homewood, Ill.: Richard D.
Irwin, 1964.
Hicks, John R. "Economic Foundations of Wage Policy." Econ. J. 65
(September 1955):389-404.
Keynes, John M. A Treatise on Money. Vol. 1. London: Macmillan, 1958.
Mundell, Robert A. "A Fallacy in the Interpretation of Macroeconomic
Equilibrium." J.P.E. 73 (February 1965):61-66.
Olivera, Julio H. G. " On Structural Inflation and Latin-American 'Structuralism,"' Oxford Econ. Papers 16 (November 1964): 321-32.
Patinkin, Don. Money, Interest and Prices. 2d. ed. New York: Harper & Row,
1965. (First edition, 1956.)
Quirk, James, and Ruppert, Richard. "Qualitative Economics and the
Stability of Equilibrium." Rev. Econ. Studies 32 (October 1965):311-26.
Samuelson, Paul A. Foundations of Economic Analysis. Cambridge, Mass.:
Harvard Univ. Press, 1955.
Tooke, Thomas. An Inquiry into the Currency Principle. London: Longman,
Brown, Green, Longmans, 1844.
Wicksell, Knut. Geldzins und Guterpreise. Jena: G. Fischer, 1898.