Credit Cards, Economization of Money, and Interest Rates

2836 users shared this document! click

Credit Cards, Economization
of Money, and Interest Rates Frank G. Steindl An interesting , standard exercise in monetary-macroeconomics inquires into the effects of a reduction in the demand for money. One purpose is to demon-
strate that such effects are qualitatively identical to those associated with an
increase in the money supply, that is, each occasions an excess supply of money.
In those exercises, the avenue through which the reduced money demand occurs
is generally left unspecified, principally because it appears to be irrelevant. In this article , I deal with the issue of the effect on interest rates of increased use of credit cards, say, by households. This is not a purely hypothetical concern;
in recent years, there has been a marked shift to credit instruments, particularly
credit cards, to finance consumption. The question is what are the effects? The
corollary to increased credit card usage is a reduction in the demand for money,
specifically checkable deposits, a phenomenon in accord with Duca and White-
sells finding that for every 10 percent increase in the probability of owning a
credit card, checking balances are reduced by 9 percent (1995, 621). As an
empirical matter, checkable deposits fell 22 percenta 5.4 percentage annual
ratefrom the first quarter of 1994 through the third quarter of 1998. 1 At fi rst glance, the answer seems stra i g h t fo r wa rd. The fi rst round effect of e c o n o m i z ation of money balances because of increased use of credit cards is a
reduction in the interest rat e, a result identical with the liquidity effect of an
i n c rease in the stock of money, i n a s mu ch as each gives rise to an excess sup-
p ly of money. 2 But there is something wrong here. Gre ater credit card usage is one manife s t ation of an increase in the demand for cre d i t , wh i ch increases the Frank G. Steindl is Regents Professor of Economics and Ardmore Professor of Business Adminis - tration at Oklahoma State University (e-mail: steindl@okstate.edu.). The author thanks Scott Weir
and Ron Moomaw for comments in the departmental workshop, where an earlier version of this arti -
cle was read. Summer 2000 271 In this section, the Jo u rnal of Economic Educat i o n p u blishes art i cl e s
c o n c e rned with substantive issues, n ew ideas, and re s e a rch findings in
economics that may influence or can be incorp o rated into the teaching of
e c o n o m i c s . HIRSCHEL KASPER , Section Editor Content Articles in Economics i n t e rest rat e. Sure ly, the interest rate cannot be going up and down at the same
t i m e ! 3 The two pre fe rred money demand ap p ro a ches of the ninetiesthe cash-in- a dvance (CIA) and the ove rl ap p i n g - ge n e rations (OLG) modelsdo not ap p e a r
d i re c t ly useful for dealing with the credit card issue. The constant ve l o c i t y
i m p l i c ation of the CIA model rules it out as a ve h i cle for inve s t i gating a phe-
nomenon that increases ve l o c i t y. In add i t i o n , the model takes as given the pay-
ments period and money holdings as it re l ates age n t s planning periods to the
time period with respect to wh i ch velocity is measure d. Patinkin (1989,
xxxxxxi) emphasizes these concerns when he points out that the CIA fra m e-
wo rk is therefore a most unrealistic model to describe a world in which the statement
only credit cards can buy goods is an increasingly more accurate description that
the Clower dictum that only money can buy goods. Similarly, this [CIA] approach
most unrealistically implies that the velocity of circulation is constant. As for the OLG approach, its emphasis on moneys role as a store of value dic-
tates against it being used to investigate switches between transactions technolo-
gies. Accordingly, I use several well-traveled models of a few decades past to
demonstrate that the interest rate must rise. THREE MODELS The first model is an augmented price flexibility Patinkin (1965) framework in which the bond (i.e., credit) and money markets are highlighted. Use of credit
cards is captured by a shift parameter such that increases in it increase the sup- ply of bonds and reduce the demand for money, thereby generating excess sup -
ply in each market. This of course implies that there must be excess demand in
the commodity market. The formal equations for the money and bond markets
thus are L 2 , L 4 < 0 < L 3 < 1 B 2 , B 4 < 0 < B 3 < 1, where m is initial real balances M 0 /p, r is the interest rate, and Y is income (out- put). With price flexibility, output is at full employment Y 0 ; hence there is no need to specify partial der ivatives with respect to income . These two markets are used because it is in those that the disparate (partial equilibrium) interest rate conclusions are reached, that is, when each is used in a
partial equilibrium approach to understanding the behavior of the interest rate.
The models solution is (1) dr d = r 2 [B 4 (1 L 3 ) + B 3 L 4 ] B 2 (1 L 3 ) r 2 B 3 L 2 > 0. B Y 0 1 r ,m,



= 0; L(Y 0 , r, m, ) = m; 272 JOURNAL OF ECONOMIC EDUCATION Figure 1 shows the adjustment. The initial equilibrium, given by the intersection
of the BB 0 and LL 0 curves at a, establishes the interest rate r 0 and price level p 0 . An increase in the use of credit cards reduces the demand for money, thereby
shifting LL down to LL 1 . Increased borrowing via credit cards increases the sup- ply of bonds, shifting the bond equilibrium curve up to BB 1 . These shifts are the manifestation of interpreting interest rate movements in the partial equilibrium
terms, respectively, of money and credit markets. As the figure makes clear, the
result must be that the interest rate increases, with the new equilibrium at b. If the commodity market (C C c u rve) had been fo rm a l ly considere d, the re s p e c- t ive excess supplies in the bond and money markets would imply an excess demand
for commodities. The C C re l ation in Fi g u re 1 accord i n g ly shifts to the ri g h t , as a
result of wh i ch it intersects the bond and money equilibrium schedules at b. 4 The rightward shift of the commodity equilibrium curve, hence an increased price level, is a logical implication of the model. Interestingly, noneconomists
would likely surmise that one of the principal benefits of credit cards is as a vehi-
cle to increase current consumption; hence a rightward shift is what would be
expected. Economists on the other hand are predisposed to interpreting credit
cards as simply a way of reducing money demand, with no effect on aggregate
demand, perhaps because of the Keynesian-oriented money-bonds as substitutes Summer 2000 273 FIGURE 1 Credit Cards in Price-Flexibility Model framework. Yet, falling money demand is only one element in the credit card
story; increases in the supply of bonds is another; together with the consequent
excess supplies in the money and bond markets, there must be an excess demand
for goods and services. 5 Analyses of inflation, for instance, typically proceed by underscoring velocity increases resulting from an excess supply of money result-
ing from inflationary-expectations-induced economization of real balances. They
do not similarly emphasize that the corollary is an increased demand for com-
modities, in addition to increased nominal net indebtedness. See for instance
Mundell (1963), whose model is explicitly investigated later. The increase in aggregate demand, the CC curve, is a general result; it is nei- ther model- nor case-specific. Assume for instance that the increased use of cred-
it does not increase net indebtedness, that is, the increased short-term credit card
debt is offset by greater long-term bond holdings. This is equivalent to having the
BB curve unchanged. Hence, the increased use of credit cards would not shift the
BB curve up. The reduction in money demand because of greater usage of cred-
it cards shifts the LL curve down along the BB 0 curve. And the excess supply of money shifts the CC curve to the right so that it intersects the other curves at a
higher interest rate and price level. 6 That the use of credit cards serves in gener- al to increase consumption, that is, aggregate demand, is the expected result. 7 The second model is a heuristic IS-LM framework, augmented by a credit mar- ket. The goods and money market equations are familiar except that each now
includes , which captures the effect of credit cards. Here, as before, d reduces the demand for money and increases the demand for goods. The bond market is
modeled simply as D b 1 , D b 3 < 0 < D b 2 < 1 S b 1 , S b 3 > 0, 0 < S b 2 , The excess demand for bonds B d minus B s thus can be written as B 1 < 0, B 2 = ?, B 3 < 0. The sign of B 2 depends on the relative strength of income on the respective sup- ply and demand for bonds. Assume for convenience that B 2 = 0; increases in income increase the demand and supply of bonds by the same amount. 8 In the IS- LM quadrant, the bond market equilibrium curve BB would thus be horizontal. 9 An increase in the use of credit cards, d > 0, shifts the bond equilibrium curve up to BB 1 and the LM curve down to LM 1 , as shown in Figure 2. Regardless of the magnitude of the downward shift, the interest rate must rise. As for IS, its
rightward shift further underscores the increase in the interest rate. 10 Here again, B 1<i>r ,Y,



= 0 B d = B s . B s = S b 1 r ,Y,



B d = D b 1 r ,Y ,



274 JOURNAL OF ECONOMIC EDUCATION economization of money resulting from increased use of credit cards occasions
an increase in the demand for commodities. This result accords not only with
noneconomists perceptions of the role of credit cards; it also underscores the
monetarist transmission channel, the one that stresses the link between an excess
supply of money and its direct effect on aggregate demand, regardless of whether
the excess supply of money resulted from an increase in the money stock or a
decrease in money demand. The last model in which the influence of a shift to credit cards for financing consumption can be examined originated with Mundell (1963), the basis of the
MundellTobin effect in the economic growth literature. His model is an amal-
gam of the previous models. 11 The goods and money markets are considered in the Keynesian framework associated with the IS-LM model, but its price sticki-
ness assumption is abandoned. Here the Mundell model follows the price-flexi-
bility full employment framework of the Patinkin model. The specific setup is: I 1 < 0 < I 2 : S 1 , S 2 < 0 L 1 , L 2 < 0. The first equation is the IS curve in which there is a wealth-saving relationship. Note that plays the same role as previously, namely, that a shift to credit card m = L(r, ) I(r, ) = S(m, ) Summer 2000 275 FIGURE 2 Credit Cards in IS-LM F ramework financing of expenditure increases aggregate demandconsumption (S 2 < 0) and investment (I 2 > 0). That is, the excess supply in the money (and bond) market(s) implies an excess demand for goods and services, this being a consequence of the
reduced demand for money. Curiously, Mundells economization of real money
resulting from inflation expectations does not result in increased demands in the
commodity market. This is another example of economists treating reductions in
the demand for money as not affecting directly the demand for goods and ser-
vices. The IS curve is positively sloped in the (m, r) planean increase in real bal- ances m reduces saving. For equilibrium, investment must fall, and that occurs if
the real interest rate r rises; therefore The LM curve is negatively sloped1/L 1 < 0in the (m, r) plane, as can be seen in Figure 3. The credit (bond) market can be modeled (Steindl 1986, 25862) as an excess demand for bonds. dr dm IS = S 1 I 1 > 0. 276 JOURNAL OF ECONOMIC EDUCATION FIGURE 3 Influence of Credit Cards in Mundell Model B 1 , B 3 < 0 < B 2 , where B 3 represents the increased supply of bonds, that is, the demand for cred- it, resulting from switching from money to credit cards. In the (m, r) plane, the
BB curve is negatively sloped, To the left of the equilibrium, it lies below IS and above LM; there cannot be
excess supply in all mar kets. 12 From an initial equilibrium point a, the move to credit card financing of expenditures reduces the demand for money, thereby shifting LM down to LM 1 . Similarly, the bond curve shifts up to BB 1 as the demand for credit increases. Increased goods market expenditures push prices higher, thereby reducing the
real money stock and, of course, real money holdings m. IS accordingly shifts to
the left, to IS 1 , yielding a new equilibrium at point b, at which the interest rate has risen to r 1 . With a fixed stock of money, the rise in the price level reduces real money holdings to m 1 . The rise in the price level can thus be interpreted as owing to an increase in velocity. CONCLUDING OBSERVATIONS In this art i cl e, I have dealt with the effect on interest rates when households shift to an incre a s i n g ly popular method of financing current consumption: c re d i t
c a rds. Use of two conventional partial equilibrium fra m ewo rks gives disparat e
a n swe rs , t h at is, the rate rises in one and falls in the other. In an important sense,
this situation is homologous to the fra c t i o u s , petulant liquidity pre fe re n c e - l o a n-
able funds controve rsy that consumed the pro fession in the decades fo l l owing the
p u bl i c ation of Key n e s s G e n e ral Th e o ry in 1936. As Hicks (1946) established in
his Wa l ra s s Law n + 1 equat i o n s w i t h n + 2 go o d s , b o n d s , and money s e t u p ,
the interest rate is a ge n e ral equilibrium phenomenon, although d e t e rmining . . .
the rate of interest by the demand and supply of [bonds, i. e.,] loan funds . . . is
the most nat u ral course to purs u e, and there does not seem to be anything aga i n s t
i t ( 1 9 4 6 , 161). Just as that debate was resolved by resorting to general equilibrium consider- ations, so too can the credit card conundrum be settled using such a framework.
The tractable models used herePatinkin-based price flexibility, heuristic IS-
LM, and Mundellianestablish that the interest rate must increase when credit
cards are increasingly substituted for cash in financing expenditures. Hickss
(1946) dictum concerning loana ble funds as the most natural course to pursue,
and there does not seem to be anything against it holds true. But the shift from money to credit cards is not a zero-sum game in which the same amount of expenditures is financed. Rather, the models establish that the dr dm BB = r 2 B 3 B 1 < 0. B 1<i>r ,m,



= 0 Summer 2000 277 credit card induced economization of money also stimulates additional spending
as agents reallocate their excess money holdings to increased expenditures,
thereby increasing velocity and raising the price level. NOTES 1. Economization of checkable deposits associated with credit card use may only be a minor factor in their decline. More likely, it is the increasing popularity of sweepaccounts that underlies the
deposit decline as banks seek to avoid the reserve requirement tax. 2. The excess supply of money also generates increased demands for goods and services, thereby increasing the price level,this time via velocity inasmuch as the stock of money is unchanged. A
model formally dealing with simultaneous interest and price adjustments is presented in the last
part of the next section. 3. In fact,there is a third possibility. If the central bank tar gets interest rates,the money supply then accommodates changes in money demand, thereby leaving the interest rate unchanged. Targeting
of interest rates cannot,however, be a permanent policy because price movements induced by the
endogenous money supply changes would be encapsulated in interest rates. In fact, formal mod-
eling of the interest rate targeting strategy results in indeterminacy of the nominal money supply
and price level, though real money holdings are determinate. Copies of this result are available
from the author. 4. The equation for the commodity mar ket, CC, is F(Y 0 , r, m, ) = Y 0 ; F 2 < 0 < F 3 < 1; F 4 > 0. Use of credit cards thus shifts the curve to the right. The analysis could similarly have been done in
terms of the commodity and bond markets. In that case, the former shifts right and the latter left,
thus definitely increasing the interest r ate. 5. In terms of a money-goods model,the interest rate would fall if the commodity mar ket had F 4 = 0, the idea being that reduced money demand does not affect the commodity market directl y.
Even if it did , that is, for F 4 > 0, the interest r ate movement depends on the relative shift of the money and commodity curves, that is,it is indeterminate. It is only with the formal, explicit con-
sideration of the bond market that the interest rate movement can be definitely signed, and it is
positive. 6. Note, however, the basic implausibility of keeping net indebtedness constant. For one, such actions are special cases of the Savings and Loan problemfunding long-term assets with short-
term debt. Of more consequence, because rates on credit card debt are higher than even on junk
bonds, the constant net indebtedness strategy is equivalent to buying high and selling lo w. 7. The one case where this does not occur is when credit card debt is used solely to finance acqui- sition of longer-term bonds. Neither money nor aggregate consumption demand is affected. The
basic implausibility of this is discussed in the foregoing note. 8. The procyclicality of interest r ates suggests that the supply of bonds increases by more than the demand as income rises, thus resulting in an excess supply, B 2 < 0. The bond cur ve accordingly becomes positively sloped, 9. Because there cannot be excess demand (supply) in all markets simultaneously, the BB curve lies below the IS and above the LM curves to the left of the equilibrium and below the LM and above
the IS curves to the right of the equilibrium. In addition,stability conditions similarly require such
a relationship. 10. There are two points here. First, because economization of money increases aggregate demand, the IS relation must include a term (capturing it, so that a typical equation would be Y = F(Y, r, ) < F 1 < 1, F 2 < 0 < F 3 . Next, for the case of a positively sloped BB bond curve in conjunction with LM, the argument parallels that in Figure 1 in which the interest rate rises. The increase in
the interest rate when the BB curve is negatively sloped can be seen from the goods and bond
markets. IS shifts to the right and BB to the left. If, however, the bond and money mar kets alone
are emphasized, the movement of the interest rate appears ambiguous. Imposition of the goods
market IS establishes,however, that the rate cannot decline; in fact, it must increase. 11. His concern, as is evident from the title, is with the Fisher distinction between nominal and real interest rates owing to inflationary expectations. That is not a concern of the present article, hence
the notation r is taken as pneumonic symbol for both the nominal and real interest rate. 12. Stability requires that the BB curve be less steeply sloped , in absolute value, than the LM curve. dr dY BB = r 2 B 2 B 1 > 0. 278 JOURNAL OF ECONOMIC EDUCATION REFERENCES Duca, J. V., and W. C. Whitesell. 1995. Credit cards and money demand: A cross-sectional study. Journal of Money, Credit, and Banking 27 (May): 60423. Hicks, J. R. 1946. Value and capital. 2nd ed. Oxford: Clarendon.
Mundell, R. 1963. Inflation and real interest. Journal of Political Economy 71 (June): 28083.
Patinkin, D. 1965. Money, interest,and prices. 2nd ed. New York:Harper & Row.
. 1989. Money, interest, and prices. 2nd ed. abridged. Cambridge, Mass.: The MIT Press.
Steindl, F. G. 1986. General equilibrium models of inflation and interest rates:Specification consid- erations. Kredit und Kapital 19 (Heft 2): 25270. Summer 2000 279

Comments