THE FINANCIAL AND TAX EFFECTS OF

MONETARY POLICY ON INTEREST RATES

MICHAEL R. DARBY *
University of California, Los Angeles and
National Bureau of Economic Research

SUMMARY

Standard analysis of monetary policy effects on interest

rates in terms of liquidity, income, and expectations effects
is incomplete. After a change in monetary policy, substitution
among securities will increase as time elapses and so reduce
or eliminate financial effects caused by short-run financial
market segmentation. Also, the standard expectations effect
omits the transfer of income tax liability on that part of the
interest payment representing a return of real capital. So a
1 percentage point increase in the expected inflation rate
should increase the nominal interest rate by 1/(1 - r) per-
centage points, r being the marginal tax rate.

Nearly as rapid as the spread and adoption of the Fisher-Phillips

Curve after its rediscovery by A. W. Phillips in 1958l has been the
spread and adoption of the analytic separation of monetary policy effects
on interest rates into liquidity, income, and expectations effects. This
approach has been lucidly presented by Friedman (1968), fruitfully
applied by such authors as Cagan and Gandolfi (1969), Cagan (1972),
Gibson (March 1970; May 1970; MaylJune 1970; 1972), and Sargent
(1969; 1972), and already appears in such undergraduate texts as Baird
(1973), Kaufman (1973), and Sprinkel (197 1). Unfortunately, the empir-
ical implementations suffer from two crucial defects - one empirical
and the other analytical - that make interpretation hazardous: (1) The
theoretical model refers to an interest rate that is an index of interest
rates on a variety of loans, but empirical studies have exclusively used
interest rates on highly liquid bonds. There are plausible reasons for
these interest rates to display a different cyclical pattern from “the”

‘The author wishes to acknowledge helpful conversations with Armen Alchian, Milton Friedman,
and John Pippenger, with the usual acknowledgement that only credit for wisdom but not blame
for remaining errors can be shared. An earlier version of this paper was presented at the meetings
of the Western Economic Association, Las Vegas, Nevada, June 11, 1974. H. Irving Forman drew
the diagrams.
1. See Irving Fisher, “A Statistical Relation Between Unemployment and Price Changes,”
conveniently reprinted as “I Discovered the Phillips Curve” (1973). A constant hazard of any foray
into macroeconomics, including this one, is the significant probability that Fisher has already dis-
cussed the matter somewhere in his voluminous writings, only better. The basic idea of expected
inflations’ being reflected in interest rates derives from as early as Fisher (1896).

Economic Inquiry 266

Vol. XIII, June 1975
 DARBY: INTEREST RATES 267

interest rate. (2) Since Fisher’s original formulation of the expectations

effect, taxation of income has become quite significant, but the standard
model does not take into account the tax treatment of interest payments
and receipts. When taxation is considered, Fisher’s argument implies (on
the somewhat high assumption of a 50 percent marginal tax bracket2)
that an increase of 1 basis point in the expected rate of inflation should
cause a 2 basis-point increase in the nominal rate of interest observed
in the market.
In Section I, I review the standard argument on the effects of monetary
policy on interest rates. Differential financial effects on observed market
interest rates are introduced and analyzed in Section 11. Income taxation
is added to the model in Section 111. Section IV summarizes the revised
model.

I. LIQUIDITY, INCOME, AND EXPECTATIONS EFFECTS

Standard treatments attempt to answer the question: What will be

the effects on the interest rate, over time, of a change in monetary policy
from a constant growth rate of the money supply equal to go to a new
constant growth rate of g,? The problem is greatly simplified by the
assumption that the economy was in full long-run growth equilibrium
at time to when the monetary change is initiated. The solution is said to
depend on the interaction of the liquidity effect, the income effect, and
the expectations effect all of which display different temporal paterns.
The nominal demand for money at time t is assumed to be of the basic
form

where yt is real income, it is the nominal interest rate, and Pt is the

price level.
An increase in the rate of growth of the money supply3 will create
excess cash balances at the levels of yt, it, and Pt that would have existed
in the absence of the change in monetary policy. It is argued empirically
that the growth path neither of real income nor of the price level is much
affected immediately after a change in money supply growth and that
the nominal interest rate must fall over time to restore equality of money
supply and demand.4 This immediate impact of monetary policy on the
interest rate is known as the liquidity effect.

2. Approximately the postwar U.S. corporate tax rate

3. Precisely reverse arguments apply throughout to a decrease in the growth rate of the money
supply.
4. Note that while real income and the price level are unaffected, the nominal rate of interest
will be steadily falling not simply at a lower level. Interest rates will follow a similar pattern in
stock-adjustment models of money demand.
268 ECONOMIC INQUIRY

Over time, however, aggregate demand for goods is increased both

by direct impacts of the falling interest rate on investment demand and
of excess cash balances on purchases of consumers’ durable goods and
by indirect multiplier effects of the direct increases in spending. This
causes real income and prices to rise and so increase the demand for
money. This will be consistent with initially smaller decreases in it and
then - as the increases in the growth rates of y, and Pt become sub-
stantial - with increases in it. It is normally argued that the increases
in real income above its natural growth path are only temporary and
that in full equilibrium the level and rate of growth of the price level
will rise sufficiently to completely reflect the increased growth of the
money supply. The combined impact on the nominal interest rate of’the
temporary increase in real income above its equilibrium growth path
and the permanent increase in the growth path of the price level is
referred to as the income effect.
The expectations effect refers to Fisher’s argument that the nominal
interest rate it will exceed the real interest rate rt by the expected rate
of inflation a,. That is,

since both borrower and lender will agree to allow for inflation-induced
decreases in the real principal value at the rate a,.5 Observation of
higher rates of inflation will cause the expected rate of inflation to rise
and approach the new equilibrium rate of inflation pl . Since the increase
in the equilibrium rate of inflation ( p l - po) is equal to the increase in
the rate of growth of the money supply (gl - go), it is argued that
nominal interest rates will increase by the same amount in final
equilibrium.
In the standard presentations it is assumed that the liquidity effects
and income effects exactly cancel in their effects on the real interest
rate so that the nominal interest rate increases in final equilibrium by
exactly the increase in the rate of inflation ( p I - po = g, - go). This
assumption has not been firmly based however on either static or
dynamic analysis. If there is a negative interest elasticity of the demand
for monef and the nominal interest rate is increased, the desired ratio
of money to income will fall. This may affect both real income and the
real rate of interest.

5. These statements are based on continuous compounding. If annual compounding were used,
the right hand side would be increased by the interaction term ru.
6. This appears to depend on the conditions of money supply. Benjamin Klein (1970; 1974)
has developed an estimator of the yield paid on demand deposits. This yield usually moves in
proportion to the yield on short-term bonds. I (1972) have shown that, for the postwar U.S. data,
the net effect of changes in interest rates was zero for a fully specified stock-adjustment model,
Peltzman (1969) had previously found similar results for a stock-adjustmentmodel using consumer
expenditures as a proxy for an appropriatemix of wealth and cyclical portions of income.
 DARBY: INTEREST RATES 269

Detailed discussion of possible long run effects on the real interest

rate is beyond the scope of this paper. Static analyses arguing for a
changed real interest rate have been based on changes in the ratio of
money to bonds supposed to be induced by the money creation p r o ~ e s s . ~
Though only tangentially concerned with this issue, research on neo-
classical growth models containing money8 would seem more to the
point. Effects of alternative rates of money supply growth on the real
interest rate appear in those models to depend on the effects - if
any - on the nonmonetary savings ratio and the aggregate production
function.9 Analytical ambiguities can be eliminated - at times uncon-
sciously - by special assumptions on the impact of real money balances
on the nonmonetary savings ratio and the production function, but such
assumptions are far beyond the scope of existing evidence. In part this
is because of the imprecision of the problem: Is the increased monetary
growth replacing government borrowing or taxation or is it financing
increased government transfer payments or spending, and on what? As
a result of the compounded ambiguities of the theoretical analyses and
the lack of empirical evidence, it can be justified to assume that the
effect on the real interest rate is relatively small and of uncertain sign
and can be estimated as no change. A dogmatic attachment to an
unchanged - or decreased or increased - real interest rate in full
equilibrium cannot be maintained at the present.
The complete standard model is summarized in Figure 1. Both
nominal and real interest rates are expected to first fall below and then
rise toward their original level. The nominal rate will increase above
its original level as the expected rate of inflation rises. The real rate
may also temporarily overshoot its original level, but in any case
eventually equals its original level. When the expected rate of inflation
rises to the actual rate, the nominal interest rate will exceed its original
level by (gl - go).

7. Metzler (1951 ) and Mundell(l963).

8. See the review article on this literature by Allan Meltzer (1969).
9. Thus Tobin’s comments on Sargent’s paper (1972, pp. 477-78) erred in concentrating only
on the first change to the omission of the second. Even if as asserted by Tobin the nonmonetary
savings ratio rises, we cannot say whether the marginal product of capital falls or rises without an
explicit examination of the changes in the position and shape of the aggregate production function
implied by the reduced ratio of money to capital.
270 ECONOMIC INQUIRY

Figure 1. The Standard Model of Effectsof Monetary Policy on Interest Rates

l l t
'0

II. THE FINANCIAL EFFECT ON MARKET INTEREST RATES

Cagan and Gandolfi (1969) and Gibson (May, 1970; May/June, 1970)
have argued that the time pattern of effects of monetary policy on
interest rates can be used to obtain information about the lags in the
effect of monetary policy on nominal income.10 Since there must be
significant increases in real income and the price level to stop the decline
of the interest rate and start its increase, the trough of interest rate
changes would provide a measure of the time lag between an increase
in monetary growth and its significant effect on nominal income. This
argument assumes that the expectations effect is not operative at the
time of trough, but this may be unreasonable for economies that have
undergone large monetary swings in the past.
A more crucial difficulty - it appears to me - is that the interest rate
used is typically for short-term marketable securities." The cyclical
response of this class of interest rates can be explained by variations in
yield differentials without reference to effects of real income and the
price level on the demand for money. Bankers use marketable securities
as an adjustment asset (or secondary reserve) to absorb the shocks of

10. Note also that Cagan and Gandolfi examine the problem as stated here (a maintained increase
in the growth rate of money) while Gibson's work tends to emphasize the effects of once-and-for-all
changes in the money supply. This would seem to explain the shorter length of time for interest rates
to bottom out found by Gibson than was found by Cagan and Gandolfi. Excess cash balances will
continue to increase for some time in the continued increase case as opposed to the one-shot increase
case studied by Gibson.
11. For example, the commercial paper rate is used by Cagan and Gandolfi (1969)and Treasury
Bill rates by Gibson (May/June 1970).
 DARBY: INTEREST RATES 271

monetary policy without having to make frequent changes in the

quantity and terms of loans outstanding. That is, banks have much
lower short-run cross elasticities of demand between marketable securi-
ties and loans than exist in the long-run because of the relatively high
costs of adjustments for changes in outstanding loans. When the Fed
increases the rate of growth of the money supply by increasing its open-
market purchases of bonds, banks are initially flooded with excess
reserves. Banks use these excess reserves to purchase securities in the
open market thus forcing a multiple expansion of bank deposits and a
falling interest rate on marketable securities to induce individuals and
firms to accept the new deposits for their holdings of market securities.
As the monetary policy persists, banks begin to readjust the terms of
their loans in order to reachieve long-run portfolio balance. This will
over time relieve the pressure of bank demand on the yields of marketable
securities as banks move from an unusually large to an unusually small
net demand at the market interest rates. Put simply, an increased rate
of growth of high powered money will induce a cyclical overshooting
of bank demand for marketable securities which will be reflected in a
cyclical pattern of their interest rates. In the case considered the initial
decline will be heightened by the decreased net Federal government
issuance of marketable securities.
On the other hand the private supply of marketable debt securities
by issuing firms is probably more elastic with respect to continued dif-
ferentials between loan and market interest rates than initially. This will
enhance the eventual rise in market interest rates.
While it is true that these financial effects refer only to temporary
effects of monetary growth, they could very well cause market interest
rates to fall further and begin to rise before there is any significant
impact of income effects on the overall average of all “interest rates.”
It thus seems inappropriate to argue that the cyclical pattern of market
interest rates is superior to national income data as an indicator of the
lag in the effect of monetary policy on nominal income.
This analysis implies a significant market segmentation effect between
different interest rates which is eliminated in whole or part as the
financial markets adjust to the changed fraction of total wealth repre-
sented by government bonds. Whether the marginal product of capital
is changed in full equilibrium depends on the issues discussed in section I
with respect to money and growth models. Whether the yield differentials
among different classes of securities are altered depends on the long-run
cross elasticities of demand and supply for the different securities.
111. INCOME TAXES AND THE EXPECTATION EFFECT

The standard model of the effects of the expected rate of inflation on

the market rate of interest is based upon solving for that value of the
interest rate at which the amount of interest paid in terms of real goods
272 ECONOMIC INQUIRY

is constant in the long-run at r =r7 This is interpreted as implying that

the nominal interest payment i must be made up of the real interest
payment 7 plus a return of capital at the expected rate of inflation a.12
But this pattern of payments leaves the real costs of borrowing and real
rewards of lending constant only in a country without income taxes.
In a country with proportional income taxes at the rate T and for which
interest receipts are taxable and interest payments are deductible, the
real after-tax receipts and costs per dollar of loan are equal to

(3) r;“ = it - ri, - a,.

The first term on the right hand side is the gross receipts per dollar, the
second term is the tax liability shifted from borrower to lender, and the
third term is the change in the real value of the principal sum. If the
real after-tax interest rate is constant at 7 *in long-term equilibrium,
then the nominal rate of interest, for a constant T, is given by

(4) it = f * + a,) / (1 - T).

Hence, if the expected rate of inflation goes up by one percentage point,

the nominal interest rate must increase by enough to cover not only the
loss of principal due to inflation, but also by enough to pay the taxes
on that return of principal, that is by 1/(1 - T) percentage points.
Borrowers are just as well off in real terms paying this amount since
their tax savings and the lower real value of principal exactly offset the
rise in the nominal interest rate.
With progressive taxation, this simple formula holds only if the
marginal tax rate of marginal traders is constant. Clearly there would
be advantages in high-tax-rate individuals increasing their borrowing
and low-tax-rate individuals increasing their lending relative to equity-
type financing. Were the corporate tax rate of about .5 applicable at
the margin, a one percentage point increase in expected inflation would
be associated with a two percentage point increase in the observed
nominal interest rate. This is too high however. The implicit marginal
tax rate indicated by the yield difference between municipal bonds and
taxable securities currently appears to lie between .33 and .4,though
this may be depressed by the recent “tax reform” agitation and legisla-
tion. Even at the low end of the estimate, the nominal interest rate would
increase by 150 basis points for every 100 basis point increase in the
expected rate of inflation. When the correction for income taxation is
made, it is not at all clear that the rise in nominal interest rates during
the last ten years has anything like kept up with actual rates of inflation.

12. Seeequation (2).

 DARBY: INlEREST RATES 2 73

Fisher’s view that expected inflation rates are quite sluggish appears
more plausible.
Empirical estimates of the formation of expectations based on the
simple Fisher equation (2) instead of the corrected version (4)are subject
to considerable doubt. Specification bias due to constraining the coeffi-
cient of the expected rate of inflation to unity will exist for some methods.
Other methods use the equation (2) only for the interpretation of uncon-
strained results, so that the interpretation - but not the estimates
themselves - is affected. These problems are completely separate from
the specification difficulties of models which interpret correlation of
short-lag rates of inflation and the nominal interest rate as capturing
the expectations effect instead of the ‘income effect.’
If we consider the standard data on disposable income under expected
inflation there will be two main types of upward bias: (1) The real burden
of the inflationary tax on money will be added to the “true” income and
savings data. (2) Net interest payments by government and consumers
will increase stated income and savings by the amount of the return of
capital portion of interest payments.

IV. THE COMPLETE MODEL OF THE EFFECTS OF MONETARY

POLICY ON INTEREST RATES

An increase in the rate of growth of the money supply leads to a

pattern of changes in the interest rate that reflects the interaction of the
liquidity, income, and expectations effects which vary over time in
relative strength. In addition, interest rates on (especially short-term)
marketable securities will be influenced by financial effects due to
banks’ behavior with respect to their portfolio of investments and loans.
The liquidity effect causes “the interest rate” to fall in order to main-
tain (or move toward) equality of the demand for and supply of money
when the rate of growth of the money supply is increased suddenly with
only trivial immediate effects on the growth rates of real income and
the price level. As aggregate demand begins to increase relative to what
it would have been under the old monetary policy the growth rates of
real income and the price level are increased. This acceleration of the
growth of nominal income increases the demand for money. Eventually
these effects of more rapid nominal income growth are sufficient to
reverse the fall in the interest rate and cause a rise back toward its
original level. Since increased growth in the monetary base causes banks
initially to expand their investments in marketable securities and only
gradually to restore their desired ratio of loans to investments, there is
a temporary large increase in bank demand for marketable securities.
If the increase in the growth of monetary base in due to increased
Federal Reserve open market purchases, this will coincide with a
decreased net government supply of marketable securities. The resulting
relative fall in interest rates of marketable securities will later be reversed
274 ECONOMIC INQUlRY

as banks begin to increase their loans and decrease their demand (at least
in a flow sense)for marketable securities.
On the other hand, as time passes and nominal income adjusts to its
new steeper growth path, the rate of inflation will increase, then over-
shoot, and eventually settle at a rate (gl - go) above the original rate.
This increased rate of inflation will cause the expected rate of inflation
to rise and the difference between nominal and real interest rates to
increase.13 The nominal interest rate must rise by 1/(1 - r ) basis points
for each basis point increase in the expected rate of inflation, where T is
the marginal income tax rate, in order to leave borrowers’ and lenders’
expected payments and receipts unaffected in real terms. Achievement
of full adjustment takes many years in the postwar U.S.
Whether the real rate of interest is unaffected in full equilibrium is
not known. It could be unaffected, rise, or fall depending on whether
the desired ratio of money to income is altered and how this effects the
aggregate production function and real savings ratio. Convincing empir-
ical evidence is lacking. Real interest rates on marketable securities may
also be affected if there is a significant segmentation of the capital market
remaining in the long-run which alters yield differentials. The initial
segmentation implied by the financial effect is consistent with either no
or some residual long-run capital market segmentation.
The idealized pattern of the cyclical adjustment of interest rates to
an increase in the rate of growth in the money supply is illustrated in
Figure 2.
Figure 2. The Complete Model of Effects of Monetary Policy on Interest Rates

‘0

13. Indeed expectations may adjust before inflation is experienced under certain informational
conditions in which the change in the growth rate of the money supply is observed or announced
that leads to direct predictions of future inflation. See Darby (forthcoming) for a detailed analysis
of the use of such exogenous information in the formation of expectations.
 DARBY: INTEREST RATES 275

REFERENCES

D. W. Baird, Macroeconomics: A n Integration of Monetary, Search, and Income Theories. Chicago:

Science Research Associates, 1973.
P. Cagan, The Channels of Monetary Effects on Interest Rates. New York: Columbia University
Press, 1972.
P. Cagan and A. Gandolfi, “The Lag in Monetary Policy as Implied by the Time Pattern of Interest
Rates,” American Economic Reuiew, 59 (May 1969), 277-84.
M. R. Darby, “The Allocation of Transitory Income Among Consumer Assets,” American Economic
Reuiew, 62 (December 1972), 928-41.
, “Rational Expectations Under Conditions of Costly Information,” Journal of Finance,
forthcoming.
I. Fisher, “Appreciation and Interest,” Publications of the American Economic Association, 2
(August 1896). 33 1-442.
, “A Statistical Relation Between Unemployment and Price Changes,” International
Labor Reuiew, 13 (June 1926). 785-92. Reprinted as “I Discovered the Phillips Curve.” Journal of
Political Economy, 81 (MarchiApril 1973), 496-502.
, The Theory oflnterest. New York: Macmillan, 1930.
M. Friedman, “Factors Affecting the Level of Interest,” in Saving and Residential Financing: 1968
Conference Proceedings, D. P. Jacobs and R. T. Pratt, eds. Chicago: United States Savings and Loan
League, 1968.
W. E. Gibson, “Interest Rates and Inflationary Expectations: New Evidence,” American Economic
Review, 62 (December 1972). 854-65.
, “Interest Rates and Monetary Policy,” Journal of Political Economy, 78 (MayiJune
1970). 431-55.
, “The Lag in the Elffect of Monetary Policy on Income and Interest Rates,” Quarterly
Journal ofEconomics, 84 (May 1970). 288-300.
, “Price-Expectations Effects on Interest Rates,” Journal of Finance, 25 (March 1970),
19-34.
W. E. Gibson and G. G. Kaufman, “The Sensitivity of Interest Rates to Changes in Money and
Income,” Journal of Political Economy, 76 (MayiJune 1968), 472-78.
H. G. Johnson, “Money in a Neo-Classical One-Sector Growth Model,” Essays in Monetary Eco-
nomics. London: George Allen & Unwin, 1967.
G. G. Kaufman, Money, the Financial System and the Economy. Chicago: Rand McNally, 1973
B. Klein, “Competitive Interest Payments on Bank Deposits and the Long-Run Demand for Money,”
American Economic Reuiew, 64, (December 1974), 931-49.
, “The Payment of Interest on Commercial Bank Deposits and the Price of Money: A
Study of the Demand for Money,” Ph.D. dissertation, Univ. of Chicago, 1970.
A. H. Meltzer, “Money, Intermediation, and Growth,” Journal of Economic Literature, 7 (March
1969),27-56.
L. A. Metzler, “Wealth, Saving, and the Rate of Interest,” Journal of Political Economy, 59 (April
1951). 93-116.
R. A. Mundell, “Inflation and Real Interest,”Journal ofPolitical Economy, 71 (June 1963), 622-26.
, “The Public Debt, Corporate Income Taxes and the Rate of Interest,” Journal of
Political Economy, 68 (December 1960),622-26.
S. Peltzman, “The Structure of the Money-Expenditures Relationship,” American Economic Review,
59 (March 1969), 129-37.
A. W. Phillips, “The Relation Between Unemployment and the Rate of Change of Money Wage
Rates in the United States Kingdom, 1861-1957,” Economica, 25 (November 1958),283-99.
276 ECONOMIC INQUIRY

T. J. Sargent, “Anticipated Inflation and the Nominal Rate of Interest,” Quarterly journal of
Economics, 86 (May 1972). 212-26.
, “Commodity Price Expectations and the Interest Rate,” Quarterly Journal of Eco-
nomics, 83 (February 1969), 127-40.
, “Rational Expectatons, the Real Rate of Interest, and the Natural Rate of Unemploy-
ment,” Brookings Papers on Economic Activity, No. 2, 1973, 429-80.
B. W. Sprinkel, Money and Markets. Homewood, Ill.: Irwin, 1971
F. G. Steindl, “Price Expectations and Interest Rates,” Journal of Money, Credit, and Banking, 5
(November 1973), 939-49.
J. Tobin, “Money and Economic Growth,” Econometrica, 33 (October 1965),671-84.
S. Turnovsky, “Empirical Evidence on the Formation of Price Expectations,” Journal of the
American Statistical Association, 65 (December 19701, 1441-54.
W. P. Yohe and D. S. Karnosky, “Interest Rates and Price Level Changes, 1952-69,” Federal Reserve
Bank of St. Louis Reuiew, 51 (December 1969). 18-29.