Full text of Alternative Approach to the Monetary Mechanism
The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
[SUBCOMMITTEE PRINT]
AN ALTERNATIVE APPROACH TO THE
MONETARY MECHANISM
SUBCOMMITTEE ON DOMESTIC FINANCE
COMMITTEE ON BANKING AND CURRENCY
HOUSE OF REPRESENTATIVES
88th Congress, 2d Session
Printed for use of the Committee on Banking and Currency
U.S. GOVERNMENT PRINTING OFFICE
34-474
WASHINGTON : 1964
COMMITTEE ON BANKING AND CURRENCY
WRIGHT PATMAN, Texas, Chairman
CLARENCE E. KILBURN, New York
ALBERT RAINS, Alabama
WILLIAM B. WIDNALL, New Jersey
ABRAHAM J. MULTER, New York
EUGENE SILER, Kentucky
WILLIAM A. BARRETT, Pennsylvania
PAUL A. FINO, New York
LEONOR K. SULLIVAN, Missouri
FLORENCE P. DWYER, New Jersey
BENIIY S. REUSS, Wisconsin
SEYMOUR HALPERN, New York
THOMAS L. ASHLEY, Ohio
JAMES HARVEY, Michigan
CHARLES A* VANIK, Ohio
OLIVER P. BOLTON, Ohio
WILLIAM S. MOORHEAD, Pennsylvania
W. E. (BILL) BROCK, Tennessee
ROBERT G. STEPHENS, JR., Georgia
ROBERT TAFT, JR., Ohio
FERNAND J. ST GERMAIN, Rhode Island
JOSEPH M. McDADE, Pennsylvania
HENRY B. GONZALEZ, Texas
SHERMAN P. LLOYD, Utah
CLAUDE PEPPER, Florida
JOSEPH G. MINISH, New Jersey
BURT L. TALCOTT, California
CHARLES L. WELTNER, Georgia
DEL CLAWSON, California
RICHARD T. HANNA, California
BERNARD F. GRABOWSKI, Connecticut
CHARLES H. WILSON, California
COMPTON I. WHITE, JR., Idaho
JOHN R. STARK, Clerk and Staff
Director
JOHN E. BARRIERE, Professional
Staff
Member
ALVIN LEE MORSE,
ORMAN S. FINK, Minority
Counsel
Staff
Member
SUBCOMMITTEE ON DOMESTIC F I N A N C E
WRIGHT PATMAN, Texas, Chairman
HENRY S. REUSS, Wisconsin
WILLIAM B. WIDNALL, New Jersey
CHARLES A. VANIK, Ohio
JAMES HARVEY, Michigan
CLAUDE PEPPER, Florida
OLIVER P. BOLTON, Ohio
JOSEPH G. MINISH, New Jersey
W. E. (BILL) BROCK, Tennessee
CHARLES L. WELTNER, Georgia
ROBERT TAFT, JR., Ohio
RICHARD T. HANNA, California
CHARLES H. WILSON, California
ROBERT E. WEINTRAUB, Senior
Economist
ROBERT A . SCIIREMF,
Investigator
HARVEY W . GEIST,
Investigator
STEPHEN D. KENNEDY, Research
Assistant
TL
LETTER OF TRANSMITTAL
AUGUST 17,
1964.
To the Members o/ the Subcommittee on Domestic Finance:
Transmitted herewith for the use of the subcommittee is the third
and final part of a staff analysis of the Federal Reserve System's
policy action pertaining to the System's determination of the volume
of money and credi\
This part I I I contains an appendix which relates to parts I and I I as
well as to this part. It is contemplated that at a future date this
study may be published in its entirety in one volume.
Sincerely yours,
W R I G H T P A T M A N , Chairman.
m
LETTER OP TRANSMITTAL
TT
_
H o n . WRIGHT PATMAN,
AUGUST 14, 1 9 6 4 .
Chairman^ Home Banking and Currency Committee,
House of Representatives, Washington, D.C.
D E A R M R . C H A I R M A N : Transmitted herewith is the last portion
of a three-part study on "An Analysis of Federal Reserve Monetary
Policy Making," that has been prepared for the committee. The first
part, chapter I I of the completed study, was published on February
10 and described "Some General Features of the Federal Reserve's
Approach to Policy." It discussed the generally inchoate nature of
the Federal Reserve's conception of monetary processes. Some major
consequences of this situation, revealed by systematic misinterpretations of policy actions, were presented. The evident disregard of rational methods of policymaking was partially explained in terms
of the Federal Reserve's inherited procedures.
Among the many disconnected strands that apparently constitute
the Federal Reserve's conception of monetary mechanisms, we found
a frequently recurring notion that occupies a dominant position. This
notion emphasizes the central role of free reserves in the causal process shaping the behavior of the monetary system. The Federal
Reserve's choice of free reserves as an indicator of monetary situations and of their prevalent policies has been greatly influenced by
this notion. The occasional choice of free reserves as a target of
policy actions was similarly affected by the Federal Reserve's "free
reserve doctrine."
The pervasive occurrence of this doctrine among the Federal Reserve's policy statements suggested a more detailed investigation of
the particular notion. The second part of our study containing chapters I I I , I V and V, published on May 7, was therefore devoted to the
"The Federal Reserve's Attachment to the Free Reserve Concept."
The discussion traced some of the prevailing confusion to an inadequate distinction between three logically independent ingredients,
viz, free reserves as a centerpiece or the causal process transmitting
policy actions to the monetary system, free reserves as an indicator or
situations and policy positions, and free reserves as a target of policy
actions. Evidence from published statements and overt actions was
presented in support of our contention about the central role of the
'free reserve doctrine." Moreover the doctrine was tested in order to
appraise its relevance. We found that, for the data available? it had
almost no explanatory value. More specifically, our appraisal repealed almost no connection between "credit" or money supply on the
<>ne side and free reserves on the other. If this doctrine were held to
J* a true description of the monetary process, one would therefore be
forced to admit that monetary policy is indeed a futile exercise*
feince the doctrine is not the only conception of the monetary media-
YX
LETTER OF TRANSMITTAL
nism, there is an alternative interpretation of the results: viz, the free
reserve doctrine seriously misconstrues the true nature of the money
supply process. But this can only be judged appropriately by formulating an alternative conception of this process and exposing it to
observations in competition with the Federal Reserve's doctrine.
The present and last part of our study (sees. I - I I I of the present
report), containing chapters VI* V I I , and V I I I of the full study,
pursues this theme. I n chapter V I we describe in some detail an alternative conception that we label the "modified base doctrine." Under this conception the money supply (and the banks' portfolio of
earning assets) is determined by the monetary base (i.e. the volume
of high-powered money issued by the government sector of the economy), the legal requirement ratios, the public's behavior revealed by
its division of money balances between currency and checking deposits and total deposits between checking and time accounts, and
the banks' desired reserve positions. The characteristics of these determinants are explained and their interaction in the process
described. A number of implications for monetary policy and the
behavior of the money supply process are presented. Among the
major implications may be mentioned:
(1) The monetary multiplier, which measures the System's magnifying power in response to injection of additional reserves, is substantially below the reciprocal ot the average reserve requirement ratio,
the measure of the monetary multiplier usually used by spokesmen for
the System.
(2) The major reason for the comparatively low multiplying power
exhibited by the System is the systematic occurrence of currency
drains from banks to the public, typically associated with the banks'
portfolio responses to changes in their reserve position.
(3) I n addition to these systematic currency drains other features
of the public's demand for currency also have an important position
in the monetary process.
(4) The single most important determinant of the money supply
is the extended base. This magnitude summarizes the actual policy
of the Federal Reserve authorities. The monthly averages of the
extended base can be adjusted to any desired level by suitable Federal
Reserve actions.
(5) The analysis developed also implies the existence of a continued
effective connection between policy actions and the money supply
throughout the depression of the 1930's.
The development of an alternative conception of the money supply
process is not sufficient to indicate its relevance. The exposure of
the Federal Reserve's conception to observation should be matched
by similar exposure of the alternative presented. Chapter VII,
therefore, summarizes some results of a detailed appraisal of the
"modified base doctrine," A few of these results are briefly indicated:
(1) The variations in the annual growth rate of the extended base
account for approximately two-thirds of the variations in the growth
rate of the money supply during the 1950's.
(2) 'The.completely specified base doctrine accounted for 95 percent
of the variations in the growth rate of the money supply observed
J
1 J
during the 1950's:
LETTER OF TRANSMITTAL
VII
(3) The orders of magnitude implied by the modified base conception for the. response in the money supply's growth rate to accelerations in the base, the accumulated sum of "liberated reserves,"
the public's currency and time deposit behavior and the banks' reserve
adjustment are confirmed by observations. In particular a dollar's
worth of open-market purchases induces on the average the same
response as a dollar's worth of reserves liberated by lower requirement
ratios.
(4) The monetary multiplier is substantially below the reciprocal
of an average of the requirement ratios. Our estimates indicate that
per dollar change in the extended base, or per dollar of open market
operations, the money supply changes by $2.50 to $2.70, on the
average.
(5) The banks' reserve adjustment is found to be sensitive to variations in open market rates and the discount rate. But this sensitivity
does not prevent an effective connection of monetary policy, summarized by the extended base, and the money supply. Observations
for the 1930's yield quite negative results for the notion that large
excess reserves broke the link between the money supply and policy
actions. Variations in the base continued to be transmitted effectively
to the money supply.
(6) The modified base conception was used to predict the money
supply for 16 quarters beyond the period used to obtain estimates.
The average predictive error for the money stock was one-half of
1 percent and for the changes in the money supply 10 percent.
Moreover, the association of predicted and actual changes and predicted and actual accelerations was evaluated. The results show a
strongly marked association between predicted and observed values.
Finally the chapter covers some additional material bearing on the
public's currency behavior and a further discussion of selected policy
problems in the context of the modified base doctrine. It is found
that the results of a number of policy actions, that are inexplicable
in terms of the ruling Federal Reserve notion, can be interpreted
usefully in terms of the conception presented.
We wish to emphasize that the material in chapters V I and V I I is
not presented with any sese of finality or conclusiveness. We are
convinced that diligent research will yield substantial improvement
in our future knowledge. Rational policymaking will undoubtedly
benefit from such procedures. The purpose of these chapters is not
to close a discussion, but on the contrary, to open a discussion by
posing a challenge to the Federal Reserve authorities. We have criticized their policy procedures and their disregard for substantiated
analysis. Our investigation of their conceptions and tests of their
leading notions have found them seriously inadequate. We have submitted an alternative frame of analysis that is both more explicitly
formulated and better substantiated than the Federal Reserve's
ruling conception. We hope the Federal Reserve will respond to the
challenge and carry the analysis and evidence further.
Our recommendations in chapter V I I I must be viewed in the context of the above remarks. We present 8 general recommendations
and 10 specific proposals. Among the major recommendations we
emphasize the importance of deliberate efforts to apply systematic
Vni
LETTER OF TRANSMITTAL
analysis to improve the policymaker's understanding of monetary
processes. I n particular, the policymaking bodies should learn to
exploit their research facilities more effectively. Moreover, the
Federal Reserve authorities should once acknowledge fully and without reservation that they have a responsibility to control the growth
rate of the money supply and that they possess the technical means to
avoid its past gyrations.
Monetary policy is not a panacea. But appropriate monetary
policy can contribute substantially to avoid both mass unemployment
and substantial inflation. There is no reason or excuse for permitting
inappropriate monetary arrangements or policies to aggravate economic fluctuations as they have in the past. Neither is there reason
to permit inadequate analysis and unverified assertion to continue
to furnish the basis for policy decisions.
KARL BRUNNER.
A L L A N H . MELTZER.
CONTENTS
^Financc t r a i l S m i U l 1 1 t 0 t h ° m c r a b c r 8 o f t h e Subcommittee on Domestic Pag§
Lettcr of trunsmittal"to~lion'Wri gh~t~l>a tman~bv Karf I3run"ner and"Allan"
11. Moltzer
v
section I.—An alternative approach to the monetary mechanism
I
l
An outline of the money supply process.2
The basic ingredients of the process
3
I he two major mcchanisms constituting the money supply process
7
The multiplier mechanism..
7
The processes injecting surplus reserves
9
The monetary base
9
Table VI-1.—Sources and uses of the monetary base
12
Relative size and variation of the sources of the base
13
Table VI-2.—Range of variation of annual percent change of base
and of the contributions made by major source components.14
Table VI-3.—Average monthly changes of source components of
the monetary base between adjacent months during half cycles.
16
Table VI-4.—Extreme values of average monthly changes in postwar half cycles
-16
Reserve requirements.
16
Table VI-5.—Release of required reserves occurring between
corresponding months of adjacent years, attributable to the
redistribution of demand deposits between different classifications of member banks.
19
Currency drains that are independent of banks' portfolio adjustments
20
Variations in the public's demand for time deposits independent
of the bank's portfolio adjustment
25
The role of interbank deposits
27
Variations in banks' desired excess reserves independent of their
portfolio adjustments
28
Table VI-6.—Annual rate of growth of the money supply before
and after changes in reserve requirements monthly 1936 and
1937.
30
A summary of the money supply process and an indication of some
evidence..
32
Some comments on policy action
34
Discount policy
34
The Federal Reserve's check collection facilities
35
The administration of Treasury deposits
——
-36
Treasury deficit and surplus
-36
Some conflicting implications of the base and free reserve doctrines
38
Conclusion
41
Section II.—Some evidence on the relation of the base mechanism to the
supply of money
43
The relation of the modified base conception to the money supply--43
Table VII-1,—Measures of association between changes in the
money supply and changes in the extended base
45
CONTENTS
x
Section II.—Some evidence on the relation of the base mechanism to the
SUf P
i he effect^ofTh^pr^ndpal determinants on the money supply. . .
Table VII-2.—The response of the change in the money supply to
changes in the extended base
-Table VII-3—The response of the money supply to the separate
components of the extended base, first quarter 1948 to fourth
quarter 1959
--Table VII-4—The magnitude of the change of the money supply
in response to changes in each of the elements of the modified
base conception—period, first quarter 1949 to fourth quarter
1962
The banks' liquidity trap
-------Table VII-5.—The comparative magnitude of the push of the
extended base and the pull of income on the money s u p p l y .
Some predictions of the money supply.
--Table VII-6.—The predictive performance of the modified base
doctrine
-----Table VII-7.—A measure of the closeness of association between
predicted changes and observed changes in the inonev supply-Table VII-S — Measure of the association of the changes between
adjacent quarters of changes in the predicted values of the
money supply between corresponding quarters, acceleration and
deceleration...
Currency behavior and money supply
Table VII-9.—Currency patterns in the postwar period..
A discussion of selected policy actions in 1946-51.
Pegged bond yields and inflationary pressures...
Table VII-10.—Reserve bank credit outstanding, Federal Reserve holdings of Government securities and gold in the prcaccord period
Changes in reserve requirements
7-Table VII-11.—Compensation of changes in reserve requirements, by date and compensation percentage
Table VII-12.—The association between changes in the base and
changes in the accumulated sum of liberated reserves during
periods surrounding changes in reserve requirements.
Has the Federal Reserve had a countercyclical policy in recessions?
(1) The recession from November 1948 to October 1949
(2) The recession from July 1953 to August 1954
(3) The recession from July 1957 to April 1958
(4) The recession from May I960 to February 1961
(5) The Federal Reserve's evaluation of calendar year 1963Preliminary conclusion
The stock of money, velocity, and national income
Chart VI-1.—National income and predicted income, quarterly
1954-IV to 1959-1V
-Chart VI-2,—Deflated net national product and price deflator,
1954-IV to 1959-1V
_
J
Conclusion
Section III.—Some suggested changes in policymaking procedures.
Chart VIII-1.—First differences between adjacent months of the
extended monetary base (paster)
Facing
Some general suggestions for policymaking
___
Some specific suggestions
.
_
Conclusion
IIIII'IIIIIIIIIIII"-..
APPENDIX I
Questions and responses to questionnaire mailed to each member of the
Board of Governors and each president of a Reserve bank who held the
position on August 21, 1963
_
^
45
4
'
^
ou
5ob6
Jjj
w
""
"b
68
69
W
^
ij{
701
y
£0
77
79
©J
^
87
9o
APPENDIX II
Scaling of the Federal Reserve's policy decisions.
119
CONTENTS
XT
APPENDIX III
The relation of changes in money to changes in bank portfolios ("credit")APPENDIX IV
Charts:
Three-week moving average of free reserves for all member banks 194662 (paster)..
face page
Annual changes in the extended monetary base, monthly 1945-62
(paster)....
face page
Annual changes in currency between corresponding months 1946-62
(paster)
face page
The ratio of the money supply in a given month to corresponding
month, 1946-62 (paster)
face page
The ratio of the money supply plus time deposits in a given month to
corresponding month, 1946-62 (paster)
face page
127
132
132
132
132
132
SECTION I — A N
ALTERNATIVE APPROACH TO T H E MONETARY
MECHANISM
Two possible interpretations of the observations presented in our
earlier chapters 1 may be suggested. First, we might conclude that
Federal Reserve policies, particularly open-market operations, have
very little effect on the level or rate of change of the stocks of money
and credit. Even if changes in money have a decisive influence on
the pace of economic activity, the practical usefulness of discretionary
monetary policy is severely reduced if this conclusion is correct. "We
have seen that the FOMC and the manager correctly evaluated future
or current changes in the pace of economic activity in the postwar
years. And they generally changed the level of free reserves in the
direction that was appropriate in terms of their understanding of the
monetary mechanism. But the evidence suggests that there is very
little relation between the elements incorporated in the modified free
reserve doctrine and monthly or annual changes in the supply of
money. The relation of free reserves to changes in member bank
credit is weaker still. Since there is little assurance that Federal
Reserve operations designed to induce changes in the supply of money
will have much effect on the money stock, or on "credit," there must be
doubt about the effect of their policy on income, whatever the response
of income to variations in the stocks of money or "credit."
A second interpretation of the observations is that the Federal Reserve's approach is based on an incorrect conception of the process
by which monetary policy alters the stock of money. In this view,
the evidence presented in the last chapter tells us little about the
effectiveness of discretionary monetary policy. I t suggests only that
the Federal Reserve has not developed a valid understanding of the
monetary mechanism. To appraise the relation of monetary policy
to changes in the stock of money, an alternative view is required.
The choice that is made between these two interpretations has important consequences. Monetary policy has been relied upon in the
postaccord period to reduce unemployment and to prevent inflation.
At times, reliance on effective monetaiy policy has included the apparent belief that restriction of the growth of the stock of money
could maintain price stability during periods of economic expansion
the face of deficits in the Government budget. If monetary policy
operations have only an uncertain effect on the stock of money, this
confidence seems to be misplaced.
.
In this chapter, an alternate view of the monetary mechanism is
discussed. Selected policv actions will be analyzed in terms of the
alternative conception. Evidence supporting the approach and the
Power of monetary policy will be presented below as a means of
judging the usefulness of the suggested approach.
Drln?^
Reserve's Attachment to the Free Reserve Concept/' a subcommittee
b> Karl Brunner and Allan H. Meltter.
2
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
The mechanisms to be outlined here connect Federal Reserve policy,
the behavior of the banks, and the public with the stock of money.
The theory presented is a theory of the money supply, reflecting an
approach that emphasizes the importance of the stock of money as a
determinant of the pace of economic activity and the importance of
Federal Reserve policy as a determinant of the stock of money. Because of the persistent confusion between money and "bank credit,
we note, first, that we do not regard these terms as synonyms. Statements that apply to the stock of money do not necessarily apply to
the stock of "credit" Second, it is not a valid procedure to dismiss
analysis and evidence with the statement often made by spokesmen
for the Federal Reserve system, "we prefer to view the process in
terms of ' c r e d i t ' W i t h o u t analysis and evidence relating Federal
Reserve policy to "credit" and "credit" to the pace of economic activity, there is no reason to regard these statements as more than an
announcement of the speaker's preference that is without bearing on
any substantive issue. Third, we do not present very much detailed
analysis or evidence supporting the statement that the money supply
is an important determinant of the pace of economic activity. Such
analysis goes well beyond the scope of this report, concerned primarily
with the relation of policy action to the money supply. But we wish
to point out that the use of monetary policy as a means of promoting
employment of resources and preventing inflation presupposes a connection between the supply of money and the pace of economic activity.
If the two are unrelated, there would be much less reason to be concerned with Federal Reserve actions.
AN OUTLINE OF THE MONEY SUTPTA" PROCESS
The behavior of the money stock emerges f r o m the interaction of
the public and the banks, responding to the conditions of the Government sector's monetary accounts. These accounts are a c o n s o l i d a t i o n
of the Federal Reserve banks' balance sheet and a statement summarizing the Treasury's position and activities in currency issues and in
gold. The monetary base emerges from the consolidated statement
The base is the volume of high-powered money directly issued and
controlled by the monetary division of the economy's G o v e r n m e n t
sector. Base money consists of (1) currency issued by the Treasury
(coin, silver certificates, and other Treasury currency issues); (2)
notes issued by Federal Reserve banks; and (3) those deposits at Federal Reserve banks that are assets of member banks.
If all the base money were held by the public and if liabilities of
private economic units were rarely used to settle obligations, the Nation's money stock would coincide* with the monetary base. But such
identity between monetary base and money stock is observed only in
countries with exceedingly primitive economic organization. At some
stage of economic development, liabilities of private economic units
become generally accepted as a means of payment and form part of
the money stock. In any countiy with sufficiently complex economic
organization, the money stock is a multiple of "the monetary base.
The existence and peculiar functioning of banks is the sine qua non
for such multiplying power; part of the monetary base is held as currency by the public, the remainder is held by banks as a reserve against
demand and time liabilities of banks. Both currency and deposits at
a central bank may be used as a part of bank reserves/
3
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
The basic ingredients of the process
Under our present arrangements the Board of Governors can effectively control the monetary base with the aid of open market operations and suitable adjustments of the discount rate. Given the
magnitude of the base, the behavior patterns of the public and banks
determine the division of the base into reserves at banks and currency
balances in possession of the public. These crucial behavior patterns are summarized by the public's demand for currency and time
deposits and the banks' demand for reserve balances.
The public's desired portfolio of time deposits depends foremost
on the interest rate offered on time accounts, the interest rate available on savings accounts with mutual savings banks, the interest rate
on the share accounts at savings and loans associations, and the bill
rate. The service charges on checking deposits quite likely affect
the public's desired allocation of wealth to time deposits also. I n
addition, this allocation depends on the publics wealth in the form
of money balances and nonmoney assets.
The behavior of the public's currency balances is somewhat more
difficult to explain. But the costs and yields associated with specific
financial assets, especially the service charges on checking deposits,
and wealth—held in the form of money balance and nonmoney assets—appear to shape the public's currency demand. Both demand
for currency and time deposits may be usefully partitioned into two
components. One component depends on monetary wealth (i.e.,
money stock plus time deposits), the other is dependent on nonmoney
wealth and a spectrum of costs and yields associated with the range
of financial assets under consideration.
The banks' demand for reserves can be similarly partitioned into
a component depending on deposit liabilities and a component depending on prevailing market interest rates and the discount rate.
The first component covers the volume of required reserves that banks
must hold under the reserve requirements imposed by the Federal
Reserve authorities. Given these requirement ratios, a bank's required raserves depend completely on its time and net demand deposits.
But a bank will usually hold reserves beyond the volume of required
reserves. Excess reserves are "excess" only in a strictly legal sense.
I t is dangerously misleading to use this legal notion to interpret the
actual behavior of banks. Although large banks may at times hold
few reserves in excess of requirements, the mass of member banks
generally hold assets in the form of excess reserves. In the thirties
and forties banks held more than $1 billion in this form for many
years.
Analysis of the banks' reserve adjustment process reveals both
a systematic demand for excess reserves resulting from the adjustment process outlined in a previous chapter, and the influences shaping the banks' desired excess reserve balances. These desired reserve
balances depend partly on the banks' deposit liabilities. But foremost they seem to depend on short-term open market rates, the discount rate, and some other factors to be specified shortly.
During any particular settlement period, a bank continuously adjusts its portfolio under conditions of uncertainty with respect to
the net balance of currency flows to the public, the net balance of
claims against other banks arising from checks drawn by its depositors,
and the net balance of time deposit inflows. The daily fluctuation
4
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
in the bank's reserve position can be very large relative to the average
reserve balance held during the settlement period. B u t pervasive
uncertainty surrounding a bank's flow of reserve funds does not mean
that the bank is without relevant information. Bankers acquire
"experience" which permits them to judge the nature of the observed
variations in their reserve positions and to attempt to separate systematic from random events.
The concept of a probability distribution may be used to approximate the judgments and "experience" on which bankers rely. The
spread of the distribution indicates the variability of the reserve position to be expected on the basis of the banker's past experience. Relatively large positive or negative changes are most unlikely, i.e., have
relatively small probabilities in the absence of a systematic change
in underlying market conditions or the bank's portfolio. The center of
the probability distribution indicates the amount of reserves that the
banker may expect to hold. This is the average reserve position around
which deviations in reserves will occur. The anticipated average
reserve position and the anticipated variability are determined primarily by past observations about the credit market, the behavior of
the Federal Reserve authorities, and a bank's reserve flows. The average and the variability around the average frequently provide an
adequate description of the uncertain environment in which the banker
operates.
The wide spread of the probability distribution reflected by the
substantial variability of reserve positions, reflects the peculiar uncertainty facing a bank's decision makers. There is always some likelihood that any particular portfolio decision will be accompanied by a
loss of reserves. Of course, the probability with which losses of different size occur depends on the portfolio decisions made. At any
particular reserve position, there will be portfolio decisions that would
rnncler reserve losses very likely, while other decisions could lower the
probability of reserve losses to negligible proportions.
Within the context of uncertain prospects, characterized by a given
probability distribution, expressed by a specific anticipated average
and expected variability, a definite association emerges between a
bank's portfolio adjustments and the likelihood of a reserve deficiency; that ; s, a volume of reserves on the settlement day below the
required level. Reserve deficiencies involve specific costs to a bank.
These costs are both of a pecuniary and nonpecuniary nature. In
order to understand the banks' behavior with respect to their reserve
positions, the costs of reserve deficiencies require some further attention. These conditions may be usefully traced bv considering the
courses available to a bank with a reserve deficiency at the end of a
settlement period.
If the reserve deficiency does not exceed 2 percent of required reserves, a bank may carry over the deficiency to the next settlement
period. This carryover imposes additional constraints on the bank's
portfolio adjustments in the following period. Such constraints involve costs of both a subjective and a pecuniary type. A bank may
instead decide to cover the deficiency by borrowing Federal funds
from other commercial banks or from its Federal Reserve bank. If
the deficiency exceeds the critical 2-percent limit and is judged to be
only temporary, the bank will usually borrow from one of these
sources.
5
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
The interest on borrowing and the constraint on subsequent portfolio decisions do not form the only costs associated with potential
reserve deficiencies. Such deficiencies give rise to additional costs
to a bank, particularly the allocation of valuable resources, the time
of skilled and expensive labor, to the arrangement and implementation of the technicalities involved in the borrowing transaction and
the subsequent readjustment of the short end of portfolios to repay
the borrowed funds and to assure an adequate volume of reserves.
Costs also arise if a bank obtains Federal funds, i.e., reserves, by
withdrawing correspondent balances at other commercial banks. The
costs incurred in this manner very likely depend on the relative frequency with which a bank pursues this course and the relative size of
its correspondent balances. Reserve deficiencies thus generate specific costs which depend on size, frequency, and duration of the
deficiencies.
The previous discussion indicated that a bank can control the
expected costs generated by potential reserve losses through the choice
of its portfolio. Arranging portfolios to hold excess reserves is one
means of lowering these costs. The larger the volume of excess reserves held by a bank, the smaller the probability of reserve deficiencies
and thus the smaller the costs (on the average) associated with such
deficiencies. But excess reserves, while lowering one type of cost to
banks, increase another type of cost. Allocation of assets to excess
reserves involves a loss of revenues in proportion to the interest rate
on earning assets that would have been acquired. If assets have been
allocated to Treasury bills, for example, the bank would have obtained
revenues equal to the amount of reserves multiplied by the prevailing
bill rate. Revenues on excess reserves are zero; an allocation to excess
reserves is a sacrifice of potential income. If no cost were associated
with holding excess reserves, all banks would find it most profitable
to hold excess reserves at a level sufficiently large to render the occurrence of a reserve deficiency totally improbable. But we frequently
observe reserve deficiencies on settlement days, thus confirming the
assertion that holding excess reserves is not without very specific costs
to a bank.
The optimal volume of excess reserves emerges from the cost conditions associated with the holding of excess reserves and the cost conditions characteristic of reserve deficiencies. The two types of cost
operate to influence a bank's optimal excess reserve balance in opposite
directions. The marginal cost of excess reserve holding is closely dependent on short-term rates on the open market. The marginal cost
of reserve deficiencies depends on size, frequency, and duration of such
deficiencies and the nature of a bank's "production function," that is,
the technology associated with the production of banking services.
These technological aspects very probably explain major differences
in the marginal cost of reserve deficiencies between different types of
banks. A l a r g e Reserve city bank most likely has much lower marginal costs of reserve deficiencies than medium or small banks, particularly country banks. No matching differential seems to occur in
the marginal costs of excess reserve holding. I t follows from optimizing behavior that country banks and smaller banks will desire to
hold a larger volume of excess reserves per dollar of deposit liabilities.
The optimal excess reserves that a bank desires to hold on the average thus emerges from the interaction of the two types of cost, the
34—474—64
2
6
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
costs of holding excess reserves and the costs generated by reserve
deficiencies. If a bank attempts to minimize cost, its desired excess
reserve balance depends on the volume and structure of its deposit
liabilities, interest rates on the open market, technological characteristics shaping the operational costs of its reserve adjustment processes,
and s o m e characteristics (anticipated average and variability) expressing the nature of the uncertain prospects bearing on the flow of currency and the net withdrawal or accrual of deposits. A bank's total
demand for reserves thus depends on the factors just listed and on the
requirement ratios imposed by the Federal Reserve.
Apart from mergers and substantial changes in institutional arrangements prohibiting or limiting branches, we deem it unlikely
that technological aspects significantly modify the marginal costs
of reserve deficiencies in the shorter term. We expect such modifications to work their mass eiFects quite slowly and gradually. The
other factors listed exert, on the other hand, a decisive short-run
effect on the banks' desired volume of excess reserves. In particular,
variations in short-term rates and the discount rate substantially
alter the banks' desired volume of excess reserves. Ranks tend to
hold considerably larger excess reserves in periods with low open
market, rates than in periods of comparatively high rates.
A major portion of the difference observed between excess reserves
in the 1930's and 1950's can be attributed to the difference between
open market rates prevailing in the two periods. Still, the large
difference in market rates does not completely explain the difference
in the behavior of excess reserves in the two periods. The probability distribution summarizing a bank's uncertain prospects with respect to the flow of reserve funds seems to he of importance in explaining the difference also. The shocks exerted bv the accumulating bank
failures, and the accelerated conversions oi deposits into currency
experienced after the late fall of 1930, raised both the expected average and variability of reserve flows.
The failure of the Federal Reserve to offset the currency d r a i n or
to stem the rise in bank failures substantially raised the probability
of given reserve losses and thus contributed to raise the marginal
costs expected from potential reserve deficiencies. This change in
costs alone, independent of accompanying movements of market
rates, would have induced banks to hold substantially larger excess
reserves.
The experiences of the early thirties, combined with the events
observed at financial crises occurring before the operation of the
Federal Reserve System, s t r o n d v suggest that the anticipated average and variability of reserve flows immediately respond to accelerating bank failures and unexpected large deposit withdrawals. The
same observations also suggest that, once the initial shock has passed,
its eftects wear off only gradually and are distributed over n considerable period. The anticipated average and variability of reserve
flows appear to adjust, only slowly to a new set of comparatively
stable circumstances. The residual effects of the terrible shocks,
shattering the U.S. banking system from late 1030 to early 1033,
seem to have persisted well into the subsequent war period. The relatively low volume of excess reserves observed during the 1950's is
thus the result of comparatively higher interest rates and the elimi
7 ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
nation of the shock effect on the anticipated average and variability
of reserve flows.2
T I I E T W O ' M A J O R ' M E C H A N I S M S C O N S T I T D T I N G T H E M O N E Y S U P P L Y PROCESS
The basic behavior elements required to understand the operation
of the money supply process have been introduced in the last section.
We presented for this purpose the public's demand for currency and
the public's demand for time deposits. We also emphasized the dependence of the public's wealth allocation to currency and time deposits on a spectrum of costs, yields, money and nonmoney wealth.
Furthermore, the banks' behavior with respect to excess reserves
was summarized by the concept of demand for excess reserves or
'desired7' excess reserves. A detailed analysis of the banks' reserve
adjustment process was outlined. This analysis demonstrated the
emergence of a demand for excess reserves dependent on interest
rates, deposit liabilities, and some other entities considered in the
previous section.
With the basic behavior elements at our disposal we can proceed
^'ith the description of the money supply process. This process may
he usefully visualized in terms of two interrelated mechanisms. One
mechanism describes the response of the banking system to surplus
reserves by suitable portfolio adjustments; the other describes the
process injecting surplus reserves into the system. The generation
of surplus reserves and their absorption by the banks' portfolio adjustments thus form the focal point of our description of the money
supply process.
The ni ultipIier viechan-Usm
Surplus reserves are the difference between measured excess reserves and desired excess reserves. While the volume of measured
excess reserves is a result of legal or administrative requirements,
the volume of surplus reserves is a behavioral construct reflecting
the bankers' demand for reserves. Our previous discussion and the
assumption that banks seek to maximize wealth imply that bankers
adjust their portfolios to eliminate surplus reserves. By suitable
acquisition of earning assets, positive surplus reserves are exhausted,
and by appropriate unloading, or running off, of earning assets negative surplus reserves are removed. A bank unavoidably loses surplus
reserves by acquiring earning assets. Such acquisition is necessarily
accompanied by at least one of the following events: (1) deposits
newly created by the acquisition of earning assets are converted into
currency; (2) newly created deposits are checked away to other
banks; (3) newly created deposits stay with the bank as demand or
time deposits. In the first two cases surplus reserves are lost because
of a loss in reserves to the public and other banks. In the last two
cases surplus reserves are lost because of an increase in required reserves necessitated by larger deposit liabilities. Appropriate port* Useful m a t e r i a l p e r t a i n i n g t o t h e b a n k s demand f o r excesspreserves c a n b e ' f o u n d to
unpublished d o c t o r a l thesis by George Morrison; Liquidlty P r e f e ^
B a n k s / * j u m » 1062, University of Chit&tfo. Morrison investigated the dependence of excess
reserves on i n t e r e s t r a t e s and t h e *A>f>ek of s p r e a d i n g baak f a i l u r e s and
w i t h d r a w a l s . T h e previously discussed r e s u l t s of Meigs also, support t h e assertion t h a t
excess reserves depend on i n t e r e s t rates. Additional m a t e r i a l coupled with a detailed
a n a l y s i s will be contained In o u r f o r t h c o m i n g book on money supply a n d money demand.
8
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
folio adjustments thus always enable a bank to remove surplus
reserves*
Emerging surplus reserves induce banks to respond with suitable
portfolio adjustments. Such portfolio adjustments eliminate the
surplus reserves initially experienced by the adjusting banks but do
not eliminate, immediately, the surplus r e s e r v e s for the system as a
whole. Surplus reserves are distributed to other banks. This redistribution is caused by the relative frequency with which the deposits
of a bank, and particularly the newly created deposits, are checked
away to other banks. This "spillover" to other banks is the major
limitation imposed on a single bank's portfolio adjustment in response to available surplus reserves.
The redistribution of surplus reserves over the system, accompanyr
ing the portfolio adjustments induced by the initial surplus reserves,
lowers the volume of surplus reserves available to the system. The
portfolio adjustments generated in the first round are typically associated with outflows of currency to the public. The loss in surplus reserves through the currency drain is reinforced by increases in required
reserves attributable to the portion of the newly created deposits not
converted into currency. The surplus reserves available after the first
round of portfolio adjustments is thus necessarily smaller than the
initial volume that triggered the sequence of portfolio adjustments.
I n every round banks respond to accruing surplus reserves by appropriate portfolio adjustments. These adjustments partly absorb some
of the remaining surplus reserves into currency held by the public or
into required reserves, and redistribute the remaining surplus reserves over the system. The redistributed volume triggers a further
round of portfolio adjustments with the train of consequences already
described. The chain reaction to initially emerging surplus reserves
proceeds until the initial volume of surplus reserves has been absorbed
through a series of portfolio adjustments generating spillovers into
currency or increases in required reserves. Our tentative investigations into this process suggests that it works quite rapidly and that
most of the surplus reserves that triggered the process are absorbed
within a month.
The sequence of portfolio adjustments, associated with the continuous redistribution and absorption of the surplus reserves initially
injected into the system, yields an increase m total bank earning
assets and the money supply that is a multiple of the original volume
of surplus reserves. The money stock rises in every round of the process by (1) the amount of the currency spillovers to the public, and
(2) the concomitant increase in checking deposits. Both magnitudes
mirror the public's demand for currency and time deposits. The
banks portfolio adjustments generate a matching amount of new deposits that are distributed by the public between currency, time, and
demand deposits according to the public's marginal propensity to
hold currency and time deposits with respect to money wealth. "The
response of the public's demand for currency and time deposits to
variations m money wealth ( ^ ^ m o n e ' j stock plus time deposits)
thus forms a crucial e l e m e n t ^ S l J ^ l ^ t o l i e r mechanism.
The currency s p i l l o v e r a ^ u ^ ^ ^ p u b l i c ' s systematic reaction
to larger money w e a l t h , t h e magnitude of the system's multiplier responseUff£$&
such spillovers the mufti-
9
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
plier effect of surplus reserves on the money stock would be inversely
proportional to a weighted average of reserve requirements against
demand and time deposits. Reserve requirements on time deposits
have been 0.05 until recently and the average requirement ratio
against demand deposits has been in the neighborhood of 0.15 f o r a
lengthy period in the fifties. Without the occurrence of systematic
spillovers of currency, as an element of the portfolio adjustment sequence, the monetary multiplier would have been at least 6% and
would not have exceeded 20. A multiplier substantially below six could
only occur if currency drains, reflecting the public's currency demand behavior, are typically associated with the series of portfolio
adjustments constituting the multiplier mechanism of the monetary
system.
The processes injecting surplus reserves
We discussed the system's multiplier mechanism; i.e., its response to
an initial volume of surplus reserves, without questioning how surplus
reserves emerge. Our description of the money supply process must
be extended at this point to describe the wav in which surplus reserves
are injected into the system. The eventual combination of the multiSlier mechanism with the injection mechanism yields a complete
escription of the money supply process.
The joint operation of the two mechanisms provides information
about the response of the money stock to open-market operations, to
variations in reserve requirements and the discount rate. The injection mechanism consists of six distinct processes which operate more
or less independently to create or absorb surplus reserves. These
processes are independent of the portfolio adjustments that constitute the multiplier mechanism. They center on (1) variations in the
monetary base, (2) variations in requirement ratios and the redistribution of already existing deposits among banks with different requirement ratios, (3 and 4) changes in the public's demand for
currency and time deposits that are independent of modifications in
money wealth and consequently independent of the bank's j>ortfolio
adjustments, (5) variations in the system's interbank deposit structure, and (6) changes in the banks' desired excess reserve position
attributable to variations in interest rates. The subsequent sections
develop the characteristic nature of these processes.
The monetary base
I t is most useful to visualize new injections of base money as operating initially through bank reserves. Given the volume of required reserves and the existing demand for excess reserves, injections of base
money increase the amount of available surplus reserves. Variations
in the base thus constitute a channel through which surplus reserves
are injected into the system with the consequences traced in our description of the multiplier mechanism. The latter mechanism distributes the new base money between cash assets held by banks and
currency held by the public. Cash assets of banks, in particular bank
reserves, thus emerge jointly with the money supply from the operation of the multiplier and injection mechanisms. The base is thus
allocated by these processes between the banks and the public. I t follows that the base can be measured as the sum of currency held by the
public, bank reserves and vault cash- But this summation does not
10
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
reveal the mechanism generating the monetary base. To investigate
the latter channel in more detail, we focus on the sources of the monetary base, the magnitudes and processes that determine the volume of
base money issued.
.
.
Every issue of the Federal Reserve Bulletin carries a table with the
title "Member Bank Reserves, Federal Reserve Bank Credit, and
Related Items'- at the beginning of the statistical section of the Bulletin. The table includes all the principal asset and liability accounts;
from the consolidated balance sheets of the 12 Federal Reserve banks.
I n addition, liabilities of the Treasury that are money—currency and
coin issued by the Treasury—are included with the Federal Reserve
accounts. The table, as presented in the Bulletin, is divided into
"Factors Supplying Reserve Funds" and "Facton; Absorbing Reserve Funds." By a simple rearrangement, of the items that appear
in the table, we can construct a statement of sources and uses of reserves plus currency from the statement of sources and uses of reserves that appears in the table. The sources and uses of reserves
plus currency will be referred to as the monetary base, or more simply
the base.
To obtain the uses of the base, we add the principal liability accounts on the Federal Reserve balance sheet—member bank reserves
plus note issue—to the principal monetary liability of the Treasury,
Treasury currency outstanding. The sum of these three items is equal
to Federal Reserve and Treasury coin and currency held by the public,
currency and coin held by member banks, plus member bank deposits
at Federal Reserve banks. These three items are the total uses of the
base. Any increase or decrease in the base must affect one of these
items. Actually, our procedure is more restrictive, as we noted above.
Injections of base money initially modify available reserves and generate surplus reserves. The process set in motion distributes the new
injections between bank reserves and currency held by the public.
The sources of the base, that is its determinants, combine the remaining factors supplying and absorbing reserves that appear in the table
in the Bulletin to which reference was made above. The first source
of the base, reserve bank credit outstanding, is divided into three component parts: (1) U.S. Government securities held by the Reserve
banks, (2) discounts and advances, and (3) float.
(1) U.S. Government securities held by the Reserve banks: Increases or decreases in this account reflect open market operations,
repurchase agreements, and Federal Reserve decision* to replace
maturing securities retired by the Treasury. This item can be com^
pletely controlled by the Federal Reserve'authorities through their
open market operations. By purchasing or selling securities in the
market, the Federal Reserve can reinforce or counteract any other
change that affects the monetary base.
(2) Discounts"and advances'show changes in borrowing bv m e m b e r
banks. An increase in member bank borrowing increases the base and
a decrease contracts the base. Unlike the Federal Reserve's Government security portfolio, this item is not completely controlled by the
Reserve banks. Indirect control is exercised through changes in the
discount rate, but the banks' demand f o r borrowed reserves depends on
market rates as well as the discount rate. Moreover, borrowing is a
privilege and not a right under present arrangements. Although the
11
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
Federal Reserve banks rarely refuse to grant the privilege, they may
put pressure on individual banks to retire indebtedness, thereby inducing other banks to borrow as noted earlier. Hence the total volume of
borrowing is not completely determined by the Reserve banks. But
reserves supplied by borrowing can be offset by compensating sales in
the open market that reduce the Government security account.
(3) Float is an interest-free loan from the Reserve banks to the
member banks. I t is neither a necessary consequence of monetary
policy nor an unavoidable feature of monetary arrangements. I t
occurs because of the fixed time schedule that the Federal Reserve has
adopted for payment of checks cleared through the Reserve banks.
When a check is presented for payment, the Federal Reserve gives
"deferred availability credit" to the collecting bank. The deferred
credit is later replaced by a credit to the bank's reserve account. If
the check is not completely processed within the fixed time schedule,
the bank presenting the check for collection receives an increase in
reserves at a Federal Reserve bank before the reserve position of the
paying bank has been charged with the check. The difference between
these "uncollected items" and "deferred availability credit" is float.
If the Federal Reserve should decide to reduce the time schedule
underlying deferred availability credit without any compensating
change in the technical means of collecting checks, float would increase.
For example, if all checks were credited immediately, deferred availability credit would be abolished and float would equal the volume of
checks in process of collection. In effect, the Federal Reserve would
be making interest-free loans to banks up to the amount of checks not
yet collected. Conversely, float could be abolished by making the
credit and debit to reserve accounts coincide with the actual collection
period. If this were done, the uncertainty underlying the process
generating float would not be reflected in the shortrun variations of the
base.
To the sum of these three items that compose reserve bank credit we
add the gold stock. The next item in the table, "Treasury currency
outstanding," is the value of the currency and coin issued by the Treasury. I t measures the Treasury's direct contribution to the money supply. This item appears in the table as a factor absorbing reserves. Our
analysis in terms of the base includes total currency held by banks
and the public as a part of the base rather than as a "factor absorbing
reserves." An increase in the outstanding volume of Treasury currency, whether used by banks for till money (now counted as part of the
reserves of member banks) or held by the public, expands the base.
Treasury currency is therefore a positive source of the base.
Total currency in circulation is the next column heading. Included
in the value of this item are all the currency holdings of the banks and
the public whether issued by the Treasury of the Federal Reserve
banks. We pass over this item since it has been incorporated as a use
of the base above. I t is a part of the magnitude to be explained by the
sources or determinants considered in this section.
The remaining column heads designate additional factors absorbing
reserves. I n the statement describing the sources of the monetary base,
these items are treated as negative sources, that is, as a reduction in
the amount of the base supplied by total reserve bank credit, gold stock,
and Treasury currency outstanding. We consider each briefly. The
12
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
first, Treasury cash holdings, is of minor importance (except for the
period following the devaluation in 1934), both because of its small size
and its small monthly and annual variation. I t is a technical item that
arises because of accounting procedures related principally to gold
and need not concern us here. The second item, deposits other than
member bank reserves with Federal Reserve banks, includes three
liabilities from the consolidated Federal Reserve balance sheet. One
of these is the highly volatile Treasury cash balance at the Reserve
banks. To reduce the impact of its operations on the banking system, the Treasury adheres to the practice of depositing at commercial
banks the checks that it receives and writing checks against balances at
the Reserve banks. Balances are periodically transferred from the
commercial banks to the Reserve banks to increase the relatively small
Treasury balance prior to drawing checks. This process withdraws
reserves from member banks temporarily. When checks are written
against and collected from the Treasury balance at the Federal Reserve, the reserves are restored to the banking system. The generation and absorption of reserves caused by movement of the Treasury
balance from commercial banks to the Reserve banks had a large
and important influence on the monetary base in the early postwar
years.
The second and third components of the item "Deposits other
than member bank reserves with Federal Reserve banks" are principally the deposits of foreign central banks and the# deposits
of commercial, nonmember banks. Each of these is treated in a manner analogous to the Treasury balance; i.e., as a negative source of reserves. Reductions in these Federal Reserves liabilities expand the
monetary base; increases in these items reduce the base. Thus a
transfer of deposits from member banks to nonmember banks, accompanied by a shift from member bank reserves to deposits of nonmember banks at the Federal Reserve, reduces both the sources and
uses and hence reduces the monetary base. The exchange of foreign
central bank deposits at the Federal Reserve banks for gold and the
withdrawal of gold from the United States has no direct effect on the
monetary base. But an exchange of foreign deposits at c o m m e r c i a l
banks for gold and a withdrawal of gold reduces both the sources and
uses sides and hence reduces the monetary base.
The final item in the table, "Other Federal Reserve accounts," is the
sum of all other Federal Reserve liabilities plus net worth minus all
other Federal Reserve assets. This small magnitude is a balancing
item that assures that assets equal liabilities plus net worth or, in the
case at hand, that total sources equal total uses of the base.
TABLE VI-1,—Sources and uses of the monetary
base
Sources
Uses
Total Reserve bank credit, including Government
~ securities, discounts and advances* plus float.
P l u s gold stock,
_
Plus Treasury currency outstanding
Minus Treasury cash holdings
M i n u s deposits, other than member bank reserves, a t
Federal Reserve banks.
M i n u s other Federal Reserve accounts.
Equals total sources
_
Member bank reserves with Federal Reserve
banks.
Plus member bank reserves in currency and coin.
Plus currency held b y the public.
Equals total uses.
13
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
We can recapitulate the above discussion by presenting a table summarizing the sources and uses of the base. Table V l - 1 shows the
rearrangement of the Federal Reserve's table "Member Rank Reserves * *
into the monetary base.
Relative size and variation of the sources of the base
The source components of the base exhibit neither a uniform relative size nor a variability of similar order. From the beginning of
the Federal Reserve System both gold stock and Reserve bank credit
have dominated the base. But changes occurred over the decades in
the composition of Reserve bank credit. Until the early thirties, discounts and advances were a significant proportion of Reserve bank
credit. Thereafter. Reserve bank credit almost coincided with the
portfolio of securities until the early fifties. Around the start of the
first postwar decade requests for discounts and advances at Federal
Reserve banks accelerated and this particular source contributed once
again to the base. The volume of Reserve bank credit remains, however, completely dominated by the portfolio of Government securities.
Some information about the comparative variability of the major
source components is collected in table VI-2. To construct this table
the percentage change of the base for each month relative to the
corresponding month in the preceding year was computed beginning
^Jth January 1919 and terminating with December 1962. The total
into four periods,
per
tC?le-WaS
distinguished according to
rows in the table. The annual percentageTchange of the base
ecorded for each month was partitioned into the portions contributed
thA T? i
Federal Reserve's portfolio of Government securities, (2)
Reserve>s
fltol/fi
portfolio of discounts and advances, (3) the
toc
? *i (4) Treasury currency issued, and (5) the sum of the regaining items listed on the sources side of table VI-1. Maximal and
chft i m • v a ' l l e s W e r e t ' i e R selected from each period for the percentage
ange m the base and, similarly, extreme values were selected from
ne contributions made by the five major source components to the
percentage change in the base.
m taI)
ran
le V*- 2 exhibit the extreme values defining the
nge of variation in the growth rate of the base and the extreme
aiues of the contributions made by the various sources. An inspec°n of the table indicates that the range of variation of the base,
jneasured by the distance between maximal and minimal percentage
nange per annum was almost the same for the first three periods
aistmginshed. We note that the growth rate of the base was as high
loio P e r c e n t P e r annum and fell as low as minus 15 percent between
and 1929, a range of 32 percent. During the thirties and forties
. is range was maintained. The size of the variations is remarkable
indeed.
Even more remarkable is the occurrence of negative growth rates,
such growth rates reflect a deflationary posture of the Federal Reserve authorities. We contend that there is little reason to lower the
jnoney supply, even in the aftermath of an inflationary episode; e.g.,
wie period 1941-46. Once the inflationaiy injections have been made,
nothing is gained by the subsequent imposition of a deflationary
monetary p^icy. Termination of the inflationary process can be assured by retrenching the growth rate of the money supply and can be
achieved by suitable adjustments in the growth rate of the base,
such adjustments rarely require a negative growth rate in the base.
14
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
The deflationary posture assumed by the Federal Reserve authorities in the thirties and revealed by the minimal growth rate of the
base (minus 5.7 percent) was even less appropriate, in the light of
prevailing circumstances, than the deflationary policies pursued after
1918 and again after 1945. Negative growth rates of the base occurred
again in the past decade. A later section will indicate, however, that
these reductions in the base were typically the result of "compensatory
open-market operations" designed, in principle, to modify the sudden
impact of changes in reserve requirements and to redistribute their
effect over time. More important is the compression of the range of
variation, recently exhibited by the growth rate of the base, relative
to the range observed in all the previous decades. The exaggerated
variability has been conspicuously lessened, and fluctuations in the
growth rate of the base have been reduced by more than 50 percent.
An inspection of the remaining columns in table Y I - 2 yields information concerning the major sources of the variability of the base.
I t is quite apparent that from 1919 to 1929 the variations of discounts
and advances dominated the variability of the base. At their maximal
rate, discounts and advances contributed to an increase of 23.9 percent
in the base and at their minimal rate they generated, by themselves,
a decline of 25.1 percent in the base per annum. The close association
between the money supply and the base determines a similar variability in the money supply attributable to the gyrations in discounts and
advances. I t follows that the Federal Reserve's mode of operating
the discount window essentially introduced a systematically destabilizing feature into our monetary framework. The variability of the
gold stock was substantially smaller during the 1920'S than the large
swings in the base attributable to discounts and advances. The variations in the portfolio of securities appear to be even more moderate in
comparison, and Treasury currency and other items had only minor
significance.
of variation of annual pcrcent
contributions made bit major source
TABLE V I - 2 . — R a n g e
change of base and of the
components
[In pcrcent]
Period
January 1919 to December 1929:
Maximum
Minimum.
January 1930 to December 1940:
Maximum
Minimum
January 1941 to December 1950:
Maximum
_
Minimum. ____
January 1951 to December 1962:
Maximum
Minimum
Portfolio
of Gov- Discounts
ernment and adBase (B) securities vances Gold
(including float)
(Fi)
Treasury
(A) currency
(C)
+16.6
-15.5
+7.7
+23.9
-25.1
+10.9
-7.1
+2.4
-3.7
+26.6
+18.2
-2.0
+9.5
-12.7
+52.1
-14.6
+3.4
-.5
-39.8
+23.1
-9.3
+28.1
+1.4
+9.8
+13.6
-5.7
—3.2
—11-6
-2.6
- 1 0 0 + ^ £ 100+M
+3.1
-4.6
+21.9
-1.0
—4.5
+3.1
-.002
+U
—7.6
+2.4
-2.5
+3.4
-5.8
+•>4
-1.5
+2.0
-.06
Remarks: The table is based on tbe following formula-
Other
items
on
g . , 0 0 - ^ f • 100
15
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
Iii tire second period, the thirties, gold moved to the center of the
stage. Variability of both gold stock: and "other items'* dominated
the base; and the changes of "other items" were due principally to
the behavior of the portion of gold included as part of Treasury cash.
Still, the dominant role assumed by the behavior of gold should not
veil the fact that, the Federal Reserve's portfolio of securities and discounts exhibited variations of a similar order. The movements in the
base attributable to Treasury currency were again comparatively
insubstantial.
The next period, 1941-50, saw another shift in the role of particular
sources in the movement of the base. The Federal Reserve's portfolio
of Government securities became the central actor, closely followed
by the gold stock. The contribution of discounts and advance almost
vanished. Even the "other items" showed a larger range of variations.
The Federal Reserve's portfolio of securities and gold also dominated the variations of the base in the last decade. Discounts again
varied over a smaller range. Treasury currency and other items again
exhibited the least variation.
The four decades of Federal Reserve operation can be summarized
in three points: (1) In most of the periods, the variations in the base
assume proportions difficult to justify in terms of a rational monetary
policy; (2) the minimal growth rates exhibited by the base mirror a
deflationary posture not necessitated by the circumstances prevailing
in the periods; and (3) the gyrations of gold and the variability
of Reserve bank credit dominated the variations in the annual growth
rate of the base in all periods.
Table VI-2 supplies information concerning the range of variability associated with the major source components of the base. Table
supplements this information for the postwar period. The interval from October 1949, the first lower .turning point of the postwar
period, to December 1963 is divided into seven half cycles. Each half
cycle, other than the last, represents either a complete expansion phase
or a complete recession phase. The average change between adjacent
months of all the major source components of the base was computed
for each period indicated in the table. The table thus reveals both the
average relative importance of the various source components and the
variability of this relative importance for different half cycles. A
cursory glance at the table suffices to establish the comparatively small
significance of the last two components specified: namely Treasury
currency outstanding 0, and the residual agglomeration TF (Treasury
deposits plus foreign deposits, plus Treasury cash, plus other deposits,
plus other Federal Reserve accounts). The major contribution to the
changes in the base emanate from the Federal Reserve's portfolio of
securities, F1, discounts and advances F\ or the gold stock A. But the
table also reveals the shifting importance of these source components.
1
V a r i a t i o n s in F dominated the changes in the base during the first two
half cycles, with gold clearly assuming a position of secondary
importance.
The next two half cycles exhibit a radically different pattern. Variations of the Federal Reserve's discounts and advances and changes in
tf* gold stock dominated the changes in the base. The negligible
changes in F1 reveal that open-market operations contributed almost
nothing to the longer run changes in the base over the interval from
August 1954 to April 1958. The pattern was once more reversed in
16
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
the last two half evcles and the recent, incomplete upswing phase.
Changes in the gold stock and the Federal Reserve's portfolio of securities emerged again with decisive importance, while the discounts
and advances slipped into a range of minor significance. The reader
should also note the remarkable cyclic behavior of the Federal Reserve's discounts and advances. They grow during expansionary
phases and decline in recession phases.
The information presented in table V I - 3 may be further summarized. Table V I - 4 specifies the maximal and minimal average growth
rate per month of the five source components considered.
Table V I - 4 clearly reveals the relative order of variability—the
Federal Reserve's portfolio of securities shows the largest range of
variation. The changes in the gold stock show the next largest range
of variation, and the Federal Reserve's discounts and advances the
smallest range of the three. The remaining two components; i.e.,
Treasury currency and the host summarized by IF, generate comparatively minor perturbations.
TABLE VI-3.—Average monthly changes of source components of the
base between adjacent months during half cycles
monetary
[In millions of dollars]
Half cycle
October 1949 to July 1953
July 1953 to August 1954
August 1954 to July 1957
July 1957 to April 1958
April 1958 to May 1960
M a y 1960 to February 1961
February 1961 to December 1963-
in
171.0
-80.8
-2.7
.3
92.9
124.7
242.6
A
6.5
-16.8
20.2
-87.4
14.8
-39.5
6.2
TABLE "VI-4.—Extreme values of overage monthly
W
C
-49.6
-38.1
21.5
-51.0
-112.4
-217.1
-54.1
5.8
8.2
4.2
9.2
6.3
6.1
5.2
changes in postwar
5.2
-16.2
-10.2
-3.6
-20.6
19.2
7.9
half cycles
[In millions of dollars]
Maximal value
Minimal value
pi
pt
242.6
—80.8
20.2
—87.4
A
21.5
-217.1
W
C
9.2
4.2
19.2
-20.6-
Reserve requirements
The institution of reserve requirements for member banks (extended by State regulation to nonmember banks) contributes in two
distinct ways to generate or absorb surplus reserves. Until the middle
thirties the Federal Reserve authorities had no power to change the
prevailing requirement ratios. These ratios were fixed at a specific
level for each class of member banks. The volume of required reserves
for member banks could be computed as an average of the unchanging
reserve requirement ratios weighted by the volume of deposits to which
the particular requirement ratio was applicable. A shift in the distribution of deposits among banks in different classifications modified
the volume of required reserves. A redistribution of deposits from
banks with higher requirements to banks with lower requirements
created surplus reserves for the system similar to the injections accom-
17
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
plished by variations in the base. On the other hand, a redistribution
of already existing deposits from banks with lower to banks with
higher requirements, absorbed surplus reserves from the system and
induced a contraction of bank portfolios. Variations in the distribution of deposits among classes of banks thus altered the volume of surplus reserves because of the differences in the applicable requirement
ratio. As we have indicated, changes in the volume of surplus reserves
alter the current response of the monetary system, mirrored in the
behavior of money stock and the system's portfolio of earning assets.
Uniform reserve requirements would, of course, completely remove
any connection between the redistribution of deposits and the money
supply.
When the Federal Reserve authorities acquired the power to change
the level of reserve requirements against both demand and time deposits, an additional source of variation in surplus reserves was introduced. Fiat changes of the legal requirement ratios announced by the
Board of Governors immediately subtract or inject surplus reserves
for the system. The banks respond in the manner described in our
discussion of the multiplier mechanism. Changes in requirement
ratios do not modify the volume of total reserves, but thev do modify
the volume of excess reserves supplied to the system relative to the
desired excess reserve position. A change in the requirement ratios,
Per se, exerts at most only a marginal effect on desired positions.
However, excess reserves are immediately released or absorbed. Consequently the supply of excess reserves shifts relative to the banks7
demand for excess reserves: surplus reserves are created or destroyed.
The effect on surplus reserves of both the redistribution of deposits
among classes of banks and fiat changes in legal ratios can be summarised in terms of variations of the average requirement ratios on
demand and time deposits. A weighted average has been constructed
for this pun>ose. The legal ratios applicable to a class of deposits
are weighted by the volume of deposits in that classification. Because
the ratio on all member banks' time deposits has rarely differed between classes of member banks, we obtain an average time deposit
ratio dependent on the uniform requirement ratio imposed on member
banks and the ratios assigned to nonmember banks. These ratios are
Weighted by the relative magnitude of total time deposits held by
member and nonmember banks. The average requirement ratio on
demand deposits depends essentially on the legal ratios characterizing
member bank classifications and the requirements imposed on nonmember banks. The relative magnitude of demand deposits is used
as the weight for each classification.
.
t To summarize, changes in the average requirement ratios may occur
because of changes in the values of one or more of the requirement
ratios or because of shifts in deposits from one class of banks (or dePosits) to another. The first cause of changes m the average^fleets
Policy action by the Federal Reserve authorities (or the State banking
agencies); the s e c o n d mirrors a multitude of underlying factors
Raping the distribution of existing deposits
different requirement ratios. Both causes have been generally
acknowledged as influences on the average requirement ratios.
18
A L T E R N A T I V E APPROACH. TO T H E M O N E T A R Y M E C H A N I S M
Some writers, including spokesmen for the Reserve System*
assign a level of importance to the redistribution of deposits that
cannot be substantiated.. There is little doubt that volatile redistri1
bution of deposits among classes of banks could seriously impair the
degree of control exercised by the monetary authorities over the
money supply. Of course, such impairment could be removed even
in the face of severe instability of distributional pattern, for example,
by imposing uniform requirement ratios on all commercial banks.
Investigation of the variations generated in the average requirement
ratio on demand deposits attributable to a shifting distribution of
existing deposits, however, yields no support for the contention that
volatile shifts in deposit distribution actually impairs the degree of
control over the money stock. Some results of this investigation are
collected in table V I - 5 for periods during which the r e q u i r e m e n t ratios
for each classification remained unchanged. F o r each month in the
period we computed the change in required reserves, relative to the
same month in the preceding year, that was strictly attributable to the
redistribution of already existing deposits. The table presents
averages and extreme values of these changes for every period
selected.
We note that the average value of required reserves released via
deposit redistribution has been declining over the postwar period.
With a monetary multiplier of approximately 2.5 the average increment in the money stock per annum, induced solely by deposit redistribution in the context of differential requirement ratios, reached 175
million in the early postwar period and fell to 87 million in the early
sixties. Deposit redistribution thus contributed approximately 0.16
percent to the growth in the money stock immediately after the war
and only about 0.06 percent in the early sixties. These p e r c e n t a g e s
are of a magnitude that does not exceed the effect of random influences operating on the money supply process. Moreover, a caref u l inspection of the sequences of monthly values reveals the
occurrence of a clearly marked cycle in the operation of the deposit
redistribution process' The values of required reserves released do not
oscillate at random between the extreme values recorded in the table.
They move in a regular pattern exhibiting sharply separated
upswings and downswings. The random component of the process
is thus only a fraction of the magnitudes in the table, and the increment in the money stock attributable to this random component in
the deposit redistribution process can reasonably be expected to be a
fraction of the 0.16 and 0.06 percent mentioned previously. Only
a comparatively large random component could potentially impair the degree of control over the money stock. Since our i n v e s t i gation indicates that the operation of the random component has no
relevant effect on the money supply p r o c e s s , it follows that the removal of differential requirement ratios cannot be justified in terms of
the degree of control over the money supply. But elimination of this
argument offers no justification to continue differential requirements.
The discussion will be resumed in the final section.
ALTERNATIVE APPROACH TO T H E MONETARY M E C H A N I S M
19
TABLEI V U — R e l e a s e
months
between
of required
reserves
occurring
between
corresponding
of adjacent
years, attributable
to the redistribution
of demand
deposits
different
classifications
of member
banks
Period
Average of
monthly
values (in
millions of
dollars)
June 1945 to January 1948
69
February 1952 to May 1953...
68
August 1955 to February 195S.
50
May 1959 to September 19G2..
35
Average of absolute values
43
Range of.
monthly
values (in
millions of
dollars)
1141
* 12
i 123
1
104
3 14
i 141
>-111
1
a
Maximum.
Minimum.
^Remarks: The underlying data were obtained from the computation of the ^magnitude described in
K. Brunner, "A Schema for the Supply Theory of Money 7 " International Economic Review, January 1961.
Changes in the requirement ratios not only release—or absorb—•
surplus reserves, they also modify the magnitude of the multiplier
response expressed by the size of the monetary multiplier. This
multiplier, revealing the banking system's power to magnify injections of surplus reserves, depends partly on the level of the requirement ratios. An increase in requirement ratios tends to compress,
and a reduction to raise, the magnitude of the monetary multiplier.
Changes in legal requirement ratios are traditionally acknowledged
to exert an effect on the money stock via two essentially distinct
channels. One channel transmits the impact of surplus reserves released ; the second channel modifies the magnitude of the multiplier
response. Our preliminary studies of the money supply process indicate the dominant operation of the first channel—the change in
surplus reserves. The effect on the current response in the money
stock transmitted via the second channel is of comparatively small
order and may be neglected without serious error. For example,
a reduction in reserve requirements by 1 percentage point can be
expected to raise the multiplier by approximately 2.5 percent of
its existing magnitude, i.e., from 2.50 to 2.56. Exaggerated views
concerning the impact of requirement changes on the system's magnifying power appear to have resulted from a prevalent Federal
Reserve attitude that equates the money multiplier with the reciprocal
of the System's average requirement ratios. According to the latter
notion the given percentage change in the average requirement
ratio modifies the multiplier by an equal percentage change in the
opposite direction. Our investigation indicates that the presumed
equality of the twTo changes overstates the effect on the multiplier of a
change in the requirement ratios.
A crucial and frequently overlooked fact explains the comparatively
attenuated sensitivity of "the multiplier to requirement changes, viz,
the multiplier depends only partly on the legal requirement ratios. I t
also depends on the public's marginal propensity to hold currency
20
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
relative to monetary wealth, the public's marginal propensity to hold
time deposits relative to monetary wealth, the banks' marginal propensity to hold excess reserves relative to their deposit liabilities, and
some institutional features expressing the system's degree of centralization and the interbank deposit structure. T h e first factor assumes
a particular significance in this context. I t s operation generates the
currency drains typically associated with the System's portfolio adjustments. According to our analysis such currency drains form the
single most important reason why the change in the multiplier resulting f r o m changes in the requirement ratios is much smaller than is
typically asserted in Federal Reserve publications or the standard textbooks on money and banking. Moreover, the currency drain and the
other factors listed explain why t h e value of the multiplier is
substantially below the reciprocal of the average requirement ratio.
This point will be f u r t h e r pursued in a subsequent section.
Our detailed analysis of the money supply mechanism also yields
information about the comparative response of the money stock to
variations in the base and the legal reserve ratios. I t appears that,
on the average, the same volume of surplus reserves is generated by
a dollar injection of base money and a dollar reduction of required
reserves achieved by means of fiat changes in reserves requirements or
the deposit redistribution process. I t follows that, per dollar of operation, open market operations and changes in legal reserve ratios induce,
on the average, the same response in t h e money supply.
A policy change of minor importance can be effectively i n c o r p o r a t e d
in the monetary mechanism through changes in the accumulated sum
of "liberated" reserves. A t times, t h e Board of Governors changes
the classification of banks within cities. F o r example, small banks
in Reserve cities may be designated as "country" banks. Such changes
in classification are identical in their effects to other changes in reserve requirements. The weighted average of reserve requirements is
raised or lowered by the reclassifications. Reserves are absorbed or
liberated and the rate of monetary expansion is consequently increased
or decreased I n all respects, this minor policy change is equivalent
to a redistribution of deposits.
Currency drains that are independent of hanks'' portfolio adjustments
I n a previous section, the public's demand f o r currency was partitioned into two components. One component, the public's m a r g i n a l
propensity to hold currency relative t o money wealth, was shown to
play a significant role in the System's multiplier response to emerging
surplus reserves. This role is reflected in the spillovers of currency
to the public associated with the banks' portfolio a d j u s t m e n t s .
T h e systematic prevalence of such induced currency drains, occurring
as p a r t of the multiplier process, is revealed by a value f o r the money
multiplier that is substantially below the reciprocal of the average
reserve requirement ratios.
T h i s section turns our attention to t h e role of the o t h e r c o m p o n e n t
in the public's demand for currency. T h i s second component describes
the response of the public's currency demand to nonmoney w e a l t h
and a spectrum of costs and yields associated with currency and related
financial assets. This demand component does not change in response
to the new deposits created or destroyed by the adjustment of bank
portfolios. T h e currency drains reflecting Variations in this d e m a n d
21
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
component are not induced as an integral part of the response to surplus reserves expressed by the multiplier mechanism. These currency drains result from changes in nonmoney wealth and modifications in the above-mentioned costs and yields. They are imposed on
the multiplier mechanism and, along with variations in the base and
changes in the average requirement ratios, help to determine the volume of surplus reserves available at the banks.
To clarify the effect of this component, consider the following
sequence. Reserve positions are in equilibrium—desired reserves equal
actual reserves. The public converts a dollar of existing demand deposits into currency, draining a dollar of reserves from the banks.
The supply of excess reserves is lowered by less than $1, to be precise by
(1-r) w here " r " designates the average requirement ratio on demand
deposits. The lowered volume of deposit liabilities also reduces the
banks' desired excess reserves. The banks' demand to hold excess
reserves per dollar of deposit liabilities, however, is substantially
smaller than the amount (1-r). It follows that a dollar's worth of
currency drain, generated by an increase in the second demand component for currency, lowers the supply of excess reserves decisively
more than it lowers desired excess reserves. Consequently, surplus
reserves have been siphoned off, a deficiency has been created; the
multiplier mechanism is thus set in motion since desired reserves are
now greater than actual reserves.
The opposite process occurs if the second demand component for
currency declines. A dollar's worth of currency converted into checking deposits increases fhe supply of excess reserves by the fraction
(1-r) , whereas the desired excess reserves increase by a smaller fraction. The volume of surplus reserves generated are slightly less than
(1-r). We note that a smaller volume of surplus reserves is created
(or destroyed) by a dollar change in the public's second demand
component for currency—that is, by an imposed currency accrual (or
drain)—than by a dollar injection of base money, or by a dollar of
reserves liberated through changes in average requirement ratios.
Currency drains of the second kind thus induce a smaller response
in the money stock than open market operations and liberations in
required reserves through changes in legal ratios. The difference in
the order of the responses is approximately equal to the fraction (1-r).
It is reasonable to expect, therefore, that the response to a currency flow
of the second type, per dollar, is less than (1-r) times the response to
a dollar change in the base, or a dollar of required reserves liberated.
The significant role played by the currency flows described in this
section can be explored further. As we have emphasized, these flows
are imposed on the multiplier mechanism, through variations in the
public's second demand component for currency. Several patterns
can be usefully distinguished among these variations. One such pattern, generally referred to as seasonal variation, was acknowledged
students of monetary affairs long before the Federal Reserve System was established. Many complaints were heard in the period
before the Federal Reserve System existed about the s e r i o u s inconveniences created by the juxtaposition of an "inelastic currency" and
the seasonal pattern in currency demand. The problem has traditionally been formulated in terms of an insufficiently elastic currency
supply. This formulation is both uninformative and misleading.
34-474—64
3
22
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
The problem faced at the time by the financial community was the
result of a seasonal change in currency demand in the context of a
base, determined by processes unrelated to this seasonal pattern, and
not accessible to immediate policy actions. Before the Federal Reserve System began to operate, the base consisted only of^ currency
issued directly by the Government sector and currency issued by
banks. The latter type of currency was covered by collateral, Government securities, deposited by the issuing bank with the Treasury.
The seasonal variations in currency demand thus unavoidably generated a seasonal pressure on bank reserve positions that was reflected
in banks' supply behavior on credit markets and consequently in
the behavior of interest rates. The seasonal element in currency
demand forced a seasonal readjustment of financial portfolios in
major parts of the economy. The seasonal readjustments of financial
portfolios, forced by the seasonal pattern of currency demand (in the
context of a base unrelated to these processes) was reflected in the
seasonal variation of interest rates.
The organization of the Federal Reserve System in 1014 introduced an institutional framework that permitted a lessening of the
consequences induced by the seasonal fluctuations of currency demand. Under the Federal Reserve System, both the sources and uses
of the base were restructured. The asset portfolio of Federal Reserve
banks, especially Reserve bank credit, became an important source
of the monetary base. And this source, controlled by the Federal
Reserve authorities, can be easily modified at almost zero social cost
Thus variations in the base, initiated by appropriate Federal Reserve
action, could be used to offset the seasonal fluctuations in currency
demand at much lower social costs after 1014 than before. The
seasonal increase in currency demand can be met by an i n c r e a s e d
supply of base money. Neither the reserve positions of banks nor
their desired portfolio of earning assets need be affected. The necessity for widespread readjustment of financial portfolios that had been
a feature of the prevailing system was eliminated. Resources originally allocated to the seasonal readjustment of portfolios were thus
freed for other opportunities.
I t should be noted, however, that %ve remain uncertain about both
the exact, shape of the seasonal movement in currency demand and
the precise response in the money stock to variations in the base and
to currency drains of the second kind. I t is therefore u n r e a s o n a b l e
to expect a perfectly matching adjustment of the base to the seasonal
gyrations in currency demand. We contend that in the absence of
any endeavor by the Board to penetrate the structure of the monetary
mechanism, it has almost no systematic and validated k n o w l e d g e
at its disposal concerning the relative effect of variations in the base
or currency drains of the second kind. Considerably more effort has
been applied to measurement of the seasonal pattern in currency
demand. But such information, even when quite reliable, is not
sufficient to guide policy designed to minimize the impact of s e a s o n a l
variations.
Seasonal demands are the source of variations in the second com*
ponent of the public's demand for currency that has dominated the
Federal Reserve's attention. But the history of monetary events in
this country provides some dramatic evidence of the relevant opera*
23
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
tion of cost and yield factors operating in the public's allocation of
wealth to currency, expressed by nonseasonal elements in the public's
second demand component for currency. A few of these events are
outlined briefly to indicate the serious consequences of ignoring all but
the seasonal element.
After the peak of 1920 the public's currency demand behaved according to the characteristic pattern for most comparatively mild
downswings. Currency, released from the public's holdings, flowed
into the banks until late in the year 1030. This reduction in the
second component of the public's demand for currency, frequently
encountered in a downswing, generated surplus reserves and thus
partly compensated the decline in surplus reserves resulting from
reductions in the base. The public's behavior thus attenuated the
nrst phase of the downswing, partly offsetting the Federal Reserve's
deflationarv posture.
By X ovember 1030 a radically new element had emerged. The
eruption of serious bank failures suddenly shifted the balance of
relative advantages in the direction of increased currency and smaller
demand deposit holdings. The risk attached to demand deposits
increased substantially, lowering the relative inconvenience of holding and using currency. The currency drain generated by this jump
in the public's demand for currency created negative surplus reserves,
the decline in surplus reserves set off the multiplier mechanism in a
deflationary direction, compressing the banks' portfolios and the
money supply. The first of a series of currency drains abated in the
winter of 1931. A second series, with increased intensity, began in the
spring, further reducing the volume of surplus reserves. Each increase in the public's demand for currency pushed available excess
reserves below the desired level, thus forcing adjustments that lowered
flie money supply. Increases in the base, reflecting some expansion of
Reserve bank credit, modified the deflationary impulses emanating
irom the public's fearful response to bank failures to some extent.
But the actions taken were too small to offset fully the decline in surplus reserves.
From January 1032 to the following April the currency drains
abated. Unfortunately, the Federal Reserve authorities permitted a
eduction in the monetary base during this period. Thus the surplus
reserves injected into the svstem by the public's declining currency
demand were destroyed. When the currency drains resumed again in
April 1032, new deflationary pressures were imposed on the system.
Once more the base expanded and attenuated, somewhat, the deflationary shock emanating from the public's demand for currency.
But again, the expansion was insufficient to offset the public's increased
demand for currency and thus break the deflationary impact on the
monetary system. This third series of currency drains tapered off
m July. The public's currency demand at first declined slowly but
accelerated later in the year. The flow of currency into the banks,
accompanied by a maintained base, sharply curtailed the deflationary
momentum and even induced a hesitant increase in the money supply,
during the fall of 1032. But currency drains emerged again m
December, accelerating to catastrophic proportions by March 1933.
-The monetary base was expanded during this period by an increase
of Reserve bank credit, but this increase was quite inadequate to stem
fc
ue deflationary impact of the public's currency behavior.
24
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
The monetary events from October 1030 to March 1033 form a
terrifying sequence of missed opportunities to break the deflationary
process by means of an effective monetary policy. During this period
the behavior of the monetary system and the money supply was dominated by the public's (second) demand component for currency,
responding to the public's repeated reappraisals of the risk attached
to checking deposits relative to currency holdings. The crisis was
met by the Federal Reserve authorities with a singularly pallid attitude. Every new'spurt of the public's currency demand, mirrored
in accelerating currency drains, was accompanied by inadequate expansions of the base. The subsequent return flows of currency to
the banks were either offset by a compression of the base or by the
banks' increased desire to hold excess reserves in response to a greater
anticipated variability of reserve flows induced by the public's currency behavior. A number of ingredients in the Federal Reserve's
conception most likely shaped their attitude during this dramatic
period. But we submit that the prevailing inchoate view of the
monetary process, the absence of any clear and tested understanding
concerning the interaction of base and currency demand in the monetary mechanisms, contributed to the repeated sequence of inadequate
or omitted expansions in the base. Contrary to the assertion made
by spokesmen for the System, the Federal Reserve authorities had
the technical capabilities to engage in such expansion.
The upswing, initiated in the spring of 1033, was accompanied by
a persistent decline in the second component of the public's currency
demand that lasted until the spring of 1036. The return flow of currency to banks generated by this process worked iointly with the
expansions of the base, resulting from the gold inflow, to raise the
money supply by substantial proportions. This behavior of the
money supply contributed, we contend, quite decisively to the duration and magnitude of the recovery. It is revealing that this monetary expansion was not the result of any actions taken by the Federal
Reserve authorities. I t occurred entirely as a result of the decline in
the public's currency demand and international capital transfers. The
net effect of Federal Reserve actions, subsequent to the spring of
1033, remained deflationary. The recovery was not aided by Federal
Reserve policy. In the face of large-scale unemployment and output # levels substantially below any reasonable measure of "full
utilization," the Federal Reserve's* manor concern was focused on
"inflation."
We skip some interesting features of currency behavior in the later
thirties and turn our attention to the war and the early postwar
period. The use of checks or currency involves costs. One^cosh associated with the use of checking deposits, is the cost of obtaining information about the issuer. In periods of limited soc : al dislocation, this
cost remains relatively constant. The maior dislocations, a s s o c i a t e d
with the wartime mobility of the population, raised this cost by disrupting the regular sources of information applicable to a more stationary population. The cost of accepting payments by check increased markedly. Some of these costs were shifted to the issuers.
The public responded by increasing its demand f o r currency. This
increase in the public's currency demand was reflected by a currency
drain that reduced the size of the continuous injections of surplus
25
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
reserves resulting from the explosion in Reserve bank credit and the
monetary base.
The demand for currency did not abate immediately after the war.
It rapidly decelerated, however, during the year 1946, and declined
from the start of 1017, until the end of 1050. The return flow of
currency to tlie banks injected surplus reserves into the monetary
system. This action by the public was of particular importance in
1919 when the Federal Reserve authorities compressed the extended
base (i.e., base plus cumulated sum of liberated reserves). A major
portion of the decline in the public's demand for currency was very
likely attributable to the operation of the factors previously considered, now working their effects in the opposite direction. Adjustment
to peacetime conditions was accompanied by lower mobility and more
stable sources of information. The frequency of exposure to unfamiliar surroundings declined, and potential check users wTere less frequently forced into a choice of higher transaction costs associated with
check payments or holding a larger currency inventory. The cost of
check payments thus declined in the postwar period, lowering the pubhe s demand for currency and leading to a return flow of currency.
This backflow and the motion of the base dominated the behavior of
the money supply during most of the period prior to the accord between the Treasury and Federal Reserve authorities in March 1951*
. If we look at the broad contours of the money supply process during the war and preaccord years, we detect a peculiar pattern. The
inflationary increase of the money supply during the war was essentially due to the expansion of the base induced by the continued
expansion of Reserve bank credit. The public's behavior on the other
hand, expressed by its currenc}r behavior, attenuated the Federal Reserve's inflationary policy. After the war, most specifically in 1948
and 1949, Federal Reserve policy turned in a deflationary direction.
The growth rate of the money stock slowed (in 1948) and declined (in
1949). The public once again attenuated the Federal Reserve's policy
by substantial return flows of currency. Thus, Federal Reserve policies during and after the war were mitigated by the public's behavior
w
i t h respect to its desired currency inventories. In the decade from
1W0 to 1950 the public reduced the destabilizing impulses emanating
from the Federal Reserve's policy operations.
The relevant operation of the currency drains in the money supply
process has not disappeared with the accord. Additional material
clarifying this long neglected aspect of our monetary mechanism
^"ill be presented in a later section, where evidence is presented in
support of the conception described in the present chapter.
Variations in the public*8 demand for time deposits independent of the
bankys portfolio adjustment
The public's demand for time deposits is partitioned into two components in a manner similar to our treatment of the demand for cur^ n c y . The component involving the marginal propensity to hold
time deposits with respect to money wealth is related to specific features of the monetary system's multiplier process, viz, the spillover of
newly created deposits into time deposits. The other component
contributes to the current flow of surplus reserves. The latter demand
operates independently of the multiplier mechanism. By generating
or absorbing surplus reserves, it contributes to the impulses maintain
26
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
ing the mechanism in motion and thus plays a role in current variations of the money supply and bank portfolios.
The second component of the public's demand for time deposits depends primarily on interest rates and nonmoney wealth. An mcrease
in the rate offered on time deposits relative to other interest rates
raises the public's desired portfolio of time deposits. The public re*
sponds by shifting funds from checking accounts to time accounts at
a rate determined by the increase in the demand for time deposits.
This shift releases reserves and increases the supply of excess reserves.
At the present time, the shift of a dollar from checking account to
time account raises excess reserves by more than 10 cents. We surmise
that the banks' demand for excess reserves is more sensitive to variations in the volume of checking deposits than to changes in the volume
of time deposits. Thus the shift of deposits toward time accounts,
induced by higher rates offered on time accounts, lowers the banks'
desired level of excess reserves. This decline is probably quite small,
at most a few cents per dollar of funds shifted. Still, the reallocation
of deposits between demand and time account lowers, comparatively
slightly, the banks' desired excess reserves and simultaneously raises
the supply of excess reserves. Surplus reserves are generated, and
the multiplier mechanism is set in motion. Bank portfolios and the
money stock are modified as a result of the process.
We noted that per dollar of deposits reallocated, at least 10 cents of
surplus reserves would be created. Suppose we fix the amount at 15
cents. This specification will enable us to determine in more detail
the effect of the deposit reallocation on the banks' portfolios and the
money supply. The evidence summarized in a later section indicates
that, on the average, the response in bank portfolios measures approximately $2.80 per dollar of surplus reserves, while the money supply
changes on the average by about $2.50. (The difference of 30 cents is
due to the spillover into time deposits with respect to money wealth.)
The deposit reallocation would thus expand bank portfolios by approximately 42 cents (15 cents of surplus reserves times a multiplier of
2.8) per dollar shifted. The net effect on the money supply, on the
other hand, emerges from the joint operation of two distinct components working in opposite directions. First, there is the conversion of
checking deposits into time deposits, initiating the process. F o r every
dollar transferred, the money supply immediately falls by $1. But
the multiplier process triggered by the 15 cents of surplus reserves
generated per dollar transferred adds approximately 38 cents to the
money stock (15 times 2.5). The deposit reallocation thus lowers the
money supply by approximately G2 cents per dollar transferred.
A similar analysis holds for events that lower the public's demand
for time deposits and induce a reallocation of deposits toward checking accounts. Such reallocation will lower the banks' portfolio by
36 cents and raise the money supply by 02 cents per dollar transferred. Variations in the public's demand for time deposits thus contribute to explain the phenomenon noted in chapter I I , viz, the differential behavior of "bank credit" and money supply. Once again, we
note that under an appropriate conception of the money supply?
money and "credit" are not "two sides of the same coin."
The events observed in the recent past dramatically reveal the
significant role played in the money supply process by reallocations
between demand and time deposits. The relaxation of regulation Q
27
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
in January 1062 enabled banks to raise the rates offered customers
on time accounts. An extensive reallocation of deposits toward time
accounts, induced by the change in regulations, prevailed throughout
the subsequent year. This reallocation reflected the increase in the
public's demand for time deposits attributable to the higher yields
obtainable on time deposits. Surplus reserves were generated by
the switch from demand to time deposits and absorbed by the portfolio
expansion of the banking system. The reallocation of deposits also
explains the difference in the behavior of the money supply and the
sum of the money supplv plus time deposits. The growth rate of the
money supply declined from December 1061 to October 1062, whereas
the larger sum, including time deposits, grew at an accelerating rate
until April 1062. The declining growth rate of both money supply
and the extended sum in the late spring and summer of 1062 was
attributable to a spurt in currency drains and the deceleration in the
growth rate of the extended base.
The role of interbank deposits
The money supply is a measure of the amount of money balances
on the public's balance sheets. Interbank deposits are therefore excluded from the money supply. Nevertheless, these intrasystem
claims, that cancel in a consolidated statement of the banking system,
play a role in the money supply process. This role has been largely
neglected in dicussions* of the monetary system. Analysis of this
role, however, involves unavoidable technical complexities that render
it inadvisable to push our discussion beyond a summary of the analytical results developed. Moreover it can be demonstrated that changes
in the interbank deposit structure, even if large relative to those
observed, exert a comparatively small effect on the monetary system.
Our previous discussion of the bank's desired portfolio of excess
reserves can be usefully extended to cover all cash assets of banks.
Such extension requires some adjustments that w^ill be glossed over
at^ this point. The incorporation of interbank deposits into a detailed analysis of the money supply process yields three principal
conclusions: (1) The banks' marginal propensity to cover net checking
balances by debiting or crediting interbank deposits or to transfer
Federal funds affects the structure and magnitude of the multiplier
response to emerging surplus reserves. (2) Variations in the banks'
desired holdings of deposits at other commercial banks (induced by
changing interest rates or modifications in the anticipated average
or variability of cash asset flows) generate surplus reserves setting
off the multiplier mechanism. Changes in the desired level of interbank deposit assets thus exert the same effect as changes in the desired
level of excess reserves. (3) Shifts in the distribution of interbank
deposits over the system release or absorb surplus reserves.
The first proposition refers to the multiplier mechanism and the
others to the mechanism injecting surplus reserves. Our third proposition acknowledges that variations of the interbank deposit structure influence the supply of cash assets relative to the desired portfolio. But omission of this process from further consideration imparts no serious error, into our analysis. The second proposition,
however, assigns a substantial role to the scale effect of interbank
deposits. This scale effect is similar to the effect of variations m
desired excess reserves. Our next section terminates the discussion
28
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
of the injection mechanism and considers this source of surplus reserves as a part of the banks' demand for reserves.
Variations in hanks' desired excess reserves independent of tlieif
portfolio adjustments
The banks' second demand component for excess reserves was introduced into our discussion above. I t was argued that this demand
component depends on interest rates, particularly on open market
rates, the rediscount rate and the anticipated average and variability
of reserve flows. Variations in open market rates, the rediscount
rate and changing anticipations about the behavior of reserve flows
modify the banks' desired excess reserve position. This modification
changes the desired position relative to the actual position and creates
surplus reserves. An increase in open market interest rates induces
banks to compress their desired excess reserve position. The release
of surplus reserve sets off the sequence of portfolio adjustments that
we have called the multiplier process. A reduction in interest rates
induces the opposite response, a reduction in surplus reserves, bank
portfolios, and the money supply. Raising the rediscount rate induces banks to hold a larger volume of excess reserves, on the average,
by increasing the marginal cost of potential reserve deficiencies without affecting the marginal cost of holding excess reserves. The optimal portfolio of excess reserves rises. If this rise is not matched by
an increased supply of excess reserves, the discrepancy between actual
and desired position pushes the multiplier process in a deflationary
direction. A similar argument applies to variations in the anticipated average or variability of reserve flows.
The dependence of the banks' desired reserve position on interest
rates connects the money supply with the credit markets and thus with
the public's asset supply behavior on these markets. Such asset supply
has at times been introduced as the centerpiece of the money supply
process, particularly by spokesmen for the Federal Reserve* System.
Our investigations confirm the dependence of the money supply on
interest rates via the banks' desired reserve position, but they also
indicate the comparatively small order of this influence. The behavior
of the money supply has been dominated by the base, the requirement ratios and the currency patterns. Variations in interest rates
do, however, modify the basic contours. They inject an element into
the money supply process that creates a minor dependence of the
money supply on the behavior of national income under existing
institutional arrangements.
The role of interest rates in the money supply process also prevails
in deep depressions. I t has often been asserted that in periods of large
deflation a "liquidity trap" opens. The " t r a p " is said to appear in the
form of a demand for cash assets by banks that is so sensitive to interest rates that rates cannot decline. The rationale for the floor has
rarely been developed explicitly. Presumably it is determined by the
marginal costs associated with credit market transactions. In any
case when interest rates have reached this floor, the banks' desired
portfolio of excess reserves is said to become indefinitely large. Any
operation that injects surplus reserves into the System—open m a r k e t
operations, gold inflows, return flows of currency to banks, etc.—is
supposed to be offset by an increase in the banks' desired excess reserves. With interest rates at the lower floor, variations in the base,
29 ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
the second components of the public's currency and time deposit demand, and changes in the requirement ratios would exert no effect
on the money supply. The modifications in the supply of excess reserves engineered by these policy actions or induced by the public's
behavior, would be immediately matched by a change in the banks'
desired reserve position.
With interest rates at the level of the "floor" or "trap," discrepancies
between actual and desired positions become impossible. The crucial
link between monetary policy and the money supply is broken. The
behavior of the money supp\v would have to be explained in terms
entirely different from those that have been presented. At the floor
level of interest rates the banks' portfolio of earning assets would only
mirror the public's asset supply decisions. Banks would be passive
agents without independent response to inflows or outflows of reserves.
Their portfolio changes would depend solely on the public's demand
for loans. Under these conditions, the money suj^ply is completely
determined by the public's asset supply to banks.
Considerable attention has been given in the literature to the conjecture that a liquidity trap operated in the thirties and rendered
monetary policy futile and useless. The large volume and persistent
accumulation of excess reserves seemed to strengthen such notions.
On occasion, the behavior of excess reserves was accepted as prima
facie evidence for the operation of a liquidity trap. Unfortunately,
vague, impressionistic evidence of this kind was accepted by the Federal Reserve authorities and given added weight by their references
to monetary policy as a matter of "pushing on strings."
The observed behavior of excess reserves is quite compatible with a
persistently effective monetary policy and can be adequately explained
in terms of the banks' demand for excess reserves developed in an
earlier section. The persistent decline of interest rates accompanied
l>y the substantial rise of anticipated average and variability of reserve
flows (induced by the shocks experienced between October 1930 and
March 10:33) raised the banks' desired level of excess reserves. The
decline of interest rates lowered the marginal cost of holding excess
reserves, and the induced increase of anticipated average and variability of reserve flows raised the expected marginal cost of reserve
deficiencies. Both types of cost thus moved in a direction that contributed to the expansion of the banks' desired excess reserves.
Our discussion of an alternative explanation of the events that have
been widely interpreted as evidence for the "trap" or "string pushing"
does not dispose of the alternative notions. Neither does the assertion
that the "trap" rendered monetary policy futile dispose of our explanation of the events. Evidence is required to discriminate between the
rival explanations. Such evidence is of importance for deciding
whether the Federal Reserve and others erred when they accepted the
v
»cw that monetary policy was inoperative and ineffective ni the dePassion of the thirties.
, ,
™ * *•
Three pieces of evidence will be discussed here. The first is an
appraisal of the effects of doubling reserve requirements m a series
of
steps. Under the "liquidity trap" notion, this action would Have
»o effect on the monev supply. The volume of excess reserves
would be lowered, and "the volume of required reserves raised, by
this action. Under the liquidity trap notion, banks are assumed to
30
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
accept passively sucli changes in reserve position. The alternative
explanation that we have presented implies that an increase in the
requirement ratios absorbs surplus reserves, sets off the multiplier process and leads to a reduction in the money supply and the
banks' earning assets. The changes in reserve requirements became
effective on August 16, 1936, March 1 and May 1, 1937. The money
supply had been growing at the annual rate of 18 percent during most
of the months of 1935. The rates of growth in the money supply for
May 1936 to December 1937 are shown in table VI-6. The rates shown
are annual rates measured from the same month in the previous year.
TABLE VI-6.—Annual rate of growth of the money supply beforjz and after
changes in reserve requirement*, monthly 1936 and 1937
Month
May 1936
June
July
August
September
October
November
December
January 1937
February
Growth
rate
(percent)
16.8
17.6
17.1
10. 8
14.5
12.9
11.6
14.1
13.1
12.1
Growth
rate
Month
March 1937
April
May
June
July
August
September
October
November
December
(percent)
12.6
7
3.3
2.4
2.1
•!
—2.0
—3.5
The National Bureau of Economic Research records a peak in economic activity in May 1937. The Federal Reserve's index of industrial production also peaked in that month, although the annual rate
of growth of output remained positive throughout the summer. Thus
the growth rate of the money supply and the stock of money began to
decline before the peak in economic activity. Since output was rising,
the decline in the money supply cannot be/attributed to a reduction in
the public's supply of assets to banks, as would be expected from the
liquidity trap notion. The evidence is inconsistent with the "trap." 4
A second source of evidence is the behavior of interest rates during
the period. Under the "trap" notion, interest rates reach a "floor'
below which they will not decline. Bond yields on U.S. Government
securities declined from approximately 3.31 percent in the second
quarter of 1933 to 1.9 percent at the time of the U.S. entry into the
war. This decline proceeded during most of the 8-year period, interrupted on occasion by increases in interest rates. An even more persistent decline can be observed in the rates charged by banks on commercial loans. The average rate on commercial loans charged by
banks m "principal cities" fell from approximately 4.53 percent in
the second quarter of 1933 to 2.41 percent in December 1941. Yields
on short-term paper declined sharply in the early phases of the recoveiy and oscillated considerably around a very low average in later
years. F o r example, in 1939 3-month Treasury bills were quoted at
yields of from .02 to .05.
Inspection of the behavior of interest rates during the period reveals
either a persistent downward trend for the longer maturities or a
sharp decline ^ followed by substantial monthly variation for the
shorter maturities. Only stock exchange loans and banker's acceptances show constant yields that remained at a low level for extended
periods. But these securities were, at the time, quite unimportant.
31
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
According to the liquidity trap conception, interest rates reach a
floor at winch banks passively accepted the variations imposed on
their excess reserves. The behavior of interest rates during the period
is inconsistent with the explanation offered by proponents of the
'trap." Again, the evidence casts doubt on the relevance of the asserted "trap" as an explanation of the events in the thirties.
This doubt is reinforced by examination of the data on bank portfolios, a third source of evidence. The liquidity trap notion suggests
that port folios reflect only the actions taken by the public. The banks'
response is said to be passive. I t is difficult to conceive of the mechanism underlying this passive adjustment. Banks caught in the liquidity trap would have a loan portfolio dependent only on the public's
demand for loans. But banks also have portfolios of Government
securities that are bought and sold on an organized market. An expansion of the security portfolio cannot be interpreted as an event occurring because the public dumped packets of securities on the banking
system. Under prevailing institutional arrangements, securities purchased by banks necessarily involved a choice by bankers confronted
with costs and yields on alternative assets.
p I t is noteworthy that in the first year of recovery from the depression, the expansion of "bank credit" took the form of an increase in
the securities portfolio that more than offset the decline in loans.
This portfolio expansion resulted primarily from an inflow of gold
that generated surplus reserves. The accelerated demand for securities by banks contributed to the persistent decline in yields on securities. Under the base conception of the money supply process, the
decline in .yields helps to explain the increased holding of excess reserves by the banks and the attenuation—but not the elimination—
of the magntitude of the monetary multiplier. Thus, according to
the base conception, the explosion of the monetary base in the middle
thirties occurred in the presence of a smaller monetary multiplier.
One should also note that, during the recovery from 1933 to 1937,
the portion of the increase in the base that flowed into excess reserves
declined persistently. This observation is consistent with the attenuation of the anticipated mean and variance of reserve flows as
the catastrophic events of the early thirties faded into the background.
Detailed inspection of the events of the thirties casts serious doubt
on the relevance of the liquidity trap notion. These doubts are reinforced by our results bearing on the interaction of the demand for
and supply of money. Moreover, almost no evidence beyond casual
references and impressionistic appeals has ever been advanced in
support of the "trap" conjecture. 3 We thus dismiss this conjecture
as poorly substantiated and conclude that variations in the banks'
desired portfolio of cash assets, in response to interest rates and
anticipated average or variability of reserve flows, do inject surplus
reserves into the multiplier mechanism. But these flows do not systematically offset the surplus reserves cast up by the other processes
previously discussed. The existence of a demand for excess reserves,
dependent on prevailing credit market conditions, thus modifies the size
J A r e c e n t a t t e m p t c l a i m i n g t o p r o v i d e evidence s u p p o r t i n g t h e ' ' t r a p " c a m e to o u r
a t t e n t i o n a f t e r t h i s section w a s w r i t t e n . Cf. G. H o r w i c h , " E f f e c t i v e Reserves, Credit, a n d
C a s u a l t y in t h e B a n k i n g S v s t e m of t h e T h i r t i e s , " in D. Carson, ed., B a n k i n g a n d M o n e t a r y
s t u d i e s . H o w e v e r , H o r w i c h ' s t e s t s a r e i n c o n s i s t e n t w i t h t h e h y p o t h e s i s he f o r m u l a t e s In
ttoe a p p e n d i x .
32
ALTERNATIVE APPROACH TO T H E MONETARY M E C H A N I S M
of the response in money supply to policy actions during periods of
comparatively low interest rates. But the response is not obliterated.
Policy actions were connected to the money supply even during the
phases of collapse and insufficient recovery of the 1930's. I t follows
that monetary policy cannot be absolved from responsibility for both
events. Inappropriate policies, guided by fundamental misconceptions about the structure of the monetary process, accelerated the
deflationary process in 1929-33, and subsequently prevented the large
monetary expansion required to absorb all idle resources.
A SUMMARY OF T H E MONEY SUPPLY PROCESS AND AN INDICATION OF
S03IE EVIDENCE
According to the conception outlined above, variations in the
money supply are explained in terms of the extended m o n e t a r y
base (i.e., the base plus the cumulated sum of reserves "liberated" from or "impounded" into required reserves by fiat changes
in requirement ratios or due to the redistribution of existing deposits
between different classifications), the public's desired wealth allocation to currency and time deposits, and the banks' desired portfolio
of excess reserves. The major implications of this conception can
be summarized in the following propositions:
(1) Variations in the base, requirement ratios and the public's currency behavior dominate the behavior of the money supply.
(2) The public's asset supply to banks, especially the public's loan
demand, affects the money supply to the extent that the banks' desired
volume of excess reserves is sensitive to interest rates. There is little
doubt that excess reserves depend on interest rates and are highly
sensitive to interest rates in a low interest rate regime. Nevertheless,
the base and the public's currency behavior dominated the money
supply process during the middle and late thirties as in other periods.
(3) A dollar change in the base induceSj on the average, a multiple
change of the money supply in the same direction. The multiplier is
substantially below the reciprocal of a weighted average of r e q u i r e m e n t
ratios. Depending on the marginal allocation of newly created deposits to demand or time accounts, this reciprocal would be at least 7
and at most 25 (at present). If 20 percent of the newly created deposits were allocated to time accounts, the reciprocal would be 10.6.
On the other hand, if 80 percent of the newly created deposits were allocated to time accounts, the reciprocal of the suitably weighted average
requirement ratios would be 21.4. According to the base conception,
the multiplier effect of the base is much smaller than the lower boundary of the reciprocal mentioned. The single most important reason for
this comparatively reduced magnitude of the monetary multiplier is
the spillover into currency. This spillover is typically associated with
the banks' series of portfolio adjustments triggered by surplus reserves. These spillovers mirror the public's positive marginal propensity to hold currency relative to money wealth.
(4) Changes in reserve requirement ratios induce changes of the
money supply in the opposite direction. A dollar change in r e q u i r e d
reserves, attributable to changes in the average requirement ratios induces, on the average, a response in the money supply that is s i m i l a r
33
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
in magnitude to a dollar change in the base. The accumulated sum
of all reserves liberated from or absorbed into required reserves by
changes in the average requirement ratio has been computed for our
analysis. This sum is added to the monetary base to form the "extended monetary base," the single most important determinant of the
money supply. The extended base is completely controlled by the
Federal Reserve authorities via open market policy, reserve requirement policy, and discount policy.
(5) The extended monetary base, currency drains and reallocations
of deposits between demand and time accounts affect the money supply,
per dollar of operation involved, in a descending order of magnitude.
A dollar change in the base has a larger effect than a dollar's worth
of currency drain, and the latter has a larger effect than a dollar's
worth of deposit reallocation. The difference in the magnitude of the
responses is determined by the size of the prevailing requirement
ratios and the banks' marginal propensities to hold excess reserves
with respect to demand and time deposits.
(6) The size of the monetary multiplier depends on the reserve requirement ratios, the public's marginal propensity to hold currency and
time deposits with respect to money wealtn, the banks' marginal propensity to hold excess reserves with respect to deposit liabilities, and
the pattern of interbank payments through the correspondent banking
system. While the magnitude of the multiplier is responsive to variations in requirement ratios, it is much less sensitive to the level of the
legal ratios than is usually suggested. The elasticity of the multiplier
with respect to the reserve ratios is substantially below unity (in magnitude) and is approximately minus one-third. A 20-percent increase
in requirement ratios from 15 to 18 percent would lower the multiplier
by about 7 percent; i.e., from 3 to 2.8.
(7) The conception describing the money supply process clearly
reveals that "monetary expansion" and "credit expansion" are not
"two sides of the same com." Instead the conception implies that
the joint operation of the base, the requirement ratios (via the cumulated sum of liberated reserves), the public's allocation of payment
money between currency and checking deposits, and its allocation of
deposits between checking and time accounts yields substantially different patterns for money supply, total deposits and the banks' portfolio of earning assets (i.e., one*sense of bank credit). A dollar of
liberated reserves induces the same response in the banks' portfolio
as in the money supply plus time deposits and a smaller change
in the money supply. 6n the other hand, a dollar change in the base
exerts, on the average, a smaller effect on the banks' portfolio of earning assets than on the money supply. An important result that we
asserted above to hold for the money supply (with or without time
deposits) does not hold for the banks' earning assets. The response
of earning assets to variations in the base is smaller on the average
than the response to liberated reserves.
(8) Variations of the interbank deposit structure within the observable range exert only a negligible effect on the money supply.
Moreover, if required reserves were assessed against total demand
deposits (unadjusted for claims against other banks), the connecting
link between the interbank deposit structure and the money supply
would be broken.
34
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
Some comments on policy action
Monetary policy centers on open-market operations, reserve requirement ratios and the discount rate. But other courses of action remain available to both Federal Reserve authorities and the Treasury.
Among such courses of action affecting the money supply should be
mentioned: (1) the reclassification of member banks, (2) the Federal
Reserve's check collection arrangements and the behavior of float,
(3) the administration of the discount window revealed by the cyclical
behavior of discounts and advances among the source components of
the base, (4) the Treasury's administration of its money balances,
particularly the division of the total balance between Federal Reserve
accounts and tax and loan accounts at commercial banks and the relative variability of these two components, (5) the division of the total
gold stock between reserve account and the general fund, and (6)
the extent to which the Treasury uses a surplus to retire G o v e r n m e n t
debt held by the Federal Reserve banks or finances a deficit (directly
or indirectly as in 1917-18) by borrowing from Federal Reserve banks.
The effect of both open market operations and changes in reserve
requirement ratios have been discussed sufficiently in previous sections. Open market operations immediately modify the base and thus
set off the multiplier mechanism; changes in the legal ratios also inject surplus reserves in the manner described above. But the operation of the rediscount rate and the effect of other policy actions requires further discussion.
Discount policy
Variations in the discount rate work their effects on the money
supply through two different channels, the banks' desired volume of
excess reserves and the base. (1) An increase in the discount rate
raises the banks' expected marginal cost of potential reserve deficiencies
and induces a rise in the desired volume of excess reserves. This response in bank behavior is comparatively small. But, surplus reserves
are reduced, since desired excess reserves are pushed above the supply
of excess reserves. To the extent this occurs, the multiplier mechanism works to lower both the monev supply and bank nortoMos
until desired and actual excess reserves are equated. (2) The effect
on the base results from the occurrence of discounts and a d v a n c e s
among the sources of the base. The borrowing of commercial banks
(mostly member banks, but occasionally foreign central banks)
usually dominates the behavior of this fxnirce component, B a n k s
respond to the discount rate announced by the Federal Reserve
authorities according to a demand pattern that is partly influenced
by the Federal Reserve System's administrative pressures, expressed
by the "tradition against borrowing" so persistently advocated by
spokesmen f o r the System. Under this demand pattern, banks'
indebtedness to Federal Reserve banks is essentially d e t e r m i n e d
by the discount rate, the Federal funds rates and other rates of
interest. The available evidence strongly supports the c o n t e n t i o n
that an increase in market rates relative to the discount rate
raises the banks' volume of indebtedness and thus raises the F e d e r a l
Resenre banks' discounts and advances. Conversely, an increase in
the discount rate, relative to prevailing market rates, reduces the incentive to borrow, lowering the Federal Reserve's portfolio of d i s c o u n t s
and advances and the base. A n increase in the discount rate thus
A L T E R N A T I V E APPROACH TO T H E M O N E T A R Y M E C H A N I S M
35
induces a negative response in the money supply via the banks' desired
volume of excess reserves and the volume of discounts and advances.
A reduction in the discount rate would similarly operate to raise the
base and compress (slightly) the banks' desired volume of excess
reserves.
The administration of the discount window contributed both in the
twenties and the fifties to the cyclical variability of the money supply.
The discount rate typically lags behind the movements of the market
rates. A cyclical upswing, generated or reinforced by nonmonetary
factors, pushes market rates ahead of the discount rate, and induces
banks to expand their borrowing. The rising volume of discounts
and advances increases the base and consequently increases the money
supply. A reverse operation occurs in a downswing. The cyclical
variability of the money supply is thus amplified by the operation
of the discount window. A feedback mechanism is introduced
into
the monetary system that amplifies economic fluctuations.4
Neither the amplifying feedback nor the administrative pressures
applied to borrowing banks are necessary features of the discounting
mechanism. Administrative pressure could be replaced by a discount
rate that is always higher than open market rates and lean rates. The
discount rate has traditionally been low in the United States relative
to short-term asset yields. This alone would have induced, on the
average, large scale borrowing by banks. Thus a "tradition against
borrowing" had to be fostered by administrative procedures that
complicate a bank's "life with the Federal Reserve.;" If a penalty
rate replaces the traditional "inducement rate," administrative pressures to bottle up commercial banks' borrowing requests become unnecessary. Furthermore, rapid adjustments of this penalty rate to
evolving market situation would eliminate the amplifying feedback
traditionally associated with the Federal Reserve's discount operations.
The Federal Reserved check collection facilities
The portfolio of securities and discounts dominates the longrun
movements of Reserve bank credit But variations in Federal Reserve
float frequently exert a decisive effect on Reserve bank credit and the
base in the very short run, e.g., from week to week. Float arises as part
of the Federal Reserve's check collection process that was described in
the section devoted to the base. It consists of checks drawn on commercial banks, collected via the Federal Reserve System, for which
collection time has exceeded the fixed time schedule. Every dollar
of additional float adds a dollar to the base.
Federal Reserve float is determined by the volume of checks presented to Federal Reserve banks for collection, the time period fixed
for deferred availability credit and the probability distribution governing the time required by Federal Reserve banks for processing and
collecting these checks. Both the volume of checks and the actual
collection time are subject to substantial shortrun vagaries that create
the erratic shortrun behavior of float. These erratic movements, difficult to foresee from day to day or week to week, are a major source of
shortrun variation in the base. The fluctuations attributable to float,
' T h e result of discount policy that has just been described reinforces
X^hSSt!?
m
o
described In an earlier chapter. We^notedtin t h e w r i t e r chapter
u
h
tfa n P t a
at postwar monetary policy has been procyclical rather than countercyclical on the
average.
36
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
however, tend to average out and in the longer run are submerged
by the changes in the base due to gold or other components of Reserve
bank credit.
.
.
,
The operation of the check collection facilities thus contributes substantially to the feeling that "defensive" action is required. Float
enhances the uncertainties confronting the account manager and the
pressures on him to concentrate on the shortest run horizon that we
discussed in chapter I I . The erratic shocks in the base emanating
from the check collection process could be attenuated by lengthening
the period governing deferred availability credit. Moreover, if
reserve accounts of banks presenting checks for collection were credited
at the time the reserve accounts of the drawers' banks were debited,
float would disappear and with it the erratic short run element in the
base.
The administration of Treasury deposits
The Treasury holds tax and loan accounts with thousands of commercial banks. I t also maintains deposits at Federal Reserve banks.
The Treasury has to decide how to partition its total balances between
the two types of accounts. Variations in this division, and particularly variations in the Treasury deposits at Federal Reserve banks,
have monetary consequences.
A transfer of Treasury funds from tax and loan accounts to Federal
Reserve deposits raises a component of the base that is included as a
negative source. The base declines and deflationary impulses are sent
through the multiplier mechanism when such transfers occur. The
base rises when the Treasury disburses funds from its Federal Reserve
account. The administration of the Treasury's money balances can
exert a substantial effect on the monetary system. The Treasury's
receipts from taxes and borrowing do not have a timing pattern matching the expenditures made by drawing on its Federal Reserve deposits.
A t times in the past, considerable shortrun variation in Treasury deposits at Federal Reserve banks have occurred as a result. These
variations either induced a corresponding shortrun instability in the
base or added to the uncertainties facing the account manager. A
readjustment in the Treasury's procedures and improve c o o r d i n a t i o n
between the account manager and the Treasury now minimize the
destabilizing impact of variations in Treasury balances. Most of the
Treasury's receipts accrue on tax and loan 'accounts at c o m m e r i c a l
banks and are transferred to Federal Reserve deposits under a formula arrangement. Treasury deposits at the Federal Reserve banks
are kept at a comparatively very low level. Moreover, transfers are
geared sufficiently close to disbursements to contain fluctuations in
Treasury deposits at the Reserve banks within relatively narrow
limits. The effects on the base and the money supply have thus been
similarly confined.
Treasury deficit and surplus
Numerous discussions during and after the war about the interrelation of fiscal and monetary policy attached great importance to the
monetization of debt by the monetary system. The inflationary effect
of a deficit seemed to depend on the distribution b e t w e e n " public
and banks of the securities issued to finance the deficit. Monetization of debt by commercial banks was alleged to raise the money
37
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
supply and contribute to inflationary pressures. Application of a surplus, on tlie other hand, to retire debt from banks was often characterized as a deflat ionary policy action designed to lower the money supply.
Our analysis of the money supply process clearly implies that a coordination of Federal Reserve expansion of the base with Treasury
deficits exerts a substantial effect on money supply and bank portfolios.
}Ve contend, however, that the concept of debt monetization by banks
involves serious misconceptions. Such misconceptions may have been
fostered by the unfortunate terminology used in Federal Reserve publications. The analytical frame outlined in previous sections permits
us to clarify the issues involved.
The crucial distinction in the case of financing a deficit is not between securities acquired by banks and securities acquired by the public but between securities acquired by Federal Reserve banks and
securities acquired by the banks or the public. The first method of
classification has no relevance for money supply behavior. The second
classification permits consideration of the significant difference in
the response of the money supply and bank portfolios to the Treasury's
fiscal procedures. If the Treasury's issue of new securities is matched
by an equal purchase of marketable securities by Federal Reserve
bunks, then Reserve bank credit expands and the base rises immediately
by a matching amount. Money supply and bank portfolios respond
correspondingly. On the other hand, if the Treasury's new issues are
offered on the market for placement in the public's and the banks'
portfolios, neither the base nor any other major determinant of the
money supply or total bank earning assets is affected directly. However, prevailing interest rates are altered. The Treasury's new offering
to the public or the banks raises the (stock) supply of Treasury securities confronting the public's and banks' demand. Interest rates, therefore, rise. This rise in interest rates lowers the banks' desired excess
reserves, raises their desired borrowing from Federal Reserve banks,
and compresses the public's relative demand for time deposits. The
induced increase in the commercial banks' borrowing from Federal
Reserve banks expands the base and consequently expands the money
supply.
The induced reallocation of the banks' assets between cash assets
and earning assets, or the public's deposits between time and checking
accounts, also contributes to the increase in the money supply. The
Espouse of the money supply to new Treasury issues sold to the banks
the public thus depends on the operation of the interest mechanism.
The magnitude of this response is determined by the sensitivity to
interest rates (i.e., the interest elasticity) of the banks' desired excess
reserves, the banks' indebtedness to Federal Reserve banks, and[the
public's demand for time deposits. A pronounced sensitivity of both
banks and public to interest changes (expressed by lar^e interest
elasticities) generates comparatively large responses of the money
supply to new Treasury issues sold to the public and the banks. An
interest insensitive behavior of the public and the banks, on the other
hand, renders the money supply q u i t e insensitive to variations m the
Treasury's debt operations.
The effects of interest rates on the banks demand for excess reserves
and borrowing and the public's demand for time deposits influence the
money supply-in the same direction. This is not the case for the comO4-474—64
4
38
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
mercial banks' portfolio of earning assets. The banks' and the public's response to interest rates work in opposite directions on this
magnitude. I t follows that the working of the interest mechanism,
set in motion by the market's absorption of new Treasury issues or the
Treasury's net retirement of outstanding debt, generates a larger response in the money supply than in the banks' portfolio of earning
assets.
The base conception of the money supply process thus implies that
the Treasury's debt management operations affect the money supply.
But this conception also implies the need for a careful distinction between variations in the distribution of outstanding debt that raise or
lower the Federal Reserve's holding and those that alter only the
distribution between commercial banks and the public. Reshuffling
the ownership of the public debt between commercial banks and the
public has no significant effect on the money supply.
When the Treasury's debt operations are matched by a corresponding increase in the base, the stock supply of securities on the market
is unchanged and the money supply expands. If the Treasury's operations only shift the stock supply of securities on the market, the
response of the money supply is comparatively small. This should
not suggest that the redistribution of securities bet ween the b a n k s and
the public has no economic effect. The redistribution affects the structure of bank assets—the composition of portfolios between loans and
securities—in a manner determined by the public's and the bank's demand for financial assets. But this redistribution does not affect the
money supply. The monetary effect of the shift in securities between
the public and the banks depends, in a first approximation, on the
change in the size of the outstanding debt. I n short, m o n e t i z a t i o n
or demonetization of Government debt by commercial banks has no
relevent connection wTith the behavior of the money supply and the
total volume of banks' earning assets, in the conception p r e s e n t e d .
Some conflicting implications of the base and free reserve doctrines
The free reserve conception has been discussed in some detail in
earlier chapters. 5 AVe contend that this conception has held the dominant position among the Federal Reserve's notions. The radical difference between the base conception, outlined in this paper, and the
Federal Reserve's view is explored in this section.
I n the Federal Reserve's view emphasizing the central r o l e played by
free reserves, increased borrowing by banks is relatively deflationary,
and a larger supply of excess reserves is relatively inflationary. A
completely different interpretation of borrowing and excess reserves
emerges under the conception that assigns a central position to the
extended base and the public's currency behavior. An increase of
member bank borrowing from Federal Reserve banks raises the monetary base. This holds equally for funds obtained by banks t h r o u g h
discounting eligible paper or through loans based on collateral deposited at Federal Reserve banks. The leiral and administrative technicalities have no relevant effect on the base. I n either case, the base
increases and so does the money supply. Similarly, a c o n t r a c t i o n in
the Federal Reserve's discounts and advances lowers the base and comp r e s s the money supply.
6
Published separately as "The Federal Reserve's Attachment to the Free Reserve
Concept" a subcommittee print
39
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
The conception outlined in the present study thus reverses the Federal Reserve s traditional interpretation of member bank borrowing.
The Federal Reserve's expanding portfolio of discounts and advances,
typically observed during an upswing, does not, according to the conception centered on the base, gradually decelerate the money supply
and inject "less ease" or "more tightness" into the monetary system.
On the contrary, member bank borrowing contributes to the persistent
growth in the money supply observed during periods of economic expansion. Similarly, the reduction of member bank borrowing during
a recession is not an event that "eases" the monetary system by accelerating the money supply or slowing its deceleration. A decline in
discounts and advances lowers the base and retards the growth rate of
Jhe money supply. Under the base conception, reductions of member
bank borrowing are a deflationary and not an inflationary move.
The same argument applies of course to any component of Reserve
bank credit. An increase in Reserve bank credit, whatever the precise
form and technicality involved, raises the base and consequently expands the money supply. In particular, securities purchased by Federal Reserve banks from security dealers under repurchase agreements
also affect the base and thus the money supply. Under the modified
base doctrine presented in this chapter, member bank borrowing and
repurchase agreements have exactly the same effects on the money
supply.
According to the free reserve doctrine essentially trivial differences
in technicalities lead to opposite evaluations of member bank borrowings and repurchase agreements. Both events involve loans made by
Federal Reserve banks on the basis of collateral left with the Federal
Reserve banks. When a member bank borrows by obtaining a loan
from a Federal Reserve bank, collateral is pledged to the Reserve bank
as a guarantee of repayment. Many banks maintain collateral in the
form of short-term Government securities at the Federal Reserve to
simplify the mechanical side of this transaction. If the bank has not
pledged the securities to the Reserve bank as collateral for a loan, they
can be used for transfers from bank to bank over the private wires of
the Federal Reserve System. But if they have been pledged as collateral for a loan from*the Reserve bank, they are no longer available
for this purpose. Thus the effect of the most commonly used form of
borrowing from a Reserve bank is a temporary reduction in the amount
of short-term securities available in the market, although technically
the title to the securities remains with the member bank.
A loan to Government security dealers is technically treated as a
dealer repurchase agreement. The dealer sells short-term Government
securities to the Reserve bank and agrees to repurchase the same securities at the end of a specified number of days. The effect on the monetary base of the dealer repurchase agreement is identical in every way
with the effect of a collateral loan. The dealer receives bank reserves
(Federal funds), and the Federal Reserve makes a loan to the dealer
that is in effect collateralized by Government securities. A liability
item, reserves of member banks, increases on the consolidated balance
sheet of the 12 Federal Reserve banks. The asset item Federal Reserve
bank credit rises. The monetary base increases temporarily, i.e., until
the transaction is reversed.
The Federal Reserve recognizes that dealer repurchase agreements
are a means of supplying reserves and easing the prevailing pressure
40
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
on the monetary system. Indeed, the rationale for the use of repurchase agreements is that it is a means of easing a money market situation that has become temporarily more restrictive than is deemed desirable. We must ask, therefore, how two transactions that have an
identical effect on the consolidated balance sheet of the Reserve banks
can have a different effect on the market. The answer in terms of the
monetary base is, of course, that they do not have different effects.
Collateral loans to member banks and dealer repurchase agreements
have an identical effect on the monetary base. But an increase in borrowed reserves has no effect on free reserves because total reserves
and borrowed reserves both rise, while a dealer repurchase agreementraises total reserves and is not counted as an increase in borrowed
reserves. Judged in terms of the concept of free reserves, borrowed
reserves and dealer repurchase agreements are different. Only the
latter expands free reserves and accelerates "credit expansion. 7 ' It is
not unlikely that the incorrect appraisal of the role of borrowing by
the Federal Reserve has been an important contributing factor to the
procyclical movement of the money supply in the postwar years.
An arbitrary technical distinction thus yields a strange result u n d e r
the free reserve doctrine. An increase in member bank borrowing
decelerates the money supply, whereas an expansion of r e p u r c h a s e
agreements with security dealers accelerates the money supply. Loans
made on collateral are deflationary and loans made by simultaneously
purchasing spot and selling long are inflationary. Of course, the
Federal Reserve might concede this point and adjust the notion of free
reserves by lumping repurchase agreements with member bank borrowings. Expansion of either repurchase agreements or bank borrowings, excess reserves unchanged, would lowrer "free reserves"
and decelerate the money supply. But, the puzzle has only been
shifted. Loans made under the form of a repurchase a g r e e m e n t ,
combining a spot purchase with a long sale at a specified future date,
would be deflationary under the adjusted free reserve conception at
the time of purchase. On the other hand, spot purchases of s e c u r i t i e s
by Federal Reserve banks combined with a definite decision to u n l o a d
the securities at an uncertain future date, would be expansive at the
time of purchase and contractive at the time of sale.
The base conception of the money supply also yields an interpretation of the role and position of excess reserves that differs sharply
from the view presented by the free reserve doctrine. The latter interprets expanding excess reserves to mean accelerated movements in
money supply and "bank credit." Falling excess reserves are a s s i g n e d
a deflationary interpretation. Once more, the modified base doctrine
reverses this interpretation. Expanding excess reserves, typically associated with falling interest rates, actually slow the growth rate of
the money supply and bank portfolios. And declining excess r e s e r v e s ,
usually accompanied by rising interest rates, accelerate the g r o w t h
in money supply and bank portfolios.
I t should be noted, however, that all Federal Reserve statements
do not conflict with the implications of the base doctrine. M o r e o v e r ,
at times it is possible to interpret some of the statements made by the
Federal Reserve in terms of the base conception. B u t such interpretations yield no support for the more frequently occurring statements
that justify policy operations in terms of the free reserves c o n c e p t i o n .
41 ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
CONCLUSION
In section I of this study, we have developed an alternative conception of the6 money supply process that we have called the modified
base doctrine. This conception emphasizes the role of the extended
monetary base and the public's currency behavior as the principal
elements that must be incorporated in an analysis of money supply
behavior. We have also pointed to some other elements that must
be incorporated as part of a more complete explanation. Specifically,
the public's demand for time deposits, the banks' demand for excess
and borrowed reserves, and the role of interbank deposits, and the
check collection process have been indicated as ingredients in the
process.
Specific policy actions and the events recorded in monetary history
have been discussed using the framework centered on the base. We
have noted that the conception developed is an extremely useful tool
for separating the policy actions taken by the Federal Reserve to
influence the money stock from the other factors affecting the money
supply. The analysis reveals that the failure of the Reserve authorities to appreciate the importance of the public's demand for currency
and the demand by banks for excess reserves
contributed in major ways
to the serious policy errors of the thirties.7
The base conception established a connection between the comparative size of the excess reserves in the banking system and the strength
of the feedback from nonmonetary processes to the money supply.
TJnder some conceptions, this feedback from economic activity, via
the demand for loans, is the principal determinant of the money supply. The base doctrine acknowledges the feedback but denies that
it is the centerpiece of the money supply process. Instead, the feedback is viewed as operating principally through the effects of variations in interest rates on the banks' demand for excess reserves. In
periods of low interest rates and large excess reserves, an increase in
the pace of economic activity that stimulates the demand for loans
and raises interest rates compresses the demand for excess reserves
by banks and increases the money supply. Our investigation suggests that this effect is stronger in periods of low interest rates.
The effectiveness of monetary policy, operating through the extended base, is not impaired by the existence of the feedback. Operations on the base or on the reserve requirement ratios could offset the
induced expansion. And policies to expand the money supply need
not patiently wait for the operation of the feedback. Moreover, the
importance of the feedback could be further reduced by institutional
rearrangements that reduce the cost of holding excess reserves or the
cost of potential reserve deficiencies.
T h e monev supplv theorv outlined h a s been developed in more detail in a paper published bv one of t h e a u t h o r s In "A Schema f o r t h e Supply Theory of Money, I n t e r n a t i o n a l
Economic Review, J a n u a r y 1001. More m a t e r i a l bearing p a r t i c u l a r l y on the public s
d e m a n d f o r c u r r e n c y and time deposits w a s contained in " T h e S t r u c t u r e of t h e Monetary
System and t h e A r c r e s a t e Monev Supply F u n c t i o n , " which w a s presented a t a session of
t h e E c o n o m e t r i c Societv in Decrmber 1960. The reader m a y also consult our p a p e r
" S o m e F u r t h e r Investigation of Demand and Supply F u n c t i o n s f o r Money, J o u r n a l of
F i n a n c e , May 1064. Our f o r t h c o m i n g book will contain several c h a p t e r s dealing w i t h
m o7n e y supply theories.
A more detailed a n a l y s i s of t h e F e d e r a l Reserve's policy d u r i n g the recovery phase of
t h e g r e a t depression can be f o u n d in K a r l B r u n n e r ' s / ' A Case S t u d y of U.S. M o n e t a r y
P o l i c y : Reserve Requirements and t h e I n f l a t i o n a r y Gold Flows of t h e Middle Thirties,
Schweizerisehe Z e i t s c h r l f t f u r V o l k s w i r t s e h a f t u n d Statistife, 1958.
S E C T I O N I I — S O M E E V I D E N C E ON T H E R E L A T I O N O P T H E
M E C H A N I S M TO T H E S U P P L Y O F M O N E Y
BASE
A major contention in our criticism of the Federal Reserve System
is that they have failed to develop an adequate analysis of the money
supply process and have failed to support their conjectures
with evidence. The evidence presented in chapter V 1 provides ample
reason for doubting the relevance of the modified free reserves doctrine.
But neither our doubts nor the existence of an alternative conception
are sufficient to reject the Federal Reserve's view centered on free reserves. The alternative conception must be appraised in competition
with the free reserves doctrine. In this chapter, some evidence had
been gathered to2 indicate the comparative relevance of the modified
base conception.
T H E R E L A T I O N OF T H E M O D I F I E D BASE C O N C E P T I O N TO T H E M O N E Y S U P P L Y
The discussion in the previous chapter assigned a dominant role in
the monetary process to the effect of the extended base on the money
supply. We noted there that the extended base is the single most important determinant of the stock of money. However, our analysis
implies that other factors must be included along with the base for a
more complete explanation of monetary behavior. When these elements are incorporated with the extended base, we obtain the relation
that we call the modified base conception. To recapitulate the earlier
discussion, the modified base conception incorporates: (1) the monetary
base plus the accumulated sum of reserves liberated or impounded by
changes in reserve requirements; (2) the public's second demand component for currency; (3) the public's second demand component for
time deposits; (4) interest rates; and (5) the rediscount rate. The
two last elements reflect the operation of the interest mechanism on
the banks' desired reserve position and provide a link between the
money supply and the credit markets.
The coefficient of determination is used to measure the relation of
changes in the base or in the elements of the modified base conception
to changes in the money supply. This measure was introduced in
chapter V 3 when we appraised the evidence for the modified free
reserve conception. As we noted there, the coefficient of determination must be between zero and 1. The closer the computed coefficient of variation comes to 1, the larger the percentage of variation
in one entity that is explained by concomitant variations in the others.
Table V t l - 1 presents the coefficients of determination for the extended base and the modified base conception. The underlying data
^ 1 P u b l i s h e d s e p a r a t e l y a s sec. I l l of " T h e Federal Reserve's A t t a c h m e n t t o t h e F r e e
Reserve
C o n c e p t ." a subcommittee p r i n t , by Karl B r u n n e r and Allan H. Meltzer.
s
O t h e r evidence is provided in our paper "Some F u r t h e r Investigation of t h e Dem a n d nnd S u p p l y F u n c t i o n s f o r Money," J o u r n a l of Finance, M a y 1964. Additional findings will be r e p o r t e d in o u r f o r t h c o m i n Tg book.
_
,
. •. . „ „
•
..
* PubUshed s e p a r a t e l y a s sec. I l l of 'The F e d e r a l Reserve's A t t a c h m e n t • * V op. c i t .
43
44
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
are tlie changes in each of the variables from a given quarter to the
corresponding quarter in the following year. This is similar to the
procedure that we used in some tests of the free reserve conception.
Two principal advantages of measuring changes from year to year
recommended the use of annual changes between corresponding quarters of adjacent years rather than changes between adjacent quarters.
First, there is seasonal variation in the data for the change in the
money supply and the other entities. The use of annual changes avoids
the problem of choosing some alternative procedure for eliminating
seasonal variations. Second, and most important, we have indicated
repeatedly that the so-called defensive operations, practiced by the
Federal Reserve, impart substantial variation to the changes in the
money supply. The use of changes between corresponding quarters in
adjacent years dwarfs the importance of these h a p h a z a r d variations.
The computations were made for two overlapping periods. Both
begin with the first quarter of 1940. One ends in the fourth quarter
of 1958, the other in tlie fourth quarter of 1962. The s e p a r a t i o n into
two periods permits an evaluation of the effect of an institutional
change that occurred during 1959-60. Congress granted to the Federal Reserve the power to vary the proportion of member banks' vault
cash that was counted as reserves. Shortly after this p o w e r was
granted in 1959, the Federal Reserve, in a series of steps, p e r m i t t e d
all vault cash to be counted as part of reserves. The modified base
doctrine indicates that this institutional change should be i n c o r p o r a t e d
by adding the volume of vault cash released to the extended base.
When vault cash is incorporated in this way, the effect on the money
supply of each of the elements in the modified base conception should
be unchanged. The division into two periods permits a check on the
specific interpretation of the change in institutional arrangements and
the implied statements about the response of the money supply to the
change.
Columns 1 and 4 in table Y I I - 1 indicate the gross association between changes in the extended base and changes in the money supply
or the monev supply plus time deposits. I t is apparent that the gross
association between the changes m these magnitudes is substantially
larger than any of those observed when we considered the modified free
reserves doctrine. Thus the data suggest the comparative relevance of
the base doctrine and the comparative irrelevance of the modified free
reserve conception.
The coefficients of determination in columns 2 and 5 go beyond
the simple association between changes in the money supply and
changes in the extended base. The extended base is now a u g m e n t e d
by all of the determinants of the money supply that are a part of the
modified base conception. At. least 95 percent of the variations in
changes of the money supply are explained by the changes in the varia b l e s included in the modified base conception. Substantially superior
results are thus obtained when changes in the elements r e p r e s e n t i n g
the behavoir of the banks and the public are incorporated in the mechanism. By f a r the larger part of the changes in the money supply
appear to be responses to movements in the proximate d e t e r m i n a n t s
that have been specified.
Additional information on the operation of the extended base in
the money supply process is given in columns 3 and 6. The i n d i c a t o r
45 A L T E R N A T I V E A P P R O A C H TO T H E M O N E T A R Y M E C H A N I S M
used is the partial coefficient of determination. This measure shows
the association between changes in the extended base and changes in
the money supply, after removing the effect of other money supply
determinants included in the modified base doctrine. We usually find
that the extended base and the currency factor are the two most important elements in the process. In the samples used here, this result
is obtained again. Changes in the extended base account for 91 percent of the changes in the money supply afer removing the effect of
other variables operating on the money supply.
Moi cover, we note that the coefficients in columns 2 and 3 are quite
similar to those in columns 5 and 6. This suggests that the institutional change in 1959-60—the release of vault cash—has not altered
the explanatory power of the doctrine considered here.
TAHLE V I I - L . — M e a s u r e s of association
between
changes
and changes in the extended
base
Period: 1st quarter of 1940 to 4th
quarter of 1953
Coefficient of determination using—
Changes in the money
supply
Changs in money supply
Pins time deposits
The
extended
base only
The
modified
base doctrine
(1)
(2)
in the
money
supply
Period: 1st quarter of 1958 to 4th
quarter of 1962
Partial
Coefficient of determinaPartial
tion using—
coefficient
coefficient
of deterof determination for
mination for
The
The
modified
modified
modified base doctrine
base doctrine extended
base only base doctrine
(4)
(3)
(6)
(5)
0.663
0.956
0.914
0.602
0.957
.607
.064
.914
.SO
.985
0.S08
.908-
T H E EFFECTS OF T H E PRINCIPAL DETERMINANTS ON T H E M O N E T SUPPLY
The treatment of the public's demand for currency is one of the
principal differences between the modified base conception and the
views expressed by spokesmen for the Federal Reserve. Their discussion of currency movements is usually restricted to seasonal variation in currency flows. At times, references are made to secular
changes in currency demand. But the Federal Reserve apparently
has never noted either the cyclical component in currency movements
or the spillover into currency that is a part of the multiplier
mechanism.
Statements by Federal Reserve officials clearly deny the relevant
operation of the currency drain as a part of the multiplier process.
Their statements assert that the monetary multiplier for the banking
system is the reciprocal of the average reserve requirement ratios.
This implies that the monetary multiplier has been at least 6 and no
more than 24 during most of the postwar period, the precise value
depending on the marginal allocation of newly created deposits between demand time accounts as we have previously indicated.
The results in tables V I I - 2 and V I I - 3 present information on the
size of the monetary multiplier and hence on the operation of currency spillovers. I n all cases, we find a multiplier value substantially
4 6
ALTERNATIVE APPROACH TO T H E M O N E T A R Y
MECHANISM
less than the minimum value of six, implied or stated in Federal
Reserve publications. The values thus support our contention that
portfolio adjustments by the banks, induced by changes in surplus
reserves, are typically associated with a partial conversion of the
newly created deposits into currency. A billion dollars of open market purchases, or the release of a billion dollars from required reserves
by reduction in the requirement ratios, cannot be expected to raise
the money supply by more than $6 billion as the Federal Reserve has
contended.
The computed values of the multiplier are smaller for the money
supply than for the money supply plus time deposits. This, again,
is in accord with the underlying modified base conception. The multiplier mechanism incorporates a spillover into time deposits as well as a
spillover into currency. The effect of this spillover is to lower the expansion of the money supply, and to raise the expansion of the money
supply plus time deposits, per dollar of surplus reserves.
The computed multipliers show the effect of a dollar of base money
(table V I I - 3 ) or a dollar change in base money (table VII-2) on
the two measures of the money supply or changes in the supply.
The extended base incorporates the open market portfolio of the
Federal Reserve banks, the volume of member bank borrowing, float,
and the cumulated sum of liberated reserves. In the past few years,
the amount of vault cash liberated under the new i n s t i t u t i o n a l arrangements is included in the extended base. Table V I I - 2 provides
estimates for the two overlapping periods described earlier. The
values of the multipliers computed from the simple base c o n c e p t i o n
are relatively close for the two periods. This suggests that the change
has had little or no effect on the magnitude of the monetary multiplier,
as the base conception implies.
TABLE V I I - 2 . — T h e response
of
the change
the extended
in the
base
money
supply
to changes
in
Values of the monetary
multipliers
Period: 1ft
quarter 1949
to 4th nuarter
1958
Chantie in the money supply
Change in the money supply plus time deposits
2.
2.818
Period: 1st
quarter 1M9
to 4th quarter
1962
2.939
The estimates in tables VTI-2 and V I I - 3 permit a comparison of
the value of the multipliers computed from the relation of the base to
the money supply with those obtained from computations of the effect
of changes m the extended base on changes in the money supplv. F° r
the money supply, the two sets of estimates provide almost identical
values; for the money supply plus time deposits, the estimates differ
slightly but suggest that the multiplier is larger when time deposits
are added to the money supply. This comparison again indicates the
relevance of the modified base doctrine. Moreover, we note that the
partial coefficients of determination in table V I I - 3 reveal the important influence of the base on the stock of money.
47
ALTERNATIVE APPROACH TO T H E MONETARY M E C H A N I S M
TABLE VII-3.-—TFTC response of the money
the extended
base, 1st quarter
supply to the separate components
1948 to 4th quarter 1959
of
Multiplier—
For the money supply:
Multiplier resr>on$e
Partial coefficient of determination...
tor the money supply plus time deposits:
Multiplier response
Partial coefficient of determination...
For the base
For the sum
of liberated
reserves
2.50
.884
2.53
2.64
.884
2.67
.823
An important feature of the modified base doctrine is the response
of the money supply to the base and the accumulated sum of reserves
liberated or impounded by changes in reserve requirements or redistributions of deposits between classes of banks. In most of our tests these
two components are combined to form the extended base. Table V I I - 3
provides estimates of the multipliers applicable to each component^
a test of the procedure that we have used. The results support our
assertion that the effect on the money supply of a dollar of base money
or of a dollar of liberated reserves are the same. A similar result is
found when time deposits are added to the money supply. Moreover,
the partial coefficients of determination suggest that variations in the
money supply explained by synchronous variations in each of the
separate components of the extended base, other elements in the modified base conception unchanged, are of approximately the same order
of magnitude. Again, the evidence supports the procedures used in
the development of the modified base doctrine.
We noted in chapter I I that a leading spokesman for the Federal
Reserve has indicated that changes in reserve requirement ratios have
only seven-eighths of the effect of open market operations that supply
the same volume of excess reserves.4 They have presented no evidence in support of this assertion. Our evidence denies their contention and suggests that the effects of the two are the same, as the base
conception implies.
Some additional implications of the modified base conception can be
evaluated. Our discussion indicated that the public's currency and
time deposit behavior influenced the money supply in two separable
w
ays. One component has been incorporated in the multiplier mechanism ; the second set of demand components for currency and time deposits occur independently of the banks' portfolio adjustment and the
^multiplier. Six statements, or propositions, summarize some of the
main implications of the modified base conception with respect to the
second demand components for currency and time deposits.
(1) The reallocation of $1 from demand deposits to currency lowers
both the money supply and the money supply plus time deposits.
(2) Tlie negative effect of the increased demand for currency is
smaller for the money supply than for the money supply plus time
deposits.
„ 4 w . W. Riefler. "Open Market Operations In Long-Term Securities," Federal Reserve
Bulletin, vol. 44, No. 1U
48
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
(3) A dollar shifted from currency to demand deposits has a smaller
effect on both measures of the money supply than a billion dollar
change in the extended base.
,
(4) A dollar of deposits shifted from demand to time deposits has
a negative effect on tlie money supply and a positive effect on the
money supply plus time deposits. The effects of the shift on the two
measures, though in opposite directions, will be of approximately equal
size.
. _
(5) The reallocation of a billion dollars from demand to time deposits has a smaller effect on either measure of the money supply than
a billion dollar shift from demand deposits to currency.
(6) If a billion dollars of demand deposits is reallocated to the time
deposit account, the sum of the decrease in the money supply (ignoring
signs) and the increase in the money supply plus time deposits will
be slightly greater than $1 billion. According to the base conception^
the sum of the two responses will be approximately $1,050 million
under prevailing institutional and behavioral arrangements.
These six propositions implied by the modified base conception can
be compared to the results shown in table V I I - £ . To test the propositions, we have used annual changes in the relevant magnitudes from a
given quarter to the corresponding quarter in the following year.
The evidence supports the six propositions quite well. We will discuss the evidence for each of the six propositions in the sequence in
which the propositions were presented.
(1) An increase in the public's demand for currency lowers both the
money supply and the money supply plus time deposits. This confirms
proposition 1.
(2) A $1 bilVon increase in the demand for currency lowers the
money supply plus time deposits by approximately $2.7 billion on the
average. This is almost $300 million larger than the average effect on
the money supply as indicated by proposition 2.
(3) Only a $1 billion change in currency demand has a smaller
effect on each measure of the money supply than a $1 billion change in
base money, as stated in proposition 3.
(4) A $1 billion shift of deposits from demand to time account
lowers the money supply $590 million and raises the money supply
plus time deposits by $450 million on the average. The effects are in
opposite direction for the two measures and are of a p p r o x i m a t e l y
equal size. This result clearly reveals that when a dollar is reallocated
from demand deposits to time deposits, the money supply ( e x c l u s i v e
of time deposits) contracts by less than $1. The reallocation releases
surplus reserves that stimulate the expansion of additional money by
the banking system, as the base conception implies. This e x p a n s i o n
partially restores the money supply (or the change in money supply)
to its previous position.
(5) The descending order of the responses to changes in the extended base, changes in the demand f o r currency and changes in the
demand for time deposits are supported by the e v i d e n c e for both
measures of the money supply. This is a major implication of the
modified base conception and reveals the power of the base doctrine
to indicate the approximate quantitative effects of various monetary
operations on the money supply.
(6) Ignoring signs, the sum of the effect of a $1 billion change in
the demand for time deposits on the two measures of the change in
49 A L T E R N A T I V E A P P R O A C H TO T H E M O N E T A R Y M E C H A N I S M
the money supply is $1,043 million on the average. Again, the approximate value is extremely close to the magnitude implied by the
theory.
Thus the evidence supports some of the major qualitative
and quantitative implications of the base conception. Where the elements of the modified free reserves conception have at best a tenuous
link with the change in the money supply, the results in this section
reveal that the major changes in the money supply in the postwar
period conform to the patterns implied by the modified base doctrine.
TABLE V I 1 - 4 . — T h e magnitude
changes
in each of
1st quarter
lD.)9-},th
of the change
the elements
of the
quarter
VJW
of the money supply in response
to
modified
base
conception—Periodf
[Billions of d o l l a r s ]
Response
gesponse
response
gespon.se
response
to
to
to
to
to
a
a
a
a
a
billion dollar chanpe in tlie extended base
billion dollar shift irom demand deposits to currency
billion dollar real! cation r {depesits to time account
change in t ( i e Treasury bill rate bv l percentage point
change in the discount rate by 1 percentage p o i n t . . .
Money supply
Money supply
plus time
deposits
$2.478
~2.423
-.592
4*. 656
-.2.9
$2,755
-2.692
+.451
+.728
-.243
inf! l h c o r y ? f t h e money supply that we have presented incorporates
1,1 i n U e s 111 t l i e monetary mechanism along with the factors that
1 considercd in
for
this section; the extended base, the demand
^or currency, and time deposits. Variations in interest rates operate
renll ™ o m y s l l PP!y through three distinct channels: (1) via the
eailocation of deposits between demand and time accounts; (2) via
tne
desired excess reserve position; (3) via the banks' desired
•Borrowing from the Federal Reserve banks. The operation of the
Jirst and third channels has been implicitly incorporated in the mone«try process by procedures that were discussed earlier. The operation of interest rates on the allocation of deposits between demand
and time account is implicitly recognized through the second demand
component for time deposits. The working of interest rates on the
remand for borrowed reserves has been incorporated through the
volume of member bank borrowing that is treated as a source of the
monetary base. Under an alternative procedure, the base is adjusted
D
y the removal of member bank borrowing as a source and borrowed
reserves as a use. The demand for borrowed reserves is then incorporated through the interest mechanism.
i n the present case, interest rates operate through the second channel only—tl ie demand by banks for excess reserves. The results are
shown in lines 4 and 5 of table V I I - 4 above where the computations
f f 6 based on changes in interest rates. We find that an increase in
Treasury bill rate lowers the banks' demand for excess reserves and
a
dds to both the money supply and the money supply plus time deposits. A rise in the rediscount rate lowers both measures of the
money supply. However, the results suggest that an equal increase in
the Treasury bill rate and the rediscount rate leads banks to reduce
ex
2?f s r e s erves and increase the money supply.
. -I he response of the change in the money supply to changes in
mterest rates permits a comparison between policy operations dei g n e d to raise interest rates with changes in the monetary base. For
50
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
example, an increase of Treasury bill rates by 1 percent, for example
from 3 to 4 percent, has an effect on the money supply that is approximately equivalent to a $250 million increase in the base. A rise in the
rediscount rate by 1 percent is approximately equivalent to a reduction
in tho monetary base of $80 million.
.
#
These numerical examples bring out some essential points. Ihe
Federal Reserve opposes flexible discounting arrangements and favors
infrequent flat changes in the discount rate. A t times their spokesmen
have suggested that variations in tlie discount rate have important
effects, particularly announcement effects, and that a floating discount
rate would lessen their degree of control. Yet they do not hesitate to
increase and decrease the monetary base by several hundred million
dollars through their defensive operations. These defensive operations have a much larger effect on the money supply than the conceivable effect imposed by the likely variation in the discount rate.
Moreover, our results suggest that the net effect of an increase in
Treasury bill rates and the rediscount rate by 1 percent raises the
money supply by less than $450 million. If the Federal Re*
serve is concerned with the behavior of the money supply, the effect
of these interest rate changes on the money supply can oe offset by a
small reduction in the monetary base, one that is well within the typical daily or weekly variation in their open market operations.
Viewing the same example in another way makes clear that the
Federal Reserve's attempt to reduce the seasonal variation in interest
rates by relatively large variations in the base lias consequences for
the behavior of the money stock. This should not suggest that the
allocative effects of seasonal changes in interest rates should be permitted. But it does suggest that analysis and evidence is required
to appraise the effects of the operations that are undertaken.
The estimates of the response of the money supply to variations in
Treasury bill and rediscount rates can be used' to derive estimates of the
response of excess reserves to changes in the two rates. The sensitivity
of changes in excess reserve to interest rate changes can be expressed in
term of a concept called elasticity. The elasticity of the banks' desired
excess reserves with respect to Treasury bill or discount rates describes
the percentage change in desired excess reserves induced by a given
percentage change in either rate. The estimated value of the elasticity
of desired excess reserves with respect to Treasury bill rates is approximately minus nine-tenths and with respect to the discount rate
plus three-tenths percent. This means that a 10-percent rise in the bill
rate (from 3.3 to 3.63 percent) lowers desired excess reserves by 0 percent on the average. A similar increase in the discount rate raises desired excess reserves by 3 percent. Thus, a 10-percent change in both
rates changes desired excess reserves by approximately 6 percent on the
average. The resulting accrual of surplus reserves at c o m m e r c i a l
banks induces changes m the money supply as we have noted.
The banks* liquidity trap
The operation of the interest mechanism introduces a d e p e n d e n c e
of the money supply on the pace of economic activity. This dependence modifies the influence of the policy magnitude, expressed bv the
extended base, m the money supply process. But it neither removes
the dominant role of the extended base nor breaks the effective connection between the money supply and its principal determinant As
we have just seen, the dominant role of the extended base r e m a i n s
51
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
when the effect of interest rates is incorporated in the monetary
mechanism.
It has been asserted at times that the effective connection between
Federal Reserve policy and the money supply is broken during periods of comparatively low interest rates. This is the contention of
those who suggest that monetary policy can be a matter of pushing
on strings. In this view, the attempt to expand the base by increasing the supply of reserves is matched by an insatiable willingness by
banks to acquire and hold reserves. Each addition to the supply of
reserves is matched by an identical increase in quantity demanded. Increases in reserves do not set off the multiplier mechanism leading to
portfolio expansion and an increased money supply. Reductions in
available reserves do not lead to contractions in the stock of money.
The banking system passively holds any additional reserves generated
by an expanding base or an outflow of currency to the banks, and passively surrenders reserves absorbed by a contracting base or a currency
outflow to the public. Monetary policy becomes entirely ineffective;
the money supply is determined by the public's asset supply to banks;
the monetary system is caught in the liquidity trap.
The public's supply of assets to banks depends on the pace of economic activity. If the liquidity trap is operating, the money supply
becomes dependent on the public's desired supply of assets to banks;
hence it too depends on the pace of activity. The onset of a liquidity
trap thus implies that the "push" exerted by monetary policy operating through the base is replaced by the "pull" of national income as
the principal determinant of the money supply. Indeed, as Ave have
noted, the liquidity trap implies that the Federal Reserve authorities
have no control over the money supply. Under these circumstances,
the Federal Reserve must be absolved from responsibility for the
growth rate of the money supply.
The liquidity trap or "pushing 011 strings" notion is frequently
invoked to describe or explain the monetary events of the thirties. In
the previous chapter, we produced some evidence relevant to the question and showed that the events of the period are incompatible with
the asserted break in the link connecting Federal Reserve policy with
the money supply. Our evidence suggested that the doubling of
reserve requirements in 1936-37 was responsible for the reduction in
the extended base and the consequent reduction in the money supply.
Additional evidence in this chapter reinforces that conclusion.
Table V I I - 5 summarizes some of the evidence from detailed tests
of the liquidity trap notion. By incorporating the effect of the public's asset supply to banks—assumed to depend on national income—
as a part of the money supply process, we obtain a relation indicating
the dependence of the money supply on national income and the extended base. If the liquidity trap notion is well-founded, the money
supply depended exclusively on national income during the thirties.
The extended base had 110 role in the money supply mechanism. However, if the modified base doctrine is the more appropriate theory, the
extended base exerted a dominant influence on the money supply; the
influence of national income reflected only the feedback from income
to the money supply and was of little consequence in the monetary
mechanism.
I t should be noted, however, that the modified base doctrine implies
that the response of the money supply to policy action will depend on
the prevailing level of market interest rates. The monetary multiplier
52
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
for the extended base is smaller in periods of low interest rates than
in periods when interest rates are higher. A given decline in interest
rates will induce banks to add a larger amount to desired excess
reserves when interest rates are low than when rates are at higher
levels. The response of desired excess reserves to variations m deposits is dependent on the level of interest rates also. Additions to
excess reserves per dollar of additional deposits are smaller in periods
of comparatively high interest rates, larger in periods of relatively
low rates. Both implications operate to reduce the expected change
in the money supply per dollar of surplus reserves in a period of low
interest rates.
TABLE V I I - 5 . — T h e
comparative magnitude of the push of the extended
the pull of income on the money supply
base ani
[In billions]
Deflationary environment
1st quarter 1929 to 4th
quarter 1940— Response
of money supply to a
billion dollar change in—
Extended
base
Money supply
Money supply plus time deposits
$1.25
1.32
National
income
&X06
.07
Nondeflationary environment 1st quarter 1951 to
4th quarter 1959-Hesponge of money supply
to a billion dollar change
inExtended
base
$2.74
2.89
National
income
$0.03
.03
The data in table V I I - 5 permit a comparison of the policy multipliers computed for the thirties and the fifties and of the c o m p a r a t i v e
effects on the money supply of the extended base and national income.
Note that the "push" of the extended base dominated the "pull" of
national income in both periods. An increase in the extended base of
$1 billion raised the money supply by $1.25 billion during the thirties.
The pull of income was much smaller. A $1 billion increase in national income called forth only $60 million of money supply on the
average. In the high-interest-rate regime of the fifties, the*push of
the extended base is much larger, and the pull of income much smaller,
as the base conception implies.
The evidence is incompatible with the "liquidity t r a p " notion. The
money supply in the thirties does not appear to have been d o m i n a t e d
by national income operating on the public's asset supply to banks.
The influence of Federal Reserve policy was not eliminated. Although
the Treasury bill rate remained at a level only slightly above zero for
long periods of time, the Federal Reserve remained capable of inducing
a multiple expansion of the money supply by increasing the monetary
base. Thus the evidence denies the relevance of the liquidity trap
notion and suggests that monetary policy remained effective despite
persistently low interest rates in the thirties.
Some predictions of the money supply
The results presented earlier in this chapter reveal that the modified
base conception is an approximative theory of the money supply that
is capable of explaining the major changes that have taken place. Such
evidence is an important means of distinguishing between valid and
invalid theories, but it is not the only tvpe of evidence that can be
adduced. This section presents some additional data to suggest the
usefulness of the modified base conception as a tool for predicting
values of the money supply or the change in the money supply.
53
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Observations for the change in the money supply and its approximate determinants for corresponding quarters in adjacent years were
used to infer the relation between the change in the determinants and
the change in the money supply. Data for the period beginning in the
first quarter 1949 and ending in the last quarter 1958 were used for
this purpose. Predictions of the change in the money supply and the
value of the stock of money were then made.
Two measures summarize the results of the predictions. The average percentage error indicates the magnitude of the error expressed
as a percentage of the change in the money supply or its level. The
standard deviation measures the extent to wThich the percentage errors
cluster around the average. The larger the standard deviation, the
looser the cluster. A large standard deviation and a small average
percentage error indicate that on the average the percentage error is
not large but that in any particular quarter one might expect to find
an error substantially larger than the average. For example, a sequence of estimates that first overestimate and then underestimate the
change in the money supply by approximately equal percentages would
produce a small average percentage error and large standard deviation.
Table V I I - 6 is divided into two parts. The first section shows the
average percentage error and the standard deviation for the predictions of the money supply, the money supply plus time deposits, and
the changes in the two measures of the stock of money. The second
part of tlie table presents the percentage errors for each of the 16
quarters following the period used to infer the estimates.
TABLE V I I - G . — T h e
predictive
performance
of the modified base doctrine
A. A V E R A G E AND STANDARD DEVIATION OF P E R C E N T A G E E R R O R S FOR P R E D I C T I O N B E G I N N I N G 1ST QUARTER OF 1959 AND E N D I N G LAST QUARTER
OF 1962. T H E E S T I M A T E S USED FOR T H E P R E D I C T I O N W E R E D R A W N F R O M
T H E P E R I O D 1ST QUARTER 1949 TO 4TH QUARTER 1958
Average percentage error,
error of predictions
Standard deviation of percentage error
Without
proper
inclusion of
vault cash
release
With proper
inclusion of
vault cash
release
Without
proper
inclusion of
vault cash
release
-0.5
-.3
-25.1
-21.3
0.5
.4
10.4
.3
1.7
1.1
231.0
48.1
Money supply plus time deposits...*
Change in money supply...
Change in money supply plus time deposits—
With proper
inclusion of
vault cash
release
B. S E Q U E N C E OF P E R C E N T A G E E R R O R S I N P R E D I C T I O N FOR 16 QUARTERS
FOLLOWING T H E P E R I O D USED TO I N F E R T H E ESTIMATES
Money
supply
Period
0.4
.5
.1
.6
.4
.5
1st quarter, 1959.
2d quarter 1959..
3d quarter 1959,.
4th quarter, 1959.
1st quarter, 1960.
2d quarter, I960..
3d quarter, I960..
4th quarter, 1960.
34—474—64
-.1
1.2
5
Money supply plus
time
deposits
-0.1
.2
.1
.5
.3
.3
.3
.1
Period
1st quarter, 1961.
2d quarter, 1961..
3d quarter, 1961..
4th quarter, 1961.
1st quarter, 1962.
2d quarter, 1962..
3d quarter, 1962..
4th quarter, 1962,
Money
supply
-0.1
-.2
.6
-1.0
.6
1.4
1.6
1.4
0.6
.4
65.9
29.4
54
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
The period for which the predictions were made contained an important change in institutional arrangements. In a scries of steps
in 1959-60, the Federal Reserve permitted banks to count some, and
later all, vault cash as a part of reserves. These changes in arrangements occurred after the period used to infer magnitudes for the
monetary multiplier and other values (parameters) that link the
behavior of the public and the banks to the money supply. An additional test of the modified base doctrine can, therefore, be made since
the base doctrine indicates the manner in which the release of vault
cash should be incorporated. Two sets of estimates are provided.
One set includes the released vault cash as a part of the extended
base as is required by our theory; the other ignores the influence on
the stock of money of the release of vault cash. In effect, the second
procedure assumes that the Federal Reserve's action had no influence
on the money stock while the first implies that the volume of reserves
released through the institutional change is equivalent to a reduction
in reserve requirements or an open market purchase that supplied
the same amount of surplus reserves.
Inspection of table VI1-6 shows that the average predictive error
for the money supply did not exceed one-half of 1 percent. The
proper inclusion of vault cash did not affect the size of the average
error, but it substantially lowered the variability of the errors in
predicting the money supply measured by the standard deviation.
The standard deviations are three times smaller when the released
vault cash is included as a part of the extended base after the institutional change.
The substantial improvement achieved by incorporating the vault
cash release as a part of the extended base is most clearly revealed by
the percentage errors applicable to predictions of the change in the
two money supply measures. The average error falls from 25.1
to 10.4 percent for the change in the money supply and from
21.3 percent to a negligible three-tenths of 1 percent for the money
supply plus time deposits. The standard deviations of the p e r c e n t a g e
errors also fall markedly.
The complete sequence of errors for the 16 quarter!v predictions
is shown in the second part of table VII-6. We note that the errors
tend to be somewhat smaller for the money supply plus time deposits.
There also appears to be a tendency for'larger than average errors
to occur near the end of the period. "Nevertheless, the r e s u l t s indicate
that the modified base conception supplies a substantially more reliable
conception of the money supply process than has been suggested by
the Federal Reserve authorities.
Two additional sets of computations have been used to e v a l u a t e
the predictive accuracy of the modified base doctrine. Table VII-7
compares the direction of predicted and actual changes in the money
supply. Table V I I - 8 compares the acceleration and deceleration of
predicted and actual money supplv. The hitter l-esults pertain to the
changes in the quarterly changes m the money supply. That is, we
attempt to predict the rate of change of the change in the money
supply. If the money supply rises by $500 million in one period and
by $300 million m the next, the change is positive in both periods.
55ALTKKNATIVKAPPROACH TO TIIF. MONETARY MECHANISM
but the change has decelerated since the money supply is rising at a
slower rate.
The rank correlation coefficient has been used to express the degree
of association between predictions and observations. This measure
of association is based on the orders of magnitude rather than the
numerical values. If the extended base conception predicts a large
positive change in the money supply or its rate of change and such a
change occurs, the rank correlation is increased. Conversely, if large
positive predictions occur when there are small positive or large negative changes, the rank correlation is reduced. A rank correlation of
1 indicates a perfect association j i.e., the largest predicted change occurred in the quarter experiencing the largest observed change; the
second largest prediction is made in the quarter with the second
largest, change, etc. A value of —1 indicates perfect inverse association. The largest predicted change is made in the quarter experiencing
the smallest actual change. If the rank correlation is zero, there is no
association between predicted and actual changes or predicted and
actual acceleration and deceleration.
Once more the predictions were made with and without proper incorporation of the vault cash release. The incorporation of vault cash
as a part of the extended base substantially raises the rank correlation.
This is particularly true for the predictions of the money supply that
is defined exclusive of time deposits. But for both measures, the association is considerably closer when the implication of the base conception with respect to vault cash is followed.
Table V I I - 7 indicates the presence of a strong positive association
between predicted and actual changes in the money supply, when
vault cash is properly incorporated. This association is somewhat
stronger when time deposits are included as a part of the money
supply, but in both cases, the data suggest that predictions of large
positive and negative changes occur in periods experiencing such
changes. Moreover, we note that the association between predicted
and actual changes is approximately twice as large as the association
between the changes in the two measures of the money supply. This
suggests that the relatively high rank correlation for the two measures
of the money supply is not the result of a common movement in the
money supply and the money supply plus time deposits. That the
modified base conception is able to predict the changes more accurately
than the simple explanation that the two move together, adds to our
confidence in the underlying conception.
measure of the closencss of association between
changes and observed changes in the money supply
TABLE V L I - 7 . — A
predicted
K E N D A L L ' S R A N K C O R R E L A T I O N IS U S E D AS OUR M E A S U R E O F A S S O C I A T I O N
Without proper
Inclusion of
vault cash
release
Association between predicted change and actual change of money
Association between predicted and actual change of money supply plus
time deposits _
„
_
With proper
inclrsion of
vault cash
release
0.56
0.81
.76
.89
NOTE—Association between change in the money supply and changes in the money supply plus time
deposits Is 0.43.
56
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
The values in table VII-6 were computed to predict the annual
change from a particular quarter to the corresponding quarter of the
following year. These predictions were based on estimates of the
structure of the money supply relation computed from quarterly data
for the period 1949 to 1958. The change in the predicted change between adjacent quarters was then computed and compared to the
change in the actual change (acceleration or deceleration). These
data were then ranked, and the rank correlations was used to measure
the association. The results are presented in table YII-8.
TABLE VII-8.—Measure of the association
of the changes between adjacent
quarters of changes in the predicted and observed values of the money supply
between corresponding
quartersf acccleratio71 and
deceleration
Without
proper inclusion of
vault cash
release
For the money supply
For the money supply plus time deposits.
*
-
0.58
.70
With proper
inclusion of
vault cash
release
0.79
.83
NOTE.—Association between the acceleration of the money supply a n d t h e money supply plus time deposits is 0.68.
Again we find a relatively close association between predicted and
actual observations. The clear, positive association between changes
of changes reveals the strong relation between the money supply and
the elements of the modified base conception. As before, the degree of
association is markedly larger when the vault cash release is incorporated in the extended base.
Currency behavior and money supply
The description of the injection mechanism and multiplier process
emphasized the importance of currency flows between the public and
the banks. The multiplier process was seen to generate a spillover of
newly created deposits into currency. This spillover helps to explain
why the observed magnitude of the monetary multiplier is substantially below the reciprocal of a weighted average of reserve requirement ratios. Currency flows also result from variations in the public's
demand for currency that occur independently of changes in monetary
wealth. These flows generate or absorb surplus reserves and trigger
the multiplier mechanism. The evidence adduced in previous sections
confinned the operation of currency patterns in both the multiplier
and injection mechanisms. Our observations thus support the contentions implicit in the modified base doctrine that ascribe a substantial role to currency behavior in the money supply process. This role
has not been fully appreciated by the Federal Reserve. I t appears
useful, therefore, to investigate more fully some major behavior features of the public's demand for currency.
Under the prevailing institutional arrangements the volume of currency outstanding, the amount issued by the Treasury and the Federal
Reserve, is determined by the public's demand. The supply quantity
is thus identical with the demand quantity, which is shaped by the
public's monetary and nonmonetary wealth, and the costs and yields
associated with currency holding or currency using. These arrange
57
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
ments have not always prevailed. A t times, the supply of currency
has had an important influence on the monetary mechanism. But
under our contemporary institutions, the observed changes in "currency outside banks77 can be safely attributed to the operation of the
public's demand for currency.
Throe aspects of the public's demand for currency can be usefully
separated—a seasonal movement, a cyclical pattern, and a secular
drift. The seasonal movements have attracted substantial attention
from the Federal Reserve. Passing references are made to secular
drift in the public's currency demand, but a clear description of the
nature of the movement has not been provided. The cyclical pattern
has been disregarded. We found no references or descriptions of the
cyclical demand for currency in Federal Reserve discussions.5 Nevertheless, the cyclical pattern is an important feature of the public's
demand for currency and has been incorporated in both the multiplier
and the injection mechanisms. The observation of a cyclical component in the movement of "currency outside banks" thus provides
important evidence on the operation of currency flows in the monetary
process through the two mechanisms.
In the absence of both cyclical and secular behavior components the
public's demand for currency would remain constant, except for seasonal variations. Under these conditions, changes in the volume of
currency between corresponding months of successive years would tend
to vanish, unless there is a change in the seasonal pattern. But we
observe the regular occurrence of significant changes between corresponding months. This indicates that either secular or cyclical forces
(or both) operate on the public's demand for currency.
Changes in wealth, costs, and yields associated with currency holding or currency use modify the public's demand for currency. For
example, we notice a sweeping secular drift before World War ! expressed by a persistent decline in the demand for currency relative to
the demand for total deposits. This decline can very likely be attributed to a reduction in the cost of banking, particularly the cost of
access to and information about banks. The reduction m the "cost
of banking" lowered the yield associated with currency holding or
currency using and led to a relative decline in the public's currency
demand. These longrun influences on the public's currency demand
worked their effects through the injection mechanism of the money
supply process and contributed to an acceleration of the longrun
growth in the money supply.
Scrutiny of the data for more recent years provides no evidence of
the continuation of the r e l a t i v e decline in the relative demand for currency. Both W o r l d W a r s raised the yield associated with the noMing
of currency and thus u n l e a s h e d a remarkable increase m the public s
relative (and absolute) demand for currency. The high yield on
currency subsided rapidlv with the termination of the tyro
both wars were followed by a speedy decline m the relative (and even
t h e B o a r d of Governors n n d t h e «
J ^ M ^
a h n u t t h e c u r r e n c y problem as a p a r t of t h e J ^ L ™ ™ ^
™ „ mixture of
Their answers t o question 1. pt. B. f o r e s t t h a t t h e y ™ £
w l I c H chnnres in the
a n d secular cbantres. T h e y make no
t*1^
f o r rnrrencY. Moreover, the Board'? replv Isiqu te « p l f « t
rV^rvp positions
to offset " w i d e b u t f a i r l y predictable ^ r ^ ' L ^ I n d f o r c n r r e p c r mlnus the reduction In
a r e increased bv t h e a m o u n t of increased demand for• a i r w c y «« p I a c e d CTL t h e free
s q u i r e d reserves
Thi* Is additional evidence of the reliance t h a t is pmcea
r e s e r v e s d o c t r i n e in d e t e r m i n i n g policy actions.
58
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
absolute) demand. But this relative decline was quickly attenuated
and eventually disappeared. I n recent years the public's relative
demand for currency was more than 60 percent above the corresponding demand in 1929. Thus there exists no evidence for the persistence
of the secular decline of the public's relative demand for currency in
this century. The disappearance of the secular d r i f t implies that the
changes in the public's currency holdings between corresponding
months of successive years must be attributed to the operation of
cyclical forces.
To pursue our investigation of the role of currency patterns in the
money supply process, we obtained annual changes of currency outside banks and of the extended base for each month of the year. The
change in currency for any month was computed as the difference between the amount outstanding in that month and the amount outstanding in the corresponding (same) month of the previous year.
Identical methods were used to compute the annual change in the base
from month to corresponding month. The ratio of the change in currency to the change in the base was then obtained, and this ratio was
used to observe the prevailing patterns. Data for each postwar year
and for each postaccord half-cycle arc presented in table VII-9.
These figures are averages of the ratios obtained for each month.
A negative ratio for any month can only occur when the change in
currency and the change in the base are in opposite directions. Occurrences of negative ratios are, however, more usefully described for our
purposes in terms of "overcompensation of currency changes by
changes in the base." Such overcompensation is characterized by two
conditions: (1) changes in currency and bank reserves exhibit opposite signs, and (2) the magnitude of the change in bank reserves exceeds the magnitude of the change in currency. This specification implies that overcompensation of currency changes is a necessary and
sufficient condition for the occurrence of a negative value for the ratio.
I t follows that annual averages or half-cycle averages of the ratios
are negative only if overcompensation dominates the changes in the
base. Moreover, it can be shown that overcompensation is typically
associated with a reallocation between currency and demand deposits
that reinforces the effect on the money supply of the simultaneous
change in the base.
Some examples may clarify the foregoing discussion. October, November, and December 1953 show increases in currency holdings of the
public from the corresponding months of 1952. The ratio of the
change in currency to the change in the base is negative for these 3
months. This means that the decline in member bank reserves from
1952 to 1953 in each month during the autumn of the year was larger
than the rise in currency. The base contracted while its currencv component expanded. The fall in the base occurred despite the fact that a
recession had started in the summer of 1953. The reduction in reserve
requirements in July 1953 explains some of the observed change in
member bank reserves. But the change in reserve requirements does
not justify the negative change in total reserves if the Federal Reserve
is trying to offset the effect of seasonal changes in the supply of money.
Neither does the onset of recession explain the reduction in reserves of
member banks. Indeed, one would expect that the change in reserves
would be positive for that reason alone, if the Federal Reserve pursues a countercyclical policy.
59
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
The 3*ear 1054 provides a somewhat different example. The annual
change in currency from month to corresponding month was negative
between autumn 1053 and autumn 1954. Currency held by the public
declined from October, November, and December 1953 to the same
months in 1954. But the ratio of the change in currency to the change
in the base was positive during this period indicating a decline in bank
reserves. I n fact, the Federal Reserve reduced bank reserves from
each month of 1953 to the corresponding month of 1954. This added
further to the annual reduction m the monetary base between corresponding months of the 2 years. Again, there were reductions in reserve requirements; in this case reserve requirements were reduced for
both time and demand deposits.
The increase in the demand for currency in autumn 1953 and the
decline in 1954 is surprising, if the demand for currency is constant
except for seasonal variations. Although there has been little or no
recognition of cyclical changes in the demand for currency by the
Federal Reserve, the data for the autumn of 1953 or 1954 suggest
that such cyclical changes do occur.
TAHLE VI1-9.—Currency patterns in the postwar
period
P A R T A. ANNUAL AVERAGES
Year or period
1946..
1947..
1948..
1949..
1950.
1951.
1952.
1953..
The ratio
of changes
in currency
to changes
in the base
Direction of
change in the
monetary base
through most
of the year
+0. 443 Rising
do
-.519
-.599 .-...do
-.117 Falling
do
-.103
+ . 138 Rising
do
+.514
+.S64
do
19M..
1955..
+ . 303
+.742
Falling
No change.
1956..
1957..
+.548
+.329
Rising
do
1958..
1959..
+ . 187
+.860
FallingRising. .
1900.
1961..
+.C45
+.044
Falling
do
1962..
+.5S6
Rising
Remarks
11 months only.
ratio is 8.200.
For 12 months
11 months only.
ratio is -0.987.
For 12 months
11 months only.
ratio is -6.365.
For 12 months
11 months only.
ratio is 1.370.
For 12 months
11 months only. For 12 months
ratio is —0-574.
P A R T B. CYCLICAL AVERAOES
November 1948 to October 1949..
Octobcr 1949 to July 1953„
July 1P53 to August 1954
August 1954 to July 1957
July 1957 to April 1958
April 1953 to May 1960
May i960 to February 1961
-0.183
+.340
+ . 019
+ . 564
+.394
+ . 369
+.077
P e a t to trough.
Trough to peak.
Peak to trough.
Trough to peak.
Peak to trough.
Trough to peak.
Peak to trough.
The modified base conception implies that the ratios in the table
^vould be approximately one-half if the spillover into currency
through the multiplier process dominates the currency movements. A
positive ratio of less than one-half indicates that the reallocation between currency and demand deposits creates or destroys surplus
60
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
reserves in a direction opposite to the concurrent change in the base.
The public's demand for currency partially offsets Federal Reserve
policy operations summarized by the movement of the base. But in
the table we find that a ratio of approximately one-half for 1946
is followed by a series of negative ratios in the next 4 years. In 2 of
these years the base rose, and in 2 years the base fell. Since 1950,
overcompensation has not occurred to an extend sufficient to be apparent in the average ratio for the year as a whole. This observation
suggests that after the postwar readjustment, currency flows were
dominated, on the average, by the spillovers associated with the
multiplier mechanism. However, the reallocation between currency
and demand deposits appears to explain the variations in the ratio and
its deviation from the neighborhood of one-half.
An examination of the ratios in table V I I - 9 B yields additional
information about the cyclical role of currency in the money supply
process. If currency grows at a steady rate, without any cyclical
variation, the ratio would exhibit an inverse relation with the growth
rate in the base. The observed procyclical movement of the base
would imply that the ratio is lower in upswings than in downswings.
This implication of a steady growth rate in the public's currency
demand is quite inconsistent with the data in part B of table
VII-9. The average ratio for each upswing exceeds the average
ratio for the subsequent downswing.
Inspection of the ratios in part B of the table thus r e v e a l s some
disturbing indications of Federal Reserve policy. The ratio of
changes in currency to changes in the base is generally larger in
periods of expansion than in adjacent periods of recession. # The
change in currency is often negative in recession, but the ratios in
Sart B are generally positive. This again suggests that the Federal
teserve permits total member bank reserves to decline in periods
of recession. The fact that the ratio shown in the table is small
and positive further suggests that the decline in reserves is much
larger than the decline in the demand for currency by the public.
Examination of the data for reserves and currency confirms that during recession both components of the base often decline. For e x a m p l e ,
during the recession of 1960-61, member bank reserves and the monetary base were permitted to decline in every month from the level
maintained in the corresponding month of the previous year. This
behavior of the base is difficult to reconcile with an active, countercyclical monetary policy.
The cyclical pattern'in currency flows resulting from both the spillover and reallocation effects is revealed again in the chart a p p e n d e d
to this study. I n general, the movements observed during the postwar years have reduced the cyclical variation in the monev supply
generated by the procyclical policy pursued bv the Federal*Reserve.
Increased demand^ for currency puts increased pressure on member
bank reserve positions during expansions, since a rise in the public's
demand for currency reduces the expansion of the money supply permitted by a given value of the monetary base. I n view of the cyclical
variations in currency shown in the table and in the appended chart,
it is surprising to find no mention of cyclical changes in the d e m a n d
for currency in the Federal Reserve reply to question 1, part 5.
61
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
A D I S C U S S I O N O F S E L E C T E D P O L I C Y ACTIONS I N
1946-51
Two s t r i k i n g features of the actions taken by the Board of Governors and the Open Market Committee in the postwar period prior
to the accord can be analyzed in terms of the modified base doctrine.
These are the so-called pegging of Treasury bond prices and the use of
the power to change reserve requirements. Discussion of these topics
furnishes additional support for the view taken here that the Federal
Reserve has failed to analyze adequately the mechanism connecting
their actions with the supply of money or to consider seriously the influence of the stock of money on prices and employment. The Federal
Reserve's usual appraisal of the two policies is found to be seriously
deficient.
The two policies are closely associated. During much of the preaccord period, the Federal Reserve complained about the effect of
pegged interest rates and the reduction in (or elimination of) its
power to reduce inflationary pressures. Other devices such as the
reduction of Treasury balances at the member banks, retirement of
securities from the Reserve banks, and increases in required reserves
were relied upon to control the inflationary pressures. Yet during
much of the preaccord period, the rate of change of the money supply
was not large. Indeed, examination of the data for the period reveals
that the annual percentage change in the money supply declined almost continuously from the middle of 1946 through the end of 1948,
remained negative for the five quarters starting in the fall of 1948, and
did not exceed a 3-percent rate of increase from the beginning of 1947
to the start of the Korean war. F a r from suggesting that the Federal
Reserve was converted into an engine of inflation, the data suggest
that during much of the period, and particularly before and during
the recession of 1949, the Federal Reserve was pursuing a policy of
contraction.
Lest our discussion be misunderstood, we are not advocating inflation. The indexes of recorded prices did rise during the period. But
much of the price rise was a readjustment of recorded prices to the inflationary policy of war finance that had produced increases in the
money supply of as much as $5 billion per annum during the war.
These exceedingly large increases in the supply of money put upward
pressure on prices. With the removal of price controls, recorded prices,
began to increase at a faster rate. Most of this rise occurred between
the date the removal of price controls and the middle of 1948. In 1949,
the wholesale price index (base 1926) fell by 10 points. A renewed increase in prices began with the advent of the Korean war. But a
strong statement of intention by the Federal Reserve in the fall of
1950, culminating in the famous accord and the removal of the peg
from bond prices in early 1951, did not prevent a much more rapid expansion in the money supply than had occurred during the period of
pegged bond prices.
We do not wish to suggest that the policy of pegged bond prices
was either appropriate or desirable. Quite the contrary. Pegging
bond prices created a serious distortion in the allocative mechanism.
But our intention in this report is to focus primarily on the larger
questions raised by the failure of the Federal Reseire to> analyze,
understand, or examine the mechanism connecting their actions with
the stock of money and the economy. These failures are the major
62
ALTKKNATIVK APPROACH TO TIIF. MONETARY MECHANISM
reasons for the inappropriate policies that they pursued during the
preaccord period and afterward. Unless the Federal Reserve is
equipped to more adequately appraise their actions, critical comments
based 011 hindsight do little to improve future policy. In short, we
do not wTish to concentrate on the specific features of the inappropriate
policy of pegging yields or the reasoning that was used to support
that policy. Rather, we wish to examine briefly their statements and
the evidence during the period to appraise the repeated assertion that
the policy of "pegging*' was a major source of inflation and to evaluate
the use that was made of the power to change reserve requirements.
Pegged bond yields and inflationary
prcsswcs
During much of the preaccord period, yields 011 Government bonds
with 15 years to maturity remained below 2y 2 percent. Indeed bond
yields declined slowly during 1948 and quite steadily in 1049 fmd
early 1950. A t the time of the accord, intermediate "and long-term
Government bond yields were below the average level that had prevailed in 1948. We have already indicated that the annual percentage change in the money supply was below the percentage changein population during most of the period and that the return to "flexibility" in monetary policy in March 1951 wTas not accompanied by an
abatement of inflationary pressures emanating from an increased
stock of money. On the contrary, the annual percentage change in
the money supply from month to corresponding month was always
above 4 percent, and often above 5 percent, in the 18 months from
Janxiary 1951 through June 1952. Thus there does not appear to be
a prima facie case that the pegging of bond yields converted the Federal Reserve into an engine ofinflation or that the removal of the peg
eliminated the danger.
Turning to the specific details of Federal Reserve operations during the period, we find very little evidence that open market purchases
were feeding the inflationary forces. Table V I I - 1 0 shows total Federal Reserve holdings of Cxovernment securities during the period.
These data do not suggest that open market operations c o n t r i b u t e d
substantially to the inflationary pressures. Indeed, the System reduced
its holdings annually from month to corresponding month during
most of the period. The most notable exception occurs after the meeting of August 1950 when the problem of inflation was described as
a matter of "critical importance and urgency," in the Record of Policy
Action.
TABLE V I I - 1 0 . — R e s e r v e
Government
bank credit outstanding,
Federal Reserve holdings of
securities and gold in the preaccord
period
lln millions]
Date
December 1945
.
December 1 9 4 6 - . . . . . . . * . . . „ . „ . _ _
Jane 1947
December 1947
June 1948
.
December 1948
June 1949
December 1949^_„
June 1950.
December 1950
„
February 1951
Total Reserve bank
credit outstanding
$25,091
24,093
22,170
23,181
21,900
24,097
19,696
19,499
18,703
22,216
23,188
Free reserves
$743
650
763
752
663
658
685
699
885
297
Total System holdings
of Government securities
$24,262
23,350
21,872
22,559
21,366
23,333
19,343
18,885
18,331
20,778
21,881
Gold
stock
$20,065
20,529
2L2*
22,754
2532
24,244
24,466
24,427
24.231
22.70&
22,06*
63
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Interest rates on long-term Government securities declined slightly
during 1948, as we noted earlier. The Federal Reserve greatly increased its bondholdings during the year and more than offset the
decline in its holding of bills, certificates, and notes. The Board's
annual report for the year refers to the abatement of inflationary pressures. Credit is given to the Treasury surplus as the principal limiting element. In the opinion of the Board, the budget surplus was reinforced by the monetary policies pursued.6 This is in fact a correct
statement in terms of the modified base mechanism. The small positive change in the base, reflecting primarily the inflow of gold, was
more than offset by increases in reserve requirements. The net effect
was a decline in the monetary base plus the accumulated sum of liberated reserves. This decline occurred, however, after the middle of
the year and continued throughout 1949 when the Federal Reserve
described its policy as one of "ease."
Some of tlie confusion generated within the System during this
period is reflected in the remarks made in the annual reports for 1948
and 1949. Read in tandem these statements strongly suggest the
unwillingness of the Federal Reserve to seek foundation for its statements. The 1948 report notes: 7
The Federal Reserve System is also much better equipped
than ever before to meet the credit needs of the economy in
a period of downward readjustment. Through open market
operations, the System has virtually unlimited means of supplying the money market with additional reserves, if the situation should call for such action.
A f t e r the recession of 1949 had started, the annual report noted: *
Because the System had not been in a position to exert
greater monetary restraints, it had less scope for the reversal
of policy when the time came to relax credit restraints.
Notwithstanding these limitations, the System acted
promptly to adjust monetary and credit policy to the changed
conditions of early 1949.
I n fact the System did not pursue a countercyclical policy. The
annual percentage change in the money supply was negative throughout the year. The same conclusion holds even if time deposits are
included in the money supply. Moreover, had it not been for the
continued reduction in the demand to hold currency by the public,
the decline in the money supply would have been greater. The monetary base and the accumulated sum of liberated reserves declined
throughout the year, and on balance the System sold or retired $4.5
billon of Government securities.
A t the meeting of the FOMC on May 3, 1949, the recession was
described as desirable. Policies were to be directed not toward reversing the direction of the economy but at "keeping the movement
from going too far." This perhaps explains the reference in the
1948 report 9 to the need for offsetting a reduction in reserve requirements by sales of securities in the open market. # The FOMC policy of
supporting Government security rates was cited as the reason for
7• Annual Report of the Board of Governors for 1048, p.
Ibid., p. 7.
• Annnnl Report of the Board of Governors for 1949, p. 4.
»r»
• Op. c!t., p. 2. The 1948 annual report Included a discussion of the developments In
64
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
tightening the monetary system in the midst of recession. This policy
was continued until the announcement of June 28 recognized that "the
maintenance of a fixed pattern of rates has the undesirable effect of
absorbing reserves from the market at a time when the availability of
•credit should be increased." Despite this acknowledgment there is no
^perceptible change during the last half of 1949 in the rate of decrease
in the monetary base plus the accumulated sum of liberated reserves.
However, when we look at the volume of free reserves, we find a sudden
jump from the level $650 million to the level $850 million that was
maintained from July through October.
With the benefit of hindsight, the Board summed u p the postwar
period 1946-49 in the following terms: 1 0
I n the transition period from a war to a peacetime economy
the inflationary problem become more acute, notwithstanding
the termination of heavy Government deficits. The development of inflation was made possible by the large volume of
liquid assets built up during the period of war finance, * * *
but it was augmented by postwar expansion of credit to
private borrowers. Liquidation of Government securities
was an important source of funds f o r current spending and
for credit expansion, and the Federal Reserve found it necessary to purchase securities in order to maintain a stable and
orderly market for Government securities. These purchases
supplied additional bank reserves. Under the circumstances
action for counteracting inflationary developments had to be
limited to relatively moderate measures.
There is little evidence that recognition was given to the principal
features of their policy. Much more attention was spent throughout
the period on the need for more controls and more powers. In the
section that follows, we will discuss the use that was made of the additional powers to raise reserve requirements that the Congress granted
in 1948.
Early in 1950 the annual change in the extended monetary base and
the annual percentage change in the money supply became positive.
The rise in the base and the money supply accelerated through 1950
and continued into 1951. The annual report recognized the rising
money supply and discussed the factors contributing to the increase.
The Board noted that "Reserve positions of commercial banks were
under greater pressure in 1951 than in any other postwar vears/' la
fact, the monetary base increased by approximately $2 billion, the
largest annual increase in the postwar period between 1947 and 1963.
But the monthly average of free reserves declined in the course of the
year to the lowest levels that had been achieved in the postwar period
up to that time. I t seems likely that this reduction in free reserves
from $885 million in December 1950 to $169 million in D e c e m b e r 1951
encouraged the Board to believe that an anti-inflationary program was
m effect when in fact their policy was expansive.
^ I t is particularly important to note the very different m o v e m e n t s
m the monetary base and the level of free reserves. The base and the
money supply were rising, while the level of free reserves was rising
om°i? Report*ptVp'857*
Manngement of the
Public Debt» Joint Committee on t b e
65
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
and falling during the year but declining on balance. This is not the
only time that the level of free reserves and the base moved in opposite directions. But a critical judgment about the monetary system
appears to have been based on the movement of free reserves. The
Federal Reserve interpreted the data as showing that the banks were
under greater pressure. Had they looked at the monetary base as an
indicator of the position of the monetary system, it is unlikely that
they would have reached the conclusion that they did.
Prices soared during the year 1950 and remained high in 1951. The
BLS Wholesale Price Index (1947-49 base 100) rose from an average
103.1 in 1950 to 114.8, a rise that was surpassed in the postwar only in
1947. Most of the increase occurred during late 1950. Very similar
evidence is obtained if the Consumer Price Index is used as a guide.
If the pegging of bond prices was a serious handicap to an effective
anti-inflationary program, one would expect that the removal of the
peg would eliminate the root of the difficulty. In our interpretation
of the postwar events, the principal inflation that is attributed to Federal Reserve postwar policy occurred after the start of the Korean war.
True, prices rose in 1946 and 1947 following the removal of commodity
price controls. But it is likely that much of the rise during this period
was an adjustment of quoted prices to increases in the money supply
that occurred during the war. Federal Reserve policy during 1947-49
w-as not actively expansionist. Indeed, monetary policy exercised a
deflationary influence on the economy in late 1948 and throughout 1949
despite the presence of recession in tlie latter year. The Federal Reserve made no attempt to "roll back" prices in 1951 after the peg was
removed from bond prices. Instead they permitted a more rapid expansion of the money supply than they had permitted during the
previous 4 years. There is, therefore, little reason to conclude that
monetary policy would have been used to reverse the direction of price
changes in 1946 and 1947 had the support policy been repealed. 11
Federal Reserve policy in the postwar, preaccord years made substantial use of the power to vary member bank reserve requirements.
On 18 separate dates reserve requirements were altered for one or more
classes of banks or types of deposits. Reliance was placed on these
changes both as a device for controlling inflationary pressures in 1948
and early 1951 and as a means of easing the reserve position of the
banking system in 1949. Within the modified base conception, fiat
changes in reserve requirements operate principally by altering the
accumulated sum of liberated reserves. A multiple expansion or contraction of the money supply should follow the liberation or absorption
of reserves brought about by fiat changes in reserve requirements. But
we have already noted that the money supply moved in the opposite
direction to tlie one suggested by the changes in reserve requirements.
I n 1949, the money supply contracted, despite the liberation of reserves
through reductions in reserve requirements; the money supply expanded in 1951 despite the increases in reserve requirements during
January and early February. The following section discusses in more
detail the Federal Reserve policy of changing reserve requirements
and evaluates the use that was made of existing and augmented power
to control inflationary developments.
11
T h i s s h o u l d n o t be t a k e n a s a s u g g e s t i o n t h a t a " r o l l b a c k " of prices w o u l d h a v e been
a desirable poller.
66
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Changes in reserve requirements
The modified base doctrine indicates that changes in the sum of
liberated reserves and changes in the monetary base are of approximately equal significance as determinants of a change in the money
supply. If the Federal Reserve does not offset tlie change in reserve
requirements by moving the monetary base in the opposite direction,
such changes are capable of providing a multiple expansion or contraction in the supply of money. But Federal Reserve policy can offset
the changes in reserve requirements by altering the monetary base.
Before presenting the evidence on compensation of particular
changes in reserve requirements, it should be noted that tlie Federal
Reserve recognizes that compensation occurs. Generally, they have
regarded postwar changes in reserve requirements as events of major
significance in their effect on available bank reserves. To reduce the
impact of the change, open market operations in the opposite direction
have withdrawn or furnished some of the reserves released or absorbed
by the change in reserve requirements. But it has not been made clear
that, on many occasions, compensation completely offsets the effect on
reserve positions of the fiat change in required reserves. (See response
to question VI in the appendix.)
One measure of compensation is the monthly change in the reserves
of banks divided by the amount of reserves absorbed or liberated
through the fiat change in reserve requirements. In table VII-11
these changes have been expressed as percentages. The larger the
percentage, the greater the amount of compensation. ^ A change of
more than 100 percent indicates that the reduction or increase in reserve requirements was overcompensated, i.e., more than completely
offset.
TABLE
VII-11.—Compensation of changes in reserve requirementsf
compensation
percentage
by date and
1. Increase in reserve requirements, Feb. 27,1948:
March 1948
April
May
2. Increase in reserve requirements, June 11, 1048:
June 1948
August
3. Increase in reserve requirements, Sept. 1G-24,1948:
October 1948
November
December
_
_
January 1949
February
4. Decrease in reserve requirements, May 1-5, June 30^Ju1y"lTfoii)"'
May 1949
June
_
July
November
90.6
71.1
81.8
SI. $
II_IIIIIII~IIII~~IIIIIII__ 81.0
II_III~~IIII ~I_I
6. Increase in reserve requirements, Jan. 1-Febll7l9?)l:
January 1951
I_I
February
March
"II
April
__
May
I
June
II
100.6
107.4
115.2
11*. 2
94.1
6-". 4
IIIIIIIIIIIIIIIIIIIII
5. Decrease in reserve requirements, Aug. 1-Sept ~1 194lf*
September 1949
'
*
October
4
IIIIIIIIIIIIIIIIIIIIIIIIIIII-- 82.6
August
September
1*2
19-6
_
120.9
80.9
34,5
75.0
89.9
95.7
74.3
94.9
67
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Tho table makes clear that, vritli the exception of the first increase
in reserve requirements in 1048, all of the fiat changes is reserves during the preaccord years were compensated by a change in bank reserves in tlie opposite direction. In some cases, the change in reserves
overcompensated the effect of a change in reserve requirements. Moreover the policy of compensation did not stop at the time of the
accord. During 1951, 90 percent of the increase in required reserves
was offset within 2 months. Very similar findings pertain to changes
in reserve requirements in later years. The reduction in reserve requirements in July 1953 was more than 60 percent compensated by
mid-August. A further reduction in reserve requirements in midJune 1954 was 80 percent compensated within 1 month and greatly
overcompensated at the end of 2 months. Nevertheless, another reduction in reserve requirements followed at the end of July. In that
case, compensation never exceeded 72 percent so that a substantial
amount of additional reserves was liberated.
Compensation of the reserve requirement changes in 1948 may be
defended by reference to the pegging policy. But that defense cannot
explain the Federal Reserve's willingness to void more than 80 percent of each set of reductions in reserve requirements in 1949. Nor
does tho compensation in 1951 and 1954 follow from the inability of
the Federal Reserve to pursue a more "flexible" monetary policy.
One wonders at the rationale for a policy that introduces a change
of major importance to the monetary system and then almost completely reverses the policy within a few short weeks.
To pursue the matter one step further, a correlation was computed
between the annual change in the base and the annual change in the
accumulated sum of reserves liberated or impounded. The annual
changes wTere computed for corresponding months starting with each
change in reserve requirements and terminating 1 year after the
change. For example, to investigate the relation between the change
in the base and the change in the accumulated sum of liberated reserves associated with the increase in reserve requirements in January 1948, we computed the change in both policy variables starting
with the change from January 1947 to January 1948. Because of
tlie frequent changes in the requirement ratios during the preaccord
years, tlie annual changes between corresponding months terminate
with September 1950, 1 year after the reduction in reserve requirements in September 1949.'
Two measures are used to indicate the Federal Reserve's compensatory policy. Both are showm in table VII-12. Column 1 indicates
the average amount by which the monetary base changed per dollar of
reserves liberated or impounded by reserve requirement changes. A
negative number in this column indicates compensation of reserve
requirement changes. Values close to minus 1 reveal that compensation was almost perfect on the average—each dollar of reserves liberated or impounded was accompanied by a change in the base in the
opposite direction. A positive value in this column indicates p policy
of reinforcement. Each dollar of reserves liberated or impounded
is accompanied by a movement in the base that effects the money
supply in tlie same direction as the change in the requirement
ratios. The numbers in parentheses are measures of the variation
around the average, technically known as "t statistics." The larger
68
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
the values shown, the smaller the variability in the Federal Reserves
compensatory policy. Column 2 is the computed coefficient of determination that we have used several times as a measure of association.
association
accumulated sum of liberated
reserve
requirements
TABLE V I I - 1 2 . — T h e
between
reserves
changes in the base and changes in the
during periods surrounding changes in
Period
T h e a m o u n t of
compensation
per dollar of
reserves liberated or
Impounded
(1)
F e b r u a r y 1948 to September 1950
J a n u a r y 1951 to J a n u a r y 1952 plus J u l y 1953 to J u l y 1955 plus F e b r u a r y
1958 to April 1959
September 1900 to December 1961 plus October 1902 to Octol>cr 1963
All above periods combined
-
- 0 . 9 6 (32.55)
— 1. 26 (28.63)
2.22 (6. 76"
-.992(16.41!
Coefficient of
determination
(2)
0.97
.941
.629
.706
The first period in the table is the preaccord period. Changes in
reserve requirements were almost completely compensated by movements of the monetary base in the opposite direction. On the average,
the changes in the base offset the changes in reserve requirements so
that the latter had little effect on the money supply. The Federal
Reserve's open market policy operated against the policy of raising
reserve requirements in 19-48 and against lowering them m 1949. As
we noted previously, the compensation in 1949 was not forced on the
Federal Reserve by the "pegging" operation then in effect
The data for the second period reveal that the compensatory policy
continued throughout the fifties, long after the accord. Reserve requirement changes in these years were more than offset by open market
operations, on the average. Tlie evidence from the fifties strongly
reinforces our interpretation of Federal Reserve policy. Compensation appears to be a relatively persistent feature when reserve requirements are raised or lowered.
Between 1958 and 1960, the policy appears to have changed. The
reductions in reserve requirements in I960 and 1962 were not offset by
reductions in the base. Judged by the coefficient of determination, the
changes in the base were less closely associated with the change in
liberated reserves. Nevertheless, open market policy must be interpreted as reinforcing rather than compensating reserve r e q u i r e m e n t
policy, on the average, during the early sixties.
In the postwar years, there have been increases and decreases in
reserve requirements. On balance, however, the requirements have
been reduced. This is particularly true for the period since the accord.
All changes after 1951 have been reductions in the requirement ratios.
In addition, the admission of vault cash as a part, of the m e a s u r e d
reserves of the banking system was the equivalent of a further liberation of reserves.
There is no important difference between a reduction in reserve
requirements compensated by a sale of s e c u r i t i e s to the banking svstem
and a subsidy to the earnings of the commercial banks. The earning
assets of the banking system are increased relative t o the cash assets
when reserves are first liberated by changes i n reserve r e q u i r e m e n t s
69
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
and then absorbed by open market operations. Tlie expansion of
banks' earning assets is 110 greater than would be permitted by an appropriate increase in the monetary base with no change in the accumulated sum of liberated reserves.
Eos the Fedcral Reserve had a countercyclical policy in recessions?
On numerous occasions, Federal Reserve spokesmen have described
their policy during recessions as one designed to stimulate the economy.
Our discussion in earlier chapters has shown that they quite generally
use the level of free reserves as the principal indicator of their policies.
But we have found almost 110 relation between the level of free reserves and changes in money and credit. Moreover, our analysis clearly
reveals that the stock of money grew more slowly during postwar
months of contraction than during months of expansion in the economy. I t is difficult, therefore, to reconcile their statements 12 with
the available evidence.
Wo contend that Federal Reserve spokesmen and policymakers
seriously misinterpret their policy actions. There is little evidence
that they pursued a stimulative policy in periods of recession. In
this section we present a brief discussion of Federal Reserve policy
during recessions in terms of the movements of the extended monetary
base. This magnitude is the single most important determinant of
the money supply and is controlled to a close approximation by Federal Reserve policy. The extended base is, therefore, an excellent indicator of the actual policy pursued by the Federal Reserve.
A chart in the appendix shows the change in the extended base
between corresponding months of successive years from June 1945
through December 1962. Our discussion of the four postwar recessions and the upswing that started in February 1961 will refer to
the chart. To date the cyclical peaks and troughs, we will use the
dates provided by the National Bureau of Economic Research. We
are aware of the uncertainty that surrounds the dating of turning
points. However, the National Bureau dates are quite likely to be
within a few months of the true peaks and troughs and are sufficiently
accurate for the discussion. A change of a few months in the dating
of the peaks and troughs would not affect the main conclusions.
(1) The recession from November 1948 to October 1949
The annual growth rate of the extended base was about $800 million
at the start of 1948. I t rapidly fell, became negative in July, and
reached a postwar low of —$800 million in January 1949. Thereafter, the extended base rose slightly, but it remained below -$500
mill on throughout the remainder of 1949. The growth rate of the
base d ; d not"become positive until March 1950. Throughout the
recession, tho extended base was declining and exerting a contractive
effect on the money supply. In fact, the growth rate during the first
postwar recession was smaller than in any period since the recession of
1936-37. Thus there is no indication that Federal Reserve policy
became more expansive relative to the policies pursued before the
recession. On the contrary, the behavior of the extended base reveals
that policy became more deflationary than it had been in the downswing.
^ 13 F o r a n e x a m p l e , see p t . 2 of t h e a n s w e r t o question V In t h e a p p e n d i x w h e r e t h e l 2
P r e s i d e n t s i n d i c a t e t h a t d u r i n g recession ' S y s t e m policy is m o r e likely t o be pos tlvely
stimulative r a t h e r t h a n to remain the s a m e " and t h a t " a t such times the System s h i f t s to
a n a c t i v e a n t i r e c e s s i o n a r y poUcy."
34—474—64
6
70
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
(2) The recession from July 1953 to August 195k.
During the year 1952, the growth rate of tlie extended base reached
levels not achieved since 1946. I n the second half of 1952, the annual
growth rate never dropped below $2 million. From January 1953
to the peak of economic activity in July, the base decelerated and the
annual growth rate fell to $1.2 billion. A f t e r a brief pause, the deceleration resumed until the low point of $119 million was reached
in September 1954, after the trough of the recession. The chart
clearly indicates that the growth rate in 1953 was below the growth
rate in 1952 and above the growth rate in 1954. The recession was
thus accompanied by a declining growth rate in the e x t e n d e d base—a
growth rate that was substantially smaller than the rate maintained
before the recession. Once more we conclude that there is no evidence
of a switch in policy in a more stimulative direction. Policy shifted
in a less stimulative direction.
(3) The recession from July 1957 to April 1958
With the exception of March 1957, all months between January 1956
and the peak in July 1957 showed annual growth rates between $400
and $800 million. The growth rate declined rapidly after July and
became almost zero at the end of the year. In February 1958 the
extended base accelerated with impressive rapidity. An inspection
of the chart clearly demonstrates that the growth rate of the e x t e n d e d
base f e l l during {lie recession until February, i.e., 2 months before
the trough of economic activity was reached. During the period from
August 1957 to February-March 1958 the growth rate not only declined
but was lower in every month than it had been since the recession
period of 1954 (with one exception). Again, the pattern o b s e r v e d indicates no shift in Federal Reserve policy toward expansion until 2
months before the trough was reached. Monetary policy was comparatively deflationary even after the System recognized the onset of
the downswing. The" System effectively reversed its comparatively
deflationary policy in February 1958, and its decisive reversal very
likely contributed to the rapid termination of the downswing.
(4) The recession from May I960 to February 1961
The same patterns occurred in the most recent recession. For most
of the downswing the growth rate of the extended base was lower than
in the previous 18 months. Relative to the policy pursued in 1959
most of the downswing exhibited a lower growth rate of the extended
base. The Federal Reserve thus followed a comparatively more deflationary policy during the recession. However, it should be acknowledged that the System reversed the direction of policy in the fall of
1960. This shift in direction, mirrored by the increase in the growth
rate charted in the appendix, occurred much earlier, relative to the
previous peak, than in any other postwar recession. The extended
base, including the released vault cash, had an annual growth rate of
—$125 million from April through August 1960. As a result of the
increasing importance of the released vault cash, the growth rate of
the extended base began to rise in September. By the trough of the
recession, the annual growth rate had reached $800 million.
(5) The Federal Reserve's evaluation of calendar year 1963
While our analysis in this section is concerned with policy action
during recessions, a similar comparison of Federal Reserve s t a t e m e n t s
71
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
and events during periods of expansion would reveal a divergence between the facts expressed by the extended base and the statements of
the policymakers. Some of ihcse periods have been discussed elsewhere
in this study. However, as our report was about to go to press, the
Federal Reserve released its annual report for 1003. We comment
briefly on some statements made in that report to indicate that the
problem to which we have repeatedly referred, remains. The absence
of a validated conception as a basis for policy actions or statements is
readily apparent in the most recent annual report.
The report notes:
At its last meeting in 1002 the Federal Open Market Committee had concluded that it was appropriate to reduce a little
the degree of ease existing at that time. Accordingly, it redirected its actions toward accommodating moderate further
increases in bank credit and the money supply. * * * The Committee made no further change in policy until mid-May, when
it moved to reduce reserve availability slightly further. 13
Elsewhere the report notes that in the second half of 1903 reserves
became "less readily available" and comments on three distinct shifts
toward "less ease." 14
The growth rate of the extended base yields no support for the
Systenrs appraisal of its policy. The growth rate expanded markedly
throughout the year and reached a level in the second half of 1963 that
had not been observed since 1952. We find no evidence of the shift
toward less ease in the extended base. On the contrary, the Federal
Reserve pursued a substantially more expansionary policy in 1963 than
in 1962.
Once again, the extended base clearly indicated the expansionist
monetarv policy. The level of free reserves gave the wrong indication. In chapter I V of this study, written before the appearance of
the annual report, we noted the decline in the level of free reserves in
February and May and predicted that the annual report would record
a movement toward less ease in these months. The confirmation of
these predictions (see footnote 14 of this chapter) reveals quite clearly
that free reserves remain the principal indicator of Federal Reserve
policy action.
The systematic difference between the Federal Reserve's description
of its policies and the actual policies requires an explanation. The
basic ingredients for the explanation have been discussed elsewhere
this study. Their misconceptions result from the continued
reliance on free reserves as an indicator of policy. Our analysis
of this doctrine indicates that even if one accepts the central
role assigned to free reserves in the causal process, it does not follow
that free reserves are a useful indicator of policy or prevailing monetary situations. 15 As a result of the indicator function mistakenly
„ " 0 ™ t T l * ^ S o e ^ i V t h ^ ' R o v i e w of Open M a r k e t O p e r a t ^
On p. 120. J a n u a r y and F e b r u a r y a r e listed a s showing a slight s h i f t tow-ard leibs ease.
M a y t h r o u g h J u l y a r e a g a i n listed a s showing a s h i f t toward less ease. The same s h i f t is
indicated again f o r A u g u s t a n d September.
ISWIAI...I
_
T h i s a n a l y s i s h i s been developed In our underlying pnper ' ' f
Conceptions of the Money Supply t r o c o s s . " that will be published
t h a t a completely specified f r e e reserves doctrine, acknowledging .the f u l l
Jj
t h e behavior relntlons c o n s t i t u t i n g the m o n e t a r y p r o c e s s , agnln yields ^ J ^ " / ^
a s t h e m a j o r d e t e r m i n a n t of t h e money supply. When a proper analysis I«carrie(Iotit, the
b a s e — n o t t h e f r e e reserves—emerges a s t h e policy Indicator even if f r e e reserves are
given the c e n t r a l position in t h e causal r e l a t i o n .
72
ALTERNATIVE APPROACH TO T H E MONETARY MECHANISM
assigned to free reserves, increasing free reserves are interpreted as
a reflection of a policy of more "ease" or less "restraint."
PRELIMINARY CONCLUSION
We do not regard the above analysis of the monetary mechanism
as complete. The results reported are but the bare beginnings of an
adequate appraisal of the role of particular factors and their influence on the money supply. Yet these results go f a r beyond those
that have been obtained for the modified free reserves doctrine in
providing an understanding of the monetary process. And they
strongly suggest that control of the change in the money supply rests
largely in the hands of the Federal Reserve if they pursue appropriate policies and attempt to understand the mechanism connecting
their policy operations with changes in the supply of money.
Our study indicates that the most important single factor operating to change the money supply is the change in the monetary base..
Through open market operations the change m the monetary base can
be brought under the control of the Federal Reserve authoritiesSuch control is a precondition for appropriate policy actions to offset
changes in the public's demand for currency and time deposits orchanges in the banks' demand f o r cash assets. But control cannot
be achieved as long as the Federal Reserve allows day-to-day money
market events to obscure its view of the money supply mechanism.
This study began by pointing out some of the errors that prevent a
useful approach to the understanding of the monetary mechanism by
the Federal Reserve. We have tried to show that many of these
errors are embodied in the concept of free reserves and the modified
free reserves doctrine. Most of all this study has been directed toward
presenting evidence to support the statement that in 50 vears the
Federal Reserve has not obtained an understanding of the basic factors that must be incorporated in an analysis of the money supply
process.
Before turning to some suggestions designed to improve Federal
Reserve monetary operations, it is useful to consider one last set of
questions. An effective monetary policy must do more than control
the stock of money. I t must also operate on the level of income with
sufficient force to change the direction of income or at least mitigate
recession and inhibit inflation. I t is appropriate therefore to c o n c l u d e
this chapter by considering the question: Has the stock of money exercised an important influence on the postaccord economy?
T H E STOCK OF MONEY, VELOCITY, AND NATIONAL INCOME
The discussion in this section shifts attention from the supply of
money to the demand for money by the public. I n several recent
studies we have attempted to formulate and test a relation between
the demand for money and its determinants. Closely related to the
concept of a demand for money is the measure known as the income
velocity of money. Once we know the factors affecting the demand
for money, we know the determinants of income velocity* Knowledge
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
73
of either one of these is tantamount to knowledge of the other. If
we can forecast the velocity of money and control the supply of money,
we can have substantial impact on the level of national income through
monetary policy.
To obtain the predicted level of income from the prediction of velocity, all that is required is that the predicted velocity be multiplied by
the money supply made available through Federal Reserve operations.
W e hasten to add that controlling income is not a simple, mechanical
operation. Nor is monetary policy a panacea that guarantees that we
can make the level of national income whatever amount we desire.
First, the analysis does not separate changes in real income from
changes in prices. Second, increases or decreases in the quantity of
money alter interest rates, one of the major factors determining the
movement of income velocity.
Nevertheless, the analysis and the forecast are useful. Evidence that
we have gathered clearly indicates that prediction of virtually all of
the major turning points in the level of national income since 1910
can be made by this means.16 In recent work, we have begun to explore the possibility that forecasts of national income made by this
procedure would be an aid to policymakers and would add to our
understanding of the economy. We have, therefore, included here
some very preliminary results based on our attempt to forecast the
quarterly value of national income during part of the postaccord
period.
The results, though preliminary, are encouraging. They suggest
that the average error in forecasting quarterly velocity is approximately 1 y2 percent. These forecasts of velocity have been converted
into forecasts of national income by multiplying the forecast by the
actual supply of money. I n this way, the forecast of national income
is obtained that has the same error of forecast, 1% percent, as the
velocity forecast. Ultimately, we will wish to modify this procedure
"by using a forecast of the quantity of money in place of the actual
money supply. Our predictions of the money supply earlier in this
chapter clearly indicates that the error will not be enlarged substantially. But, the use of a forecast of the money supply is of little
relevance until the Federal Reserve concentrates attention on control
of the quantity of money and its rate of change rather than on the
day-to-day details of the money market or the stock of bank credit.
The quarterly forecasts of national income at annual rates and the
actual level of national income are shown in chart VI-1. We observe
that in addition to the reasonably close agreement between the forecast values and the actual values, there is reasonable agreement in the
two series at turning points in economic activity. I t should be noted
that the forecasts of velocity were made using only information that
would have been available at the time that the^forecasts were made.
I n the case at hand, each forecast used information for a date no later
than the quarter preceding the forecast value. These values were
then converted into forecasts of income by using the actual values of
money supply, as noted above.
"See "Predicting Velocity: Implications for Theory and Policy," Journal of Finance,
3Iay 1963.
CHABT V I - 1
Billions of
Dollars
67-A
450
440
430
national Income and Predicted Incocw
Quarterly 1954-IV to 1959-IV
420
410
400
390
national
's
Prodietad National
Incooa
380
370
360
350
340
330>
/
320
4-54 1-55 2-55 3-55 4-55 1-5* a-56 3-56 4-56 1-57 3-57 3-57 4-57 1-5« 2-5« 3-58 4-5« 1-59
4-5?
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
75
The dominant factor is the quantity of money and not velocity.
That is, an increase in the quantity of mone}^ adds to the level of national income; it is not completely offset or reversed by changes in
velocity. These findings strongly suggest that if monetary policy
can control the stock of money more closely than it has, the degree of
control over cyclical variations in the pace of economic activity can
be improved greatly. But all changes have not been forecast completely; control of the stock of money is an aid in promoting a full
employment economy with stable prices. I t is not a guarantee of
perpetual "full employment."
One major lacuna in the forecast is the omission of a mechanism
capable of predicting separately the change in prices and the change
in real income. Our confidence in the forecasts of economic activity
would be much greater if these two components of money income had
been separated. F o r it is clearly important in policy formation to
predict the extent to which an attempt to increase the pace of economic activity will produce inflation. The studies that we have
underway have not reached the point at which this can be done. But
we can survey a part of the postaccord record to see the way in which
changes in national income were divided between changes in the price
level and changes in real income during the period for which our forecasts have been made. These data are shown in chart VI-2.
CHART
VI-2
Deflated Nat National Product
and Price Deflator
1954-17 to 1959-IV
BUlica* of
Dollars
O
H
H
50
s
>w
^
w
o
>
H
O
aw
o
H
t>
W
M
oR
a3
£/
K
ir-54 V-55 2-55 3-55 ^-55 1-56 2-56 3-56 fr-56 1-57 2-57 3-57 4-57 1-53 2-58 3-58 4^38 V59 2-59 3-39 y-«
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
77
CONCLUSION
This chapter lias presented some of the accumulated evidence 011
the ability of the Federal Reserve to ell'ect desired changes in the
stock of money and in the economy through monetary policy operations. We have found that the degree of control over the money
supply can be greatly improved if analysis is focused on what we have
called the modified base doctrine. And we have seen that the preliminary evidence seems to indicate that changes in the stock of money
affect the level of income in a predictable way. Thus it would appear
that additional analysis of monetary factors, and a reorientation of
Federal Reserve thinking, would have desirable consequences for economic stability.
Moreover, we have found that many of the implications of the modified base doctrine have been confirmed by our tests. These findings
suggest that to obtain more useful knowledge about the monetary process, analysis should focus on the monetary base, the demand for currency and time deposits by the public, the demand for reserves by
banks and other components discussed in the text. Our work is a
small step in the direction of improving understanding of the process.
Much remains to be done. But the results are sufficiently encouraging
to suggest that further work along these lines would be useful.
The central concept presented here, the monetary base, is obtained
from a regularly published table in the Federal Reserve Bulletin,
Most of the other data used in the analysis are regularly provided inother published tables. Yet there is little or no indication that the
appropriate steps have ever been taken within the Federal Reserve to
link the base to the money supply. I n an earlier chapter, we commented on some of the reasons for this failure. Organizational
factors, particularly Greshman's law of planning and the banker's
orientation, seem to be responsible for the attachment of the Federal
Reserve to an inadequate conception of the monetary process. In addition, what is sometimes referred to as "selective perception" seems
to be an important element.
Observations that confirm the prevailing view are accepted as evidence that the prevailing view is correct, while observations that contradict the prevailing view are neglected or made to fit into some
specially contrived framework. These special categories are never integrated with one another to obtain a unified view of the monetary
mechanism. Thus erroneous views are perpetuated.
Analysis and theory are required to fit the parts into a coherent
framework capable of explaining the details of monetary life. Tests
of the conception are required to assure that the framework is adequate and that the specific factors fit together in the manner envisaged
by the theory. This is a task for a research staff, not for policymakers. But the judgment of policymakers who are not gifted with
clairvoyance cannot be much better than the analysis on which their
conclusions rest. I t is therefore of the utmost importance to make
the analysis explicit, to fit the pieces together, to test the framework
against the facts.
, ,
,
The major failure of the Federal Reserve System has been the unwillingness to take these steps. I f our analysis leads to any conclusion, it is the conclusion that changes in the method of making monetary policy are required if Federal Reserve monetaiy operations are
to achieve their potential usefulness in the economy.
SECTION
III—SOME
SUGGESTED CHANGES I N POLICTOXAEJNG
PROCEDURES
What is the primary purpose of Federal Reserve policy? Is it
designed to smooth the adjustment of banks to the inevitable random
changes in the money market? Or is it primarily a means of controlling, as best we can, the movements of the stock of money? The
operations designed to smooth the reserve adjustments for bankers,
the day-to-day operations that often dominate System policy, introauce a large amount of variation in the monetary base. As a result,
such operations add to tlie variation of the money stock, weaken or
reduce the "degree of control" over the money supply, and introduce
substantial changes in the monthly rate of change of the money stock.1
-kven if the modified free reserve doctrine is replaced by the more
appropriate base theory of the determination of the supply of money,
aaiJy and weekly changes in the base for "defensive" reasons would
continue to hamper the effective transmission of Federal Reserve polt^'-f ™tever the lags in the relation of money to income, whatever
the factors affecting the speed of transmission, continuous defensive
operations cause the signals emitted by Federal Reserve policy to be
extremely variable. The variability of the monetary signal will regain as long as defensive operations and random changes create the
esteilc e
VIII™
* d monetary base that is observed in chart
If the analysis presented in the two preceding chapters has validity,
as the evidence presented there and elsewhere suggests, it follows that
federal Reserve policy is summarized by the movements of the extended base shown in the chart. It is extremely difficult to rationalize
the observed movements of the base as part of a coherently formulated,
systematic policy for controlling the supply of money. Even if Federal Reserve operations affected the money supply instantaneously,
it would be difficult to defend the repeated changes in the direction of
policy operations. One would still wish to know about the effect on
economic activity of expectations engendered by the variations in the
money supply.
Of course, it may be suggested that some of the movement in the
extended base is introduced as the result of a conscious policy of removing seasonal changes in interest rates. The Federal Reserve is
charged with the problem of removing interest changes of a seasonal
nature to avoid the effects of such changes on the allocation of resources. We noted earlier that, consistent with their recognition of
this responsibility, the "a" section of the directive has been interpreted
as an instruction to the Manager to remove seasonal disturbances
affecting interest rates. But, before we accept this explanation, two
* W e h a r e p r e v i o u s l y noted t h a t t h e c o r r e l a t i o n between changes i n t h e money snppl.^
•currency plug d e m a n d deposits, in a d j a c e n t m o n t h s is negative. ^ rise in t^he money
?»pply In a p a r t i c u l a r m o n t h i s n o i n d i c a t i o n that i t will be followed by a f u r t h e r rise i n
t h e succeeding m o n t h .
79
80
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
points must be noted: (1) The monthly variability of the extended
base is the result of more than seasonal operations. Removing seasonal variations in interest rates would introduce such changes into
the extended base. No consistent seasonal pattern remains in the
changes of the extended base plotted in chart V I I I - 1 . This suggests
t h a t other short-term operations explain the exhibited monthly variability. (2) Attenuation of seasonal variations is not a purely mechanical affair. As we noted previously, a validated conception is
required. The amount by which the base is changed to offset a given
seasonal change depends upon the theory that we use and the goal
t h a t we pursue.
This becomes clearer if we suppose that we desire to offset the seasonal increase in the demand for currency. We choose as a goal the
elimination of seasonal influence on Treasury bill rates. Our t h e o i j
must tell us the amount of reserves to supply to accomplish that end.
A theory based on free reserves will not yield the same answer as a
theory based on total reserves. Nor can we solve the problem by deciding to keep the interest rate unchanged for the particular season,
f o r in doing so we remove more or less than the seasonal influence.
Clearly, a theory is required and all theories do not provide equivalent
answers. Hence it becomes important to choose among competing
theories or conceptions before we can decide on the meaning of "appropriate" in the expression "appropriate policy action" to remove
seasonal influences.
A very similar issue has been raised quite recently by two members
of the banking community. 2 They argue that, contrary to the Federal
Reserve view that "defensive" open market operations are a major
force stabilizing the economy or the money market, such operations
interfere "unnecessarily with private security markets." Moreover,
they suggest that prevailing methods of settling reserve balances "magn i f y the impact of random deposit fluctuations" and, in periods of
monetary ease, "destabilize trading in Federal funds particularly to<
the disadvantage of country banks. 3
A n unofficial reply by a staff member of the Federal Reserve Bank
of New York makes light of these objections. 4 This reply, though unofficial, is of interest since it reveals some basic features of the Federal
Reserve conception to which we have referred at several places in
this report. For example, Sternlight argues that if banks were allowed
longer settlement periods, some banks might "expand credit more
rapidly then was justified by the degree of reserve availability s o u g h t
by the authorities, * * * transferring the burden of adjustment to
others." This would "significantly delay and, in effect, blunt the
impact of Federal Reserve influence on bank credit." 5 Moreover, the
prospect that the Federal Reserve would no longer have weekly averages of reserves for Reserve city banks is viewed with alarm. The*
"System would not have nearly as good an idea as it does now of
where the banking system stood at any particular time with r e s p e c t
to reserve availability in relation to requirements." 8
» Albert H. Cox Jr., and Ralph F. Leach. "Defensive Open Market Operations and the
a
Settlement Periods of Member Banks," Journal of finance, March 1964.
* Ibid., p. 93.
r t> teTnJlffht
«?bW
* M R e s e r ™ Settlement Periods of Member Banks: Comment* IbM0|
«Ibid." p.* 96.*
CHART V I I I - 1
i
5" I
i m
7
8 9
m i
81
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
J % f i r S t ^ u ° t a t i o n f r o m Sternlight once again suggests the single
Dank trame of reference that dominates Federal Reserve thinking:
the second amply demonstrates that extremely short-term daily Sr
weekly events, are the events that the Federal Reserve watclies and
to winch they respond. Let us consider these statements in a bit
more detail.
Sternlight's argument that longer settlement periods would "blunt
me impact of Federal Reserve influence" is not supported by any
analysis or evidence. I t is a pure assertion that shows little understanding of the operation of the monetary mechanism. If the Federal
Reserve controls the size of the monetary base, why should it matter
mat a particular bank might be able to borrow Federal funds to meet
xne reserve settlement and thereby transfer the pressure to another
d o e s tIlis differ f r o m t h e
vif1' u •
P r e v a i l i n g arrangements under
winch individual banks buy and sell reserves in the Federal funds
market ?
The answer is that longer settlement periods would make for a
more efficient use of existing reserves. When the Federal funds rate
is comparatively high, a bank might be more inclined to sell Federal
lunds (reserves) with the expectation of reacquiring the reserves
later m the month at a lower rate. The longer settlement period
would encourage such operations. I n this way, the volume of reserves would be able to support a larger volume of aggregate deposits,
-to short, the costs to the banking system of reaching a settlement
would be reduced because a given volume of total reserves would sup^ t V 1 * a r £ e r v o ^ u m e of deposits for the banking system as a whole.
I he Federal Reserve can control the supply of reserves. The increased efficiency in the banking system's use of the supply of reserves
can be easily offset by a reduction in the volume of reserves. While
the degree of ease or restraint" associated with a particular volume
of reserves would change, the change need not produce one additional
dollar of monetary expansion if the Federal Reserve reduces the supPly of reserves by an appropriate amount. In principle, the proposal f o r monthly reserve settlement periods is akin to the revival
of the Federal funds market in the past decade. The introduction of
the Federal funds market reduced the cost of bank operations by permitting an individual banker (1) to reduce the amount of reserves
held as a contingency for sudden deposit withdrawals and (2) to earn
a
return on surplus reserves that are expected to remain available for
°nly a short period.
Contrary to the view taken by Sternlight, the existing Federal
funds market already performs the function of spreading the impact
of Federal Reserve policy from a single bank with a reserve deficiency
to other banks. Federal Reserve policy is not, or should not be, aimed
at maintaining the pressure on a single bank or group of banks. As
long as the Federal Reserve has control of the supply of reserves, the
desired "degree of restraint" can be maintained. The spread ot pressure to other banks is the first step in the mechanism transmitting
monetary policy from a localized influence on banks in money market
centers to banks in other locations*
,
^ The principal problem that arises in the o p e r a t i o n of the leaeraJ
funds market as a part of the reserve adjustment mechanism. is quite
different f r o m the one suggested by Sternlight. Because bankers will
82
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
not pay more than the rediscount rate to buy Federal funds, the Federal funds rate does not rise above the discount rate. Federal Reserve
discount rate policy thus imposes a restriction on the price that prevails in the Federal funds market. This restriction prevents the market from allocating reserves efficiently in some of the periods when
the Federal funds rate and the discount rate are equal. This problem
is discussed in more detail below.
Here, it should be noted that Sternlight uses a single-bank approach to policy operations. What we would regard as the mechanism transmitting monetary policy through the banking system, he
refers to as a means of blunting the impact of policy. His argument
may be true—for the single bank. But if the Federal Reserve attempts to control the rate of monetary expansion by operating on the
monetary base, it will not be true for the banking system.
The second quotation from Sternlight provides additional evidence
for our view that the Federal Reserve is concerned with daily or weekly
occurrences at the expense of longer term influences. Commenting on
the present system, he notes that "there is often some uncertainty as
to how the country banks stand in the first week of their 2-week reserve periods, and as to how this position might affect their willingness and ability to supply reserves to the central money market in the
latter part of the 2-week period" (i.e., in the following week).7 We
readily acknowledge that such weekly movements may have a decisive importance for the Federal Reserve in carrying out the m a n d a t e
given by the Congress. But we insist that neither analysis nor evidence has been presented to show that this is the case. Our own
attempts to find supporting evidence for this conception have produced none. Instead, the evidence seems to support the view that
concern with transitory intramonth variations in reserve positions produce the pattern of monthly changes in the monetary base that is
exhibited in the chart VIII—1, Such changes are partly reflected in
the money supply and produce a pattern that increases uncertainty.8
Again, the important point to be emphasized is the need for a validated theory or conception that serves as a basis for Federal Reserve
action. Continuous repetition of existing dogma, unsupported by
analysis or evidence, is not a substitute. I t may turn out that there
is an underlying rationale and substantial evidence for many of the
positions taken by the Federal Reserve spokesmen and officials. But
this can only be ascertained by analysis and evidence that has, as yet,
not been provided.
This study is concerned with methods for improving policy operations and arriving at "appropriate policy actions." t o that'end, we
suggest a number of general and specific recommendations. The list
presented is by no means exhausti ve, and no attempt is made to suggest
an ideal set of administrative arrangements. Rather our interest is
concentrated on some general features of an improved set of policy
arrangements that would be consistent with any one of a number of
alternative administrative structures. A few specific changes in policy
procedures are then discussed briefly. Xo attempt is made to present a
detailed analysis supporting each proposal.
* Ibid., p . OGw
• W e e m p h a s i z e t h e q u a l i f y i n g w o r d " p a r t l y " i n t h e s e n t e n c e i n t h e t e x t . As we noted
e a r l i e r In t h e r e p o r t , t h e c h a n g e s in r e s e r v e s m a y b e i n t e r p r e t e d a s s y s t e m a t i c o r random
( t r a n s i t o r y ) by b a n k e r s . T h e i r r e s p o n s e t o e a c h w o u l d b e d i f f e r e n t .
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
83
SOME GENERAL SUGGESTIONS FOR POLICYMAKING
Our principal recommendation follows directly from our analysis:
I. The Federal Reserve should develop and test a theory incorporating the essential elements of the money supply process—policy operations, currency, and time deposit behavior of the public and the demand
for reserves by banks. While we regard our own efforts in this direction as a useful beginning, we do not claim to have developed a definitive theory of the supply of money. But we insist that the evidence
strongly supports our contention that the theory that we have presented performs substantially better than the alternative conception
that is the dominant Federal Reserve view.
I t must be emphasized that until the Federal Reserve improves its
understanding of the money supply mechanism, policy operations are
more akin to an attempt to "steer a ship without a rudder" than the
'rudder that causes the monetary ship to lean against the wind."
Without a validated conception of the monetary process connecting
policy operations with the money supply, there is no basis for the belief
that the Federal Reserve operates to carry out the congressional mandate.
I I . The immediate aim of monetary policy should be control of the
stock of money.—Our analysis and the evidence in chapters I I and V
strongly support our contention that money and credit are not "two
sides of the same coin," as the Federal Reserve has maintained. In
the past, policies focused on bank credit have produced larger increases
jn the money supply during months of economic expansion than during recession periods. The analysis and the evidence that we have
resented supports the view that there is a close and predictable link
etween Federal Reserve policy and the stock of money in terms of
the conception centered on the monetary base; evidence from validated
economic theory supports the view that there is a relation between the
stock of money and the pace of economic activity measured by national income. To our knowledge, there is no substantial body of evidence supporting the contention that total "bank credit" is closely
related to the pace of economic activity. However appealing the latter
association may be at the level of intuition, neither analysis nor evidence has been presented to support this association.
I t must be noted, therefore, that the argument that our approach is
based on money, while the Federal Reserve prefers to view the process
in terms of "bank credit" is irrelevant. In the present state of knowledge, such statements indicate only that the speaker prefers to disregard the accumulated evidence and substitute his own unsupported
views.
I I I . The free reserves or "modified free reserves" doctrine should be
abandoned,—We find little evidential support for the asserted relation between the conception centered on free reserves and changes m
the stock of monev. We find even less evidence for the view that tree
reserves are related to changes in bank credit as specified in the modified free reserves conception. Indeed, the free reserves conception
does little better in explaining the behavior of the stock of money than
the simple statement: If the money supply fell last month, it will rise
next month and fall in the following month.
IV. The free reserves conception should he replaced by the conception centered <m, the monetary base.—Again, we emphasize that this
t
84
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
conclusion is based on a comparative appraisal of existing theories of
the money supply. "We do not wish the base theory, as we have presented it here and elsewhere, to become a rigid dogma. Continuous
appraisal, reappraisal, and comparisons with alternative conceptions
should be a part of the work of the Research Division at the Board
and at the Reserve banks. These efforts will lead to improved understanding of the money supply process that the Board once acknowledged as its primary concern. However, in the present state of knowledge, the conception centered on the monetary base, while far from
perfect, is the most highly validated conception of the money supply
process. Moreover, this conception incorporates the essential elements of the process—Federal Reserve policy actions, the behavior of
the banks and the public. Until modifications or tests against alternative theories show that another conception performs better, the
theory centered on the monetary base should serve (1) as a guide for
policy action and (2) as the conception used by the Federal Reserve to
analyze the money supply process.
V. The Federal Reserve should be required to report periodically
on the progress that it is making toward the development of improved
understanding of the monetary process.—The reader of this report is
well aware that after careful reading of published statements and
many pages of discussion, we cannot be certain about the Federal Reserve's views on many fundamental issues. The reason is that their
views are rarely articulated clearly or fully developed. They are more
in the nature of impressions obtained bv selective perception of events.
The requirement to report periodically on the work that is being done
to develop a validated conception of the monetary process will lead to
a better utilization of the Research Division. More important, it will
lead to a better understanding of the monetary process, and the relation of policy actions to the money supply and the relation of the
money supply to the pace of economic activity. Such knowledge is
the basic foundation for effective policy action and for the a v o i d a n c e
of errors and the inappropriate actions that have so frequently occurred in the past.
The report that is envisaged would be one that meets high professional standards. The conceptions should be clearly stated and evidential support, including tests against alternative conceptions, should
be provided. Such reports would provide a greatly improved understanding and increased discussion of the monetary process by the
staffs of the executive departments, including the President's Council
of Economic Advisers, the Congress, the interested members of the
academic community, and the public. Discussion and critical examination by outsiders would be of substantial assistance in suggesting
modifications and additional tests.
Moreover, the requirement to provide such detailed reports will
focus attention on the development and testing of alternative conceptions within the Federal Reserve. To those who would question the
need to force attention to this issue with the Federal Reserve, we ask
only: Why has the Federal Reserve failed to develop a validated conception after 50 years? Why has it retreated from the useful beginning made by Rietler more than 30 years ago ?
VI. The desired growth rate of the money supply should be explicitly
chosen for a 6-month or longer period and policy operations should be
85
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
directed toward achieving that growth rate by explicit choice of a
growth rate for the monetary} base.—We have found that the Federal
Reserve has demonstrated in the postwar years that it is capable of
judging turning points in economic activity quite accurately. The
major problem in policy operations has been the failure to take appropriate action. This failure is closely related to the invalid conception of the monetary process and the concern with extremely shortrun
operations. There is no clear reason why shortrun technical problems
of the money market should dominate Federal Reserve policy discussions. The proposal made here would require discussion of the
appropriate growth rate of the monetary base and the money supply
over a longer period of time. This would focus the attention of policymaking officials on the problems that have been entrusted to their control rather than on the technical position of the financial markets.
We do not suggest that the decision taken should be made once and
for all. The demonstrated ability of the Federal Reserve in the postwar years to judge turning points in economic activity suggests that
discretionary monetary policy can make an important contribution to
achieving tlie goals that Congress has established for the economy.
The choice of a desired growth rate for the base and the money supply
is a means of (1) providing a clear statement on the part of the policymakers of the short-term objective of policy; (2) freeing the Federal
Reserve from the dominant influence of the Manager of the System
Open Market Account, who, as we have seen in chapter IV, has
played a major role in making policy decisions: (3) furnishing a
criterion by which policymakers can judge the action of the Manager
that is related to the aims of monetary policy decided by the Congress
m the Emplovment Act of 1046 and elsewhere: (4) providing more
consistent, less variable, policy operations and less interference in
market processes and operations.
Needless to say, if events occuring in the economy indicate that a
slower or faster growth rate in the monetary base is required to increase emplovment or forestall inflation, the desired growth rate
should be altered. But such decisions should be made by specifying
an alternative monetary growth rate to be achieved at a relatively
steady rate over a period of months rather than in terms of a vague
statement about desired "ease or restraint" for a 3-week period
This report has been dominantly concerned with problems of domestic monetary policy. We have not specificallv discussed the international balance of payments. At this point, however, some comments
on the relation between the international balance of payments and
monetary policy are unavoidable. In particular, the connection between the optimal growth rate of the money stock and the choice between two alternative arrangements governing international transactions requires some comment.
. ..
. Adjustments of international transactions openite through changes
m relative prices (including interest rates) and income. ; Changes m
relative prices are induced either by modifications of exchange rates
or
w by
oy suitable
suitame changes
cnanges in domestic
aoniwsuu price levels under
— — fixed exchange
. •:
,
- • v * • —
—^aoc-PhI rvnarsrhnn. that
rates* A fixed exchange rate r e a m ^ ,
ftjg^^SSSh&SS
in a relatively decelerated growth rate in the base and a correspond
34-474—64
86
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
inff relative deceleration of tlie money stock. This retardation must be
sufficiently large to generate the relative fall in domestic prices and
income required to restore equilibrium. Similarly, in case of a persistent surplus, the base should accelerate and the money supply should
expand more rapidly. The induced rise in domestic prices and income would eventually remove the surplus.
Under a fixed exchange rate serious constraints are thus imposed
on the growth rate of the monetary base. On the other hand, these
constraints are relaxed under a regime of flexible exchange rates.
Such a regime absorbs the transmission of deflationary or inflationary
impulses from one country to another through variations in the exchange rate. This permits the choice of a monetary policy designed
to stabilize prices and dampen fluctuations in output.
The choice between a flexible and a fixed exchange rate thus affects
the selection of policy patterns guiding the growth rate of the base.
I t is important to acknowledge this dependence of a rational monetary
policy on the type of international exchange system. Disregard of
this dependence creates persistent difficulties and eventually forces
difficult decisions.
We submit that the Federal Reserve authorities should face the
choice between fixed and flexible exchange rates. The choice is sufficiently important to justify a careful analysis and an assessment of
the two exchange systems with more than the slogans usually supplied. Moreover, in case the balance of rationally rounded judgment
should tip in favor of a fixed exchange rate system, the Federal Reserve authorities would have to acknowledge that their policies contribute to variations in price levels, output and employment. However, even with the growth rate of the base properly adjusted to the
evolving balance of payments under fixed excliange rates, there would
be no reason for the shortrun gyrations observed in the past. The
gyrations were not induced by the balance of payments. T h e y emanated from the misconceptions of the Federal Reserve about the structure of the monetary process.
V I I . Delegate^ to the Manager the responsibility for a limited
amount of technical operations provided that these operations do not
alter the steady growth rate of the monetan; base—We have stressed
that there is little need for the continuous buying and selling operations carried on by the Federal Reserve under the guise of "defensive
operations." This view has been supported by members of the banking
community, as we noted earlier in this chapter. But technical market
considerations may at times require limited "defensive" operations
to improve the technical position of the market. For example, payment for a Treasury issue may fall on a "double settlement" day.
However, we believe that arrangements, such as those suggested in
the following section, will eliminate the need for many "defensive"
operations.
VOT. Separate the problems of bank regulation from the problems
of monetary controh—The skills and knowledge required to determine
the desirability off mergers, branching, supervision, and other questions
of bank regulation are not the same as those required for making appropriate judgments about monetary policy. There is little relation
between the duties :or the information acquired in the two t y p e s of
activity. Questions of supervision, merger, and branching presently
87
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
occupy a substantial portion of the time of Board members. Such
operations should be separated f r o m monetary policymaking and
should become the responsibility of some other group. 9
The above recommendations are of a general nature. They reflect
the view that the principal problems in Federal Reserve policymaking
are those that have been emphasized in this report: (1) The absence
of a coherent, validated conception of the money supply process and
(2) the (related) failure to carry out the congressional mandate
through countercylical policy operations. These failures of the Federal Reserve could be eliminated without any change in the prevailing
administrative arrangements. Administrative rearrangement, by itself, will neither provide an appropriate conception nor lead to better
policy decisions in the absence of a validated conception, although it
may increase the probability of obtaining both.
Our primary interest is in improving discretionary monetary policy.
We regard the general recommendations—particularly the development of a validated theory and the elimination of concern with extremely shortrun phenomena—as the most important means of
achieving that end. Nevertheless, some specific changes in legal and
administrative arrangements contribute to improved policymaking
procedures provided that a coherent, validated conception of the monetary process replaces the prevailing collection of largely unsubstantiated views. The following section, therefore, suggests some specific
legal and administrative changes that may be useful.
SOME SPECIFIC SUGGESTIONS
I X . The Federal Reserve's power to alter reserve requirement ratios
should be abolished—Discretionary
power to alter the requirement
ratios was given to the Board of Governors in the middle thirties as a
means of increasing the power of the Federal Reserve to control the
monetary system. This increased power was given despite the failure
°f the Federal Reserve to use its existing power to prevent the destruction of a large p a r t of the banking system and the stock of money in
the early thirties. We have found little evidence that suggests that
the Federal Reserve has used this power wisely or well. Furthermore,
our
analysis in this study suggests that most of the c h a n g e s in resen e
requirements in the postwar period have been largely or fully compensated by o f f s e t t i n g open market operations
.
r The analysis in chapters V I and V I I also suggests that there is littie
d i f f e r e n c e f o r the money supply between an open market operation or
magnitude and the liberation or absorption of the same volume
°f reserves through changes in reserve r e q u i r e m e n t s .
Su
ggests that the power to alter reserve r e q u i r e m e n t s is largeJj're
is accomplished by r e s e r v e r e q u i r e m e n t changes
°u d be achieved by open market operations. Unless a
*?ti v e conception of the money supply process is d e v d o p ^ A ^ t t ^
^shes a rationale for such d i c t i o n a r y power*. we v r g f i t ^ t ^ ^
abolished. The principal advantage of i ^ ^ ^ ^ ^ r s
°J
power reduces the Federal Reserve^ ability ^ ^ ^ ^ t s
as those t h a t were made in 1936-37 when reserve requireme
We
re doubled.
dundant—whatever
c
thi7? e p r ° P e r resting place for responsibility
since it raises questions quite unrelated to monetary, pv w
Digitized forreport
FRASER
^
88
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
X. The power to alter vault cash requirements should be abolished.—At present, the Board of Governors is able to vary the percentage of vault cash that is counted as a p a r t of reserves within the
range zero to 100 percent. Variations in the proportion of vault cash
that is included as a part of reserves is equivalent to a change in reserve
requirements. Liberation or absorption of reserves by this means is
an alternative to open market operations.
The Federal Reserve has presented no analysis or evidence that the
power to alter this requirement adds to the "degree of control." Our
analysis indicates that the power is largely redundant in terms of the
"degree of control." Moreover, member banks now hold more than
$3 billion in vault cash that is counted as a part of reserves. A sudden reversal of position by the Federal Reserve, one that eliminated
vault cash as a part of reserves, would have a sizable impact on the
money supply. Such action would produce a decline in the money
stock of approximately $8 billion if not offset by open market
operations.
X I . The discount rate should be a penalty rate.—Many of the shortterm problems in the money market that are used to justify "defensive"
operations^ result from deficiencies in the existing arrangements for
reserve adjustment by individual banks. One such problem is created
by the failure of the Federal funds rate to rise above the discount rate.
If the Treasury bill rate is above the discount rate, many banks are
induced to hold Treasury bills rather than sell Federal funds. The
efficiency of the Federal funds market as a place for adjusting reserve
positions is seriously diminished.
The failure of the Federal funds rate to rise above the discount
rate leads to a rapid increase in borrowing at the Reserve banks, generally followed by a fiat change in the rediscount rate to a level above
the Treasury bill rate. Sudden changes of this kind in the r e d i s c o u n t
rate produce "unsettled conditions" in the financial markets since
they alter the cost and yield relations on those assets that banks use
for short-term adjustment. A penalty rate that removed this source
of difficulty would permit a smoother adjustment to prevailing conditions.
4 Our analysis suggests t h a t the demand by banks for excess reserves
is dependent on prevailing market rates and the rediscount rate.
When market rates are comparatively high, banks reduce the a m o u n t
of reserves that they desire to hold. I t is precisely during such
periods that the Federal funds market could play a particularly
useful role in redistributing reserves among bankers and in
assisting in the weekly or biweekly settlement. All bankers do
not experience the identical rate of expansion (or c o n t r a c t i o n )
lft deposits. Some have reserves that can be profitably lent in the
market. However, the rediscount rate operates as a ceiling above
which banks will not buy Federal funds. Moreover, when the Treasury bill rate js sufficiently above the Federal funds rate, it is profitable
for banks with surplus reserves to pay the higher transaction costs
and acquire bills even if the transaction must be reversed in a few days.
ITie quantity of Federal funds offered on the Federal funds market is
reduced.
These considerations affecting both the quantity supplied and the
quantity demanded, a t times, restrict the role of the Federal funds
89
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
market as a mechanism for adjusting reserve positions. As a result
banks borrow more heavily from the Reserve banks. The Federal
Keserve thus supplies through the discount window some of the reserves that it has attempted to absorb through the open market
account
If one adopts the Federal Reserve view that increased borrowing
by bankers restricts the expansion of the banking system, there is no
conflict between supplying reserves through the window and removing
them through open market operations. But our analysis and evidence
strongly suggest that this interpretation is incorrect and that it confuses the effect on an individual bank with, the effect on the monetary
system.
Under our proposal the rediscount rate would always be above the
Treasury bill rate. Although the two need not be tied together, we
envision a rather close relationship between the two in the direction
and timing of movements. As the Treasury bill rate rose, the rediscount rate would rise also. This would remove the ceiling presently
imposed on the Federal funds rate by the "sticky" rediscount rate.
The rediscount rate would become more of a market determined rate.
Removing the ceiling imposed on the Federal funds rate by the rediscount rate would increase the information content of the latter rate.
Bankers in various parts of the country would have better information on the daily reserve position summarized by the price that is being
paid for reserves, or Federal funds, in the market.
Further, the removal of a restriction on the Federal funds market
as an allocative device would reduce the problems of short-term maldistribution of reserves. At present, total reserves (or total free reserves) may be sufficient^ on a given day, from the Federal Reserve's
view, but the distribution may be poor. "Defensive" open market
operations are conducted to supply reserves and avoid the distribution
problem. A part of the distribution problem occurs because of the
implicit upper boundary imposed on the Federal funds rate. By removing the restriction on the Federal funds rate, the solution to the
roblem of reserve distribution could be given to the individual
ankers. By paying a sufficiently high rate, bankers could attract surplus reserves through the Federal funds market.
We strongly believe that such technical problems should be resolved
by individual bankers insofar as possible. By removing restrictions
imposed on the market for Federal funds, the Federal Reserve could
facilitate solution of the distribution problem through the market
mechanism.
X I I . The discount window should be "open" at a penalty rate.—
The spirit of the original Federal Reserve Act envisioned that bankers
would adjust actual to desired reserves by borrowing from the Reserve
banks. Since the twenties, the Federal Reserve has attempted to impose administrative restrictions on the use of the discount window,
particularly in periods when banks desired to borrow. The rationale
for these administrative restrictions is open to question. If banks do
not borrow for profit but only for "need," and r e p a y borrowing as
promptly as possible as the Federal Reserve has maintained, why
should the Federal Reserve impose administrative restrictions designed to reduce the volume of borrowing to meet these 4 needs ? If
banks borrow for profit, the profitability of borrowing can and would
be eliminated by the use of a penalty rate.
90
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Our proposal to keep the discount window "open" would mean that
borrowing at a penalty rate would be a "right" and not a "privilege"
of membership m the System. A major reason for advocating this
step stems from the recognition of the desirable role that member bank
borrowing can play in the individual bank's reserve adjustment
Many smaller banks cannot participate in the Federal funds market
regularly, if at all, because the minimum size of transactions in that
market is too large relative to the size of their excess reserves. Generally such banks hold excess reserves and do not borrow or lend.
However, even the smallest member banks may at times have reserve
deficiencies. The discount window permits them to adjust their position. The proposal would not change this adjustment procedure other
than granting the member bank the right to expect accommodation.
A second reason for advocating the "open window policy" stems
from our concern with so-called defensive operations. Many of these
operations are designed to smooth the adjustment of individual banks
or sectors of the banking system to the inevitable random movements
that affect their reserve position for short periods. We propose to
make the adjustment to such movements a matter to be resolved by
the individual banker. I f there is a maldistribution of reserves or
some other random event, the banker may, if he wishes to pay the rate,
avoid this problem by borrowing. Such borrowing would solve his
temporary problem and would not involve any substitution of judgment by the Federal Reserve f o r judgment by the individual banker.
I t may be suggested that under prevailing arrangements, bankers
are rarely refused the discount privilege. This might lead many to
believe that this proposal would not provide any new mechanism other
than the penalty rate. Our response to that potential comment is contained in the next recommendation.
X I I I . The discount window should replace open-market operations
as the principal means 6y which the Federal Reserve handles "defensive" operations.—This proposal would transfer the judgment about
"defensive" operations to the individual bankers. Under present arrangements, when the Treasury withdraws balances from an individual bank, the bank may have a reserve deficiency. W h y s h o u l d the
Federal Reserve correct this temporary deficiency by a "defensive"
open-market purchase? I f collection schedules are disrupted so that
Federal Reserve float rises, why should open-market operations at the
initiative of the Federal Reserve be used to adjust reserve positions?
We recommend that, by paying a penalty rate, the individual bank
be allowed to adjust reserve deficiencies at its own descretion. Our
proposal ( X I ) envisions that many of the deficiencies and surpluses
in reserve positions would be resolved through the Federal funds
market. Those that were not resolved in that way could be handled
through the discount window, if the individual banker is willing to
pay the price. I f the individual bankers prefer to take the risk of
acquiring reserves later in the settlement period, we see no reason for
the Federal Reserve to substitute its collective judgment for the decisions of the bankers.
X I V . The effect of member bank borrowing on the growth raft
of the monetary base should be offset—Our proposal V I calls for the
choice of a growth rate for the monetary base to be achieved over
a period of time at a steady rate. Proposal VTI is designed to p r e v e n t
91
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
the Manager from altering the growth rate by large temporary injections of reserves for defensive reasons. The present proposal
makes more explicit the procedure that we have in mind. The Mana ei !
£ * should observe the increase in member bank borrowing and offset
the boirowing by open-market operations. Thus defensive operations
designed to maintain the growth rate of the base that was
decided upon by policymaking officials. As above, banks would be permitted to borrow at a penalty rate to ease the problems associated
with tlie distribution of reserves. But such private decisions should
not interfere with the decision of the body that Congress has entrusted
to control tlie money supply.
Among the advantages of the proposed system, we single out three
for comment : (1) The policymaking officials would have a clear set
of criteria by which to judge the actions of their Manager. (2) The
economy would experience much smaller fluctuations in the money
supply over short periods of time with attendant reduction in the uncertainty presently engendered by Federal Reserve procedures. (3)
The information content of market indicators would be substantially
increased. In particular, the movement of the Federal funds rate
and the size and location of member bank borrowing would convey
quite useful information about the state of the financial markets.
Such information is presently obscured by the ceiling imposed on the
Federal funds rate by the rediscount rate and by the variety of administrative restrictions imposed on member bank borrowing.
XV. Float should be abolished.—A major source of short-term variation in member bank reserve balances and a major reason used to justify defensive open-market operations is the existence of float. This is
purely a manmade, or more correctly a Federal Reserve made, problem.
We suggest-that float should be abolished by making credits and debits
to reserve balances simultaneous.
Under present arrangements the Federal Reserve determines a fixed
schedule governing deferred availability credit, i.e., the schedule of
dates on which the reserve balances will be made available to the banks
whether or not the checks have been collected. The difference between
the amount of uncollected items (or cash items in process of collection)
and deferred availability credit is "float." Float is subject to large
intramonth and seasonal variation owing to the variability of the
volume of checks that must be processed, the failure to meet collection
schedules due to weather and other, often random, events.
Float is an interest-free loan that the Reserve banks make to the
member banks. I t arises solely because the payment schedule determining the conversion of deferred availability credit to actual credit,
that the Federal Reserve imposes on itself, does not match the actual
collection schedule.
We see no clear reason why this problem should be used to justify
a large volume of defensive open-market operations and interference
in the market by the Federal Reserve. If banks are temporarily under pressure because mail schedules or collection schedules are not met,
does it follow that the Federal Reserve should make an interest-free
loan to the banks? The alternative proposed here would abolish the
interest-free loan by replacing the predetermined sruaranteed collection
schedule with the actual collection schedule. This would eliminate
the rationale for many of the defensive open-market operations required to adjust for changes in float.
92
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Sizable intramonthly variation in reserves would occur because of
the intramonth variation in the volume of checks drawn. We suggest that bankers be permitted to solve this problem on their own,
without Federal Reserve interference, by anticipating reserve drains
in advance, by purchasing or selling Federal funds, or by borrowing
at a penalty rate from the Reserve banks. W e are pleased to note that
some bankers partially share our view that bankers would be able to
solve some of these problems, without Federal Reserve assistance, under alternative institutional arrangements.
X V I . Rmnove restrictions on the payment of interest on demand
deposits and regulation of interest rates pa-id on. time deposits.--These
restrictions interfere with the allocative efficiency of the financial markets. They have an important effect on the rate at which demand
deposits are converted into time deposits. The zero nominal rate of
interest on demand deposits is particularly important in this regard.
Permitting banks to pay interest on demand deposits would reduce this
source of instability m the growth rate of the money supply, currency
and demand deposits.
Ceiling rates of interest on demand and time deposits encourage
the growth of nonbank financial intermediaries when interest rates
are comparatively high. More important, little analysis and evidence
has been presented to support this form of price control. The only
detailed study of the effect of payment of interest on demand deposits
that we know 1 0 suggests that many of the asserted dangers arising
from the payment of interest on demand deposits are without evidential support.
X V I I . We suggest that the Federal Open Market Committee be
abolished.—The committee system as a means of regulating the volume
of open-market operations developed from an attempt to coordinate
the semi-independent operations of a number of Reserve banks. This
form was retained in the modifications introduced into the Federal
Reserve Act during the middle thirties.
The committee system is somewhat unwieldy. While there are
officially only 12 members of the committee at any one time, the seven
members of the Board and the 12 Reserve bank presidents participate
in most of the meetings. I n addition, there are a number of vice
presidents, advisers, and staff present and participating in the meeting. While such large meetings may contribute to the development of
an active and at times heated discussion, it is not clear that such discussion contributes to the formation of rational monetary policy.
Indeed, the evidence that we have surveyed suggests the opposite. A
variety of unsubstantiated judgments and unsupported opinions replaces analysis and evidence as the basis for policy operations.
We don't wish to infer that the Reserve bank presidents are more
responsible than the members of the Board f o r the absence of analysis
and evidence as a basis for policy decisions. We suggest, instead, that
the committee system and the discussion procedure is itself a major
contributing factor to the present state of affairs.
Further, we are convinced that much of the information that the
regional officials bring to the meetings can be obtained i n alternative
ways. Under the proposed arrangements t h a t we have s u g g e s t e d , the
J 1 S S 2 E STM*00'
i,Interest
on Demand IXeposits" (multllHbed), University
93
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
volume of member bank borrowing in the various regions and the
Federal funds rate would be a more useful indicator of regional banking situations than they are at present. Moreover, we have indicated
that the primary policy matter to be decided is the rate of growth
of the monetarv base. This is a national, not a regional, matter that
should be decided on the basis of analysis and evidence—not as a matter of imsubstantiated judgment.
X V I I I . Sertoli consideration should be given to the replacement
of the Board of Governors with a single administrative official.—Our
reasons for suggesting this change are much the same as those discussed briefly in connection with recommendation X V I I . We note, in
addition, that the Commission on Money and Credit and the author of
the Hoover Commission Report on the Federal Reserve both suggested
a smaller Board. 11
Our own suggestion assumes that many of our other suggestions
^ill be adopted. There will no longer be a problem of deciding on
branches, bank mergers, etc. There will be no problem of deciding on
reserve requirements or interest rates on time deposits. Problems
connected with the supervision of banks and the inspection of assets
will have been moved elsewhere. The principal problem that will
remain is that of deciding on the appropriate growth rate of the
monetary base after interpretation of the events occurring in the
economy. We know of no compelling reason why this decision requires a Board. Our reading of past history suggests that at crucial
points in our monetary history, the committee system has prevented
appropriate action from being taken because of indecision within the
Board.
CONCLUSION
The principal conclusion of this study has been stated many times
in the present and preceding chapters. *To repeat once more, we find
that the Federal Reserve has failed to develop an understanding of
monetary processes that is adequate for carrying out the mandate
given to" it by the Congress. The recommendations in this chapter
are largely designed to correct that deficiency and other failures of
the present system that prevent or hamper discretionary monetary
policy from having maximum effectiveness in promoting employment
and forestalling inflation.
Our "general" recommendations are our principal suggestions.
Among these, we suggest most strongly the urgent need for development of a coherent, validated framework connecting Federal Reserve
operations with the monetarv process. Without a validated analytical
foundation, the errors that have reduced the usefulness of discretionary monetary policv are likely to occur and reoccur with tragic
consequences similar to^those that havp been experienced on occasion
in the past.
We have not attempted to investigate the complete range of monetary phenomena and have not analyzed some features of policymaking. Consequently, we have not recommended changes in some existing
procedures that are a part of the prevailing system. This failure
" " M o n e y a n d C r e d i t . " t h e r e p o r t of t h e Commission on Money a n d C r e t t t ( ^ p i e w o o a
C l i f f s : P r e n t i o f - H n l l . l f l f l l , pp. 87, 9 0 ) . G. L. Bach, " F e d e r a l Reserve Policymaking' (New
York: Alfred Knopf. 1950).
94
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
should not be construed as an acceptance of these procedures. Many
present restrictions—e.g., the 25-percent gold reserve requirement—
might be usefully abolished. B u t our discussion and analysis have not
taken us in that direction, and we have refrained from making recommendations where we have not supplied analysis.
Finally, we wish to note that we do not regard the type of "monetary rule" that has been suggested by a number of economists as a
viable alternative to discretionary monetary policy and the development of a coherent framework for policymaking. The Federal Reserve does not directly control the money stock but must operate
through the banking system and the public- Without a valid understanding of the monetary process, the Federal Reserve is unlikely to be
able to deliver a constant rate of growth in the money supply. Without prior analysis of the effect of their policy operations on the money
supply, attempts to-control the monetary growth rate seem fruitless.
Once the Federal Reserve has attained a validated understanding of
the monetary process, their power to control the rate of monetary
growth and their demonstrated abilitv to judge turning points seem
to argue in favor of discretionary procedures.
APPENDIX
I
Q U E S T I O N S AND R E S P O N S E S TO QUESTIONNAIRES M A I L E D TO E A C H
M E M B E R OP THE B O A R D OF G O V E R N O R S AND E A C H P R E S I D E N T OF
A R E S E R V E B A N K W H O H E L D T H E POSITION ON AUGUST 2 1 , 1 9 6 3
B O A R D OF G O V E R N O R S OF THE
F E D E R A L R E S E R V E SYSTEM,
O F F I C E OF THE CHAIRMAN,
Washington, September 12, 1968.
H o n . W R I G H T PATMAN,
Chairman, Committee on Banking and Currency,
House of Representatives, Washington, D.C.
D E A R M R . C H A I R M A N : On August 2 1 , 1 9 6 3 , you addressed letters
to each of the members of the Board enclosing questions that had been
prepared by Drs. Brunner and Meltzer relating to a study of several
aspects of the effectiveness of monetary policy.
The members of the Board have considered these questions and
have indicated their substantial agreement on the answers they
would supply. Accordingly, the members of the Board join in submitting the enclosed set of answers to the eight questions you
presented.
Sincerely yours,
W M . M c C . MARTIN, J r .
Enclosures.
C O N F E R E N C E OF P R E S I D E N T S OF T H E
FEDERAL RESERVE BANKS,
September 10, 1963.
H o n . WRIGHT
PATMAN,
Chairman, Committee on Banking and Currency,
House of Representatives, Washington, D.C.
D E A R M R . C H A I R M A N : I am writing to you with reference to your
recent letter addressed to each of the Federal Reserve bank presidents
and enclosing seven questions prepared by Drs. Brunner and Meltzer
in connection with a study of several aspects of the effectiveness of
monetary policy.
The Federal Reserve bank presidents have had an opportunity to
discuss these questions, and they have concluded that they are in
substantial agreement with respect to the answers to these questions.
They have, therefore, concluded t h a t it would be expeditious to
respond to your request by submitting a single set of answers which
they have worked out among themselves.
95
96
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Accordingly, I am transmitting herewith the joint answers on behalf
of the presidents of all 12 Federal Reserve banks.
Yours sincerely,
W A T R O U S H . IRONS, Chairman.
Enclosures.
Question
I. The annual reports of the Board of Governors frequently refer to
the desire to accommodate the credit needs of commerce and industry.
1. Are certain classes of assets or liabilities used to measure
the credit needs of commerce and industry? If so, which are
they?
2. Are other specific measures used? If so, which are they?
3. What specific criteria do you use to judge whether these
needs are adequately met? Cite the relevant data for a period
when inadequate accommodations obtained?
4. Does the requirement under the Federal Reserve Act to
accommodate the needs of commerce and industry involve more
than a seasonal adjustment of the items in question 1? If it
does, what other adjustments are involved? A concrete example
covering a given time period will be helpful.
5. Does the requirement under the Federal Reserve Act to
furnish an elastic currency involve more than a seasonal adjustment of the amount of currency outstanding? If it does, what
other adjustments are involved? Do you understand currency
in this context to be only Federal Reserve notes? If not, how
do you define currency?
6. During 1949 and 1961 (as well as at other times) the sum of
currency held by the public plus demand deposits adjusted
grew at a lower rate than its longrun average growth rate. Is
this behavior of the money supply appropriate under the Employment Act of 1946? (If your answer suggests that time
deposits of commercial banks should be included in the money
supply, kindly indicate the principal reasons for this choice of
definition as part of your answer.)
Answer Supplied by the^Board of Governors
Parts 1-4
Credit needs of commerce and industry cannot be measured solely
by any specific classes of assets or liabilities. In addition to the
direct credit needs of manufacturers and merchants to finance outlays
for inventories and equipment, credit is needed to finance c o n s u m e r
purchases of the products of commerce and industry, credit is n e e d e d
by agricultural and other suppliers of raw materials and services used
in commerce and industry, and credit is needed by g o v e r n m e n t s to
provide the community services essential to the proper functioning of
the production and distribution processes.
The Federal Reserve is not empowered to and does not a t t e m p t to
allocate credit to any of these specific uses; in our free e n t e r p r i s e
economy such allocation is a function of the marketplace. Nor is
the Federal Reserve the principal source from which these credit
needs are met; the bulk of all credit arises from the flow of c o n s u m e r
and business financial saving. It is the responsibility of the Federal
Reserve to insure that the marginal element in the total credit supply—
97
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
bank credit—when combined with the flow of financial saving, is
adequate to meet the aggregate credit needs of an expanding economy.
Adequacy in this context is not a matter of satisfying all credit
demands at all times nor, on the other hand, is it a matter of meeting
only seasonal fluctuations in business credit demands. Judgments as
to the adequacy of credit availability must balance the overall credit
needs for longer term economic growth against the current availability
of productive resources to accommodate additional credit-financed
expenditures without upward pressure on prices. The impact on
credit supply and demand of other governmental policies must also
be assessed. All of these factors are weighed in arriving at judgments
as to whether the credit needs of commerce and industry are being
met adequately, and the record of policy actions taken by the Board
of Governors and the Federal Open Market Committee (published in
detail in the annual reports of the Board of Governors) indicates the
wide range of economic data reviewed in arriving at such judgments.
Part 6
In the context of the question, currency is understood to mean
currency in circulation outside of the Treasury and the Federal
Reserve banks. Of the total of such currency in circulation (S35.5
billion at midyear 1963), $30.3 billion is in Federal Reserve notes,
and the remaining $5 billion is principally in the form of silver certificates and subsidiary silver coin.
The public obtains additional currency through the commercial
banking system by exchanging demand deposits for currency. I t is
free to do so at will. Technological and social changes, such as those
that have made checkbook banking so popular in this country, have
limited the secular growth in demand for currency, but there are
still sharp seasonal fluctuations in the public's currency needs.
Commercial banks meet these fluctuations by adjusting their own
holdings of currency or by obtaining additional currency in exchange
for their reserve balances at Federal Reserve banks. An increase in
the public's preference for currency as against deposits increase these
reserve needs of the banking system, for each dollar of increase in
currency demanded by the public represents an equivalent drain on
member bank reserve positions offset only in part by the reduction
in the fractional reserves required to be held against deposits. Along
with other factors affecting reserves which also may have wide but
fairly predictable seasonal variations, seasonal changes in cuirency
demands are anticipated and provided for in the implementation of
monetary policy. Secular growth in currency needs is accommodated
within the framework of providing an adequate reserve base to meet
the credit and liquidity needs of an expanding economy growing at a
sustainable rate.
Part 6
Changes in the money supply result from the prevailing posture of
monetary policy as well as many other factors. Among the most
important of these other factors are the economy's demands for bank
credit, public preferences for holding liquid assets in particular forms,
and the incentives for banks to make loans and purchase investment.
I n recession periods, credit demands and bank incentives to acquire
loans and investments tend to decline, while public preference for
liquid assets tends to rise. In 1949 and 1961, for example, evidences
98
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
of recession were apparent in the course of the money supply and
other financial magnitudes as well as in real economic activity. The
Federal Reserve was following an active countercyclical monetary
policy in both years, as would be appropriate under the Employment
Act of 1946.
Resulting stimulation to economic activity was not such as immediately to create vigorous expansion. That takes time and monetary policy cannot do the job by itself. In any case, the average
longrun rate of growth in the money supply is not a measure of the
desirable addition to the supply in each individual year, for conditions
influencing the supply of and demand for money vary greatly from
time to time and affect the rapidity with which the economy responds
to shifts in policy.
Amswer Supplied by the Presidents
1. In the current context, the languago of the Federal Reserve
Act 1 must be interpreted in a broad sense and the goals of monetary
policy defined in terms transcending purely monetary and credit
variables. The credit needs of commerce and industry cannot be
isolated and considered apart from the credit needs of the economy
as a whole, and these, in turn, are related to the performance of the
economy. Credit needs cannot be related to specific classes of assets
or liabilities. Nor are all credit needs of any segment of the economy
met by the commercial banking system alone, and, indeed, bank
credit plays a marginal role in our credit structure.
Monetary policy is concerned with the overall availability of credit,
but the allocation of credit among competing needs occurs in the
marketplace. The process of credit allocation is studied closely by
System officials in order to ascertain the uses made of credit r e s o u r c e s
and in particular of the marginal amounts made available as a result
of System policy actions. At times, the monetary authorities may
want to influence credit flows, but their policy actions do not aim at
identifying or meeting specific credit needs of individual segments of
the economy.
Data relative to the assets and liabilities of the banking system
contribute importantly to an understanding of the credit needs of
the economy, but by themselves they are not sufficient. Moreover,
they cannot be interpreted mechanically; they must be interpreted
in the light of the lessons of the past and the anticipated problems of
the future; this requires both skilled analysis and judgments based on
detailed knowledge of the economy as a whole (with due regard to
regional differences and needs).
Within this framework, changes in bank assets and liabilities must be
examined closely. Changes in member bank reserves and in borrowings from Federal Reserve banks provide a measure of the capacity of the banking system to expand credit. Similarly, changes in
bank loans, by category, provide some measure of the strength of
demands for the various types of credit, nationally and by region,
and of the willingness or ability of the banking system to a c c o m m o d a t e
them. Changes in batik holdings of Treasury obligations, and especially changes in the maturity distribution of these holdings, when
viewed against movements in deposit liabilities, shed some light on
whether increases in the loan totals of the banks are being limited by
of coarse
'
Of "accommodating commerce and business." See,
sees. 12A (3){c)
99
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
liquidity and portfolio considerations or by lack of strong demand from
borrowers. Similarly, changes in savings and time deposits, when
compared with changes in mortgage Joans and other long-term
assets, help to provide a clearer picture of the bank activity in these
sectors of the capital market. Relative rates of growth of demand
deposits and time deposits help to provide some measure of shifts in
the public's preferences for various types of liquid assets, which in
turn is relevant to the economy's need for expansion in the money
supply, as defined conventionally. Loan-deposit ratios are examined,
because these help in the assessment of the liquidity position of
banks.
Changes in credit extended by other financial intermediaries and
iu interbusiness credit are other important considerations. Developments in the banking area are viewed against the broader
context of a flow-of-funds analysis encompassing the entire network
of financial relationships.
2. Statistical measures which help to gage the rate of actual and
potential growth of the economy, the availability and degree of utilization of labor and other physical resources, the various sources of
savings and of credit, the sources and nature of pressure on prices,
and forces shaping the balance of payments are examined in the process of determining the credit needs of the economy.
It should be stressed, however, that no statistical data, however
complete and up to date, can provide a precise determination of the
economy's credit needs. Statistics never tell what the situation will
be in the period immediately ahead; they merely provide the underpinnings for interpretation and assessment.
3. Monetary policy is concerned with the national economy as a
whole and with its relation to the outside world. The national
economy, of course, is made of many parts and sections and overall
policy can be determined intelligently only with an understanding of
developments in these sectors. Nevertheless, the overall objectives of
policy have to do with the economy as a whole rather than the
problems of specific areas or groups.
The principal criteria employed in judging whether the credit needs
of the economy are being adequately met are aggregative and national.
By concerning itself with the overall supply of money and credit,
monetary policy relies on the marketplace for distributing this
su
pply. Needs are never absolute, and therefore cannot be
"measured" except through the process of market allocation usmg
the price mechanism. In an economic system which aims at preserving monetary stability, credit is. among those resources the supply-of
which is limited. Our economy is much too complex for its credit
and liquidity needs to be gaged solely by statistical criteria, including
norms and patterns derived from past performance.
.
4. Only an unreahstically narrow interpretation # of the federal
Reserve Act would confine System actions to meeting the seasonal
demands for credit. The broad goals of national economic poncy
require an active and imaginative policy to make a m a ^ m n contribution toward.achieving high levels of employment of the Nations
resources, stable prices, a balanced and sustainable rate ot growm,
and equilibrium in our international payments. Meeting recurrent
and largely predictable seasonal needs does not normally w p j g i a
significant challenge to monetary -policy.:. Major policy probems
usually arise in connection with the System's endeavors to attenuate
100
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
movements and to accommodate long-term growth needs
of the economy. In general, these other adjustments require supplying reserves in~amounts either greater than or less than seasonal needs
alone would appear to require. For example, in the fourth quarter
of 1960 monetary policy was one of expansion, and reserves were
supplied in amounts exceeding normal seasonal needs.
5. In our type of monetary system, the public makes the ultimate
decision as to the distribution of its cash holdings as between currency
and demand deposits, and as between money holdings and holdings of
other types of liquid assets.
The Federal Reserve System makes no distinction between Federal
Reserve notes and Treasury notes and subsidiary coin, supplying the
public with the amounts and denominations of currency desired.
Obligations of the Reserve banks, which constitute by far the largest
segment of currency, and that of the Treasury, are all legal tender.
The System supplies currencv to the member bunks to meet the wants
and needs of the public and accepts currency from the banks when
the seasonal demand has passed. In addition to seasonal needs, the
System must take into consideration the need for additional currency
to support the growth of the economy.
6. The public's demand for money is not a stable function of time.
It is subject to many forces (such as the intensity with which nonbank
financial intermediaries seek to increase their liabilities) and to many
sets of expectations, especially regarding prices and interest rates.
The System constantly endeavors to assess these foices and expectations, including the use made by the banking system of reserves
recently supplied, and through such a process of analysis and judgment reaches its policy decisions, including those "regarding the
appropriate increase in the money supply.
We, therefore, do not regard a fixed Vate of growth in the money
supply, defined as currency outside banks plus demand deposits adjusted, as necessarily an appropriate objective in all circumstances.
Circumstances may at times require money supply to exceed and
sometimes to fall below the long-term trend line. The demand for
means of payment will depend, among other things, on relative
preferences for other types of financial assets, of which time deposits
(e.g., 1961) and Treasury bills are among the most important. If the
Federal Reserve were to attempt to force an increase in the money
supply at a faster rate than the public was willing to add to its cash
balances at prevailing price levels, the result would be rising prices
and aggravation of the balance-of-payments situation, rather than a
promotion of sustainable economic growth.
Question
II. The Annual Report of the Board of Governors for 1961 contains
the following statements in a discussion of the FOMC meeting on
July 11, 1961. (Similar statements are found elsewhere; e.g., March
25, 1959.)
" I t was the consensus that open-market operations
should be designed to maintain approximately the same
decree of ease that has recentlv prevailed, associated
with a free reserve level of around $500 to $600 million."
1. Are the level of free reserves and the distribution of free
reserves significant factors in judging ease and restraint? If'SO,
how is each related to the degree of ease?
cyclical
101
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
2. Are the level and distribution of free reserves the principal
factors shaping the degree of ease or restraint? If not, what
are the principal factors?
3. Under the conditions prevailing on July 11, 1961, were
there any factors that would have made the existing situation
tighter than indicated by the prevailing level of free reserves?
If the answer is "yes," indicate the factors.
4. Has the level of free reserves of S500 to $600 million always
been considered easy? If not, what specific factors have modified
your interpretation of the prevailing level of free reserves? When
did this occur?
Answer Supplied by Board oj Governors
Part 1
Ease and restraint are relative terms. Taken by itself, without
reference to existing economic and financial conditions, the absolute
level of free reserves; that is, the excess reserves of member banks
less their borrowings from the Federal Reserve banks is not a reliable
or significant factor in judging the relative degree of ease or restraint.
If, however, it is related to existing conditions—and in particular to
what may be called "the desired level of free reserves" by member
banks—the level of free reserves provides a useful indication of the
degree of ease or restraint. By the desired level of free reserves of
member banks, is meant essentially the desired level of their borrowings from the Reserve banks, for the desired level of their excess
reserves changes only infrequently.
At a time when customer loan demands are slack and short-term
market rates of interest are below the discount rate, the desired level
of free reserves tends to be high; that is, member banks have little
incentive to borrow from Reserve banks and they prefer a relatively
high level of excess reserves. At such a time, an effort by the Federal
Reserve to maintain a low level of free reserves would lead to actions
by the banks, as they attempted to restore the desired level, which
would tend to reduce "the availability of bank credit and raise interest
rates.
On the other hand, if loan demands were strong and short-term
market rates were near or above the discount rate, banks in the aggregate would be willing to maintain a higher level of borrowing and
a somewhat lower level of excess reserves. In such circumstances,
maintenance of the same level of free reserves specified in the previous
example would probably involve a condition of relative ease. A
given level of free reserves can therefore be associated with either a
rising or declining volume of total reserves, bank credit, and bank
deposits, depending upon economic and financial conditions as they
are reflected in demands for bank credit.
Thus the level of free reserves is related to the degree of ease'in the
same sense that supply of a commodity is related to price pressures
on that commodity: the effect of supply on price depends also on the
state of demand. Interaction between supply of and demand for free
reserves gets reflected in the rate of expansion in total and required
reserves. For instance, as the supply of free reserves is maintained
large relative to demand, total and required reserves will expand as
banks p u t newly available reserves to work. Thus, the actual level
of free reserves may be unchanged for a period, or may even be declining some, while total reserves and bank credit are expanding,
34-474—64
8
102
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
because banks are utilizing the flow of new reserves to add to their
loans and investments, and thereby to their deposits.
The distribution of free reserves is at times significant in interpreting
very short-term developments. Excess reserves are ordinarily concentrated largely at country banks; Reserve city banks maintain very
low levels of excess reserves. Occasionally, however, excess reserves
may shift between country and Reserve city banks, depressing or
elevating the level at one group and causing the opposite temporary
movement at the other group of banks. Because country banks react
less quickly than city banks to changes in their reserve positions, a
given level of free reserves may temporarily have different implications for the degree of ease or tightness in the money market, depending on distribution.
Part2
The level of free reserves is, as noted above, an indicator of the
degree of ease or restraint if interpreted in the light of prevailing
demand conditions. The principal factor shaping the degree of ease
or restraint is the total supply of loanable funds in relation to the
demand for such funds. Federal Reserve policies are directed at
influencing one element in the supply of credit and money through
the bank reserve mechanism. Federal Reserve policy, as a result of
its significant influence over one blade of the demand-supply scissors,
is a principal factor in shaping the degree of ease or restraint. Federal
Reserve policy with respect to the supply of bank reserves wall be
reflected in the level of free reserves; furthermore, the Account management may at times receive instructions from the FOMC in terms
of a level of free reserves, as in July 1961. Whatever the technical
means of implementing FOMC policy decisions, the principal factor
on the supply side subject to Federal Reserve policy is the flow of
reserves to member banks.
Part 8
No. There is no evidence of a change in the factors that might
have influenced banks' preferences regarding excess reserves or borrowings at that time.
Part 4
In recent years, a level of free reserves of S500 to S600 million
has been considered relatively easy. This is so because the desired
level of excess reserves by member banks as a group has seldom
risen above this range. Since borrowings seldom decline much below
$100 million, a level of free reserves of $500 to S600 million implies
that Federal Reserve policy is providing a level of excess reserves
significantly above the level that banks desire to hold in that idle
form. The result is t h a t banks normally put these additional reserves
to work by acquiring earning assets. As the Federal Reserve continues to maintain free reserves at this level, total reserves will rise
and banks will continue rapidly to increase their loans and investments. Periods in which free reserves have been at the level of §500
to $600 million have usually been periods of rapid credit expansion.
Answer Supplied by the Presidents
1. Free reserves are one of a number of indicators used by the Open
Market Committee and the Manager of the System Open Market
103
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Account to determine the impact of market forces and of monetarypolicy on credit conditions during the relatively short time intervals
between Open Market Committee meetings. The weekly figures on
free reserves are also followed by the informed public as an early
indicator of possible shifts in Federal Reserve credit policy. Almost
never can they be used as the single, and usually they are not even
the most important indicator of credit conditions, even in the short
run, as no given level can be interpreted out of context. Free reserves
(or their negative counterpart, net borrowed reserves) are most
useful if analyzed as part of a complex of factors which includes
bank credit, short-term interest rates and the other indicators of
money market conditions (the "feel" of the market). Free reserves
have their greatest usefulness in appraising conditions over relatively
short periods; over longer periods the significance of marginal reserve
availability is superseded by other indicators that throw more light
on the behavior of bank credit and related variables in response to
past levels of marginal reserve availability.
Different patterns of distribution of free reserves tend to have
different credit and money market effects. I t is helpful in understanding reactions to credit policy actions on a day-to-day basis to
analyze the distribution of free reserves and the factors which account for it, such as an unexpected bulge in float due to weather
conditions. Knowledge of the distribution is also helpful in projecting
near-term developments when marked changes occur in the level of
free reserves owing to changes in required reserves (for example,
as a result of changes in applicable percentages or bulges in Treasury
financing), or to recurring factors such as the completion of reserve
settlement periods.
A given level of free reserves, such as the range given in this question, may at one time be associated with a faster or slower rate of
growth in total bank reserves and bank credit than the same level at
a different time.
The directive to the Manager of the Account is generally not
written in terms of free or net borrowed reserves. At the FOMC
meeting of July 11, 1961, after a broad review of the business and
credit situation including developments in demand deposits and total
money supply and related monetary data and an extended discussion
of these factors, the Committee members concluded that the directive
could be carried out over the subsequent 3 weeks by attempting
"* * * to maintain approximately the same degree of ease that had
recently prevailed, associated with a free reserve level of around
$500 to $600 million." 1 This analysis was summarized in the statement from which the passage quoted in this question was taken.
The formal directive given to the M a n a g e r , however, was stated
in broader terms.
I t is not unusual for individual members of the Committee in
discussions held in the presence of the M a n a g e r (which helps him to
understand the full range of considerations which have led to the
adoption of the necessarily short formal directive) to mention a level
of free reserves or other short-run guidelines such as market "tone"
or "feel," Federal funds rates, or Treasury bill rates. But when it
i Annual Report ot the Board of Governors, 1061, p. 69.
104
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
has been judged desirable to mention specifically bank reserve positions in the formal directive, the reference has been usually to total
reserves, which, indeed are also the magnitude more relevant from
a longer term point of view.
2. The level and distribution of free reserves are merety one of the
indicators of current money market conditions. They have to be
interpreted in the light of what is happening generally to the total
credit extension by member banks. The degree of ease or restraint
in the credit markets is a product of the interaction of the supply of,
and demand for, reserves. The effects of changing degrees of ease or
restraint are discussed in answers to another question.
3. Money market conditions prevailing at the time of the July 11,
1961, FOMC meeting were easy and there were no factors that would
have made the existing situation tighter than indicated by the prevailing level of free reserves. Compared with the then discount rate
of 3 percent, the effective rate on Federal funds that week averaged
1.36 percent (the month of July as a whole averaged 1.10 percent,
against 1.73 percent for June). On July 10 the 3-month Treasury
bdl sold at an average rate of 2.322 percent, only slightly above the
average for the previous week, while the average for July as a whole
was 2.268 percent. Total commercial bank credit expanded 0.8 percent during June and 0.7 percent during July to levels 7.6 and 7.8
percent, respectively, above those prevailing a year earlier. Other
measures also indicated a general condition of ease.
4. As noted earlier, it is not possible to associate a particular
level of free reserves uniquely with a particular credit condition.
The credit situation may at times be easier at a lower level and less
easy at a higher level.
However, since the "Accord," whenever free reserves have been for
some time in the area of $500 to $600 million, the money market and
bank reserve positions have been easy and credit policy expansive.
The months from the beginning of 195*4 to date in which the monthly
average level (based on daily figures) of free reserves fell in this range
are shown in the accompanying table. The table also suggests a few
of the problems associated with the use of free reserves as an indicator
of the monetary situation.
Some of these months were periods of active, possibly even "aggressive" ease (for example, 1954); others represent periods when policy
was beginning to become less easy (for example, mid-1958). The
average level of free reserves for individual weeks within these months
fluctuated from a level of less than $300 million to more than a billion.
The difference between monthly average Federal funds rates and the
discount rate ranged from less'than one-half of 1 percent to almost
2 percent. Changes in bank credit on a month-to-month (seasonally
adjusted) basis ranged from a decline of 1.0 percent to an increase of
2.3 percent and on a year-to-year basis from + 2 . 5 to +8.4 percent.
The outstanding average level of borrowing was in the range of $37
to $189 million. Conceivably, higlier levels of borrowing (and correspondingly higher levels of excess reserves) could be associated with
free reserves of between $500 and $600 million
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
Months when free reserves have averaged $500,000,000 to $600,000,000,
credit market indicators9 1954-63
Free reserves
Month and year
Weekly
average
range
1954—March,
April..
May
1953—March.,
April...
May...
June
July....
1960—Octobe r
November...
1961—February...
March
April
May
-Time
July
August
September..
October
November...
December...
1962—January
1
Borrowing
(daily
Monthly average)
dally
average
Change in total
bank credit 1
Year
to
year
Month
to
month
105
and other
Average Average
auction effective
3-month Federal
Treasury
funds
bills
rate
Discount
rate
XfiUionxXftUiontMillion)PercentPercentPercentPercentPercent
$2S6- $557
460- $37
535- 625
470- 524
416- 552
511- m
404- 669
45*V- 704
287- 802
37S-1.084
359- 618
333- 629
42*- 604
406- 523
504- 558
499- 634
402- 560
484- 630
429- 492
3S2- 620
344- 527
365- 657
$503
6 26
561
495
492
547
4S4
547
4S9
614
517
476
551
453
549
530
537
547
442
517
419
546
$189
139
155
133
130
119
142
109
149
142
137
70
56
96
63
51
67
37
65
105
149
70
+2.5
+4.1
+5.3
+5.2
+6.2
+6.1
+8.4
+7.3
+3.8
+4.2
+6.9
+6.7
+6.0
+7.2
+7.6
+7.8
+7.6
+8.3
+7.5
+7.9
+7.8
+7.8
-0.4
+.9
+.7
+1.8
+1.5
+.4
+2.3
-1.0
+.9
+.3
+1.6
-.2
-.4
+L4
+.8
+.7
+.3
+1.3
+.2
+.6
+.6
+.5
1.053
1.011
.782
L354
1.126
1.046
.881
.962
2.426
2.384
2.408
2.420
2,327
2.288
2.359
2.268
2.402
2.304
2.350
2.458
2.617
2.746
1.26
.85
.86
1.20
1.26
.63
.93
.68
2.47
2.44
2.54
2.02
1.50
1.98
1.73
1.16
2.00
1.88
2.26
2.62
2.33
2.14
1U
1H
2J4-3
IK-2K
1%-2K
1u
1%
3
3
3
3
3
3
3
3
3
3
3
3
3
3
Seasonally adjusted.
Question
I I I . The April 1963 Federal Reserve Bulletin contains an article by
Mr. Robert Stone, surveying Federal Reserve open market operations
in 1962. The article states that money was easy in the first 5% months
of 1962, slightly tighter for the next 5 to 6 months, and still tighter
at the end of the year.
During 1962, long-term interest rates fell, short-term interest rates
were relatively stable, and the money supply increased. At the end
of the year member bank borrowings rose and free reserves declined.
1. Was the increase in member bank borrowings and the associated decline in free reserves the sole indication of the relatively
greater tightness?
2. If not, what other specific indications of relative tightness
are applicable?
3. Did the increase in relative tightness retard the asset expansion of member banks under the conditions of 1962? If not,
what did the relative increase in tightness achieve?
Answer supplied by the Board of Governors
Federal Reserve open market operations contributed to expansion
throughout 1962, but at times the contribution was lessened slightly
because of concern about the persistence and severity of the balance-ofpayments problem. Nevertheless, money and bank credit was at no
time during the year "tight."
The change in monetary policy that developed over 1962 was
reflected mainly, if not entirely, in money m a r k e t conditions, particulariy short-term interest rates. There was some increase in member
106
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
bank borrowings from tlic Federal Reserve banks and some reduction
in free reserves—excess reserves less borrowings -particularly late in
the vear. There was also some increase in the Federal funds rate—
the interest rate on excess reserve balances lent among banks—as well
as in most other short-term rates of interest. The movements of
short-term rates also reflected continued Federal Reserve efforts to
supply reserves to member banks in ways that minimized downward
pressures on such rates.
The aggregate volume of reserves supplied to the banking system
increased significantly throughout 1002. with the pace of increase
slightly slower in the second* half of the year than in the first half.
Similarly, the bank credit increases that accompanied such reserve
expansion tended to be larger prior to mid-1902 than thereafter.
Over the year as a whole, with reserve funds continuing to be readily
available and with very large inflows of time and savings deposits
loans and investments outstanding at banks ro>e SI9 billion, a record
annual amount for the postwar period.
Whether the second half growth in bank credit was retarded in any
meaningful way by the lessening of monetary ease is a question that
cannot be answered conclusively. Some individuals fell that there
was such an effect; others did not. In any case, many felt there was a
need for raising short-term interest rates to aid in reilucing incentives
to outflows of short-term capital abroad and that this warranted a
moderation of the prevailing degree of monetary ease.
Answer Supplied by the Presidents
Mr. Stone's April Bulletin article did not refer to Svstem policy as
being "slightly tighter" after the first 5!J months of *1WJ2 and "still
tighter" at the end of the year. Rather, the characterization was
that policy shifted toward "slightly less ease" after mid-June while
a similar shift occurred in December. (See pp. 421) and 431 of the
April 1903 Federal Reserve Bulletin.) The difference in wording is
small but still significant in that the policy changes that occurred
during 1903 were minor shifts within the context of a broadly easy and
stimulative policy.
The modest sliifts toward lesser ease (subquostions 1 and 2) wore
reflected, as the article notes, in a slightly lower range of free reserves
resulting from a lower level of excess reserves and a slightly higher
average level of member bank borrowings. In addition, the modest
firming was reflected in higher Federal fund rates. The lessened
availability of reserves relative to the demand for them meant that
banks in need of additional reserves had to pav slightlv more to
acquire the less freely available supplies from 'banks with excess
reserves, and there were more frequent occasions when demand for
reserves could not be satisfied in the Federal funds market and were
met instead through member hunk borrowing at the discount window.
1 he modest reduction in monetary ease was also reflected in Treasury bill rates. Similarly, rates on other very short-term money
market instruments, such as rates on directlv placed finance company
paper and rates charged on loans to Government securities dealers by
Aew lork City banks also moved up slightlv during the vear.
As suggested by the modest change in the indicators cited above,
there was only a slight reduction of ease in the money market during
1902, and accordingly it is not surprising that no significant retardation of bank credit growth showed up (subquestion 3).
107
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
The slight shift in the posture of monetary policy in the course of
1962 was not intended to retard the asset expansion of member banks
in a business climate overshadowed bv an unemployment rate stubbornly hovering around 5% percent of tlie labor force. What the
modest firming of System policy during 1962 was intended to accomplish was u slightly higher level of short-term interest rates to help
reduce tho balance-of-pnyinents deficit and to signal our determination
to use monetary policy tools to defend the international position of the
dollar. The outflow of short-term capital to foreign financial centers,
where higher rates of return may be earned, has been an important
contributing factor in our baiance-of-payments problem. While
higher short-term rates in tlie United States did not cure the problem
in 1962, and still have not done so. they have made a useful contribution without which the present problem would probably have been
worse.
Question
IV. During periods of Treasury refunding or other Treasury debt
operations, tho Federal Reserve has often been described as attempting to operate on "an even keel."
1. What specific factors constitute "an even keel"?
2. Does the "even keel" mean that interest rates are kept
constant?
3. Does tho "even keel" mean that free reserves are kept
const ant?
4. Does the "even keel" mean that reserves are supplied in
sufficient quant it v to absorb the debt into nondealer positions?
5. Has the "even keel" policy introduced conflict with seasonal
or long-run objectives of monetary policy? If "yes," indicate
some specific examples.
Answer supphd by the Board of Governors
Stated in general terms, the"Federal Reserve practice of operating
"on an even keel" during periods of Treasury financings implies that
the System avoids any actions which might be interpreted by purtiein u i i t s !« i l — Ti
*
—
1
-
. . financings no changes :
rat^s or in momber bank reserve requirements and that changes in
oank reserve t 0availability
and in money market conditions are not
ex mil
credit p^
P fions of a possible shift in money and
An "even keel" policy, however, involves no commitment to stabilize
prices and yields on Treasury securities. Accordingly, such prices
and yields may fluctuate as a'result of market influences other than
monetary policy shifts, such as changes in demand and supply of
funds in tho capital markets or developments affecting the outlook for
interest- rates. In this way tlie Treasurv is better able to prico and
oiler its new issues on the basis of vields determined by market- forces
of u
s PPly and demand rather than by official intervention. To this
end the System also ordinarily avoids operations in "rights'' to
treasury xefundings, in "when issued" securities, or in other issues
' maturities adjacent to those involved in Troasury financings.
While the "even keel" policy does not involve any commitment
to stabilize interest rates, the System would intervene to correct a
120
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
"disorderly" market- situation involving cumulative and disruptive
movements in prices and yields of U.S. Government securities.
The "even keel" policv does not mean that "free reserves are kept
constant." Although large changes in free reserves may be precluded
bv the need to maintain a stable tone 111 tlie money market and by ihe
desirability of not giving apparent signals of a monetary policy shift
lo the market, the*maintenance of an unchanged monetary position
can be associated with fluctuations in free reserves, as indicated in
the answer to question II.
The "even keel" practice does not mean that reserves are supplied
in sufficient quantities to absorb the debt into nomlealer positions.
Rather, it means that the System provides reserves 011 a temporary
basis to assure the successful distribution of new Treasury issues bv
private underwriters, including banks. Thus, increases m renuired
reserves needed to cover expanded tax nnd loan accounts at banks are
usually supplied by the System in order to keep money conditions
from tightening on the payment date for the new Treasury issue.
In addition, money market "pressures resulting from the need of nonbank Government securities dealers to finance holdings of new
Treasury securities may also require some action by the System in
providing repurchase agreements.
In recent years tlie "even keel" policy has not conflicted seriously
with seasonal or longrun objectives of monetary policy. On occasion,
particularly in the 1950's. "even keel" considerations have affected
the timing'of Federal Reserve policy shifts, but they have not imposed
a severe limitation on monetary policy execution" in the longer run.
Answer Supplied by the President*
1. Essentially, the maintenance of an "even keel" at times of
Treasury financing means that tlie System normally seeks to avoid
taking action that would tend to alter basicfinancialmarket conditions
for a period of time shortly before, during, and shortly after Treasury
financing operations. Treasury financings, which are typically
large in size, fully engage the capacity of the Xation'* financial markets.
Any major change in background conditions during a period of financing could cause serious disruption in the Nation's financial mechanism.
In exercising its monetary powers, the Federal Reserve System strives
not to complicate the debt management operations of the Treasury.
^ The Treasury, on its part, is aware of the need to minimize difficulties for monetary policy which may arise from "even keel"
considerations and it has been making a Mgnilicant eirort to arrange
its financing schedules in such a way a- to leave the System's freedom
of action as little impaired us possible.
At the same time, the maintenance of "even keel" conditions does
not mean that the System would so conduct open market operations
as to insure the success of any and every Treasury financing operation.
Rather, the System's role is one of eschewing new actions on its own
part that might interfere with Treasury financing. Thi* means that
Treasury financings must "meet the test of the market", with the
System maintaining as neutral a stance as possible.
Specific "even keel" criteria are not readily defined, since every
market situation is unique, but- a few illustrative references may help
to clarify the point. Almost certainly, it would be contrary to
"even keel" criteria to change the discount rate or undertake a major
109
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
shift m reserve availability during a period of major Treasury financing. On the other hand, a modest shift in policy emphasis during a
period of minor Treasury financing would not necessarily be ruled out.
2. As already indicated, the System would avoid taking any action
in an "even keel" period that would tend to change the existing
atmosphere, but the System would not ordinarily resist any upward
or downward rate tendency that emerged as a result of market flows
and expectations or represented the continuation of trends that had
been at work in rccent months. Only if rate changes threatened to
become very abrupt, and beyond the market's absorptive capacity,
might the System intervene to cushion the adjustment.
3. An "even keel" period would not necessarily mean that free
reserves were "kept constant," but the System would usually try to
avoid significant changes in net reserve availability in such periods.
In practice, if in the period before a Treasury financing free reserves
had been fluctuating, say, in a range of $200 to $400 million, there
would be some effort to remain within that range through the financing
period, unless some unusual developments in terms of reserve distribution called for a different reserve level in order to maintain reasonable
continuity in the money market atmosphere. However, rather than
seeking constancy of some specific indicator such as free reserves, the
System would usually be aiming at avoiding .sharp changes in the
general money market atmosphere. To the extent that the free
reserve level is an indication of that atmosphere—and it is a highly
imperfect one—then the System would be aiming for continuing
approximately the same range of free reserves.
4. Basically, it is the terms of the Treasury financing, set in the
light of current market conditions, that bring about the absorption
of the new securities. The Federal Reserve supplies whatever volume
of reserves is required to achieve the overall objectives of monetary
policy. Hence, in the broad sense, as already stated above, an "even
keel" policy does not contemplate providing a volume of reserves
necessary to accomplish complete digestion in the market of issues
included in Treasury financing. However, in the narrower sense, the
initial increase in required reserves that may occur as issues move
out into the hands of dealers, banks, and nonbank investors, is one
of the many influences on reserves taken into account in determining
the proximate reserve objectives and the day-to-day conduct of open
market operations.
5. Implementation of longrun objectives of monetary policy and
offsetting seasonal fluctuations in the availability of reserves will
meet with difficulties only if "even keel" periods follow each other in
rapid succession. At times, the maintenance of "even keel" has
tended to delay the implementation of changes in System policy, and
on a few rare occasions the System has had to step in more positively,
cast aside immediate reserve objectives, and temporarily give overriding attention to the state of the market. Yet, the "even keel"
constraint has not seriously interfered with the implementation of
general policy objectives as a long-term matter.
To cite, however, specific examples of temporary interferences: In
mid-1958, severely adverse developments in the bond market meant
that System reserve objectives temporarily gave way to correcting
disorderly market conditions and helping to restore confidence m the
functioning of the market. As System policy again moved from ease
110
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
toward restraint against the background of the 1958-59 recover}- and
expansion, there were Additional instances when System action was
delayed because of major Treasuryfinancingoperations—for example,
in early 1959 and again in the spring of that year.
Question
V. When an increase in the degree of ea*e occurs, in your judgment,
does the increased degree of ease—
1. Cause a change in interest- rates?
2. Cause a change in the composition of bank portfolios?
3. Cause a change in the total of bank earning assets?
4. Cause a change in the volume of demand deposits and currency outstanding?
5." Causa a change in commercial banks' time deposits outstanding?
6. Cause a change in the position of dealers in the government
securities market?
Answer supplied by the Board of Gorernors
The effect of an increase in the degree of credit ease depends in
part on the surrounding economic andfinancialcircumstances. Moreover, the effects of the easing itself may either intensify or offset the
effects of other developments simultaneously taking place in the economy.
A shift of policy from restraint toward ease may be expected to take
place when business activity is showing less strength and loan demand
is slackening. Such developments are likely to follow a period in
which banks have drawn down their liquid assets and increased their
loan-deposit- ratios. Under such circumstances, an increase in the
degree of credit ease is likely to be reflected in an increase in bank
credit and deposits, or at least a moderation of the decline that would
otherwise have taken place: an increase in the proportion of assets
held in the form of short-term open market paper and a subsequent
increase in holdings of longer term securities: and a general reduction in
interest rates. The easing of credit conditions at a time of rapidly
increasing loan demand, or a prolonged period of verv great ease, on
the other hand, might lead to excessive credit expansion in unsound
directions.
Tn a period of slackening loan demand following a period of credit
restraint, commercial banks are likely to use reserves becoming
available in roughly the following order: first, to rcnav indebtedness
to the Federal Reserve banks and, to some extent, to"build up cash
assets; second, to replenish depleted portfolios of Treasury bills and
other liquid assets; and third, to increase holdings of longer tenn
L.:>. trovermnefit securities and State and local government securities.
Banks will also show increased willingness to make loans, except in so
far as deterred by rising delinquency rates. Because of the reduced
loan demand, however, most types of loans, with the possible exception of real estate loans, are likely to decline or show little change
pending an improvement in the business outlook.
i tie demand for loans and investments on the part of banks,
coupled with the reduction in supply of some types of assets, is likely
to be reflected in a reduction in interest rates and security yields.
In the absence of sharp changes in supply conditions, the reduction
m rates tends to be earliest and most marked for short-term open
111
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
market paper but spreads also to longer term securities and to many
types of loans.
b The expansion of banks' earning assets tends to be reflected directlv
in an expansion of demand deposits as banks make payments to
sellers of securities or to borrowers. The extent to which recipients
of such deposits, or holders of existing deposits, choose to convert
them into time and savings deposits with commercial banks depends
on a variety of circumstances. The reduction in market rates of
interest, makes consumers and businesses more willing to hold demand
deposits than formerly. Interest and dividend rates on savings deposits and shares, however, have typically responded only slowly,
if at- all, to an easing of credit conditions, and the reduction in market
rates of interest together with reduced expenditures by consumers and
businesses may cause a shift to time and savings deposits at commercial and mutual savings banks and to shares of savings and loan
associations and credit unions. Commercial banks would probably
receive a particularly large share of any temporary increase in time
deposits oil the part of businesses, governments, or wealthy individuals.
Reduced rates of interest accompanying easier credit- conditions
affect both the yields on securities held by Government security
dealers and the cost of funds borrowed to cam- such securities. The
assurance of a ready supply of borrowed funds, however, might
encourage some, expansion of'dealers' trading positions. In addition,
if dealers expect a further reduction in interest- rates and an increase
in security prices, they may temporarily expand their demand for
securities," thus tending to accelerate the' reduction in interest- rates.
/IH^WYT Supplied
by the
Presidents
An increase in the degree of ease may bo reflected in any or all of
the processes enumerated in the question. Basically, an increase in
the degree of ease represents an increased availability of reserves
relative to the economy's demand for credit. Such a change in the
degree of ease will be reflected immediately in money market conditions.
Obviouslv, an increase in the degree of case—and the resulting
effects on interest rates, commercial bank earning: assets, deposits
and other financial variables—can come about either through an
increased availability of reserves to the banking system or through a
decline in the intensity of credit demands.
I- If an increase in ease took place solely as a consequence of an
increased willingness on the part of the System to supply reserves,
the probable effects are as follows. With more reserves available,
banks would tend to add to their
earning
rapid pace
*assets tat a more
— l..i«i/ii>lh«
llftCV
W i l l i i i u * OUIlv OI UU5 l i i c r t ' i u M H *
c-
o^
Tn
be diverted to securities and p o s s i b l v mortgages (subquebt on
in
tho process of acquiring such securities the d e i n a r i d s exerted bj the
banks would tend to put downward pressure on interest r»U. (sub
question 1). A stopper-up acquisition of earning a & e t ^ the banK
in* system implies, of c o » i 4 a more rapid increase ™ ^ s i t liabijiitu»
(subquestions 4 and 5), In genera .both the inonesuppb C"
112
ALTKKNATIVK APPROACH TO TIIF. MONETARY MECHANISM
influenced by other factors. Thus, for example, if the increase 111
tho degree of ease leads to a considerable expansion of economic
activity, then the rise in the money supply may be relatively rapid.
On the other hand, if economic activity responds only sluggishly,
then ihe public's need for more money to finance activity may be
relatively slight.
2. The second possible cause of an increase in the degree of ease
would involve a reduced need for reserves on the part of banks.
This could occur, for example, if the pace of an advance in economic
activity slowed (or if the economy moved into recession) with a
consequent weakening in loan demands. If t he availabilii v of reserve?
remained unchanged, banks would tend to divert at least part of
their funds into securities and mortgages. At tho same time, a
slower advance in economic activity' would probably reduce the
public's demand for money, while the growth of time deposits may
continue or even accelerate. If the System continued to supply
reserves at the same rate (and jriven the lower leseive requirements
against time deposits) the banking system would be in a position to
increase total bank earning assets "more rapidly. As in the first case,
interest rates would be under downward pressure.
An increase in the degree of case will also have an effect on dealer?
in Government securities (subquestion i>). First, with credit conditions somewhat easier, the availability of boirowed funds to such
dealers will increase, in itself this would lend toward an increase in
dealer positions. A more important consideration, howe\er. is the
possible effect of the increase in ease oil market expectations. If the
change in credit, conditions gives rise to expectations of declining
rates, dealers may be induced to add to their positions; however, expectations axe influenced bv many other developments so lhat there
is no certain relationship between an increase in the degree of ease
and an increase in dealer positions.
The causal interconnections outlined above appear most clearly
during business recessions. At such times System policy is more
likely to become positively stimulative rather than to remain the
same. At such times, the effect of business recession in easing credit
conditions is reinforced by the Systems shift to an active antirecessionary policy.
The effects of an incicase in the degree of tightness need not be the
exact opposite of the effects of an increase in ease. For example,
under boom conditions marked by inflationary pressures, the degree
of tightness typically increases, both because'of an increase in commercial bank demand for reserves, reflecting strongly rising loan
demands, and a reduced willingness of the Svstem to supply reserves.
Under such conditions banks would tend to sell off Governments in
order to finance a part of the additional loan demands, and interest
rates would tend to rise. However, the strength of loan demands
might be such as to result in some further growth of earning assets
and of the money supply. In effect, then, credit conditions might
appear more restrictive if attention were focused on interest rates
alone, and less restrictive if stress were placed on the behavior of
bank credit and the money supply. ]t is worth noting, however,
that interest rates probably would "have been lower, and the pace of
the advance in bank credit and the money supply more rapid, in the
absence of the reduced willingness of the"Svstem to supply reserves.
113ALTKKNATIVKAPPROACH TO TIIF. MONETARY MECHANISM
Tho behavior of earning assets and the money supply just described
may, of course, be entirely appropriate to the situation; tho immediate
goal of policy under such conditions may be to slow the grow! h of
bank credit and the money supply below what it would otherwise
have been and thus prevent an excess of expansion. If a rate of advance in bank credit and the money supply WOT, already excessive,
additional tightening could bring about an even slower advance.
QttcxlUm
VI. In the postwar period, changes in member bank reserve requirements have generally boon hugely offset by compensating open
market operations.
1. What is the rationale for coupling the change in reserve
requirements with an open market operation instead of effecting
the desired increase or decrease in member bank reserves through
open market operations alone?
An-iwr Supplied by the Board oj Governors
Changes in reserve requirements have generally been made at
times when the changing need for bank reserves has been more than
temporary or of large magnitude, or when it has been desirable to
provide reserves at once on a countrywide basis. However, open
market operations are the more usual method of supplying and absorbing reserve funds because they arc more flexible, can be applied more
gradually, and are more easily reversed. Thus, reserve requirement
changes are not made very frequently because open market operations
are better adapted to the continuous adjustments required of monetary
policy us it responds to the economy.
Because the impact on reserves of a change in reserve requirements occurs all at ouce rather than over a period of weeks or months,
partially off-citing open market operations are sometimes undertaken
when the new requirement first becomes effective. These operations
are temporary, however, and are designed in part to cushion the
market impact of sudden large changes in bank reserve positions.
Answer Supplied by the President
Changes in reserve requirements are ordinarily made in steps of not
less thaii one-half of 1 percent. But even a reduction as small as this
produces an immediate change in reserve availability of sizable proportions (under current conditions, a reduction of one-half of I percent
in the requirements against demand deposits would bo, about 8550
million) and the resulting change in reserve availability can, therefore,
readily be greater than required by basic economic and credit conditions at that- exact point of time, or greater than the money market
can absorb quicklv without undue disturbance. Thus, it is often
desirable to cushion the initial impact of a reserve requirement reduction by partially (usually temporarily) offsetting it by open market
action*
"
, .„
It is onlv in this limited sense that one can speak of "coupling the
changc in" reserve requirements with an open market operation."
The two operations should not rightly be thought of as substitutes to
each other but- as complementary. Each method of putting reserves
into the market, has special advantages at times. For example, m a
recessionurv period, a reduction in reserve requirements has tue advantage of putting reserves immediately into the hands of all classes of
114
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
banks whose reserves are lowered. The impact of open market operations, while immediate on banks in the financial centers, takes more
time to be transmitted throughout the country. Reserve requirement
changes also seem to have a more immediate psychological impact,
which may be important under certain conditions.
In a broad sense, changes in member bank reserve requirements
during the postwar period as a whole must be viewed as a reduction
from unprecedentedly high levels established during the 1930's to
absorb the excess reserves flowing into the country from abroad and
subsequently to counteract excessive bank liquidity generated by
World War I I financing. The postwar reductions (except for the
Korean period and disregarding the November 1960 increases for
country banks to compensate for the inclusion of vault cash in
required reserves) have served to move toward a level and structure
of reserve requirements that would be more appropriate in the long
run, provide part of the increase in the money supply which the
expansion of the economy has necessitated, and place the banks in a
better competitive position vis-a-vis other similar financial institutions.
The successive postwar reductions were in general undertaken in an
amount and at a time when they could be expected to support the
broad aims of monetary policy. Primary reliance, in most cases, for
the achievement of policy objectives was placed on open market operations and the discount rate.
From the end of World War I I to the time of the Accord, open market operations on balance withdrew reserves (in several steps in 1949)
from the banking system in an attempt to reduce excessive liquidity.
After the outbreak of the Korean war, open market operations provided reserves to support war financing, and the increase in reserve
requirements in January 1951 was designed to offset some of the effects
of war finance on bank liquidity. While total Reserve bank credit
remained about level for 5 years from the beginning of 1951, another
opportunity for a constructive reduction in reserve r e q u i r e m e n t s
occurred as the economy moved into a recession in 1954.
Since the emergence of balance-of-payments considerations as a
major element in the formulation of monetary policy, further reductions in reserve requirements for demand and time deposits were made,
together with the reduction of requirements for the banks previously
classified as central Reserve city banks. The reserve requirement tool
is particularly well suited to a situation in which it seems d e s i r a b l e
both to add to reserve availability and to minimize downward pressure on short-term rates. Indeed, because of the relative thinness of
the market for long-term U.S. Government securities, the provision
of heavy reserve needs solely through open market operations would
have to be accomplished primarily by purchases of Treasury bills and
other short-term investments. This method of supplying reserves
would have the presently undesirable effect of putting downward pressure on short-term rates. Accordingly, in a period when policy is
directed simultaneously toward discouraging short-term capital outflows and stimulating the economy, a reduction in reserve requirements, may offer an important advantage over providing an equal
amount of reserves through open market operations because of their
different effect on short-term rates.
Thus, over the postwar period as a whole, changes in reserve requirements as well as open market policy have been used to broaden signifi
115
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
camly tlio credit base of the economy in a way consistent with tlie
overall objectives of sustained growth and international stability.
Question
YIT. The four most likely causes of higher free reserves are:
1. A desire 011 the pari of business and individuals to decrease
their borrowing at commercial banks.
2. A desire on the part of individuals to hold a larger proporl ion
of their money balances in the form of checking deposits.
3. A desire on the part of banks to hold a larger proportion of
their assets in reserves.
4. An increase in excess reserves brought about bv open market
operations.
(a) Are there significantly different effects on economic
conditions if the change in*free reserves originates in one
source rather than another? Do you attempt to deal
differently with each of them?
(b) Do you attempt to distinguish among these sources
of change m evaluating the change in free reserves? If so,
are there explicit measures you employ to distinguish anion":
these sources of changes in free reserves? If so, which
measures do you use?
Answer Supplied by tlie Board oj Governors
As is suggested in the answer to Question II, it is useful to appraise
the significance of free reserves in terms of the supply of reserves
made available by Federal Reserve actions and the demands for reserves by the banks, which are affected by prospects for bank profits.
Such prospects are. in turn, affected by the first two factors noted in
this question, as well as the cost of alternative methods of reserve
adjustment. In response to the question under (a), it does make a
difference whether a change in five reserves originates, on the one
hand, in a desire by the banks to reduce or increase their borrowings
or. on the other, in action by the Federal Reserve to change monetary
policy.
In'response to the question under (ft), the Federal Reserve does
attempt to distinguish between sources of change. It recognizes,
for example, that in a period of recession when loan demands are weak
and interest rates relatively low, member banks will wish to reduce
sharply their borrowings at Federal Reserve banks, in other words,
banks'will desire a higher level of free reserves. The movement
toward larger actual free reserves in such a period ordinarily reflects
two factors: (1) adaptation by the Federal Reserve to this change in
bank preferences and (2) adjustment in Federal Reserve policy to a
more stimulative posture. Although factors influencing the banks'
demand for free reserves cannot be measured with precision, indications of this demand are provided, for example, by changes in bank
loans, especially business loans, and by the level of interest rates on
short-term securities and the Federal funds rate and their relationship
to the discount rate.
Answer supplied by the presidents
Different monetary factors which affect, among other things, free
reserves mav of course have different implications for general monetary
and economic conditions. The factors listed are not always the "most
116
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
likely causes of higher free reserves." Free reserves are affected by a
variety of factors including those stated in the question. The relevance of these influences to monetary policy lies in what they indicate
with respect to economic and credit conditions. There may be significantly different effects on economic conditions if the changes originate in certain sources rather than others. But, more importantly,
the changes originating in certain sources mirror effects that are more
significant to monetary policy formulation than those originating in
other sources. Thus, in this sense, we do attempt to distinguish
among these sources of change.
The factors of change listed (and others) are reflected first in total
reserves and excess reserves and then in borrowings and free reserves. Depending upon the current posture of policy, actions may
be taken to offset or not to offset changes in reserves stemming from
the so-called money market factors such as Treasury operations, gold
and foreign account operations, float, currency, etc. Thus, with specific reference to factor 2 (a desire on the part of individuals to hold a
larger proportion of their money balances in the form of checking deposits) we would read normally no deep economic significance into
this. Like movements of other market factors affecting reserves, the
impact of these flows on free reserves is at times substantial. These
movements, however, are generally a reflection of purely seasonal and
random factors, rather than of any significant change in the e c o n o m i c
atmosphere. For this reason, we normally seek to offset changes in
the banks' reserve position stemming from such sources—subject to
the needs of policy goals—through open market operations. ;
Factor 3, a desire on the part of banks to hold a larger proportion
of their assets in reserves (an increase in the demand for free reserves)
might reflect a number of influences. In a broad sense, it would be
interpreted as a reduction in the. willingness of banks to expand credit
and deposits on the same basis, relative to reserves, as had prevailed.
Again, given a particular policy goal this might be allowed to hold or
it might be offset by injecting more reserves so that the new demand
schedule for free reserves might be met without contraction of bank
credit. Incidentally, this is one reason why a given dollar a m o u n t
of free reserves cannot be assumed to have an identical meaning over
time in terms of the pressure on banks to expand or contract credit.
The immediate impact of purchases of securities through open
market operations (factor 4) is, of course, also to raise excess r e s e r v e s ,
unless such purchases are undertaken to offset market factors. A
reserve injection due to open market purchases will result in a more
"permanent" increase in the level of free reserves if the banks wish to
use the new reserves to repay borrowings from the Federal Reserve
banks or to maintain excess reserves at a higher level. It is likely,
however, that at least a part of the reserve gain stemming from open
market purchases (or from any other source of nonborrowed reseives)
will be used to advance credit, thereby increasing required reserves
and tending to use up free reserves.
In any case, the injection of reserves through open market o p e r a t i o n s
is a step in the direction of credit and deposit expansion, although
the quantitative magnitudes do depend upon the banks' d e m a n d for
excess reserves. While it is difficult to estimate this demand directly,
as indicated below, the pattern of the response of credit and deposit
expansion is usually discernible—except over very short periods—and
117
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
if further adjustments in reserve availability appear to be appropriate,
they can subsequently be made. Of course the appropriateness of
any monetary variable or group of variables must, in turn, be determined in the context of the larger aims of policy.
The immediate impact of a decline in individual and business loan
demand (factor 1) is a fall in required reserves and a rise in excess
and free reserves. The longer term impact on free reserves is less
certain. The banks are likely to respond by attempting to offset the
loan decline t hrough purchases of securities from the public—including
newly issued securities—as indicated in our answer to question V.
Required reserves will thus tend to rise again and there would be a
tendency for free reserves to revert to their original level. On the
other hand, this tendency could be partly offset, should open market
interest rates decline and if banks were, stimulated merely to hold
higher levels of excess reserves. Whatever the ultimate impact on
free reserves, a decline in loan demand of any substantial size is most
likely to occur during periods of weakness in economic activity when
policy is actively seeking to counter recessionary forces through
generous injections of reserves.
I t is, to be sure, very difficult, if not impossible, to isolate with
any degree of confidence the influence of changes in the banking
system's demand for free reserves from the other factors that may be
operating on actual changes in free reserves at any moment. This is
particularly true in view of the occasionally erratic movements of
certain supply factors such as float. At times it may be possible to
get some understanding of the reasons for shifts in the demand for
free reserves, however,"by examining the tone of the money market
(as evidenced, for example, by the rate on Federal funds) in relation
to the estimated level of free reserves in the banking system. Abnormalities in the relationship between money market tone and the
nationwide level of free reserves are more likely to be indicative of
temporary shifts in the distribution of reserves than of any fundamental shift in bank demands.
Question
VIII. Identify some specific occasions during your membership on
the Board of Governors when the Federal Reserve and Treasury
positions on a particular issue were not the same. State the issue and
the nature of the conflicting views.
1. In which cases did the Treasury view prevail?
2. In which cases did the Federal Reserve view prevail?
Answer Supplied by the Board oj Governors
The nature of the arrangements for cooperation and coordination
between the Treasury and the Federal Reserve in monetary and debt
management policies and operations does not generally produce
conflicts of the sort implied by this question. Typically prospective
actions are discussed freely and frankly and there is an interchange
of judgments as to the desirable and undesirable effects that might
be expected to flow from them. In the course of these discussions,
differences in emphasis and judgment will emerge a m o n g the individuals concerned, but after further discussion a consensus is reached
that is understood and acceptable to both agencies.
These interagency discussions form part of the background against
which decisions are made by the appropriate authority—that is, the
34-474—64
0
118
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
Treasury on questions of debt management and the Federal Reserve
on questions of monetary policy. Since the discussions are basically
an exchange of views and information, there is usually no sense in
which the agency not responsible for the action can be said to have had
a "position" that had "prevailed" or been overruled.
The only specific instance which might fall within the scope of
this question occurred in the spring of 1956, when the then Secretary
of the Treasury felt that a discount rate increase proposed by several
Federal Reserve banks was premature. This increase was approved
by the Board, and Secretary Humphrey subsequently indicated publicly his feelings in the matter. This "incident" was subsequently
explored in some detail at hearings before the Subcommittee on
Economic Stabilization of the Joint Committee on the Economic
Report, held June 12, 1956.
(The presidents were not asked to respond to this question.)
A P P E N D I X
II
S C A L I N G OF T H E F E D E R A L R E S E R V E ' S P O L I C Y
DECISIONS
The scaling represents the authors' consensus about Federal Reserve^
decisions to change policy. The procedure is described more fully
in the text of chapter IV. A positive value indicates a move toward
"greater ease" or "less restraint"; a negative value indicates a move
toward "less ease" or "greater restraint." Large values, whether
positive or negative, are used to indicate more decisive changes in
desired policy. Note, however, that the values in the table relate
to a particular decision made for the period between two meetings.
Scaling is not cumulative.
Scaling of the Federal Reserve's policy decisions
1. 1946—Mar. 1
* • The Committee should continue the existing open
market policy * * V
Before this date "action had been taken * * * to discontinue
the preferential discount rate. * * * "
Oct.
0
* * there had been no new development which presented
reasons for a change in the policies adopted * * *."
Comment: No emphasis on restraint to be found. Policy directives refer to stabilization of
security prices only.
Z 1947—Mar.
0
"* * * the Committee continued to be of the opinion that
monetary * * * policy should be directed * * * toward restraining the further expansion of bank credit.* * * "
June 5 - .
0
The "Committee therefore decided that * * * no change
should be made in existing open market policies.* * * "
tj
Aug. 8 . .
—>4
"* * * to provide for supporting the present Issuing rate on
Treasury certificates instead of the H rate previously prevailing."
Oct. 6, 7 . . .
—H
*'* * * directed the Federal Reserve banks to terminate the
policy of buying and to terminate the repurchase option privilege
on Treasury bills."
" I t was tbe view of the Committee that * * * the situation was
such as to justify the Tresaury and the Federal Reserve System
taking such actions as were available to them to eliminate * * *
excessive credit expansion."
Dec. 9
.
0
" • * * It was felt that the existing open market policy should
be continued.* * * "
Comment: The Committee mentions in March Its opinion to restrain credit expansion.
In October a detailed,anti-inflationary program is formulated.
Dec. 1947<j
—H
«•*'*-* the policy of combating inflation has been further
implemented by additional retirement of maturing government
debt, a downward revision * * * of the prices at which the
system would purchase Treasury bonds * * an increase in the
issuing rate of 1-year Treasury certificates * * * "
3. 1948—January
—M
Increase in discount rate.
February
Increase in reserve requirements at CRCB.
Mar.
— " T h e existence of a large Treasury surplus during the first
quarter of the year would be the principal weapon available
in the credit field for combating Inflation."
May 20
*
0
"* * * its existing anti-inflationary policy of beeping pressure
on the money market for the purpose of restraining the expansion
of bank reserves should be continued."
July-August
—1
Raise in reserve requirements effective Sept. 1.
Aug. 9
-M
Raise of discount rate and increase of short-rate.
Oct. 4
—W
"* * * the Treasury continued to manage its balances * • •
to further the policy of keeping pressure on bank reserves. * * •
The effect on bank reserves of securities purchased * * * had
been offset by the sale of Treasury bills. * *
Nov. 30
0
"* * * to continue to cope with conditions as they develop."
Comment: .The report states that "monetary expansion was more effectively restrained
in 1948 than in any other year since before the war."
119
June 1 0 . . ,
0
—
120
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
Scaling of the Federal Reserve's policy
decisions—Continued
4. 1949—Mar. 1
"It would replace the existing policy of exercising restraint on
credit expansion with a policy which would rd"X such restraint
without following an aggressive easy money poliry."
April—
Reduction in reserve requirements "offset * * * because * * *
policy of maintaining * * * short-term rates required sales of
securities "
May 3
"* * * that the policies of the system should be directed toward somewhat easier credit conditions.* * *"
June 28
+1
Reductions in reserve requirements were made. Sales of
bonds were discontinued. Only partial offsetting of reduction
in requirement ratio by sales of bills.
Aug. 5
0
"* * * to reduce the system's holdings of securities to offset the relepsod reserves * * * so that the reduction would not
result in a general further lowering of short-term rat^s."
N o v . 22 to Dee. 1.
— }4
* * in order to indicate a change from a policy of monetary easo to a policy of mild restriction.* * *" Bill rates were
allowed to rise and l}4 established for I-year paper.
6. 1950—Mar. 1
«* * * mildly restricting the availability of bank reserves."
"* * * some rise in those (long-term rates) * * * sales of bonds
_
„
b y the System. * *
Early June
* * Treasury refunding required substantial Federal
Reserve purchases."
June 13,14
»• * * the System's policies should be directed towards restricting increases in bank reserves."
18
-IK
"It was felt b y the Committee that immediate action to restrain credit expansion should be taken as an essential part of the
broad anti-inflationary program of the Government."
"* * * the prevention of inflation was a matter of critical Im*
portance and urgency. * * *"
"Treasury borrowing * * * should be as much as possible
from savings and as little as possible from banks."
n
.
* * use all the means at their command to restrain *
.
Sept. 28
0
"* * * the System should endeavour to hold down purchases
of securities to the minimum consistent with maintenance of an
_ . „
orderly market.
11
g
* * proceed with policies decided upon. * * *"
uct
0
* * continued flexibility in the short-term money market
was
essential
* * V'
Nov 27
»
0
"Continuation of the policies of credit restraint through open
,, - , market operations * * * was decided upon. * * *"
^ ,„
mu
c o m m e n t : The pollcyjjf tight restraint was not put into effect. Purchases exceeded sales.
See annual report.
6. 1951—January-February.
Increase in reserve requirements. The increase ahsorbs some
additional reserves. Incomplete compensation by open market
,
„t
purchases.
Jan
* 61
0
* * continuing in effect * * * the existing policy of
a Q
„0
restraint on further expansion of bank credit. * *
f
* * no change should be made in the-existing general
direction of the Committee of restraint of further expansion or
»,
- _
bank credit. * *
JMUT, l, J
« T h e change i n p o i i c y (j C f accord with Treasury) did not
require any change in direction for the reason that the direction
issued at the meeting of Aug. 18,1950, was changed in the light of
the policy adopted at that time. * *
"First * * * bonds could be taken off the market by Treasury
offer * * * nonmarketable * * * to minimize the monotization
of public debt * * * a limited volume of open market purchases
would be made * * * and open market purchases, if any, would
be made on a scaledown of prices."
"Third * * * immediately reduce or discontinue, purchases
of sh ort-term securities * * • * banks would depend upon borrow
o
_
tag
at the Federal Reserve. * * *»
M a r
0
*
* * * continue to carry out the policy approved at the
Anril
lVfft^
,
meeting on Mar. 1 and 2. * *
p
1 y
~
Reserve positions of commercial banks were under greater
M a yy 7
n
in 1951 than in other postwar years."
...
0
—
"Reduced Federal Reserve buving of Government securities
after April was an important factor limiting bank reserves
expansion. As bank reserves became less readily available than
0Ct
they hada been
previously * * V '
N o v 14"
n
« ! I I m ° r e restrictive policy seemed unnecessary."
.
• * * n o c h ^ g g i n e s i s t i n g objectives of credit policy was
u
neede<
7* 1952—Mar 1
.
June 19
n
o
n
Sept. 25
j.iy-
Dec. 8_»
1953-Jaiwarr
0
n
u
0
I"* * * to continue the policy which had been pursued for
several months."
** * * the general policy of limiting the availability of bank
reserves
that had been pursued b y the System since October of
£ l was still appropriate."
* * additional reserves should be a p p l i e d through open
Purchases in order to avoid undue restraint."
# #
tf * did not call for action to change the existing p o ^ y
19
grease in discount rate.
.
Committee agreed * * * that it would pursue a policy
wmch woind maintain abo"t the same degree of restraint on
month eX*a*1S1°,n
t h a t h a d 1)6611 f o H o w e d m r e c e u t
P recedulg
121ALTKKNATIVKAPPROACH TO TIIF. MONETARY MECHANISM
Scaling of the Federal Reserve's policy
decisions—Continued
8. 1953—Continued
J u n e 11
+1
q ti+ oj
i>epr.
•Dec-
15
+H
_ .
9. 1954 February
Mar 3
0
0
June 23
_ , A
hept.
22
iJec
- *
+H
_
iu. law—Jan. 11
1n
Mar
*2
0
M a y 10_„
_
June 22
o
July 12_
o
Aug. 2
—1
Aug. 23
—M
Sept. 14—
—yi
Sept. 26—
Oct. 4
Oct. 25
—
„
N o v . 16
-f J£
N o v . 30.
^
Dec. 13
„,
11. 1956—Jan. 16
Jan. 24
0
-f
Feb. 1 5 - . , .
0
Mar. 6„—
0
Mar. 27
-H
* * the change in policy at this meeting reflected recent
developments * * V '
"It was the view of the Committee * * * that policy should
be one
o f aggressively supplying reserves to the market. * *
«* * * that further easing would be needed to assure ready
availability of credit. * * *"
* * in carrying out operations for the System account
there would be more emphasis on a program of actively maintaining a condition of ease in the money market."
Reductions in discount rate.
"* * * the policy of actively providing reserves to the money
market to facilitate credit expansion should be continued during
the spring of 1954."
Reductions in reserve requirement and only partial compensation by open market sales.
* * resolve doubt on the side of ease. • * *"
"A reexamination of the policy of 'active ease' * * * led the
Committee to the conclusion that the developing economic
situation did not warrant continuing as active a program of
supplying reserves to the market as had been followed during
the preceding year. * * *"
"* * * easy credit w n s n o longer needed. * •
a farther step away from the poJicv of 'active ease' ** V
«* * * ^ aereed that, although Increased ease should be
avoided,
• + +>> further measures towaid restraint should be deferred.
«* * * a further shift In emphasis toward a policy that would
discourage undue crcdit oxpansion."
* * the Committee concluded that for the immediate
future it should not alter the course it had been following recently which had had a restraining influence on credit expan*
sions. * * *"
*'* * * the Committee aereed that it should maintain substantially the degree, of restraint that had existed. * * *"
* * shift to a policy of restraining inflationary development * * V
"* * * It agreed that the wording of its directive should be
changed * * * to show that increased monetary restraint on
credit expansion was now clearly appropriate."
* * the Committee's policy should be one of gradually
increasing pressure. * * *"
"it was the judgment of the Committeo that this situation
called at least for the maintenance of, and preferably for some
slight increase, in the restraining pressure it had been exerting
"* * * it should aim at maintaining about the same degree of
credit pressure that had existed, with the understanding however, that doubts need not be resolved on the side of greater
restraint."
* * restraint on credit expansion, with the understanding
that doubts should be resolved on the side of increased restraint."
"Continuation of the policy * * * seemed to be called for with
the understanding that doubts should be resolved on the side of
dispelling anv idea of an easing."
* * while it was trying to move in the direction of maintaining tightness, it would not be concerned if operations in the
open market during the immediate future die not achieve as
great a degree of tightness as had existed recently." This statement is balanced partly by an increase in the discount rate.
N o change in policy, but emergency action to support Treasury refinancing: "The volume of cash redemption would bo
considerably larger than had been anticipated." Purchases up
to S100,000,000.
..
s
"With the passage of that difficult period, it seemed desirable
to attempt to regain as far as possible the level of pressure that
had existed * * * just prior to the announcement of the Treasury's refunding."
.
.
* * the general policy of r e p a i n t followed in the recent
months should be reaffirmed. • •
.
"* * * that a continuation of restraint on credit expansion
was required."
, t
. * * the Committee did not feel that the general level of
restraint should be increased beyond that which existed in tne
autumn of 1955."
* * a shift in emphasis seemed desirable."
* * some relaxation of restraint appropriate in the near
"* *' * a relaxation of pressures * * * was not indicated,
although no increase in restraint appeared to be called for
.
* * to continue the existing policy without overt action
toward increasing or lessening the degree of restraint
.
"The supplementary clause which was i n t r o d u c e d In Jan. g,
was eliminated * * *. The Committee concluded that its instructions to take into account deflationary tendencies
was not consistent with the existing situation * *
122
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
Scaling of the Federal Reserve's policy
11. 1956—Continued
A p r . 17
+M
May 9
M a y 23
June 5
0
+H
-
0
June 28
+H
July 17--
0
Aug. 7
Aug. 21
Sept. 11—
—H
Sept. 25
+l£
Oct. 16
o
N o v . 13
0
Nov. 27
0 to 4-H
io mr-T t c , r t 1 0
IZ. 1057—Jan. 8
Jan
23
*
^H
1C
J>A3r
rt
c
*0
*
010
,
+W
s.,
26
.
A
Pr-16
x
M a
0
y
7
y
23
0
J u n e 18
0
Aia
0
„
T .
jujy w
T
A
0
i on
OA
* "u
, ,9
+H
,
+H
AUg
bept
n
in
*10
,
A
0
f w «o
"" ^
N
n
v
-
.
"H*
+H
,
6
+M
n
17
13. 1958—Jan 7
an. 7
J a n <*
J
0
decisions—Continued
^
,
,
„
* * no relaxation of pressures. However, t h e restrictive
policies should not be pressed too strongly."
"* * * to make no change in the existing policy."
The qualifying phrase deleted from the instructions in Mar. 27
was reintroduced "* * * the Committee agreed that during the
immediate future additional reserves should be supplied to take
care * * * of growth needs."
"The Committee did not wish policy to become more restrictive."
"The committee agreed that, within the framework of the
restrictive policy it had been following, doubts should be resolved
on the side of case during the next few weeks."
"* * * no significant change in credit policy should be
made. * * * "
The qualifying phrase re-introduced in M a y 23 was deleted.
Instructions required attention be given only to inflationary developments.
"* * * should be somewhat more restrictive but Federal
Reserve banks * * * considering discount rate increases. * *
"* * * to maintain substantially the existing degree of stability in the market, with doubts resolved on the side of tightness
rather than ease. * * * "
"* * * to maintain pressures of about the same degree that
had existed recently, but that in the case of doubt operations
should be resolved on the side of ease. * * * "
* * to maintain substantially the present degree of restraint."
"* * * the degree of pressure in the money market should be
substantially unchanged. * * * "
In spite of the modification in the directive, "the committee
did not intend an overt change away from a policy of restraint; * * V*
"A qualifying phrase was added (to the directive) requiring
'recognition' of additional pressures. * * * "
***** the existing policy should not be changed."
»* * * the directive issued at its first meeting of 1957 maintained the policy of restraint upon credit expansion that had been
in effect for approximately 2 years, but it represented an
adjustment from the program followed in the last few weeks of
19.ri6."
* * restoring approximately the degree of restraint of the
late November-December period * * *."
"The current relative easo was unintended."
* * continuation 0 f tho status quo. * * *"
"The committee sought to continue about the same pressure
on credit expansion that had been intended by the action taken
at the la$t several meetings * * *" The directive was changed
to acknowledge emerging uncertainties.
«•* * * doubts should he resolved on the side of greater rather
than less restraint that had existed in recent months."
* * a stable situation should be maintained for the next
few weeks."
* * current events made continuation of substantially the
existing degree of restraint appropriate. * * *"
* * the Committee sought to have the same situation continue for future weeks. * * *"
***** t tje Committee's conclusion was that a firm policy of
restraint should be continued for the present."
"* * * to maintain but not to increase the existing degree of
pressure."
11V * t 0 k e e P the banking system under substantial pressure."
The directive was renewed and "* * * the System account
would have flexibility in providing reserves. * *
,
^
* * that in carrying out the policy of restraint * * * doubts
would be resolved on the sido of less rather than greater restraint."
" * * * to continue the same degree of pressure that had been
soufht d u r i n g t h e previous 3 weeks."
* * * although general policy was not to be changed appreciably, it should tend on the easier side from where it had been in
recent weeks."
"* t* * there should be a moderate relaxation of the degree of
restrictive pressure."
* * there should be further moderation of restrictive pressures * * *."
,
directive was modified "to contribute further by monetary ease to resumption of stable growth * * *."
V A more positive approach to recovery. * *
' ! * * a ^ g h t easing in the reserve position of banks would
b e desu
*at>le. * * *"
* * there should be no change * * * in the policy of supplying reserve funds. * * *"
* * maintaining approximately the same condition in the
money market that had existed immediately prior to this
meeting."
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
Scaling of the Federal Reserve's policy
123
decisions—Continued
13. 1958—Continued
11
-M -a r
+H
4
-
-M
__
iVAar
^>
.
—
,,
Ma 07""
iviay / /
A p r i0
+H
,9
—^
T
J u n e 17
.
Y
0
0
TniX to"
-tY%
,^
-H
-M
r i ' oj
i"\y
be )t
I *
o
from
-
—K
9
0
0
n ? oi
uct. 2 1 —
..
INioy 10
J->ec. 2
^
JJec 16
*
„
T
« . 19J9—Jan. 8
0
0
—M
—H
-H
_
Jan
- 27
0
Feb. 10
o
v
Ma
*". 3
—H
Mar. 24
0
Apr. 14
0
May 5
0
M a y 26
-U
J u n e 16
July 7
0
July 28
0
Aug. 18
0
Sept. 1
0
Sept. 22.
-Hi
Oct. 13
Nov. 4
0
-fH
N o v . 24
0
Dec. 15
0
* t h e r e s h o u I d be n o cIlan^e i n Policy."
,A
J U i y 10
.
Aug. 4
.
19
* * continue to follow an 'even keel' policy stopped on the
side of ease. * * *"
The directive was modified "to contribute further by monetary ease atom resumption
of stable growth. * *
ore
!!! ! *
Positive approach to recovery. * * *"
* * operations in the System account should be directed
toward maintaining a slightly larger volume of free reserves
and money market conditions slightly easier. * *
"Easing" was "contemplated" in form of lower discount
rates and reserve requirements.
* * the prevailing policy of ease should be continued. * * * "
"* * * maintain the current posture of monetary ease without
further depressing Treasury bill rates. * * *"
"* * * no change in Federal Reserve credit policv * • * no
action should be taken to cause the tone of the market to get
materially easier or tighter."
"Disorderly conditions" developed on the securities market
the System account was authorized to engage in purchases,
Termination of authorization passed on July 18.
The Committee felt that the reserves generated by recent
emergency purchases of securities were now redundant and
should be absorbed.
"* * * the policy to be followed * * * should be one of keeping
having redundant reserves."
* * that the rate of expansion in the money supply * * •
should be tempered and that operations for the System Open
Market Account should move in the direction of lower free
reserves.*
* *"
* * maintaining
substantially the same tone In the money
market.* * *"
* * t 0 raaintain a n < e v e n
fa
the market.* * *"
"* * * it would be undesirable to aim toward a greater degree
of restraint on reserve availability.* * *"
"* * * to maintain conditions in the market about as they
were at present."
"* * * the Committee's conclusion contemplated letting
market developments tend to increase restraint.* * *"
"* * * it was believed that a move toward somewhat greater
restraint on the availability of reserves would be appropriate."
"* * * that the degree of restraint on credit expansion * * *
should be about the same as in the immediate past, but that any
deviation should be on the side of restraint. * * *"
*'* * * the current degree of restraint on bank reserves * * *
should be continued. * * *"
• • to maintain the same degree of pressure on bank reserves
position that had been exerted recently. * * *"
•** * * about the same level of restraint should be maintained
on bank reserves. * * *"
"* * * that any doubt * * * should be resolved on the side of
restraint."
* * maintenance of about the degree of restraint that had
prevailed. * * *"
"* * * maintenance of about the same degree of restraint as
had existed. * * *"
* * it would be desirable to move toward greater restraint
on credit expansion ,f* * * after the current Treasury financing
was complete. * * *
"* * * the first paragraph of the Committee's policy directive
was revised * * * to provide for increased restraint on credit expansion."
, ,, . . . „
* * that an intensification of restraint was required.
,
"
*
*
*
continuance
of
the
present
*
•
*
policy
of
restraint.
* • *"
* * doubts should be resolved on the side of the ease during the period of Treasury financing."
"* * * aiming as far as practicable at the same degree of restraint on credit expansion as currently prevailed."
"The consensus * * * favored continuing the present degree
of restrictiveness. * * *"
« • • • maintenance of the existing degree of pressure on
reserve positions of banks. * * *"
, ,
._
"* * * any deviations preferably should be on the side or
less restraint. * * * *"
^
"* * * to continue the existing directive. * *
, .. „
* * operations * * * should aim at maintaining a feeling
of stability in monetary and credit conditions and assuring w e
availability of funds for seasonal credit needs,"
"* * * the current open market position * * * should
continued. * * *"
,
„_
n * r*.
"* * * a consensus favoring maintenance or the degree m restraint on credit expansion. * * *"
and
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
Scaling of the Federal Reserve's policy
15. I860—Jan. 12
0
Jan. 26
0
Feb. 9
+H
Mar. 1
+X
Mar. 22
0
Apr. 12
May 3
M a y 24
+H
Juno 14.
+H
July G
0
July 20
0
Aug. 16
+%
Sept. 13.
+M
Oct. 4
Oct. 25
+H
0
N o v . 22
0
D c c . 13
0
16. 196 W a n . 10
0
Jan. 24
Feb. 7
0
Mar. 7
0
Mar. 28
0
Apr. 18
0
May 9
0
June 6 .
June 20
0
July 11
0
Aug. 1
0
A u g . 22
Sept. 12
0
Oct. 3
0
Oct. 24
—J6
N o v . 14
0
Dec. 5
0
Dec. 19
17. 1962—Jan. 9
Jan. 23
Feb. 13
Mar.
0
0
0
decisions- -Continued
"• * * tbe consensus favored n o change In credit and monetary policy. * * *»
* * continue substantially the same degree of restraint. * *
*•* * * any tightening in the degree of restraint should b&
avoided. * *
"The Committee concluded that it would be appropriate to
supply reserves to the banking system somewhat more readily."
"* * * a policy of moderately less restraint.. * *
"* * * maintaining about the existing situation. * *
"* * * the consensus favnred easing further the reserve position of member banks. * *
"* * * moving moderately in the direction of increasing the
supply of reserves available to the hanking system."
"The consensus * * * favored a further supplying of reserves. * *
"* * * any deviation should be on the side of ease. * *
"* * * to continue to provide reserves at approximately the
present rate. • * **'
1
' continue
to make reserves '
readily available.
"For several months Committee policy had been directed
toward providing reserves needed for moderate bank credit
expansion, and the consensus of the meeting was that thi*
objective should be emphasized."
«* * * doubts sho"ld be resolved on the side of ease. • * *"
"* * * d o r b t s resolved on the side of ease."
* * a conflict should arise between providing additional
reserves and further decline in the bill rate * * * the first of
these considerations should take precedence. * * *"
"The directive * * * was renewed. * * *"
"* * * a continuation of the current degree of ease would be
the preferable objective. * * *"
"* * * to maintain approximately the same amount of ease
in the market. * * *•»
* * there should be n o change in the existing degree of
monetary ease. * *
Close attention urged to the bill rate
because of balance of payments.
"The consensus of the Committee favored n o change in open
market policy. * * +"
"The consensus of tho Committee was that the existing
monetary policy of ease should be followed. * * *"
' * * seeking to maintain about the existing degree of
ease. * * *»»
"* * * revision of the directive carried with It no intent to
modify open market policy in any significant extent at this
stage.
"* * * aimed at maintaining approximately the same degree
of ease that had prevailed. * *
•** * * maintaining the same degree of ease that had prevailed
in recent weeks. * * *"
wollld
V
desirable to maintain approximately the
same degree of ease as had prevailed recently, resolving any
doubts on the side of ease. * *
'
* maintaining substantially the same degree of reserve
availability as had prevailed recently. * *
"
to maintain approximately the same degree of ease that
had recently prevailed. * *
" T h e consensus favored continuation of approximately the
same degree of ease that had been maintained recently."
A
U e c o n s e T l s u s favored continuing the period in early
August when a confluence of market factors contrived to produce
more firmness than had otherwise been the case."
vailed * c * 0 I 2 h m u a t i 0 n o f t h e s a m e degree of ease that had preease * * continuation of approximately the same degree of
1
to
resolve
any
doubts •
' on
the
side of less
* * to produce approximately the same degree of ease that
had prevailed. * * *"
* • maintaining * * * approximately the same policy In
respect to the supplying of reserves that tbe Committee had
been pursuing. * * *"
for " * * n 0 t s u t ) s t a n t i a i change from recent policies was called
* * a somewhat slower rate of increase i n total reserves
than during recent months. * * *'*
Maintain e v e n keel. N o change in basic policy,
o n balance no change. Emphasis in steady money market. ^
Domestic developments did not • * * require shift toward
ease. Hold the posture. * * *
T h e majority favored no change. (But) "Promote further
expansion of bank credit. * * *"
125
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
Scaling of the Federal Reserve's policy
17. 1962—Continued
M a r . 27
•Apr.
May
May
June
17
8
29
19...
0
0
0
—H
June 21
July 10
July 31
0
—
0
Aug. 21
September
0
o
Oct. 2
Oct. 23
0
0
Nov. 13
0
Dec. 4
Dec. IS
0
—\i
decisions—Continued
Slightly more expansion In reserve availability than had
developed.
N o change Indicated.
Current posture appropriate: reissued directive.
N o change in policy despite stock market decline.
"Slightly less easy policy indicated."
"Avoid redundant reserves."
Foreign currency purchases only.
Consensus for continuing degree of ease of June 19. No further
reduction in degree of case.
A majority favored greater ease. However • * * concluded
even keel.
N o change from previous directive.
Evidence of adequate liquidity; policy should remain unchanged.
Supply reserves for seasonal purposes; continue current policy.
N o change in policy seemed called for re Cuba. Reduced
requirements on time deposits. Maintain status quo in mone*
tary policy.
Free reserves at level of $400,000,000. A majority felt no
change called for.
Majority valid; no change.
Somewhat less easy policy favored by majority to firm Treasury bill rates.
APPENDIX III
T H E R E L A T I O N OF C H A N G E S I N M O N E Y TO C H A N G E S I N B A N K
PORTFOLIOS ( " C R E D I T " )
At several points in the text of the study, we have stated that
changes in money and "bank credit" have not been the same for the
monetary system as a whole. This appendix provides a more technical treatment of the problem than the one presented in the text and
indicates the sources of the discrepancy between the two rates.
We start from the balance sheet of the commercial banks
R+E= Dp+ T+ D<
+D'+A+N
where R denotes the banks' cash assets in the form of base money
(i.e., currency and Federal Reserve deposits), E designates total
earning assets, Dp indicates
demand deposits adjusted, T refers to
total time deposits, Dl denotes
the Treasury's tax and loan accounts
at commercial banks, and Df measures net foreign deposits at U.S.
commercial banks. A designates the indebtedness incurred to Federal
Reserve banks and N refers to the banks' net worth. The statement
describes the consolidated position of commercial banks; all interbank
items have been suitably canceled.
Total earning assets E are subdivided into two components El and
2
E such that
E=El+E2
and defined by the relations
R+E1=D*+T+A
and E2=Dt + Df+N
The relative change in total earning assets is thus the weighted sum
of the components El and E2:
ae_ae1
e1 ae2 e*
1
E
E * E
E2 ' E
The relative change of E1 can be usefully transformed into an
expression consisting of the currency ratio k} the time d e p o s i t ratio t,
the average requirement ratio r, and the free reserve ratio/. It is
easily demonstrated that
~(r+f)(l+t)+k
a
where B is the base adjusted for "discounts and advances." The
relative change in El can thus be approximated by a linear combination of relative changes in the adjusted base B", the average require-
127
128
ALTERNATIVE APPROACH TO THE MONETARY MECHANISM
ment ratio r, the free reserve ratio/, the currency ratio k, and the timedeposit ratio t, i.e.—
-AL+«<„,
r + f )
*>
«) f
The coefficients e of the linear combination designate the elasticities
of the asset multiplier e with respect to r+f, k} and t. The asset
multiplier e is defined by
_ (1—r—y)(l+0
(r+y) (1+0+fc
The elasticities are thus exhibited as rational functions of the parameters indicated.
A similar approximation can be derived for the money supply
M^=CP+DP where Cp denotes the volume of currency held by the
public:
AM
AB a . ,
, ,,
Ar
,
,
, «
Af . ,
Air ,
.
A
M
The coefficients € are now the elasticities of the monetary multiplier
__
1+k
(r+/)(l+t)+*
with respect to the behavior parameters indicated as second arguments
of the expressions. These elasticities are also rational functions of the
parameters. Detailed investigations covering the postwar period indicate that the variability of the elasticities over h a l f - c y c l e s , expressed
by the coefficient of variation, is quite small when compared to the
corresponding variability of the relative changes of Bt k} r, 1, and t
The coefficient of variation of e(m. A), for example, is always less than
m
0.03.
The linear approximation to E 1 is used to replace the relative
change in El in the formula for E} and then we subtract the relative
change in the money supply. We obtain in this manner the difference
between the relative change in the banks' total earning assets and the
money supply:
^
AJDf D* AN N^
U(*> r+f)-e(m,
+Y Me,t)~e(m,
r+f)}+~~
Ar
r
,
,^
[«(«, k)~t(m,
,
„,
k)]
0]
The differences between the elasticities occurring in the four bracketed
expressions can be converted into rational functions of the underlying
behavior parameters. These functions were evaluated at the values
129
ALTKKNATIVK
APPROACH TO TIIF. MONETARY MECHANISM
prevailing in the recent past. The difference between the relative
change in E and M can thus be written
AE
E
AMi_AD' AP' AP' Dr , AN N
M
Dl E + D'' E+ N' E
r+jf
'
r+y
This formula supplies the frame for an explanation for the differential
patterns observed for earning asset and money supply.
The formula repeals that "credit-expansion" and "monetary expansion" are not "two sides of a coin." "Credit" and money supply
would exhibit identical relative changes if and only if the relative
changes in r, /, k, t, Z?', Dff and N were restricted in such manner
that any one of these relative changes were perfectly explainable in
terms of the remaining relative changes with the aid of predetermined
coefficients. But our observations indicate that this special restriction
does not hold. Therefore, earning assets and money supply respond,
in general, with substantially different patterns to changes in the
'public's allocation of "payment-money" or total deposits,
variations
in reserve requirements and changes in the banks7 desired reserve
position.
APPENDIX IV
Charts
o
131
Three W e e k M o v i n g Average of Free Reserves for All Member Banks
V a l u e s are Placed in the last w e e k of the three weeV period
•*••••«•••SSI
••••••••Miia
YXOO
\oou
I;II ions
Annual Changes in the Extended Monetary Base, Monthly June 1945 - December 1962
4 0
The Values on the Graph are chances in the extended monetary base between
corresponding months in adjacent years. The value for June, 1945 is the
change in the extended base between June, 1944 and June 1945, etc.
5-0
t+.o
3.0
2. O
I. 0
3 - 0
-hO
-JL-0
-1.
i i i i i i i i i i i i
I
ilflfJNNfii
19.47
i i i l i M i i l i l l i
19
J I I J I L I L L I L L U I N I I L L S L M ^ ^ ^ Y ^ ^
i9.rjr
19
a
if
r7
i m m i i i i U j ^ i
0
,
Do/Alri
11!,ill!
Anr.ual Changes in Currency between Corresponding Months in Adjacent Years
1946-62
(The value shown on the chart for any month is the change in currency fron the
sane rsonth in the previous yeafc to the month indicated.)
i
M
/
s
t U
t n
3r
/ ?<
O
>
/
& 7
i
/
W
cf ,o " n
/ ? t'7
She ratio of the Money Surrly plus Time Deposits in a Given Month to the
Money Supply plus Tine Deposits in the Corresponding Month of the Previous
Year. Monthly January 1946-December 1962.
(The graph indicates the percentage increase or decrease in the money supply
plus tir.e deposits, afa annual rates, fror. month to corresponding month.)
@FedFRASER