to be in charge. Knowing the name, the background,
and the personal qualities of the chairman of the Fed,
for example, is of little use in judging what happened
to monetary growth during his term of office.
If the present monetary structure were producing
satisfactory results, we would be well advised to leave
it alone. Tactics would then be the only topic. Howev-
er, the present monetary structure is not producing
satisfactory results. Indeed, in my opinion, no major
institution in the United States has so poor a record of
performance over so long a period yet so high a public
reputation as the Federal Reserve.
The conduct of monetary policy is of major impor-
tance; monetary instability breeds economic instabil-
ity. A monetary structure that fosters steadiness and
predictability in the general price level is an essential
precondition for healthy noninflationary growth. That
is why it is important to consider fundamental changes
in our monetary institutions. Such changes may be
neither feasible nor urgent now. But unless we consider
them now, we shall not be prepared to adopt them
when and if the need is urgent.
The tactics for avoiding a crisis
Three issues are involved in the tactics of monetary
policy: adopting a variable or variables as an interme-
diate target or targets; choosing the desired path of the
target variables; and devising procedures for achieviny
that path as closely as possible.
The intermediate targets. The Fed has vacillated
between using one or more interest rates or one or
more monetary aggregates as its intermediate targets.
In the past decade, however, it joined monetary au-
thorities in other countries in stressing monetary
growth. Since 1975, it has been required by Congress
to specify explicit numerical targets for the growth of
monetary aggregates. Although many proposals have
recently surfaced for the substitution of other targets-
from real interest rates to sensitive commodity prices
to the price of gold to nominal GNP-I shall assume
that one or more monetary aggregates remains the
intermediate target.
In my opinion, the selection of a target or of a target
path is not and has not been the problem. If the Fed had
consistently achieved the targets it specified to Con-
gress, monetary growth would have been highly stable
instead of highly variable, inflation would never have
become the menace it did, and the United States would
have been spared the worst parts of the punishing re
cession (or recessions) from 1979 to 1982.
The Fed has specified targets for several aggregates
primarily, as I have argued elsewhere, to obfuscate the
issue and reduce accountability. In general, the differ-
ent aggregates move together. The exceptions have
essentially all been due to the interest-rate restrictions
imposed by the Fed under Regulation Q and the associ-
ated development of new forms of deposit liabilities.
And they would not have arisen if the Fed had achieved
its targets for any one of the aggregates.
The use of multiple intermediate targets is undesira-
ble. The Fed has one major instrument of monetary
control: control over the quantity of high-powered
money. With one instrument, it cannot independently
control several aggregates. Its other instruments-pri-
marily the discount rate and reserve requirements-are
highly defective as instruments for monetary control
and of questionable effectiveness in enabling it to con-
trol separately more than one aggregate.
It makes far less difference which aggregate the Fed
selects than that it selects one and only one. For sim-
plicity of exposition, I shall assume that the target
aggregate is M, as currently designed. Selection of
another aggregate would alter the desirable numerical
targets but not their temporal pattern.
The target path. A long-run growth rate of about I
to 3 percent per year for M, would be roughly consis-
tent with zero inflation. That should be our objective.
Actual growth in M, was 10.4 percent from fourth
quarter 1982 to fourth quarter 1983; 5.2 percent from
fourth quarter 1983 to fourth quarter 1984. A crucial
question is how rapidly to go from such levels to the I
to 3 percent range. In my opinion, it is desirable to
proceed gradually, over something like a three- to five-
year period, which means that the rate of growth
should be reduced by about 1 to 1.5 percentage points a
year-a very different pattern from the erratic ups and
downs of recent years.
The Fed has consistently stated its targets in terms
of a range of growth rates. For example, its initial
target for M, for 1983 was a growth rate of 4 to 8
percent from the fourth quarter of 1982 to the fourth
quarter of 1983. That method of stating targets is seri-
ously defective: it provides a widening cone of limits
on the absolute money supply as the year proceeds and
fosters a shift in base from year to year, thereby frus-
trating accountability over long periods. This is indeed
what happened. In July 1983, Chairman Volcker an-
nounced a new target of 5 to 9 percent for the second
quarter of 1983 to the second quarter of 1984 but from
the second-quarter 1983 base, which is 3 percent (6
percent at an annual rate) above the top of the earlier
range.
A better way to state the targets is in terms of a
central target for the absolute money supply plus or
minus a band of, say, 1.5 percent on either side-about
the range the Fed has specified for annual growth
rates. [Since this was written and initially published,
the Council of Economic Advisers has made the same
suggestion, and Chairman Volcker has expressed sup-
port for such a change.]
Procedures for hitting the target. There is wide-
spread agreement both inside and outside the Federal
Reserve System that current procedures and reserve
regulations make accurate control of monetary growth
over short periods difficult or impossible. These pro-
cedures and regulations do not explain such long sus-
tained departures from the targets as the monetary
explosions from April 1980 to April 1981 or July 1982
to July 1983 or the monetary retardations from April
1981 to October 1981 or January 1982 to July 1982.
However, they do explain the wide volatility in mone-
tary growth from week to week and month to month,
which introduces undesirable uncertainty into the
economy and financial markets and reduces Fed ac-
countability for not hitting its targets.
There is also widespread agreement about the
changes in procedures and regulations that would en-
able the Fed to come very much closer to hitting its
targets over fairly short periods. The most important
such change was the replacement of lagged reserve
accounting, introduced in 1968, by contemporaneous
reserve accounting comparable to that prevailing from
1914 to 1968. The obstacle to controlling monetary
growth posed by lagged reserve accounting has been
recognized since 1970 at the latest. Unfortunately, the
Fed did not act until 1982, when it finally decided to
replace lagged by contemporary reserve requirements.
However, it delayed implementation until February
1984-the longest delay in implementing a changed
regulation in the history of the Fed. There was no
insuperable technical obstacle to implementing the
change more promptly.
The other major procedural changes needed are:
1. Selection by the Fed of a single monetary target
to end the Fed's juggling between targets;
2. Imposition of the same percentage reserve re-
quirements on all deposit components of the selected
target;
3. The use of total rather than nonborrowed re
serves as the short-term operating instrument;
4. Linking of the discount rate to a market rate and
making it a penalty rate (neither this change nor the
preceding was feasible for technical reasons under
lagged reserve accounting; they are now feasible, but
neither has been adopted);
5. Reduction of the churning in which the Fed en-
gages in the course of its so-called defensive open-
market operations.
Even without most of these changes, it would be
possible for the Fed to put into effect almost instanta-
neously a policy that would provide a far stabler mone-
tary environment than we have at present, even though
it would by no means be ideal. The obstacle is not
feasibility but bureaucratic inertia and the preservation
of bureaucratic power and status.
A simple example will illustrate. Let the Fed con-
tinue to state targets for M, growth. Let it estimate the
change in its total holdings of U.S. government securi-
ties that would be required in the next six months, say,
to produce the targeted growth in M,. Divide that
amount by 26. Let the Fed purchase the resulting
amount every week on the open market, in addition to
any amount needed to replace maturing securities, and
make no other purchases or sales. Finally, let it an-
nounce this schedule of purchases in advance and in
full detail and stick to it.
Such a policy would assure control over the mone-
tary aggregates, not from day to day, but over the
longer period that the Fed insists is all that matters. It
would enable the market to know precisely what the
Fed would do and adjust its own actions accordingly. It
would end the weekly guessing game that currently
follows each Thursday's release of figures on the mon-
ey supply. The financial markets have certainly dem-
onstrated that they have ample flexibility to handle
whatever day-to-day or seasonal adjustments might be
needed. It is hard to envisage any significant adverse
effects from such a policy.
A few numbers will show how much difference
such a policy would make to the Fed's open-market
activities. In 1982, it added an average of $176 million
a week to its total holdings of government securities-
an unusually high amount. In the process of acquiring
$176 million, it purchased each week an average of
$13 billion of securities and sold nearly as much.
About half of these transactions were on behalf of
foreign central banks. But that still leaves roughly $40
of purchases or $80 of transactions for every one dollar
added to its portfolio-a degree of churning of a custo-
mer's account that would send a private stockbroker to
jail, or at least to limbo.
Increased predictability, reduced churning, the loss
of inscrutability-these are at the same time the major
reasons for making so drastic a change and the major
obstacles to its achievement. It would simply upset too
many comfortable dovecotes.
A framework for basic reform
The chief problem in discussing the framework of
monetary policy is to set limits. The subject is old, yet
immediately pertinent; numerous proposals have been
made, and few, however ancient, do not have contem-
porary proponents. In view of my own belief that the
important desiderata of structural reform are to reduce
the variability of monetary growth, to limit the discre-
tion of the monetary authorities, and to provide a stable
monetary framework, I shall limit myself to proposals
directed at those objectives, proceeding from the least
to the most radical.
Imposing a monetary rule on the Fed. I have long
argued that a major improvement in monetary policy
could be achieved without any significant change in
monetary institutions simply by imposing a monetary
rule on the Fed. From an economic point of view, it
would be desirable to state the rule in terms of a mone-
tary aggregate such as M, that has a close and consis-
tent relation to subsequent changes in national income.
However, recent years have demonstrated that the Fed
has been unable or unwilling to achieve such a target,
even when it sets it itself, and that it has been able to
plead inability and thereby avoid accountability. Ac-
cordingly, I have reluctantly decided that it is prefer-
able to state the rule in terms of a magnitude that has a
somewhat less close relation to national income but
that unquestionably can be controlled within very nar-
row limits within very brief time periods, namely, the
Fed's own non-interest-bearing obligations, the mone-
tary base.
In Free to Choose, my wife, Rose, and I proposed a
specific form of rule as a constitutional amendment:
"Congress shall have the power to authorize non-in-
terest-bearing obligations of the government in the
form of currency or book entries, provided that the
total dollar amount outstanding increases by no more
than 5 percent per year and no less than 3 percent.
"It might be desirable to include a provision that
two-thirds of each House of Congress, or some similar
qualified majority, can waive the requirement in case
of a declaration of war, the suspension to terminate
A constitutional amendment would be the most ef-
fective way to establish confidence in the stability of
the rule. However, it is clearly not the only way to
impose the rule. Congress could equally well legislate
it, and, indeed, proposals for a legislated movroan,:
rule have been introduced in Congress.
I remain persuaded that a monetary rule that leads to
a predictable long-run path of a specified monetary
aggregate is a highly desirable goal-superior either to
discretionary control of the quantity of money by a set
of monetary authorities or to a commodity standard.
However, I am no longer so optimistic as I once was
that it can be effected by either persuading the mone-
tary authorities to follow it or legislating its adoption.
Congressional attempts in the past decade to push the |
Fed in that direction have repeatedly failed. The Fed I
has rhetorically accepted monetary targets but never a
firm monetary rule. Moreover, the Fed has not been
willing even to match its performance to a rhetorical
acceptance of monetary targets. All this suggests that a
change in our monetary institutions is required in or-
der to make such a rule effective.
Separating regulatory from monetary functions . A
modest institutional reform that promises considerable
benefits is to separate the regulatory from the mone-
tary functions of the Fed. Currently, regulatory func-
tions absorb most of the Fed's attention. Moreover,
they obscure accountability for monetary control by
confusing the two very separate and to some extent
inconsistent functions.
As has recently been proposed in a study of the
Federal Deposit Insurance Corporation, the Fed
should be stripped of its regulatory functions, which
would be combined with the largely overlapping func-
tions of the FDIC, the Federal Savings and Loan Insur-
ance Corporation, and the comptroller of the currency.
Such a combined agency should have no monetary
powers. It also might well include the operating func-
tions of the Federal Reserve Banks-the monitoring of
reserve requirements, issuance of currency, clearing
of checks, reporting of data, and so forth.
A separate monetary-control agency could be a very
small body, charged solely with determining the total
quantity of high-powered money through open-market
operations. Its function would be clear, highly visible,
and subject to effective accountability.
Ending the independence of the Fed. An approach
that need involve relatively little institutional change-
although it is far more drastic than the preceding-and
that could be implemented by legislation would be to
end the independence of the Fed by converting it into a
bureau of the Treasury Department. That would end
the present division of responsibilities for monetary
and fiscal policy that leads to the spectacle of chairmen
of the Fed blaming all the nation's ills on the defects of
fiscal policy and secretaries of the Treasury blaming
them on the defects of monetary policy-a phenom-
enon that has prevailed for decades. There would be a
single locus of authority that could be held responsible.
The immediate objection that arises is that it would
make monetary policy a plaything of politics. My own
examination of monetary history indicates that this
judgment is correct, but that it is an argument for, not
against, eliminating the central bank's independence.
I examined this issue at length in an article pub-
lished more than two decades ago entitled "Should
There Be an Independent Monetary Authority?" I con-
cluded that it is "highly dubious that the United States,
or for that matter any other country, has in practice
ever had an independent central bank in [the] fullest
sense of the term.... To judge by experience, even
those central banks that have been nominally indepen-
dent in the fullest sense of the term have in fact been
closely linked to the executive authority.
"But of course this does not dispose of the matter.
The ideal is seldom fully realized. Suppose we could
have an independent central bank in the sense of a
coordinate constitutionally established, separate orga-
nization. Would it be desirable to do so? I think not, for
both political and economic reasons.
"The political objections are perhaps more obvious
than the economic ones. Is it really tolerable in a de-
mocracy to have so much power concentrated in a
body free from any kind of direct effective political
control? . . .
"One [economic] defect of an independent central
bank . . . is that it almost inevitably involves dispersal
of responsibility....
"Another defect . . . is the extent to which policy
is ... made highly dependent on personalities....
"A third technical defect is that an independent
central bank will almost inevitably give undue empha-
sis to the point of view of bankers.
"The three defects I have outlined constltute a
strong technical argument against an independent cen-
tral bank."
The experience of the past two decades has led me to
alter my views in one respect only-about the impor-
tance of personalities. They have on occasion made a
great deal of difference, but additional experience and
study has impressed me with the continuity of Fed
policy, despite the wide differences in the personalities
and backgrounds of the persons supposedly in charge.
For the rest, experience has reinforced my views.
Anna Schwartz and I pointed out in Monetary History
that subservience to congressional pressure in 1930
and 1931 would have prevented the disastrous mone-
tary policy followed by the Fed. That is equally true
for the past fifteen years. The relevant committees of
Congress have generally, though by no means invari-
ably, urged policies on the Fed that would have pro-
duced a stabler rate of monetary growth and much less
inflation. Excessively rapid and volatile monetary
growth from, say, 1971 to 1979 was not the result of
political pressure-certainly not from Congress, al-
though in some of these years there clearly was pres-
sure for more rapid growth from the Administration.
Nonetheless, no political pressures would have pre-
vented the Fed from increasing M, over this period at,
say, an average annual rate of S percent-the rate of
increase during the prior eight years-instead of 6.7
percent.
Subordinating the Fed to the Treasury is by no
means ideal. Yet it would be a great improvement over
the existing situation, even with no other changes.
Competitive issue of money. Increasing interest
has been expressed in recent years in proposals to
replace governmental issuance of money and control
of its quality by private market arrangements. One set
of proposals would end the government monopoly on
the issuance of currency and permit its competitive
issue. Another would eliminate entirely any issuance
of money by government and, instead, restrict the role
of government to defining a monetary unit.
The former set of proposals derives largely from a
pamphlet by F. A. Hayek entitled Choice in Currency:
A Way to Stop Inflation. Hayek proposed that all spe-
cial privileges (such as "legal tender" quality) at-
tached to government-issued currency be removed,
and that financial institutions be permitted to issue
currency or deposit obligations on whatever terms
were mutually acceptable to the issuer and the holder of
the liabilities. He envisaged a system in which institu-
tions would in fact issue obligations expressed in
terms of purchasing power either of specific commod-
ities, such as gold or silver, or of commodities in
general through linkage to a price index. In his opin-
ion, constant-purchasing-power moneys would come
to dominate the market and largely replace obligations
denominated in dollars or pounds or other similar units
and in specific commodities.
The idea of a currency unit linked to a price index is
an ancient one-proposed in the nineteenth century by
W. Stanley Jevons and Alfred Marshall, who named it
a "tabular" standard-and repeatedly rediscovered. It
is part of the theoretically highly attractive idea of
widespread indexation. Experience, however, has
demonstrated that the theoretical attractiveness of the
idea is not matched by practice.
I approve of Professor Hayek's proposal to remove
restrictions on the issuance of private moneys to com-
pete with government moneys. But I do not share his
belief about the outcome. Private moneys now exist-
traveler's checks and cashier's checks, bank deposits,
money orders, and various forms of bank drafts and
negotiable instruments. But these are almost all claims
on a specified number of units of government currency
(of dollars or pounds or francs or marks). Currently,
they are subject to government regulation and control.
But even if such regulations and controls were entirely
eliminated, the advantage of a single national currency
unit buttressed by long tradition will, I suspect, serve
to prevent any other type of private currency unit from
seriously challenging the dominant government cur-
rency, and this despite the high degree of monetary
variability many countries have experienced over re-
cent decades.
The recent explosion in financial futures markets
offers a possible new road to the achievement, through
private market actions, of the equivalent of a tabular
standard. This possibility is highly speculative-little
more than a gleam in one economist's eye. It involves
the establishment of futures markets in one or more
price indexes-strictly parallel to the markets that have
developed in stock-price indexes. (The Commodities
Futures Trading Commission has authorized the Cof-
fee, Sugar, and Cocoa Exchange to begin futures trad-
ing in the Consumer Price Index as of June 21, 1985.)
Such markets, if active and covering a considerable
range of future dates, would provide a relatively cost-
less means of hedging long-term contracts against
risks of changes in the price level. A combination of an
orthodox dollar contract plus a properly timed set of
futures in a price level would be the precise equivalent
of a tabular standard, but would have the advantage
that any one party to a contract, with the help of specu-
lators and other hedgers in the futures market, could
have the benefit of a tabular standard without the
agreement of the other party or parties.
Recent changes in banking regulations have opened
another route to a partial tabular standard on a substan-
tial scale. The Federal Home Loan Bank has finally
authorized federally chartered savings and loan associ-
ations to offer price-level-adjusted mortage (PLAM)
loans. Concurrently, the restrictions on the interest rate
that can be paid on deposits by a wide range of financial
institutions have been eased and removed entirely for
deposits of longer maturities.
This would permit financial institutions simulta-
neously to lend and borrow on a price-level-adjusted
basis: to lend on a PLAM and borrow on a price-level-
adjusted deposit (PLAD), both at an interest rate speci-
fied in real rather than nominal terms. By matching
PLAM loans against PLAD deposits, a bank would be
fully hedged against changes in inflation, covering its
costs by the difference between the interest rate it
charges and pays. Similarly, both borrowers and lend-
ers would be safeguarded against changes in inflation
with respect to a particular liability and asset.
As yet, I know of no financial institutions that have
proceeded along these lines. I conjecture that no major
development will occur unless and until inflation once
again accelerates. When and if that occurs, PLAMs
and PLADs may well become household words and
not simply mysterious acronyms.
Freezing high-powered money. The final propos-
al combines features from most of the preceding. It is
radical and far-reaching, yet simple.
The proposal is that, after a transition period, the
quantity of high-powered money-non-interest-bear-
ing obligations of the U.S. government-be frozen at a
fixed amount. These non-interest-bearing obligations
now take two forms: currency and deposits at the Fed-
eral Reserve System. The simplest way to envisage the
change is to suppose that Federal Reserve deposit lia-
bilities were replaced dollar for dollar by currency
notes, which were turned over to the owners of those
deposits. Thereafter, the government's monetary role
would be limited to keeping the amount constant by
replacing worn-out currency. In effect, a monetary
rule of zero growth in high-powered money would be
adopted. (In practice, it would not be necessary to
replace deposits at the Federal Reserve with currency;
they could be retained as book entries, so long as the
total of such book entries plus currency notes was kept
constant . )
This proposal would be consistent with, indeed re-
quire, the continued existence of private institutions
issuing claims to government currency. These could be
regulated as now, with the whole paraphernalia of re
quired reserves, bank examinations, limitations on
lending, and the like. However, they could also be
freed from all or most such regulations. In particular,
the need for reserve requirements to enable the Fed to
control the quantity of money would disappear.
Reserve requirements might still be desirable for a
different though related reason. The new monetary
economists argue that only the existence of such gov-
ernment regulations as reserve requirements and pro-
hibition of the private issuance of currency explains the
relatively stable demand for high-powered money. In
the absence of such regulations, they contend, non-
interest-bearing money would be completely dominat-
ed by interest-bearing assets, or, at the very least, the
demand for such money would be rendered highly
unstable.
I am far from persuaded by this contentlon. It sup-
poses a closer approach to a frictionless world with
minimal transaction costs than seems to me a useful
approximation to the actual world. Nonetheless, it is
arguable that the elimination of reserve requirements
would introduce an unpredictable and erratic element
into the demand for high-powered money. For that
reason, although personally I would favor the deregu-
lation of financial institutions, thereby incorporating a
major element of Hayek's proposed competitive finan-
cial system, it would seem prudent to proceed in
stages: first, freeze high-powered money; then, after a
period, eliminate reserve requirements and other re-
maining regulations, including the prohibition on the
issuance of hand-to-hand currency by private insti-
tutions.
Why zero growth? Zero has a special appeal on
political grounds that is not shared by any other num-
ber. If 3 percent, why not 4 percent? It is hard, as it
were, to go to the political barricades to defend 3
rather than 4, or 4 rather than 5. But zero is-as a
psychological matter-qualitatively different. It is
what has come to be called a Schelling point-a natural
point at which people tend to agree, like ''splitting the
difference" in a dispute over a monetary sum. More-
over, by removing any power to create money it elimi-
nates institutional arrangements lending themselves to
l discretionary changes in monetary growth.
Would zero growth in high-powered money be con-
sistent with a healthy economy? In the hypothetical
long-long-run stationary economy, when the whole
economy had become adjusted to the situation, and
population, real output, and so on were all stationary,
zero growth in high-powered money would imply zero
growth in other monetary aggregates and mean stable
velocities for the aggregates. In consequence, the price
level would be stable. In a somewhat less than station-
ary state in which output was rising, if financial inno-
vations kept pace, the money multiplier would tend to
rise at the same rate as output, and again prices would
be stable. If financial innovations ceased but total out-
put continued to rise, prices would decline. If output
rose at about 3 percent per year, prices would tend to
fall at 3 percent per year. So long as that was known
and relatively stable, all contracts could be adjusted to
it, and it would cause no problems and indeed would
have some advantages.
However, any such outcome is many decades away.
The more interesting and important question is not the
final stationary-state result but the intermediate dy-
namic process.
Once the policy was in effect, the actual behavior of
nominal income and the price level would depend on
what happened to a monetary aggregate like M, rela-
tive to high-powered money and what happened to
nominal income relative to M,-that is, on the behav-
ior of the money multiplier (the ratio of M, to high-
powered money) and on the income velocity of M, (the
ratio of nominal income to M,).
Given a loosening of the financial structure through
continued deregulation, there would be every reason to
expect a continued flow of innovations raising the
money multiplier. This process has in fact occurred
throughout the past several centuries. For example, in
the century from 1870 to 1970, the ratio of the quantity
of money, as defined by Anna Schwartz and me in
Monetary History, to high-powered money rose at the
average rate of I percent per year. In the post-World
War II period, the velocity of M, has risen at about 3
percent per year, and at a relatively steady rate. This
trend cannot, of course, continue indefinitely. Above,
in specifying a desirable target for the Fed, I estimated
that the rise in velocity would slow to about 1 or 2
percent per year. However, a complete end to the rapid
trend in velocity is not in sight.
There is no way to make precise numerical esti-
mates, but there is every reason to anticipate that for
decades after the introduction of a freeze on high-
powered money, both the money multiplier and veloc-
ity would tend to rise at rates in the range of historical
experience. Under these circumstances, a zero rate of
growth of high-powered money would imply roughly
stable prices, though ultimately, perhaps, slightly de-
clining prices.
What of the transition? Over the three years from 1979
to 1982, high-powered money grew an average of 7.0
percent a year. It would be desirable to bring that rate
to zero gradually. As for M, growth, about a five-year
period seems appropriate-or a transition that reduces
the rate of growth of high-powered money by about 1.5
percentage points a year. The only other transitional
problem would be to phase out the Fed's powers to
create and destroy high-powered money by open-mar-
ket operations and discounting. Neither transition of-
fers any special problem. The Fed, or its successor
agency, could still use part of the existing stock of
high-powered money for similar purposes, particular-
ly for lender-of-last-resort purposes, if that function
were retained.
The great advantage of this proposal is that it would
end the arbitrary power of the Federal Reserve System
to determine the quantity of money, and would do so
without establishing any comparable locus of power
and without introducing any major disturbances into
other existing economic and financial institutions.
I have found that few things are harder even for
knowledgeable nonexperts to accept than the proposi-
tion that twelve (or nineteen) people sitting around a
table in Washington, subject to neither election nor
dismissal, nor close administrative or political con-
trol, have the power to determine the quantity of mon-
ey-to permit a reduction by one-third during the
Great Depression or a near-doubling from 1970 to
1980. That power is too important, too pervasive, to be
exercised by a few people, however public-spirited, if
there is any feasible alternative.
There is no need for such arbitrary power. In the
system I have just described, the total quantity of any
monetary aggregate would be determined by the mar-
ket interactions of many financial institutions and mil-
lions of holders of monetary assets. It would be limited
by the constant quantity of high-powered money avail-
able as ultimate reserves. The ratios of various aggre-
gates to high-powered money would doubtless change
from time to time, but in the absence of rigid govern-
ment controls-such as those exemplified by Regula-
tion Q, fortunately being phased out-the ratios would
change gradually and only as financial innovations or
changes in business and industry altered the propor-
tions in which the public chose to hold various mone-
tary assets. No small number of individuals would be
in a position to introduce major changes in the ratios or
in the rates of growth of various monetary aggre-
gates-to move, for example, from a 3 percent per year
rate of growth in M, for one six-month period (January
to July 1982) to a 13 percent rate of growth for the next
six months (July 1982 to January 1983).
Conclusion
Major institutional change occurs only at times of cri-
sis. For the rest, the tyranny of the status quo limits
changes in institutions to marginal tinkering-we
muddle through. It took the Great Depression to pro-
duce the FDIC, the most important structural change
in our monetary institutions since at least 1914, when
the Federal Reserve System began operations, and to
shift power over monetary policy from the Federal
Reserve Banks, especially that in New York, to the
Board in Washington. Since then, our monetary insti-
tutions have been remarkably stable. It took the severe
inflation of the 1970s and accompanying double-digit
interest rates-combined with the enforcement of Reg-
ulation Q-to produce money-market mutual funds
and thereby force a considerable measure of deregula-
tion of banking.
Nonetheless, it is worth discussing radical changes,
not in the expectation that they will be adopted prompt-
ly but for two other reasons. One is to construct an
ideal goal, so that incremental changes can be judged
by whether they move the institutional structure to-
ward or away from that ideal.
The other reason is very different. It is so that if a
crisis requiring or facilitating radical change does
arise, alternatives will be available that have been
carefully developed and fully explored. International
monetary arrangements provide an excellent example.
For decades, economists had been exploring alterna-
tives to the system of fixed exchange rates-in particu-
lar, floating exchange rates among national currencies.
The practical men of affairs derided proposals for
floating rates as unrealistic, impractical, ivory-tower.
Yet when crisis came, when the Bretton Woods fixed-
rate system had to be scrapped, the theorists' impracti-
cal proposal became highly practical and formed the
basis for the new system of international monetary
arrangements.
Needless to say, I hope that no crises will occur that
will necessitate a drastic change in domestic monetary
institutions. The most likely such crisis is continued
monetary instability, a return to a roller coaster of
inflation about an upward trend, with inflation acceler-
ating to levels of 20, 30, or more percent per year. That
would shake the social and political framework of the
nation and would produce results none of us would like
to witness. Yet, it would be burying one's head in the
sand to fail to recognize that such a development is a
real possibility. It has occurred elsewhere, and it could
occur here. If it does, the best way to cut it short, to
minimize the harm it would do, is to be ready not with
Band-Aids but with a real cure for the basic illness.
As of now, I believe the best real cure would be the
reform outlined above: abolish the money-creating
powers of the Federal Reserve, freeze the quantity of
high-powered money, and deregulate the financial
system.
The less radical changes in policy and procedures
suggested in the section on tactics seem to me to offer
the best chance of avoiding a crisis. These tactical
changes are feasible technically. However, I am not
optimistic that they will be adopted. The obstacle is
political: as with any bureaucratic organization, it is
not in the self-interest of the Fed to adopt policies that
would render it accountable. The Fed has persistently
avoided doing so over a long period. None of the
tactics that I have proposed is new. The proposed
changes would have made just as much sense five or
ten years ago-indeed, if adopted then, the inflation
and volatility of the past ten years would never have
occurred. The proposals have had the support of a
large fraction of monetary experts outside the Fed.
The Fed has resisted them for bureaucratic and politi-
cal, not technical, reasons. And resistance has been in
the Fed's interest. By keeping monetary policy an ar-
cane subject that must be entrusted to "experts" and
kept out of politics, incapable of being judged by non-
experts, the Fed has been able to maintain the high
public reputation of which I spoke at the outset of this
article, despite its poor record of performance.
One chairman after another, in testimony to Con-
gress, has emphasized the mystery and difficulty of the
Fed's task and the need for discretion, judgment, and
the balancing of many considerations. Each has
stressed how well the Fed has done and proclaimed its
dedication to pursuing a noninflationary policy and has
attributed any undesirable outcome to forces outside
the Fed's control or to deficiencies in other compo-
nents of government policy-particularly fiscal policy.
The testimony of the four most recent chairmen of the
Fed documents their pervasive concern with avoiding
accountability-a concern with which it is easy to sym-
pathize in view of the purely coincidental relation be-
tween their announced intentions and the actual
outcome.
Clearly the problem is not the person who happens
to be chairman, but the system.
to be in charge. Knowing the name, the background,
and the personal qualities of the chairman of the Fed,
for example, is of little use in judging what happened
to monetary growth during his term of office.
If the present monetary structure were producing
satisfactory results, we would be well advised to leave
it alone. Tactics would then be the only topic. Howev-
er, the present monetary structure is not producing
satisfactory results. Indeed, in my opinion, no major
institution in the United States has so poor a record of
performance over so long a period yet so high a public
reputation as the Federal Reserve.
The conduct of monetary policy is of major impor-
tance; monetary instability breeds economic instabil-
ity. A monetary structure that fosters steadiness and
predictability in the general price level is an essential
precondition for healthy noninflationary growth. That
is why it is important to consider fundamental changes
in our monetary institutions. Such changes may be
neither feasible nor urgent now. But unless we consider
them now, we shall not be prepared to adopt them
when and if the need is urgent.
The tactics for avoiding a crisis
Three issues are involved in the tactics of monetary
policy: adopting a variable or variables as an interme-
diate target or targets; choosing the desired path of the
target variables; and devising procedures for achieviny
that path as closely as possible.
The intermediate targets. The Fed has vacillated
between using one or more interest rates or one or
more monetary aggregates as its intermediate targets.
In the past decade, however, it joined monetary au-
thorities in other countries in stressing monetary
growth. Since 1975, it has been required by Congress
to specify explicit numerical targets for the growth of
monetary aggregates. Although many proposals have
recently surfaced for the substitution of other targets-
from real interest rates to sensitive commodity prices
to the price of gold to nominal GNP-I shall assume
that one or more monetary aggregates remains the
intermediate target.
In my opinion, the selection of a target or of a target
path is not and has not been the problem. If the Fed had
consistently achieved the targets it specified to Con-
gress, monetary growth would have been highly stable
instead of highly variable, inflation would never have
become the menace it did, and the United States would
have been spared the worst parts of the punishing re
cession (or recessions) from 1979 to 1982.
The Fed has specified targets for several aggregates
primarily, as I have argued elsewhere, to obfuscate the
issue and reduce accountability. In general, the differ-
ent aggregates move together. The exceptions have
essentially all been due to the interest-rate restrictions
imposed by the Fed under Regulation Q and the associ-
ated development of new forms of deposit liabilities.
And they would not have arisen if the Fed had achieved
its targets for any one of the aggregates.
The use of multiple intermediate targets is undesira-
ble. The Fed has one major instrument of monetary
control: control over the quantity of high-powered
money. With one instrument, it cannot independently
control several aggregates. Its other instruments-pri-
marily the discount rate and reserve requirements-are
highly defective as instruments for monetary control
and of questionable effectiveness in enabling it to con-
trol separately more than one aggregate.
It makes far less difference which aggregate the Fed
selects than that it selects one and only one. For sim-
plicity of exposition, I shall assume that the target
aggregate is M, as currently designed. Selection of
another aggregate would alter the desirable numerical
targets but not their temporal pattern.
The target path. A long-run growth rate of about I
to 3 percent per year for M, would be roughly consis-
tent with zero inflation. That should be our objective.
Actual growth in M, was 10.4 percent from fourth
quarter 1982 to fourth quarter 1983; 5.2 percent from
fourth quarter 1983 to fourth quarter 1984. A crucial
question is how rapidly to go from such levels to the I
to 3 percent range. In my opinion, it is desirable to
proceed gradually, over something like a three- to five-
year period, which means that the rate of growth
should be reduced by about 1 to 1.5 percentage points a
year-a very different pattern from the erratic ups and
downs of recent years.
The Fed has consistently stated its targets in terms
of a range of growth rates. For example, its initial
target for M, for 1983 was a growth rate of 4 to 8
percent from the fourth quarter of 1982 to the fourth
quarter of 1983. That method of stating targets is seri-
ously defective: it provides a widening cone of limits
on the absolute money supply as the year proceeds and
fosters a shift in base from year to year, thereby frus-
trating accountability over long periods. This is indeed
what happened. In July 1983, Chairman Volcker an-
nounced a new target of 5 to 9 percent for the second
quarter of 1983 to the second quarter of 1984 but from
the second-quarter 1983 base, which is 3 percent (6
percent at an annual rate) above the top of the earlier
range.
A better way to state the targets is in terms of a
central target for the absolute money supply plus or
minus a band of, say, 1.5 percent on either side-about
the range the Fed has specified for annual growth
rates. [Since this was written and initially published,
the Council of Economic Advisers has made the same
suggestion, and Chairman Volcker has expressed sup-
port for such a change.]
Procedures for hitting the target. There is wide-
spread agreement both inside and outside the Federal
Reserve System that current procedures and reserve
regulations make accurate control of monetary growth
over short periods difficult or impossible. These pro-
cedures and regulations do not explain such long sus-
tained departures from the targets as the monetary
explosions from April 1980 to April 1981 or July 1982
to July 1983 or the monetary retardations from April
1981 to October 1981 or January 1982 to July 1982.
However, they do explain the wide volatility in mone-
tary growth from week to week and month to month,
which introduces undesirable uncertainty into the
economy and financial markets and reduces Fed ac-
countability for not hitting its targets.
There is also widespread agreement about the
changes in procedures and regulations that would en-
able the Fed to come very much closer to hitting its
targets over fairly short periods. The most important
such change was the replacement of lagged reserve
accounting, introduced in 1968, by contemporaneous
reserve accounting comparable to that prevailing from
1914 to 1968. The obstacle to controlling monetary
growth posed by lagged reserve accounting has been
recognized since 1970 at the latest. Unfortunately, the
Fed did not act until 1982, when it finally decided to
replace lagged by contemporary reserve requirements.
However, it delayed implementation until February
1984-the longest delay in implementing a changed
regulation in the history of the Fed. There was no
insuperable technical obstacle to implementing the
change more promptly.
The other major procedural changes needed are:
1. Selection by the Fed of a single monetary target
to end the Fed's juggling between targets;
2. Imposition of the same percentage reserve re-
quirements on all deposit components of the selected
target;
3. The use of total rather than nonborrowed re
serves as the short-term operating instrument;
4. Linking of the discount rate to a market rate and
making it a penalty rate (neither this change nor the
preceding was feasible for technical reasons under
lagged reserve accounting; they are now feasible, but
neither has been adopted);
5. Reduction of the churning in which the Fed en-
gages in the course of its so-called defensive open-
market operations.
Even without most of these changes, it would be
possible for the Fed to put into effect almost instanta-
neously a policy that would provide a far stabler mone-
tary environment than we have at present, even though
it would by no means be ideal. The obstacle is not
feasibility but bureaucratic inertia and the preservation
of bureaucratic power and status.
A simple example will illustrate. Let the Fed con-
tinue to state targets for M, growth. Let it estimate the
change in its total holdings of U.S. government securi-
ties that would be required in the next six months, say,
to produce the targeted growth in M,. Divide that
amount by 26. Let the Fed purchase the resulting
amount every week on the open market, in addition to
any amount needed to replace maturing securities, and
make no other purchases or sales. Finally, let it an-
nounce this schedule of purchases in advance and in
full detail and stick to it.
Such a policy would assure control over the mone-
tary aggregates, not from day to day, but over the
longer period that the Fed insists is all that matters. It
would enable the market to know precisely what the
Fed would do and adjust its own actions accordingly. It
would end the weekly guessing game that currently
follows each Thursday's release of figures on the mon-
ey supply. The financial markets have certainly dem-
onstrated that they have ample flexibility to handle
whatever day-to-day or seasonal adjustments might be
needed. It is hard to envisage any significant adverse
effects from such a policy.
A few numbers will show how much difference
such a policy would make to the Fed's open-market
activities. In 1982, it added an average of $176 million
a week to its total holdings of government securities-
an unusually high amount. In the process of acquiring
$176 million, it purchased each week an average of
$13 billion of securities and sold nearly as much.
About half of these transactions were on behalf of
foreign central banks. But that still leaves roughly $40
of purchases or $80 of transactions for every one dollar
added to its portfolio-a degree of churning of a custo-
mer's account that would send a private stockbroker to
jail, or at least to limbo.
Increased predictability, reduced churning, the loss
of inscrutability-these are at the same time the major
reasons for making so drastic a change and the major
obstacles to its achievement. It would simply upset too
many comfortable dovecotes.
A framework for basic reform
The chief problem in discussing the framework of
monetary policy is to set limits. The subject is old, yet
immediately pertinent; numerous proposals have been
made, and few, however ancient, do not have contem-
porary proponents. In view of my own belief that the
important desiderata of structural reform are to reduce
the variability of monetary growth, to limit the discre-
tion of the monetary authorities, and to provide a stable
monetary framework, I shall limit myself to proposals
directed at those objectives, proceeding from the least
to the most radical.
Imposing a monetary rule on the Fed. I have long
argued that a major improvement in monetary policy
could be achieved without any significant change in
monetary institutions simply by imposing a monetary
rule on the Fed. From an economic point of view, it
would be desirable to state the rule in terms of a mone-
tary aggregate such as M, that has a close and consis-
tent relation to subsequent changes in national income.
However, recent years have demonstrated that the Fed
has been unable or unwilling to achieve such a target,
even when it sets it itself, and that it has been able to
plead inability and thereby avoid accountability. Ac-
cordingly, I have reluctantly decided that it is prefer-
able to state the rule in terms of a magnitude that has a
somewhat less close relation to national income but
that unquestionably can be controlled within very nar-
row limits within very brief time periods, namely, the
Fed's own non-interest-bearing obligations, the mone-
tary base.
In Free to Choose, my wife, Rose, and I proposed a
specific form of rule as a constitutional amendment:
"Congress shall have the power to authorize non-in-
terest-bearing obligations of the government in the
form of currency or book entries, provided that the
total dollar amount outstanding increases by no more
than 5 percent per year and no less than 3 percent.
"It might be desirable to include a provision that
two-thirds of each House of Congress, or some similar
qualified majority, can waive the requirement in case
of a declaration of war, the suspension to terminate
annually unless renewed."
A constitutional amendment would be the most ef-
fective way to establish confidence in the stability of
the rule. However, it is clearly not the only way to
impose the rule. Congress could equally well legislate
it, and, indeed, proposals for a legislated movroan,:
rule have been introduced in Congress.
I remain persuaded that a monetary rule that leads to
a predictable long-run path of a specified monetary
aggregate is a highly desirable goal-superior either to
discretionary control of the quantity of money by a set
of monetary authorities or to a commodity standard.
However, I am no longer so optimistic as I once was
that it can be effected by either persuading the mone-
tary authorities to follow it or legislating its adoption.
Congressional attempts in the past decade to push the |
Fed in that direction have repeatedly failed. The Fed I
has rhetorically accepted monetary targets but never a
firm monetary rule. Moreover, the Fed has not been
willing even to match its performance to a rhetorical
acceptance of monetary targets. All this suggests that a
change in our monetary institutions is required in or-
der to make such a rule effective.
Separating regulatory from monetary functions . A
modest institutional reform that promises considerable
benefits is to separate the regulatory from the mone-
tary functions of the Fed. Currently, regulatory func-
tions absorb most of the Fed's attention. Moreover,
they obscure accountability for monetary control by
confusing the two very separate and to some extent
inconsistent functions.
As has recently been proposed in a study of the
Federal Deposit Insurance Corporation, the Fed
should be stripped of its regulatory functions, which
would be combined with the largely overlapping func-
tions of the FDIC, the Federal Savings and Loan Insur-
ance Corporation, and the comptroller of the currency.
Such a combined agency should have no monetary
powers. It also might well include the operating func-
tions of the Federal Reserve Banks-the monitoring of
reserve requirements, issuance of currency, clearing
of checks, reporting of data, and so forth.
A separate monetary-control agency could be a very
small body, charged solely with determining the total
quantity of high-powered money through open-market
operations. Its function would be clear, highly visible,
and subject to effective accountability.
Ending the independence of the Fed. An approach
that need involve relatively little institutional change-
although it is far more drastic than the preceding-and
that could be implemented by legislation would be to
end the independence of the Fed by converting it into a
bureau of the Treasury Department. That would end
the present division of responsibilities for monetary
and fiscal policy that leads to the spectacle of chairmen
of the Fed blaming all the nation's ills on the defects of
fiscal policy and secretaries of the Treasury blaming
them on the defects of monetary policy-a phenom-
enon that has prevailed for decades. There would be a
single locus of authority that could be held responsible.
The immediate objection that arises is that it would
make monetary policy a plaything of politics. My own
examination of monetary history indicates that this
judgment is correct, but that it is an argument for, not
against, eliminating the central bank's independence.
I examined this issue at length in an article pub-
lished more than two decades ago entitled "Should
There Be an Independent Monetary Authority?" I con-
cluded that it is "highly dubious that the United States,
or for that matter any other country, has in practice
ever had an independent central bank in [the] fullest
sense of the term.... To judge by experience, even
those central banks that have been nominally indepen-
dent in the fullest sense of the term have in fact been
closely linked to the executive authority.
"But of course this does not dispose of the matter.
The ideal is seldom fully realized. Suppose we could
have an independent central bank in the sense of a
coordinate constitutionally established, separate orga-
nization. Would it be desirable to do so? I think not, for
both political and economic reasons.
"The political objections are perhaps more obvious
than the economic ones. Is it really tolerable in a de-
mocracy to have so much power concentrated in a
body free from any kind of direct effective political
control? . . .
"One [economic] defect of an independent central
bank . . . is that it almost inevitably involves dispersal
of responsibility....
"Another defect . . . is the extent to which policy
is ... made highly dependent on personalities....
"A third technical defect is that an independent
central bank will almost inevitably give undue empha-
sis to the point of view of bankers.
"The three defects I have outlined constltute a
strong technical argument against an independent cen-
tral bank."
The experience of the past two decades has led me to
alter my views in one respect only-about the impor-
tance of personalities. They have on occasion made a
great deal of difference, but additional experience and
study has impressed me with the continuity of Fed
policy, despite the wide differences in the personalities
and backgrounds of the persons supposedly in charge.
For the rest, experience has reinforced my views.
Anna Schwartz and I pointed out in Monetary History
that subservience to congressional pressure in 1930
and 1931 would have prevented the disastrous mone-
tary policy followed by the Fed. That is equally true
for the past fifteen years. The relevant committees of
Congress have generally, though by no means invari-
ably, urged policies on the Fed that would have pro-
duced a stabler rate of monetary growth and much less
inflation. Excessively rapid and volatile monetary
growth from, say, 1971 to 1979 was not the result of
political pressure-certainly not from Congress, al-
though in some of these years there clearly was pres-
sure for more rapid growth from the Administration.
Nonetheless, no political pressures would have pre-
vented the Fed from increasing M, over this period at,
say, an average annual rate of S percent-the rate of
increase during the prior eight years-instead of 6.7
percent.
Subordinating the Fed to the Treasury is by no
means ideal. Yet it would be a great improvement over
the existing situation, even with no other changes.
Competitive issue of money. Increasing interest
has been expressed in recent years in proposals to
replace governmental issuance of money and control
of its quality by private market arrangements. One set
of proposals would end the government monopoly on
the issuance of currency and permit its competitive
issue. Another would eliminate entirely any issuance
of money by government and, instead, restrict the role
of government to defining a monetary unit.
The former set of proposals derives largely from a
pamphlet by F. A. Hayek entitled Choice in Currency:
A Way to Stop Inflation. Hayek proposed that all spe-
cial privileges (such as "legal tender" quality) at-
tached to government-issued currency be removed,
and that financial institutions be permitted to issue
currency or deposit obligations on whatever terms
were mutually acceptable to the issuer and the holder of
the liabilities. He envisaged a system in which institu-
tions would in fact issue obligations expressed in
terms of purchasing power either of specific commod-
ities, such as gold or silver, or of commodities in
general through linkage to a price index. In his opin-
ion, constant-purchasing-power moneys would come
to dominate the market and largely replace obligations
denominated in dollars or pounds or other similar units
and in specific commodities.
The idea of a currency unit linked to a price index is
an ancient one-proposed in the nineteenth century by
W. Stanley Jevons and Alfred Marshall, who named it
a "tabular" standard-and repeatedly rediscovered. It
is part of the theoretically highly attractive idea of
widespread indexation. Experience, however, has
demonstrated that the theoretical attractiveness of the
idea is not matched by practice.
I approve of Professor Hayek's proposal to remove
restrictions on the issuance of private moneys to com-
pete with government moneys. But I do not share his
belief about the outcome. Private moneys now exist-
traveler's checks and cashier's checks, bank deposits,
money orders, and various forms of bank drafts and
negotiable instruments. But these are almost all claims
on a specified number of units of government currency
(of dollars or pounds or francs or marks). Currently,
they are subject to government regulation and control.
But even if such regulations and controls were entirely
eliminated, the advantage of a single national currency
unit buttressed by long tradition will, I suspect, serve
to prevent any other type of private currency unit from
seriously challenging the dominant government cur-
rency, and this despite the high degree of monetary
variability many countries have experienced over re-
cent decades.
The recent explosion in financial futures markets
offers a possible new road to the achievement, through
private market actions, of the equivalent of a tabular
standard. This possibility is highly speculative-little
more than a gleam in one economist's eye. It involves
the establishment of futures markets in one or more
price indexes-strictly parallel to the markets that have
developed in stock-price indexes. (The Commodities
Futures Trading Commission has authorized the Cof-
fee, Sugar, and Cocoa Exchange to begin futures trad-
ing in the Consumer Price Index as of June 21, 1985.)
Such markets, if active and covering a considerable
range of future dates, would provide a relatively cost-
less means of hedging long-term contracts against
risks of changes in the price level. A combination of an
orthodox dollar contract plus a properly timed set of
futures in a price level would be the precise equivalent
of a tabular standard, but would have the advantage
that any one party to a contract, with the help of specu-
lators and other hedgers in the futures market, could
have the benefit of a tabular standard without the
agreement of the other party or parties.
Recent changes in banking regulations have opened
248 Section Five Monetary Theory
another route to a partial tabular standard on a substan-
tial scale. The Federal Home Loan Bank has finally
authorized federally chartered savings and loan associ-
ations to offer price-level-adjusted mortage (PLAM)
loans. Concurrently, the restrictions on the interest rate
that can be paid on deposits by a wide range of financial
institutions have been eased and removed entirely for
deposits of longer maturities.
This would permit financial institutions simulta-
neously to lend and borrow on a price-level-adjusted
basis: to lend on a PLAM and borrow on a price-level-
adjusted deposit (PLAD), both at an interest rate speci-
fied in real rather than nominal terms. By matching
PLAM loans against PLAD deposits, a bank would be
fully hedged against changes in inflation, covering its
costs by the difference between the interest rate it
charges and pays. Similarly, both borrowers and lend-
ers would be safeguarded against changes in inflation
with respect to a particular liability and asset.
As yet, I know of no financial institutions that have
proceeded along these lines. I conjecture that no major
development will occur unless and until inflation once
again accelerates. When and if that occurs, PLAMs
and PLADs may well become household words and
not simply mysterious acronyms.
Freezing high-powered money. The final propos-
al combines features from most of the preceding. It is
radical and far-reaching, yet simple.
The proposal is that, after a transition period, the
quantity of high-powered money-non-interest-bear-
ing obligations of the U.S. government-be frozen at a
fixed amount. These non-interest-bearing obligations
now take two forms: currency and deposits at the Fed-
eral Reserve System. The simplest way to envisage the
change is to suppose that Federal Reserve deposit lia-
bilities were replaced dollar for dollar by currency
notes, which were turned over to the owners of those
deposits. Thereafter, the government's monetary role
would be limited to keeping the amount constant by
replacing worn-out currency. In effect, a monetary
rule of zero growth in high-powered money would be
adopted. (In practice, it would not be necessary to
replace deposits at the Federal Reserve with currency;
they could be retained as book entries, so long as the
total of such book entries plus currency notes was kept
constant . )
This proposal would be consistent with, indeed re-
quire, the continued existence of private institutions
issuing claims to government currency. These could be
regulated as now, with the whole paraphernalia of re
quired reserves, bank examinations, limitations on
lending, and the like. However, they could also be
freed from all or most such regulations. In particular,
the need for reserve requirements to enable the Fed to
control the quantity of money would disappear.
Reserve requirements might still be desirable for a
different though related reason. The new monetary
economists argue that only the existence of such gov-
ernment regulations as reserve requirements and pro-
hibition of the private issuance of currency explains the
relatively stable demand for high-powered money. In
the absence of such regulations, they contend, non-
interest-bearing money would be completely dominat-
ed by interest-bearing assets, or, at the very least, the
demand for such money would be rendered highly
unstable.
I am far from persuaded by this contentlon. It sup-
poses a closer approach to a frictionless world with
minimal transaction costs than seems to me a useful
approximation to the actual world. Nonetheless, it is
arguable that the elimination of reserve requirements
would introduce an unpredictable and erratic element
into the demand for high-powered money. For that
reason, although personally I would favor the deregu-
lation of financial institutions, thereby incorporating a
major element of Hayek's proposed competitive finan-
cial system, it would seem prudent to proceed in
stages: first, freeze high-powered money; then, after a
period, eliminate reserve requirements and other re-
maining regulations, including the prohibition on the
issuance of hand-to-hand currency by private insti-
tutions.
Why zero growth? Zero has a special appeal on
political grounds that is not shared by any other num-
ber. If 3 percent, why not 4 percent? It is hard, as it
were, to go to the political barricades to defend 3
rather than 4, or 4 rather than 5. But zero is-as a
psychological matter-qualitatively different. It is
what has come to be called a Schelling point-a natural
point at which people tend to agree, like ''splitting the
difference" in a dispute over a monetary sum. More-
over, by removing any power to create money it elimi-
nates institutional arrangements lending themselves to
discretionary changes in monetary growth.
Would zero growth in high-powered money be con-
sistent with a healthy economy? In the hypothetical
long-long-run stationary economy, when the whole
economy had become adjusted to the situation, and
population, real output, and so on were all stationary,
zero growth in high-powered money would imply zero
growth in other monetary aggregates and mean stable
velocities for the aggregates. In consequence, the price
level would be stable. In a somewhat less than station-
ary state in which output was rising, if financial inno-
vations kept pace, the money multiplier would tend to
rise at the same rate as output, and again prices would
be stable. If financial innovations ceased but total out-
put continued to rise, prices would decline. If output
rose at about 3 percent per year, prices would tend to
fall at 3 percent per year. So long as that was known
and relatively stable, all contracts could be adjusted to
it, and it would cause no problems and indeed would
have some advantages.
However, any such outcome is many decades away.
The more interesting and important question is not the
final stationary-state result but the intermediate dy-
namic process.
Once the policy was in effect, the actual behavior of
nominal income and the price level would depend on
what happened to a monetary aggregate like M, rela-
tive to high-powered money and what happened to
nominal income relative to M,-that is, on the behav-
ior of the money multiplier (the ratio of M, to high-
powered money) and on the income velocity of M, (the
ratio of nominal income to M,).
Given a loosening of the financial structure through
continued deregulation, there would be every reason to
expect a continued flow of innovations raising the
money multiplier. This process has in fact occurred
throughout the past several centuries. For example, in
the century from 1870 to 1970, the ratio of the quantity
of money, as defined by Anna Schwartz and me in
Monetary History, to high-powered money rose at the
average rate of I percent per year. In the post-World
War II period, the velocity of M, has risen at about 3
percent per year, and at a relatively steady rate. This
trend cannot, of course, continue indefinitely. Above,
in specifying a desirable target for the Fed, I estimated
that the rise in velocity would slow to about 1 or 2
percent per year. However, a complete end to the rapid
trend in velocity is not in sight.
There is no way to make precise numerical esti-
mates, but there is every reason to anticipate that for
decades after the introduction of a freeze on high-
powered money, both the money multiplier and veloc-
ity would tend to rise at rates in the range of historical
experience. Under these circumstances, a zero rate of
growth of high-powered money would imply roughly
stable prices, though ultimately, perhaps, slightly de-
clining prices.
What of the transition? Over the three years from 1979
to 1982, high-powered money grew an average of 7.0
percent a year. It would be desirable to bring that rate
to zero gradually. As for M, growth, about a five-year
period seems appropriate-or a transition that reduces
the rate of growth of high-powered money by about 1.5
percentage points a year. The only other transitional
problem would be to phase out the Fed's powers to
create and destroy high-powered money by open-mar-
ket operations and discounting. Neither transition of-
fers any special problem. The Fed, or its successor
agency, could still use part of the existing stock of
high-powered money for similar purposes, particular-
ly for lender-of-last-resort purposes, if that function
were retained.
The great advantage of this proposal is that it would
end the arbitrary power of the Federal Reserve System
to determine the quantity of money, and would do so
without establishing any comparable locus of power
and without introducing any major disturbances into
other existing economic and financial institutions.
I have found that few things are harder even for
knowledgeable nonexperts to accept than the proposi-
tion that twelve (or nineteen) people sitting around a
table in Washington, subject to neither election nor
dismissal, nor close administrative or political con-
trol, have the power to determine the quantity of mon-
ey-to permit a reduction by one-third during the
Great Depression or a near-doubling from 1970 to
1980. That power is too important, too pervasive, to be
exercised by a few people, however public-spirited, if
there is any feasible alternative.
There is no need for such arbitrary power. In the
system I have just described, the total quantity of any
monetary aggregate would be determined by the mar-
ket interactions of many financial institutions and mil-
lions of holders of monetary assets. It would be limited
by the constant quantity of high-powered money avail-
able as ultimate reserves. The ratios of various aggre-
gates to high-powered money would doubtless change
from time to time, but in the absence of rigid govern-
ment controls-such as those exemplified by Regula-
tion Q, fortunately being phased out-the ratios would
change gradually and only as financial innovations or
changes in business and industry altered the propor-
tions in which the public chose to hold various mone-
tary assets. No small number of individuals would be
in a position to introduce major changes in the ratios or
in the rates of growth of various monetary aggre-
gates-to move, for example, from a 3 percent per year
rate of growth in M, for one six-month period (January
to July 1982) to a 13 percent rate of growth for the next
six months (July 1982 to January 1983).
Conclusion
Major institutional change occurs only at times of cri-
sis. For the rest, the tyranny of the status quo limits
changes in institutions to marginal tinkering-we
muddle through. It took the Great Depression to pro-
duce the FDIC, the most important structural change
in our monetary institutions since at least 1914, when
the Federal Reserve System began operations, and to
shift power over monetary policy from the Federal
Reserve Banks, especially that in New York, to the
Board in Washington. Since then, our monetary insti-
tutions have been remarkably stable. It took the severe
inflation of the 1970s and accompanying double-digit
interest rates-combined with the enforcement of Reg-
ulation Q-to produce money-market mutual funds
and thereby force a considerable measure of deregula-
tion of banking.
Nonetheless, it is worth discussing radical changes,
not in the expectation that they will be adopted prompt-
ly but for two other reasons. One is to construct an
ideal goal, so that incremental changes can be judged
by whether they move the institutional structure to-
ward or away from that ideal.
The other reason is very different. It is so that if a
crisis requiring or facilitating radical change does
arise, alternatives will be available that have been
carefully developed and fully explored. International
monetary arrangements provide an excellent example.
For decades, economists had been exploring alterna-
tives to the system of fixed exchange rates-in particu-
lar, floating exchange rates among national currencies.
The practical men of affairs derided proposals for
floating rates as unrealistic, impractical, ivory-tower.
Yet when crisis came, when the Bretton Woods fixed-
rate system had to be scrapped, the theorists' impracti-
cal proposal became highly practical and formed the
basis for the new system of international monetary
arrangements.
Needless to say, I hope that no crises will occur that
will necessitate a drastic change in domestic monetary
institutions. The most likely such crisis is continued
monetary instability, a return to a roller coaster of
inflation about an upward trend, with inflation acceler-
ating to levels of 20, 30, or more percent per year. That
would shake the social and political framework of the
nation and would produce results none of us would like
to witness. Yet, it would be burying one's head in the
sand to fail to recognize that such a development is a
real possibility. It has occurred elsewhere, and it could
occur here. If it does, the best way to cut it short, to
minimize the harm it would do, is to be ready not with
Band-Aids but with a real cure for the basic illness.
As of now, I believe the best real cure would be the
reform outlined above: abolish the money-creating
powers of the Federal Reserve, freeze the quantity of
high-powered money, and deregulate the financial
system.
The less radical changes in policy and procedures
suggested in the section on tactics seem to me to offer
the best chance of avoiding a crisis. These tactical
changes are feasible technically. However, I am not
optimistic that they will be adopted. The obstacle is
political: as with any bureaucratic organization, it is
not in the self-interest of the Fed to adopt policies that
would render it accountable. The Fed has persistently
avoided doing so over a long period. None of the
tactics that I have proposed is new. The proposed
changes would have made just as much sense five or
ten years ago-indeed, if adopted then, the inflation
and volatility of the past ten years would never have
occurred. The proposals have had the support of a
large fraction of monetary experts outside the Fed.
The Fed has resisted them for bureaucratic and politi-
cal, not technical, reasons. And resistance has been in
the Fed's interest. By keeping monetary policy an ar-
cane subject that must be entrusted to "experts" and
kept out of politics, incapable of being judged by non-
experts, the Fed has been able to maintain the high
public reputation of which I spoke at the outset of this
article, despite its poor record of performance.
One chairman after another, in testimony to Con-
gress, has emphasized the mystery and difficulty of the
Fed's task and the need for discretion, judgment, and
the balancing of many considerations. Each has
stressed how well the Fed has done and proclaimed its
dedication to pursuing a noninflationary policy and has
attributed any undesirable outcome to forces outside
the Fed's control or to deficiencies in other compo-
nents of government policy-particularly fiscal policy.
The testimony of the four most recent chairmen of the
Fed documents their pervasive concern with avoiding
accountability-a concern with which it is easy to sym-
pathize in view of the purely coincidental relation be-
tween their announced intentions and the actual
outcome.
Clearly the problem is not the person who happens
to be chairman, but the system.
MILTON FRIEDMAN, a 1976 Nobel laureate, is a senior research
fellow at the Hoover Institution, Stanford University,
and Paul Snowden Russell Distinguished Service Professor of
Economics at the University of Chicago. This article is adapted
from "Monetary Policy for the 1980s'' in To Promote Prosperity,
edited by John H. Moore and published by the Hoover Institution Press.
1984 by the Board of Trustees of the Leland Stanford Jr. University.
