Full text of Review (Federal Reserve Bank of St. Louis) : May 1970

View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

FEDERAL RESERVE BANK
OF ST. LOUIS
MAY 1970

Transition to Reduced In fla tio n .................... 2

EIGHTH D ISTRICT
•
* M EM PHIS
LITTLE RO CK

V o l. 5 2 , No.

Let’s N ot Retreat in the Fight
A gain st Inflation..................................... 7
Neutralization of the M oney S to c k .............. 12
—

C o m m e n t......................... T5

Transition to Reduced Inflation

( GROW TH OF TOTAL SPENDING slowed fur
ther in the first quarter of this year, following a
significant moderation late last year. This reduced
expansion of total spending has been accompanied by
a substantial decline in real economic activity with
little change in the upward trend of prices. The goal
of a reduction in the rate of increase of prices re
mains to be achieved.

The Effects of Restraint
Total spending is estimated to have increased at
only a 3.6 per cent annual rate from the third quarter
of 1969 to the first quarter of this year, markedly
slower than the 8.5 per cent rate during the previous
two years. The reduced growth of total spending since
last fall includes moderation of spending for consumer
durables and for inventories. Spending for consumer
durables, normally one of the first activities to feel
the pinch of public policy restraint on total spending,
fell slightly from the third quarter, after increasing at
a 10.7 per cent rate in the previous two years. Inven
tory accumulation slowed sharply in the first quarter
of this year in response to sluggish sales. Total non
farm inventories increased at a $0.4 billion annual
rate following an increase of $7.8 billion in 1969.

Reduction in the rate of inflation typically follows
a slowdown in the growth of total spending with a
substantial lag. It took several years to significantly
reduce each of the other postwar inflations, and the
current inflation is the strongest w e have experienced
in the last twenty-five years. As a result, the rate of
increase of prices can be expected to come down
slowly in response to a moderated expansion of total
spending. Real product will probably continue for
some time to bear most of the impact of the reduced
rate of growth of spending, but further reduction in
real economic activity seems likely to be more mild.

Prices — The GNP deflator, often used to measure
changes in general prices, showed a 6.3 per cent
annual rate of increase in the first quarter
of this year, but is not an accurate indi
cator of price trends in this instance. The
Ratio Scale

D e m an d an d Production
Ratio Scale

Quarterly Totals at Annual Rates

T rillio n s o f D o lla r s

1.1

1.0
2 9 5 9 .6

Total Spending n

+9.67

+ 5,3V

+93
^724.3

^ ~ + 2
rv

+6.1%
+4.9%

.6
' +827o

0.7%

-1.7°

+53%

|
1
Real Produ C l *

+4.6%
cr

•£
t

Istqtr.

t
1962

1963

1964

t
1965

t
1966

.fc

Page 2

1967

1968

Q G N P in current dollars.
12 G N P in 1958 dollars.
Percentages are annual rates of change for periods indicated.
Latest data plotted: Istquarter

~o
c
CN

t
1969

?
1970

Source: U.S. Department of Commerce

accounting procedure used to measure

total spending treats Government pay
raises as increases in the prices paid by
the Government sector. Thus the recent
Government pay increase, retroactive to
January, was entered into the first quarter
data as an increase in the GNP deflator.
Therefore, the general price index for the
quarter must be received cautiously.
Without the pay increase, the index rose
at a 5.3 per cent annual rate in the first
quarter, about the same rate as in the
previous year. The distortion o f the price
data caused by the handling o f the
Government pay increase is indicated
by the behavior o f the other price meas
ures in the same period. Consumer prices,
for example, rose at a 6.3 per cent
annual rate from late fall to late winter,
compared with a 5.8 per cent rate in the
preceding year. Prices of wholesale in
dustrial commodities rose at a 4.3 per

FEDERAL RESERVE BANK OF ST. LOUIS

MAY, 1970

cent rate during the previous two years. Population
of labor force age has been increasing at a 1.6 per
cent rate.
The recent slowdown of real economic activity was
inevitable if the inflation rate of the past four years
is to be reduced. Over the past few years production
increased at unsustainable rates under the pressure
of excessive growth of total spending. Real product
increased at a 5 per cent rate from mid-1967 to
late 1968, compared with an estimated 4 per cent
rate of increase in capacity. During that period the
rate of unemployment fell to 3.4 per cent of the
labor force, and shortages of skilled labor became
widespread. Employment of progressively less efficient
workers contributed to a marked reduction in growth
of labor productivity. Output per man-hour, after
increasing at a 3.6 per cent rate from 1961 to 1965
and a 2.8 per cent rate from 1965 to 1968, decreased
slightly during 1969.

cent rate from late fall to late winter, compared with
a 4 per cent rate in the previous year. It seems reason
able to conclude, therefore, that the rate of increase
of general prices did increase from the fourth quarter
to the first quarter, but not by as much as the general
price index indicated.
While care must be taken in analyzing the appar
ent acceleration of inflation in the first quarter, one
must also remain aware that the pace of inflation has
not yet moderated. Furthermore, there is little evi
dence that the rate of increase of prices will decline
sharply in the near future. Instead, the rate of infla
tion will more likely decline slowly, in response to
moderated growth of total spending.
Output — Real product fell at a 1.7 per cent annual
rate from the third quarter of 1969 to the first quarter
of this year. Industrial production declined at a 3.2
per cent annual rate from July to April. Paralleling
the decline in the growth of total spending, the de
mand for labor, especially in manufacturing, has weak
ened. Total employment was about unchanged from
January to April, compared with a 2.7 per cent rate
of increase during the previous two years. Payroll
employment has increased at about a 1 per cent an
nual rate since last fall, compared with about a 3 per

The economic slowdown since last summer has
been quite mild, however, compared to other periods
of slowdown in the last twenty-five years. The recent
1.7 per cent rate of decrease of real product compares
with a 4.7 per cent average rate of decline during
the first two quarters of other contractions. On
average, the unemployment rate increased from 4
per cent to 6 per cent in the first two quarters of
other postwar slowdowns, yet the average quarterly
rate increased from 3.6 per cent of the labor force to
4.2 per cent in the first two quarters of this slowdown.
Corporate profits after taxes have declined at about
a 14 per cent annual rate since the third quarter of
last year, markedly less than the average 30 per cent
annual rate of decline experienced in the first two
quarters of previous contractions. Most other meas
ures of economic activity have also moderated much
less than in similar periods in the past.
Interest Rates — Short-term interest rates declined
rather sharply early this year, probably reflecting
a shift in market expectations as it became apparent
that economic activity was slowing. The yield on 4to 6-month commercial paper fell from an average
of 9.00 per cent in early January to 8.03 per cent in
late March. The rate on Treasury bills declined from
8.02 per cent in January to 6.16 per cent in late March.
Subsequently, short-term rates have risen some
what. The rate on commercial paper averaged 8.33
per cent in early May and the rate on Treasury
bills averaged 6.80 per cent. Long-term interest rates
have changed little on balance since early in the year.
Page 3

FEDERAL RESERVE BANK OF ST. LOUIS

Yields on H igh e st-G rade Corporate Bonds

MAY, 1970

When the rate of growth of total spending increases
and demand for credit pushes market interest rates
above the rates savings institutions can pay on
deposits, the supply of residential mortgage funds is
restricted. Similarly, when moderation of demand for
credit pushes interest rates down toward the level
paid on time deposits, the flow of funds into time
deposits increases, and expands the supply of resi
dential mortgages.
Since there are only limited amounts of saving and
loan funds available in the economy at any one time,
shifts in the demand for funds for nonresidential con
struction have caused changes in the amount of credit
supplied for housing. The accompanying chart com
pares residential housing starts with the spread be
tween market interest rates and the ceiling rates on
time deposits under Regulation Q.3 There is a very
close correspondence between periods of weakness
in residential construction and periods when market
interest rates were significantly above the Regulation
Q ceiling rates, such as in 1966 and 1969. This sug-

on current a n d la g g e d price c h a n g e s a n d on va ria b le s thou gh t to influence " r e a l " interest rates

S p re a d Betw een
4 - to 6 -M o n th Com m ercial Pap e r Interest R a t e 1
an d R egu lation Q C e ilin g Rate 12
Per Cent
Per Cent

(i.e., the level of a n d c h a n g e s in output a n d c h a n ge s in the d eflated m o n e y stock). Se e W illia m P.
Y o h e a n d D e n is S. K a rn o sk y, "In tere st Rates a n d Price Level C h a n g e s, 1 9 5 2 -6 9 ," R eview . Fe deral
R eserve B a n k of St. Louis, De ce m b e r 1969.
Latest d a ta plotted: A p ril

\
J

P r e p a r e d l y Fe de ra l Reserve Ba n k o f St. Louis

A ,/*

The yield on seasoned corporate Aaa bonds averaged
8.04 per cent in early May, up slightly from a 7.90
per cent yield in early January. Failure of long-term
interest rates to decline as economic activity has
weakened suggests continued strong demand for
long-term funds, probably bolstered by expectations
of continuing inflation.1
In the past, interest rates have generally declined
along with economic activity, and were apparently
expected to fall with moderation of spending this
year. Never before in the postwar period, however,
have interest rates been so influenced by inflation as
since 1966. Expectations of inflation have accounted
for most of the rise in interest rates since 1966, and if
rates are to fall significantly, these expectations must
first be reduced. There is little prospect of reducing
expectations, however, until the actual rate of infla
tion begins to slow.
Trends in housing are influenced considerably by
Government regulations which set the maximum in
terest rates savings institutions can pay on deposits.2
’ See this Review (Decem ber 1969), pp. 18-38.
2See this Review (June 1968), pp. 5-12.
4
Digitized forPage
FRASER

1.754

f
■*

Z 1\

-1
Total Private H o u sin g Starts
Thousond5 of Units

S e a s o n a l ly A d ju st e d at A n n u a l Rates
Thousands of Units
_L_!—-------------------------------2000
2000------------------------r

1500

/v A

r \ r "s

K
-—

1000

1500

1181
1000

500

0
196 3

196 4

1965

1966

1967

1968

1969

197 0

So u rc e s: U.S. D e p a rtm e n t of C om m erc e a n d the B o a rd o f G o v e rn o rs o f the F e d e ra l
R eserve Syste m
Q .B o n d e q u iv a le n t yield.
12 M a x im u m interest rate o n certificates of d e p o sit o f $ 1 0 0 ,0 0 0 o r m ore, m a t u rin g in 9 0 to
179 d a y s.
Latest d a ta plotted-. A p ril

gests that with current Regulation Q ceilings, interest
rates in 1970 would have to be significandy lower
3Under the provisions of Regulation Q, the Federal Reserve
sets the maximum interest rates commercial banks can pay
on time and savings deposits. Similar regulations on savings
institutions are administered by the Federal Home Loan
Bank Board and Federal Deposit Insurance Corporation.

FEDERAL RESERVE BANK OF ST. LOUIS

MAY, 1970

than they were last year if residential construction
is to increase significantly.

sure on interest rates from increased demand for
credit.

The Government has recently been attempting to
induce large institutional investors to place more
funds in the housing market in order to encourage
residential construction. Such a program would have
little effect on total credit available in the economy.
While it could lead to more housing construction,
this plan would tend to reduce the flow of funds into
other credit markets.

Reduction of money growth tends to have the
opposite effect on interest rates. In the short run,
interest rates may rise in response to monetary re
straint. After a few months, however, the slower
growth of money causes growth of total spending to
slow and the demand for credit to moderate. The
net result is downward pressure on interest rates.

Monetary and Fiscal Actions in 1970
The rate of inflation probably cannot be reduced
rapidly, but a gradual decline is possible. Should
spending continue to grow at a reduced pace, the
rate of resource utilization will continue to fall and
contribute to slowing in the rate of inflation. If the
growth of total spending were to accelerate, how
ever, the pressure to cut back on pro
duction would ease significantly. This
would mean that the pressure to hold
back on price increases would also be
reduced.

Fiscal Actions — In January the Federal Govern
ment proposed a budget for fiscal 1971 which sug
gested that fiscal actions would remain restrictive. A
budget surplus of $1.3 billion was planned, about the
same as was expected for the current fiscal year.
Since January there have been several fiscal actions
which threaten to shift the budget for both fiscal 1970
and fiscal 1971 from surplus into deficit. The recently
enacted 6 per cent pay increase for all Federal em

The strength of total spending through
the rest of this year depends in large
part on the past and future course of
monetary and fiscal actions. The effect of
last year’s restrictive policies continues to
be a strong force in the economy, while
recent and future actions will have pro
gressively more influence as the year
proceeds.
Monetary Actions — The trend of mon
etary actions has apparently shifted in
the first quarter of this year. The money
stock rose markedly in March and April
and has increased at a 5.7 per cent rate
since December, following a six-month
period of no growth.
Resumption of monetary expansion
should not be expected, however, to pro
duce sharp and sustained declines in in
terest rates. An increase in the rate of
monetary expansion can restrain interest
rates for a short period by increasing
liquidity and the supply of credit, but
the effect is not long lasting. Increased
monetary expansion also stimulates total
spending with a brief lag, which in turn
results in higher prices, stronger expecta
tions of inflation, and in upward pres

Page 5

FEDERAL RESERVE BANK OF ST. LOUIS

MAY, 1970

ployees seems likely to increase budget expenditures
above the January estimates by $1.2 billion in the fiscal
year ending June 30 and by $2.6 billion in fiscal 1971.
The budget picture is further clouded by the
probable course of economic activity. Growth of tax
revenue normally slows in periods of business slow
down, and if the economy continues on its current
path or even picks up moderately, the Government
can expect only slow growth in revenue. This would
serve to push the budget further into deficit, unless
Federal Government expenditure growth is curbed
beyond current plans.
The prospect of a budget deficit suggests to some
observers that the Government would be contributing
to further inflation. Care must be taken, however, in
attaching too much significance to budget figures in
analyzing the effect of fiscal actions on economic
activity. The budget tends toward deficit in periods
of economic slowdown as revenue growth automatic
ally slows in response to moderation of total spending
growth in the economy. This response does not repre
sent expansionary fiscal actions or policy but reflects
the effect of slower economic expansion on the budget.
The high-employment budget was developed to
eliminate some of the misleading information about
fiscal actions which arises from actual budget figures,
and was first popularly used in the early 1960’s to
argue that fiscal actions were actually restrictive de
spite large deficits in the actual budget. It removes
most of the effect of real economic activity on the
budget and attempts to indicate only the effect of
the budget on economic activity.

which develop in response to continued moderation
of economic activity might be as large as those of
the 1961-63 period. As was true then, however, they
need not indicate a shift in the influence of fiscal
actions on the economy toward stimulus.

Growth in Government spending, on a highemployment basis, has moderated considerably from
the high rates of increase experienced in 1966 and
1967. Although the pay increase for Federal employ
ees and increased Social Security benefit payments
will cause a large increase in the second quarter of
this year, the growth in expenditures from late 1969
to late 1970 is expected to be at about the same rate
as in the previous year.

The mild slowdown experienced since last fall has
been the logical consequence of restrictive monetary
and fiscal actions. Businesses have responded to
moderation of total spending growth by cutting pro
duction, but only later will the effect of reduced
growth of spending be reflected in a slower advance
in the level of prices.

Government spending and revenue in fiscal 1971,
as proposed in January, would result in a highemployment budget surplus of $13 billion in the
second half of this year. The pay increase provisions
enacted since January are estimated to decrease this
surplus by about $4 billion in the second half, to
about the level of fiscal 1970. Actual budget deficits

There is no reason to expect sudden improvement
in the rate of inflation. Eliminating substantial infla
tion has never been a quick or painless process, and
the current inflation is the strongest we have experi
enced in the postwar period. There appears to be
little danger that the current slowdown will develop
into a full-blown recession. It appears that the danger,
if any, is in the other direction.

Page 6

Conclusions

Let’s Not Retreat
in the Fight Against Inflation
A speech by DARRYL R. FRANCIS, President, Federal Reserve Rank
of St. Louis, before the Mississippi Rankers Association Convention,
Ruena Vista Hotel, Riloxi, Mississippi, May 19, 1970

I HIS IS A W ELCOM E OPPORTUNITY to dis
cuss my view of the state of our economy with friends
of long standing, the bankers of Mississippi. As busi
ness leaders in Mississippi and in your local communi
ties, it is, of course, always important that you keep
in touch with economic stabilization efforts to pro
mote a high level of employment and relatively stable
prices. At this particular time, I want also to discuss
with you some pitfalls which could threaten the suc
cess of those efforts and defeat their objectives.
By way of background, I will examine two topics,
tracing first the development of our inflation since
1965, and next, some reasons for the extremely slow
response of inflation to monetary and fiscal restraint
of the past two years. This background is essential to
my principal point which is this — a possible threat to
the success of current stabilization actions. This threat
comes from some frequently expressed desires to
achieve several good but incompatible objectives by
year’s end — namely, a markedly lower rate of infla
tion, little further rise in unemployment, and a sharp
reduction in market interest rates. I say actions to
accomplish these short-run objectives constitute a
threat because attaining any one of them would re
quire extreme monetary actions, leading to later con
ditions quite contrary to desired policy objectives.
Moreover, these near-term objectives cannot be
achieved simultaneously.
In developing the background topics and outlining
the possible impediments to achieving current policy
objectives, my remarks will draw heavily on recent
research at the Federal Reserve Bank of St. Louis.
For the past two years our economists have been at
tempting to quantify the response of total spending,
real output, the price level, the unemployment rate,
and market interest rates to monetary and fiscal ac
tions. Monetary actions in this research are measured
by changes in the nation’s money stock —that is, de
mand deposits and currency held by the nonbank

public. Fiscal actions refer to changes in spending and
taxing provisions of the Federal Government budget.
One important conclusion suggested by these stud
ies is that actions of the Federal Reserve which change
the rate of monetary expansion exert a relatively
quick and pervasive influence on total spending, and
changes in Federal Government expenditures rela
tively less, unless accompanied by accommodating
changes in the money stock. Changes in Federal tax
ing provisions are found to have an insignificant influ
ence on total spending.

Current Inflation
I turn now to my first background topic — an ex
amination of our inflation since 1965. After six years
of relative price stability from 1958 to 1964, we have
since experienced accelerating inflation. The general
price level rose at a three per cent annual rate from
late 1965 to mid-1967, then at a four per cent rate to
the end of 1968, and finally, during the past five
quarters, at a five per cent rate. The inflation rate
shows few signs of abating up to now.
This five-year record of accelerating inflation re
sulted from the pressure of total spending on the
ability of our economy to produce goods and services,
particularly since early 1966. From the first quarter
of 1966 to mid-1968, total spending rose at a 7.5 per
cent annual rate, while output of goods and services
grew at about a four per cent rate, or approximately
the rate of growth of the economy’s productive poten
tial. At full employment of our resources, expansion
of real output depends on growth in the labor force,
capital plant, and technology. In recent years these
factors have fostered growth of production potential
at about a four per cent annual rate.
By 1968 and 1969, inflation had developed a very
strong momentum which has complicated greatly the
problem of reducing the rate of increase of prices.
Page 7

FEDERAL RESERVE BANK OF ST. LOUIS

This momentum is the result of households, busi
nesses, and labor unions attempting to protect their
economic positions by building anticipated price in
creases into contracts for goods, services, and loans.
In this manner, the “demand pull” inflation of 1965
to 1968 was subsequently changed into “cost push”
inflation. I want to point out, however, that excessive
total spending was the basic cause of our present
inflation problem, and that the so-called cost push in
flation is also a result of earlier excessive total demand.
Where did the excessive increase in total spending
come from? Mainly it was a result of overly expansive
monetary actions. The money stock increased from
April 1965 to April 1966 at a six per cent annual rate,
at that time the fastest rate since the inflationary
period of the Korean War. Following 1966 when the
money stock remained unchanged for eight months,
money grew at a seven per cent rate during 1967 and
1968, the most rapid rate since W orld War II.
That period when the money stock remained
unchanged during the last eight months of 1966 set
the stage for curbing inflation. This could have led to
a balanced rate of spending if it had not been fol
lowed by resumption of expansion in money at a very
rapid rate in 1967 and 1968. Our studies indicate that
if expansion in money had been maintained at a
moderate four per cent rate instead of the seven per
cent rate actually recorded in 1967 and 1968, the rate
of inflation since late 1966 most likely would not
have surpassed 3.5 per cent, instead of reaching five
per cent as it did last year. Moreover, if the four per
cent growth in money had been maintained up to
the present, the rate of inflation would be receding,
and if that moderate rate of monetary expansion
were to be continued through 1972, price increases
would be down to about a 1.5 per cent rate by the
end of that year.
Excessive total spending has not only been the
cause of price inflation but also of the great increase
of market interest rates during the past four years.
Our research indicates that market interest rates are
highly responsive to anticipated price changes. Past
increases in the price level, such as those during the
last five years, cause participants in the money and
capital markets to expect a continued high rate of
inflation. An inflationary premium is thus built into
market interest rates. W e attribute almost all of the
sharp rise in market interest rates since 1966 to an
accelerating inflation fostered by excessive monetary
expansion.
As was the case with the general price level, the
monetary restraint of 1966 set the stage for lower

Page 8

MAY. 1970

interest rates. Our studies indicate that a moderate
four per cent growth in money from the end o f 1966
to the end of 1969 would have produced a peak in
short-term interest rates, as measured by the rate on
four- to six-month commercial paper, of around 5.5 per
cent, and these interest rates would have been about
4.5 per cent this spring, instead of the present eight
per cent or more. Further continuation of this mod
erate growth in money would have produced short
term interest rates heading to below four per cent by
late 1972. Long-term interest rates would have moved
in a similar manner. With a four per cent growth in
money, seasoned corporate Aaa bond rates would have
probably peaked at about 6.25 per cent, would likely
have been about 6 per cent this spring compared
with the actual level of almost 8 per cent, and would
be moving to about 5 per cent in late 1972.
It must be evident to everyone that our failure to
take advantage, during 1967-1968, of the eight months
of restraint in 1966 was a golden opportunity lost.
Had the period of restraint been followed by a mod
erate, instead of rapid monetary expansion, the many
economic dislocations caused by the continuation of
high and accelerating rates of inflation after 1966
could have been prevented. Commercial banks and
savings institutions could have done very well with
short-term market interest rates not in excess of 5.5
per cent, as these institutions would not have under
gone the problems caused by the disintermediation of
the last three years.
Furthermore, the housing industry would have been
in much better condition throughout this period. Labor
contract negotiations today would have been less
acrimonious and disruptive. And, of course, the whole
of society would have benefited by a lesser rate of
inflation.
A logical question to be asked is, “W hy was this
opportunity to control inflation lost?” The published
record and statements of prominent economists in
dicate several reasons. First, there was the mistaken
belief at the time that easing actions of monetary
authorities could prevent increases in market interest
rates in the short run or, as some argued, actually
lower them permanently. Such actions were deemed
desirable in order to shelter savings institutions and
the housing industry from market forces set in motion
by the excessive total spending. Second, many argued
that monetary actions, as indicated by changes in the
money stock, have little influence on total spending.
As a consequence, those holding this view were little
disturbed by the exceedingly rapid growth in the
money stock. Third, in contrast with the previous

FEDERAL RESERVE BANK OF ST. LOUIS

view, many believed that rapid growth in money
was desirable in early 1967 to avoid an anticipated
recession. Finally, the national debt was increasing,
and it was thought desirable by many that the Fed
eral Reserve “even keel” the money markets at times
of Treasury financings.
All of these reasons have proven to have been
spurious. The resumption of rapid monetary growth
in 1967 and 1968 gave us higher interest rates, not
lower; less funds for housing, not more; greater
strains in the financial markets, not less; and more
difficulty with managing the Federal debt, not less.

Slow Response to Recent
Stabilization Actions
With inflation mounting, restraining actions have
been adopted since mid-1968, but the response of
inflation has been agonizingly slow. People naturally
ask why. The answer is fairly simple — as a result of
avoiding monetary actions to curb inflation until 1969,
an inflationary momentum was allowed to develop.
As a result, the general price level has continued to
rise rapidly up to the present time, and market in
terest rates remain near their extremely high levels
of late 1969. This is the legacy of the excessive total
spending from 1965 to 1968, which requires more
restraint and patience to overcome now that inflation
is moving under its own momentum.
As a step toward restraint, monetary expansion was
reduced to a four per cent rate during the first half
of 1969. Further restraint was applied in the second
half of 1969 when there was no growth in money.
The impact of such monetary actions has fallen pri
marily on total spending and real output of goods
and services and not, as yet, to any appreciable ex
tent, on the price level.
Some have begun to question whether monetary
restraint will result in slower growth in the price level
in a reasonable period of time. But our research in
dicates that a marked move to monetary restraint,
such as we had in 1969, generally slows total spend
ing with only a two- to three-quarter lag, and this
was the case in 1969. Such a change in the rate of
growth of total spending is accompanied by a simul
taneous decrease in the rate of growth of output. And
so it was in the last year. It is not until a further
two or three quarters that prices respond appreciably
to the slower growth in spending. So we should not
have expected price restraint in 1969. The course of
the price level depends not only on total spending
but also on anticipated price movements. The greater
the anticipated rise in prices, the longer delayed is

MAY, 1970

the response of the price level to monetary restraint.
This is what we mean by the problem of inflationary
momentum.
So here we are again in 1970, with the stage set
for reducing the rate of price increase, just as was
the situation at the beginning of 1967. But 1970 is
not exactly like 1966: inflation has built up a longer
and stronger momentum since then. Consequently, it
is more difficult to curb inflation this time; and the
public, as well as economic policy makers, must be
patient in waiting for the results of monetary restraint
to appear.
Many have become concerned that the extreme
monetary restraint of 1969 may result in excessive
retardation of economic growth and have recom
mended a resumption of monetary expansion. I, too,
share these concerns, and I favor a moderate rise in
the money stock. W e should avoid, however, a repeat
of the 1967-1968 experience when concern over a
possible recession was one of the major bases for ex
cessively stimulative monetary actions. This effort will
take time — longer than it would have taken if pur
sued to completion following 1966. Now, as many as
three more years will probably be required for the
rate of price advance to fall below two per cent,
assuming a moderate rate of growth in the money
stock.
While moderate growth in money will reduce price
increases to a tolerable rate by late 1972, this achieve
ment will not be without some transitional costs. Dur
ing the next three years, growth of real output would
remain below the economy’s productive potential,
and, as a result, the unemployment rate would con
tinue to increase. If our measurements of the re
sponse of prices and unemployment to stabilization
actions are reasonably correct, and I believe they are,
the excesses of 1965 to 1968 cannot be corrected with
out temporary costs in terms of lost output and
employment opportunities.

Some Threats to a Successful Fight
Against Inflation
I turn now to my final subject — some possible
threats to a successful fight against inflation. Many
may not be satisfied with the price level, output,
unemployment, and interest rate movements between
now and late 1972 that I have just indicated are
likely to follow from a moderate rate of monetary
expansion. Many recommend that present stabiliza
tion actions be altered so that in 1970 the rate of
inflation be reduced to below four per cent. Others
argue that the unemployment rate should not be al
Page 9

FEDERAL RESERVE BANK OF ST. LOUIS

lowed to reach five per cent this year. Some propose
that market interest rates be reduced markedly in the
near term. It is argued that once these immediate
objectives have been achieved, moderate monetary
growth can safely be resumed in 1971 and 1972.
But these desired accomplishments are not mutu
ally compatible. To achieve any one of them this year,
we are probably not willing to consciously pay the
costs in terms of the other two. In addition, achieving
any one of these short-run objectives may set in mo
tion forces which would lead to unacceptable con
sequences at the end of two or three years.
I have already indicated that a policy of moderate
four per cent rate of monetary expansion during the
next three years will most likely produce reasonably
stable prices by late 1972, along with lower interest
rates. Let us now examine the implications for late
1972 of alternative monetary policies over the balance
of this year which would be designed to achieve the
three short-run objectives I have just outiined. In
each case, I will assume, after 1970, a four per cent
rate of growth in money. Given the existing inflation
ary momentum, extreme monetary actions in terms of
growth in money would be required to achieve any
one of the three objectives by the end of this year.
Let us first examine the proposal that the rate of
inflation be reduced below four per cent by the end
of this year. Many have actually forecast a rate of
price increase in the 3.5 to 4 per cent range. In
order to accomplish this objective — a rate of inflation
below four per cent — the money stock would have to
be decreased at about a four per cent rate from the
first to the fourth quarter. The price situation would
be very good in 1972, when the price level would be
rising very slowly. Such an action would result now,
however, in an extremely severe recession. Output
would probably decrease sharply during the next five
quarters, and the unemployment rate would be mark
edly higher in 1972 than now. In my opinion, the
employment and output costs of attaining rapid price
level restraint in 1970 would be far too high for it
to be given serious consideration.
The next short-run proposal to be examined is the
one calling for actions to avoid further recession and
to hold the unemployment rate below five per cent
during the remainder of this year. This proposal is
based on the same kinds of fears of a recession as, in
early 1967, led to a high rate of growth in the money
stock. Accomplishment of this objective, according to
our studies, would require a ten per cent rate of
monetary expansion during the last three quarters of
this year.
Page 10

MAY, 1970

Such a course of monetary action would provide
little reduction in the rate of price advance this year
and a rate of inflation still in excess of three per cent
in late 1972. It could be said that this would be very
slow progress in curbing inflation, and I would agree.
This course of monetary expansion would result in
only a temporary spurt of growth in real output. By
1972, as a result of the shift back to a moderate rate
of monetary expansion, real output would be growing
at about half the increase in full employment poten
tial. Consequently, the unemployment rate would
most likely increase to above 5 per cent by late 1972.
Finally, I would like to consider the possibility of
achieving a sharp and immediate reduction in market
interest rates. Such an objective has been suggested,
just as in 1967 and 1968, in order to help savings
institutions and the housing industry. With respect to
long-term interest rates, because the inflation pre
mium incorporated in them is so great, the rates
could be affected only slightly by year’s end even with
extremely rapid monetary expansion. Furthermore,
if rapid monetary expansion were used to reduce long
term rates this year, these rates would remain at
relatively high levels through 1971 and into 1972.
With respect to short-term rates, we may expect some
declines this year if money supply increases only
moderately. More rapid monetary expansion could
bring slightly greater declines, but at the expense of
higher rates in 1971 and 1972.
Pursuit of such an interest rate policy would result
in no headway in controlling inflation this year and
only slight improvement by 1972. As a result of the
continuing high rate of inflation, short-term interest
rates would soon return to their present levels, or
higher, and long-term rates would rise further from
their present levels. The year 1972 would still be one
of high interest rates. But that is not the whole pic
ture; the shift back to a moderate rate of money
growth after this year would result in very slow in
creases in output in 1972 accompanied by a rising
unemployment rate.
The preceding analysis suggests several implica
tions. First, given the existing momentum of inflation,
relatively stable prices cannot be achieved in a short
period of time, unless we are prepared to accept
very high costs in terms of reduced output and em
ployment. Second, monetary actions in 1970 to achieve
the short-run employment and interest rate objectives
mentioned are self-defeating over the longer run.
Third, delaying moderate monetary expansion until
after the end of this year, in order to achieve these
unemployment and interest rate objectives, would

FEDERAL. RESERVE BANK OF ST. LOUIS

seriously impede efforts to curb inflation within the
next three years. Finally, if we are to contain infla
tion, there will be accompanying output and employ
ment costs. Such costs can be postponed this year
by high growth rates in money, but they cannot be
avoided if we are ever to achieve relative price
stability.

Conclusion
In conclusion, it is my opinion that the current
resumption of monetary expansion be kept moderate
and maintained for at least the next three years. Such
a course, in my view, is optimal — it would produce
relative price stability by 1972 without incurring as
high a cost in terms of output and employment as
would a more restrictive course of action. Although
unemployment would rise, this problem in the long
run cannot be treated by monetary and fiscal policy

MAY, 1970

and should be treated by other means. For example,
better approaches to ameliorate unemployment would
be to remove the many impediments to the free func
tioning of our labor markets, to improve the mobility
of our labor force, and to upgrade the skills of the
disadvantaged.
As at the beginning of 1967, the stage is now set
for achieving relatively stable prices. Let us firmly
resolve to seize the opportunity. Let us further resolve
that our patience will be equal to the time required.
Above all, let us not throw away this opportunity for
achieving price stability, as we did a few years ago.
If we do, not only will our efforts to date go for
nothing, but the battle against inflation will be more
difficult and more costly the next time we attempt
to make a stand. So this time, let’s not retreat in the
fight against inflation.

Page 11

Neutralization of the Money Stock
by PATRIC H. HENDERSHOTT *

T h e AUGUST 1969 ISSUE of this Review con
tained three papers dealing with the adequacies of
the observed money stock as an indicator of Federal
Reserve policy actions. In the first paper1, Emanuel
Melichar asserted, on the basis of my analysis2, that
the money stock is an inaccurate measure of policy
actions. He suggested as an alternative my neutral
ized money stock —the observed money stock after
removal of the impact of the business cycle. In the
second paper3, Michael Keran argued that observed
money is a better indicator than neutralized money
because the Federal Reserve offsets the impact of
the business cycle on the money stock.4 Finally, in
the third paper,5 Leonall Andersen examines em
pirically the argument that the money stock is influ
enced by the business cycle. He concludes that it
is not.
In this short note, I first consider Keran’s theoretical
argument against neutralizing the money stock.
Keran’s argument is a familiar one that I had hoped
“ Patric Hendershott is Associate Professor of Economics and
Management at Purdue University.
1Emanuel Melichar, “ Comments on the ‘St. Louis Position,’ ”
this Review (August 1969), pp. 9-14.
2Patric Hendershott, The Neutralized M oney Stock: An Un
biased Measure of Federal Reserve Policy Actions (Richard
D. Irwin, Inc., Homewood, Illinois, 1968).
3Michael Keran, “ Reply,” this Review (August 1969), pp.
15-18.
4He also mistakenly criticizes Melichar for using the neutral
ized money stock as a measure of the impact of all monetary
influences on the economy. There is no basis for Keran’s
criticism. Melichar states explicitly in numerous places in
his paper that the neutralized money stock is used as a
measure of monetary policy actions only ( see particularly
pp. 11-12). Further, Melichar’s use of a policy, rather than
a total monetary, measure is appropriate because Rowsher
and Kalish, in the paper that induced Melichar’s response,
were quite straightforward in their identification of changes
in the rate of change in the money stock with changes in
Federal Reserve policy actions. See “ Does Slower Mone
tary Expansion Discriminate Against Housing?,” this Review
(June 1968), pp. 5-6.
BLeonall Andersen, “Additional Empirical Evidence on the
Reverse-Causation Argument,” this Review (August 1969),
pp. 19-23.
Page 12

my book would put to rest. I then point out that the
results of Andersen’s empirical work are neither in
consistent with my results nor very surprising. Ander
sen defines his monetary policy variable so broadly
that there is scarcely any room left for an endogenous
money stock.

Keran s Critique of the Neutralized
M oney Stock6
Keran’s principal argument is that money is an
exogenous variable controlled by the Federal Reserve,
not an endogenous variable, and thus that it is the
best measure of Federal Reserve policy actions. He
seems willing to acknowledge that the banking sys
tem’s demand for free reserves, foreigners’ demand
for the U.S. gold stock, and the public’s demand for
commercial bank time deposits are all negatively re
lated to U.S. interest rates. He contends that money
is not endogenous because the Federal Reserve acts
to offset the impact of these responses on the money
stock.
6I want to correct one exception to Keran’s otherwise quite
accurate summary of my work. Keran asserts (p. 16) that
I constructed the “ modified-neutralized” money stock, which
implicitly treats gold flows as if they were offset by Federal
Reserve actions, because the Federal Reserve likely offsets
such flows. My real reason for calculating this is clearly
stated in the paragraph immediately preceding the figure
containing the modified-neutralized money stock:
The appropriateness of a comparison o f Federal Reserve
policy actions with the expressed intent o f policymakers
depends on whether the definition of policy actions em
ployed is the same as that which the policymakers had in
mind when they discussed their actions and issued
directives.
. . . Since it is likely that policy statements refer to
actions net of offsetting gold movements, a modifiedneutralized money stock, which differs from the neutral
ized money stock in that the impact of the business cycle
is not removed from the gold reserves component, is
calculated and compared with the expressed intent of
policymakers (p . 132).
(A footnote is attached to this paragraph pointing out that a
“ neutral” policy was defined in the FOM C minutes as staying
out of the market after offsetting gold flows.) Thus, Keran’s
statement (p. 16) that there “ is no reason why Hendershott
should have stopped with allowing only for offsetting actions
with respect to gold” is incorrect.

FEDERAL RESERVE BANK OF ST. LOUIS

The concept of an endogenous money stock can be
formalized as follows. The money stock depends upon
the actions of the Federal Reserve, denoted by the
vector MP, and a vector of interest rates R:
(1 )

M = f(M P ,R ),

where MP is defined so that increases in it lead to
increases in M. Since an increase in interest rates
leads to increases in free reserves and time deposits,
which are uses of reserves, and to a decline in the
U.S. gold stock, which is a source of reserves, money
is unambiguously related to R in a positive manner.
The MP vector includes a source-of-bank-reserves
variable, legal reserve requirements, the discount rate,
ceiling rates on time and saving deposits, and other
selective controls sometimes employed. In my book
I treated the Federal Reserve’s portfolio of govern
ment securities plus various minor reserve compo
nents as the source-of-bank-reserves variable.7 Here I
adjust this variable for changes in legal reserve re
quirements, denote it by P*, and substitute P° for
MP, thereby capturing the principal monetary policy
instruments in one variable:
(2 )

M = g (P * ,R ).

Let us make the unlikely assumption that the Fed
eral Reserve always varies P° so as to offset exactly
the impact on M of changes in R. For example, if R
falls, the Federal Reserve raises P* by precisely
enough to hold money constant. This would, indeed,
remove money from the class of endogenous varia
bles.8 But it would hardly make the money stock an
accurate indicator of Federal Reserve actions. In fact,
we have explicitly assumed, following Keran, that
every time interest rates change, the Federal Reserve
takes actions that, on net, are not reflected in the
money stock. And these actions are quite interesting.
Since interest rates have tended to fall just prior to,
or concurrently with, the onset of U.S. recessions,
Keran implicitly admits that the Federal Reserve has
taken expansive actions at this crucial juncture of the
cycle. Moreover, because the money stock is un
changed, his position forces him to conclude that the
Federal Reserve is essentially doing nothing.9 Since
7The minor reserve components are those Keran denoted by
Ct and O.
8The endogenous tendency of money would, of course, still
remain. That is, if the Federal Reserve ceased to follow its
offsetting policy, money would behave endogenously.
°Keran’s views are quite similar to those expressed by Cul
bertson in “ Reply,” Southern Economic Journal, (April
1963), pp. 330-35. For a detailed critique of Culbertson’s
position, see Hendershott, pp. 99-102.

MAY. 1970

only those actions over and above the offsetting ones
are attributed to the Federal Reserve, this procedure
is clearly biased toward an unfavorable interpreta
tion of anti-recession policies. In contrast, I have
argued that all Federal Reserve actions should be
credited to the monetary authority.10 Thus, in order
to obtain an unbiased measure of Federal Reserve
policy actions, I removed the impact of the business
cycle from the money stock, leaving a series whose
cyclical movement reflects only Federal Reserve ac
tions (and other exogenous forces).
For an illustration of the implications of Keran’s
view, consider a business recession that leads banks
to sell securities to the public and repay its borrowing
from the Federal Reserve. Since the public gives up
deposits in this exchange with banks, the stock of
money declines. If the Federal Reserve offsets this
decline by purchasing securities (in particular, by
purchasing the securities the banks wish to sell,
thereby preventing interest rates from rising and
money demand from falling), Keran would interpret
the Federal Reserve as doing nothing; if the Federal
Reserve does nothing, Keran would interpret it as
taking restrictive actions; if the Federal Reserve off
sets only part of the decline by purchasing a portion
of the securities banks are selling, Keran would inter
pret it as selling securities.
Keran’s defense of the observed money stock as
the best indicator of Federal Reserve policy actions
is very reminiscent of the argument of those who use
observed interest rates as the indicator of policy ac
tions. The latter would view a decline in interest
rates during a recession as indicative of an easy
monetary policy, even if the Federal Reserve were
partially offsetting a decline in private security sup
ply by selling securities. Keran views a decline in
the money stock during recessions as indicating re
strictive actions, even if the Federal Reserve were
partially offsetting a decline in bank demand by
purchasing securities. The views are, of course,
equally erroneous.11
Perhaps an analogy with fiscal policy will make
my argument even more compelling. Say that the
lnHendershott, pp. 93-99.
1'In addition to the discussion in The Neutralized Money
Stock, pp. 1-5, see Patric Hendershott and George Horwich,
“ Money, Interest, and Policy,” Institute Paper No. 250,
Krennert Graduate School of Industrial Administration,
Purdue University Qune 1969), pp. 21-23 and 29-31.
(This paper was presented at the U.S. Savings and Loan
League Conference on Saving and Residential Financing
in May 1969 and will be published in the proceedings
of the conference.)
Page 13

FEDERAL RESERVE BANK OF ST. LOUIS

MAY, 1970

Federal Government raised tax rates during reces
sions in order to maintain a constant, balanced budget.
W ould we view this fiscal policy as being contrac
tionary or not? The “old view” is that since the budget
is still balanced, policy must be neutral. The “new
view,” which is based on the “high-employment budget
surplus” concept and to which the St. Louis Federal
Reserve Bank subscribes,12 says that policy is restric
tive because the Government actually raised tax rates.
If we were to apply Keran’s analysis, we would be
led to the old view. That is, Keran would “not count”
the increase in tax rates because it was an automatic
offsetting response to the decline in tax receipts ac
companying the recession. Thus, the fiscal policy of
raising tax rates during recessions would be inter
preted as a neutral policy with respect to the business
cycle.
As I pointed out in my book, the neutralized
money stock measure of monetary policy is analogous
to the full-employment budget surplus measure of
fiscal policy; the impact of the business cycle is
absent from both. To accept one measure and not
the other is inconsistent and, I suspect, quite reveal
ing of one’s biases.

Andersen and Endogenous M oney
Andersen has taken a quite narrow view of the
endogenous money stock concept. In particular, he
views the money stock as being related to a monetary
policy variable and gross national product ( GNP ). In
light of the free-reserves, gold, and time-deposits
12See Federal Reserve Bank of St. Louis Review
1 9 6 9 ), p . 4.

14
Digitized for Page
FRASER

(August

responses noted above, the money stock should be
related to a policy variable and interest rates. And
such a distinction is important. For example, I con
cluded that GNP has had only a small impact on
money, where interest rates have had a large impact.13
In addition to relating the money stock to the
“wrong” endogenous variable, Andersen defines the
monetary policy variable so broadly that his inability
to estimate successfully an endogenous money stock
relation is hardly surprising. In particular, the two
reserve components that 1 found to be primarily re
sponsible for the strong stock-interest rate relation —
member bank borrowings from the Federal Reserve
and the U.S. gold stock — are treated as policy-de
termined by Andersen.14 The only interest rate rela
tions that Andersen allows are the admittedly weak
excess-reserves relation and a stronger time deposit
relation which has, however, only a small impact on
the money stock.
In conclusion, Andersen’s empirical estimates are
based on a model which, by choice of the policy and
endogenous variables, rules out the expected money
stock links to the economy. Thus, the estimates should
not be interpreted either as a criticism of my work or
as an adequate treatment of the subject.

13Hendershott, pp. 140-41.
14Hendershott, p. 117 and Keran, p. 16. Subsequent dis
covery of a computational error in the neutralization of the
money stock reveals that the gold relation was not as
strong as initially believed. For a discussion of the error
and an analysis of the correctly neutralized money stock,
see Patric Hendershott, “ A Quality Theory of Money,”
Nebraska Journal of Economics and Business (Autumn
1969), or Hendershott and Horwich, pp. 25-28.

The Comment to this article begins on the next page.

Neutralization of the Money Stock —Comment
by M ICHAEL W . KERAN

ENDERSHOTT raises some interesting issues
with respect to my earlier critique of his neutralized
money stock concept. Before considering the specific
issues he raises, it would be useful to define certain
relevant terms. This will allow us to more sharply
focus the debate with respect to areas of agreement
and disagreement.

Terminology
1) Federal Reserve A ctions—A most comprehen
sive measure of Federal Reserve actions are changes
in Federal Reserve holdings of government securities
adjusted for changes in reserve requirements. Hendershott calls this ( P ° ), and uses this as his measure
o f Federal Reserve policy actions. This measure simul
taneously captures the two major policy instruments
of the Federal Reserve: open market operations, and
changes in reserve requirements. Other Federal Re
serve policy instruments, such as the discount rate,
are generally assumed either to move in line with
( P ° ), or to be of lesser importance.
2) Defensive and Dynamic Operations — Federal
Reserve actions ( P ° ) , as described in ( 1) , can be
divided into defensive operations and dynamic op
erations. Defensive operations are those Federal Re
serve actions which are designed to prevent undesired
changes in member bank reserves (or some other
intermediate financial target) as a result of changes
in factors not under the control of the Federal Re
serve, such as offsetting gold flows or changes in Fed
eral Reserve float. Dynamic operations are Federal
Reserve actions designed to change the desired level
of member bank reserves ( or some other intermediate

financial target) in response to changes in monetary
policy. As dynamic operations represent the Federal
Reserve economic policy actions, they are assumed to
vary systematically over the business cycle.
3)
Monetary Influence on the Economy — This
should reflect the net impact of all monetary influ
ences on the real sector of the economy, that is, em
ployment, income, and prices. This influence may or
may not be under the dominant control of the Federal
Reserve, depending upon how the Federal Reserve
has actually operated. The appropriate measure of
this monetary influence depends upon the linkage
between monetary variables and the rest of the econ
omy. One’s concept of these linkages depends upon
one’s assumptions about economic behavior. Keyne
sian income-expenditure theory has typically meas
ured this influence by market interest rates, while the
modem quantity theory has typically measured it by
changes in the money stock.
Hendershott’s concern is with the first point. He
wishes to construct an unbiased measure of Federal
Reserve policy actions. This is a useful exercise in its
own right, but it cannot be considered as providing
evidence or insight into point (3 ) listed above. The
best measure of Federal Reserve policy actions is not
necessarily the best measure of monetary influence
on the economy. This second question requires a
separate theoretical and empirical verification which
is not attempted either by Hendershott or by me in
this article.1
'T he author has considered the question of monetary influ
ences on the economy in other articles. See Michael W .
Keran, this Review, November 1969 and February 1970.
Page 15

FEDERAL RESERVE BANK OF ST. LOUIS

Areas of Agreement and Disagreement
Hendershott points out that I am inconsistent in
rejecting the neutralized money stock as a measure
of monetary policy actions when I have presumably
accepted the principle of neutralization with respect
to government spending and tax receipts as a measure
of fiscal policy actions. He then attempts to demon
strate that the error one makes in analyzing policy
actions of the Federal Reserve without a neutralized
money stock is of the same character as in analyzing
policy actions of the Federal Government without
“neutralized” receipts and expenditures.

I do accept the conceptual desirability of a neutral
ized money stock as an unbiased measure of Federal
Reserve policy actions, with the previously mentioned
caveat that this should not be considered as providing
any information with respect to monetary influences
on the economy. My disagreement with Hendershott
is on the empirical relevance of his particular neu
tralization process.
I tried to make that point in my original article
when I said, “. . . what if open market operations had
not been conducted in a way to offset the influence
of borrowings and gold on the money stock? In that
case Hendershott’s neutralized money stock would
have been a superior measure of Federal Reserve
(policy) actions.”2
My disagreement with Hendershott is with respect
to the interpretation of Federal Reserve actions, ( P° ) .
I assert that this is a measure of both policy actions
and non-policy actions related to offsetting non-controlled sources of member bank reserves. Stated in a
somewhat different way, the Federal Reserve engages
in both defensive operations and dynamic operations,
and only the latter should be considered as policy
actions. Hendershott, on the other hand, asserts that
( P * ) is an appropriate measure of “just” policy
actions.
The issue which separates us is not theoretical but
empirical in nature. As such, it is subject to standard
statistical tests. In the original article I presented such
a statistical test.3 It indicated that a substantial share
of the changes in P ' (I used the symbol SA) could
be explained by defensive operations designed to
offset influences on member bank reserves from
changes in non-controlled sources of reserves.
2Michael Keran, “ Reply,” this Review (August 1969), p. 17.
3lbid., p. 17.
Page 16

MAY, 1970

If the Federal Reserve had not acted in this sys
tematic way to offset non-controlled sources of re
serves, then the question of whether the neutralized
money stock was a superior measure of Federal Re
serve policy actions would depend on how well Hen
dershott’s explanation of public influences on these
non-controlled sources of reserves stood up under
critical analysis. Because Hendershott’s results had
not passed the first test, I did not examine his results
in detail to see whether they had passed the second
test.

Defensive versus Dynamic Operations
Hendershott argues that the public, through its in
fluence on market interest rates, will influence certain
sources of member bank reserves (specifically gold
flows and member bank borrowing), and through
this the observed money stock. When this public
influence is estimated and removed, we have in the
neutralized money stock an unbiased measure of
Federal Reserve policy actions. My position is that
whatever the cause of the gold flow or changes in
member bank borrowing, the Federal Reserve has
acted to systematically offset their influence on mem
ber bank reserves through the standard and long
standing procedure of defensive operations. On the
basis of empirically verifying the existence of defen
sive operations to offset the influences of gold and
borrowings on member bank reserves, I asserted that
the observed money stock is superior to the neutral
ized money stock as a measure of Federal Reserve
policy actions.
Hendershott does not question the statistical re
sults presented. On what basis then could he continue
to press this position that P° measures Federal Re
serve policy actions? He must assume that Federal
Reserve defensive behavior is systematically different
during periods when the Federal Reserve is following
a tight money policy than during periods when it is
following an easy money policy. That is, when the
Federal Reserve wishes to follow an easy money pol
icy it will not engage in net defensive operations
which would tend to reduce total reserves of member
banks. When the Federal Reserve is engaged in a
tight money policy, it would not engage in net de
fensive operations which increase total reserves.

Statistical Tests
Only if Federal Reserve defensive operations are
systematically different between periods of tight
money and periods of easy money can Hendershott

FEDERAL RESERVE BANK OF ST. LOUIS

assert that all Federal Reserve actions ( P*) be
evaluated in a uniform way rather than being divided
into defensive and dynamic operations. Two tests of
this possibility are made. The first is to compare
Federal Reserve defensive operations during periods
of tight money policy with defensive operations dur
ing periods of easy money policy. A “Chow” test will
tell us whether the data for these subperiods were
drawn from different behavior populations.
The second test focuses on gold flows and member
bank borrowings from the Federal Reserve. Because
these are the major factors causing the neutralized
money stock to deviate from the actual money stock,
it is desirable to see if the value of these defensive
coefficients changes between periods of tight and
easy money policy.
To make the tests in such a way as to provide
the greatest chance to validate Hendershott s position,
periods of tight money policy and easy money policy
are constructed according to the breakdown given by
Hendershott.4
Tight Periods

Easy Periods

April 1955 - Nov. 1957
Aug. 1958 - March 1960
Jan. 1962 - Oct. 1962

Aug. 1953 - April 1955
Nov. 1957 - Aug. 1958
March 1960 - Jan. 1962
Oct. 1962 - Dec. 1964

The results of the first test are presented in Table I.
The dependent variable is Federal Reserve holdings
of Government securities adjusted for a change in
reserve requirements ( A P ° ). The independent varia
bles are all other factors which influence member
bank reserves and are not directly controlled by the
Federal Reserve. If the Federal Reserve engaged in
defensive operations, the sign of the gold, float, and
borrowing coefficients would be negative, and the
sign of the currency coefficient would be positive. The
sign of the “other” coefficient is indeterminant.5
The first column shows the estimated coefficients
for defensive operations in tight money periods, as
designated by Hendershott, and the second column
shows the estimated coefficients for his designated
easy money periods. The Chow test, which is designed
to test for a shift in the structure between these pe
riods, was not significant at the 95 per cent level of
confidence. There is no statistical evidence that the
4Patric Hendershott, “ A Quality Theory of Money,” Nebraska
Journal of Economics and Business (Autumn 1969).
5This is because it is a combination of sources and uses of
member bank reserves.

MAY, 1970

T a b le

FEDERAL RESERVE D E F E N SIV E O P E R A T IO N S
( M o n t h ly C en tral D ifferences — B illio n s of D o lla rs)
A P * = OCo “b CCi A G ~ \~ OC2 AF + OC;{A B + OC4 A O + CX5 A C 0
T ight M o n e y
P eriods
G o ld

(A G )

Float ( A F )
B o rro w in g s ( A B )
O th e r ( A 0 )
C urren cy in H a n d s
of Public ( A C o )
C o n sta n t Term
R2
D -W

E a sy M o n e y
P eriods

.53
(2 .7 2 )
.0 9
(.4 4 )
— .63
(2 .8 8 )
.95
(3 .1 3 )

.58
(3 .2 2 )
— .2 7
(1 .6 8 )
— .91
(2 .9 5 )
.26
( 71)

1.22
(8 .6 1 )
.03
.81
1.81

1 .2 7
(1 0 .8 8 )
.01
.68
1 .3 6

N o te : Regression coefficients are the top fig u re s; their “ t” statistics
appear below each coefficient, enclosed by parentheses. R 2 is
the per cent o f variations in the dependent variable which is
explained by variations in the independent variable. D -W is
the Durbin-W atson statistic.

Federal Reserve conducted defensive operations dif
ferently during periods of easy money than during
periods of tight money.6
Even though there was no shift in the general
structure of Federal Reserve behavior between sub
periods of tight and easy money, it is possible that
Federal Reserve actions with respect to particular
variables could have changed between subperiods.
Because gold and member bank borrowings were
found by Hendershott to dominate the difference
between the actual and the neutralized money stock,
it is important to see whether there was a shift in the
value of these coefficients between periods of tight
and easy money. This is done in the second test, and
the results are presented in Table II.
The second test estimates the value of coefficients
for the same variables as in Table I for the entire
period, and compares them to the coefficients for just
the tight money periods.7 If, between periods of tight
and easy money, the Federal Reserve had engaged
in different defensive operations with respect to any
6Perfect defensive operations could have implied coefficients
with absolute values close to 1.0. In all cases the coefficients
in Table I are significantly different from one (1 ) for both
periods. Only defensive operations are accounted for in this
regression. By introducing variables to account for dynamic
operations, the estimations would have been more efficient
For an example of both defensive and dynamic operations
as an explanation of A P °, see Michael W . Keran and Christo
pher T. Babb, “ An Explanation of Federal Reserve Actions
(19 33 -68)” , this Review (July 1969) pp. 7-20. In this latter
case, the values of the estimated defensive coefficients are
not significantly different from one in absolute value.
"This approach uses multiplicative dummy variables which
have the property of allowing for shifts in the slopes of
the independent variables between the two periods. See
Arthur S. Goldberger, Econometric Theory (John W iley
and Sons, 1964), pp. 224-227.
Page 17

FEDERAL RESERVE BANK OF ST. LOUIS

MAY, 1970

T a b le II

FEDERAL RESERVE D E F E N SIV E O P E R A T IO N S
( M o n t h ly Central Differences — B illio n s of D o lla rs)
(J a n u a r y 1 9 5 2 —— Decem ber 1 9 6 4 )
Total
Period
G o ld

(A G )

F loat ( AF)
B o rro w in g s ( A B )
O th e r { A O )
C urren cy in H a n d s
of Public ( A C o )
C o n sta n t term
R2
D -W

T ight M o m
P eriods

—

.6 3
(4 .1 7 )
— .31
(2 .1 7 )
.9 8
(3 . 2 8 )
.23
(6 6 )
1 .3 3
(1 2 . 7 1 )
— .01
.7 4
1.48

.0 9
(.3 0 )
.5 7
(1 -8 4 )
.33
(.8 0 )
1.15
(2 .2 4 )
—

.22
(1 .0 2 )

N o te : Regression coefficients are the top figures ; their “ t” statistics
appear below each coefficient, enclosed by parentheses. R 2 is
the per cent o f variations in the dependent variable which is
explained by variations in the independent variable. D -W is
the D urbin-W atson statistic.

of the variables specified, the second series of esti
mated coefficients would be statistically significant.
If the Federal Reserve had not responded in different
ways between subperiods, then the second group of
coefficients would not be statistically significant.
In Table II the t statistics in the tight money period
for gold and borrowings are statistically insignificant,

Page 18

which indicates that the Federal Reserve’s response
to these independent influences on member bank
reserves was not significantly different during pe
riods of tight and easy money policy. There is no
statistical evidence that the Federal Reserve had
responded to changes in gold and borrowings dif
ferently during the subperiods.
The statistical tests represented in Table I and
Table II are consistent with each other. The results
in Table I show that total Federal Reserve defensive
operations did not change between the two periods.
Table II indicates that Federal Reserve behavior,
with respect to gold and borrowings, did not change
between subperiods. To state these results in statisti
cal jargon, Hendershott’s results have passed neither
the F test (Table I), nor the t test (Table II) of
statistical significance. Therefore, it must be con
cluded that Federal Reserve defensive operations
are not sensitive to changes in Federal Reserve pol
icy, and that Hendershott is not justified in treating
( Ps ) as responsive to “just” policy changes. Because
the neutralized money stock does not consider the
interaction between Federal Reserve holdings of
Government securities ( P ° ) and other sources and
uses of reserves, it is not an unbiased measure of
Federal Reserve policy actions.

The above two articles are available as Reprint No. 56.

FEDERAL RESERVE BANK OF ST. LOUIS

MAY, 1970

Publications of This Bank Include:
Weekly

U. S. FINANCIAL DATA

Monthly

REVIEW
MONETARY TRENDS
NATIONAL ECONOMIC TRENDS
SELECTED ECONOM IC INDICATORS - CENTRAL
MISSISSIPPI VALLEY

Quarterly

QUARTERLY ECONOM IC TRENDS
FEDERAL BUDGET TRENDS
U. S. BALANCE OF PAYMENTS TRENDS

Annually

ANNUAL U. S ECONOMIC DATA
RATES OF CHANGE IN ECONOMIC DATA
FOR TEN INDUSTRIAL COUNTRIES
(QUARTERLY SUPPLEMENT)

Copies of these publications are available to the public without charge, including
bulk mailings to banks, business organizations, educational institutions, and others.
For information write: Research Department, Federal Reserve Bank of St. Louis,
P. O. Box 442, St. Louis, Missouri 63166.

Page 19

S u b s c r i p t i o n s to this bank’s

e v ie w

are available to the public without

charge, including bulk mailings to banks, business organizations, educational
institutions, and others. For information write: Research Department, Federal
Reserve Bank of St. Louis, P. O. Box 442, St. Louis, Missouri 63166.