View original document

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


December 1977

The Threat of Inflation
Forecasting the EconomyThe Stock Market Versus Econometric Models
Federal Reserve MembershipEstimates of the Cost to Member Banks
Of Reserve Requirementa

This publication was digitized and made available by the Federal Reserve Bank of Dallas' Historical Library (

The Threat of Inflation
Remarks by
David M. Ully
Member of the Board of Governor.
Federal Reserve System
before a
Business Conference
Sponsored by Federal Reserve Bank of Dallas
Dallas. Texas
November 10, 1977

It is a pleasure for me to be here in the Eleventh
District and to appear before so distinguished a
group of banking and business leaders. Since I
joined the Board of Governors last year, and even
before that time, I have focused particular attention on the need to improve our economic education
and understanding-an improvement that is sorely
needed in our society and should be actively advanced.
Because your positions place you in the forefront
of economic education in your communities, I would
like to discuss with you today an issue that, while
perhaps the greatest threat to our economic society
as we know it, has not yet received the public attention that it deserves. That issue is inflation. It's
a disease that could change the entire fabric of our
life unless it is brought under control. And I must
say to you in all frankness that we at the Federal
Reserve feel very lonely when the Government institutions that worry about inflation are asked to stand
up and be counted. We live in almost virtual isolation, a condition that I hope will change, and
change dramatically, as time goes on.
The seriousness of inflation cannot be overestimated. Except for a brief period after World War II,
prices in this country in recent years have been
rising faster than in any other peacetime period in
our history. Most recently, there has been some
respite from double-digit inflation, but a resurgence
of inflation could jeopardize the future of our
country, if past experience is any guide. No country
in history has been able to maintain widespread economic prosperity, once inflation got out of hand. And
Renew I December 1977

the unhappy consequences are not just of an economic
nature. Inflation inevitably causes disillusionment
and discontent. It robs millions of people who have
set aside funds for education or retirement, and it
hits the poor and the elderly especially hard.
Inflation has also been a contributing factor to
the fall of governments. Even in this country, the
distortions and injustices created by inflation have
contributed to a distrust of government officials
and governmental policies. History teaches us that
the discontent bred by inflation can provoke disturbing social and political change. The ultimate consequences of inflation could well be a significant
decline of economic and political freedom.
To appreciate the threat of inflation fully, one
must understand how it has infected our economy.
There is in our society a built-in inflationary bias
that has grown more intense over the last deCade.
Our long-range problem of inflation has its roots in
the structure of our economic institutions and in the
financial policies of our Government. This basic
fact is aU too frequently obscured by outside events
that increase the rate of inflation noticeably, such
as a crop failure or action to raise oil prices. The
plain truth is, however, that the United States-and
many other countries throughout the world-has developed an underlying bias toward inflation that has
only been magnified by these special influences.
Where does this leave our traditional tools of policy? Fiscal and monetary policies still have impor·
taut roles to play in our fight against inflation, but
they can no longer do the entire job. Removal-or
even a reduction-of structural barriers will help us
squeeze the inflationary bias from the economy and
remove the distortions that we, ourselves, have made
a part of our system.
One indication of these distortions is the constant
penchant of our society to run budget deficits. I
have no desire to delve deeply into the pros and cons
of budget deficits-nor to criticize what is our very
commendable national willingness to provide for the
less fortunate members of our society. The record
shows, however, that since 1950, deficits have outrun surpluses by 23 to 5. That is a runaway score in
anybody's language, and it has tended to put the
Government's stamp on inflation.

But let me give you some examples of how the Government has contributed indirectly to the inflationary
bias in our economy. Let me explain, first, some of
the structural problems we have created and the structural reforms that must be made to supplement traditional monetary and fiscal policies as the best hope
of unwinding inflation.
For many years, the Federal Government has regulated various forms of transportation-trucks, railroads, airlines, and ships. A certain amount of regulation is undoubtedly good, but analysts estimate that
regulation of transportation in this country has raised
costs anywhere from 10 to 30 percent.
Tariffs, together with import barriers, are another
factor in this bias. Granted tariffs and import barriers have benefited certain industries and groups
of Americans. But they have also served to increase
prices of some commodities or to prevent a drop for
The list of examples is legion-restrictive building codes, agricultural price supports, the market
power of business and unions, pollution controls,
safety regulations, pension costs, mandatory retirement, and the minimum wage laws. Taken together
or individually, they increase the economy's bias for
Let me give you a current example. It is our estimate that the rate of inflation for next year will be
1 percent more because of the new payroll taxes
(social security and unemployment insurance), the
new minimum wages, and the proposed energy taxes.
I am not malting a moral judgment on these programs. I am only suggesting that Congress may well
have offset them by other deflationary actions.
Let me take a moment to explain one bias that is
particularly troublesome to me. That is the innovation that has crept into our economic thinking in
recent years of indexing price increases. Since the
economy experienced double-digit inflation several
years ago, there has been more and more emphasis on
indexing. The "cost-of-living adjustment" has frequently been called an engine of inflation by government economists. Nevertheless, more and more labor
contracts now include the same type of provisions.
There are some economists who believe a cost-ofliving escalator-or indexing-is a perfectly rational
approach to the nagging problem of inflation. As
costs go up, incomes also go up by the same amount
and the effects of inflation are nullified, or so the
argument goes. But indexing also has one destructive

feature that makes it an unacceptable approach to
inflation. It serves to take away the incentive of
people to fight the inflation itself. It is an admission of defeat. It is,literally, throwing in the
towel, and that is one thing I and my colleagues at
the Federal Reserve will never do.
While I am on this point,let me say that I have
been particularly distressed of late by business'
tacit acceptance of these new sorts of escalators and
arrangements-almost as a matter of course nowand by their effects on the structure of the labor market. For one reason or another-fear of strikes or
delays perhaps-management is simply not pressing
as hard as it should be or as it once did to hold down
labor costs. This is also a form of giving up the fight;
I don't think there is another area in which structural
inflation so clearly has a life of its own, or in
which we have acquiesced so easily to expectations of
continuing inflation. We need now to examine much
more diligently the structure of the labor marketto look closely at the fact that we have permitted
high unemployment to exist while concurrently granting high wage increases. We need to look at other
such distortions that have been allowed to occur.
These practices are inimical to our goals of full
employment and stable prices.
What we as a people have done, in summary, has
been to subject our available resources to increasing
intense demands. At the same time, we have sought
to ensure that incomes do not get eroded when excessive pressures generate inflation. This amounts to
creating upward pressures on costs and prices, and
then arranging to perpetuate them. It is an awesome
combination and one which we at the Federal Reserve
must cope with.
I become very disappointed when I see the Federal
Reserve alone in its struggle against inflation. Too
often, our perseverance is attributed to callousness
about the answer to the following question: Won't
the unemployment picture worsen if we try to reduce
inflation? But the answer to that question happens
to be an emphatic NO. There is no longer a meaningful trade-off between unemployment and inflation.
Inflation is one of the major causes of unemployment
in the current environment. It leads to hesitation
and a sluggish buying. Today's high unemployment
rate is, fundamentally, the legacy of an inflation
that surely could have been avoided if proper steps
had been taken in time.
I hope you have not become discouraged by my
remarks thus far. Now that I have painted a rather
gloomy picture of inflation,let me offer some measures designed to bring it under control. And I must

emphasize that inflation can be brought under control
if we work together. Only a few years ago, we had
double-digit inflation. The present rate of inflation is
half what it once was, but we must do better and
we can.
Conventional thinking about the economy, however, is out of date. Ways must be found to reduce
unemployment while at the same time avoiding a new
wave of intI.ation. The areas that must be explored are
many and difficult, but we must open our minds if we
are to have any chance of ridding our economy of its
inflationary bias.
The first step in an anti-inflationary program is
prudent monetary and fiscal policy. This is essential.
Any heightening of inflationary expectations will
erode business and consumer confidence and impair
the economic expansion now underway. Thus, any
attempt to make a "quick fix" of our major economic
problems by a heavy dose of economic stimulation
may be counter-productive. Only a steady, moderate
policy will provide the foundation for both a lessening
of inflation and a return to full use of our labor force
and productive capacity.
But prudent monetary and fiscal policy is not all
that is required. Inflation is more than an evil of
the business cycle. As I have emphasized, it is a
structural phenomenon as well. We must improve the
efficiency and flexibility of our markets. Structural
changes in our economy will be hard to come by. But
we must make the effort if we are to have any success
in our struggle.
For example, steps must be taken to improve competition among businesses through increased emphasis
on antitrust policy, changes in regulatory procedures,
and an easing of barriers to international trade. Local
building codes that do not reflect modern construction techniques should be reexamined and changed
where necessary. We should begin to substitute economic incentives for the morass of costly and inefficient safety and environmental regulations we have
imposed on certain industries.

As I mentioned, I have a special interest in our
labor market policies. Our job training programs
should be strengthened to increase productiveness of
workers, particularly minorities and teenagers. These
programs should provide a better match between
worker skills and business requirements. We also need
expanded job bank programs. The Federal minimum
wage law is pricing many teenagers out of the job
market. It has been estimated that a 25-percent rise
in the minimum wage is associated with an average
of a 3- to 6-percent reduction in teenage employment
When the minimum is raised, employers tend to
Review I December 1977

ignore the less skilled worker and seek the more productive employee. Employers may also attempt to
get along with fewer employees. Some Federal lawssuch as the Davis-Bacon Act, which requires the payment of "prevailing wages" on goverrunent construction contracts---<:ontinue to escalate costs. As I
indicated, we must find alternatives to automatic
cost-of-living adjustments that have the same effect.
There is also a belief that the unemployment compensation laws provide such generous benefits that
incentives to work are reduced. Reform of the welfare
system to increase work incentives, a rethinking of
mandatory retirement policies-these are also
needed. I could go on, but I think. you get the picture.
A restructuring of our economic system-along the
lines I have outlined-would improve the efficiency
and effectiveness of our conventional tools. During
periods of excessive demand, for example,less
restraint would be required to bring inBation under
control. Consequently, I am convinced that structural reforms deserve more attention from Congress
and, yes, from the general public than they have been
receiving. Too many people have tended to concentrate on overall fiscal and monetary policies. These
traditional tools are, of course, useful and essential;
but once inflationary expectations have become widespread, the tools must be used with great care and
moderation. We must work to remove the in1lationary
bias of our economy.
On our part, we at the Federal Reserve will continue to provide money and credit sufficient to
accommodate a continuation of the orderly economic
expansion. We will strive to avoid excessive growth in
money and credit that would stimulate further in6ationary pressures. The Federal Reserve System,
moreover, is committed to a gradual reduction over
the longer run as one of the conditions that must be
achieved in order to bring about an end to the inflation. As I have tried to emphasize, however, other
conditions are also necessary to reach this happy end,
including progress in reducing the structural distortions that have added to our economy's inflationary
But these policies require public support and
understanding. People must understand that this
problem of inflation is a serious long-term threat to
our social and economic system. People must understand that government programs, no matter how
worthy, are likely to have inflationary consequences
unless they are accompanied by an appropriate means
of financing. Even programs that are superficially
costless may still impede the efficient workings of markets and the Bexible adjustment of prices. Such programs may impose very heavy burdens of their own.

The Federal Reserve, in this struggle against inflation, has a unique role to play. Congress has enabled
the Federal Reserve System to operate independently
of day-to-day political pressures. Considerable scope
for independent judgment has been permitted by providing for the appointment of members of the Federal
Reserve Board for long, staggered tenns. Moreover,
Congress has placed the Federal Reserve System
outside the usual appropriations process.
This provides us with a unique environment especially conducive to adopting a longer-run point of
view. And it is this longer-run perspective that
leads us to our concern about the problem of
inflation. We hope to convince others that this concern is wananted and that this struggle against
inflation is worthy of their support.
Let me offer one idea that might have appeal for
Americans who are worried about inflation. Many
people who have been worried about the destruction
of our environment succeeded in pushing through
requirements for "environmental impact statements"
for many types of development. Although many
businesses have complained about the "red tape"
produced by these statements, they have served a
purpose in helping to protect our environment. In


some cases, projects have been canceled or modified
because of their impact on the environment. We
should, perhaps, borrow a page from this book and
require "inflation impact statements" to be filed by
the Government in connection with changes in the
minimum wage, import restrictions, and other regulations that give our economy an inflationary bias.
True, this would add another layer of red tape to
already cumbersome procedures, but it would awaken
the public to the cost that must be borne when new
legislation is adopted and new procedures are instituted. Some Federal agencies have already mandated
such an approach.
Somehow, inflationary pressures have to be
uprooted. But that cannot be done without changing
a number of ideas and attitudes that have been popular for a generation or more. Consequently, we should
have no illusions about finding a quick and easy
solution. It will take years of prudent and vigorous
public policies and restraint on the part of consumers,
workers, and business firms alike. Even so, I am confident that with the imagination you and other business leaders across the country are displaying, we
can work to put an end to inflation and restore the
conditions essential to a stable prosperity .

Forecasting the Economy-

The Stock Market Versus Econometric Models
By Wallace H. Duncan
There is a traditional wisdom. often heard along
Wall Street and in other financial circles, that
the future course of the economy is to be read in
the trend of the Dow Jones averages. Just as the
Crash of '29 was the harbinger of the Great Depression, it is said, so also does the stock market signal the less cataclysmic expansions and contractions
of today's economy.
At times, this view is also coupled with strong
skepticism of efforts to predict economic trends
through the assembly and interpretation of massive
amounts of statistics using computers and econometric models. For example, in a recent article in the
Wall Street Journal, a case was made for the stock
market's forecasting ability. The author arrived at
the following conclusion:

So who is right? Well, though the economists give us
"hard numbers," while the stock market gives us
nothing but massive intuition, it is the market that
is more often in the right.... there does not exist an
econometric model which, over these recent years,
has consistently outperformed aU the other models
- nor do all the models taken together give us more
accurate predictions than the stock markel l

If this view is correct, then a great many economists and statisticians could be assigned to work
on other problems, and the makers of national economic policy could simply take their signals from
the stock market.
The contention that the stock market is at least
as good an economic forecaster as are large-scale
econometric models certainly should not be dismissed
out of hand. Fortunately, it is not difficult to
devise comparative tests of forecasting capabilities.
But the evidence from the tests reported here is
that, though not without some capability, the stock
market is nevertheless inferior to large-scale econometric models as an economic forecaster.
The view that the stock market serves as a precursor of economic activity rests on two quite different
types of analysis. The first contends that stock
market participants as a group are able to foresee

1. Irving Kristol, "The Fozes VB. the Hedgehog," June 14,
1977. See also the lead editorial in the October 20,1977,

Review I December 1$77

the future path of the economy with a substantial
degree of accuracy and that the movement of stock
prices reflects this infonned consensus. The second
is that changes, for whatever reasons, in stock
prices actually cause subsequent changes in economic
activity through influencing the spending decisions
of consumers and businessmen. These two views-the
stock market as economic seer and the stock market
as economic cause--are considered in general terms
before turning to an analysis of the evidence of the
stock market versus econometric models as economic

The evidence from the tests reported here
is that, tlwugh not witlwut some capabil.
ity, the stock market is nevertheless
inferior to large·scale econometric mod·

els as an economic forecaster.

Stock market as a predictor
A cursory look at the accompanying graph reveals
that the tops and bottoms of stock price swings are
related to those of the economy and tend to precede
them. And there is ample reason to expect such a relation. Corporate profits are affected by production
and sales, and in turn, stock prices reflect, among
other things, investors' expectations of future corporate profits. If investors are able to guess the future
with some measure of success, stock prices should
provide a leading indication of changes in the state
of health of the economy.

Standard and Poor's index of prices of 500 common stocks is one of the 12 components making up
the Commerce Department's composite index of leading indicators. This index of stock prices is also given
the most weight of all the components in the calculation of the composite index, since it has the highest
score as an individual leading indicator according to
six objective criteria.' Since the 12 components of the
composite index of leading indicators were selected
as the best among 150 cyclical indicators studied,
and stock prices receive the highest score of these


Economic recessions and recoveries are usually signaled in advance by the stock market



60 -

50 -

40 -

30 -

20 -


NOTE; Shaded a reas Indicate economic recessions as dated by the National BUfeau 01 Economic Research.
SOURCE; Standard to Poor's Corporation .



best 12, it is evident that a strong case exists for
looking to the stock market for clues as to where the
economy is headed.
However, there is also some evidence that stock
prices do their job as predictor of economic downturns too well. Paul Samuelson, a winner of the Nobel
Prize in economics, has remarked dryly that the
"stock market has correctly predicted nine of the last
five recessions."
Samuelson is alluding to the stock market's most
serious drawback as a leading indicator-its tendency to produce false as well as dependable signals.
While the exact count of false signals depends on how
one wishes to define them, there can be little serious
doubt that in 1962 and in 1966, stock market declines
incorrectly indicated the approach of recession. In
those instances, Standard and Poor's index of 500
stocks declined about 28 percent and 22 percent,
respectively. These stock market declines were considerably greater in magnitude than those that had
correctly signaled the onset of the previous three
Thus, it is difficult to say with much confidence
just what any particular stock market decline may be
signaling. For example, the present market decline
from September 1976 is greater than some prior declines that have correctly signaled approaching recessions, and at the same time, it is less than some that
have turned out to be false signals.
Stock market as cause of economic fluctuations
The stock market's role as a leading indicator implies
no necessary causality. Both the stock market and
the economy may be responding to other forces that
are in fact causal agents, with the stock market
leading the economy simply because it responds more
quickly to these other forces. For example, changes
in monetary or fiscal policy may have an immediate
impact on the stock market as investors assess the
implications for future profits, while the actual
impact on the economy may not be evident for several
2. See Victor Zarnowitz a nd Charlotte Boschan, "Cyclical
Indicators: An Evaluation and N ew Leading Indexes,"
Busin ess Conditw ns Digest, May 1975, p. xv. For a
critical view. see Saul H. Hymans. "On the Use of
Leuding I ndicators to P redict Cyclical Turning Points,"
BrookinGS Popers on Economic Activity. 1973. no. 2.
IJ. 346. HymalllJ finds that in a multiple regre8Sion of an
index of current economic activity on 12 leading indicators. s tock priceJ are insignificant and have the wrong
s ign. However. Hymans' conclusion is probably fl awed by
the existence of strong mul ticollinearit y among t he
explanatory variables.
Review I December 1977

In recent years economists have been devoting
more attention to the possibility that changes in stock
prices may actually cause subsequent changes in aggregate demand and, consequently, in employment
and output of the economy. Such influence may operate through consumption or investment, or both.
The most generally accepted view of how changes
in stock prices affect consumption is that consumption depends, among other things, on wealth. Consequently, other things being equal, an individual's
consumption increases when his wealth increases and
declines when his wealth declines. Thus, stock price
changes cause changes in wealth, which in turn cause
changes in consumption-and this in turn causes
changes in employment and production.

In recent years economists have been
devoting more attention to the possibility that changes in stock prices may
actually cause subsequent changes in
aggregate demand and, consequently, in
employment and output of the economy.

This view is consistent with the life-cycle theory
of consumption, which is built into the giant econometric model of the economy known as the MITPENN-SSRC model. 3 This model contains several
hundred mathematical equations, consumes oceans of
economic statistics, and is used extensively by the
Federal Reserve Board and others to simulate the
economy. As with other gadgetry devised by man, it
is ever in need of update and repair. In its latest version, for any change of $1,000 in wealth, annual consumption spending changes about $47. While a complete adjustment takes place over a period of a year
and a half, almost two-thirds of the full change in
consumption spending occurs within two quarters of
a change in stock prices.

3. Albert Ando and Franco Modigliani, "The 'Life Cycle'
Hypothesis of Saving: Aggregate Implications and Tests,"
American Economic Review 53 (1963): 55·84.
4. See Frederic S. Mishkin, "I\Iiquidity, Consumer Durable
Ex~nditure, and Monetary Policy," American Economic
Review 66 (1976) : 642-54; and Mishkin, "What Depressed
the Consumer? The Household Balance Sheet and the
1973-75 Recession," Brookings Papers on Economil:
Actiuity, 1977. no. 1, pp. 123-64, and the comments and
discussion following that article.

An alternative way of looking at the relationship
between stock prices and consumption is the liquidity
hypo~hesis. 4 In Mishkin's view, consumer behavior
responds primarily to the composition of wealth
rather than its total amount. The consumer's balance
sheet is divided into illiquid assets (housing and consumer durables), financial assets, and liabilities. Financial assets provide a buffer against forced liquidation of illiquid assets during periods of adversity.
When financial assets are reduced by a decline in
stock prices, the consumer shifts his demand away
from housing and durables in an attempt to restore
the balance between illiquid assets and the buffer
of financial assets. Conversely, rising stock prices
cause the financial assets to be more than needed to
perform the buffer function, and, consequently, consumers increase their demand for durables and housing. Mishkin estimates that for every SI,OOO change
in the value of financial assets, spending on consumer
durables changes $63 per annum and on housing $32
per annum. This estimate of the effect on consumer
durables spending is about a third greater than the
estimated effect on consumption in the MIT-PENNSSRCmodel.
Business investment in new plant and equipment is
a second category that may be affected by changes in
stock prices. One theory of how stock prices affect
investment is that fixed capital provides a continuous
flow of services to the firm and these services are
paid for by a continuous flow of "rental" payments
in the form of dividends to equity sources and interest to lender sources. 5 When a rising stock market

5. See Dale W. Jorgenson, "The Theory of Investment
Behavior," in Robert Ferber, ed., Determinants of lnuest.
ment Behavior. Universities-National Bureau Conference
Series, no. 18 (New York and London: Columbia University Press (or National Bureau of Economic Research,
1967); and Robert E. Hall and Dale W. Jorgenson,
"Application of the Theory of Optimum Capital Acx:umulation," and Charles W. B ischoff, "The Effect of
Alternative Lag Distributions," both in Gary Fromm,
ed .. Tax lncentives and Capital Spending, Brookings
Institution Studies of Gover nment Finance (Washington,
D .C., 1971).
6. See William C. Brainard and James Tobin, "Pitfalls in
Financial M odel Building," American Economic Review
58, no. 2 (May 1968): 99-122; and Tobin and Brainard,
"Asset Markets and the Cost of Capital," in Bela Balll8!!a
and Richard Nelson, eds., Economic Progress, Private
Values. and Public Policy .' Essays in Honor of William
Fellner (Amsterdam and New York: North-Holland
Publishing Company, 1977), pp. 235-62.
7. James Tobin, "Monetary Policy in 1974 and Beyond,"
Brookings Papers on Economic Activity. 1974, no. 1, p. 223.

causes the dividend-stock price ratio to fall, the
"rental price" of capital from equity sources declines
and managements are stimulated to invest in new
plant and equipment. Conversely, a falling stock mar·
ket increases the rental price of capital and inhibits
new investment.
There are other theories of course. One compares
the market value of a firm (as represented by the
total value of outstanding equity shares) with the
replacement cost of its capital. s In this view,
whenever a rising stock market causes the market
value of a firm to exceed the replacement cost of its
capital, investment will be stimulated since "a cor·
poration can sell paper claims to physical capital for
more than the capital costs.'" Conversely, invest·
ment is inhibited when finns would raise less money
by selling paper claims to capital than the physical
capital actually costs.
What, then, might be the total effect that a change
in stock prices could have on the economy, operating
through both consumption and investment? Barry
Bosworth, using the MIT-PENN -SSRC model, esti·
mates that about a fourth of the decline in national
output during the 1974-75 recession can be attributed
to the prior decline in the stock market. 8 In other
words, had the market remained at the level it
reached in late 1972, the subsequent economic decline
would have been only three·fourths as severe as ac·
tuaDy occurred. Employing the alternative relationships based on the liquidity hypothesis, Mishkin
estimates that the decline would have been only half
as great as it actually was, had stocks remained at
their prerecession level.'
Stock market and econometric models
Useful information for forecasting is embodied in
stock prices, whether the stock market tends only to
foresee the economy's future path or actually helps
to cause it. The question is, How much information?
Does this information enable one to forecast the
economy with greater, or less, precision than can be
obtained with large-scale econometric models currently in use?
This question is of particular importance at the
present time, since the stock market is currently
signaling a different economic outlook than that of
most forecasters using econometric models. A model

8. "The Stock Market and the Economy," Brookings Papers
on Economic Activity. 1975 no. 2, pp. 257-90.
9. '"What Depressed the Consumer? The Household Balance
Sheet and the 1973-75 Recession," pp. 156-61.

T8bl. 1

(Annual percentage rates of growth
of real gross national producl)
Mean ebsol"ta
error r

0 ..


Root-m .. /In ·
,,,u..e error'







Stock price model
Conference Board
Chase Econometrics
Data Resources
Naive model .. ....... .





oll~a .I>solule valuu of Iha erro,$.
I~e s quired valuu 01 Ihe .moll;
simrl.r 10 • ,Ian(!ar(! (!e'fillion.
NOTE: Foreca.!ll, are lor each Quaner Ind for each tWO-Quarter .,e,lod Irom
Ihe Ihird qUlrter 011972 through Ihe .econd qUB,lor 011977.
SOURCES: Conleranea Board.
U.S Oeparlment 01 Comme'ce.
Fed"al Rne rve Bank 01 OaJlu.


The .verJge

2. The SQua .. roOI of 1M average 01

based on past stock prices, to be described subsequently, is forecasting relatively weak economic
growth for both the present quarter and the first
quarter of 1978, with real gross national product
(GNP) projected to increase at annual rates of 2.6
percent and 3.3 percent, respectively. This is quite
close to Chase Econometrics' forecast of 2.5 percent
for the present quarter but somewhat stronger than
the 1.8 percent of growth projected for the first
quarter of 1978. However, most comparable current
predictions of other recognized forecasters tend to
be significantly more optimistic than either of these.
At least a tentative answer to the question of relat ive forecasting capabil ities can be provided by
analysis of past experience. The rate of growth of
real GNP was compared with past rates of change of
stock prices for a 25-year period from mid-1947 to
mid-1972 using conventional regression analysis. 10
A positive and significant relationship was found
between the rate of change of G NP in any quarter and
the rate of change of stock prices during the two
immediately preceding quarters. No consistent relationship was found for quarters further removed in
time. This suggests t hat the stock market is a relatively short-term forecasting device, providing
little useful information beyond a forecast horizon
of about six months. II
The observed relationship was used to forecast real
GNP growth rates at quarterly intervals over the subsequent five-year period from mid-1972 to mid-1977.
Review I December 1977

Obviously, such forecasts were not precisely "on target," as is common with nearly all forecasts. The forecast errors of the stock price model were compared
with errors of other forecasts for the same period.
These included t he forecasts of Chase Econometric
Associates and of Data Resources, Incorporated,
based on their econometric models; a primarily subjective economic forecast made by the Conference
Board; and a naive model. The Conference Board,
Chase Econometrics, and Data Resources
were selected from a number of widely recognized
forecasts because the data were available for the
entire forecasting period and were most closely comparable to t he stock price model in timing. The
naive model foreca sts that next quarter's growth rate

10. A number of alternative specifications of t he r elationship
i>ctween GNP and stoc k prices were tried. The results
reported here are bllser! un th e relationsh ip t hat provided the best fit. thu s giving stock prices the best chance
of fo recflSting with errors of minimum size. H owever , even
the best fi t was cap~,ble of explaini ng only about a fourth
of Ihe qU1lCterly variation in the percentage nlte of
growth of GNP. The UPllenrlix to th is article (."Ontains 11
det.ailed descri ption of the statistic:al procedures a nd
11. T his result is consist.ent wilh the view that investors
have some Ilbility to sec into the economic future for a
period of about six months but have little ability to do
so beyond thut. Alternatively. it also suggests that most
or Ihe casu a l effect t ha t changes in stock p rices may have
011 th e aggrega te demand for goods and services tends to
occur wi thin a rela tivel y short time lag.


will be the same as last quarter's and that the growth
rate over the next t wo quarters will be the same as
that over the previous two quarters.

The forecasting errors are summarized in Table 1.
The smallest errors over the five-year period were
achieved by Chase. Its average absolute errors were
about half those of the naive model for both one-quarter and two-quarter forecasts. The forecasts based
on past stock prices resulted in errors that were
about three-fourths as large as t hose of the naive
model. Thus, using the naive model as a reference
point, Chase was able to reduce the size of forecasting
errors about twice as much as the forecasting model
based on past stock prices. While Chase achieved the
best results, bot h of t he large-scale econometric
models forecast more accurately than did the stock
price model.'z

Table 2

(Annual percentage rates of growth
of rea l gross national product)
One_q".rt.r 10reculS


Stock price model
Conlerence Board
Chase Economet ri cs
Data Res ources
Naive model

e rror'

s qua re








"ver age 01 '~e absOlute vatuel 0 1 t~ e e rrors.
s quare fOol 01 tM e ave rag e 01 Ihe s qu ""d
values 01 the erro ,.; $lm,larlO a I,and a,d
~OT e , T~e umpl" pe,IOd I. 1973·~. 19H..Ql.
1975·01. and 1975-02. T~".e a ,. the
Quarte rs ,m~dlat" surtoundlftg t~" uPPe<
Bnd lowe, turninll points 01 In, mOlt rac ent
bu s ine ss cycle.
SOURCES, Conle rence Board.
U S Oe pa rtment 01 Commerce
FeCle,nl Reserve B an~ 01 Dallas .


T ~e

12. Additional statistical tests were conducted to determine
whethe r the diffe rences between the mean absolute
errors of the va rious forecasti ng models we re significant
or were simply due to random chance. As a result, the
possibility that the differences between the Chase errors
and those of the stock price model were due to random
chance-implying that the mean absolute errors of the
two models would not differ if the samples were infinitely
large instead of onl y 20 quarte~ould be rejected with
a 9S-percent level of confidence. These forecastin g
results thus indicate that a t least one of the large-scale
econometric mode ls performs s ignifica nUy better than
forecasts based on past stock prices.


Suppose, however, t hat one compares the forecast·
ing errors of different models for only the turning
points of economic expansion and contraction. Would
past stock prices forecast better t han large-scale
econometric models at such turning points? For the
period studied, the answer is no. The Chase and Data
Resources econometric models both have smaller
average absolute errors in the turning-point quarters
than the Conference Board's subjective forecast or
the stock price model. The average absolute error of
the stock price model is about midway between
average errors of the best econometric model and the
naive model. However, because there were only four
turning-point quarters during the period studied, it
is not possible to state with assurance that this relationship would be likely to be maintained in
other periods.
The evidence from both the 20-quarter sample
period and the 4-quarter turning-point sample thus
suggests that economic forecasts based on past stock
prices are not able to outperform the forecasts of the
large-scale econometric models. For the full five-year
period and for the turning points, the stock price
model was able to reduce t he size of the errors of the
naive model by only about haIf the amount accomplished by the best-performing large-scale econometric model.
One might t hus conclude that while the stock market does provide infonnation that is useful in forecasting the economy, such infonnation alone does
not pennit forecasts as accurate as those attainable
with the large-scale econometric models. A secondary conclusion is that these models appear to be
capable of forecasting with errors about haH as large
as those of a naive model, which simply projects the
most recent trend.

The stock price model
In seeking the best econometric relationship
between real gross national product and
lagged stock prices, severe problems of serial correlation of residuals were encountered
when the data were used in the form of levels of GNP and stock prices. Use of a Cochrane-Orcutt iterative estimation technique
allowing for first-order autocorrelation did
not adequately cope with the difficulty;
hence, the relationship was estimated in the
form of percentage rates of change. A comparison of the standard deviation of the regression equation with the root-meaD-square
efror (3.87 and 3.76, respectively) indicates
that there was no deterioration of the estimated relationship in the post-sample-forecast period.
The best fit was obtained by regressing
quarterly data of annual rates of change of
real GNP on past a nnual rates of change of
nominal stock prices. The latter were calculated from quarterly averages of stock
prices as measured by Standard and Poor's
composite index of 500 common stocks,
with historical data obtained from the Commerce Department's Business Conditions
Digest. Regressions were also run using real
stock prices (deflated by the GNP implicit
deflator), using monthly rather than quarterly rates of growth of past stock prices,
and using log-linear forms. But no improvement in fit could be obtained.
The best relationship was chosen as the
one having the smallest standard deviation
of the residuals, subject to the constraint
that no negative coefficients on lagged stock
price terms were pennissible. The best relationship estimated was;
y , = 2.77 + .064SPt_l + .038SPt . 2 + e t
(4.BB) (3.4B)



.238 (2.45)

R2 = .245
Durbin-Watson statistic = 2.04
Standard deviation of the estimate = 3.87
where Y = annual percentage rate of change
of real GNP at quarterly intervals, and SP
= annual percentage rate of change of
nominal stock prices at quarterly intervals.
Numbers in parentheses denote t statistics.
All the coefficients are significant at the
5-percent level, with the constant teno and
Review I December 1977

the coefficient on stock prices lagged one
quarter being highly significant at the I-percent level. The Cochrane-Orcutt iterative
estimation technique was used, employing
100 quarterly observations from 1947-Q3
through 1972-Q2, inclusive.
It is apparent from the adjusted R 2statistic that only about a fourth of the quarterly
variation in annual growth rates of GNP
can be explained by lagged stock prices.
And the standard deviation of 3.87 percent
is relatively large compared with the mean
real GNP growth rate of 3.81 percent during the period.
The estimated relationship was then used
to make 1-quarter-ahead forecasts in the
20-quarter post-sample period of 1972-Q3
through 1977-Q2, inclusive. The infonoation contained in the preceding error term
was used in the forecasts.
A similar estimated relationship deleted
the one-quarter lagged stock price term, and
this regression was employed to make onequarter forecasts two quarters ahead in the
post-sample period. The one-quarter-ahead
forecasts and the two-quarter-ahead forecasts were then combined to obtain the twoquarter, or half-year, forecasts from which
the two-quarter errors in the table were
Other forecasts. The other forecasts-those of the Conference Board, Chase Econometrics, and Data Resources--were selected because the data were available
throughout the 1972-77 period and the forecasts were usually available in the first
month of each quarter. Thus, they are comparable in timing to the forecasts of the
stock price model, which could have been
made in the first month of each quarter, as
soon as a preliminary estimate of the previous quarter's real GNP growth rate became available. Some alternative forecasts
- such as those of the Wharton, Michigan,
MAPCAST (General Electric), and Merrill
Lynch econometric models--were not used
because either they were not available as
early as 1972 or they were released late in
the quarter or at irregular intervals.
Even with the three forecasters included
in the tests, there were a few instances in
which no forecast was available for the first

month of the quarter (three times for the
Conference Board and one time each for
Chase Econometrics and Data Resources).
The latest prior forecasts were used in these
instances. Forecast data for these three forecasters were taken from various issues of
the Conference Board Statistical BuUetin
and were used in the original form of actual
percentage growth rates of real GNP.
The naive forecasts were constructed as
simple extrapolations of the most recent
growth rates of real GNP, using data that
were actually available at the time the forecasts could have been made. They were
thus constructed from preliminary estimates and first revisions of real GNP as
issued by the Commerce Department.
Biases in forecast comparisons. The stock
price model is favored in the tests in two
ways. First, as mentioned previously, earlier
forecasts for the other models Were used in a
few instances when no forecast was available for the first month of the quarter. In
these instances, the stock price model had
the advantage of a shorter forecast horizon.
Additionally, the stock price model was
estimated from GNP data containing all
revisions to date, rather than the corresponding data actually available in 1972. In
the forecasting period, the forecast made
use of information in the previous error
term, and this error term was calculated
from current data, including all revisions,
rather than from the preliminary estimates
that would actually have been available at
the time. Thus, the stock price model has
had some measurement error purged from
it, while the other forecasts have not had
the advantage of this ex post refinement.
The econometric forecasts, however, are
favored by the test procedure in at least one
way. Throughout the five-year forecasting
period, the stock price model was not revised. In contrast, the other forecasters were
at liberty to adjust their models throughout
the period, incorporating new information
about changing relationships as it was perceived. In contrast to the stock price model,
they were not required to forecast in 1977
with the same model developed to forecast
in 1972.


The effects of these biases incorporated
in the tests are partially offsetting. And the
net effect is probably small enough not to
place the conclusions in doubt.
Significance of differences in mean absolute errors. One might consider determining
whether the mean of the signed errors of
any specific forecaster differs significantly
from zero, in order to test for forecaster
bias. However, accuracy is probably better
measured. by the size of the absolute errors
rather than by the degree of bias, since the
latter appears to be comparatively small in
all cases.
It is plausible to assume that the signed
errors of any specific forecaster are normally
distributed about a mean value somewhere
in the vicinity of zero. If so, then the absolute errors will have a distribution that is
quite skewed and that will resemble the
right half of a normal distribution to the extent that the mean of the signed errors is
close to zero. Significance testing of mean
absolute errors based. on the normal distribution would therefore not be valid for
small samples.
However, the central limit theorem indicates that regardless of the distribution of
the parent population, the distribution of
sample means will approximate a nonnal
distribution for large samples. Significance
testing using Student's t distribution would
thus be permissible. While a sample size of
20 is, at best, only marginally suitable as
a large sample, it was nevertheless considered. to be close enough to warrant the tests.
An alternative nonparametric significance
test, which does not depend on any population distribution assumptions, was also conducted but did not yield any interesting
The null hypothesis tested is:
H: Ul-U2 = 0,
where U l and lb.! are the mean absolute errors of two different forecasting models. A
two-tailed test is indicated, since it is not
known a priori which of the two forecasts
may be better. Since the forecast errors
cover the same sample period for both forecasters, they are not independent but may
be paired by specific quarters. The paired-

sample test is thus appropriate.1
The null hypothesis that the stock price
model errors do not differ from those of
Chase Econometrics can be rejected at a
confidence level of .95, suggesting that one
can be quite confident that the difference
in forecasting accuracy between the two
models during the sample period does in
fact represent a true difference in forecasting capabilities. On the other hand, in a
similar hypothesis test between the stock
price model and the naive model, the null
hypothesis can be rejected only at a conn-

Rmew I December 1977

dence level of .50. It cannot be rejected at a
confidence level of .60. This suggests that
one can be much less certain that a true
difference exists between the forecasting
capabilities of the stock price model and the
naive model.

1. See, for example, Thomas H. Wonnacott and
Ronald J. W01Ul8cott, Introductory SUlfuric.
lor Business tmd Economics (New York: John
Wiley & Sona, 1972). pp. 171·78.


Federal Reserve Membership-

Estimates of the Cost to Member Banks
Of Reserve Requirements
By Edward E. Veazey
Membership in the Federal Reserve System has declined continuously in recent years. In the Eleventh
Federal Reserve District, which includes Texas and
parts of Louisiana, New Mexico, and Oklahoma, 54
percent of all commercial banks were Federal Reserve
members in 1965; in 1976, 46 percent were members.
The continuing erosion of membership, long a matter
of concern to Federal Reserve officials, has drawn
increased attention from bankers, banking organi~
zations, and the Congress in recent years.
All agree there is continuing need (or effective
monetary policy. Most agree also that this can be
realized most effectively if a broad cross section
of banks of all sizes, types, and geographic areas
are members or, at least, are subject to comparable
monetary control. Moreover, simple competitive
equity requires that firms doing the same kind of
business in like situations be subject to like
Surveys and studies have revealed a number of
reasons why banks choose to be or not to be Federal
Reserve members. The reasons given for choosing
to be a member usually include one or more of the
services available from Federal Reserve banks, or
the more general philosophical view that all banks
have a responsibility to participate directly in
the monetary mechanism of the country and the regulation thereof.
Members have assured access to Federal Reserve
credit for certain purposes; they may send their
checks to their Federal Reserve banks for collection;
their securities may be held in safekeeping in
Federal Reserve vaults; money and U.S. Government
securities may be transferred to banks anywhere in
the country on the Federal Reserve's wire-transfer
facilities; currency and coin are available from
their Federal Reserve banks, and excess currency
and coin may be delivered to the Federal Reserve
for credit; periodic examinations are performed of
members operating under state charters. These services are provided without charge, as a privilege
of membership.
Banks choosing not to be members have cited a
variety of reasons, but with few exceptions the

"cost of membership" has been given as an important
if not overriding consideration. Member banks are
required to hold capital stock of their Federal
R eserve bank equal to 3 percent of the member bank's
capital. A 6-percent statutory dividend is paid
on such stock; hence, the cost is nominal, except
in periods of exceptionally high interest rates. Of
more importance, member banks must comply with
capital requirements established by t he Federal
Reserve and provide numerous statistical reports to
the Federal Reserve bank.

The continuing erosion of membership.
long a matter of concern to Federal
R eserve officials, has drawn increased
attention from bankers, banking organizations, and the Congress in recent y ears.

The major cost of membership, however, is the
requirement of the Federal Reserve Act that member
banks hold reserves equal to specified proportions
of their demand and time deposits. These reserves
may consist of currency and coin in the member
banks' vaults and deposits with their Federal Reserve
bank. No interest is earned on these reserves. Hence,
if required reserves exceed the cash balances member
banks would normally hold for ordinary operating
needs, member bank earnings are lower than they
would be in the absence of the reserve requirements.
In this article we examine the comparative costs
of reserve requirements for member and nonmember
banks in Texas, New Mexico, Louisiana, and Oklahoma. Nonmember banks in the Southwest are subject to high reserve requirements rela tive to percentages applicable to nonmember banks in other states
and relative to Federal Reserve requirements on
member banks. However, the percentages by themselves do not accurately reflect the relative cost of
reserve requirements because of important differences
in the assets that qualify as reserves and differences

among banks in their need to hold these assets to
conduct banking business.
Table 1


Deposit inl.rv.1
(Mtttton dott .... )

Net demand deposits
Less than $2

$2 to $10.
$10 to $100

$100 to $400
Over $400 ..
Time deposits'
Savings deposits
Other time deposits
Maturing in 30 to 179 days

Less than $5
Over $5
...... . .. .
Maturing in 180 days to 4 years
Maturing in 4 years or more

(Percent or








Demand deposUs s ub retltO .eQuOfementa are groas
dlm ~1Id dep,)l,r. m,nue ClaI> hem, In p'Deeu 0 ' cotttctron
Ind demand due 'Iom dom..Uc DInk •. ~"II'.
menl ...:hedu'n Ir. g.adU'"d. _ .ach de_lIlnllrv.1
1!)t01... to thlll p.,t 01 tha diP_II or each bIInlt
:2 The ave rave 01 __ NIII on ..vlng. end O!her tome depoall,
m~'1 o.;lt Ie .., 3 ~n:..". the mInImum epeel"ed by lew.
SOURCE Fwler;l/ If,..,.,. Buill/;n

Reserve requirements in District states
As shoml in Table 1, reserve requirements for mem·
ber banks are graduated by the amount of net
demand deposits-from 7 percent of the first S2
miUion of net demand deposits to a maximum of 16 Y4
percent of net demand deposits over $400 million.
The structure of reserve requirements on time and
savings deposits at member banks is more complex.
The requirement varies from 1 percent to 6 percent.
depending on the amount of such deposits and their
maturity. But the average of reserves on all time and
savings deposits is about 3 percent. the minimum
level authorized by the Federal Reserve Act.
Reserve percentages for nonmember banks in
Texas. New Mexico, Louisiana, and Oklahoma are all
at least as high as the Federal Reserve requirements
for member banks. In Texas. state reserve require·
ments are considerably higher. Nonmember banks in
Texas are required to hold cash assets equal to 15
percent of total demand deposits and 5 percent of
time and savings deposits. By comparison, the small·
est member banks are required to hold only 7 percent
of net demand deposits and 3 percent of time and
savings deposits. Only the four largest member banks
Review I December 1977

in Texas are subject to the highest marginal rate of
1614 percent of net demand deposits. And even for
these banks. the average reserve requirement is less
than the 15-percent state requirement because lower
requirements apply over the first $400 million of
net demand deposits.

Thus, in tenns of percentages alone. reserve
requirements in Texas clearly favor member banks.
not nonmember banks. Furthennore, member banks
deduct balances "due from" other banks and checks
and other cash items in process of collection (CIPC)
from gross demand deposits before computing re·
quired reserves. Nonmember banks in Texas are
required to hold reserves against gross demand
It is the regulations regarding the composition
of reserve assets that reduce the impact of reserve
requirements for nonmember banks (Table 2). In
Texas. nonmember banks are permitted to include as
reserves checks in the process of clearing through
a clearinghouse association. The volume of these
CIPC items varies greatly with bank size, accounting
for a minor proportion of total deposits at the
smallest nonmember banks to over 6 percent at the

Nonmembers also may include as reserves any
funds on deposit with other conunerciaI banks. Texas
law is not specific as to the type of deposit accoWlt
that can be included as reserves or whether the
reserves can earn interest. But as a matter of
practice. norunember banks are permitted to hold
time and savings deposits as reserves against their own
time and savings deposit liabilities and, thus, earn
interest on this portion of their reserves. Most
banks hold these reserves in certificates of deposit.
However, under legal and regulatory changes made in
late 1975. banks and other businesses are now permitted to hold savings accounts; and a few nonmem·
ber banks maintain reserves in this form.
State reserve percentages in New Mexico are also
higher than Federal Reserve requirements. New
Mexico requires norunernber banks to hold cash as·
sets or U.S. Government securities equal to 12 per·
cent of gross demand deposits and 4 percent of time
and savings deposits.' The percentage requirements
L Only one bank in New MeJ:ico holds a large enough
volume of demand deposits to be subject to the marginal
rate of 12% percent that the Fed applies to net demand
deposits between $100 million and $400 million. And
because lower requirements a pply to the first $100 million. even that bank would be subject to a lower average
requirement as a member bank than the 12 percent
imposed on nonmembers by state law.

Teble 2
Membe, bankl,

S,.te nonme mbe, banks




Basic percentage
Net demand deposits
Time deposits

S~' . m

7% to 1SV
1% to 6%

Lou l,llna

7Yz% to 14%

New Mulco




7% to 16V.%




Deductions from gross
demand deposits
in computing net
deposits subject
to reserve

CIPC (cash
Items In
process of
Balances due
from other

Balances due
from other


Balances due
from other
banks In
excess of


Reserve period






Assets eligible
as reserve assets

Vault cash
Reserves at

Vault cash
Balances due
from other
U.S. Government
and state 01
up to 50% of

Vault cash
Balances due
from other
U.S. Government
up to 50% of

Vault cash
Balances due
tram other
In 24 hours

Vault cash
Balances due
from other
Time deposits
(up to the
amount of
time deposit

SOURCES: FHler.' RII''''' Sul/etl".
SI.,e banking dlPartmenl • .

apply to gross demand deposits and to daily balances
rather than weekly or semimonthly averages. The
task of meeting requirements on a daily basis is
more expensive and complicated than averaging over
longer periods because over time many of the minor
day-to-day fluctuations tend to offset one another.
Reserve requirement percentages in Louisiana are
quite similar to those of the Federal Reserve.
Requirements are graduated over ranges of deposits
and, with one insignificant exception, were the same
as those at member banks until the reduction of
Federal Reserve requirements at the end of December 1976, 2 The calculation of net demand deposits
subject to reserve requirements is also the same
as for member banks. Cash items in process of
collection and balances due from other banks are
subtracted from gross demand deposits in computing
the balance subject to reserve requirements. But
nonmember banks may hold half their reserves in
U.S. Government securities and securities of the
state of Louisiana, in effect cutting in half any
"cost" of holding reserves,

In Oklahoma, reserve percentages at nonmember
banks are the same as those imposed on member
banks by the Federal Reserve. However, in comput·
ing net demand deposits subject to reserve requirements, nonmember banks deduct balances due from
other banks only to the extent that they exceed the
amount required as reserves in the preceding week.
Reserve requirements are based on the average of
deposits over a semimonthly period..
The existence of relatively high reserve percentages
for norunember banks could suggest that the burden
of Federal Reserve requirements for member banks
would be relatively low in the Eleventh District.
However, this turns out not to be the case. Nonmem·
ber banks generally need more than the legally required reserve assets for ordinary operating purposes.

2. The single exception is a 14·percent, rather than 16%·
percent, requirement on net demand deposits over $400
million, and no nonmember bank in Louisiana is large
enough to be subjed to this marginal ~uirement.

Consequently, the relatively high reserve percentages
do not significantly affect the amount of cash assets
held by them. The difference in the cost of reserve
requirements for member and nonmember banks in
the Eleventh District is about the same as the nationwide average. The decline in membership in the District also appears to be about the same.

Nonmember banks generally need more
than the legally required reserue assets
for ordinary operating purposes. Consequently, relatively high reserve percentages in the four southwestern states do
not significantly affect the amount of
cash assets held by them.

Cost of Federal Reserve requirements
in District states
The cost of Federal Reserve requirements to member
banks can be estimated by taking the difference
between the proportions of "nonearning" cash assets
held by member and nonmember banks. The result of
such comparisons nationwide are given in Table 3.
The first component is the larger amount of vault
cash held at member banks. The level of vault cash
held by a nonmember bank is assumed to represent
the necessary working balances for each size bank.
Member banks typically hold more vault cash than
nonmembers of similar size. This is probably because
vault cash qualifies as part of member bank reserves.
The average vault cash held by member banks in excess of the amount held by comparable size nonmember banks nationwide is given in line 1 of Table 3 as
a percentage of total deposits.

3. A bank can only invest funds that have actually been
collected, so due·to balances and due-from balances must
be adjusted to reflect the amount that has been collected
and is available to invest. In this study, 56 percent of
due·to and due·from balances are assumed to be collected.
The same sdjustment was made in Robert E. Knight,
"Comparative Burdens of Federal Reserve Member and
Nonmember Banks," Monthly Review. Federal Reserve
Bank of Kansas City, March 1977, and in John Paulus et
aI., "The Burden of Federal Reserve Membership, NOW
Accounts, and the Payment of Interest on Reserves," a
Federal Reserve Board staff memorandum of June 1977.
4. Due·to balances are deposits or other banks-that is,
deposits that could be withdrawn by other banks.
Review I December 1977

The second component of the imputed cost of Federal Reserve requirements is excess holdings of correspondent-type balances. Member banks receive correspondent-type services from Federal Reserve banks
as well as from correspondent banks. So for member
banks, total correspondent-type balances are calculated as the sum of balances at Federal Reserve banks
and collected balances at commercial banks. The
amount by which this total exceeds collected correspondent balances of nonmember banks of the same
size is assumed to reflect the impact of Federal
Reserve requirements. This is shown in Table 3,line
2, as a percentage of total deposits of member banks. 3
Commercial banks that provide correspondent
banking services to other banks find it advantageous
to be Federal Reserve members. A number of the
correspondent services-such as collecting checks,
transferring money and securities, and obtaining
currency and coin--are facilitated by having access
to Federal Reserve services. Payment to correspondent banks for such services is largely in the fonn of
compensating balances. This is reflected in relatively
larger due-to balances at member than nonmember
banks. 4 We assume, as have some other analysts,
that 25 percent of the collected correspondent
balances, or due-to balances, is available for profit
to the correspondent banks. Thus, 25 percent of the
excess of "due-to's" at member banks over those at
nonmembers of comparable size is assumed to be
attributable to Fed membership. This amount, expressed as a percentage of total deposits, is assumed
to offset imputed "costs" of Federal Reserve
The net "cost" is obtained by adding the first two
components and subtracting the third. Estimates of
this cost for various size banks nationwide are given
in line 4 of Table 3. In most size categories it is
the income forgone on 2 Y.z to 3 percent of total
deposits. However, at the very largest member banks,
the ability to attract correspondent balances reduces
this burden significantly.
The cost in the four states of the Eleventh District
is given in Table 4. Individual items differ from
the national averages, but the overall burden is
roughly comparable.
Vault Cash. Both member and nonmember banks
in the Eleventh District hold vault cash in amounts
generally comparable to the nationwide averages.
Louisiana banks and small Texas banks hold slightly
more than the national averages, whereas Oklahoma
and New Mexico banks and large Texas banks hold
slightly less.
Also corresponding to nationwide patterns, member banks hold slightly larger proportions of vault

Table 3
(Percentage 01 total deposits as of December 31, 1976)



. .. ..................

1. Excess vault cash
2. Excess correspondent-type balances
3. Less: Excess due to balances avaifable
for profit at member banks




............ ...............

TOI" d,poe!l alII (MInion doll,,.)










l UI
























SOURCES: Feder.' OfoPOIIt InlUrlnce COll'oretlon.
F,dersl RileI'Ve Blnk 01 Daltll.

cash than nonmembers. Nonmember banks have an
economic incentive to deposit excess vault cash at a
correspondent bank, where in addition to serving as
reserves in meeting state reserve requirements, the
deposit earns an implicit return in the fonn of
services from the correspondent. Member banks, on
the other hand, have no particular incentive to economize on vault cash. Both vault cash and deposits
at Federal Reserve banks qualify as reserves. Neither
earns interest.
On balance, the excess vault cash at member banks
in the Southwest, as shown in line 1 of Table 4,
is roughly comparable to nationwide averages and
amounts to a fairly minor portion of the total cost
of Federal Reserve requirements.
Due-from balances. Both member and nonmember
banks hold correspondent balances with other banks
as working balances for clearing checks, transferring
money, and other ordinary banking business. In addi·
tion, banks usually hold balances with correspondent banks as compensation for less tangible services,
such as investment advice and help from expert
For the nation as a whole, nonmember banks hold
balances due from other banks that equal between 5
and 9 percent of their total deposits. On average,
nonmember banks in all four states of the Eleventh
District hold more than these due-from balances than
do banks in the rest of the nation, Banks in Texas
hold significantly higher average balances-ranging
from 9 to 1112 percent of total deposits-while
banks in the other three states hold only slightly
If these holdings were attributable to state reserve
requirements, the relative burden of Federal Reserve
requirements could be significantly reduced and per·
haps even eliminated. However, three types of evi18

dence suggest that the state reserve requirements are
not responsible for the higher than average due-from
balances at most nonmember banks. First is the existence of a large volume of excess reserves. Even
in Texas, with the highest percentage reserve require·
ments, nonmember banks generally hold far more
cash assets than the law requires. 6 As shown in Table
5, more than two-thirds of these banks hold at least
half again as many reserves as required.
Second, nonmember banks in Texas have the abi!·
ity if they were to exercise it, to eliminate all the
burden of state reserve requirements by creating
the necessary reserves through reciprocal balances.
Two banks could simply credit each other's accounts.
Bank A could increase the balances due to Bank B,
while Bank B increased the amount due to Bank A.
Of course, each bank would have to hold reserves on
the newly created balances, but 85 percent of such
balances would be available to meet reserve requirements. The fact that reciprocal balances are not
maintained in significant amounts suggests that
banks hold enough deposits for banking purposes
to meet the requirement.
A third piece of evidence suggesting that rela·
tively high levels of state reserve requirements in
the Southwest do not constrain the activities of
nonmembers is the relatively high demand for correspondent balances of member banks. On average,
member banks in all four southwestern states hold

5. Corroborating evidence for other states is presented in
Paulus et a I., "The Burden of Federal Reserve Membership," pp. 77-85. Another study of data from all states
suggests that state reserve requirements could constrain
the cash holdings of nonmember banks to a minor extent.
See Lawrence G. Goldberg and John T. Rose, "Do State
Reserve Requirements Matter?" in Journal of Bank
Research 8, no. 1 (Spring 1977).

Table 4
(Percentage of total deposits as 01 December 31, 1976)

. "

Totat deposit sl u (MUlion dolla..)




1. Excess vault cash

. . . . . .. . . . . .
.. ...... ........ ..........
............. ..........

New Mexico
2. Excess correspondent·type balances
New Mexico ......... ........
Texas .. ........
3. less: Excess due to balances available
for profit at member banks
New Mexico
.. ...... ... . ...
Oklahoma . .
Texas .....
New Mexico
Texas . .
















.. ........ .. .......
... .. ... .... .. ........




....... ... ...........




.. .. ....................
.................. ........
...... ....... ... .... ......







n.a .

.. .......... ....












n.a .




n.a .



n.a. Data not lValiabl. from anough bank, to provide a dependable eomparilOn.
SOURCES: Fadef,l Oapo'it In,uranee CorpOratiOn.
Fad ..ral Reltt .... Bllllk of Dall ....

about 2 percentage points more correspondent-type
balances than member banks in the nation-the same
amount of extra balances that nonmembers hold.
Since state requirements do not apply at member
banks, this is additional evidence that the higher demand for correspondent balances in the Southwest is
not related to state reserve requirements.
To calculate the excess correspondent-type balances of member banks, their collected balances due
from other conunercial banks and their reserves at
Federal Reserve banks are added together. The
amount by which this total exceeds the collected
"due· from's" of nonmember banks of the same size
equals the excess, which is the largest component of
the "cost" of Federal Reserve requirements. It is
shown in line 2 of Table 4.
As one might expect, considering the relatively
heavy demand for correspondent balances in the
Southwest, this component of the reserve cost turns
out to be as high in the four states as the nationwide average. The reason is the higher than average
due-from's at nonmember banks are matched by
higher than average due-from's at member banks.
Due-to balances. If banks in the Southwest hold
higher than average balances due from other banks,
Review I December 1977

Results for the four southwestern states
are similar, on average, to those for the
nation but show somewhat more variation. The cost of Federal Reserve requirements for most of these member banks is
the income forgone on 2 to 4 percent of
total deposits.

the other side of the coin is higher than average
balances due to other banks. Banks in Texas have
more due-ta's than average banks in the nation in
nearly every size category, With a few minor exceptions, these same results hold for Louisiana, New
Mexico, and Oklahoma. (Because a large number of
correspondent balances are deposited at banks in
New York City and other financial centers, total
due-to's fall significantly short of due·from's in
all four states.)
As member banks have an advantage in providing
correspondent banking services and attracting carre·
spondent balances, measurement of the cost of Federal Reserve requirements includes an adjustment to


member banks, there might even be no additional
cost attached to Federal Reserve requirements.

Table 5
I, ptlrClnt


01 t.M.......

rltQulred 1011n.'
1110111'11 IIIPOIl,.


nonmember Cumu ',, (vt

Less Ihan 0 .. ...
10% ..
10% 10 20%
20% 10 30%
30% to 40%
40% to 50%
50% 10 60%
60% to 70%
70% to 80%
80% to 90%
90% to 100% ....
100% 10 110% ..
110% to 120% .•.
120% to 130% ..
130% or more . . .

o 10



SOURCES, o..PQsil Insurance

Federal Rcur.ft

B~n k

01 Oall,.

reflect the additional resources of member banks
arising from these due-to balances. Following earlier
studies, it is assumed that 75 percent of collected
due-to's is enough to compensate the correspondent
bank for services provided the respondent so that 25
percent of collected due-to's is available for profit.
Thus, the burden of member banks is reduced by 25
percent of the excess of collected due-to's at member
banks over those at nonmembers of comparable siZe.
The resulting cost of Federal Reserve requirements
is given for each state in line 4 of Table 4.
Results for the four southwestern states are similar, on average, to those for the nation but show
somewhat more variation. The cost for most of these
member banks is the income forgone on 2 to 4 percent
of total deposits.
Summary and conclusion
Reserve requirement percentages for nonmember
banks in Louisiana, New Mexico, Oklahoma, and
Texas are higher than the average percentages for
nonmember banks nationwide. Because of this, it is
possible that the burden of Federal Reserve requirements in these states might be lower than elsewhere.
If the state reserve requirements constrained nonmembers to hold as large a proportion of cash assets as

However, this is not the case. Nonmember banks
in the four states tend to hold more correspondent
balances than nonmembers in the nation, but so do
member banks. The extra balances appear to be necessary to the conduct of banking business rather
than being a constraint imposed by state reserve
requirements. Member banks maintain higher than
average due-from's even while holding balances at
Federal Reserve banks. They do not behave as
though Federal Reserve balances are a good substitute for correspondent balances.
With member and nonmember banks both holding
higher due·from's than the nationwide average, the
difference is roughly the same as the average difference nationwide. The resulting cost of Federal
Reserve requirements is also about average.
The extent of this cost can be translated from
percentage of deposits, as calculated above, to dollar amounts. Taking 2 percent of deposits as an estimate of t he burden and 6 percent as an estimate of
the return the resources would yield if banks could
invest them in short-term,low-risk, readily marketable assets, the cost to member banks nationwide is
roughly $700 million. 6 About $52 million of this
would be attributable to member banks in Texas, and
about $19 million to member banks in Louisiana,
New Mexico, and Oklahoma. 1 Of course, the cost rises
or falls with the rate of return on securities.
Higher interest rates in recent years have raised
the imputed cost of idle reserves, with a predictable
impact on membership in the Federal Reserve System. In the past ten years, membership has declined

6. The Paulus study estimated the earnings burden. before
adju'Jt ment to account for the implicit value oC the discount window for member banks. to lie between $531
millio n a nd $690 million. The lower estimate assumes
excess cash as!lets could be invested at an interest rate of
5 percent. while the higher one assumes a rate of 61Al
percent. After adjustment, t he burden was estimated to
lie between $431 million and $590 million. See "The
Burden of Federal Reserve Membership," pp. 21-23.
7. One study of a sample of Texas banks suggests that the
cost of Federal Reserve requirements is passed on to the
communities served by member banks. The study found
t hat membership did result in larger holdings of cash
assets a nd fewer loans. H owever, the reduction in revenue
resu lting Crom this asset alloc.'l.tion was genera lly offset by
higher prices charged {or bank services. As a result, profitability was little affected by membership. See Donald
R. Fra!ler, Peter S. Rose, a nd Gary L. Schugard. "Federal
Reserve Membership and Bank Perlonnance : The Evi·
dence Cra m T e xas," Journal of Finance 30, no. 2 (May

substantially. The proportion of total bank. deposits
accounted for by member banks has dropped from 84
percent to 75 percent nationwide. In the four south·
western states the proportion of total deposits that
is held by member banks has dropped from 79 per·
cent to less than 68 percent. The roughly comparable
decreases are consistent with the evidence of an
average membership burden in the Southwest.

Federal Reserve discount window, erosion of membership complicates the Federal Reserve's role as
lender of last resort in times of banking emergencies.

The erosion of membership is of concern to the
Federal Reserve principally because it affects or
threatens to affect monetary policy. The more non·
member banks there are, the more difficult it is for
the Federru Reserve to Ct:Itimate the amount of open
market operations that will be necessary to maintain
the Federru fun ds rate within a given range or to
achieve desired growth rates of money. A higher proportion of nonmember banks also ma kes it more
difficult for the Federal Reserve to adopt a proposed
alternative approach to monetary control that would
focus on the level of bank reserves rather than on
the Federal funds rate.

A number of ways to alleviate the membership burden are being considered by Federal Reserve officials,
bankers, banking organizations, and members of the
Congress. Each of the proposals provides member
banks a way to earn interest on some of the assets
they now hold as reserves. Reserve requirements
could be reduced or eliminated, although there is
some concern that this could introduce additional
instability in money growth. Alternatively. member
banks could be authorized to hold a portion of
their required reserves in U .8. Government securities
or other interest·bearing securities. as is done by
many states.' Another alternative, which has been
detailed in legislation currently before the Congress,
would authorize the Federal Reserve to pay interest
on reserve balances at Reserve banks and would
authorize reductions in reserve requirement percentages (or smaller member banks.

The threat of further declines in membership also
constrains the Federal Reserve's ability to use
changes in reserve requirements in the conduct of
monetary policy. Since the "cost" of the higher
requirements falls on member banks, the resulting
inequity and threat of further decreases in member·
ship deter use of this instrument. In addition, since
nonmember banks do not have equal access to the

8. One tuch propoeal that could be implemented without
threatening any 1088 of monetary control has been detailed
in "A Proposal for Enhancing the Attractiveness of Membel'8hip in the Federal Reserve System," a report by the
Board of Directol"l of the FederaJ Reserve Bank of Kansas
City, June 1977.

Review I December 1977


Statistical Releases of Federal Reserve Bank of Dallas

Available without charge from: Research Department
Federal Reserve Bank of Dallas
Station K
Dall.a.s, Texas 75222
Commercial and Industrial Loans Outstanding at Weekly Reporting Commercial Banks, by
Industry-Eleventh Federal Reserve District
Condition Report of Weekly Reporting Commercial Banks-Eleventh Federal Reserve District
Weekly Department Store Sales in Eleventh District Areas (Total Eleventh District, Austin,
Dallas, EI Paso, Houston, and San Antonio)
Commercial and Industrial Term Loans Outstanding at Weekly Reporting Commercial Banks,
by Industry-Eleventh Federal Reserve District
Louisiana Industrial Production
Texas Industrial Production and Manufacturing Capacity Utilization

Condition and Income of Member Banks-Eleventh Federal Reserve District
Employees' Salary Survey (Member banks in Eleventh District)
Officers' Salary Survey (Member banks in Eleventh District)
Operating Ratios of Commercial Banks-Eleventh Federal Reserve District
Survey of Trust Department Income and Expenses (Commercial banks in Eleventh District)


New member banks
First National Bank, Copperas Cove, Texas, a newly organized institution located
in the territory served by the Head Office of the Federal Reserve Bank of Dallas,
opened for business November 16, 1977, as a member of the Federal Reserve
System. The new member bank opened with capital of $500,000 and surplus of
5500,000. The officers are: J. C. Sloan, Chairman of the Board; David I. Mountain,
President: Kenneth J. Ambler, Vice President: and B. M. Sutton, Cashier.
South Texas National Bank, Laredo, Texas, a newly organized institution located
in the territory served by the San Antonio Branch of the Federal Reserve Bank of
Dallas, opened for business December 5, 1977, as a member of the Federal Reserve
System. The new bank opened with capital of $550,000, surplus of $550,000, and
undivided profits of $550,000. The officers are: Raymond M. Keck, Jr., Chairman of
the Board: Edward H. Starr, Jr., President; Alnoso S. Rios, Vice President: and
Nella R. C..... Cashier and Secretary.

New par banks
Bank of Pleasant Hill, Pleasant Hill, Louisiana, an insured nonmember bank
located in the territory served by the Head Office of the Federal Reserve Bank of
Dallas, began remitting at par November 10, 1977. The officers are: James A.
Ta rver, President; W. O. Woods, Vice President; Lurline W. Dykes, Cashier.
Sabine State Bank & Trust Company, Many, Louisiana, an insured nonmember
bank located in the territory served by the Head Office of the Federal Reserve Bank
of Dallas, and its branches at Zwolle and F1orien, Louisiana, began remitting at
par November 14, 1977. The officers are: Horace Tompkins. President; James R.
Cole, Executive Vice P resident; Don Zick, Vice President, Zwolle Branch; Stan
Russell, Cashier; C. E. Smith. Assistant Cashier: Beatrice Hughes, Assistant
Cashier: Kenneth Freeman. Assistant Cashier, Florien Branch; and H . A. Prewitt,
Assistant Cashier.

Revlew I December 1977



Federal Reserve Bank of Dallas
December 1977

Eleventh District Business Highlights

Growth of total108ns at large
banks in the Eleventh District in
1977 may be the strongest in a
number afyears. Cumulative data
show that total1080s advanced 17
percent in the first 11 months of the
year. That was more than twice the
average growth rate for comparable
periods in the preceding ten years.
On a seasonal basis, the rate of
growth in total bank lending picks
up considerably in the final month
of the year. Consumer borrowing
during the Christmas season and
borrowing by businesses to meet
quarterly tax and dividend payments and operating expenses are
usually very strong sources of loan
demand toward year-end. In view of
the high rate of growth in borrowing already recorded and the
expected seasonal increase, it is
likely that 1977 will be a record
year for borrowing at large banks.
The bank lending category with
the fastest rate of increase in 1977
has been real estate loans. Reflect-

ing the rapid growth in District
construction activity and a greater
emphasis on real estate lending,
these loans have increased at record
or near-record rates alm06t every
month and advanced 32 percent
through November. Moreover, they
accounted for slightly more than a
fourth of the growth in total loans.
Consumer loans at large banks
advanced a record 23 percent
through November. That was
almost twice as fast as the previous
record growth rate for the period
and more than three times the
average growth for comparable
periods in the past ten years.
As in the nation, consumers have
accounted for much of the growth
in the District economy in 1977.
Department store sales advanced
about 14 percent in the first 11
months, and sales of new automobiles and other consumer durable
goods have also been strong. Thus,
use of bank credit cards has
increased sharply, and consumers
stepped up their bank borrowings



19 68-76 AVERAGE









Loans to nonbank financial institutions advanced 24 percent in the
first 11 months of the year. That
compared with an average growth
of less than 4 percent in corresponding periods of the prior ten years.
Despite sizable deposit inflows at
savings and loan associations, the
sharp demand for residential mortgages led these institutions to
increase their short-tenn borrowing
from large banks. Finance companies also stepped up borrowings at
banks to accommodate their own
higher level of loan demand.
Loans to businesses also rose at a
near-record rate in the first 11
months of the year, even though
they declined moderately in the
first three months. Spurred by
higher demand from transportation, public utilities, construction, and mining industries, total
business loans through November
rose almost 13 percent. That was
more than twice the average growth
for the same period in the previous
ten years and almost matched the
record rate of growth for the first 11
months of 1972.
In the past ten years, total loans
at the District's largest banks
increased an average of 2.6 percent
in December. With almost all major
loan categories sharing in the
overall advance in total loans in
1977 and with continued strong
growth in the District economy in
prospect, the growth in total loans
at large banks should be at a record
rate for the year.
Consumer loan demand is likely
to remain well above nonnal,
reflecting continued strength in
automobile and retail sales. And
business loans could account for a
larger share of the overall increase
in total loans in December.
(Continued on back page)





". - - - - - - - - - -



". - - - - - - - - - -



, ---------~1 50 .9

....":·" :.r '49 . 9

140 -

r ... ...f

. I
: ..• ,
..•.•. ..., ,



130 -

120 -







110 -

' 393




r ' ! 120.8

.. ../


....•...../ ....-~.......






.'1 ....... /



..................,r-:... "'. i:.

120 -

. ····f; '!i...



•..\/,·\.••.. 90.7

90 -



8O .-=:-r-:-::::-0:::::-,.

" • ...,.l..I:=:-r-=.,-.,.,.,.:::-T'
1 ~75


SOURCES: Bon d 01 Gov.,nor•• F. d.,. 1,v. Sy. tem.
F.d.,al R... rv. Bank of DaU ...




' ,000


-----:::===::-:c:-:c:-:-:::::-:::----r--- '00



.00 ---------------,




'" ••











1111" •

!I!! hili! Iii; '"

: ='11 !II!!I hilI! !II!
o i I: I: Ii j;:i!! j: ii Ii !I:::;:;; iiI:::











~--::.-' ...... '




.................. ..
--.-~-~- -=-~-~-.-~-~o.=~~~------L-.


600 -





200........... -"
---.-. =.=.~.~.~.
.~~~~~~=.=~~.~--. . . . . . .
. . . ---. . .
.. .


............ '978"' -



-,OO ---,r__r,-"-~'-,r__r,-"-"r__r'-T'-',-"r­


SOURCES: Bu,ea", 01 Bu. in ... R••• a,ch, Univeraity of TeK• •.
Fe d. ,al Home Loa n Bank of Littl. Rock.




2. . - - - -_ _ _ _ _ _ _ _ ____

-...... -...
'80-_... "... ..
:: =~~.~:.::.:~.~ ........ . .

- U.S,
..... DALLAS

, ...

. ,.....•..•...


SOURCE: U.S. Department 01 A9rlcllltu, •.

... 192.7

, 182,4













,,/ 197 8

",- - '






-'" -,----::::::---,---:c=---.--=:---.


: __~.'~·.::"~;~·r: "·__~,---rj;-·-;r;-~· ·'T---~"--~-:"r'--.'--',. .
.'.c·'c:·'" ·
. - · ·~r-'-· ', '

.------~~----------------TIME DEPOSITS


---- .....

-- ................



. ....... ....






e. ·
............. ........ . .

"050 -



150 -




1~.::;:;-r.-::.":'..::-.:·;':'·::-·,..·.:'·';':·::·· ..ig75


2.' - - - - - - - - - - - - - - - 1.8_

1.5 _



1.2 -








••... -;


, __~~==?.~~~.~~Z- "-- ' --... ..~ ... ~-.
.... ~~







, II





1. Loul".n., N.w M.. lco, Okl.hom., .nd Te"...
SOURCE: F. W. Dodge, McGr.w-HiII, Inc.




2.2 - - - - - - - - - - - - - - - - - - - - - - - - - - -

8." _
8.2 _

••~................\ •••••••/ . \...........~




7 .8 -



8 .0 -




'-' ,---::~-,---:::::-,---:-=-,1975



Loul.lene, Naw M..ico, Old.ho"" , .nd T ......
SOURCE: Slate '",plo~.nt''''

1.9 -



SOURCE: U.s. Dep.rt",enl of Comm.rca.

Business e:l.penditures are likely
to continue to rise more rapidly
than funds available from internal
sources-largely profits and depredation allowances. Moreover, the
balance shee.ts of many businesses
are in better condition to support
short-tenn borrowing. Liquidity
positions are much improved over
most recent years, as corporations
acquired substantial long-term
funds in 1975, 1976, and early 1977
to repay short-term debt and
increase their liquid assets. These
improved liquidity positions may
make short-term bank loans a more
attractive source of funds to corporations.

Although preliminary figures show
industrial production in Texas
slipped 0.2 percent in October,
fairly steady gains in output have
been made since the trough of the
recession in February 1975. The
only significant setback during the
recovery has been in the second
quarter of 1976, when industrial
output fell slightly for three consecutive months.
At 141.2 percent of the 1967 base,
industrial production in October
was 5.7 percent above the level a
year earlier. The biggest contributor to that gain was manufacturing,
which makes up two-thirds of the
production index. Durable goods
manufacturing rose 8.9 percent,
while the production of nondurable
goods rose 5.5 percent.
The biggest increase compared
with a year earlier-l1.7 percentwas posted by public utilities. But
since that sector comprises 4 percent of the index, the increase made
only a small contribution to the
overall gain in industrial production.
Mining had the smallest
increase-O.8 percent-of the three
major sectors. And although its
share of the production index is 29
percent, mining added little to the
gain in total output since October

Much of the increase in durable
goods manufacturing has been in
materials and equipment supplied
to construction and drilling industries. And a significant portion of

that increase has come since the fall
of 1976.
The boom in construction activity has raised the output levels of
furniture and fixtures, lumber and
wood products, and stone, clay, and
glass products. And the expanded
search for crude oil and natural gas
has continued to boost the production of oil field equipment, which
accounts for nearly half of all nonelectrical machinery output.
Production of fabricated metals
and electrical machinery has also
grown substantially. And a share of
that growth has been due to the
increased demands of the construction and drilling industries, although that share is not as great as
in the case of nonelectrical machin-

Production of primary metals
and transportation equipment
industries since the fall of 1976 has
been mixed. Output of steel, used in
both the construction and drilling
industries, has bolstered the primary metals industry, while production of nonferrous metals has
lagged. Output of transportation
equipment slipped throughout most
of the period; only in September
and October were significant gains
In nondurable goods manufacturing, the food and kindred. products industry has recorded the
steadiest gains in production since
October 1976. But output in printing and publishing and in petroleum refining has also shown steady
upward trends.
The patterns of output of the
apparel. paper, and chemical industries have been uneven. Apparel
production, for example, rose
sharply in late 1976 and showed no
further strength in 1977. Paper production has also had its ups and
downs. Paper output in October,
although above a year earlier. was
down from the first quarter of this
year. Similarly, chemical production in October was ahead of a year
earlier but down from last spring.

The only nondurable goods
industry that has drifted down is
textiles. Output of the industry
grew in the first year of the
l"ecovery but has been slipping for
nearly two years.
Much of the sharp rise in output
of public utilities was weatherrelated. Last winter in Texas was
colder than normal, and heating
requirements raised fuel
demands-especially for natural gas
and, to a lesser extent. electricity.
And with record-high temperatures
last summer, air-conditioning
demands increased electric loads
The smaller gain in mining output resulted as increases in drilling
activity and in production of metal,
stone, clay, and earth minerals
barely offset declines in crude oil
and natural gas production. Despite
the steady climb in drilling activity,
there is little indication that the
decline in oil and gas production
will abate.
• Total deposits at member banks
in the Eleventh District continued
to increase sharply in October, as
both demand deposits and time and
savings deposits were up.
• The value of total building contracts, seasonally adjusted, in the
four states of the Eleventh District
fell in October after four months of
The number of housing starts in
Texas also declined from September
and totaled 11,400 units, seasonally
adjusted. in October.
• Total employment in the Eleventh District states in October rose
0.4 percent over the previous
month. Unemployment fell to 5.3
percent of the civilian labor force
from 5.6 percent in September.
• Consumer prices in Houston rose
0.6 percent from July to October to
a level 5.9 percent above a year earlier. The latest rise was the smallest
increase in that dty's consumer
price index in nearly five years.

Eleventh District Business Highlights is publi!>hed monthly by
the Research Department. This issue of Highlights was prepared by Mary Grandstaff, Jean Adeler. and Carolyn Hamilton
under the supervision of Edward L. McClelland.