View original document

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

The R e v i e w is published monthly by the Research Department of the Federal Reserve Bank of
St. Louis. Single-copy subscriptions are available to the public free of charge. Mail requests for
subscriptions, back issues, or address changes to: Research Department, Federal Reserve Bank of
St. Louis, P.O. Box 442, St. Louis, Missouri 63166.
Articles herein may be reprinted provided the source is credited. Please provide the Bank’s Re­
search Department with a copy of reprinted material.




The Value-Added T ax—A Review
of the Issues
R. W. HAFER and MICHAEL E. TREBING

1 ^ . ECENTLY many groups have called for major
tax reform. These groups maintain that the existing
U.S. tax system has altered economic incentives and
negatively influenced the performance of the econ­
omy. One possible solution to this problem is for the
United States to adopt a value-added tax (VAT)
similar to that used by European governments.1 This
article provides a general economic framework for
examining the current debate over the introduction of
a VAT into the U.S. tax system.

BACKGROUND
Value added is the difference between a firm’s re­
ceipts from the sale of a product and the payment
iSee, for example, “Tax Restructuring Act of 1979” proposed
by Rep. Al Ullman, chairman of the Joint Ways and Means
Committee. Fo r an analysis of this proposal and a discussion
of the value-added tax, see also “Special Supplement,” Daily
Report for Executives (October 22, 1 9 7 9 ), pp. 2-26.

for the resources or raw materials used in producing
it. In other words, value added is equivalent to a
firm’s payment to the factors of production — land,
labor, and capital — in the form of rent, wages, in­
terest, and profits.2 These payments represent the
“base” to which a VAT would be applied.
A simplified example of how value-added can be
determined is shown in table 1, which illustrates a
hypothetical four-step process of converting wheat
into a loaf of bread. The miller purchases wheat from
the farmer at a price of 10 cents and sells the result­
ing flour to the baker for 20 cents. Since one-half of
the price to the baker represents the farmer’s receipts,
10 cents of the miller’s receipts represent the value
added to the raw wheat during this stage of the pro­
duction process. After converting the flour into bread,
the baker sells the loaf of bread to the retailer for
2For a brief discussion of the value-added concept, see John M.
Blair, Economic Concentration (N ew York: Harcourt, Brace,
Jovanovich, Inc., 1 9 7 2 ), pp. 68-69.

Table 1

Calculating the Value Added
Stage
1
Farmer

.10

2
Miller

.10

+

.10

3
Baker

.10

+

.10

+

.10

4
Retailer

.10

+

.10

+

.10

+

Receipts

Value-Added

.10

.10

.20

.10 ( = .20 - .10)

.30

.10 ( = .30 - .20)

.10 = .40

.10 ( = .40 - .30)

Final sales price to consumer = $.40




Total value added = $.40

3

JA N U A R Y

F E D E R A L R E S E R V E BANK O F ST. LO U IS

30 cents. Since the cost of the flour was 20 cents,
the value added at this stage (the baker’s) is again
10 cents. Finally, out of the 40-cent retail price for
the loaf, the remaining 10 cents represents the re­
tailer’s value added.
The usefulness of this example is twofold. First, it
illustrates the fact that the final retail price to the
consumer is really nothing more than the sum o f the
value added at each stage of production. Second, it
provides a format by which a sales tax can be dis­
tinguished from a VAT. A sales tax involves a single
payment by the consumer at the retail level and is
levied at a constant rate on the purchase price. The
VAT, on the other hand, is assessed at each stage of
production. If the sales tax rate and the VAT rate are
equal, the total tax revenue in the above example is
the same irrespective of the collection process.
There are two basic methods of calculating the
VAT.3 In the additive method, the tax base for the
VAT is the sum of the firm’s payments to the factors
used in producing the good. Under this scheme, the
VAT rate, say 10 percent, is applied to the firm’s
costs in terms of wages, interest, rents, and profits.
Table 2 illustrates a hypothetical example of how the
additive procedure is used to calculate the VAT due.
Table 3 provides the information needed to calcu­
late the VAT payable under the second method,
the subtractive procedure. Essentially, this approach
makes the base to which the VAT rate is applied
equal to the difference between a firm’s sales receipts
and its cost of production (again, its value added).
The actual tax liability using this procedure is iden­
tical to that derived using the additive approach.

The Tax Base and Its Calculation
As noted before, the VAT base is usually calculated
as the difference between a firm’s sales receipts and
its cost of purchases from other firms. A debate exists,
however, over whether all of a firm’s costs should be
included in the calculation of the base or whether
special exemptions should be made. Exemptions from
the tax base are important because they determine
many of the economic effects of a VAT. Of particular
concern is the treatment of capital goods. There are
three general approaches to the treatment of capital
3The following discussion is found in a number of writings.
See, for example, Advisory Commission on Intergovernmental
Relations, T he Value-Added Tax and Alternative Sources
of Federal Revenue (Washington, D .C .: ACIR, 1 9 7 3 ); John
F . Due, “The Value-Added Tax,” The W estern Economic
Journal (Spring 1 9 6 5 ), pp. 165-71; and, Richard W . Lindholm, Value-Added Tax and Other Tax Reforms (Chicago:
Nelson-Hall, 1 9 7 6 ), pp. 30-35.


4


1980

Table 2

Calculating VAT Liability Via
Additive Procedure
$500,000

W ages
Interest

50,000

Rent

25,000

Profit

25,000
$600,000

Total
Value added = $600,000
Tax payable =

$600,000 X 10% (VAT rate) =

$60,000

SOURCE: Richard W . Lindholm, Value-Added Tax and
Other Tax Reforms (Chicago: Nelson-Hall,
1 9 7 6 ), pp. 30-31.

goods: the gross product, the income, and the con­
sumption approaches.
Under the gross product approach, firms are not
permitted to deduct the purchase price of capital
goods when calculating their VAT base, nor are
they allowed to deduct the depreciation of existing
capital. The tax base at each stage of production,
therefore, includes wages, interest, rent, profits, and
depreciation. Since there is no deduction for either
capital goods depreciation or their purchase, the VAT
applies to all income earned in the production of the
firm’s output. Thus, in an economy-wide sense, the
tax base is associated with the value of gross national
product — the total current market value of all goods
and services produced or the total income derived
from their production.
The income approach, on the other hand, does
allow firms to deduct capital depreciation from their
tax base. Since firms “use up” a percentage of their
existing capital stock each year, this approach taxes
firms on their net instead of gross income. The tax
base of the income-type VAT is analogous, therefore,
to the net national product of the economy.4
4An algebraic expression similar to the national income account
identities taught in elementary economic textbooks can aid in
distinguishing the income base of each type of VAT discussed.
Ignoring government expenditures, taxes, and the foreign sec­
tor, the expression
GNP = C + I
shows the equivalence of total output or income earned in the
economy ana the sum of consumption ( C ) and investment
( I ) expenditures. W ith no VAT exemptions, GNP is the base
of the gross income VAT. W e define
I - D = I»
NNP = In + C
NY = NNP,
where In is net investment, D is depreciation, NY is national

JA N U A RY

F E D E R A L R E S E R V E BAN K O F ST. L O U IS

1980

CRITERIA FOR EVALUATING THE VAT

Table 3
Calculating VAT Liability Via
Subtractive Procedure
Total sales of goods
and services

$1,000,000

Total interest, dividends, and
rents received

200,000
$1,200,000

Total (A)

$

T axes

200,000

Purchase of materials, services,
power, and capital on which
VAT has been paid

300,000

Interest, dividends, and rents
paid to other firms

100,000
600,000

Value added = A - B
= $600,000
Tax payable =

$600,000 X 10% (VAT rate) = $60,000

SOURCE: Lindholm, Value-Added
Reforms, pp. 30-31.

Economists use several criteria to judge the desira­
bility of a given tax or tax system. These criteria in­
volve both efficiency and equity considerations.

Neutrality
$

Total (B )

The effect of taxation on economic efficiency is
linked to the concept of potential output. Potential
output, as it relates to a market-type economy, may
vary according to incentives available in the market­
place. The economist defines these incentives in terms
of income and relative prices. To the extent that taxes
can alter economic incentives or the return to pro­
ductive activity, they induce changes in economic
behavior.

Tax and Other Tax

The third approach to the treatment of capital
goods, called the consumption approach, is the most
widely used VAT. It is the type that much of Europe
has adopted, that Congress is currently considering,
and that will be analyzed for the remainder of this
article. Under this scheme, each firm may deduct its
capital expenditures on plant and equipment in addi­
tion to depreciation. The tax base becomes the firm’s
gross receipts less its purchases of materials and capi­
tal outlays (plant and equipment).5 In an aggregate
sense, the VAT base under the consumption approach
corresponds to the output of consumer goods or,
equivalently, income earned in producing consumer

income, and NNP is net national product. The base of the
income-type VAT is GNP - D = NNP = NY, and the base
of the consumption type VAT is GNP - I = C, where I =
In + D. See Richard Musgrave and Peggy Musgrave, Public
Finance in Theory and Practice (N ew York: McGraw-Hill,
1 9 7 3 ), p. 338.

Price theory suggests that prices established in a
competitive market are reflections of both consumer
preference for goods and services and the least-cost
combination of inputs used to produce those goods.
A tax system is considered optimal if it does not
interfere with the price allocation mechanism estab­
lished in the market — that is, it is neutral toward
economic behavior. Although all taxes distort or alter
economic behavior to some degree, certain taxes can
be judged “superior” on efficiency grounds.7 VAT pro­
ponents claim that the value-added tax would achieve
the objectives of neutrality toward both economic be­
havior and allocative efficiency.
A tax on each firm’s value added can be thought of
as a proportional tax on the firm’s use of the factors of
production. If each firm combines land, labor, and
capital in the most efficient (least-cost) manner, re­
sources are bid into their most productive use. Pay­
ments to these factors in a competitive market are
approximately equal to each factor’s contribution
to the market value of each firm’s output. Thus,
a uniform VAT on all firms imposes the same
proportional tax cost — with respect to each firm’s
payments to the sum of factor payments (value
added) — and is, therefore, neutral toward the choice
of production methods or the use of productive
resources.

5Note that the dollar amounts of the income and consumption
bases are equivalent over time. “The main difference between
the two types is that, under the consumption concept, the
tax base is smaller the first year and slightly larger in the
following years. The total base added over the full deprecia­
tion period is identical under both methods.” Tax Foundation,
“A Value-Added Tax for The United States?” (June 1 9 7 9 ),
p. 1.

value added by capital is subject to the tax. F o r a given
interest rate, the tax on the goods produced with capital
raises the amount of net income firms must derive from a
capital asset in order to justify its purchase. In other words,
the tax reduces the demand for funds and, other things re­
maining the same, the interest earned by savers. See Charles
E . McLure, Jr. and Norman B. Ture, Value-Added Tax: Two
Views (Washington, D. C .: American Enterprise Institute for
Public Policy Research, November 1 9 7 2 ), pp. 88-92.

6Norman Ture argues that the “consumption-type” label has
been misnamed. The name promotes the view that it is a
tax exclusively on consumption. Savers, who directly or in­
directly are owners of capital, do not escape the tax because

7Fo r a treatment of the welfare costs of taxation, see Arnold
C. Harberger, “Taxation, Resource Allocation, and Welfare,”
Taxation and W elfare (Boston: Little, Brown and Company,
1 9 7 4 ).




5

F E D E R A L R E S E R V E BAN K O F ST. L O U IS

JA N U A RY

1980

VAT: T he European Experience
T h e V A T is by no m eans a new idea. Follow ing W orld
W a r I, the possibility of im plem enting a V A T -type
system w as discussed in certain E u ro p ean countries and
in the U nited S tates.1 A lthough th e d eb ate continued
over th e n ext several d ecad es, it was not until the 1 9 5 0 s
th at a V A T system w as in trod u ced in F ra n c e and in th e
state of M ichigan.2

Table A1

VAT Implemented
Country
Belgium

Effective
Year
1971

Denmark

1967

France

1954

Ireland

1972

Italy

1973

Luxembourg

1970

Netherlands

1969

United Kingdom

1973

West Germany

1968

In an attem p t to foster trad e betw een its m em bers,
th e Council of th e E u rop ean E co n o m ic Com m unity
( E E C ) issued a d irective on April 1, 1 9 6 7 , w hich re ­
quired all m em ber nations to establish a V A T system
by Jan u ary 1, 1 9 7 0 .3 A lthough th e d irective con cerned
only th e cu rren t m em bers — Belgium , F ra n c e , G er­
m an y, Italy , L uxem bourg, and th e N etherlands — it
w as exten ded to the U nited K ingdom , D enm ark, and
Irelan d upon th eir joining the E E C in 1 9 7 3 (ta b le A l ) .
Since m ost p op ular discussions of th e V A T refer to the
“E u rop ean exp erien ce,” it m ay be useful to briefly ex­
am ine the basis of its adoption and its record to date.

iThe idea of the VAT originated with a German industrial­
ist by the name of F . Von Siemens. In 1918 he advocated
the substitution of the VAT for the newly implemented
turnover tax. T. S. Adams suggested using such a tax
in the United States in 1921. In the same year, the tax
was included in an amendment to the Revenue Act of
1921 proposed by Sen. Reed Smoot. For a general discus­
sion of the development of the VAT, see Due, “The ValueAdded Tax,” p. 165.
2France adopted the VAT as a replacement for existing turn­
over taxes. Michigan used a modified form of the VAT
from 1953 to 1967 when it formed the Single Business
Tax. This tax was introduced primarily to replace existing
business income taxes.
3This directive is reprinted in G.S.A. Wheatcroft, ValueA d ded Tax in the Enlarged Common Market (N ew York:
John Wiley & Sons, 1 9 7 3 ), Appendices I and II.


6


Table A2

Standard VAT Rates in EEC Countries
Country

Rate

Belgium

16.0%

Denmark

20.2

France

17.6

Ireland

20.0

Italy

14.0

Luxembourg

10.0

Netherlands

18.0

United Kingdom

8.0

Germany
SOURCE: Value-Added
1 9 7 9 ).

12.0
Tax

(Arthur

Anderson

& Co.,

T h e V A T w as in trod u ced in various E u ro p ean cou n ­
tries as an altern ative to tu rn over taxes. T h ese tu rn over
taxes, som etim es referred to as cascad in g tu rn over taxes,
are ch aracterized b y th e p aym en t o f taxes a t each level
of production w ithout reg ard to relief for th e tax paid
a t a prior stage. B ecau se of th e w ay in w hich these
taxes w ere im plem ented, th ere w as considerable in­
centive for a firm to control all the stages of production
(v ertical in teg ratio n ). Also, th ere w as a variety of
rules and regulations about the taxation of certain ex­
ports and im ports, w hich red u ced th e foreign trad e
potential betw een th e countries. T h e V A T provided for
th e general tax exem ption of exp ort goods and services.
Thus, the goal of the E E C in introducing V A T w as tw o­
fold: to “h arm onize” the m em bers’ tax systems and to
en courage in tra -E E C trade.
W h ile introduction of th e V A T system has increased
foreign trad e am ong th e E E C nations, harm onization of
th e various tax systems is still far from com plete. As
tab le A 2 indicates, th e “stand ard ” V A T ra te varies con ­
siderably am ong the various E E C countries. M oreover,
each cou ntry m ay choose to ap ply different rates to
various groups of goods and services. F o r exam ple,
Belgium has tw o basic rates: a stand ard ra te of 1 6 .0
p ercen t on necessities, and a 2 5 .0 p ercen t ra te on lux­
ury goods. Likew ise, G erm any and th e U nited K ing­
dom each h ave different rates fo r necessity and luxury
items.
In addition, each cou ntry exem pts certain items from
taxation. A lthough a com p lete listing w ould point out
the w ide diversity of possible exem ptions, th ere are
several areas com m only exem p ted from V A T. As m en ­
tioned earlier, exports are fully exem p t from V A T . Also,
sales of securities, stocks and bonds, d octors’ services,
financial services such as in su ran ce and banking, and
postal services are generally exem pt.

F E D E R A L R E S E R V E BAN K O F ST. L O U IS

A question remains regarding the neutrality of the
consumption or the income approaches. Economic
theory suggests that the consumption approach is less
distorting since it does not alter the individual’s
choice between consumption and saving (i.e., current
vs. future consumption) ,8 The income approach, how­
ever, discriminates against saving since capital ex­
penditures are not deductible from the tax base.

Capital Formation and Growth
A common complaint about our tax system is that
it discourages economic growth. VAT proponents
claim that a consumption-type VAT would not. Since
they believe that a reliance upon income taxation has
altered individual consumption-savings choices through
the income tax’s impact on relative prices, proponents
argue that the VAT would support both a higher
level of savings and investment.9 Behind this argu­
ment lies the “double-tax issue,” one of the oldest
controversies in economics. In other words, “Should
taxes be based on consumption or on income?” Con­
sumption-base tax advocates claim that a tax levied
on income is also a tax on savings and is therefore
inefficient because it raises the “price” of saving (fu­
ture consumption) relative to current consumption.
Since a saver eventually is taxed on the interest earned
from present savings, some analysts argue that the
income is taxed twice. Hence, the double-tax issue.
The exemption of capital expenditures with a VAT
would offset the non-neutral aspects of the current
tax system, which relies heavily on income taxation,
and may result in a higher level of overall savings
and investment.10
In addition, VAT would directly affect labor sup­
ply.11 Some analysts contend that existing income and
8Despite the positive economic attributes of the VAT in its
more favorable consumption-type form, the tax is likely to end
up being less neutral than its theoretical construction. Critics
claim that political necessity requires that the tax be applied
with numerous exemptions. It has often been asserted that
exemption of certain categories of food, clothing, housing,
medicine, and a long list of other special categories would be
a political necessity. In addition to narrowing the overall tax
base of the VAT, numerous exemptions would erode the neu­
trality of the tax, distorting the allocation of resources.
9See, for example, Musgrave and Musgrave, Public Finance,
pp. 468-69.
10Of particular concern is the responsiveness of savings to the
after-tax rate of return. Preliminary evidence indicates
that this response may be greater than was once thought.
See Michael J. Boskin, “Taxation, Saving, and the Rate
of Interest,” Journal of Political Economy (April 1978)
Part 2, pp. S3-S27. For an additional discussion of this
issue, see E . Phillip Howrey and Saul H. Hymans, “The
Measurement and Determination of Loanable Funds Saving,”
Brookings Papers on Econom ic Activity ( 3 : 1 9 7 8 ), pp.
655-705.
n Musgrave and Musgrave, Public Finance, pp. 483-88.




JA NUA RY

1980

social security taxes have altered the relative price of
work and leisure. These taxes, they claim, have in­
duced individuals to reduce their supply of labor at
prevailing market wages below which might otherwise
have been supplied without these taxes. Whether
VAT would correct these distortions depends upon
the relative responsiveness of labor to the removal or
reduction of the existing taxes and the changes in
relative commodity prices that may result from the
VAT

Distribution of the Tax Burden
Another question of interest is “Who bears the bur­
den of taxation?” An economic examination of this
issue can be made through what economists call “in­
cidence analysis.” Incidence is the change in an indi­
vidual’s real income (nominal income adjusted for
changes in the price level) that results from the im­
position of a tax.12
The key to understanding incidence lies in distin­
guishing between statutory incidence (the legal lia­
bility of the tax) and economic incidence (the final
burden of the tax). The sales tax provides a clear
example of this distinction. Although its legal liability
is imposed upon the retailer, the sales tax often is
assumed to be fully passed on to the individual who
purchases goods from the retailer. However, since the
prices of the retailer’s goods are now higher, the
quantity demanded will be reduced, and the owner’s
income therefore will be changed. In this case, the
economic incidence of the tax differs from the statu­
tory liability since the former takes into account both
the final resting point of the tax as well as the
economic ramifications on others (the retailer, for
example).
Tracing the final burden of a VAT is difficult be­
cause it ultimately depends upon the number of ex­
emptions. Assuming that the tax is applied equally
to all goods and services and is an “additional” tax,
its burden would be fully shifted to consumers.13 If
12Taxes can affect real income in two ways. They affect indi­
viduals in their role as producers by affecting incomes
received from their supply of labor or capital. Alternatively,
taxes can affect real income by altering the purchasing
power of a given amount of nominal income — that is, the
tax alters prices in the marketplace.
13In other words, we ignore here the question of tax substitu­
tion — that other taxes may be reduced by the amount of
the tax increase. An analysis of “balanced-budget incidence”
— that is, the effect of the tax given an equal decline in
other taxes and the assumption that real government ex­
penditures are the same — is a more realistic approach and
will be discussed below.

7

F E D E R A L R E S E R V E BAN K O F ST. L O U IS

exemptions are allowed or if other taxes are reduced,
however, there will be changes in relative prices and
in the production mix of the economy.

Equity Criteria
Equity criteria are often used to evaluate different
taxes (and thus incidence patterns). These criteria
attempt to answer the questions, “Who should pay
taxes?” and “How should the burden of taxation be
distributed among different individuals?” Although
objective economic analysis provides few answers to
these questions, it can provide a perspective on rela­
tive costs and benefits and on the trade-offs between
economic efficiency and equity considerations.
A principal equity concern is over the relative tax
burden imposed on individuals of different income
levels. The concept behind our income tax system, for
example, is that those individuals who earn a higher
income should pay a higher percentage of their income
in taxes. This is a “progressive” tax system. A “regres­
sive” tax, in contrast, takes a lesser proportion of in­
come from individuals who earn a larger income.
Critics of the VAT claim that it would be a regres­
sive tax since it taxes consumption and since lower
income individuals spend a larger proportion of their
incomes on consumption goods relative to higher in­
come individuals. Although this argument is straight­
forward, its validity depends on several unsettled
issues.
First, one might question the preceding analysis of
regressiveness on the grounds that it ignores the bene­
fit side of fiscal policy. In other words, the net bene­
fits derived from government spending or transfer
payments is not taken into account. Thus, a more
meaningful evaluation would include the benefits re­
ceived from government-related programs as well as
the costs.14
On the other hand, the regressiveness of the tax
could be alleviated either by adjustments to the in­
come tax rate and/or by special tax credits. For ex­
ample, the exemption of necessity items such as food,
shelter, clothing, and medical care has been proposed.
These exemptions, however, would reduce the tax
14Edgar Browning argues that sales and excise taxes are pro­
gressive when analyzed in a general equilibrium context
with government transfer payments. See Edgar K. Browning,
“The Burden of Taxation,” Journal of Political Economy
(August 1 9 7 8 ), pp. 649-71. Another description of this
view, along with an empirical investigation of the burden of
the U.S. tax system, is provided in Edgar K. Browning and
William B. Johnson, The Distribution of The Tax Burden
( American Enterprise Institute, Studies in Tax Policy, 1 9 7 9 ).

8




JA N U A RY

1980

base and thus necessitate higher tax rates to insure
the same yield as under a comprehensive base ap­
proach. In general, therefore, attempts to alleviate the
regressiveness of the tax are likely to complicate the
administrative problems and interfere with the neu­
trality criteria discussed earlier.

Administrative Costs
The cost of administration is another criterion for
evaluating the merits of a tax system. The initial ad­
ministrative costs of introducing a VAT ( or any other
tax for that matter) would be relatively high.
Perhaps the most important factor affecting the
cost of using a VAT is the degree of complexity of
the tax. The use of multiple rates and numerous
exemptions, in contrast to a single uniform rate,
would raise the administrative costs. Indeed, it has
been estimated that if the VAT system used involves
more than a single rate, administrative costs may
rise by 50 to 80 percent.15 Most of this increase
would be due to increased personnel costs caused
by a rise in the amount of paperwork required of
both business and government.16
Another consideration is whether the tax is used
to replace part of an existing tax or as a supple­
mentary source of government revenue. If the VAT
replaces only part of an existing tax or is merely
added to the present system, administrative cost
savings may be negligible. In fact, adding a VAT
may increase the current cost of the government’s
tax collecting apparatus. This can be seen by consider­
ing the influence on costs from a reduction in existing
tax rates. Since the previous tax system still exists
and since there is little or no change in collecting or
reporting procedures, the time and manpower in­
volved in collecting a 5 percent tax is essentially
the same as that for a 10 percent tax.
Furthermore, adding a VAT to the existing array
of taxes would have differential cost effects on busi­
ness. For instance, soon after a VAT was introduced
in Germany in 1968, small noncomputerized firms
estimated that tax-related administrative costs in­
creased up to 20 percent. Relatively large businesses,
15National Economic Development Office, Value-Added Tax
(London: Her Majesty’s Stationery Office, 19 7 1 ) p. 41, cited
in Dan Throop Smith, et. al., What You Should Know About
the Value-Added Tax (Homewood: Dow Jones-Irwin, Inc.,
1 9 7 3 ), p. 53.
i6“It’s obviously possible to administer a VAT. But every exemp­
tion you put in makes it substantially more difficult to ad­
minister.” Deputy Assistant Treasury Secretary Emil Sunley,
quoted in Bureau of National Affairs, Daily Report for E xecu ­
tives (October 5, 1 9 7 9 ), p. K-6.

F E D E R A L R E S E R V E BAN K O F ST. L O U IS

however, reported negligible
increases.17

administrative cost

The frequency of collection is also a significant
cost factor. Most of the European countries currently
using a VAT require monthly payments.18 In the
United Kingdom, however, the collection period is
quarterly. Although a monthly payment schedule may
create cash-flow problems for some businesses, the
European experience suggests that this frequency is
feasible since most firms already record the data
needed to calculate their VAT liability on a monthly
basis.

ECONOMIC EFFECTS OF VAT
Price Effects
A major concern about the implementation of a
VAT is the resulting price effects. The introduction
of the VAT, it has been widely asserted, will lead
to a one-time increase in the general level of prices.
In fact, it can be easily demonstrated that, although
a uniform VAT rate (an identical rate for all goods
and services) applied without exemption will not
alter relative prices within the economy, it will re­
sult in a one-time increase in the overall level of
prices.19 It should be noted, however, that this con­
clusion is based on several assumptions.
Perhaps the most important assumption is that
all sellers raise their prices by the exact amount of
the VAT, thus passing the additional cost on to the
consumer. There are several reasons why this actually
may not occur.
A VAT is usually applied with a varying rate
structure — for instance, a 5 percent rate on food
and a 10 percent rate on nonfood items — and/or
with some items exempt from taxation.20 If sellers
face different cost increases due to the VAT, then
it is uncertain that all prices will be raised by an
equal amount. Since none of these conditions will
lead to identical increases in all prices, the con­
sequent relative price changes would induce a shift­
17See Alan A. Tait, Value-Added Tax (London: McGraw-Hill,
1 9 7 2 ), p. 126.
18Fo r a discussion of the various collection periods, see the
articles contained in G.S.A. Wheatcroft, Value-Added Tax
in the Enlarged Common Market (N ew York: John Wiley &
Sons, 1 9 7 3 ). This monthly scheme also appears to be favored
by some U.S. officials. A comparison of views may be found
in Daily Report for Executives (October 5, 19 7 9 ), pp. K7K8.
19Ann F . Friedlaender, “Indirect Taxes and Relative Prices,”
Quarterly Journal of Economics (February 1 9 6 7 ), pp. 12539.
20See supplement, page 6.




JA N U A RY

1980

ing in the pre-VAT pattern of demand for goods
and services. In this way, the production (and em­
ployment) decisions of producers are affected by the
introduction of a VAT.
Introducing a VAT may cause a one-time change
in the level of prices if the tax is merely a supple­
ment to existing taxes. It is generally assumed that
firms face higher costs because of the VAT and,
therefore, attempt to pass this on to consumers. How­
ever, if the tax burden is reduced in some other area—
for instance, the reduction in firms’ contributions
to social security — then this assumption is un­
warranted. To determine the final price effect of a
VAT in this case requires knowledge of the tradeoff
between reductions in existing taxes (if any) and
the cost to the firm of administering the VAT.
Finally, it is unlikely that a rise in the general
level of prices can be maintained without an in­
crease in the money stock held by the public or
an increase in the velocity of the existing money stock
(decline in real money balances). If neither of these
situations occurs, then changes in production and
employment will occur since consumers would not be
able to maintain previous consumption levels at the
higher level of prices. Thus, the reaction of the gov­
ernment to observed changes brought on by the
initial price effects of a VAT may further compli­
cate the foregoing analysis.

Interjurisdictional Intrusion
How VAT might interfere with existing state and
local taxes is also an important issue. State and
local governments likely will be apprehensive about
a VAT, given that the tax appears to be an intru­
sion into an area upon which these governments
traditionally have relied — namely, the sales tax. Con­
tinued use of the retail sales tax should not conflict
with federal use of the VAT from an economic effi­
ciency view, however. If the VAT is applied at a
uniform rate, relative prices will not be affected.
Some have argued, however, that a VAT may
change taxpayers’ perceptions of the cost of govern­
ment expenditures and may alter the level of public
services demanded. If the tax is “invisible” or hidden
in the price of goods, taxpayers may demand a
higher level of public expenditures since the costs
are not fully perceived. On the other hand, if the
tax is fully perceived at the retail level (for example,
if the VAT portion is stated separately at purchase),
consumers may object to additional retail sales taxes,
and state and local governments will find it more
difficult to tax by this method.
9

F E D E R A L R E S E R V E BANK O F ST. LO U IS

JA N U A RY

1980

Stabilization

Summary

An important feature of a tax system is its over­
all response to cyclical fluctuations in the economy.
For a given set of tax rates, the growth of the tax
base normally will change with the level of aggregate
economic activity. Provided that the change in tax
revenues does not affect the level of government
spending, these changes will cause variations in the
overall budget surplus or deficit. “Built-in flexibility”
is said to exist if tax liabilities rise (and fall) at a
faster rate than income. Our tax system, based pri­
marily on a progressive income tax, posits that the
overall federal budget has a stabilizing effect on
the economy since the tax acts to dampen fluctua­
tions in economic activity. A substitution from in­
come and social security taxes to a VAT likely would
reduce such built-in flexibility.21

The purpose of this article has been to provide
a general background for understanding the valueadded tax. Recent examinations have narrowly ad­
dressed the tax’s price effects and its potential to
reduce the existing tax burden imposed by federal
income and social security taxes. While these are
indeed important issues, a complete analysis of VAT
must go deeper, examining the tax base, its neutral­
ity, regressiveness, incidence, and effects on capital
formation. In addition, the administrative costs of the
tax itself must be considered as well.

Digitized for10
FRASER


21Stated alternatively, a tax whose income elasticity of yield is
greater than one is considered an automatic stabilizer. Elas­
ticity is defined as the percentage change in tax yield divided
by tne percentage change in income (consumption) base.

The Demand for Currency: Is the
Underground Economy Undermining
Monetary Policy?
NORMAN N. BOWSHER

^^lURRENCY holdings of the public have risen at
a brisk pace for many years, reflecting a progressive
expansion in the demand for currency.
Since 1960 the currency component of the money
stock has risen at an average 7 percent annual rate. It
is reasonable that the public, faced with higher prices,
would seek to hold more currency. However, currency
holdings adjusted for changes in consumer prices have
risen at a 2 percent rate since 1960 while real de­
mand deposits have been unchanged on balance. As
a result, currency rose from 25 percent of demand
deposits in 1960 to 29 percent in 1970, to 38 percent
in 1979 (table 1). In late 1979 the average holdings
of currency for a family of five amounted to an
astonishing $2,500.
This greater demand has been readily accommo­
dated since the supply of currency, as provided for in
the Federal Reserve Act of 1913, is elastic. In the
U.S. monetary system, any depositor can obtain more
currency by withdrawing it from deposit balances at
commercial banks. Commercial banks that are mem­
bers of the Federal Reserve System, in turn, obtain
currency to restock inventories from their Federal
Reserve District Banks by having their reserve ac­
counts (deposits) charged. Other banks obtain cur­
rency in exchange for deposits at their member
correspondent banks.
The persistent increase of currency into circulation
has been surprising in view of several developments
which discourage its use. Credit cards, for example,
have been substituted for cash in an expanding num­
ber of transactions, and the number of checking ac­
counts and traveler’s checks issued have continued
to grow. Moreover, interest rates have moved to
higher levels, raising the cost of holding wealth in
the form of currency and coin, which does not earn



Table 1

Currency and Demand Deposit
Components of Money (billions of
dollars)1
Year
1939

Currency
$

Demand
deposits

Ratio of
currency to
demand deposits

6.4

$ 29.8

.21
.25

1959

28.9

114.8

1960

29.0

114.5

.25

1961

29.1

117.4

.25

1962

30.1

119.6

.25

1963

31.5

122.6

.26
.26

1964

33.5

126.8

1965

35.3

131.8

.27

1966

37.5

137.4

.27

1967

39.4

142.5

.28

1968

42.0

152.8

.27

1969

44.8

161.7

.28

1970

47.7

166.8

.29

1971

51.1

177.7

.29

1972

54.7

190.4

.29

1973

59.3

203.9

.29

1974

64.9

212.8

.30

1975

71.0

218.9

.32

1976

77.8

227.3

.34

1977

84.8

242.6

.35

1978

93.2

259.6

.36

1979

102.3

268.7

.38

‘ All data are annual averages of daily figures except 1939
which are end-of-year data.

11

F E D E R A L R E S E R V E BANK O F ST. L O U IS

interest. In addition, the possession of currency in
large amounts is risky; it can be lost, destroyed, or
stolen. Holding liquid funds in the form of deposits,
on the other hand, has been virtually free of such
risks since the advent of deposit insurance.
Because of the sustained growth of currency out­
standing relative to demand deposits ( chart 1), ques­
tions have arisen about what purposes the currency
serves. More specifically, does this currency support
a huge and growing amount of unreported economic
activity? Moreover, since currency is one major use for
the monetary base, are there other monetary policy
implications of the alleged change in demand for
currency? This article explores the answers to these
questions.

JA N U A R Y

1980

C h a rt 1

Ratio of Currency to D e m a n d Dep o sits
Ratio

A n n u a l A v e r a g e s of D aily Figures
Ratio

The Use of Currency in Unreported
Activities
Suspicion has arisen that the pronounced growth
in currency relative to demand deposits reflects an
expansion of illegal transactions or other activities
that the participants wish to hide.1 According to Gutmann, currency lubricates a vast amount of unre­
ported income and employment, an entire subter­
ranean economy whose total product, estimated to be
almost $200 billion in 1976, exceeded the nation’s en­
tire gross national product in the middle of World
War II.2
A number of recent developments certainly have
provided increased incentives to conduct unreported
activities. Inflation has increased markedly the effec­
tive income and estate tax rates for most individuals.
For example, a family of four who earned $12,000
in 1977 and who received an 8 percent increase in
income in 1978 owed 21.8 percent more in federal in­
come taxes (using standard deductions) .3 As inflation
pushes individuals into higher effective tax brackets,
it provides an incentive to receive income in forms
that cannot be traced readily. Currency payments, of
course, are less traceable than deposit transfers. In­
creasing regulations, prohibitions, and reporting bur­
dens provide other incentives for unreported transac­
tions. Similarly, certain illegal activities such as drug
1See “The Underground Economy,” U.S. News & W orld Re­
port, October 22, 1979, pp. 49-52; also, Susan Harrigan,
“Blemished Boom -—■Miami is Prospering, Aided by Latin
Money, Illegal Drug Business,” Wall Street Journal, Novem­
ber 28, 1979.
-Peter M. Gutmann, “The Subterranean Economy,” Financial
Analysts Journal ( November/December 1 9 7 7 ), pp. 26, 27,
34. Also see Peter M. Gutmann, “Statistical Illusions, Mis­
taken Policies,” Challenge ( November/December 19 7 9 ), pp.
14-17.
3See Alfred L. Malabre, Jr., “As Salaries Climb W ith Prices,
People Pay More of Income in Taxes Despite Rate Cuts,”
Wall Street Journal, November 28, 1979.

Digitized for12
FRASER


dealing, which are almost always transacted in cur­
rency, apparently have expanded.
Nearly everyone receives a minor amount of un­
reported income in currency from “friendly” bets, per­
sonal services, garage sales, etc. Any firm or person
who receives currency in a business transaction has
the option of joining, at least in part, the subterranean
economy of unreported income. Such opportunities
exist for many engaged in the professions and those
performing repair work and personal services, where
it is traditional to compensate individuals with cash
payments. Wages for part-time work, especially for
child care and domestic workers, also are frequently
paid in cash.
Since reliable information on currency usage is
unavailable, indirect evidence has been relied on to
support conclusions about the dollar volume of un­
reported transactions. Gutmann’s estimates of “under­
ground” activity financed by currency were based on
several simple relationships and assumptions. He ob­
served that the ratio of currency to demand deposits
had risen markedly. In the five years before World
War II, which he selected as a base for comparison,
the ratio of currency to demand deposits averaged

JA N U A R Y

F E D E R A L R E S E R V E BANK O F ST. L O U IS

Treasury is seeking to obtain fuller disclosure of
sizable currency transactions.5

Table 2

Ratio of Currency to Deposits1
Year

Demand
deposits

Total
deposits

1939

.21

.11

1959

.25

.16

1969

.28

.13

1976

.34

.12

1979

.38

.13

1A11 data are annual averages of daily figures except 1939
which are end-of-year data.

21.7 percent. By 1976 this ratio had risen to 34.4
percent. If the ratio of currency to demand deposits
required to support legal activities had not changed,
the 12.7 percentage-point increase in currency (or
$28.7 billion) over and above the rise in demand
deposits represents funds devoted to increased illegal
activity. Furthermore, if each dollar of currency
used for illegal transactions supports the same amount
of activity as each dollar of money held for legal
purposes ($275.3 billion of demand deposits plus legal
currency produced $1,693 billion of reported GNP in
1976), then the $28.7 billion held for illegal activities
produced an estimated $176 billion of unreported
income.
This estimate is conservative according to Gutmann. It makes no allowance for unreported income
in the pre-World War II base period, it takes for
granted that the earlier downward trend in the cur­
rency to demand deposit ratio would not have contin­
ued into the postwar period, it assumes all demand
deposits are held for transactions and none for com­
pensating banks for services rendered, and it ignores
the growth in credit card usage and other currency
saving devices.
Several more recent studies provide support for
the broad Gutmann conclusion that the volume of un­
recorded transactions is indeed substantial. The Inter­
nal Revenue Service, from comprehensive audits of a
sample of tax returns, estimated that unreported in­
come on 1976 individual income tax returns was as
much as $100 billion from legal sources. It also re­
vealed no taxes were paid on an additional $25 billion
to $35 billion of individual income from three types of
criminal activity — narcotics, illegal gambling, and
prostitution.4 One result of this study is that the
4Estimates of Income Unreported on Individual Income Tax
Returns, Publication Number 1104, Internal Revenue Serv­
ice (September, 1 9 7 9 ).




1980

In another study, Feige estimated that unreported
transactions in 1976 totaled $369 billion (and a fan­
tastic $704 billion in 1978).6 He calculated the total
value of all transactions in the economy by using out­
standing stocks of demand deposits and currency and
estimating turnover rates for each. The derived total
transactions series has been expanding faster than
reported income (GNP). Feige used this excess as a
measure of the irregular economy of unreported ac­
tivity. His research into other possible explanations
of the faster growth in transactions than reported
income — changes in relative prices, relative volume
of financial transactions, and the degree of vertical
economic integration — concluded that, if anything,
his estimates were low.
According to Gutmann, the underground economy
for 1976 was about 10 percent of reported gross na­
tional product; the Feige estimate was nearly 22 per­
cent. Using Feige’s calculations, unreported transac­
tions had jumped to roughly 33 percent of GNP by
1978. From just before World War II until 1976,
reported GNP increased at an average 7.8 percent
annual rate; were the subterranean economy in­
cluded, the trend growth rate would have been 8.1
percent according to Gutmann and 8.3 percent ac­
cording to Feige. Feige’s research indicates that from
1976 to 1978 total nominal output, reported and un­
reported, expanded at a 16.9 percent annual rate,
compared with the 11.8 percent rate of reported GNP.
On the other hand, questions have been raised by
other analysts concerning the logic of the assumptions
used in arriving at such large estimates of the sub­
terranean economy. For example, Gutmann’s estimate
ultimately depends on the use of demand deposits
as the yardstick for judging currency growth. Were
the comparison made with total deposits, the results
would have been considerably different.7 Indeed, cur­
rency has changed little relative to total bank deposits
since 1939, which suggests that the growth of demand
deposits has slowed rather than that the growth of
currency has accelerated (table 2). Hence, the subter­
5“Treasury Study Seeks Stricter Reporting Rules F o r Large
Currency Transactions,” American Banker, November 30,
1979.
8Edgar L. Feige, “How Big is the Irregular Economy?” Chal­
lenge (Novem ber/Decem ber 1 9 7 9 ), pp. 5-13. In the report
Feige observed, “I began this investigation suspecting that
the irregular economy was smaller than previous estimates
had suggested. I am now convinced that the irregular econ­
omy is indeed of staggering proportions and growing rapidly.”
7Robert D. Laurent, “Currency and the Subterranean Econ­
omy,” Economic Perspectives, Federal Reserve Bank of Chi­
cago (M arch/A pril 1 9 7 9 ), pp. 3-6.

13

FED ERA L. R E S E R V E BAN K O F ST. L O U IS

ranean economy may not have grown relative to the
rest of the economy.
Although time deposits are close substitutes for de­
mand deposits and currency for many purposes, the
value of such criticism is uncertain. It seems reason­
able to relate currency to demand deposits for trans­
actions analysis since these two are the only mone­
tary assets that can be directly exchanged for goods
and services; other deposits must first be converted
into currency or demand deposits before they are
spent.
In another criticism Laurent suggested that in­
creases in total transfers of demand deposits and
currency are more indicative of the trend of economic
activity than increases in the outstanding stocks.8 Cer­
tain available data indicate a much larger increase
in the turnover of demand deposits than of currency.
Debits to demand deposit accounts at reporting banks
increased by more than 30 times over the period 1939
to 1976. In marked contrast, according to Laurent,
currency turnover slowed over the same period.
Although there is no firm information on currency
transactions, a rough measure of velocity was in­
ferred by observing currency redeemed and destroyed.
Currency is redeemed and destroyed when the paper
notes show signs of wear, presumably a function of
its use in transactions. From 1937 to 1941, the aver­
age life of the currency destroyed was 3.1 years; in
1976 it was 5.3 years. To some extent, this lengthening
was due to the Federal Reserve’s policy of extending
currency life by screening it less rigorously. But an
examination of data for the early 1970s, before such
policy changes occurred, also shows an increase in
currency life. The combination of the increased de­
posit turnover and a slowing of currency turnover
markedly changes the significance of the relative
growths of the currency and demand deposit stocks.
The growth in the currency stock, viewed in this light,
appears too small to accommodate both the growth in
reported transactions and large expansion in unre­
ported activities.
The Feige study, however, concluded that currency
turnover had accelerated since the forties and early
fifties, not decelerated.9 He noted that the quality of
paper used in currency was improved dramatically
by a melamine-formaldehyde resin additive in 1957.
Tests demonstrated that, after this development, each
bill could be used in nearly twice as many transac­
tions before wearing out.
In addition, Laurent suggested that the rapid
8Ibid.
9Feige, “How Big is the Irregular Economy?”

Digitized for14
FRASER


JANUARY

1980

Table 3

Ratio of Currency and Demand
Deposits to Gross National Product1
Year

Currency

Demand
deposits

1939

.07

.33

1959

.06

.24

1969

.05

.17

1976

.05

.13

1979

.04

.11

1AII data are annual averages of daily figures except 1939
which are end-of-year data.

growth in currency outstanding is due, in part, to an
increased use of currency as a store of value.10 The
increase in large denomination bills suggests hoard­
ing. In fact, there is now more money outstanding in
$100 bills than in any other currency denomination
(larger denominations are no longer issued).
On the other hand, some argue that the abundance
of large denomination notes may just as well indicate
an increased use of currency for illegal activity as
for a store of value. Furthermore, the growth of ille­
gal activity itself increases the stock of currency
hoarded. Reportedly, the large tax evader or drug
dealer frequently accumulates sizable amounts of
currency in the larger denominations. These cash
balances are maintained, despite high interest rates,
to avoid arousing suspicion in converting this cur­
rency into other assets. When it becomes profitable
to do so, the holder may purchase a legitimate busi­
ness to “launder” the illegal gains.
From a casual comparison of the growth of cur­
rency and the expansion of spending, there does not
appear to be a great unexplained rise in currency. In
fact, currency outstanding has expanded at a pace
slightly less than the increase in total recorded spend­
ing on goods and services (see table 3). The marked
rise in the ratio of currency to demand deposits, giv­
ing rise to suspicions of a rapidly expanding under­
ground economy, has reflected sluggish growth in
demand deposits. The lack of demand deposit growth
can be explained by a proliferation of alternatives for
such deposits, including repurchase agreements, NOW
accounts, telephone and automatic transfers between
savings and demand deposits, and money market mu­
tual funds.
10Laurent, “Currency and the Subterranean Economy."

F E D E R A L R E S E R V E BANK O F ST. L O U IS

In another examination of currency developments,
Garcia argues that the rapid increase in currency out­
standing should have come as no surprise to fore­
casters of currency demand.11 She notes that, since the
1950s, the amount of currency outstanding has ap­
proximated the predictions based on economic models.
Using the econometric equations developed by Goldfeld, Garcia found that currency as a function of
reported income, interest rates, and lagged currency
holdings behaved as anticipated, while demand de­
posits rose at a surprisingly slower pace.12 Hence,
she concludes that the subterranean economy operat­
ing on currency has not grown spectacularly since
the early fifties; more likely it is roughly the same
portion of GNP.13
Gutmann responded to the Garcia study by point­
ing out that an equation may do a satisfactory job of
forecasting without affording much insight into under­
lying causes.14 He faulted the currency equations for
their omission of both a variable proxying the sub­
terranean economy and the implicit yield on demand
deposits. In the latter case, he argued that including
the yield on the closest currency substitute — de­
mand deposits — would, ceteris paribus, have led to a
forecast of smaller currency growth over the period.
This is because the implicit yield on demand deposits
has risen, or, alternatively, the price of the substitute
has fallen. He claimed that the success of the Goldfeld
formulation in predicting currency demand, despite
the omitted variables, is explained by the fact that
(1) both currency and two of the model’s explanatory
variables (GNP and interest rates) rose rapidly over
the period examined and that (2) the previous
period’s currency was included as an explanatory
variable.15
^Gillian Garcia, “The Currency Ratio and the Subterranean
Economy,” Financial Analysts Journal (Novem ber/Decem ­
ber 1 9 7 8 ), pp. 64-66, 69. Johannes and Rasche also found
the currency ratio relatively stable. James M. Johannes and
Robert H. Rasche, “Predicting the Money Multiplier,” Jour­
nal of Monetary Economics (July 1 9 7 9 ), pp. 301-25.
12Stephen Goldfeld, “The Demand for Money Revisited,”
Brookings Papers on Economic Activity ( 3 : 1 9 7 3 ), pp.
577-638.
13Other studies have also concluded that the Gutmann and
Feige estimates of the subterranean economy are excessive.
See Richard D. Porter and Stephan S. Thurman, “The Cur­
rency Ratio and The Subterranean Economy: Additional
Comments” (Board of Governors of the Federal Reserve
System, January 26, 1979, processed); and Richard D. Por­
ter, “Some Notes on Estimating The Underground Econ­
omy,” (Board of Governors of the Federal Reserve System,
August 10, 1979, processed).
14Peter M. Gutmann, “Professor Gutmann Replies,” Financial
Analysts Journal (November/December 1 9 7 8 ), pp. 67-69.
15Including a lagged dependent variable in an equation is a
valid test of the adjustment of actual to desired quantities in
a partial adjustment model.




JA N U A RY

1980

In brief, it is well known that much income goes
unreported — this has been true for decades. With­
out more accurate information on currency exchanges,
however, reliable estimates of unreported activities
will be elusive. The great stock of currency outstand­
ing and its persistent expansion in the face of higher
costs of holding currency and the widespread use of
currency substitutes ( e.g., credit cards) certainly sup­
port suspicions of increasing irregular transactions.
Moreover, heavier tax burdens and government regu­
lation seem to have provided powerful incentives for
attempting to hide more activities.
On the other hand, currency has not risen as rapidly
as total spending. In addition, the velocity of demand
deposits has increased markedly in the past several
decades, while evidence on currency turnover indi­
cates only a moderate acceleration (or even a decel­
eration). It is probable, therefore, that unreported
transactions associated with currency have not risen
as fast as the relative importance of currency out­
standing in the money stock might indicate. However,
the persistent expansion in currency outstanding at a
time of strong incentives to avoid reporting could in­
dicate that unreported activities have been expanding
faster than economic activity generally.

Monetary Control
The rapid increase of currency outstanding in re­
cent years also has implications for economic stability
in addition to the unrecorded direct effects on spend­
ing, production, and employment. It is possible that
flows of currency out of commercial banks could
adversely affect the ability of the Federal Reserve Sys­
tem to control the total money stock. A useful frame­
work for analyzing the impact of currency move­
ments on monetary control is provided by the sources
and uses of the monetary base or, more specifically,
the net monetary liabilities of the Federal Reserve
System and the U.S. Treasury.
Sources of the base include the monetary gold
stock, Treasury currency outstanding, member bank
borrowings from Reserve Banks, and Federal Reserve
float (credit extended on checks received but not yet
collected). The largest and dominant source of the
base, however, is Federal Reserve holdings of govern­
ment securities.16 Fluctuations in the base caused by
irregular movements in the other sources can be read16Other more minor factors that must be considered in analyz­
ing the net sources of the base include Treasury cash hold­
ings (having a negative im pact), foreign deposits at Re­
serve Banks (having a negative im pact), other Federal
Reserve assets, other Federal Reserve liabilities, and capital
(having a net negative im pact).

15

JA N U A R Y

F E D E R A L R E S E R V E BAN K O F ST. L O U IS

ily offset as soon as detected (usually within one
business day) by System purchases or sales of
government securities. Several factors, such as data
revisions, changes in member bank borrowing (in
response to credit demands and System purchases and
sales of securities), and the two-week lagged imposi­
tion of reserve requirements (forcing member banks
to obtain reserves by borrowing from Reserve Banks
if not otherwise made available) can prevent the
System from hitting precise base targets within a
month or two. Over a period of several months, how­
ever, the growth rate of the base can be determined
within a narrow range by System purchases and
sales.17 As studies generally have shown, fluctuations
in monetary aggregates over periods of less than three
months have seldom had any measurable effect on
prices, employment, or incomes. Thus, the Federal
Reserve has the ability to adequately control the mag­
nitude and trend of the monetary base by buying
or selling securities in the open market.18
Analysis of the uses of the base is more complicated
than an examination of its sources. Although there
are just two uses of the base, currency in circulation
and reserves to support member bank deposits, the
analysis of money determination must go behind
these aggregates. The U.S. monetary system is com­
plicated by the existence of both member and non­
member banks, different classes of member banks,
different reserve requirements on different types of
deposits, and graduated reserve requirements for de­
posits of different amounts. It is thus necessary to
allocate reserves among numerous types of deposits
for both those included in the money stock and those
not included (such as government and time deposits),
as well as to estimate the amount of excess reserves
that the banking system will find expedient to hold.
In addition, a chief factor influencing the amount
of money the banking system can create, given an
increase in the monetary base, is the proportion of
currency to demand deposits that the public desires
to hold. If the public held a fixed amount of currency,
all increases in the supply of monetary base by the
Federal Reserve would remain in the banking system
as reserves and would be reflected entirely in a mul­
tiple change in deposits, the amount depending on
reserve requirement ratios for different types of
deposits.
17Albert E . Burger, “Explanation of the Growth of the Money
Stock: 1974-Early 1975,” this Review (September 1 9 7 5 ),
pp. 5-10.
18See J. Ernest Tanner, “Lag in Effects of Monetary Policy:
A Statistical Investigation,” T he American Economic Review
(Decem ber 1 9 6 9 ), pp. 794-805.

Digitized for16
FRASER


1980

When currency holdings of the public rise rapidly
as they have in recent years, however, the total mone­
tary expansion from a given monetary base injection
is greatly reduced. This results because an outflow of
currency from banks absorbs an equal amount of re­
serves. With fewer reserves, the deposit-creating po­
tential of the banking system is reduced by some
multiple.
Clearly this increased currency demand must be
considered in determining how much monetary base
must be supplied to achieve a desired increase in the
money stock. Fortunately, the System receives prompt
data on currency outstanding through frequent re­
ports, and trend movements of currency have been
projected fairly accurately by forecasting models.
Hence, prompt action to offset any undesired effects
of currency movements has been possible.
During the seventies, currency holdings of the pub­
lic rose rapidly. Unless offset, this acceleration in the
public’s demand for currency would have caused the
growth rate of the money stock to slow or even to
decline at times. The trend growth rate of the money
stock, however, was more rapid in the seventies than
in any other decade since World War II. During the
1970s there were periods in which the growth rate of
the money supply slowed long enough to depress the
growth of spending. There is no evidence, however,
that the slower money growth reflected an increased
demand for currency that could not have been off­
set by net System purchases of securities promptly
enough to avoid any material contractive effects on
spending.

Conclusions
The almost steady rise in currency in the past
decade has fostered a controversy over whether it
reflected a tremendous growth in illegal or otherwise
unrecorded activities. Although reliable evidence on
the extent of the subterranean economy is lacking, it
is no secret that numerous unrecorded transactions
take place daily. It also is probable that the volume
of such transactions has been expanding faster than
the volume of recorded transactions. Certainly, greater
tax burdens and additipnal restrictions on legal ac­
tivities provide incentives for an increase in subrosa
pursuits. A number of studies, however, indicate no
pronounced acceleration in unreported activities.
Lack of firm knowledge of the volume of un­
reported transactions does not reduce its economic
importance. The implications of a large, growing, but
unrecorded sector in our society are numerous and

F E D E R A L R E S E R V E BANK O F ST. L O U IS

affect measurements of economic activity and the eco­
nomic policies based on these measurements. For ex­
ample, because much income goes undetected, tax
revenues are smaller than they would be otherwise.19
Official statistics, which are based on recorded trans­
actions, understate both the true magnitude and
growth rates of sales, income, employment, and pro­
duction. As a result, economic policies are more apt
19See Peter M. Gutmann, “Taxes and the Supply of National
Output, Financial Analysts Journal (November/December
1 9 7 9 ), pp. 64-66.




JA N U A RY

1980

to exacerbate inflation than if all activities were
reported.
Unusually rapid growth in currency outstanding,
given our monetary system, can potentially contract
the total money stock (demand deposits plus cur­
rency). Nevertheless, since information on currency
movements into and out of banks is current, and since
offsetting transactions can be made by the Federal Re­
serve within a brief period, potentially depressing
effects on the economy from this source have been
avoided.

17

F E D E R A L R E S E R V E BAN K O F ST. LO U IS

JA N U A R Y

REPRINT SERIES
OVER THE YEARS certain articles appearing in Review have proven helpful to
banks, educational institutions, business organizations, and others. To satisfy the de­
mand for these articles, our reprint series has been made available on request. The
following articles have been added to the series in the past few years. Please indicate
the title and number of the articles you need and mail your request to: Research De­
partment, Federal Reserve Bank of St. Louis, P.O. Box 442, St. Louis, MO. 63166.
Number

Title of Article

Issue

86

A Primer on the Consumer Price Index

July 1974

87

Channels of Monetary Influence: A Survey

November 1974

88

A Primer on Inflation: Its Conception, Its Costs,
Its Consequences

January 1975

The FOMC in 1974: Monetary Policy During
Economic Uncertainty

April 1975

90

A Monetary View of the Balance of Payments

April 1975

91

A Monetary Model of Nominal Income
Determination

June 1975

Observed Income Velocity of Money: A Misunder­
stood Issue in Monetary Policy

August 1975

93

Crowding Out and Its Critics

December 1975

94

The FOMC in 1975: Announcing Monetary Targets

March 1976

95

The Link Between Money and Prices — 1971-76

June 1976

96

The Unemployment Rate as an Economic Indicator

September 1976

97

Development of Electronic Funds Transfer Systems

September 1976

98

Derivation of the Monetary Base

November 1976

99

Are You Protected From Inflation?

January 1977

100

The FOMC in 1976: Progress Against Inflation

March 1977

101

The Effects of Changes in Inflationary Expectations

April 1977

102

The Effects of the New Energy Regime on Eco­
nomic Capacity, Production, and Prices

May 1977

103

Energy Resources and Potential GNP

June 1977

104

Revision of the Monetary Base

July 1977

89

92

Digitized for 18
FRASER


1980