The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
86l s t Session 8 }
J O IN T C O M M IT T E E P R IN T
STUDY PAPERS NOS. 7, 8, AND 9
THE INCIDENCE OF INFLATION:
OR WHO GETS HURT?
BY
Seymour E. Harris
PROTECTION AGAINST INFLATION
BY
H. S. Houthakker
THE SHARE OF WAGES AND SALARIES IN
MANUFACTURING INCOMES, 1947-56
BY
Alfred H. Conrad
MATERIALS PREPARED IN CONNECTION WITH THE
STUDY OF EMPLOYMENT, GROWTH, AND
PRICE LEVELS
FOR CONSIDERATION BY THE
JOINT ECONOMIC COMMITTEE
CONGRESS OF THE UNITED STATES
NOVEMBER 26, 1959
Printed for the use of the Joint Economic Committee
UNITED STATES
GOVERNMENT PRINTING OFFICE
48084
WASHINGTON : 1959
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington 25, D.C. - Price 45 cents
JOINT ECONOMIC COMMITTEE
(Created pursuant to sec. 5(a) of Public Law 304, 79th Cong.)
P AUL H . DOUGLAS, Illinois, Chairman
W R IG H T P A T M A N , Texas, Vice Chairman
SENATE
HOUSE OF R EPR E SEN TA TIV E S
JOHN SP A R K M A N , Alabama
J. W IL L IA M F U LB RIG H T, Arkansas
JOSEPH C. O’M A H O N E Y , Wyoming
JOHN F. K E N N E D Y , Massachusetts
PR ESCO TT BUSH, Connecticut
JOHN M AR SH AL L B U T L E R , Maryland
JACOB K . JAVITS, New York
Study
of
E
m ploym ent,
R ICH AR D BO LLIN G , Missouri
H A L E BOGGS, Louisiana
H E N R Y S. REUSS, Wisconsin
F R A N K M . COFFIN, Maine
TH O M AS B. CURTIS, Missouri
C L AR EN C E E. K IL B U R N , New York
W IL L IA M B. W ID N A L L , New Jersey
Growth,
and
P r ic e L e v e l s
(Pursuant to S. Con. Res. 13,86th Cong., 1st sess.)
O t t o E c k s t e in ,
Ja m e s
n
Technical Director
W . L e h m a n , Administrative Officer
W . K n o w l e s , Special Economic Counsel
Joh n
These are part of a series of papers being prepared for
consideration by the Joint Economic Committee in connec
tion with their "Study of Employment, Growth, and Price
Levels.” The committee and the committee staff neither
approve nor disapprove of the findings of the individual
authors. The findings are being presented in this form
to obtain the widest possible comment before the com
mittee prepares its report.
in
LETTERS
O F T R A N S M IT T A L
N o v e m b e r 20,1959.
To Members o f the Joint Economic Committee:
Submitted herewith for the consideration of the members of the
Joint Economic Committee and others are study papers 7, 8, and 9,
“ Incidence of Inflation: Or W ho Gets H u rt; Protection Against In
flation;” and “The Share of Wages and Salaries in Manufacturing
Incomes, 1947-56.”
These are among the number of subjects which the Joint Economic
Committee has requested individual scholars to examine and report
on to provide factual and analytic materials for consideration in the
preparation of the staff and committee reports for the “ Study of
Employment, Growth, and Price Levels.”
The papers are being printed and distributed not only for the use
of the committee members but also to obtain the review and comment
of other experts during the committee’s consideration of the materials.
The findings are entirely those of the authors, and the committee and
the committee staff indicate neither approval nor disapproval by this
publication.
P a u l H . D o u g las ,
Chairman, Joint Economic Committee.
N o v e m b e r 17,1959.
Hon. P a u l H . D ou glas ,
Chairman, Joint Economic Committee,
U.S. Senate, 'Washington, B .C .
D ear S e n at o r D o u g l a s : Transmitted herewith are three of the
series of papers being prepared for the “ Study of Employment,
Growth, and Price Levels” by outside consultants and members of the
staff. The authors of these papers are Seymour E. Harris, Harvard
University, Cambridge, M ass.; H . S. Houthakker, Stanford Univer
sity, Stanford, C alif.; and Alfred H . Conrad, Harvard University,
Cambridge, Mass.
A ll papers are presented as prepared by the authors, for considera
tion and comment by the committee and staff.
O tto E c k s t e in ,
Technical Director ,
Study o f Employment, Growth , and Price Levels .
CONTENTS
STUDY PAPER NO. 7, “ THE INDIDENCE OF INFLATION: OR WHO GETS
HURT?” BY SEYMOUR E. HARRIS, HARVARD UNIVERSITY, CAM
BRIDGE, MASS.
Page
Chapter 1______________________________________________________________
1
General summary__________________________________________________
1
Inflation and growth_______________________________________________
2
Kind of inflation___________________________________________________
2
3
Incidence of inflation on Government______________________________
Welfare programs: Old-age insurance_______________________________
4
Unemployment compensation______________________________________
6
Other income maintenance programs_______________________________
7
Mortgages and the inflationary process_____________________________
9
Wages, other shares, and prices_____________________________________
10
Attempts to beat inflation_________________________________________
11
Some international aspects_________________________________________
12
Chapter 2______________________________________________________________
12
The problem_______________________________________________________
12
Chapter 3 ______________________________________________________________
14
Inflation and growth_______________________________________________
14
Introductory__________________________________________________
14
Inflation, a necessary cost of growth?---------------------------------------15
The case for growth___________________________________________
16
Growth and prices: the historical record________________________
18
Prices and output— by countries_______________________________
20
Prices and per capita output in recent years____________________
21
22
Creeping inflation leads to what?--------------------------------------- #-----Recent movements of prices and output in individual countries. _
23
Chapter 4 ______________________________________________________________
24
The incidence of inflation on Government__________________________
24
24
Introductory__________________________________________________
Distribution of tax burdens among different levels of Govern
ment________________________________________________________
26
Percentage of changes in assessed valuation of real and personal
property in large cities, 1940-53__________________ ___________
33
Some empirical material on the responses of taxes to income and
price changes_______________________________________________
34
Inflation and the yield of income tax___________________________
36
Some conclusions______________________________________________
37
Chapter 5 ______________________________________________________________
38
Welfare programs and inflation_____________________________________
38
Introductory__________________________________________________
38
Adjustment of benefits to rising prices and incomes------------------40
Raising maximum covered wages_______________________________
42
Minimum and maximum benefits and prices_______ _____________
43
Increased benefits through increased coverage of family members.
46
Old-age assistance and other assistance programs_______________
48
Some comments on old-age insurance and assistance__________
55
Chapter 6______________________________________________________________
56
Unemployment compensation______________________________________
56
56
Introductory__________________________________________________
Conclusion____________________________________________________
65
Chapter 7______________________________________________________________
66
Income maintenance payments with particular reference to veterans’
programs________________________________________________________
66
Workmen’s compensation______________________________________
66
Long-term commitments______________________________________
68
Veterans’ benefits_____________________________________________
69
Pension funds_________________________________________________
74
V II
VIII
CONTENTS
Page
Chapter 8 ______________________________________________________________
Assets and inflation________________________________________________
Chapter 9 ______________________________________________________________
Mortgages, installment payments, and the inflationary process______
Chapter 10_____________________________________________________________
Wages, other shares, and prices____________________________________
Introductory: the causes of inflation___________________________
Relation of price movements and other variables_______________
Some historical evidence_______________________________________
Teachers______________________________________________________
Productivity, prices, and wages________________________________
The share going to labor_______________________________________
Inflation and depreciation_____________________________________
Prices and wages in cycles_____________________________________
Conclusion____________________________________________________
Chapter 11_____________________________________________________________
Attempts to beat inflation__________________________________________
Chapter 12_____________________________________________________________
Some international aspects_________________________________________
76
76
80
80
83
83
83
86
88
89
91
96
98
99
103
104
104
109
109
TABLES
Table 3-1. Trend of gross national product and personal consumption,
1839-1959___________________________________________________________
Table 3-2. Rise of wholesale prices and industrial output, 1947-57______
Table 4-1. Tax revenues and percent Federal, State, and local tax revenues
of total tax revenues, all levels of Government, 1902, 1927, 1938, 1948,
1957_________________________________________________________________
Table 4-2. Percent tax revenue of general revenue_______________________
Table 4-3. Percent insurance and trust revenue of general revenue______
Table 4-4. Percent selected tax categories to total tax revenue, 1902, 1927,
1938, 1948, and 1957_________________________________________________
Table 4-5. Percentages of family personal income taken by the Federal
individual income tax at successive levels of income, all consumer units,
1941 and 1950________________________________________________________
Table 4-6. Percentage of family personal income taken by the Federal
individual income tax, for quintiles of consumer units ranked by size of
family personal income, 1941 and 1950_______________________________
Table 5-1. Social welfare expenditures in the United States under certain
public programs, fiscal years 1934-35, 1945-46, 1956-57_______________
Table 5-2. Average monthly payments to retired worker beneficiaries
under three social insurance programs to veterans of World War I
receiving pensions and to old-age assistance recipients, June 1948 and
June 1958____________________________________________________________
Table 5-3. Old-age, survivors, and disability insurance: Average monthly
old-age and widow’s benefits in current-payment status, in current and
in September 1958 dollars, 1940-58___________________________________
Table 5-4. Annual earnings for a full-time employee, contributions and
benefits under old-age, survivors, and disability insurance, 1935-58____
Table 5-5. Public assistance benefits: Average monthly payments in
current and 1958 dollars, 1940-58_____________________________________
Table 5-6. Legislative chronology of provision for Federal participation in
assistance____________________________________________________________
Table 6-1. Unemployment benefits for full-time work in relation to price
and wage movements— 1939, 1946, 1952, 1957_________________________
Table 6-2. Percentage changes in average monthly cash income and out
lay of household before and during unemployment of claimants in oneand four-person households___________________________________________
Table 6-3. Average monthly wages and unemployment benefits of claim
ants compared with total average monthly cash income of households
before and during unemployment of claimants in one- and four-person
households___________________________________________________________
Table 6-4. Average weekly cash outlay of household on all expenses and
selected types of expenses during unemployment and average weekly
benefit amount of claimants in one- and four-person households^_____ 61
18
24
27
29
31
34
36
37
39
40
41
45
49
52
58
59
60
CONTENTS
Table 6-5. Amount of unemployment insurance payments per dollar of
contribution--------------------------------------------------------------------------------------Table 7-1. Veterans’ Administration budget expenditures under present
laws, selected fiscal years, 1940-2000--------------------------------------------------Table 7-2. Estimated total living war veterans and veterans aged 65 and
over, selected dates, 1940-2000_______________________________________
Table 7-3. Public expenditures for income maintenance programs and
the national income, selected years, 1940-85__________________________
Table 7-4. Estimated average cost of veterans7 benefits per serviceman in
each war_____________________________________________________________
Table 7-5. Per capita income in four major wars, compared with veteran
pensions per serviceman in each war, under present laws and assuming
service pension for recent conflicts-----------------------------------------------------Table 7-6. Rise in veterans7 benefits for 100 percent disability and rise in
the cost of living, 1914-54, various years______________________________
Table 8-1. The share of national saving in changes in national wealth and
combined net worth, period totals: 1897-1949-------------------------------------Table 8-2. Comparison of personal saving and net worth changes for the
period 1901-49_______________________________________________________
Table 8-3. Ratio of intangible assets to total assets_____________________
Table 8-4. Percent price-sensitive to total assets, major saver groups,
1900-1949______ _____________________________________________________
Table 8-5. Debt ratio of major saver groups, 1900, 1929, 1949__________
Table 9-1. Characteristics of one-family new home transactions, FHA,
section 203, 1952 and 1957___________________________________________
Table 9-2. Characteristics of FHA-insured and VA-guaranteed mortgage
loans on new houses, 1948 and 1956__________________________________
Table 10-1. Behavior of selected economic variables for manufacturing
industries, 1949-56___________________________________________________
Table 10-2. Major gains by factors, various periods_____________________
Table 10-3. Cost of living and real wages (1890-99 equals 100)__________
Table 10-4. Average rates of increase in productivity, total input per
man-hour, and real hourly earnings, 1889-1957________________________
Table 10-5. Indexes of labor and nonlabor payments per dollar of real
product, prices real product per man-hour, employee compensation per
hour in current and constant dollars, private nonagricultural sector of
the economy, 1947-56________________________________________________
Table 10-6. Output, real hourly earnings, output per unit of input, and
prices, 1899-1953, (1899 equals 100)__________________________________
Table 10-7. Ratio, rise of real hourly earnings to the increase of the total
output to the total input, 1899-1953__________________________________
Table 10-8. Average hourly earnings, prices, and production in manu
facturing, 1921-38___________________________________________________
Table 10-9. Changes in gross average hourly earnings (including overtime)
wholesale prices, and production (seasonally adjusted) for selected indus
trial groups the three postwar recessions. _____ ________________________
Table 11-1. Changes in the nominal value of various types of investments,
1948-58, in current dollars and in dollars of 1948 purchasing power____
Table 11-2. Some common stock declines since 1900 versus consumer
price index___________________________________________________________
Table 12-1. Comparison of domestic inflations, autumn 1957____________
Table 12-2. North America’s trade in manufactures and total, 1955-58-_
IX
Page
65
70
70
71
72
73
74
77
78
79
79
80
81
83
84
87
88
92
92
95
95
100
101
106
107
110
112
x
CONTENTS
STUDY PAPER NO. 8, “ PROTECTION AGAINST INFLATION,” BY H. S.
HOUTHAKKER, STANFORD UNIVERSITY, STANFORD, CALIF.
Contents:
The summary--------------------------------------------------------------------------------The analysis: Protection against inflation__________________________
The burden of inflation________________________________________
Redistributive effects__________________________________________
A proposal for improved statistics______________________________
The fear in inflation___________________________________________
Hedging against inflation___________________________________________
Miscellaneous items___________________________________________
Equities_______________________________________________________
The pattern of yields----------------------------------------------------------------------In the long run________________________________________________
Some reservations_____________________________________________
Will inflation continue?________________________________________
Differences in expectations_____________________________________
Insurance policies linked to share prices________________________
The need for assets_________________________________________________
Two proposals--------------------------------------------------------------------------------Index bonds___________________________________________________
Increasing the supply of shares_________________________________
The Government bond market_________________________________
Appendix: The return from common stocks_________________________
117
118
118
118
119
120
120
120
121
122
122
124
124
126
127
128
129
129
131
133
134
STUDY PAPER NO. 9, “ THE SHARE OF WAGES AND SALARIES IN
MANUFACTURING INCOMES, 1947-56,” BY ALFRED H. CONRAD, HAR
VARD UNIVERSITY, CAMBRIDGE, MASS.
TABLES
Table 1. Income shares— All manufacturing, 1947-56-----------------------------Table 2. Frequency distribution changes in share of wages and salaries in
manufacturing value added, 1946-57__________________________________
Table 3. Profit margins, 200 large manufacturing corporations___________
Table 4. Simple correlation coefficients: annual changes, all 3-digit indus
tries, 1949-56________________________________________________________
Table 5. Simple correlation coefficients: average annual changes, all 3digit industries, 1949-56______________________________________________
Tables 6 and 7. Regression equations_________________________________ 151,
143
145
147
149
150
152
S T U D Y P A P E R NO. 7
T H E IN CIDENCE OF IN F L A T IO N : OR W H O
GETS H U R T ?
(B y S e y m o u r E. H
a r r is )
XI
STUDY
PAPER
NO. 7
THE INCIDENCE OF INFLATION: OR WHO GETS
HURT?
(By Seymour E. H arris1)
C h a p t e r 1. G e n e r a l S u m m a r y
In this paper, I am concerned with the problem of the effects of
inflation on different groups of the population. M y major interest
is, of course, the inflation that has confronted the country in recent
years. In order to get some measure of the effects of inflation on
different groups, I have also had to take into account earlier inflations
and to consider what happened as prices rose.
I should say at the very outset that it is not always easy to dis
entangle the effects of inflation from other forces. For example,
through long periods of our history we have not only had a modest
amount of inflation, though there have been important periods of
deflation, but we have also had rising gross national product (G N P )
associated in part with an increase in population and m part with the
growing productivity of our economy. Often the damage done by
inflation results from its contribution to a failure to adjust prices
or incomes of various groups to the general trends in the economy.
But if average per capita income is rising and if returns to some ele
ments in our economy lag in relation to the growth of the economy,
and if on top of the increase in productivity there is also a rise in
prices, then the failure to achieve adjustments may be aggravated by
the rising prices. Again we may find, for example, that the adjust
ment to rising prices or even to rising per capita incomes has been
more than adequate in some segments of the economy. But the
explanation may be in part that, for example, benefit payments in a
new social security program had been most adequate at first; and it
takes time to adapt benefits to the general standard of living. It
may then be found that benefits rise more than the price level and
yet a correct analysis may very well suggest that, in the absence of
inflation, benefits would have risen more in relation to the real stand
ard of living. In other words, what seems to be a response to inflation
or to a rising per capita income may, in fact, be an attempt to adapt
our benefit structures to the current standard of living.
11
am indebted to Messrs. Gordon Smith and Reginald Green for research help with this
paper. I am also greatly indebted to Mrs. Anna Thorpe, who in general supervised all the
details involved in producing this study.
1
2
THE INCIDENCE OF INFLATION
INFLATION AND GROWTH
W hat inflation does to the economy depends to some extent upon
the relationship between price movements and the growth in the
economy; and here by growth I mean largely the rise in per capita
income at stable prices, though, of course, one should also consider
to some extent the increase in the numbers on the labor market and,
therefore, relate it to the rise of population.
A historical examination of the movement of prices and the rate
of growth of the economy does not yield any very clear-cut results.
Apparently we have had growth of substantial proportions in periods
of falling prices as well as of rising prices. W e are equally surprised
by the extent of growth in periods of falling prices as we are by that
in periods of great inflation. One cannot be sure; but if we had not
had the fall of prices, for example, in the last quarter of the 19th
century, we might have had an even greater rate of growth. There
is a considerable amount of evidence that the economy was not getting
enough money, and the failure to achieve an adequate increase in the
supply of money may have contributed both to falling prices and
to a somewhat lower rate of growth than the potential.
K IN D OF INFLATION
A t any rate, the major controversy here revolves around the period
since 1950. Much more support is given to the general theory that
in the last few yearSj that is, from 1955 to 1958, we have had a
peculiar type of inflation when prices rose by 8 percent in the midst
of a so-called peaceful world. Many more economists now believe
that inflation is what is known as a cost-push inflation, that is, an
inflation that is brought about by rising wage and other costs. But
even here one must be cautious because this kind of inflation cannot
be continued without providing adequate supplies of money and ap
propriate fiscal policies. This type of inflation is in contrast to the
usual, the classic type, which is supposedly related to an excess of
demand, that is, an excess flow of purchasing power and of demand
in relation to the flow of goods. It is generally assumed that mone
tary policy can adequately deal with the excess and demand kind of
inflation, although increasingly we seem to want a larger contribution
from fiscal policy.
O f one thing we may be reasonably sure, namely, that in periods of
great unemployment it is important to expand the supply of money
and therefore the amount or spending greatly in order to achieve
an improvement in the economic situation. This was shown clearly
in W orld W a r I I , for as late as 1939, unemployment was still at 9.5
million, or about 17 percent of the labor force. By 1944 military
purchases had risen to $88 billion and unemployment had fallen to
1 million.
W e may conclude that if inflation is one of the costs of growth, then,
of course, against the inequities and adverse effects of a given degree
of inflation on the economy, we have to put any increased amount of
output and employment that result from the inflationary process. W e,
of course, all want price stability, full employment, and maximum
growth. But often we are unable to achieve all these objectives, in
part because our instruments are blunt. W e may have to choose be
THE INCIDENCE OF INFLATION
3
tween a given increase in prices or a given rise of output. The great
current debate is whether we should sacrifice stability of prices to
some extent in order to achieve larger degrees of growth and less
unemployment. O f course, we all would acknowledge a 10 percent
rise ox output and a 1-percent inflation as splendid public policy, just
as we would all denounce a policy that yields a 10-percent rise of
prices and a 1-percent increase of output.
INCIDENCE OF INFLATION ON GOVERNMENT
W hat of the incidence of inflation on Government? Undoubtedly,
because of the time consumed in the process of increasing public ex
penditures to a changing price level, there is a tendency when inflation
progresses for Government to increase its spending less than the auto
matic rise of receipts. The gains are especially large for the Federal
Government where receipts tend to rise more rapidly in response to a
given amount of inflation than for State and local governments, which
on the whole have more inflexible and less elastic tax revenues. Gov
ernment, of course, also gains qua debtor, for one of the results of in
flation is the financing of debt with dollars reduced purchasing power.
The largest losses through inflation experienced by those who hold
Government securities are likely to be felt by individuals and others
who hold long-term securities. In general, financial institutions other
than insurance companies and mutual savings banks tend to hold short
term securities, and therefore they have an opportunity frequently to
adjust their holdings to the changing price level. Insofar as individ
uals have stakes in insurance companies and financial institutions, they,
of course, suffer the losses resulting from the depreciation of the bond
holdings of the institutions. In general, we should note that the
upward adjustment of interest rates to inflation tends to lag greatly,
though by 1959 there was some evidence that interest rates were re
sponding more to the inflation. A s individuals dispose of their Gov
ernment bonds and other fixed return assets and move into equities, the
rate of interest on new issues tends to rise as the prices of old issues
depreciate. A Treasury issue of a 5-percent 5-year Federal note in
1959 is eloquent testimony of this process.
In general, it is evident that tax receipts respond much more to
inflation when the dependence is largely on direct taxes, such as in
come and corporation taxes, and on the whole receipts from sales
taxes respond more quickly to inflation than the general property tax.
In an economy with rising prices and rising per capita incomes, any
governments that are dependent on a general property tax are likely
to suffer greatly. For example, from 1938 to 1948, the yield from a
general property tax rose only from $4.4 billion to $6.1 billion, fall
ing from 34 percent of total governmental revenue to 12 percent. By
1957, however, the yield was over $13 billion and 13 percent of the
total tax revenues. This rise is explained largely not so much by the
increase of assessments and rates m response to inflation and rising
incomes of existing property, but rather to the tremendous increase of
new construction. In the 9 years from 1948 to 1957 total private con
struction rose by $241 billion. Even without reassessments of old
property and increase of tax rates, and on the assumption of a $30
rate per thousand, the new construction would yield more than $7 bil
lion of additional general property tax revenue, or roughly the in
crease from 1948 to 1957.
4
THE INCIDENCE OF INFLATION
Taxes that respond well to rising income and prices, such as the in
come and corporate taxes, are increasingly used. The ability of the
Federal Government to finance the large demands being made upon
it, is explained in no small part by its recourse to direct taxes. Cor
porate and personal income taxes of all governments accounted for
about 60 percent of total taxes by 1957 as compared to only 23 percent
in 1938, and 24 percent in 1927, and virtually no contribution m 1902.
Under the pressure of losses of general property taxes, that is, the
failure of revenue to respond to rising prices and incomes, State and
local governments increasingly depend upon sales receipts and the
like. In 1927 these taxes accounted for one-quarter of State revenue,
but by 1957, they accounted for 60 percent of total State revenue.
A s a result of the rise of prices and incomes and the failure of reve
nues from property taxes to respond adequately, and even the inade
quate help from sales receipts and taxes, the Federal Government
tends increasingly to finance State and local governments. In other
words, the flexibility of the Federal tax programs in periods of rising
prices and incomes is to some extent passed on to other governments.
In a similar way, the State governments with more responsive taxes
tended to finance local governments increasingly.
Again, we note that under our present system of taxation, inflation
tends to result in larger tax receipts as incomes rise, and therefore the
taxpayer has to pay a higher rate of tax at higher brackets, an auto
matic result of inflation on tax receipts. Quite apart from any change
in the statutory tax rates, inflation tends to shift taxpayers into higher
brackets income tax liability and therefore increases the effective tax
rates.
w elfare programs
:
o ld a g e in s u r a n c e
W hat of welfare programs and inflation ? Here is one of the costly
areas where inflation has a serious effect, for those who suffer are
generally low-income groups with little capacity to increase any sup
plementary income. One reason for this is that adjustments in their
benefits are made slowly in response to rising prices and income. In
sofar as the financing is done by State and local governments with
their unresponsive taxes, the danger is even greater.
Under our most important insurance program, old-age, survivors,
and disability insurance, the effects of inflation in the forties were
serious indeed. From 1938 to 1948, in dollars of stable purchasing
power, the average benefits dropped by about one-third. The result
was that the beneficiaries under this program experienced a reduction
of real income of about 50 percent in relation to the rest of the popu
lation, which had experienced a substantial rise in their real income.
By 1950, however, the Federal Government began to revise the pro
gram every 2 years and this has continued up through 1958, so that
by 1958 the real value of benefits under this program had increased
substantially above the prewar level, though not as much as the real
per capita income of the whole population. This experience does
suggest the need of numerous revisions of the tax and benefit programs
if the inflation is not to have a serious effect on the most vulnerable
groups in our society— namely, the old who primarily are not members
of the labor market and therefore are inhibited in any attempts to
adjust their income to the rising price level. In fact, we might argue
that if a 2 percent inflation is absolutely necessary, it would be wise to
THE INCIDENCE OF INFLATION
5
have an escalator clause under this program. Currently this would
only cost a few hundred million dollars a year if inflation were kept,
say, at 2 percent a year. (This is not meant to be suggestive of a goal
of 2 percent inflation.) I make this point even though in general I
consider escalator clauses rather dangerous, for they tend to aggravate
the inflationary process.
One other aspect of old age and survivors’ and disability insurance
program should be noted, and that is the problem of reserve financing.
W e have now accumulated more than $20 billion of reserves in this
program, and the current estimate is more than $200 billion by the
year 2020. Policies in the 1950’s to increase taxes and reserves are to
be explained in no small part by a fear that the old age account may
be unbalanced later on and recourse would have to be had to general
revenue. I find little to support a large reserve theory, and, m fact,
in the 1930’s, we had abandoned the general principle of a large re
serve on the theory that this tends to have a deflationary effect on the
economy; and also on the theory that every generation has to provide
the resources for its own old, irrespective of financing policy. I would
readily admit, however, that financing policy is not exactly an irrele
vant problem.
But these large reserves, which are on the whole underestimated be
cause they are based on the theory that present wage levels will con
tinue, are not easily supported. In fact, we might very well estimate
that the reserves on the basis of realistic anticipation of the wage
levels might very well not be $200 billion but $600 to $800 billion by the
early part of the 21st century. But what actually happens, through
the inflationary process and the rise of per capita income related to
other factors, such as productivity, is that these large reserves are
greatly eroded through the process of inflation and the rise of per
capita income. Hence, there is a considerable waste here. It would
be much better either to reduce the tax rate or, more sensibly, to in
crease benefit rates now and not build up such large reserves. In this
manner the effects of inflation might be considerably neutralized.
The management of the program is subject to one other serious
criticism— namely, that in the 1940’s, when inflation was in vogue, no
serious attempt was made to increase payroll taxes. This was just
the period when payroll taxes should have been increased. Related
also was the failure to raise the amount of wages subject to taxes.
Under the original legislation, the ceiling of wages to be covered was
put at $3,000, and this figure was unchanged until 1950, when it was
raised to $3,600. Even by 1958 the rise was only to $4,800. A ceiling
related to wage trends should be about $9,000. A t any rate, it ought
to be considerably above $4,800. Had the ceiling been raised with the
inflation, and with the rise of j)er capita income, then, of course, more
resources would have been available; and particularly in view of the
manner in which the program works, the low-income groups would
have especially gained from such increases in coverage. Benefits
would have been more nearly adequate. Even today (1959), benefits
average only about $70 monthly.
In discussing the old-age insurance program, I should add one
other point— namely, that to some extent adjustments of benefits are
not made merely through an increase in the benefits, say, for the re
tired worker, but also, as the years go on, in making benefits avail
48084— 59------ 2
6
THE INCIDENCE OF INFLATION
able to other members of the household— for example, the widow,
children, parents, and the like— who are dependent upon the bread
winner who is now retired.
A word about old-age assistance. Here the response to rising prices
and incomes is greater than under old-age insurance, all the more sur
prising since this is an aid, not an insurance program. The explana
tion in part is that the Federal Government provided increasing
amounts of funds to be matched by State and local governments.
These governments, anxious to get the maximum subsidy from the Fed
eral Government, tend to be generous in their assistance programs,
with the Federal Government playing a large and increasing part.
U N E M P L O Y M E N T C O M P E N S A T IO N
W ith inflation, the percentage of wages covered by unemployment
compensation has tended to fall over the years. O f course, inflation
alone should not be held responsible for this unfortunate trend. The
rise in average wages at stable prices is also part of the explanation,
together with the ceiling on benefits in dollars to be paid to the un
employed worker. These weekly ceilings on benefits tend to become
more restrictive as average weekly wages rise, either because of in
flation or of the rising productivity of the economy. In December
1937, the maximum benefits in most cases were $15 per week, with
95 percent of the covered workers subject to this maximum. This
maximum benefit in 1939 was in excess of 60 percent of the average
weekly covered wages in 31 States and less than 50 percent in 2 States.
But by 1952 the total was in excess of 60 percent in 2 States and less
than 50 percent in 40 States.
Y et the benefits under unemployment compensation have responded
to inflation and the rise of per capita income or weekly wages much
more than those under old age insurance. One reason for this is that in
general the program is tied to wage levels. That the results on the
whole have been disappointing is due both to the ceiling on weekly
benefits and also the spread ana effects of the merit-rating programs.
Under merit rating employers are allowed to reduce their payroll taxes
in accordance with their employment record. The result has been
that the benefits have been kept down as payroll taxes, instead of being
levied at close to 3 percent of payrolls, have been little more than 1 per
cent of payrolls. Ceilings on benefits, therefore, and the low taxes
made possible by the merit-rating program have made it difficult to ad
just benefits to the rise in prices and wage levels.
In 1939, the average weekly payments to total unemployment was
$10.66; in 1946, $18.50; in 1957, $28.21. In dollars of stable purchas
ing power there had been an increase in these years from $10.72 to
$13.26 and $13.90. Hence the increase was substantially less than the
average weekly wage. Weekly manufacturing wages from 1939 to
1957 rose by 230 percent, whereas in current dollars the average weekly
payment for unemployment rose by 170 percent.
Of course, these are not the only income maintenance programs.
There are many others—for example, workmen’s compensation and
veterans’ benefits and various private employee benefit plans.
Just a word about workmen’s compensation, with problems similar
to unemployment compensation. Here, again, the tendency has been
for the benefits to rise less than rising prices and wages. Again, the
THE INCIDENCE OF INFLATION
7
general prevalence of ceilings on benefit payments contributes toward
this failure of adjustment. In their book, the Somers conclude that
in workmen’s compensation we have approached pretty close to a flatrate payment and the adjustment to rising prices and wages is slow
indeed. Interstate competition results in large pressures on State
governments not to increase benefits which involve increased taxes on
industry.
OT H E R IN C O M E M A IN T E N A N C E PROGRAMS
Wherever these maintenance programs require long-term commit
ments, there is the possibility that expectations are likely to be dis
appointed. The point is that the trend of incomes is upward as pro
ductivity increases and prices rise and a pension program, for example,
which is based on present-day wages or income, will be inadequate at
the income levels of the future. In colleges we have found, for ex
ample, that when a pension program was set up on the theory that 50
percent of the wages at the end of the working period would be avail
able, what is actually available comes to about 25 percent. In other
words, if one sets up a pension program for young employees at the
present time, it would be wrong to assume that the average wage
would be, say, $4,000— the current level— but rather, say, $8,000, with
in a period of 25 years and substantially more at the time of retirement
of the current young worker. But few programs of this kind are based
on such projections. Public maintenance programs, which accounted
for $2 billion in 1940, were up to $11 billion in 1955 and were estimated
at $18, $24, and $30 billion in 1965, 1975, and 1985. Hence, the need
of adequate adjustments to the rise of prices and incomes in these pro
grams.
It is of some interest that the President’s Commission on Veterans’
Benefits in the United States (1956) would not adjust benefits to in
come levels but rather would base benefits on the minimum needs, as
assistance programs are, and benefits, in their view, should in the long
run be lower than benefits paid under old age insurance. The Vet
erans’ Commission was concerned that the cost of these veterans’ bene
fits would grow disproportionately in relation to G N P. But, on the
whole, they tended to underestimate the G N P and, moreover, in tying
the problem to the total charge on all maintenance programs they
leave out of account the large contribution that has been made by the
insured.
The average cost of veterans’ benefits per serviceman in each war has
been estimated at $3,700 for the Civil W ar, $12,200 for the Spanish
American W ar, $12,700 for W orld W a r I , $14,100 for W orld W ar II ,
and $14,900 for the Korean conflict. Apparently the response to ris
ing prices and incomes was large after the Civil W ar, in part because
of the very low benefits for the Civil W ar veterans; but to both rising
prices and per capita income the adjustment was most inadequate after
W orld W ar I, W orld W ar I I , and the Korean war. But should the
participants in recent wars receive ultimately the same service pen
sions as for the earlier conflicts before W orld W ar I I , then the re
spective figures would be $3,700, $12,200, $22,000, $28,700, and $34,500.
The losses in dollars of stable purchasing power since W orld W ar I
can be associated in part with inflation.
For 100-percent disability, the basic rate was $30 per month in W orld
W a r I, $80 in 1919, $100 in 1924, and $181 in 1954. The increase by
8
THE INCIDENCE OF INFLATION
1954 greatly exceeded tlie rise in the cost of living, with the cost of
living at 263 in 1954 (1914=100) and benefits for total disability at
603. These benefits, which affect a small portion of the veterans, have
stayed up reasonably well with the cost of living and the per capita
income. This holds when comparison is made with the Civil W ar.
But on a W orld W a r base, the lag in relation to W orld W a r I incomes
is large indeed.
Insurance is another matter. In 1955 there were 5,600,000 poli
cies still in force on a waiver status, with a total of $37 billion in na
tional service life insurance. Here, of course, the veterans have ex
perienced substantial losses.
Private pension funds also raise some interesting problems. The
first problem is one already mentioned— the fact that incomes at time
of payment of pension life would be much higher than the incomes on
which the plans are evolved. The Fund for the Republic estimates
that 45 to 55 percent of all the wage and salary force will be subject to
these private pensions within 10 years. Currently about 18 million
are covered, and the assets are about $40 billion. Obviously, with the
rise of prices and incomes these funds become relatively inadequate.
To some extent, the pension funds try to make up for this by investing
in common stocks. For 1957 according to one estimate they had 25
percent of their funds in common stocks, and new investments were
going into stocks at the rate of 37 percent of net receipts.
The inflationary process, of course, also influence the value of assets.
Thus, in 1945 there were $152 billion of life insurance in force; by
1958, $494 billion. Obviously, those who obtained payments from life
insurance with insurance in force before 1945 have suffered large
losses. Those involved in the large increase from $214 to $494 billion,
from 1949 to 1958, have not as yet experienced large losses, but the
continuance of inflation will seriously affect the purchasing power
of money received under these policies. It is interesting, however,
that despite the large inflation since 1940, life insurance per family
has risen roughly in the same ratio as the average family disposable
income has increased. This certainly does suggest that the public is
still not very much aware or very much interested in the process of
inflation; and one of the greatest blocks to inflation is the lag in the
general realization of its presence. Insurance companies are not
able to protect themselves against inflation through purchases of com
mon stocks in part because of legal restrictions. O f $108 billion of
assets held in 1958, the U .S. insurance companies held only $4 billion in
common stocks.
A n examination of the total assets of the Nation suggests that from
the early part of the century until 1949 savings accounted for roughly
one-half the increase in wealth, the other major factor being, of course,
the increase in the price level. Savings accounted for about twothirds of the rise of net worth. Inflation has made its largest con
tribution to the rise in asset value in real estate. Between 1900 and
1949 the current value of real estate held by individuals increased by
about $200 billion more than the owners’ savings; and this represents
well over one-half of the individuals’ total unearned increment.
The large rise of liquid assets in relation to total assets from 7.8
percent in 1900 to 19.7 percent in 1949 also suggests the vulnerability
of those that hold liquid assets. These assets remain unchanged in
dollar value as prices change. The holders suffer losses as prices rise.
THE INCIDENCE OF INFLATION
9
This increase in the proportion of liquid assets and intangible assets
in relation to total assets, points to vulnerability to inflation. House
holds, in particular, increased their share of intangible assets to total
assets from 37 percent in 1899 to 53 percent in 1949. The proportion
of the rise for business enterprise was not as large, and governments’
increase has been from 18 to 36 percent. Price-sensitive assets to total
assets for households declined from 81 to 65 percent of total assets,
for business enterprises from 50 to 40 percent, for nonfinancial cor
porations rose from 67 to 75 percent, and for governments declined
from 71 to 50 percent. Those that experienced declines in price sensi
tive assets are likely to be affected by the inflationary process to that
extent.
But we should also consider the change in debt ratios. Here, house
holds experienced losses in the sense that they did not gain as much
from rising prices with a decline of debt ratios, for their debts declined
from 10 to 8.6 percent. The debts of business enterprises rose from
54 to 57 percent, and governments from 48 to 156.2 percent. Here
governments are the main gainers and households the main losers.
M ORTGAGES A N D T H E IN F L A T I O N A R Y PROCESS
In late 1959, total mortgages were about $181 billion, and they are
held largely by financial institutions. W ith the rise of prices and the
increase in per capita incomes, the mortgagee, of course, gains with the
passage of time. His payments in dollars of stable purchasing power
tend to decline and in relation to the total income available even more.
Nonfarm residential mortgages absorb an increasing percentage of
new savings.
It is of some interest that when we examine the income distribution
of those who get into debt on installment payments of all kinds, we
find that the lowest incomes do not incur much indebtedness of this
type. For example, those with incomes of less than $1,000 in 1957 and
1958 incurred little indebtedness; for all incomes, 52 percent did not
have installment debt payments to make, but 73 percent of those with
incomes under $1,000 had no debt payments. Then the amount of
indebtedness tends to increase up to incomes of $6,000 to $10,000, where
they seem to be at a maximum, and then decline after that.
One of the peculiar developments of our Federal housing program
has been the tendency for houses to be covered under Federal guaran
tees at higher and higher values. For example, in 1952 the average
property value under F H A , section 203 guarantees, was $10,022; by
1957, the total was $14,261. Hence the appeal had to be made to higher
and higher income groups, because per capita income did not rise
nearly as much. Therefore, Federal policy tends to favor especially
the middle and high income groups who profit further from the large
mortgages currently being put on houses. In view of the increased
tendency to put large mortgages on houses, it might be assumed that
the burden of mortgages tends to rise vis-a-vis income. But this does
not seem to be true for two reasons: first, because the mortgages tend
to go to higher income groups, and, secondly, because there are longer
period mortgages and this tends to bring down the annual cost. On
reasonable assumptions concerning the rise of prices and incomes, a
30-year mortgage today might well reduce the real cost in terms of
dollars of stable purchasing power by about 20 percent.
10
THE INCIDENCE OF INFLATION
W A G E S , O T H E R S H A R E S , A N D PRICES
The distributive effects of inflation on different shares of income
depend in part upon the type of inflation that prevails. Indeed, if
the current theory of the cost-push explanation of inflation is accepted,
then to that extent it may be assumed that wages lead the rise of prices,
and to that extent gain at the expense of other shares of the national
income. In fact, if we examine the relative movements of prices and
wages in the three major wars— the Civil W ar, W orld W a r I, and
W orld W ar I I — we will find a considerable lag of wages in the Civil
W ar, a substantial lead in W orld W ar I, and a much greater lead in
W orld W ar I I . This large gain for wages in W orld W a r I I should
be written down to some extent, however, in view of the control of
spending and the unavailability of goods.
A n examination of recent trade cycles or even a comparison of trade
cycles before the war and recent trade cycles does point toward a
tendency for prices to decline less and even rise in recession periods,
and wages, instead of falling substantially in recession periods, tend
to be stabilized and wage rates even rise. Here again we note a tend
ency for wages to rise even in periods of recession, and even more, of
course, in periods of prosperity.
It is clear, of course, that as the economy has grown, real wage rates
have increased. Moreover, they have increased more than might be
suggested by the rise of product man-hour output or even by the rise
of man-hour output corrected for any change in prices. Other costs,
however, have also increased per unit of output about as much as
wages. In 1948-56 employee compensation per dollar of real prod
uct rose by 28 percent, and nonlabor payments per dollar of real prod
uct by 27 percent, and though the real product per employee rose by
26 percent, average hourly compensation rose by 61 percent, suggesting
an inflationary effect of rising wage rates. A s might be expected,
consumer prices rose, and actually by 22 percent.
In general, labor gained in the percentage of income, though this
gain is to some extent tied to the changing distribution of employ
ment ; that is to say, employments where labor’s share of total income
was large tend to become more important. The increase in the pro
portion of income going to labor is partly explained by the much larger
rise in the supply of capital than of labor.
A n examination of the rise of productivity, output, and real wages
over a period of 50 years yields some interesting results.
Variations in the rise of productivity are very large, as are those in
output. But the differences in real wages are considerably less.
W hat is striking is the lack of any close correlation between the rise
of output or productivity and the trend in real hourly earnings. For
example, in the electric light and power industry, output rose by 244
times, and output per unit of income by almost 17 times, and yet real
hourly earnings only rose by 189 percent and prices actually declined
by 38 percent. In contrast, in anthracite coal the rise of output was
only 51 percent; of productivity, 47 percent; of prices, 336 percent;
and real hourly earnings actually rose by 262 percent. I t might be
expected, when output rises little, prices would rise a great deal.
These figures do suggest that the gains of productivity are distributed
over the whole economy and that even industries that do not experi
ence large improvements share in these gains. In the short run, there
THE INCIDENCE OF INFLATION
11
is some evidence that in those industries where production rises greatly
real wages would also rise substantially. But the longrun trends do
not carry similar implications.
W e should mention one other aspect of the inflationary process,
namely, the tendency to set aside funds for replacement of equipment
and capital and inventories on the basis of costs of acquisition. This
practice, which is generally the one required by Government for tax
purposes, does mean that as prices rise the amounts allocated for de
preciation and replacement of inventories become inadequate. There
fore, profits seem to be larger than they really are and this stimulates
output and investment. One estimate put replacement costs of con
struction and equipment in manufacturing industries at 158 of acqui
sition costs in 1948, and 138 in 1955.
There can be little doubt in general that the wages are much more
likely to exceed the rise of productivity, corrected for the price move
ments than was true in earlier years. Senator Douglas5 study on
the movement of real wages from 1899 until 1926 showed this earlier
trend very clearly. There seemed to be some lag in the upward move
ment of real wages, though there were very serious differences among
different industries. For example, in this period teachers’ real income
went up very much as did real income in manufacturing, but in many
of the service industries and Government real wages tended to fall.
O f course, an explanation of these movements is partly the change in
demand for the products of the various groups. Education was on
the rise during this period, and State and local governments did not
have the grave financial problems they have today and therefore
teachers’ salaries rose. A t that time, also, there seems to have been
less of a tendency for the general gains of productivity to be spread
more evenly over the economy as currently.
A T T E M P T S TO B E A T IN F L A T I O N
In various ways, groups of our society tend to protect themselves
against inflation. For example, many pension contracts are now
made with an escalator clause. Again, institutional investors of pen
sion funds tend to put a larger proportion of their assests in common
stock. The recent tendency to desert the bond market and invest in
common stock has been reflected in the very large rise in security
prices. But there are limits to the extent to which this can be done.
A s the bond market is deserted and people increasingly invest in
equities, the tendency will be for equities to rise in price and the yield
to fall. W ith current tax advantages in financing through the issue
of bonds and the desertion of the bond market, the result tends to be
an increased yield on Government and other bonds. Hence, we can
explain a 5 percent yield on bonds by 1959 and of considerably less
than 3 percent on equities. In an economy that is growing steadily,
the gains of growth, as well as the increasing profits associated with
inflation, go especially to the holders of equities. Therefore, an in
vestor in choosing between bonds and stocks will expect to be com
pensated in the purchase of bonds by a few percent for any possible
rise of prices and also an additional few percent for growth. But
this movement into the stock market does result in much higher yields
on bonds, and the protection through the purchase o f equities is
gradually reduced as the yield on these securities is greatly reduced.
12
THE INCIDENCE OF INFLATION
One study showed that a given sum of money invested on December
13, 1948, would have the following real value in 1958 in 1948 dollars:
Cash, a loss of 16.7 percent; U .S. Treasury 2y2s of 1967-72, a loss
of 29 percent; preferred stocks, a loss of 25 percent; common stocks,
railroads, a gain of 106 percent; industrials, a gain of 225 percent;
public utilities, 125 percent; fire insurance, 19 percent; and New York
City banks, 81 percent. The selling price for a typical one-family
residence, a gain of 0.8 percent; farm real estate, a rise of 28 per
cent.
SOM E IN T E R N A T IO N A L ASPECTS
The dramatic change in the balance of payments to the United
States has occurred since 1957 and has again raised the question of
the relationship between inflation and the balance of payments. It
is generally known that a rise of prices tends to bring about an in
crease of imports and a reduction of exports and, therefore, an un
favorable balance of payments. In 1958 we lost about $3 billion in
gold and dollar claims to foreigners and in 1959 the losses might very
well rise to $4 billion additional. There is great concern about this
dramatic reversal in the balance of payments.
W hat is the explanation? Unudoubtedly, we have had a little
inflation in the last 3 years but the amount of inflation as com
pared to other countries has not been large. In fact, since the begin
ning of the postwar period our record has been unusually good as
compared with that of other countries. Indeed, to some extent, other
countries correct for their rise of prices by devaluation, though gen
erally the devaluation follows the price rise and if devaluation pro
ceeds under full employment conditions, prices rise rapidly to a point
suggested by the new external value of the currency.
I would emphasize especially the improved competitive position
of foreign countries that might be expected after a devastating war
as the major explanation of our adverse balance of payments— not
inflation here. I would also stress especially the large foreign pay
ments for military establishments abroad, foreign aid of other kinds,
private capital movements and the like. W ith an excess of exports
of $3 billion and payments of this nature of about $9 billion obviously
international difficulties may arise.
W hat is the solution to this problem? Clearly, a great deal of
inflation would damage our balance of payments position. I am not
inclined to believe that a 1 or 2 percent inflation would be a very serious
matter in view of the inflationary trends in the rest of the world,
especially with their large investment and development programs.
One solution is, of course, for foreign countries to reduce their
barriers to American goods. Another is for military aid abroad that is
not necessary to be cut, and also to stop encouraging excessive exporta
tions of private capital. Insofar as our competitive position has
deteriorated too much, as it has in some industries, the solution would
lie partly in reduced barriers abroad to our goods and partly in im
proved methods of production here.
C h a p t e r 2. T h e P roblem
M y task is largely to discuss the effects of inflation on the econ
omy, on the shares going to different groups of the population, on
THE INCIDENCE OF INFLATION
13
the Government responsible for raising the revenue and spending it,
and also on our international position. The “ real” problem which
we cannot completely disregard is the relationship between inflation
and growth. One of the effects of inflation is, of course, to what
extent it contributes to a rise or fall of output.
In this particular study, I am concerned primarily with what we
generally refer to as “creeping inflation.55 In the minds of most, this
means a rise of the prices of 1,2, or possibly even 3 percent, though the
last is such a serious increase that it should not perhaps be included
under creeping inflation. A long continued rise of prices of 3 per
cent a year bring serious results even in a generation.
I shall say very little about the larger doses of inflation, for example,
the experience during the French Revolution, when prices seemed to
have risen 200 times in a period of several years, the Russian inflation
after W orld W ar I, which was apparently brought on partly to
expropriate all those who had claims to rubles, or the Chinese infla
tion and certainly not the postwar Hungarian inflation when prices
rose 10 17— a world’s record. These are tremendous inflations and
they bring about problems and effects that are largely irrelevant for
the kind of problems we are discussing today. Indeed, much has
been said of the possibility of a creeping inflation bringing on a
galloping inflation. But when one considers that over 120 years
we have had a rise of prices averaging a little more than 1 percent,
I do not think the dangers are as serious as often made out, though
I am also convinced that the more we talk about these dangers the
greater the chance that a creeping inflation will result in a galloping
inflation.
I shall not discuss here wartime inflations. Indeed, in W orld
W ar I I the rise of prices was surprisingly moderate, an increase
of roughly 30 percent from 1940 to 1945; but the price rise was con
tained by various control measures, and once the controls were re
moved, we had a much more serious rise of prices in 1946 and 1947.
A t any rate, inflations in wartimes are likely to be much more than
creeping inflations despite the great progress made in handling in
flations during wartimes. I once estimated that, if one allows for
the proportion of economic resources used for war purposes, inflation
in the Civil W ar was about 12 times as great as in W orld W ar I I , and
the inflation in W orld W ar I was about 4 times as great as in W orld
W a r I I . Nevertheless, despite our increased capacity to handle in
flation during war times, it is unlikely that in a major war the infla
tion would proceed at a rate of less than, say, 2 percent a year. I
might even go further and say that in a modern atomic age the rate
of inflation, in view of the great destruction that would be done, would
be a matter of tertiary importance.
I also assume that the inflation we are discussing is genuine. By
that I mean there has been a good deal of discussion of the point that
the statistics on price rises overestimate the increase in prices, because
we allow inadequately for quality. This point has been made numer
ous times.
Professor Euggles has made this point effectively in the compendium
before the Joint Economic Committee. He said:
*
* * it would indeed be interesting if price comparisons, taking quality and
improvement into account, could be made between periods. This question might
be asked another way. Suppose an individual were given $1,000 and a choice of
14
the
INCIDENCE OF INFLATION
ordering goods either from an early postwar Sears, Roebuck catalog (say, 1948)
or a current (1957) catalog. If he were permitted to spend the money in terms
of only one catalog, which catalog would he choose? The 1948 catalog has sub
stantially lower prices, but also less advanced products. If the 1957 catalog
were chosen, it could not be said that the prices rose from 1948 to 1957, despite
the evidence of the price indexes. Different people would undoubtedly answer
this kind of question differently, but it is by no means certain that an overwhelm
ing majority would choose to spend their money under the 1948 price and product
conditions rather than on the 1957 conditions, despite the fact that the implicit
price indexes of consumer goods have risen about 17 percent since 1948.2
It is scarcely necessary to remind the reader that the extent of infla
tion depends, in no small part, upon the base period chosen. For
example, if we take 1913 as the base year for the Consumer Price Index,
we would find that the annual rate of inflation from 1913 to 1958 was
2.4 percent, but from 1920 the increase would be 1 percent, and from
1955, 2.5 percent. But from 1939 the rise would be 3.9 percent. We
often discuss the 50-cent dollar, meaning the reduction of value since
1939; but note that in comparison with 1929, the dollar is now worth
59 cents. In other words, part of the inflation was a correction of
deflation in the 1930’s.
The 1958 dollar was worth as follows, in relation to various base
years:
19131920
1922
1929-
Cents
- ___ 36
.
__
69
. ____ 58
_____ _______ 59
1933
1939
19451948___
_
Cents
45
_ 48
— 62
83
1950_
1951
1953
1955. -
.
Cents
83
90
93
_93
Source: Business Week, Aug. 3, 1959, p. 95.
I am to discuss later the problem of whether the current inflation
is an excess demand, that is a classic inflation, or a cost-push and ad
ministered price inflation. I have discussed this problem at various
times.3 Indeed, an excess demand inflation is likely to have a different
incidence on the distribution of the costs of inflation than a wage-push
inflation. A wage-push inflation obviously means that wages (and
other costs) rise first and then prices follow, and, therefore, one would
assume that labor does reasonably well in such an inflation. The
excess demand inflation may more likely result in a lag of wages behind
the rise of prices and, therefore, a less satisfactory position for labor
income.
C h a p t e r 3. I n f l a t i o n a n d G r o w t h
INTRODUCTORY
One of the strongest arguments used on behalf of inflation is that
it is a necessary price to pay for progress. The argument seems to be
especially valid when applied to the history of the 1950’s. But it
could also be relevant for an understanding of our history since 1939.
Only when a wartime demand was superimposed on a modest peace
time demand did this country achieve real prosperity. As late as 1939,
unemployment was still at 9.5 million or about 17 percent of the labor
2 Joint Economic Committee, “Relationship of Prices, Economic Stability, and Growth,”
compendium of papers submitted by panelists appearing before the Joint Economic Com
mittee, March 1958, p. 298.
8 For example: the “ Investigation of the Financial Condition of the United States. Pt.
7 : An Analysis of Pts. 1 -6 Held During the 85th Congress,” Senate Committee on Finance,
August 19159s, pp. 2140-214i3.
TH E INCIDENCE OF INFLATION
15
force. Then came defense and war and by 1944 military purchases
had risen to $88 billion and unemployment had fell to 1 million.
Prices rose by 25 to 30 percent and the real rise was concealed by
controls. A price rise seemed to be a necessary condition for increased
output. Growth is one aspect of the problem of incidence of inflation.
But I do not want to carry this argument too far. The historical
and geographical material summarized in this chapter does not yield
a clear-cut picture of the positive relation of price rise and output.
But it does suggest that even with very large inflation (and in some
instances a galloping inflation) output continues to rise. But the
effects of galloping inflation are so serious in other respects that few
would endorse it.
Perhaps the strongest argument for inflation as a price to be paid
for rising output can be raised by the experience in the last 10 years.
Here there seems to be a real danger that excessive concern with the
price level and restrictions on monetary supplies may bring unneces
sary curtailments of output and recessions.
IN F L AT IO N , A NECESSARY COST OF GROWTH?
The relationship between the price level and growth has had a long
history. The view widely held for many years has been that when
prices rise output also tends to rise. The explanation of this, in part,
is that with rising prices the businessman profits because costs do not
increase as rapidly. The assumption is that wages lag behind and
fixed charges also tend to become less burdensome as prices and sales
increase. Of course, as trade unions become more powerful and as
they tend to eliminate the lag and perhaps even anticipate the rise
of prices through an increase in wages, this argument loses a con
siderable amount of its force.
But in recent years, the problem has become of much greater general
interest. The charge has been made, for example, that the present ad
ministration is too much concerned with the stability of the currency,
and in order to achieve absolute stability endangers the economy by
reducing supplies of money and therefore tends to bring about a fall
of prices, or at least a reduction in the available supply of money and
thus induces unemployment.
There is no doubt but that the administration does stress greatly its
anti-inflation policy. Especially since the Democratic Party was in
power during the Great War and postwar inflation, the party in power
now tends to exploit the inflationary trends in the economy. Their
concern emerges partly in trying to protect the interests of those that
have claims to dollars and who would, therefore, suffer through a rise
of prices, and in particular those who have bonds, other kinds of prop
erty and income that do not respond to the rise in prices, and particu
larly older men and women who have relatively fixed incomes; and
since they are not generally members of the labor market, are unable
to protect themselves.
Many have been critical of the administration and monetary author
ity, particularly in 1953 and in 1955-57, for restrictive monetary pol
icies were followed by recessions. Undoubtedly there are other ex
planations of the recession; for example, the excessive rise of invest
ments from 1955 on, but many do believe that restrictive monetary
policies contributed toward the resultant recession.
16
TH E
INCIDENCE OF INFLATION
Professor Slichter, for example, argued that under current condi
tions of setting wages and with the great influence of the strong trade
unions it was not possible to prevent increases in wages that tend to
bring about higher prices. He then said that it is better to allow some
rise in prices and maintain output at a high level than to try to stop
the rise of prices associated with higher wages and, therefore, tend to
bring about reduced output and a recession. Though Slichter was
willing to admit that in the past growth had been accomplished by
both rising and f alling prices, he did not believe that this was possible
in the future; it is inevitable that growth would be accompanied by a
rise of prices.4
TH E CASE FOR GROWTH
Growth rate is important for many reasons. The greater the
growth, the larger the income and, therefore, the more available for
our needs for consumption as well as for investment. In this con
nection, it can also be pointed out that, on the whole, the present
administration has tended to favor investment as the main objective
of economic policy rather than increased consumption. Yet we all
know that one of the serious problems of maintaining an adequate de
mand for our highly productive economy is to maintain consumption
at an adequate level. There is no use turning out more and more in
vestment goods if the buying power is not available.
The higher the rate of economic growth, the more will be available
to meet the needs of the people that cannot be met through private
spending. In other words, a more rapid rate of growth means much
larger income and increased capacity to spend for schools, roads,
health, urban redevelopment, and similar services on which we are
underspending now. For example, the Rockefeller report shows that,
as compared to $86 billion Government purchases of goods and serv
ices in 1957, Government purchases of goods and services with a
growth rate of 4 percent might yield $153 billion in 1967.
On the assumption of expected increases in population and rela
tively small amounts of unemployment, and continued rise of produc
tivity, we can make reasonable guesses about our future gross national
product (GNP). For example, a study by the McGraw-Hill Co. in
1956 estimated that the GNP in 1955 was at $391 billion at that time,
and on assumptions in the table, would rise to $653 billion by 1970, or
an increase of almost two-thirds.
1955________________________________ __________
1970______________________________________ ____
Gross
national
product
Workers
Billions
$391
653
Millions
63.1
80.5
Hours per
week
39.9
36.0
Output per
man-hour
$2.99
4.33
Source: Hearings, Joint Economic Committee on ‘ 4World Economic Growth and Competition,” 1956,
p. 13.
In a study of economic growth the CED showed that the average
annual rise since 1880 was 3 percent (a doubling in 20 years as the
gains are compounded), and in the last 10 years, roughly 4 percent.
4 See hearings, Joint Economic Committee, 1959, on “Employment, Growth, and Prices,”
pp. 6-11.
TH E INCIDENCE OF INFLATION
17
The CED also points out that, should the rate of growth in GNP con
tinue at 3 percent, by 1975 the GNP would rise to $725 billion in 1956
prices; at that time, if the share of GNP absorbed by Government does
not change, American families would have an average income of $7,000
available for spending and savings after taxes as compared with
$5,300 in 1956.5
The Rockefeller report brings out dramatically the difference be
tween the rise of GNP under differing assumptions of growth. In 10
years a 3-percent rate of growth would yield a GNP of $83 billion in
1967 as compared with $434 billion in 1957 (in 1957 dollars). But
a 4-percent growth rate which had been achieved over the 10 years,
1947-57, would yield $642 billion by 1967 and, at an accelerated rate
of 5 percent which the Rockefeller report considers as possible, the
product would be $707 billion. Note the difference of $124 billion of
annual product obtained from a 5-percent growth as against a 3percent.6
Perhaps the most potent reason for a high rate growth of GNP
is our struggle with communism. Mr. Allen Dulles made this clear at
the 46th annual meeting of the U.S. Chamber of Commerce as follows:
Whereas Soviet GNP was about 33 percent of the United States in 1950, by
1956 it had increased to about 40 percent, and by 1962 it may be about 50 percent
of our own. This means that the Soviet economy has been growing, and is
expected to continue to grow through 1962 at a rate roughly twice that of the
economy of the United States.
This indicates a rate of growth of about 9 percent for the Soviet
Union. At this rate of growth, it would not be very long before the
Soviet’s GNP would be as large as ours. This is on the assumption
that our GNP does not rise much more than 2 percent a year, a rate
of growth in some recent years but less than we are likely to achieve.
Moreover, a given GNP yields much more in security outlays to the
Soviet than to this country. For example, they devote only about
one-half as much of their GNP for services relatively as this country
does, roughly 25 percent as against 50 percent by the United States.
Hence, they can, for this and other reasons, divert resource to security
much more easily than this country can.
According to an official Soviet report, the Soviet Union’s annual rise
of industrial production averaged 11.5 percent from 1928 to 1955;
according to Hodgman’s study, 9.2 percent; according to the Joint Eco
nomic Committee staff, 7.7 percent. The corresponding production
rise in the United States was 3.6 percent from 1928 to 1955; and from
1867 to 1907 when the U.S. economy was roughly at the stage of devel
opment as the Russian from 1928 to 1955, our growth was 5.2 percent
per year.7
6 C E D : “Economic Growth in the United States, Its Past and Future,” February 1958,
pp. Ii5, 36.
6 The Rockefeller Report on the U.S. Economy, “ The Challenge to America: Its Economic
and Social Aspects,” 19(58, p. 66.
7 “ Soviet Economic Growth: A Comparison With United States,” a study for the Joint
Economic Committee by the Legislative Reference Service of the Library of Congress,
1957, p. 24.
18
TH E INCIDENCE OF INFLATION
GROWTH AND PRICES: TH E HISTORICAL RECORD
Historical record on the relationship between prices and growth
does not yield clear conclusions. Dr. Goldsmith presented the follow
ing table to the Joint Economic Committee:
T a ble
3-1.— Trend of gross national product and personal consumption, 1889-1959
[Percent increase per year i]
1839-79
1879-1919
1919-59
(2)
(4)
(5)
(5)
(1)
Gross national product:
1. Aggregate, current prices.................
2. Price level.........................................
3. Aggregate, constant prices...............
4. Population........................................
5. Per bead, constant prices.................
40-year subperiods
60-year subperiods
Entire
period,
1839-1959
1839-99 1899-1959
4.85
1.15
3.66
1.97
1.64
4.13
-.1 0
4.23
2.50
1.67
(3)
5.59
2.42
3.09
1.45
1.62
4.48
.16
4.31
2.71
1.55
5.69
1.91
3.72
1.91
1.76
4.40
1.40
2.97
1.30
1.64
i Calculated from values in 1st and last year of period.
Source: Hearings, Joint Economic Committee, 1959, on “ Employment, Growth, and Price Levels,” p 271.
This table shows that with an average increase in prices of 1.15 per
cent over the years 1839 to 1959 the national product rose in constant
prices by 3.66 percent per year. A rise of population of 1.97 percent
contributed to this increase. Against the rise of price levels of 1.15,
the gross national product per head in constant prices rose 1.64.
It will be noticed from this table that in the 60 years, 1839-99, the
rice level dropped on the average of 0.10 percent, and yet output per
ead in constant prices rose by 1.67 percent. But in the period from
1899 to 1959, when the rise of productivity, that is, the rise of output
per head in constant prices was 1.62 percent, or roughly the same as
1839-99, the rise of prices was 2.42 percent per year. When these 120
years are broken into 40-year subperiods, it will be found that the
largest average rise of productivity was in the period 1879-1919, when
the increase was 1.76 per year, and this was also the period when there
was the largest rise of prices, 1.91 percent.8 Perhaps one might ex
pect a larger rise of output vis-a-vis the price level in the 19th century
than in the 20th since natural resources were more ample.
Discussing the duration of long swings and output, Professor
Abramovitz showed that these generally run from 10 to 20 years, and
that they involve a movement in the rate of growth from about 2 per
cent per annum to about 6 percent, an average of around 4 percent.
Professor Abramovitz did not go into the relationship between these
long periods of growth or the long swings in the growth with prices,
but he does suggest that there is some relationship between these and
the supply of money. He agrees with Profesor Friedman that—
E
* * * whereas in ordinary recessions, there is mere retardation in the growth of
the money supply, severe slumps were marked by an actual decline in the
money stock. These facts can be fitted into the story already related without
difficulty * * *. So long as the stock of money, corrected for business cycles,
rises at a sufficient rate, prosperity is well maintained, and output rises steadily,
subject only to minor recessions. Presumably, such steady growth would be
8 Hearings, “ Employment, Growth, and Price Level.
Rates of Production, Productivity, and Prices,” p 271.
Pt. 2 : Historical and Comparative
TH E INCIDENCE OF INFLATION
19
traceable to the stimulus which rising money balances afford to expenditures of
all kinds. But if the rate of growth of money balances falls below a certain
level, a fortiori, if money stock declines, demand ceases to rise fast enough to
absorb our growing capacity to produce, and investment expenditure then falls.®
Inadequate supplies of money obviously would tend to curtail the
rate of growth. In a recent period the Federal Reserve was criticized
for allowing the supply of money to rise by only 1 percent per year
when output was rising by several percent. Obviously, the Federal
Reserve depended in part upon an increased use of the existing sup
plies of money. Additional money is needed not only to finance rising
output, but also, because the public holds a rising part of its resources
in money in relation to income, to cover increased liquidity needs. In
other words, as income rises the increase in the supply of money should
ordinarily be of equal proportion to the rise of output and in excess in
sofar as the public increasingly tends to hold a larger proportion of its
income in cash.
In a paper before the Joint Economic Committee,10 Professor Fried
man notes that output and prices both rise in the upward swing of the
business cycle and both tend to decline during contractions. As we
shall show later, the recent tendency has been for prices to rise even
in contractions and also even for wages to rise during periods of eco
nomic contraction. In other words, we are not as likely, in our pres
ent institutional setup, to experience a fall of costs in periods of de
clining output.
Over the longer period, Professor Friedman finds it difficult to draw
any conclusions between output and prices.
What happens to a nation’s output over long periods of time depends, in the
first instance, on such basic factors as the resources available, the industrial
organization of the society, the growth of knowledge and technical skills, the
growth of population, the accumulation of capital, and so on. This is the stage
on which money and price changes play their parts as the supporting cast.
Indeed, Friedman holds that inflationary price rises seriously distort
the effective use of resources.
In the discussion of the cyclical relationship between prices and
output, Friedman acknowledges the usual arguments about the lag of
costs, inclusive of wages and rates of interest. He also points out
that rising prices tend to cut down on efficiency and also to reduce the
amount of savings. Over the longer period, he notes that from 1865
to 1879, a period of exceedingly rapid progress, prices were cut in half,
though he also notes that in general the larger growth of output was
when prices were rising slightly during this period rather than in a
period of price decline. In the years 1880-97, the period of generally
declining prices, the increase in output was not as large as in the
period of rising prices from 1897 to 1913. He also notes that in the
period of the 1920’s growth was rapid and prices were relatively
stable. It should be noted here., however, that there was a considerable
amount of profit inflation during this period; in other words, prices
did not fall with costs during this period.
Dean Bach holds that there is little evidence—
that inflation has either increased or decreased significantly total economic
output in the United States over the last two decades * * *.u
9Ibid., p. 431.
10 Joint Economic Committee,4‘Relationship of Prices to Economic Stability and Growth/’
compendium, 1958, pp. 251-255.
11 Hearings, “ Relationship of Prices, Economic Stability, and Growth,” Joint Economic
Committee, May 1A58, pp. 18-20.
20
THE
INCIDENCE
OF INFLATION
Bach’s general view is that, if wages and commodity prices are pushed
up through governmental monetary and fiscal policies faster than is
consistent with high employment and a stable price level, the results
will not be good.
Such a policy, fully relied on, will remove most of the incentives for sellers to
refrain from continually seeking ever larger income shares through higher
wages and prices. We will be continually faced with the necessity of accepting
inflation to maintain high level employment * * *.
*
* * Indeed, it may become increasingly difficult to have full employment
even with inflation, if inflation becomes increasingly accepted and expected.
Unions, businessmen, and farmers can readily increase their asking price fur
ther the next time around * * *.
In a general way, Professor Baumol’s position is a good deal like
Professor Bach’s.12 As long as the public expects that inflation is not
going to continue forever and believes it will end some day, this ex
pectation can lead businessmen to liquidate inventory and postpone
purchases. Another adverse development is the effect of rising prices
upon savings and with the reduced real value of their savings, spend
ings will be curtailed. This reduction in savings will ultimately have
adverse effects on investment. Again with inflation and the absence
of real economic pressure, Baumol reminds us, the public will tend to
reduce their standards of quality. Why produce better products, as
Professor Baumol suggests, if the commodity can be sold anyhow ?
PRICES AND OUTPUT— B Y COUNTRIES
Another approach to this problem is to study the relationship be
tween output and prices for a number of countries. On the basis of a
study of Professor Kuznets, Professor Eckstein has produced figures
for eight countries for a period of almost 100 years. These figures
generally reveal the relation between price and output movements in
one decade in relation to a preceding decade. In general these sta
tistics do not prove very much. We do have large rises of output
both when prices are falling and when they are rising. For example,
in the United States from 1879 to 1888 the output growth was 88 per
cent above that of the preceding 10 years, while prices dropped by
19y2 percent. From 1889 to 1898 the respective figures were 38.2 per
cent of growth and a price decline of 12.9. But from 1899 to 1908 the
output growth was 56.4 percent and price rise 9.3 percent. Almost
anyone would agree that a rise of output 10 times as great as in prices,
say, an annual 10-percent rise of output and 1 percent in prices, would
reflect a great achievement. But a 1-percent increase of output and
a 10-percent increase of prices would spell failure. An acceptable
and achievable policy lies somewhere in between. In later periods of
10 years, the relationship was more nearly a 1-to-l relationship of
prices and output rise. Of course, in the great depression the rise of
output was small, 6.2 percent over 10 years, as against a fall in prices
of 18 percent.
As might be expected, with growth, the relative rise of output tends
to decline by 10-year periods, though this varies to some extent from
country to country. These figures also suggest that one might have
a very large increase of prices and one that might be considered almost
catastrophic and yet output would rise.
u See the compendium, “Relationship of Prices, Economic Stability, and Growth,”
especially pp. 51-53.
TH E INCIDENCE OF INFLATION
21
In general, in the early 1950’s the increase in output prevailed
almost everywhere, and generally the rise of prices was much larger
than the rise of output. In this respect the U.S. record was remarkably
good compared to that of other countries. Output rose almost as
much as prices in the United States during these early years in the
1950’s, but in the United Kingdom the rise of prices was almost three
times as great, in Japan almost 1,900 times as great, in Norway 5 times
as large, in the Netherlands 4 times as large, in Italy almost 200 times
as large; in Sweden the record was unusually good with an increase
in output almost equal that of prices, and the increase of output was
no less than 50 percent. In Denmark the rise of prices was about four
times as large as that in output.
These statistics suggest again how difficult it is to draw any very
clear conclusions on the relationship between prices and output over
long periods of time, even if these statistics extend over many coun
tries. One important explanation of this fact is that there are so many
other relevant variables. In the latter part of the 19th century, for
example, despite the depressing effects of falling prices, a large increase
in output prevailed, in part because of the tremendous gains of pro
ductivity associated with the rising industrialization of the country
and the movement into more productive industries and in particular
the shift from the farms to the factories. In these years, also, in many
countries the rise in population was relatively large. In some of these
countries, the industrialization had just begun. For example, in
Japan the largest industrialization began in the latter part of the
19th century.13
Again one might compare, for example, some countries in Latin
America. One would find, for example, m a country like Bolivia and
Chile, the price rises have been tremendous in recent years and on the
whole the increase of output not spectacular. Here it is difficult to say
whether the explanation of the slow rise of output and the decline of
exports, for example, should be explained by the increase of prices or
primarily by the exhaustion of important sources of raw materials.
PRICES AND PER CAPITA OUTPUT I N RECENT TEARS
The above discussion centers largely around the relation of prices
and total output. Since during these years, the rise of population
varies among countries and also from period to period, a better ap
proach might well be to study the relation between price rise and in18 See Joint Economic Committee, “Compendium on Relationship of Prices to Economic
Stability and Growth,” 1958, especially pp. 361 and 362.
480S4— 59------ 3
22
the
in c id e n c e
of
in f l a t io n
crease in real product per capita. Professor Slichter has produced
such a table, which I now reproduce:
Austria ____ ___________
Finland___ ___________
France
__ _ _______
Spain__________________
Norway________________
United Kingdom_______
Sweden .■,.
Netlierlands......................
Percent
increase in
consumer
price
index,
1948-57
Percent
increase
in real
product,
per capita,
1948-57
124.0
87.5
76.7
55.7
51.4
50.6
46.8
46.2
93.9
131.4
47.4
134.5
2 12.6
22.7
2 14.6
220.2
Percent
Percent
increase in
increase
consumer
in real
product,
price
index,
per capita,
1948-57
1948-57
Denmark______________
Ireland_________________
Italy_____________ _____
Canada.............. ..............
United States. .................
Belgium_______________
Switzerland____________
43.2
41.8
27.9
26.2
16.9
12.6
9.4
16.1
14.8
48.1
20.7
18.4
123.0
U6.3
11948-55.
21950-56.
Source: Hearings, Joint Economic Committee, “ Employment, Growth, and Price Levels,” 1959, p. 11.
In general, Professor Slichter notes that the relationship is more in
the direction of higher output with larger increases in prices during
this period from 1948 to 1957. For example, Austria has the highest
increase of prices but also has the highest rise of real product per
capita, 124 and 94 percent, respectively. According to Professor
Slichter the eight countries with the largest increase in consumer price
level averaged 35 percent rise in real per capita income, and the seven
countries with the smallest increase averaged only 22.5 percent. In
general, the relationship between price rise and per capita output
movements is not very close.
CREEPING INFLATION LEADS TO W H A T ?
One of the arguments used against creeping inflation is that ulti
mately it becomes a galloping inflation, and the result of a galloping
inflation is serious curtailment with output. One can recall, for ex
ample, the results of the great inflation after World War I, when not
only did the effects of inflation have bad results on the structure of
production and the attitude toward production, but also tended to
result in a very large depreciation of the currency in Germany, for
example, much greater than would be suggested by the relative rise
of prices, say, in Germany and the United States. The result was
that foreigners could purchase German goods at such a bargain, fre
quently at 10 cents for the dollar as compared with earlier periods,
that the Germans were denuded of their goods. It became very cheap
to buy in Germany and very expensive for the Germans to buy abroad.
Ultimately, the Germans had to take measures to stop the large ex
portation of goods.
But it is not at all clear that creeping inflation does result in a
galloping inflation. I have already mentioned the point that actually
the average increase was only a little over 1 percent over a period of
120 years in American history. Most of that increase occurred in
wartimes when it is very difficult to prevent a substantial inflation.
In such periods the pressure of excess demand and the emergence of
bottlenecks, once a high level of output is obtained, result in a very
large rise of prices. Of course, if the rise of output and the increase
TH E INCIDENCE OF INFLATION
23
in demand occurs in a period when there is much excess capacity, as
happened in World War II in the United Staes, then to that extent
the increase in demand may very well be offset to a substantial degree
by a rise of output. It is of some interest, as Professor Slichter points
out, that for 11 countries the increase in the price level from 1953 to
1957 was substantially less, on the average, than the increase from the
years 1948 to 1953. The explanation of this fact might very well be
that, having once recovered from the disorganization caused by war,
these countries are able to increase output, raise their incomes and pro
vide less inflationary monetary and fiscal policies as the result.14
RECENT MOVEMENTS OF PRICES AND OUTPUT I N INDIVIDUAL INDUSTRIES
Perhaps a word should be said about the relationship between out
put and prices in individual industries.
In one respect, one might expect the greater the rise of output in
any one industry the lower its prices would be relative to an earlier
period. This would be based on a theory that increasing output re
sults in better use of capacity and lower unit costs. Costs tend to
fall in the period of rising output on the upward swing of the busi
ness cycle though in later stages with increased pressure and bottle
necks, the unit cost may tend to rise.
In the first study paper15 Professor Schultze presents effectively
the general theory that at least from 1955 to 1957 there is a fairly high
correlation between the rise of prices and that of output by industries.
His conclusion is:
During the 2 years after 1955, total aggregate output and industrial produc
tion rose very slowly, and by significantly less than the increase in productive
capacity. Aggregate demand was not excessive. The demands for capital goods,
for exports and military equipment, however, were in excess of potential supply,
while housing and auto demand fell well below the capacity of the two indus
tries. Instead of a realinement of relative prices around a stable center, prices
of almost all final goods and services rose. Price increases were generally
larger for those goods in excess demand, but were not confined to those goods
alone * * *.
He also writes:
*
* * On the average, stability of output was associated with substantial
price rise. But aggregate output rising less than capacity however, there was
no aggregate excess demand over this period * * *.10
For the years 1947 to 1957 I examined the rise of wholesale prices
and of industrial output in a number of industries where relatively
comparable figures were available.
14 Ibid., p. 12.
15 Charles L. Schultze, “ Study Paper No. 1— Recent Inflation in the United States,”
jptember 1959.
w Ibid., pp. I l l and 113.
24
the
T able
in c id e n c e
OF INFLATION
3-2.— Rise of wholesale prices and industrial output, 1947-57
Wholesale
prices rank
order
Output rank
1
Iron and steel_____________________________________________________
Tires and tubes____________________________________________________
Electrical machinery and equipment..___ _________________________
Nonferrous metals_________________________________________________
Motor vehicles____________________________________________________
Paper and pulp____________________________________________________
Coal__________ ________ _________________ _______ ____ _____ _________
Chemicals_________________________________________________________
Foods______ _______ _________________________ ____ ____ _____ _______
2
3
4
5
6
7
Textiles _________________________________ _____________________________
5
7
1
4
6
3
10
8
2
9
8
9
10
Here the association does not seem to be too close. Indeed, food
and textiles have relatively unsatisfactory rates of growth and also
experience relatively low rises of prices. The explanation in foods
is undoubtedly, in part, the excess supplies of agricultural products,
and in textiles the loss of consumer appeal in the American economy
for textiles in competition with other products and services.
In iron and steel and tires the rise of wholesale prices was more than
twice as large as in output. In electrical machinery and equipment
increase in output was about one-third more than in prices, in nonferrous metals roughly of equal proportion, and similarly for motor
vehicles. Output in paper and pulp greatly exceeded, in fact was
about twice as large relatively, as the rise in prices. In coal the
wholesale prices rose by 24 percent, whereas output declined by 17
percent. Here the explanation undoubtedly in part is the unusual
wage policy of the United Mine Workers Union, a policy that does not
take into account the declining market for coal. Wage policy is set
to achieve maximum gains for those who can still hold their jobs in an
industry where rising productivity and unsatisfactory demand cur
tail jobs. In chemicals, despite an increase in output of 84 percent,
the rise in wholesale prices was only 9 percent. Here is an industry
that is clearly gaining in terms of the changing composition of de
mand and the great advances being made in research.
C h a p t e r 4. T h e I n c id e n c e
of
I n f l a t io n
on
G overnm ent
INTRODUCTORY
How inflation affects Government is of great importance since
Government accounts for such a large part of the gross national prod
uct and performs such indispensable services as providing security
for the Nation, underwriting education, developing our natural re
sources, building our highways and the like. In 1929 the Govern
ment accounted for only 8 percent of the gross national product
(GNP), but by 1958, when Government purchases of goods and serv
ices were up to $91 billion, Government accounted for 21 percent of
the total GNP. By calendar year 1958, Government or Government
receipts were $115 billion, and expenditures $125 billion. The differ
ence between these figures and Government purchases of goods and
services is, of course, explained largely by the importance oi transfer
payments which are not reflected in Government purchases of goods
and services.
THE INCIDENCE OF INFLATION
25
In a study for the CED 17 Professor Eckstein noted that the effect
of inflation is likely to be a reduction in the real value of Government
budgets. For as inflation proceeds, it would take some time before
the Government would appropriate sums or spend sums that would re
flect the inflation. There are certain institutional delays involved
here, even if the Government wants to achieve a stable level of real
governmental expenditures. Because of the money illusion, of course,
there might well be a tendency for the Government to cut real ex
penditures more or increase items less rapidly than otherwise would
be the case just because of inflation. Contending that the response
of taxes to inflation would be more prompt and effective than the rise
of expenditure, Professor Eckstein concludes that, on the whole, in
flation tends to improve the state of the budget rather than the re
verse. This generalization, however, does not apply to State and
local budgets.
Dean Bach contends that the major gains of inflation are obtained
by the Government and the large losses by the public. Over a period
of about 20 years he calculates “that * * * over $500 million infla
tionary erosion of real purchasing power of creditors over the period”
has occurred. Since—
householders have consistently been heavy net creditors and governments (es
pecially the Federal Government) consistently net debtors, inflation has caused
a huge transfer of purchasing power from households primarily to the Federal
Government.
Since the Government debt is not likely to be repaid in the foreseeable
future, Dean Bach argues that the real gainers of the reduced real
cost of financing the debt are not the taxpayers but the nonbond
holders, who are therefore able to increase their share of total output
as their incomes increase with inflation. Government bondholders and
money holders are partially expropriated by inflation, and the benefit
is distributed over the whole population, with the biggest benefit for
those who buy the most. All households, according to Bach, hold about
30 percent of their total wealth in the form of fixed dollar value assets,
but they are in debt up to only a little over 10 percent of their total
wealth.18
We should note, however, the distribution of Government securities.
At the end of 1958, of the $283 billion outstanding, commercial banks
inclusive of the Federal Reserve banks held 33 percent, U.S. Govern
ment investment accounts held 19 percent, and other investors held
48 percent. In the last category individuals held 23 percent, State and
local governments, miscellaneous investors, and corporations 6 percent
each, insurance companies 4 y2 percent, and mutual savings banks 2y2
percent. Obviously, individuals experienced the largest losses from
inflation, particularly if we assume that the holdings of the banks
really belong to the depositors. Of course, to some extent the holders
of equities in banks would suffer, though insofar as their profits re
sponded to inflation or gained from it this would not be true. Govern
ment itself was also a very important loser, especially in view of the
fact that the excess profits of the Federal Reserve belong to the Gov
17 CED,
18 Joint
Growth,”
Economic
“Trends in Public Expenditures in the Next Decade,” April 1959, pp. 8, 9.
Economic Committee, “The Relationship of Prices to Economic Stability and
compendium of papers submitted by the panelists appearing before the Joint
Committee, 1958, pp. 41-42.
26
THE INCIDENCE OF INFLATION
ernment—but net on all issues, the Treasury gained. Inclusive of the
Federal Reserve holdings, Government security holdings accounted
for almost 30 percent of the total amounts outstanding.
Whereas commercial banks, excluding the Federal Reserve banks,
held just under one-quarter of all outstanding Federal securities, they
accounted for slightly over one-third of the total marketable debt
and had concentrated two-thirds of their total Federal debt holdings
in marketable issues maturing or recallable in less than 5 years. This
suggests that, unlike most individuals investors, the commercial banks
were in a position to adjust their holdings from time to time in re
sponse to the growth of inflation. On the other hand, little more than
one-tenth of IJ.S. Government funds’ debt holdings were in publicly
marketable securities and less than one-twentieth in marketable se
curities recallable in less than 5 years. As a result, these funds seem
especially subject to erosion by inflation. The U.S. Government also
tended to hold short-term issues much more relatively than their total
holdings. Here, when I refer to the U.S. Government, I include the
Federal Reserve banks as well.
Obviously, in a period of inflation, those who hold the short-term
securities make decisions from time to time as to whether they want
to continue to hold securities that pay a fixed return in inflationary
periods. It should be added, however, that over the years since 1939,
the adjustments of interest rates to the inflation have been very small.
Only in the last year or two has there been a substantial adjustment
so that with the interest rate at 5 percent and with prices rising 1 or 2
percent a year in the last 10 years, the increase in the rate on short
term issues to 5 percent suggests some adjustment to the inflation.
Of course, those that hold longer term securities experience losses
as the rate of interest rises in response to inflation. This would be
particularly true with those that held securities with more than 10
years to go. At the end of 1958, there were $ 8 ^ billion of these se
curities and relative to their total holdings, the mutual savings banks
and life insurance companies were especially involved and therefore
they or their depositors and policyholders experienced to this extent
losses from the inflation. This, of course, applies also to the large
holders of savings bonds that generally run for 10 years.19
DISTRIBUTION OF T A X BURDENS AMONG DIFFERENT LEVELS OF GOVERNMENT
It is important to understand that the large rise of GNP over the
years has been the result of an increase of productivity, a rise of popu
lation and capital, and also the inflationary process. For example,
from 1939 to 1959, GNP rose from $91 to $437 billion, but at 1958
prices the rise was from $209 to $437 billion. These figures point to a
rise of one-third to be explained by price changes and two-thirds by
the increase in real income. With income rising at such a rate, it is,
of course, important that Government revenue should respond. If
Government revenue does not respond to the rise of prices, then, to
that extent, the Government will perform a less effective job, and its
contribution in “real” dollars will decline.
18 For the details on distribution of Government securities, see hearings, Joint Economic
Committee, on “ Employment, Growth, and Price Level.” Pt. 6 : “ Government’s Management of Its Monetary, Fiscal, and Debt Operations,” 1959, pp. 1114-1117.
27
THE INCIDENCE OF INFLATION
But we can go even further than that. It would be important for
the Government to have sources of revenue that respond to rises in
total income, whether the increase is real, or the result of price rises.
Unless the Government does match the increase in the growth of real
income, the Government contribution to the life of the country would
relatively be reduced.
From 1952 to 1958, the increase of GNP in current dollars was $90
billion and in 1958 dollars $38 billion, or roughly the inflation ac
counted for about 58 percent of the rise of GNP as against only onethird from 1939 to 1958.
As I shall show presently, the structure of taxation varies a great
deal at different levels of government, and with this there is a varying
degree of sensitivity to price and income changes. Table 4-1, for ex
ample, shows quite clearly that the Federal Government’s share has
increased greatly, especially since 1938. In that year, the Federal
Government tax revenue as a proportion of total revenue was 41 per
cent, but by 1957 it had risen to 71 percent. During this same period.
State tax revenue declined from 24 to roughly 15 percent and local tax
revenue from 35 to 15 percent. It is of some interest that in 1902
State tax revenue was roughly 11y3 percent of the total tax revenues
and almost 15 percent in 1957, but local tax revenues, which were 51
percent of the total, or four to five times as large as the taxes of the
State government in 1902, had declined to a proportion virtually iden
tical with that of State tax revenue. One explanation of this very
significant relative decline in the contribution of local tax revenue is
the great dependence on the general property tax which on the whole
responds very slowly, if at all, to rising prices and rising income.
Under this pressure and for other reasons as well, the State govern
ments had to take on a much larger relative share of the burden of
government.
T a b le 4 -1 . — Taw revenues and percent Federal, State, and local tax revenues of
total taw revenues, all levels of government, 1902,1927,1938,1948, and 1957
Total tax
revenue
(millions)
1957 (preliminary)____
1948................................
1938................................
1927............... ...............
1902................................
$98,858
51,218
12,949
9,451
1,373
Federal tax revenue
Millions
$69,815
37,876
5,344
3,364
513
Percent
70.62
73.95
41.27
35.59
37.36
State tax revenue
Millions
$14,531
6,743
3,132
1,608
156
Percent
14.70
13.16
24.19
17.01
11.36
Local tax revenue
Millions
$14,511
6,599
4,473
4,479
704
Percent
14.68
12.88
34.54
47.39
51.27
Source: Adapted and calculated from **Governmental Finances in the”United States, 1902 to 1957,” GCGA-No. 9, March 1959, U.S. Department of Commerce, Bureau of the Census, tables 2, 3, 5, 6.
28
THE INCIDENCE OF INFLATION
Under the pressure of increased needs and inflexible tax systems,
State and local governments have tended to respond by obtaining ad
ditional revenues from other governments—the State governments, of
course, depending upon the Federal Government, and local govern
ments, in turn, depending upon both the Federal and State Govern
ments. Federal Government revenues are much more flexible than
State and local, and State than local. It will be noted that there has
been a steady decline in the proportion of tax revenue to general
revenue for both State and local government. In 1902 tax revenue
accounted for 94.5 and 82.5 percent, respectively, for State and local
government in relation to general revenue. By 1938, the figures were
82 and 67 percent and by 1957,79 and 57 percent. The major explana
tion here is the large contributions of the Federal Government to State
government and of State and Federal Government to local govern
ment. The largest gains, of course, have been for local government,
which obtained help from both Federal and State Governments. In
this manner, the Federal Government, with its tax system which re
sponds reasonably well to the rise in prices and increased income pay
ments, tends to pass on to State and local governments the advantage
of its more flexible tax system. It will be noted that the State gov
ernments obtain about $4 billion more out of total general revenues
than out of their tax revenues, and local governments obtain $11
billion additional.
THE
T a b le
4r-2 . — Percent tax revenue of general revenue
All levels of government
Tax
revenue
113,735
59,666
15,023
11,551
1,632
86.92
85.84
86.19
81.82
84.12
Tax
revenue
69,815
37,876
5,344
3,364
513
General
revenue
79,263
44,277
6,595
4,396
653
Percent
88.08
85.54
81.03
76.52
78.56
Tax
revenue 1
14,531
6,743
3,132
1,608
156
General
revenue
18,458
8,483
3,813
1,766
165
Percent
78.72
79.49
82.14
91.05
94.54
Tax
revenue1
14,511
6,599
4,473
4,479
704
General
revenue
25,639
11,373
6,651
5,903
854
Percent
56.60
58.02
67.25
75.88
82.44
i Intergovernmental transfers are included in State and local governments' general
Source: Adapted and calculated from “ Governmental Finances in the United States,
revenues.
1902-57,” G-CGA-No. 9, March 1959, U.S. Department of Commerce, Bureau of the
Census, tables 2, 3, 5, and 6.
IN F L A T IO N
98,858
51,218
12,949
9,451
1,373
Percent
OF
1957 (preliminary)_________
1948 ........................................
1938 ........................................
1927 ......... .............................
1902 ........................................
General
revenue
Local governments
State governments
Federal Government
INCIDENCE
[Revenues in millions of dollars]
30
THE INCIDENCE OF INFLATION
Before discussing the general structure of our tax system, I should
point out that an increasing proportion of general revenue comes from
insurance trust revenue, except from 1938 to 1948. All governments,
by 1957, were receiving $12.3 billion in insurance trust revenue as
against $1.6 billion in 1938 and, of course, the largest receipts of this
nature were Federal Government, with an increase from $630 million
to $8.7 billion from 1938 to 1957. For State governments the rise was
from $890 million to $3.2 billion.
These trust funds or trust revenues raise an especially serious prob
lem with the trend toward inflation and rising income payments. In
a general way, one would expect that these funds would pay out in
various benefits, for example, old age or unemployment, an amount
that would correspond roughly with the rise of income on a per capita
basis. We shall discuss these various funds much more fully in a
later chapter. I should say a word about the problem here. Generally,
whatever the program is— for example, it may be a civil service retire
ment program or an old age insurance program—the assumption is
that the benefits have been paid out on the basis of the income at the
time of retirement. Therefore, if benefits match the high income at
the years of retirement, it can be said that these programs would, in a
eneral way, provide benefits that match the rise in per capita income.
Lctually, in practice, this does not tend to happen. In some instances,
the benefits are given as a flat sum. This is perhaps more true in
private benefit programs than in public. Even when the sum is not a
flat sum, it is tied to some extent to current wage levels rather than
wage or income levels at time of retirement. The average actuary
does not generally assume that, for example, 25 years from now even
without inflation, the average per capita income will be about twice
as large as it is now. I f we assume inflation, the rise in per capita in
come would be even greater. Therefore, adjustments have to be made
frequently, but they are often greatly delayed and therefore the benefit
payments not only frequently do not keep up with the rise of prices but
even more so they fall greatly, relative to the rise of per capita income.
In this connection, it is of some interest that, in 1957, the proportion
of insurance trust revenue to total general revenue was roughly the
same as in 1938, and that despite the tremendous growth of these in
surance programs. At least in relation to the total position of gov
ernment, these funds are becoming surprisingly unequal to the demands
being put upon them and suggest the need of later periodic upper
revisions of benefits which would be financed either through general
taxation or much larger increases in payroll taxes.
In this connection, it is of some interest that under the old age and
survivors’ insurance program, the most important of all these pro
grams, the assumption made by the actuary is that wages will not
change. Obviously, insofar as they do change, benefits would have to
be adjusted to the rising price and wage level.
G
31
THE INCIDENCE OF INFLATION
T a b le 4 - 3 .— Percent insurance trust revenue of general revenue
[Revenue in millions of dollars]
All levels of govern
ment
Insur
ance
trust
reve
nue
Gen
eral
reve
nue
1957 *________ 12,305 113,735
1948................. 4,828 59,666
1938................. 1,584 15,023
1927.................
237 11,551
1902.................
1,632
Federal Government
State governments
Per
cent
Insur Gen
eral
ance
trust reve
reve
nue
nue
Per
cent
Insur Gen
eral
ance
trust reve
reve nue 1
nue
Per
cent
10.82
8.09
10.54
2.05
8,663 79,263
2,977 44,277
631 6,595
73 4,396
653
10.93
6.72
9.57
1.66
3,209 18,458
1,494 8,483
890 3,813
137 1,766
165
17.38
17.61
23.34
7.76
Local governments
Insur Gen
eral Per
ance
trust reve cent
reve nue 1
nue
433 25,639
140 11,378
63 6,651
27 5,903
854
1.69
1.23
.95
.46
1Intergovernmental transfers are included in State and local governments’ general revenues.
* Preliminary.
Source: Calculated and adapted from “ Government Finances in the United States, 1902-57," G-CGAMo. 9, March 1959, U.S. Department of Commerce, Bureau of the Census, tables 2, 3, 5, and 6.
I now turn to the problem of the tax structure of different govern
ments, as well as the trend in these structures. This is all elaborated
in table 4-4. Note, first, that total tax revenue rose from a little more
than $1 billion in 1902 to $13 billion in 1938, to $51 billion in 1948,
and $99 billion in 1957.
On the whole, it is generally assumed that the direct income taxes,
both personal and corporate, respond to rises in prices and incomes
much more than, for example, sales receipts, custom duty, or the
property tax. In a general way, the property tax may be assumed
to be the least flexible of all the taxes we have under consideration,
except for the poll tax.
Fortunately, the largest yield of revenue comes from the personal
income tax, for it yields $37 billion, almost 38 percent of all tax
revenue. This contribution toward the Federal tax receipts has
tended to grow greatly, though there was some decline during the de
pression, relatively speaking. By 1957, when the personal income
yielded more than half of the total revenue of the Federal Govern
ment, its contribution to State revenue has also tended to increase, but
not nearly so much as to the Federal Government, and even by 1957
State income taxes yielded only about 11 percent of total tax revenue.
The corporate income is also tending to become much more impor
tant, though its growth absolutely and relatively has not been nearly
as great as for the personal income tax. For example, its relative
contribution rose from 25 percent in 1938 to 50 percent in 1957, whereas
the relative contribution of the personal income tax has increased
more than 200 percent.
On the whole, the increased contribution of corporate income taxes
to State finance have been disappointing. One explanation of this
undoubtedly is the fear of interstate competition which tends to result
in a slower growth of direct taxes and therefore results in a less than
adequate use of direct taxes by State government, and to that extent,
of course, the response of tax revenues to rising prices and incomes is
disappointing.
Sales receipts and custom duties respond, to some extent, to increased
income, since as incomes rise consumption expenditures also tend to
32
THE INCIDENCE OF INFLATION
rise. Where there has been a very large increase in direct taxes, the
gains of consumption may not be as large as they otherwise would be.
This will explain, for example, to some extent, why GNP rose by 414
percent from 1938 to 1958 and consumption only rose by 349 percent.
In general, the trend has been downward in sales tax receipts and
custom duties, yielding 37y2 percent in 1902 and 20.8 percent in 1957.
The explanation of this fact lies largely in the greater importance of
Federal tax receipts and also in the unresponsiveness of sales receipts
and custom duties to the large rise of income.
But whereas the contribution of these taxes to its total tax revenue
to Federal Government tended to decline greatly, their contribution
to State revenue tended to increase, and by 1957 they accounted for 58
percent of all State tax revenue. The explanation of this fact is
partly that State governments were reluctant to increase their direct
taxes on income and corporations and partly because of the high rates
of these taxes imposed by the Federal Government and in part be
cause of the fear of interstate competition. On the whole, the sales
tax does not have as serious an impact on the interstate competition
as do the direct taxes. Another reason for this change is, of course,
the great flexibility of the income and other direct taxes and the
great need for additional revenue in the great depression and in
war. Sales taxes were an obvious source of additional revenue in
a period like the 1930’s, and it was particularly in the 1930’s that
these taxes became a much larger part of the total State tax revenue.
But the largest absolute gains of these revenues occurred since 1938.
Now I turn to the property tax. This is a tax that responds very
slowly to rising prices and rising incomes. One reason for this is that
it is difficult for administrators to increase their assessments as the
value of property rises. It requires great political courage to reassess
on the basis of changing values. Related is the fact that property
values frequently do not respond quickly and adequately to increases
in income and prices.
The importance of the general property tax lies largely in its
important contribution in financing school education. Under our
peculiar institutional setup, the major contribution of finance for the
public schools comes from the local government, which even today
provides close to 60 percent, whereas State government provides some
what less than 40 percent, and Federal Government provides only 3
or 4 percent to the total cost of public school education. The local
governments depend almost wholly on the property tax, roughly 53
percent in a recent year of the total cost of schools was financed by
the property tax.20 Undoubtedly, the property tax could be made into
a more flexible tax if assessment and rates were adjusted more quickly,
but here the pressure of the local taxpayer becomes a dominant factor.
But we note differences in the capacity of tax assessors to make adjust
ments. For example, it is difficult, on the basis of varying rates of
growth, to explain an increase in assessed valuation of real and per
20 See the National Citizen’s Commission for the Public Schools, “Financing Public
Education in the Decade Ahead,” December 1954, p. 58,
THE INCIDENCE OF INFLATION
33
sonal property from 1940 to 1953 of 323 percent for Chicago and only
2 percent for Pittsburgh, 6 percent for Boston, and 17 percent for
New York. Here, for example, are some increases in valuations from
1940 to 1953:
Percentage of changes in assessed valuation of real and personal property in
large cities, 1940-53
New York________________________ 17
Chicago__________________________ 323
Philadelphia______________________ 27
Los Angeles---------------------------------128
Detroit___________________________ 90
Baltimore________________________ 82
Cleveland________________________ 89
St. Louis_________________________ 85
Washington--------------------------------- 70
Boston___________________________
6
San Francisco____________________130
Pittsburgh_______________________
2
Milwaukee_______________________ 75
Houston__________________________261
Buffalo___________________________
9
Source: “Financing Public Education in the Decade Ahead!,” p. 59.
Under the _
with public sell
in 1929 to almost $14 billion in 1959, the need for additional revenue
has been great. To some extent the local governments have leaned
on State governments, which have contributed an increasing share of
the total amounts needed. Some indication of the inflexibility of the
eneral property tax is suggested by the fact that its yield was $4.7
illion in 1927, $4.4 billion in 1938, and $6.1 billion in 1948. It is of
some interest that in these years, from 1938 to 1948, the gross GNP
rose by more than 200 percent and yet the general property tax rose by
less than 40 percent, and its share of total tax revenue dropped from
34 to 12 percent.
But it should be added that from 1948 to 1957 the general property
tax recovered considerably. Its increased yield was 114 percent, as
compared to a rise of GNP of 170 percent. As a percentage of total
revenue, it rose from 12 to 13.2 in 19579 a rather remarkable achieve
ment. The explanation of this very large increase lies partly in the
great need for additional expenditure for education, partly in the
unavailability of other resources, and partly in the large increase of
new construction, which in turn raised total assessments.
In this period of 9 years from 1948 to 1957, total private construc
tion rose by $241 billion. Even if there were no reassessments of old
property and taxes were only $30 per thousand, this would yield more
than the $7 billion additional of general property tax revenues in 1957
as compared to 1948.21
The revitalization of the general property tax in the last 10 years
is worthy of note. But we should also make the point that this is a
most regressive tax which on the whole has a very serious impact on
the low income groups, and we may ask the question whether it is
desirable for the general property tax to yield as much as 13 percent
of total tax revenue and as much as 87 percent of total local tax
revenue. At least the State governments have gradually reduced the
contribution of the general property tax to its revenues, the yield
declining from 53 percent in 1902 to 3 percent in 1957.
g
21 Totals calculated from the “Economic Report of the President, 1959,” p. 175.
34
THE INCIDENCE OP INFLATION
T a b le 4 -4 . — Percent selected tax categories to total taw revenue, 1902, 1927, 1938,
and
[Revenues in millions’of dollars]
Tax
reve
nue
1957 1:
Total—
FederalState___
Local___
1948:
Total___
FederalState___
Local___
1938:
Total—
FederalState___
Local___
1927:
Total—
FederalState----Local___
1902:
Total___
FederalState___
Local—
Personal
income
Corporate
income
Sales, receipts,
customs
Property
Miscellaneous
and licenses
Amount Per
cent
Amount Per
cent
Amount Per
cent
Amount Per
cent
Amount Per
cent
98,856
69,815
14,531
14,511
37,387
35,620
1,563
205
37.82
51.02
10.76
1.41
22,151
21,167
984
22.41
30.32
6.77
20,588
11,127
8,436
1,025
20.83
15.94
58.06
7.06
13,097
13.25
479
12,618
3.30
86.95
51,218
37,876
6,743
6,599
19,848
19,305
499
44
38.75
50.97
7.40
.67
10,270
9,768
585
7
20.05
25.55
8.68
.11
12,092
7,650
4,042
400
23.61
20.20
59.94
6.06
6,126
11.96
276
5,850
12,949
5,344
3,132
4,473
1,495
1,277
218
11.55
23.90
6.96
1,498
1,333
165
11.57
24.94
5.27
9,451
3,364
1,608
4,479
949
879
70
10.04
26.13
4.35
1,351
1,259
92
14.29
37.43
5.72
1,373
513
156
704
5,634
1,902
3,089
663
5.70
2.72
21.12
4.57
4.09
88.65
2,881
1,243
1,340
298
5.62
3.28
19.87
4.52
3,815
2,021
1,674
120
29.46
4,440 34.29
37.82
53.45 ....... 244" ~~7~79
2.68
4,196 93.81
1,701
713
831
157
13.14
13.34
26.53
3.51
1,558
1,088
445
25
16.49
4,730 50.05
32.34
27.67 ....... 370" '23'61'
.56
4,360 97.34
862
137
631
94
9.12
4.07
39.24
2.10
515
487
28
37.51
706 51.42
94.93
17.95 ......... 82" ~52~56~
624 88.64
152
26
46
80
11.07
5.07
29.49
11.36
1 Preliminary.
Source: Calculated and adapted from “ Governmental Finances in the United States, 1902 to 1957,” GCGA-No. 9, March 1959, U.S. Department of Commerce, Bureau of the Census, tables 2, 3, 5, 6.
SOME EMPIRICAL MATERIAL ON TH E RESPONSE OF TAXES TO INCOME AND
PRICE CHANGES
In this particular study, we have used a nonparametric test which
gives the significance and the difference between two categories of a
number of paired observations, each of which may have been gen
erated under different conditions. For example, does the percentage
change from year to year in the income tax yield, when rates did not
change, exceed that in the assessed valuation of property a significant
number of times so that we may conclude that the city income tax re
sponds more than the city property tax, given constant rates? We
used six cities which have had income taxes for sufficient periods of
time, and observations were taken for St. Louis 1953-57, Philadelphia
1940-48 and 1950-55, Louisville, Ky., 1949-55, Columbus, Ohio, 194855, Toledo, Ohio, 1946-56, Dayton, Ohio, 1951-56.
We had to be content with comparing not income but income tax
receipts and not property tax receipts but assessed valuations. The
number of paired observations were 36 and the number of observations
in which the change in income tax-yield exceeded that in assessed
valuation was 31. We can be more than 98.8 percent certain that the
city income tax responds better to income changes than the property
tax, given constant rates. This is about what we might expect. These
results are, however, subject to some reservation because of the effect
of the adoption of an income tax, for example, on the collection of the
property tax. A similar study for years of declining income indicates
that we can be 99.9 percent sure that the income tax responds more to
income declines than the property tax for cities.
35
THE INCIDENCE OF INFLATION
Here are some figures that suggest the range of changes:
[Percent]
City
St. Louis, 1953-57............
Louisville, 1949-57...........
Columbus, Ohio, 1949-56.
Toledo, 1947-56................
Philadelphia, 1940-48....
Philadelphia, 1950-55___
Dayton, Ohio, 1951-56...
valuation
+5.6
+27.8
+53.6
+65.3
+8.6
+27.0
+48.7
Income tax
receipts
adjusted to
constant
rate
+11.7
+81.2
+40.8
+62.6
+91.4
+33.7
+44.2
Property tax
rates
+4.4
+21.5
-25.1
-58.7
<9
0).+17.7
i No change.
These figures also do suggest that though the immediate response of
the assessed valuation in property tax rates may not be adequate, after
a long while there are some adjustments. This fits in with the figures
that I indicated earlier for the general property tax yield for 1938-48
and 1948-57. In Los Angeles, it should be noted, the property valua
tions rose more than sales tax receipts. The explanation of this is, of
course, that this is a very rapidly growing city with large additional
construction, this in turn contributing toward a very large increase in
the yield of the general property tax.22
What is the response of the State sales tax versus the individual
State income tax ? 23
In dealing with this problem we use the States and dates indicated
in the footnote. In some instances, we had to eliminate observations
because the statistics seemed to be way out of line and might be ex
plained by changes in law. The number of paired observations were
106; the number of observations in which the change in individual in
come tax receipts exceeded that of sales tax was 73. We can be 99.7
percent certain that the individual income tax responds more to income
rises than the general sales tax. This also, of course, might be ex
pected. Somewhat similar results occur when we have years of de
clining income.
Dealing with various years for Alabama, Arizona, Colorado, Iowa,
Louisiana, and Missouri, we also find that we can be 99.4 percent con
fident that the sales tax responds more to income increases than
assessed valuations. These are presumably State taxes.
We also find, as might be expected, that the sales tax since 1948 did
not respond more to income increases than the property tax. The
reasons may be inferred from what has been said above. In discuss
ing statistics of individual income taxes and corporate income taxes
for Arizona, Arkansas, California, Colorado, Georgia, Iowa, Ken
tucky, North Carolina, North Dakota, Oklahoma, South Carolina,
and Utah, in various years since 1943, we find that the individual
personal income tax responds more to rising income than does the cor
82 For many of these statistics, we depend upon the State Tax Reporter and the Tax
Systems of the Commerce Clearing House.
28 We use the following States and periods: Arizona, 1949-54, 1955-58 ; Arkansas, 19434 9 ,1 9 5 0 -5 6 ; California, 1950-58 ; Colorado, 194 3 -4 9,1 9 51 -5 8 ; Georgia, 1952-54, 1955-58 ;
Iowa, 1950-55, 1 95 6 -5 7 ; Kansas, 1948-58 ; North Carolina, 1943-58 ; North Dakota, 194853. 1954-58 ; Oklahoma, 1949-58 ; South Carolina, 1952-58 ; Utah, 1943-58.
36
THE INCIDENCE OF INFLATION
porate income tax. The number of paired observations was 94. We
can be 87.2 percent confident that the individual income tax is more
responsive to income changes than the corporate income tax. But
this is not too significant an observation because the corporate income
tax yield is likely to respond to other factors than the personal income
tax, whereas personal income taxes are more likely to respond to per
sonal income. Of course, the State individual income tax responds
better to rising prices and incomes than the property tax. For Ala
bama, Arizona, Colorado, Iowa, and Louisiana for various years since
1942 we have 34 paired observations, and on the basis of this we can
be 98.8 percent confident that the individual income tax responds more
to income changes than the property tax base.
INFLATION AND TH E YIELD OF TH E INCOM E TAX
One of the interesting aspects of the inflationary process under our
tax situation is that as inflation proceeds there is a tendency for the
yield of the income tax to rise because of the inflationary effects. Income-tax payers tend to be shifted to higher brackets of tax liability
where rates are also higher.
*
* * This inflation shifted taxpayers into higher brackets o f income tax
liability and increased the effective tax rates to which they were subject, quite
apart from any changes in statutory tax rates. Furthermore, real incomes in
creased during this period. This factor reinforcing the effect of rising prices
tended further to shift taxpayers into higher brackets of tax liability.24
Consider the following table:
T a b le 4 -5 .— Percentages of family personal income taken by the Federal indi
vidual income tax at successive levels of income, all consumer units, 1941 and
1950
Tax rate (percent)
1941
Family personal income level (before income taxes)
1950
Under $1,000________ ______________________ ____ ______________
.............................................
$1,000 to $1,999
.................... ....................
$2,000 to $2,999.................................................................. ...................
$3,000 to $3,999............................................ ..........................................
$4,000 to $4,999............................................................. ..........................
$5,000 and over________ ________________________________________
(1) 1.9n
3.5
4.6
5.2
11.6
All incomes combined______ ______________________________
8.4
On incomes On incomes
in 1941
in 1950
dollars
dollars
(9
0.7
1.4
2.2
3.2
14.6
0.1
.7
1.1
1.6
8.8
4.6
4.6
i Less than 0.05.
Source: Review of Economics and Statistics, February 1954, p. 21.
It will be noted that the tax rate on similar incomes increases much
more in 1950 vis-a-vis 1941 incomes when stated in terms of 1950
dollars than in 1941 dollars. Should we compare the rates in the
last column, that is, on incomes in 1941 expressed in 1950 dollars, and
the rates in the first column, that is, the tax rate in current dollars
in 1950, we would then compare the effective rates on consumer units
24 Goldsmith, Jaszi, Kaitz, Liebenberg, “ Size Distribution of Income Since the Midthirties,”
Review of Economics and Statistics, February 1954, p. 22.
37
TH E INCIDENCE OF INFLATION
in identical brackets of real income in the years 1941 and 1950. It
will be noted that, when one compares the last column of this table
and the first column, the increase in rates is much greater than
if the comparison is made with the 1941 incomes expressed in 1941
dollars. In fact, at $5,000 and over, there is an actual increase from
8.8 percent on 1941 incomes expressed in 1950 dollars to 11.6 percent
for 1950 incomes in 1950 dollars. But on incomes in 1941 dollars there
is a reduction for incomes of $5,000 and over from 14.6 to 11.6 percent.
The explanation here is that with the continued need for revenues in
the war period there was a tendency to make the tax rates less pro
gressive. Much more revenue had to be obtained from the lower
incomes in the groups above $5,000. Another factor was, of course, the
split income provision which tended to cut down the rates. The split
ting provision was introduced in 1950.
One can get an even greater indication of what happened as a result
of both the rise of prices and the rise of real income by comparing
the taxes by quintiles in 1941 and 1950. Here one will find even greater
differences in tax rates in 1950 as compared to 1941, because here we
take into account not only the rise of prices but also the rise of real
incomes which result in shifting to higher income brackets.
Table 4-6.—Percentage of family personal income taken by the Federal indi
vidual income tax, for quintiles of consumer units ranked by size of family
personal income, 1941 and 1950
1950
1941
Quintile
Income range
Lowest_________________________________
2d______ _______________________________
3d.......................... ........................................
4th_______ _____________________________
Highest________________________________
Tax rate
(percent)
Under $740..........
$740 to $1.370_____
$1,370 to $2,040___
$2,040 to $2,940___
$2,940 and over___
All incomes combined_____________
0.3
.8
1.4
8.5
4.6
Income range
Under $1,840.........
$1,840 to $3,040___
$3,040 to $4,200___
$4,200 to $5,960....
$5,960 and over___
Tax rate
(percent)
1.4
3.5
4.7
5.8
12.9
8.4
Source: Review of Economics and Statistics, February 1954, p. 23.
SOME CONCLUSIONS
In general, those levels of government which depend upon the direct
income tax or the direct corporation tax have had the greatest suc
cess in increasing their revenues. This of course is the position of
the Federal Government. State governments, relying heavily on sales
and gross receipt taxes and the like, have increased their revenues
substantially though their proportion in relation to the Federal Gov
ernment has greatly declined. Local government, largely dependent
upon general property taxes, experienced the largest relative losses
of revenue and the most serious financial problems.
Income tax yield has grown especially because of both the rise of
real incomes and also because of the inflationary process. Inflation
accounted for about one-third of the rise of income since 1938 and
more than one-half since 1952.
Through grants and aids of various kinds the Federal Government
was able to pass along to other levels of government the flexibility
of its tax revenue system. Unfortunately, these allocations have
48084—59------4
38
TH E INCIDENCE OF INFLATION
not always been consistent with the needs of local and State govern
ment. For example, the largest increases have been in such matters
as welfare programs and highways, though perhaps the greatest needs
have been in education. To the last, the Federal Government makes
a very small contribution. Moreover, in making these allocations,
the result is a distortion in the spending programs of States and local
governments because of the temptation to spend more where the Fed
eral Government makes matching funds available.
It might, therefore, be desirable for the Federal Government to
raise more money through direct taxes and allocate some part to
other levels of government. Indeed, there are some disadvantages to
this method of finance, and in particular the one that State and local
governments may not be able to anticipate too accurately what their
revenues will be as a result of the yield and division of the taxes col
lected by the Federal taxes. But this particular approach has the
great advantage of eliminating the problem of interstate competition
which keeps State and local revenues at a relatively low level and also
results in starving certain essential services.
In general, it is clear that the trends in tax structure and the in
creasing participation of the Federal Government facilitate the col
lection of revenues as inflation and rising incomes continue. The
possibilities of exploiting rising income, whether due to inflation or
rising productivity and increases in the labor market, have steadily
increased since early in the century.
C
hapter
5. W
elfare
P
rogram s
and
I
n f l a t io n
INTRODUCTORY
Inflation is especially costly to those who have low incomes and
whose incomes do not respond to rising prices. For this reason we
should have a careful look at all welfare programs, and particularly
the insurance programs, for after all, insurance is an attempt to pro
vide security through prior payments. In 1934r-35 total social wel
fare expenditure in the United States, according to the Social Se
curity Board, was $7,872 million. By 1945-46 the total had risen
to close to $12 billion and by 1956-57 to almost $38 billion. These
welfare expenditures include social insurance, public aid, health and
medical services, other welfare services, education and veteran pro
grams. As a percentage of GNP these expenditures dropped from
11.5 percent in 1934-35 to 5.8 percent in 1945-46 and rose to 8.8 per
cent by 1956-57. The contribution of social insurance steadily in
creased.
As might be expected, public aid steadily declined, and largely be
cause of the improvement of the economic situation, and in turn the
burden on insurance increased. The proportion of GNP provided
for health and medical services remained fairly stable, though there
was a tendency to decline during the war and to rise after the
war.
As might be expected, during the war the percentage of expenditures
on education to GNP tended to decline, the total being 3.2 percent
in 1934-35, 1.6 percent in 1944-45, and rose to 3 percent by 1956-57.
Veterans’ programs, as a percent of GNP, also tended to fall in the
39
TH E INCIDENCE OF INFLATION
depression and early war period and then to rise dramatically after
the war and, after the veterans’ educational and other adjustment pro
grams were largely financed, to decline again.
The table below gives some indication of the trends of welfare ex
penditures.
T a b le
5-1.— Social welfare expenditures in the United States under civilian
public programs, fiscal years 19S4-S5,1945-46,1956-57
[In billion dollars]
1934-35
Total, all governments..____________ _____________________
Insurance____________________________________________
Federal, all__ ___________________________________________
Federal insurance____________________________________
State and local government, all____________________________
Insurance____________________________________________
1945-46
7.87
.38
2.97
.10
4.90
.28
11.80
2.58
4.53
.86
7.27
1.72
1956-57
37.90
12.46
17.33
8.95
20.57
3.51
Source: Adapted from Social Security Bulletin, October 1955 and October 1958.
The following few facts should be noted from this table. Insur
ance occounted for less than one-half of 1 percent in 1934-35 and rose
to 22 percent in 1945-46 and roughly one-third by 1956-57. It is clear
that the Government was putting a much greater burden on insurance,
and this trend is likely to continue.
Of all welfare expenditures, the Federal Government’s share was
slightly higher in 1956-57 than in 1934-35. But if allowance is made
for insurance payments, then the Federal programs clearly tend to
become less important. Whereas the Federal Government’s contribu
tion to the noninsurance elements was roughly 40 percent in 1934-35
and in 1945-46, by 1956-57 its share had been cut to 33 percent.
This tendency to rely more on insurance is based on the general
theory that the public should pay its own insurance bill. Insofar as
insurance takes care of welfare expenditures, so far the Federal
Government would not have to depend upon general tax revenues. A
major element in the insurance programs is, of course, the various oldage insurance programs. In 1956-57 these accounted for about threefourths of the total outlays under social insurance.
Perhaps the most important programs from the viewpoint of our
problem relate to the economic status of the old. In December 1958,
there were 15.4 million people in the country aged 65 and over, or
roughly 9 percent of the population, and their numbers, of course,
were increasing, both absolutely and relatively. At this time 3.7 mil
lion of the 15.4 million were employed, and 10.8 million were receiving
benefits under social insurance and related programs. This includes
1.24 million under veterans’ compensation. In addition, 2.5 million
were receiving public assistance, and 1.6 million received no money
income or income solely from other sources. The reliance on the oldage insurance programs was large and increasing. The total number
receiving benefits under old-age insurance programs exceeded 15 mil
lion. Of course, some received help from more than one source of
income.
The median income of families with head aged 65 or over in 1956 was
$2,550; with head employed, $2,066; with head not employed, $3,675.
The average income was roughly about one-half of the average family
40
TH E
INCIDENCE OF INFLATION
income in the Nation and suggests that on the whole, despite somewhat
lower requirements, the old had a considerably lower economic status
than the rest of the population.
I shall say something presently about the trends in old-age survivors
benefits and assistance, but an examination of the table below will give
some indication of the trends in recent years under various benefit
programs.
T a b l e 5 - 2 . —Average monthly payments to retired worker beneficiaries under 3
social insurance programs to veterans of World War I receiving pensions and
to old-age assistance recipients, June 1948 and June 1958
Program
Average momthly payment
June 1948
Retired workers under old-age, survivors, and disability insurance.................
Railroad retirement.___________________________________________________
Federal civil service retirement_________________________________________
Veterans of World War I receiving pension................... ..................................
Old-age assistance recipients____________________________________________
June 1958
$25.13
70.13
89.25
62.53
38.18
$65.66
114.92
144.53
76.14
61.92
Source: Social Security Bulletin, June 1959, p. 8.
ADJUSTMENTS OF BENEFITS TO RISING PRICES AND INCOMES
The rise in average monthly payments varied greatly from pro
gram to program. For old-age, survivors, and disability insurance,
the gain was 161 percent; for railroad retirement, 64 percent; for
Federal civil service retirement, 62 percent; for veterans of World
War I, 22 percent; and for old-age assistance, 62 percent. To some
extent the varying rates of gain are explained by the varying levels
from which the base period is recorded. Since consumer prices rose
by 20 percent during this period, it may be assumed, that each group
was compensated at least for the rise of prices during this period.
Of course, the large rise, for example, in old-age, survivors, and dis
ability insurance could be explained by the serious lag in the forties,
about which I shall write later.
During this period, in manufacturing, weekly wages rose by 54 per
cent. Hence the rise of these benefits more than matched the increase
of manufacturing weekly earnings, and on this basis was adequate for
all groups except veterans of World War I.
Let us have a look at a more comprehensive table which gives the
history from 1940, when the olci-age survivors insurance program
really began to operate, to 1958. This table gives the monthly bene
fits from 1940 to 1958, both in current and in 1958 dollars, for the
retired worker, for the aged widow, for the widowed mother and two
children.
By 1948 it will be noted that, though the average benefit did not
change very much in dollars, when correction is made for the rise of
prices, the large losses suffered by the beneficiaries are quite clear.
These losses are roughly similar for each group as the price level
increased. For example, the retired worker had his monthly benefit in
1958 dollars cut from $46.41 gradually until by December 1948 he was
receiving only $30.43, or a loss of more than one-third in his benefits
in dollars of stable purchasing power.
41
THE INCIDENCE OF INFLATION
Another approach is to compare these benefits with trends in per
capita disposable income in 1958 dollars. We find from 1940 to 1948,
a rise in per capita disposable income of 30 percent. This figure, of
course, allows for the income-tax take, and, therefore, the relative loss
of the beneficiaries was even greater than suggested by these figures;
but if we estimate that the value in real terms of these benefits de
clined by one-third and that the per capita disposable income of the
Nation in 1958 dollars rose by 30 percent, then roughly there was a
relative loss for the retired worker of 50 percent; that is, 65 is onehalf of 130. Should we allow for the tax take, then the loss would be
substantially more than 50 percent.
Obviously, the adjustments to rising prices and rising incomes were
slow indeed. Possibly in the future the adjustments will be more
rapid, for we have had amendments to the Social Security Act in 1950,
1952, 1954, 1956, and. 1958. Important gains made in 1950 are evi
dent when one compares the monthly benefits in the second column for
1948 and 1950, a rise roughly of two-thirds. These gains vary to some
extent in the different categories, but the most important category is
the retired worker because the largest part of the old-age benefits go to
him. With the inflation of 1950-51, the beneficiary lost again in real
dollars in 1951, but these losses were soon made up and there has been
a fairly steady climb up to 1958; in fact, the 1958 amendments make
possible a further gain in current dollars of about 7 percent on the
average.
5-3.— Old-age, survivors, and disability insurance: Average monthly old-age
and widow*s benefits in current-payment status, in current and in September
1 9 5 8 dollars, 1 9 W - 5 8 1
T a b le
Average monthly benefit for—
Aged widow
Retired worker (old age)
Month and year
December1940....................
1941.....................
1942.....................
1943..................
1944.............. .
1945........... .........
1946.....................
1947.....................
1948.....................
1949..............
1950.....................
1951_____ ______
1952.....................
1953___________
1954.............. .
1955.............
1956................... .
1957..................
June 1958_____ ____
September 1958_____
Widowed mother
and 2 children
In current-payment History of benefit In current-payment In current-payment
status at end of
status at end of
status at end of
that was average in
December 1940
period
period
period
Current
dollars
1958
dollars
Current
dollars
1958
dollars
Current
dollars
1958
dollars
Current
dollars
$22.60
22.70
23.02
23.42
23.73
24.19
24.55
24.90
25.35
26.00
43.86
42.14
49.25
51.10
59.14
61.90
63.09
64.58
65.71
66.17
$46.41
42.51
39.49
38.90
38.59
38.46
33.04
30.74
30.43
31.86
50.76
46.10
53.42
55.01
64.00
66.77
66.13
65.70
65.71
66.17
$22.60
22.60
22.60
22.60
22.60
22.60
22.60
22.60
22.60
22.60
41.40
41.40
46.60
46.60
51.60
51.60
51.60
51.60
51.60
51.60
$46.41
42.32
38.76
37.54
36.75
35.93
30.42
27.90
27.13
27.70
47.92
45.30
50.54
50.16
55.84
55.66
54.09
52.49
51.60
51.60
$20.28
20.22
20.15
20.15
20.17
20.19
20.22
20.40
20.60
20.82
36.54
36.04
40.66
40.87
46.27
48.69
50.14
51.09
51.56
$41.64
37.87
34.56
33.47
32.80
32.10
27.21
25.19
24.73
25.51
42.29
39.43
44.10
43.99
50.08
52.52
52.56
51.97
51.56
$47.10
46.60
46.50
46.60
47.30
47.70
48.20
48.80
49.80
50.40
93.90
93.80
106.90
111.00
130.50
135.40
141.00
146.30
148.70
1958
dollars
$96.71
87.27
79.76
77.41
76.91
75.83
64.87
60.25
59.78
61.76
108.68
102.63
115.94
119.48
141.23
146.06
147.80
148.83
148.70
i Calculated by dividing current-dollar amounts by the Consumer Price Index on a September 1958 base.
Source: U.S. Department of Health, Education, and Welfare, Social Security Administration, Research
and Statistics Note No. 12, Nov. 26, 1958,
42
TH E INCIDENCE OF INFLATION
Since minimum benefits have been increased relatively more than
the earnings base, beneficiaries whose benefits were based on low
covered earnings tended to gain more; for example, between Decem
ber 1940 and December 1958 the buying power of minimum primary
insurance benefit, which rose from $10 under the 1939 act to $33 under
the 1958 act, rose by 60 percent, while that of the maximum old-age
benefit payable went up only 3 percent. In relation to the 1935 act,
there was a substantial loss in purchasing power of primary benefits
under the 1958 act. The current $254 maximum25 per family repre
sents an increase of 15 percent in buying power, compared to the $85
maximum family benefits under the 1939 act. In January 1959 the
$254 maximum at September 1958 prices represents a 42-percent in
crease over the family maximum under the 1939 law. These trends
reflect a tendency to tie benefits increasingly to needs—for example,
the greater rise in minimum benefits and family benefits as against
the primary maximum.
By June 1959 the beneficiaries under the old-age, survivors, and dis
ability insurance numbered 13,181,000, of which number those age 62
and over were 10,792,000, the young survivors and dependents, 2,114,000, and disabled workers aged 50 to 64,275,000. Monthly benefits had
reached $805 million, or at the rate of close to $10 billion a year; the
average old-age retired worker benefit had risen to $72.19; and oldage benefits awarded in the month of June 1959 had risen to $80.32.26
That from December 1948 to September 1958 the benefits for the
retired worker had risen in 1958 dollars by 118 percent suggests that
the response to inflation has been more than adequate. It will be
recalled that prices only rose about 20 percent during these years, and
even per capita disposable income in 1958 dollars only rose by 15 percent.
It is, nevertheless, not wise to assume that the response to inflation
is automatic^ despite the great gains from 1948. One reason for the
large gains since 1948 was the general realization that the benefits were
altogether too low in relation to minimum needs. It has been said
many times that the benefit payments even today of roughly $800 to
$900 a year for the retired worker are much below the amount neces
sary for a minimum standard of living. The increase in benefits since
1948, therefore, may reflect in part the past failures to get benefits up
with rising prices, but also a realization that the benefits were alto
gether too low in relation to the needs of older members of our society.
RAISING M A X IM U M COVERED WAGES
One reason for the failure to achieve higher benefits has been the
lag in the adjustment of the amount of wages to be covered for each
worker. When the program was introduced in the 1930’s, the wages
of each worker to be covered were a maximum of $3,000. It was not
until 1958 that an increase to $4,800 was allowed. Under the 1935
act, the maximum earnings taxable and creditable were $3,000; in
1950 this sum was raised to $3,600; in 1954, to $4,200; and in 1958, to
$4,800.
25 Smaller of $254 or 80 percent of average wage (but less than the larger of 1 %
times the primary insurance amount or $20 plus primary insurance amount).
26 See release of U.S. Department of Health, Education, and Welfare, Social Security
Administration, Aug. 11, 1959.
TH E INCIDENCE OF INFLATION
43
The failure to get the maximum earnings taxable and creditable
up more rapidly was, of course, costly to the insurance program in
the sense that less was collected and, therefore, benefits tended to be
kept low. By 1950 covered wages had increased by only 20 percent,
whereas the price level had risen by 75 percent and weekly manufac
turing wages by 195 percent. Even after the amendments of 1954
and 1958 the rise in maximum earnings taxable was only 60 percent as
against an increase in prices of 110 percent, and in weekly manu
facturing wages of 315 percent.27
The failure to get the maximum earnings taxable up more rapidly
resulted in a failure to introduce an element of progressivity into the
program. Under the legislation those with relatively low wages get
larger benefits in relation to their contribution than those with
high wages. Larger coverage would have had the effect of favoring
further the transfer of benefits on behalf of the relatively low wage
groups.
A $4,800 maximum now means that roughly about 56 percent of the
workers’ total earnings that are covered now are subject to tax. Of
course, when the maximum was $3,000 in the 1930’s, this meant vir
tually complete coverage of all wages of those subject to tax. Thus in
1938 the maximum earnings base was $3,000 and only 6 percent of the
covered workers earned in excess of $3,000. In the 1950’s 61 percent of
the covered workers had wages in excess of $3,600 and under the 1954
amendments which raised the wage base to $4,200, 43 percent had
wages in excess of $4,200.28
M IN IM U M AND M A X IM U M BENEFITS AND PRICES
In a table below, I indicate the trends in wages as well as various
aspects of benefits under the old ag^e, survivors and disability insurance
legislation. This table shows quite clearly the failure of minimum
family benefits to rise as much as prices up until 1952 and the failure
of the minimum family benefit to rise as much as the increase of wages.
For example, by 1950, prices had risen by 72 percent above 1939 and yet
the minimum family benefit had risen only by 50 percent in this
same period. The average full-time wages of all employees had risen
by 150 percent. By 1952, however, the minimum family benefit had
risen about as much as prices, and by 1954 substantially more. But
even by 1958 the average full-time wages per year had risen substan
tially more than the percentage rise in minimum family benefit.
For the maximum family benefit it will be noted that one limit is
80 percent of average wage. It follows, therefore, that if the average
wage is, say, $160 a month, then the benefit cannot be more than fourfifths of $160, or $128 a month. The result of this particular provision
on the whole is that the maximum benefit expressed in dollars, say,
$85 under the 1939 act and $254 under the 1958 act, is likely to be more
restrictive for high than for low incomes. Moreover, this restriction
plays a larger part currently than it did many years ago. For ex
27 Social Security Bulletin, January 1959, pp. 18-19 and the “Economic Report of the
President,” January 1959, my calculations.
28 Social Security Amendments, Social Security Bulletin, September 1954, and House
Ways and Means hearings, “ Social Security Amendments, 1954,” p. 60.
44
TH E
INCIDENCE
OF INFLATION
ample, in 1939 the act provided that the maximum family benefit
should be the smaller of $85 or 80 percent of the average wage or two
times the primary insurance amount. Since the average wage was
$105 a month in that year, and, of course, much higher from 1939 until
1950 as wages and prices rose, the $85 maximum became a very restric
tive factor in the estimation of maximum family benefits. For ex
ample, in 1949 the average monthly wage was $238. But the limita
tion of $85 meant that, although the average wage was $238, the
limitation of $85 reduced benefits to roughly about one-third of the
average wage of that period. By 1950 the average wage was $250 and
the maximum benefit was $150. By that time the restrictive effect
of the maximum dollar benefit was less than in the late 1940’s. By
1958, 80 percent of the average monthly wages was $290 and the
maximum expressed in dollars was $254, and therefore, the restrictive
effects of the provision for a maximum dollar amount on higher in
comes was greatly reduced.
T a b l e 5 - 4 . — Annual earnings for a full-time employee, contributions and benefits under old-age, survivors, and disability insurance, 1935-58
1935
Average full-time earnings, $1,137.....................
annual rate.
Minimum family benefit-__
Maximum family benefit
$1,264..................... $3,008............................ $3,414.....................
1954
$3,670.....................
1956
$4,042
.................
$10__..................... $15................................ $18 80
$30......... ................
Smaller of $150 or 80
Smaller of~$168.75" Smaller of $200 or
Smaller of $85, 80
percent of average
percent of aver
or 80 percent of
80 percent of
average wage
age wage, or 2
wage (but not less
average wage
times primary
than $40).
(but not less
(but not less
insurance
than $45).
than the larger
amount.
of $50 or lH
times primary
insurance
amount).
Same except 2
percent rate ex
tended through
1942.
1950-53, 3 percent;
1954-59, 4 percent;
1960-64, 5 percent;
1965-69, 6 percent;
1970 on, 6H per
cent.
Self-employed pay H
of combined employer-employee
rate.
1954-59, 4 percent;
1960-64, 5 per
cent; 1965-69, 6
percent; 197074, 7 percent;
1975 on, 8 per
cent.
1958
$4,344.
$33.
Smaller of $254 or 80
percent of average
wage (but not less
than the larger of 1y2
times primary in
surance amount or
$20 plus primary
insurance amount).
1957-59, 4H per1960-64, $V2 per
cent; 1965-69,
QH percent;
1970-74,1V2
percent; 1975
on, sy2 percent.
(Increase of lA
percent is for
disability
benefits.)
1959, 5 percent; 196062, 6 percent; 196365, 7 percent; 196668, 8 percent; 1969
on, 9 percent. (In
all years, X
A percent
is for disability
benefits.)
IN F L A T IO N
No provision.........
1952
OF
(б) Self-employed.-.........
1937-39, 2 percent;
1940-42, 3 per
cent; 1943-45, 4
percent; 194648, 5 percent;
1949 on, 6 per
cent.
1950
INCIDENCE
Contribution rates:
(а) Combined employeremployee.
1939
Source: Myers, R. J., “ Old-Age, Survivors, and Disability Insurance Provisions: Summary of Legislation, 1935-58,” Social Security Bulletin, January 1959, pp. 18-19.
Oi
46
TH E INCIDENCE OF INFLATION
One will also note from this table that there has been a general
tendency for payroll taxes to rise. This was not, however, true dur
ing the 1930’s and 1940’s. In fact, the anticipated rise of rates was
postponed, and particularly in the 1930’s there was a considerable
fear that an increase in rates would have a depressing effect upon the
economy. It is a little more difficult to explain the failure to increase
rates during the war, at least the failure during the war when infla
tionary forces were at work. By that time a general view had de
veloped that the program should be on a pay-as-you-go basis and,
therefore, it would not be wise to accumulate large reserves. Be
ginning in 1950, however, the tendency to increase rates tended to
become stronger and particularly since 1954, as the table suggests.
By 1969 it is expected that the total rate for combined employeremployee would be 9 percent.
INCREASED BENEFITS THROUGH INCREASED COVERAGE OF F A M IL Y MEMBERS
So far we have discussed largely the benefits available to an in
dividual or the family. It should be noted, however, that there has
been liberalization of the program through making available addi
tional benefits to other members of the family. The original 1935
act provided 100 percent of primary insurance amount for the oldage retired worker. There were no provisions for disability until
the 1956 act and even that applies only to those aged 50 and over.
The 1939 act provided for wives’ or husbands’ benefit of 50 percent
of the primary insurance amount; a child of retired worker, 50 per
cent of the primary insurance amount; a child of deceased worker, 50
percent of primary insurance amount; widows or widowers and widow
mothers, 75 percent of primary insurance amount; parents, 50 per
cent of primary insurance amount. The last was later increased
under the 1950 act to 75 percent of primary insurance amount. The
minimum family and maximum family benefits in relation to price
and income rises are still relevant. The introduction of these addi
tional benefits to other members of the family made it possible to
provide a larger amount of benefits within these minimums and
maximums.
Under the pressure of increased benefits the costs of the program
tend to increase. For example, the level premium equivalent benefits
in 1950 were 6.05 percent of taxable payrolls and by the 1958 act they
were estimated at 8.76. Contributions in these 2 years were estimated
at 5.95 and 8.52. In recent years the Government has urged the
Congress and the Congress in turn has been anxious to balance the
accounts so that the program would be actuarially sound. This has
meant, of course, a marked rise in payroll taxes to match the increase
in benefits.
It may, therefore, be said that the old-age, survivors, and disability insurance
program is actuarially sound if the estimates show that future income from
contributions and from interest earnings on the accumlated trust funds will,
in the long run, support the disbursements for benefits and administrative
expenses * * *.29
29 Myers, R. J., “ Old-Age, Survivors, and Disability Insurance Financing Basis and
Policy Under the 1958 Amendments,” Social Security Bulletin, October 1958, p. 15,
47
TH E INCIDENCE OF INFLATION
The costs are primarily those that go to the insured, for (old age)
primary benefits accounts for 5.92 of the 8.40 percent of payrolls of
total benefits, 0.57 are for wives’ benefits, 1.23 for widow’s benefits,
and the only other major item is child’s benefits, 0.43, and disability
insurance, 0.49. These are all estimated level premium costs for
benefit payments. (The level premium cost is the average long-range
cost based on discounting at interest in relation to payroll.)
One of the most difficult problems in projecting the financing of the
old-age and survivors insurance arises from the difficulties of esti
mating what wages will be in the future. Of course, the further
ahead we go the more difficult it is also to estimate the number of
workers that will be involved because many of them are not as yet
born. Here, for example, is an actuarial estimate of the progress of
old-age and survivors insurance trust fund under the 1958 act on
high employment assumptions based on intermediate cost estimates
at 3 percent interest.
[In millions of dollars]
Contribu
tions
1951...............................................
1957...............................................
1960 (estimated)_______________
1975 (estimated)_______________
2020 (estimated)_______________
3,367
6,826
10,621
20,880
36,124
Benefit
payments
1,885
7,347
10,027
17,766
40,716
Administra
tive expenses
Interest on
fund
81
417
162
557
590
222 ............ 2,185"
426
8,379
Balance in
fund
15,540
22,393
21,794
76,432
285,282
Source: Myers, R. J., “ Old Age, Suivivors, and Disability Insurance Financing Basis and Policy Under
the 1958 Amendments,” Social Security Bulletin, October 1958, p. 17.
It will be noted that as against benefit payments of about $10 billion
today, the total would rise to about $41 billion in the year 2020, and
the balance in the fund would rise from around $20 billion to $285
billion. But even these estimates are of the roughest kind; first,
because they assume the rate of interest at 3 percent and it is not
easy to estimate what the rate of interest will be in the future; second,
because no allowance is made for the rise of prices and income.
Wages, for example, in 1975 are estimated as they are in 1958. By
that I mean the average wage. A more likely estimate would be
that wages would be about double even in the absence of all-out war
by 1975. Then if benefits lag behind, either the reserve fund would
be much larger than is estimated or else contributions would be
reduced or benefits increased greatly.
Obviously, the instability in the value of the dollar as well as the
rising productivity of the economy makes it much more difficult to
estimate what the net actuarial result will be. For this reason it
becomes very important to revise the act every few years, or possibly
even every year. Another interesting aspect of this problem is that
as the value of the fund rises and prices and income also increase,
the value of the fund in relation to the size of the economy tends to
be reduced. In other words, $285 billion at prices and incomes of
1958 is one thing, but $285 billion at the prices and average income
levels of the year 2020 is an entirely different matter. There would
be a considerable erosion of the value of these accumulations if prices
and incomes continue to rise as they have in the past. Even if we
assume a doubling of the average wage every 20 years, which is not
48
TH E INCIDENCE OP INFLATION
a reckless assumption, the increase in the average wage by 2020 would
be about eight times. We would be most conservative if we assumed
that included in this rise of eight times was a rise of prices of only
1 percent per year. Hence the $285 billion relative to 1958 wage
levels may be assumed to be only worth about $35 billion in terms of
price and income levels of the year 2020. Hence it may be assumed
that the contribution of the reserve funds to the financing of the
program would be considerably reduced as compared to the estimates
made in the actuarial statements. Of course, this may be offset by
larger rises in contributions than now contemplated. At any rate,
it is clear that the inflationary process is one argument against reserve
financing.
OLD-AGE ASSISTANCE AND OTHER ASSISTANCE PROGRAMS
Strangely enough, the response of old-age assistance and other
assistance benefits to the price and income level since 1935 has been
much greater than under old-age and survivors insurance. Since
insurance is a contributory program and the relief is a donated pro
gram, this is rather unexpected. One explanation of this fact is that
the payments under the old-age insurance program since 1939—and
the earlier the years, the more this is true—are very large compared
to the earnings to any individual’s account. In other words, and
particularly since 1939, the Government has tended to subsidize the
relatively old who receive benefits after payments for relatively brief
periods. Hence there are large gifts, so to speak, in the payments
for insurance for the current old who then qualify with relatively
small payments of their own or of their employers.
Another reason for the relatively large increase in benefits under
old-age assistance as well as under assistance to dependent children
is the fact that the Federal Government contributes a substantial
part and an increasing part of the total payments. This is an incen
tive for State and local governments to increase their contributions
since the Federal Government pays a substantial part. The trends
are given in table 5-5. It will be noted, for example, that from 1940
to 1948 the real benefits, that is in 1958 dollars, rose in excess of 20
percent for old-age assistance and for dependent children per family.
In fact, per family, the real rise for dependent children was of the
order of 30 percent. In this same period, it will be recalled, old-age
and survivors insurance benefit payments for the retired worker had
declined by one-third in dollars of 1958 purchasing power. By 1958,
however, the old-age retired worker in 1958 in dollars had achieved
a benefit of 43 percent above the 1940 level as compared to roughly
a rise of 50 percent for old-age assistance and for dependent children
per family.
T
able
5 - 5 . — Pu b lic
assistance benefits: Average monthly payments in current and 1958 dollars,1 19 40-5 8
Old-age assistance
Aid to dependent children
Aid to the blind
Aid to the per
manently and
totally disabled
General assistance
Per recipient
Per recipient
Per case
Month and year
Current
dollars
1958
dollars
Current
dollars
1958
dollars
Current
dollars
$20.26
21.27
23.37
26.66
28.43
30.88
35.31
37.42
42.02
44.76
43.95
46.00
50.90
51.50
51.90
53.93
57 99
60.68
61.79
$41.60
39.83
40.09
44.29
46.23
49.09
47.52
46.20
50.44
54.85
50.87
50.33
55.21
55.44
56.17
58.18
60.79
61.73
61.79
$32.38
33.62
36.25
41.57
45.58
52.05
62.23
63.01
71.88
74.19
72.42
77.08
83.83
84.22
$66.49
62.96
62.18
69.05
74.11
82.75
83.76
77.79
86.29
90.92
83.82
84.33
90.92
90.66
93.30
95.59
99.61
102.46
103.26
$9.85
86. 21
88.61
95.03
100. 72
103. 26
1 Calculated by dividing current dollar amounts by the consumer price index on a
September 1958 base.
2 Recipients include children plus 1 adult per family when adults are included in assist
ance group; before December 1950 partly estimated.
Per recipient2
10.21
10.93
12.36
13.41
15.15
18.11
18.39
20.92
21.70
21.13
22.36
23.98
23.77
23.96
24.35
25.79
26.90
27.44
1958
dollars
$20.23
19.12
18.75
20.53
21.80
24.09
24.37
22.70
25.11
26.59
24.46
24.46
26.01
25.59
25.93
26.27
27.03
27.37
27.44
Current
dollars
$25.38
25.82
26.54
27.95
29.31
33. 52
36.67
39.58
43.54
46.11
46.56
49.05
54.91
55.67
56.37
58.08
63.12
66.25
66.97
1958
dollars
$52.11
48.35
45.52
46.43
47.66
53.29
49.35
48.86
52.27
56.51
53.89
53.67
59. 56
59.92
61.01
62.65
66.16
67.40
66.97
Current
dollars
$45.41
49.46
53.50
53.44
54.93
56.18
58.83
60.02
60.85
1958
dollars
$52.56
54.11
58.03
57. 52
59.45
60.60
61.67
61.06
60.85
Current
dollars
$24.28
24.40
25.23
27.76
28.77
32.72
39.47
42.79
47.39
50.47
46.65
47.09
49.82
50.53
57.29
55.03
56.12
59.74
61.43
1958
dollars
$49.86
45.69
43.28
46.11
46.78
52.02
53.12
52.83
56.89
61.85
53.99
51.52
54.03
54.39
62.02
59.36
58.83
60.77
61.43
Source: U.S. Department of Health, Education, and Welfare, Social Security Admin*
istration, Research and Statistics Note No. 42, Nov. 26,1958.
IN C ID E N C E
December194 0
194 1
194 2
194 3
194 4
194 5
194.............. 6
194 7
194 8
194 9
195 0
195 1
1952...........
195 3
195 4
195 5
195 6
195 7
.
September 1958.
Per family
THE
Per recipient
50
TH E INCIDENCE OF INFLATION
The lag in payments for old-age insurance in relation to the move
ment of prices and wages was all the more unfortunate, in that grad
ually these payments tend to become much more important than public
aid for the old. For example, in 1939 they amounted only to about 1
percent of total public aid expenditures. By 1946 they were more
than 40 percent; by 1957 twice as large; and on the basis of President
Eisenhower’s estimate for fiscal year 1960, roughly three times as
large.30
When one examines the rise of the maximum monthly amounts
subject to Federal participation, one is surprised that there has been
as large an increase in dollar assistance payments as well as in real
dollars as actually occurred. Under the 1935 originl act Federal par
ticipation was at a maximum of $30 per monthly payment. By 1958
it was $65, or a rise of a little more than 100 percent. But the increase
in monthly benefits for old-age assistance was not of the order of 100
percent but in excess of 200 percent from 1940 to 1958. Actually the
increase of the maximum monthly amount subject to Federal participa
tion from 1939 to 1958 was only from $40 to $65, or an increase of a
little over 60 percent, but the actual increase in the assistance pay
ment per month was in excess of 200 percent. One explanation of this
fact is that the Federal Government’s participation tended to increase.
Federal participation for the old, therefore, increased from a maxi
mum of $15 in 1935 to a maximum of 50-65 percent (depending upon
economic strength of different States) of $65, or to from $32.50 to
$35.75. Even this increase in the Federal participation at a maximum
level does not wholly explain the large rise in old-age assistance during
these years. The increase greatly exceeded the rise of prices and was
not far from matching the rise of average wage. This is explained
partly by the fact that as the maximum covered by the Federal Govern
ment increased there was a tendency for States to increase their assist
ance in order to achieve the largest contribution of the Federal Gov
ernment.
30 Social Security Bulletin, October 1955 and October 1958 and release of August 11,
1 959; “ Current Social Security Program Operations,” June 1959; the President’s
budget, 1960.
TH E INCIDENCE OP INFLATION
51
In his budget for 1960, the President reverted to his doubts on the
trends in the social security program. He insisted that—
The Federal Government’s responsibility for income maintenance should be
mainly discharged through contributory, self-supporting social security.
He boasted of the fact that in 1946, 60 percent of the workers under
OASI were covered; in 1960, 90 percent; and that total annual benefits
had risen from $321 million to $10,510 million. He also revealed that
the average monthly number of beneficiaries had risen from 1.3 to 13.7
million. Actually, this works out as an average payment to a beneficiencary of $767 a year, a figure that is from one-half to one-third
the amount required for a minimum standard of living.
The President was also disturbed by the increased proportion of the
public assistance grants that were financed by the Federal Government.
In 1946, of total outlays of $446 million, the Federal Government’s
share was 44 percent; by 1960 the Federal Government’s share would
be 57 percent of $2,018 million. Legislation to raise the Federal maxi
mum share extending the Federal participation to new groups, he com
plained, had been enacted five times in the last six Congresses.31
*
* * i believe that this trend is inconsistent with the American system of
government. I f it continues the control of these programs wiU shift from our
State and local governments to the Federal Government. We must keep the
financing control of these programs as close as we possibly can to the people
who pay the necessary taxes and see them in daily operation.82
si Budget, 1960, pp. M 69-M70.
Ibid., p. M70.
T able
5-6.— Legislative chronology of provision for Federal participation in assistance
ASSISTANCE SUBJECT TO FEDERAL PARTICIPATION DEFINED TO INCLUDE ONLY M ONEY PAYMENTS TO RECIPIENTS
Assistance payments
Maximum monthly amounts subject to Federal par
ticipation (maximum applied to each individual
payment except where noted)
Aid to dependent children
Each
addi
tional
child
% of first $9 (average per child)
plus A balance.
% of first $12 (average per child)
plus A balance.
Aged, blind, and disabled
$30
$18
....... ........... ..............
$12
40
$18
......................... ...... ..............
12
A .......................................... ......... A -
45
$24______________ _____ _______
15
50
$27....... ................................ - ........-
18
% of first $15 (average) plus A.
balance.
% of first $20 (average) plus A
balance.
A ..................................... ................ M-
ASSISTANCE SUBJECT TO FEDERAL PARTICIPATION DEFINED TO INCLUDE MONE Y PAYMENTS TO RECIPIENTS AND PAYMENTS TO VENDORS
FOR MEDICAL AND REMEDIAL CARE
1950 amendments, effective Oct. 1, 1950, to Sept.
30,1952.
1952 amendments (temporary), effective Oct. 1,
1952, to Sept. 30, 1954.
1954 amendments (extended 1952 amendments),
effective Oct. 1, 1954, to Sept. 30,1956.
1956 amendments, effective Oct. 1, 1956, to June
30,1957.
$27 plus $27 for 1 needy relative
with whom child lives.
$30 plus $30 for 1 needy relative
with whom child lives.
$18
55
60
$32 plus $32 for 1 needy relative
with whom child lives.
23
$50
21
% of first $20 (average) plus A.
balance.
% of first $25 (average) plus A
balance.
% of first $12 (average per person)
plus A balance.
H of first $15 (average per person)
plus A balance.
% of first $30 (average) plus A
balance.
UA? of first $17 (average per person)
plus A balance.
INFLATION
Aid to dependent children
OF
First child
INCIDENCE
Aged,
blind,
and dis
abled
1935 original act, effective Feb. 1,1936, to Dec. 31,
1939.
1939 amendments, effective Jan. 1, 1940, to Sept.
30 1916.
1946 amendments, effective Oct. 1,1946, to Sept.
30, 1948.
1948 amendments, effective Oct. 1, 1948, to Sept.
30,1950.
THE
Legislation
Federal share of expenditures within specified maximums
1957 AMENDMENTS—STATES HAVE CHOICE OF 1 OF 2 METHODS FOR COMPUTING FEDERAL SHARE
M ethod 1. A ssistance Subject to F ederal Participation D efined T o Include M oney Payments to R ecipients and Payments to V endors for Medical and Remedial Care
Effective, July 1, 1957-Sept. 30, 1958________
$60
$32 plus $32 for 1 needy relative
with whom child lives.
$23
% of 1st $30 (average) plus H bal
ance.
7 of 1st $17 (average per person)
plus Yi balance.
or
M ethod 2. Separate Formulas for F ederal Participation in M oney Payments to R ecipients and Payments to V endors for M edical and Remedial Care
THE
a . separate provision for federal participation in money payments to recipients
Effective, July 1, 1957-Sept. 30, 1958............. ........
$60
$23
% of 1st $30 (average) plus
ance.
bal
1f/n of 1st $17 (averag:- per person)
plus H balance.
plus
B. separate provision for federal participation in vendor payment for medical and remedial care
1$6
I/o
} 1.7
IN F L A T IO N
1 Maximum applied to average of all payments.
1$6 per adult; $3 per child..............|........ ........
OP
Effective, July 1, 1957-Sept. 30, 1958........... .........
INCIDENCE
$32 plus $32 for 1 needy relative
with whom child lives.
ASSISTANCE SUBJECT TO FEDERAL PARTICIPATION DEFINED TO INCLUDE M ONEY PAYMENTS TO RECIPIENTS AND PAYMENTS TO VENDORS
FOR MEDICAL AND REMEDIAL CARE
Assistance payments
Maximum monthly amounts subject to Federal
participation
Federal share of expenditures within specified maximum amounts
Legislation
For first part of payments
$65 multiplied by the num $30 multiplied by the
number of recipients.
ber of recipients.
For balance of payments, all programs
Aged, blind, and
disabled
Aid to dependent
children
j i of first $30 (average) _ * # 7 of first $17 (aver
age).
INFLATION
Source: U.S. Department of Health, Education and Welfare, Social Security Administration, Bureau of Public Assistance, October 1958.
OF
Federal percent varies according to average
p*r capita income in.State for most recent
3 years except that Federal share in any
State shall not be less than 50 percent
nor more than 65 percent and Federal
share for Alaska and Hawaii is specified
as 60 percent.
Federal share is determined as follows:
Federal percent equals 100 percent minus
State percent. State percent bears same
relationship to 50 percent as square of
relationship between State per capita
income and national per capita income.
INCIDENCE
1958 amendments, effective
Oct. 1, 1958.
Aid to dependent children
THE
Aged, blind, and disabled
THE INCIDENCE OF INFLATION
55
SOME COMMENTS ON OLD-AGE INSURANCE AND ASSISTANCE
Despite the fact that the largest part of the increase of GNP from
1940 to 1950 is explained not by inflation but by the rise of productivity
and increased numbers on the labor market, the record of benefits
under old-age insurance is not a good one. Until 1950 the benefits
lagged far behind the rise of prices, even more so behind that of in
come. But the old-age assistance program responded much more
effectively to both the rise of prices and the rise of income. This fact
is explained in part by the peculiar financing mechanism. With the
Federal Government agreeing to pay some part of the total and in
creasing its contribution under each change in legislation, the States
and local governments tended to increase their benefit levels in order
to obtain the largest possible contributions from the Federal con
tribution.
Since the Federal income rises much more with an increase in prices
and income per capita than State and local revenues, there is a strong
case for the Federal Government continuing to contribute a large
part of the total assistance benefits. It was the increased willingness
of Congress to appropriate Federal funds that made possible the
rather good record for old-age assistance from 1940 on. The amounts
available even today are not adequate but they are much more nearly
adequate as compared to earlier years than could have been achieved
without the help of the Federal Government. In view of the serious
problems of State and local finance with their revenues rising by
300 percent and debt by 300 percent since the end of the war, it is very
important that the Federal Government continue to contribute gen
erously to the old-age assistance program if benefits are to respond to
rising average wages and to rising prices.
I f the resources are not available—and this is the more likely to
be true the larger the burden put upon State and local government—
then all that we can hope for is that assistance would continue to
match the rise in prices, and the relative economic status of those
receiving assistance would tend to decline relative to the improving
standards of the rest of the Nation. On the whole, this is not a de
sirable outcome.
We can also say much against the financing of old-age insurance
and the actuarial arrangements made for this program. In the 1930’s
quite rightly the Government tended to postpone the increase in pay
roll taxes on the theory that the accumulation of reserves was a depress
ing factor upon an economy which was already being deflated. But
this does not excuse the failure to increase benefits though in 1939 a
change in policy was noted. In the 1939 amendments the policy was
acknowledged that it was proper to pay larger benefits to the present
old despite their small contribution, on the theory that they had not
had sufficient time to accumulate credits. Even the 1939 amendments
were most inadequate and, therefore, in real dollars the benefits tended
to decline, reaching a minimum of one-third below the 1940 level by
1948. The great fear of putting a fiscal burden on the economy in
later years accounted in part for this determination to keep benefits
rather low in the 1940’s. This was a great mistake for another reason,
for this was an inflationary period and there was no excuse for post
poning the increase in payroll taxes in the 1940’s and keeping these
56
TH E INCIDENCE OF INFLATION
taxes at a rate of about 2 percent. The Government should have in
creased the payroll taxes during these years as the excess of purchasing
power contributed to the inflation. Also the Government was ex
cessively worried about the burdens on the economy later resulting
from increased benefits in the 1930’s and 1940’s. Actually the major
problem is one of providing resources for the old, and these resources
will be paid irrespective of the financial arrangements that are made.
Indeed, some financial arrangements facilitate this transfer to the old
who are not working, and others make it much more difficult.
In the 1950’s both taxes and benefits rose greatly and larger reserves
began to accumulate. Even in this period the benefits might have ex
panded somewhat more than they did, and there should not have been
so much emphasis on the need of large reserves. As we have noted,
an accumulation of a large reserve with the likelihood of continued
rises of prices and average wages means that the reserves would yield
much less in resources for the old than might have been expected when
accumulated. It would be much better to spend a large part of these
reserves on benefits or use them as a means of reducing old-age pay
roll taxes. Insofar as under the pressure of rising prices and incomes
the current old have received much more than they have paid in, we
might argue that there is a case here for the Federal Government
financing part of this program through general revenues. In this
manner also it would be easier to offset the rise of prices and the rise
of average incomes.
From the experience of the last 25 years or so, we can certainly con
tend that a frequent adjustment of old-age insurance is necessary in
view of the annual changes in prices and average incomes. These ar
rangements should not only provide for adjustments in benefits and
taxes but also in adjustments of general revenue contributions.
C h a p t e r 6. U
nem ploym ent
C o m p e n s a t io n
INTRODUCTORY
The trouble with our unemployment compensation program is not
merely that we have had a substantial amount of inflation. In order
to understand the shortcomings of the unemployment compensation
program we have to discuss the program briefly before we discuss the
relevance of inflation. Had the program developed as originally an
ticipated ; namely, that contribution rates would be close to 3 percent
of payrolls, and had merit rating not made such inroads by greatly
reducing the payroll tax, inflation would not have had such serious
effects.
One respect in which unemployment compensation proved inade
quate was the setting of a maximum benefit amount per week. On
December 31, 1937, that maximum benefit in most States was $15 per
week, with 95 percent of the covered workers subject to this maximum.
By 1952 more than 50 percent had maximum weekly benefit amounts
of $27 or less. The average maximum benefit had risen somewhat less
than the price level and much less than the increase in average weekly
wages. In 1939 the maximum benefit was in excess of 60 percent of
the average weekly covered wage in 31 States and less than 50 percent
in 2 States; but by 1952 the total was in excess of 60 percent in 2
States and less than 50 percent in 40 States. The average weekly un
TH E INCIDENCE OF INFLATION
57
employment payment for full-time work was 41 percent of tlie average
weekly wage of covered workers in 1939 and only 34 percent by 1952.33
The Bureau of Employment Security of the Department of Labor
which administers the unemployment insurance program is not enthu
siastic about tying benefits to the cost of living, though it does stress
the importance of a proper relationship between benefits and wages.
It is, however, vague on the latter point.
The preceding discussion has emphasized—and properly so in a
wage-loss system—the importance of maintaining proper relation
ships between benefits and wages. It is important that changes in
wages be recorded so that necessary changes can be made in the pro
visions for weekly benefits. It has been urged, also, that changes in
benefits should accord not only with changes in wages but also with
changes in the cost of living. Change in the cost of living here means
change in the Consumer Price Index of the Bureau of Labor Statistics.
Such comparisons do not seem completely valid in a short-term insur
ance system. If, by judicious amendment of maximum benefit pro
visions, a proper relationship between benefits and wages can be
maintained, it appears too much to ask that benefits be changed in
accordance with cnanges in prices as well. A worker is eligible for
benefits generally for a maximum of 26 weeks in 52. Benefits are not
payable for consecutive years unless the claimant has had sufficient
employment following his former base period to requalify. This
employment establishes new benefit rights in accordance with more
recent wages—inflated or deflated as the case may be. It may be of
some interest to see what the benefit will buy in 1939 or 1945 dollars,
but this need not be a controlling consideration in a short-term, wageloss system. For some beneficiaries, in fact, wages used as a base for
benefits may already have been changed, up or down, in response to
changes in the capitalized Consumer Price Index.
To rule out the need for variation o f benefits with average changes in con
sumers’ prices does not rule out frequent examination of the proportions of
workers’ incomes that go for “ the nondeferrable bundle.” There is an important
difference between increasing benefits because living costs on the average went
up, and increasing benefits because beneficiaries now spend relatively more of
their income for food and rent and other basic essentials. Fortunately, the
proportion of wages that goes for nondeferrable expenses does not seem to fluc
tuate as much as the Consumer Price Index itself. On the other hand, the figure
is not computed very often, or on a State basis. Much research needs to be done
in this area.84
If the Government succeeds in tying benefits to wages, then, of
course, one need not worry too much about the adjustment of prices
because wage rates generally rise more than prices, given the gains
in productivity. It is of some significance that the maximum benefit
rate would be less restrictive today if inflation had not proceeded, for
the adjustment in maximum benefit rate tended to lag. The percent
age of benefits to covered wages was 40.8 percent in 1939, 33.7 in 1952,
and, despite the great exhortation since 1952, only 34.8 in 1957.
The table below gives the benefit payments under unemployment
insurance, benefit payments adjusted for price change, average weekly
wages of covered workers and the percentage of benefits for full-time
unemployment to average weekly wages in covered employments.
83 U.S. Department of Labor, Bureau of Employment Security, “Adequacy of Benefits
Under Unemployment Insurance, 1937-52,” pp. 16 and 17.
31 Ibid., pp. 13,14.
58
T able
TH E
in c id e n c e
of
in f l a t io n
6-1 .— Unemployment benefits for full-time work in relation to price and
wage movements, 1989, 1946,1952 , and 1951
Average weekly payments for total unemployment....................
Adjusted for price change........................................................
Average weekly wages of covered workers...................................
Percentage of benefits to average weekly wage uncovered..........
1939
1946
1952
$10.66
$10.72
$26.15
40.8
$18.50
$13.26
$22.79
$11.82
$69.09
33.7
1957
$28.21
$13.90
$84.18
34.8
Source: U.S. Department of Labor, Bureau of Employment Security, “ Adequacy of Benefits Under
unemployment Insurance,” a staff report prepared for the steering committee of the Federal Advisory
Council, October 1958.
Inadequacy of unemployment compensation is partly the result of
the failure to impose adequate tax rates, in turn related to the merit
rating system. But also as inflation proceeds and incomes rise, both
because of inflation and other reasons, there tends to be a lag in the
adjustment of maximum benefits and the like. In the early years of
the program only about 25 percent of claimants were restricted by the
maximum benefit provision. In more recent years the proportion has
been about 50 percent, suggesting the needs of lifting of the maximum
benefit as the proportion rises. In the absence of inflation, these ad
justments probably would have been more substantial and, therefore,
benefits would have been greater. Merit rating has had its disadvan
tages, as we shall see, in that it tends to encourage excessive inter
state competition and to keep payroll taxes and benefits down.
Obviously, as prices rise, needs of workers increase to meet a mini
mum standard of living.
One should compare the modest budget that has been worked out by
the Department of Labor for a city worker’s family of four persons
with the average benefits paid under unemployment insurance. For
example, in October 1951, the total cost of the budget varied in 34
large cities from $3,812 for New Orleans to $4,454 in Washington, or
roughly from $320 to $370 per month. At this time, unemployment
benefits, which were likely to be exhausted after 20 weeks, averaged
less than $100 a month, or substantially less than one-third of the cost
of this modest budget.
A study has been made by the Government of the effects of unem
ployment on cash income and spending patterns.
THE INCIDENCE OF INFLATION
59
Table 6-2.— Percentage changes in average monthly cash income1 and outlay of
household before and during unemployment of claimants8 in I- and 4-person
households
Pittsburgh, Pa.
(fall 1954)
Size and type of household and item
Cash
income
1-person household (single claimant):
Before unemployment..........................
During unemployment.............. .........
Percentage change-—...........................
4-person household with claimant as head
of household:
Before unemployment.........................
During unemployment........................
Percentage change............. .................
4-person household with claimant as nonhead of household:
Before unemployment..........................
During unemployment........... ............
Percentage change................................
Cash
outlay
Anderson-Greenville-Spartanburg,
S.C. (spring 1957)
Cash
income
Cash
outlay
Portland, Oreg.
(spring 1958)«
Cash
income
Cash
outlay
$263
$103
-61
$221
$144
-3 5
$178
$92
-4 8
$153
$129
-1 6
$262
$139
-47
$227
$194
-1 5
$405
$192
—53
$374
$271
-2 8
$302
$173
-43
$276
$248
-10
$459
$229
-5 0
$458
$358
-2 2
$441
$344
-2 2
$411
$396
-4
$599
$440
-2 7
$576
$526
-9
(4)
(4)
(4)
(4)
(4)
1Income is net of Federal taxes withheld from wages and salaries.
* Claimants with 6 or more successive weeks of unemployment (8 or more weeks in Pittsburgh).
3 Preliminary data.
* Comparable data not available.
N ote.—Tables 6-2 through 6-4 present data from surveys conducted among samples of claimants chosen
from among claims filed during a selected week. Except as noted, these samples were confined to claimants
who, at the time of sample selection, were in benefit status, members of 1- and 4-person households, and
unemployed for 6 or more successive weeks. Households were excluded in which significant income was
provided by members other than the claimant or the claimant’s spouse.
Source: These tables adapted from U.S. Department of Labor, Bureau of Employment Security,
“ Adequacy of Benefits under Unemployment Insurance,” a staff report prepared for the steering com
mittee of the Federal Advisory Council, October 1958.
T able
6-3. — A v e ra g e m o n th ly w ages a n d u n e m p lo y m e n t benefits o f c la im a n t s co m p a re d w ith total average m o n th ly c a s h in co m e 1 o f ho useh olds
before a n d d u r in g u n e m p lo y m e n t of c la im a n t s 2 i n 1 - a n d 4 ~p e rso n h o u seh o ld s 3
4-person households with claimant as—
1-person households (single claimants)
Nonhead
Head of household
Before unemploy
ment
During unemploy
ment
Pittsburgh, Pa. (fall 1954)...................... ................. Anderson-GreenvUle-Spartanburg, S.C. (spring
1957)......................................- ...................................
Portland, Oreg. (spring 1958) 6...................................
Unem
ploy
ment
bene
fits *
During unemploy
ment
Wages of Total
Total
cash
cash
claim
income 1
income 1
ant 1
Unem
ploy
ment
bene
fits <
$263
$218
$103
$93
$393
$373
$196
$119
178
262
163
238
92
139
61
100
298
459
220
357
169
229
61
118
Before unemploy
ment
During unemploy
ment
Wages of Total
Total
cash
cash
claim
income 1
ant 1
income 1
(5)
$441
599
(8)
$173
205
(8)
$345
440
Unem
ploy
ment
bene
fits *
(5)
$62
81
INCIDENCE
Wages of Total
Total
cash
claim
cash
income 1
ant 1
income 1
Before unemploy
ment
THE
Area and date of survey
OF
ing weeks not compensated due to delayed filing, waiting weeks, time lapse in issuance
of benefit checks, disqualifications, etc.
* Comparable data not available.
* Preliminary data.
INFLATION
1 Wages and income are net of Federal taxes withheld.
2 Claimants with 6 or more successive weeks of unemployment (8 or more weeks in
Pittsburgh; 5 or more weeks in Albany area).
s Includes 3-person households in Albany area study.
4 Average amount of benefits received during period of unemployment studied, includ
T able 6 4.
A v e ra g e w ee k ly c a s h o u tla y o f ho useh old on a l l expenses a n d selected typ es o f expenses d u r in g u n e m p lo y m e n t a n d average w eekly
benefit a m o u n t 1 of c la im a n t s 2 i n 1 - a n d Ir p e r s o n ho u seh o ld s 3
4-person households with claimant as—
1-person households (single claimants)
Area and date of survey
Average per week during unemployment
Pittsburgh, Pa. (fall 1954)......................................
Anderson-Greenville-Spartanburg, S.C. (spring
1957 )..................... .....................: : ......... .
Portland, Oreg. (spring 1958)5_____ ______ ____
Average per week during unemployment
Cash outlay on—
Food, shelter,
All expenses
utilities,
medical care
Average
weekly
benefit
amount1
$33
$21
$26
$61
$37
$29
30
45
18
29
21
32
57
83
32
49
OO
60
OO
Cash outlay on—
Food, shelter,
All expenses
utilities,
medical care
(4)
(4)
'P'iO
00
$22
28
INFLATION
a Includes 3-person households in Albany area study,
« Comparable data not available.
5 Preliminary data.
(4)
♦poO
122
Average
weekly
benefit
amount1
OP
1 Amount of weekly benefit for which the claimant qualifies. Maximum weekly benefit
amount payable during survey period were: Pennsylvania $30: South Carolina, $26:
9el.?n- x
*
.
2 Claimant with 6 or more successive weeks of unemployment (8 or more weeks in Pitts*
burgh; 5 or more weeks in Albany area).
Average per week during unemployment
INCIDENCE
Food, shelter,
All expenses
utilities,
medical care
Average
weekly
benefit
amount1
Nonhead
THE
Cash outlay on—
Head of household
62
the
in c id e n c e
of
in f l a t io n
In these tables we present the effects of unemployment on income
and the like for four cities or regions. For example, let us consider
Pittsburgh. In a four-person household, the claimant as head of the
household had a cash income before unemployment of $405, during
unemployment $192, or a loss of 53 percent. Cash outlay dropped
from $374 to $271, or 28 percent. In the next table, it will be noted
that for a four-person household in Pittsburgh unemployment bene
fits were $119 and total cash income while unemployed was $196. The
benefits were substantially less than one-half of the cash outlays dur
ing the period of unemployment.
On the principle that food, shelter, utilities, and medical care are
absolutely essential expenditures, it will be noted that the average
weekly benefit was $29, and expenditures on food, shelter, utilities,
and medical care, $37, and expenditures on all items, $61. In other
words, the average weekly benefit for a four-person household was
not adequate to compensate for the minimum expenditures for food,
shelter, utilities, and medical care.
In a table not reproduced it is clear that expenditures per unit
household declined from $339 before unemployment to $265 during
unemployment for the Pittsburgh sample. The largest decline was
food from $130 to $101; housing utilities from $60 to $49; house fur
nishings, $9 to $6; clothing, $26 to $17; medical care, $13 to $9; trans
portation, $31 to $26; reading and recreation, $14 to $9; tobacco and
alcoholic beverages, $15 to $12; life insurance, $17 to $11.8S
From an earlier table it will be noted that in recent years general
unemployment compensation rose substantially above the real values
of these benefits in 1939, so that by 1957 they were roughtly 30 percent
as xj.S. Department of Labor, Bureau of Employment Security, “ Adequacy of Benefits Under Unem
ployment Insurance,” a staff report prepared for the steering committee of the Federal Advisory Council,
October 1958.
Average monthly cash outlay of household by type of expense before and during un
employment of claimants1 in 4-person households2 with claimant as head o f household.
Pittsburgh, Pa.
(fall 1954)
Anderson-Greenville-Spartanburg.S.C. (spring
lOK'tt
Portland, Oreg.
(spring 1958) *
Type of expense
Before
unem
ploy
ment
Total cash outlay___________
Food____________________________
Housing and utilities_____________
Household operations_____________
House furnishings________________
Clothing_________________________
Medical care_____________________
Transportation___________________
Personal care____________________
Tobacco and alcoholic beverages___
Beading and recreation___________
Gifts and contributions___________
Life insurance__ ________ ________
Other____________________________
During
unem
ploy
ment
Before
unem
ploy
ment
During
unem
ploy
ment
Before
unem
ploy
ment
During
unem
ploy
ment
$339
$265
$276
$248
$458
$358
130
60
10
9
26
13
31
5
15
14
7
17
2
101
49
10
6
17
9
26
5
12
9
6
11
2
81
48
18
12
18
12
43
5
7
9
7
14
4
75
53
16
11
11
11
32
5
7
8
6
9
3
124
98
22
15
28
26
72
7
14
21
104
80
17
13
16
14
55
7
12
21
(>) 7
23
0
6
12
* Claimants with 6 or more successive weeks of unemployment (8 or more weeks in Pittsburgh).
* Preliminary data.
« Included in “ Other.”
THE INCIDENCE OF INFLATION
63
in real dollars above the 1939 level. But there was very little improve
ment from 1946 to 1957 in real dollars. It can be argued, indeed, that
the rise of unemployment benefits was substantially less than the
increase of per capita real income, which was of the order of about
50 percent during these years. But the fact that benefits were tied
to some extent to wages raised benefits much more favorably to the
beneficiaries than, for example, under the old age and survivors’ in
surance program, where such criteria were not used in any significant
sense. Since the benefits did not rise as much as real weekly wages,
it might be expected that benefits in relation to average weekly wages
in covered industries would decline.
In theory, unemployment compensation (UC) was supposed to
provide assistance for the unemployed worker over the period during
which he would be seeking a new job. The period covered should,
therefore, be adequate and the proportion of benefits to wages be high
enough to cover minimum needs and yet not be so high as to discour
age workers from seeking new employment. It was not expected that
the UC fund would be solvent in the event of another collapse such
as occurred in the early 1930’s.
In many respects the program has failed to achieve these objectives.
Thus in two recent periods of mild unemployment, the benefits have
covered but one-quarter of the cost of unemployment, the explana
tion being the large numbers still uncovered, the small benefit pay
ments relative to wages, the exhaustion of benefits by many workers.
For the years 1938-53, there were 57 million man-years of unem
ployment; but only $12 billion of unemployment benefits were dis
bursed, or about $210 per man-year of unemployment. Li 1949, a
year of unemployment about as severe as 1954 (3.4 million unem
ployed) , wage losses amounted from $7 to $8 billion, and unemploy
ment benefits, $1.9 billion, or 25 percent of the wage losses.88
Prof. Richard Lester writes as follows:
A consequence of the
restricted coverage and
ployment insurance has
the computed earnings
low benefit levels and relatively short duration (plus
uncompensated waiting periods) has been that unem
offset (or compensated for) less than 30 percent o f
lost from unemployment during postwar recessions.
For the first 4 months of 1958, he estimates 29 percent as the com
pensated share of computed earnings lost.37
It is quite clear why the compensation under unemployment com
pensation is inadequate to deal with the large losses of wages due
to unemployment. One point is of course the inadequate duration
and another is the fact that the benefits cover about only one-third
of the average weekly wage. Then after a while the benefit rights
are exhausted. Again, covered employment is only part of total em
ployment. For example, in 1957 covered employment was less than
40 million and employment around 65 million. In October 1957,
insured unemployment was 1.3 million and the total unemployment
2.5 million; or m other words, one-half of the unemployed were
receiving benefits. In April 1958, in the depths of the recession, the
88 Figures from “ National Income,” 1954 edition; Social Security Bulletin, September
1954 : and “ Economic Report of the President,” January 1954.
87 R. A. Lester, “ Issues in Unemployment Insurance,” paper for the Social Security
Conference, East Lansing, Nov. 18, 19,58, p. 7 (mimeographed).
64
TH E
INCIDENCE OF INFLATION
unemployed numbered 5.1 million and the insured unemployed 3.4
million.38 By April 1958, 229,000 had exhausted their benefit rights,
and from January to April 1958, 713,000 had exhausted their benefit
rights.89
One reason for the disappointing results in unemployment compen
sation is the gradual decline in the tax rates. In 1938, 1939, and
1940, the average employer contribution rate was 2.75, 2.72, and 2.69
percent of the payroll. By 1954 this had been reduced to 1.12 percent
and in 1957 was 1.31.40
Undoubtedly, the major deficiency of the program results from
experience or merit rating, under which the employer is allowed to
reduce tax rates on payrolls when benefit payments charged to him are
relatively low compared to the reserve accumulated to his account.
For example, the average employer contribution rate in November
1958 was 1.31 percent. There were 13 States with rates below 1 per
cent with Virginia with 0.53, the lowest, and 2 States with rates above
2 percent: Michigan, 2.04, and Washington, 2.11.41
As of January 1, 1958, there were 15 States which had a statutory
minimum of 0 percent and 11 States with actual rates of a minimum
of 0 percent.42
Experience rating has fundamentally changed the system of unem
ployment insurance. Not only has it reduced rates by about one-half
on the average as compared with the expected rate, but it has con
tributed greatly to the kind of interstate competition that, through a
Federal scheme, the Federal Government was presumably to eliminate.
The enactment of the uniform tax and offset provisions of the Federal
law was designed, according to the Senate Committee on Finance
reporting the 1935 Social Security Act, so that—
all employers * * * will be on an equal competitive position * * *. No State
can gain any advantage through failing to establish an unemployment compen
sation program. This provision will equalize competitive conditions and thus
enable States to enact unemployment compensation laws without handicapping
their industries.48
Because the burden of unemployment is put increasingly on the
industries which suffer greatly from unemployment, as is required
under experience rating, the effect is that the program has lost much
of its insurance flavor; and in a manner, the unemployment compensa
tion has further weakened the vulnerable industries. More of the
burden should have been put upon the entire economy. Thus, in New
York State in 1954, the average tax rate per dollar of taxable payroll
was 1.59; but the extremes were 1.10 percent for the stable finance,
etc., industries and 2.38 for the unstable apparel industry.44
38 Hearings, Senate committee on “Financing Unemployment Compensation,” May 1958,
pp. 28-29,; and 1957 Supplement to “ Handbook of Unemployment Insurance Financial
Data.” p. 2.
89 Senate hearings on “Unemployment Compensation,” pp. 34-35.
40 “ Handbook of Insurance Financial Data,” p. 2.
411957 Supplement, “ Handbook on Unemployment Insurance Financial Data,” p. 5.
42 Senate hearings on “ Unemployment Compensation,” 1958, pp. 50-51.
^Hearings, Senate Committee on Finance, on “Unemployment Compensation,” 1952,
p. 18.
44 Department of Labor, State of New York, Annual Report of the State Advisory Council
on Employment and Unemployment Insurance, 1954, p. 35,
65
TH E INCIDENCE OF INFLATION
In States where unemployment is high, rates tend to be high, but
especially for the unstable industries. The unstable industries pay
heavily and withdraw disproportionately.
Table 6-5.— A m o u n t o f u n e m p lo y m e n t in s u r a n c e p a y m e n ts p e r d o lla r o f
c o n t r ib u t io n 1
Textiles_______________ ______________________ ____ _______
Manufacturing___________________________________________
__ ___________
All industries__________________ ________
1947
1948
1949
1950
$1.92
.67
.49
$0.77
.73
.60
$5.34
2.84
2.25
$1.86
1.32
1.32
1951
$1.69
.60
.56
i Payments from Letter of Mary E. Wilcox, chief supervisor of research and statistics, Division of Employ*
ment Security, Massachusetts, and employment figures from “Employment and Wages,” of Bureau of
Employment Security and Regional Statistics.
Experience rating tends to aggravate cyclical fluctuations. One of
the great advantages of unemployment compensation was held to be
that it would reduce economic fluctuations: funds would be collected
(withdrawn) in periods of high employment and spent (and hence
stimulate the economy) in periods of depression. But under experi
ence rating taxes are cut in periods of high employment (favorable
employment experience) and raised in periods of depression. Hence,
the contracyclical effects are greatly reduced. Thus, receipts from
unemployment trust funds declined only from $1.3 billion in 1948 to
$1.1 billion in 1949, and benefits rose by $1 billion.45 A greater con
tribution might have been made had experience rating not been
effective.
The result of experience rating is that in periods when employment
is high, rates tend to go down because of the effect of experience
rating. Hence, contributions do not fall as much as they otherwise
would in periods of declining wage payments. Thus, in fiscal year
1956-57, total contributions were $1,537 million; in fiscal year 1958,
$1,500 million. But the reduction is very small in view of the fact
that the fiscal year 1958 was a depression year. Again, the average
contribution rate in 1957, half of which was a depression year, was
1.81 as against a rate of 1.32 for the calendar year 1956.
CONCLUSION
It is quite clear that benefits under unemployment compensation are
inadequate when matched against wages, but the rise of benefits has
exceeded that of prices and to this extent the program may be held
to have been successful. This success is to be associated in part with
the fact that the system developed under a theory that benefits should
be adjusted to wages and to this extent the system has been more
successful than the old-age and survivors insurance program.
The failures of the system and the inadequacy of benefits both in
amount and duration are not to be explained primarily by inflationary
process. But there can be little doubt that the continued rise of prices
and hence the cost of the consumer’s budget contributed to the in
adequacy. In order to achieve adequate benefits, it was necessary to
raise the maximum benefit allowable under the various State pro« Figures from Social Security Bulletin, September 1954 (Annual Statistical Supple-
TH E INCIDENCE OF INFLATION
66
rams. But here there has been a lag, and the lag is partly explained
y the inflation. Without inflation, clearly the adjustments of maxi
fmum
benefits in relation to wage rates would have been greater.
But the greatest damage to the program has been done by the merit
rating under which employers are allowed special rates when their
unemployment record seems to be good. This follows even though
the good record may be due to, and generally is due to, outside in
fluence over which they have little control. In another sense merit
rating has been harmful. In periods of expansion and inflation, there
is a tendency for the contributions to decline, for employment records
are then good; and in such periods, the employer is forgiven taxes
instead of being forced to pay larger taxes which would be required
under a proper anticyclical policy. One correction for this particular
approach is to require minimum tax rates at all times. This approach
is, of course, from the viewpoint of offsetting inflationary and defla
tionary forces, not ideal. But at least in this maimer the program
would acquire adequate resources and higher benefits, even though its
ultimate contribution to treating the business cycle would not be as
great as it otherwise would be.
It would be helpful, also, if the administrative agencies of the
Government had some discretion in adjusting rates to changing busi
ness conditions, again assuming minimum rates as well. Therefore,
in periods of great expansion, rates might be raised temporarily and
in periods of deflation they might get down to the minimum fixed by
the Congress.
I can only conclude that, had prices been stable from 1939 on, prob
ably benefits would be a larger proportion of covered wages. But the
great damage, again I repeat, was not done by the inflationary process
but rather by the peculiar influence of the merit rating provisions in
the act.
C h a p t e r 7. I n c o m e M a in t e n a n c e P a y m e n t s W i t h P a r t ic u l a r
R eference to V e t e r a n s ’ P rogram s
According to the Social Security Board, in 1956-57 income main
tenance payments under public and private programs in 1956-57
amounted to $20.5 billion. We have already discussed several of the
programs involved and in particular the old-age and survivors and
disability insurance and unemployment insurance, as well as public
assistance.
The major items in these $20.5 billion were: Social insurance, $11.5
billion; veterans’ pensions and compensations, $2.9 billion; public as
sistance, $3 billion; and private programs, $2.8 billion, of which pen
sion plans are $1.1 billion and other employee benefit plans $1.5 bil
lion. I should also mention workmen’s compensation, $634 million,
and temporary disability insurance, $274 million.
w o r k m e n ’s
c o m p e n s a t io n
In some of these programs the response of compensation to the
rising price and wage level has not been satisfactory and some of the
same problems are raised as were suggested in discussing old-age and
unemployment insurance. For example, under workmen’s compensa
tion tne tendency to impose a maximum compensation results in a bene
TH E INCIDENCE OF INFLATION
67
fit payment that is inadequate in relation to rising prices and rising
wages. The Somerses, in an excellent summary of the situation,46
have made clear the inadequacies of these benefits:
Despite the apparent intention of compensation laws to relate benefit pay
ments to a given ratio of wage levels, usually around two-thirds, the result is
obviously something very different. Periodic liberalization of the various maximums and qualifications have utterly failed to keep pace with rising wage levels
and prices. As a result, the proportion o f wage (loss) compensation declined
substantially.47
For example, a study in 1953 for Illinois shows that the ratio of
weekly maximum compensation for an injured worker with one child
to average weekly earnings declined in that State from 98 percent in
1913-14 to 34 percent in 1952.48 The Social Security Administration
concluded in 1954—
that workmen’s compensation is probably leaving unmet on the average about
two-thirds o f the wage loss in temporary disability cases and an even greater
proportion of the aggregate loss from all disabilities of covered workers, in
cluding those fatally or permanently injured * * *.49
Apparently the average payment in 48 States for a widow and 4 chil
dren was $20 a week in 1951. That was the average payment in 48
States.50 The Somerses conclude that in workmen’s compensation we
have approached pretty close to a flat-rate payment and the adjust
ments to rising prices and wages are slow indeed.
I do not mean to give the impression that the inadequacies of work
men’s compensation are to be explained merely by the rise of prices
and incomes. But the combination of interstate competition which
results in fixed maximums and the rising price and income levels with
slow upward adjustments in benefits, and also the large diversions to
insurance companies and other intermediaries—these together help
explain the inadequacies that tend to become greater despite the rising
incomes.
An injured worker receives much less than the loss of wages in
curred. The compensation for the temporary disability is reduced by
the requirement of a waiting period; for permanent disability it is
reduced by ceilings on the period of compensation and on payments
to be made. According to one expert, in North Carolina 48 percent
of the wage loss was compensated in temporary disability costs m 1940
and only 21 percent in permanent and fatal cases; in Massachusetts
in 1933, 55 and 25 percent, respectively; in Illinois, in 1952, 30 percent
for temporary cases, 13 percent for permanent-partial cases and less
than 6 percent for fatal cases. In general^ about two-thirds of the
wage loss for temporary disability is not being compensated and less
is being recovered for the permanent disabilities.51
There are 8 States with earned premiums in excess of 1.5 percent of
payrolls and 16 with rates from 0.99 to 0.72. Substantial differences
m the premium and benefit costs of six northern and six southern
industrial States are to be noted—from 25 to 30 percent. Injuries
vary also so that the differences do not reflect equal variations in re
46H. M. Somers and A. R. Somers, “ Workmen’s Compensation,” 1954, pp. 77-82.
" ib i d ., p. 77.
148Ibid., p. 77.
" Ibid., p. 81.
60 Ibid., p. 82.
81 Social Security Bulletin, March 1954, op. dt., pp. $-10.
68
TH E INCIDENCE OF INFLATION
suits. Moreover, States vary greatly in what they get per dollar of
premium. Thus, Massachusetts pays 28 percent less than New York
but gets 13 percent more in benefits.
The President’s Commission on the Health Needs of the Nation
thus summarizes workmen’s compensation:
Since 1911 the workmen’s compensation laws o f the various States have become
keystones in America’s industrial health progress. Workmen’s compensation
systems in the States vary from excellent to grossly inadequate. In 11 States,
the law applies only to certain listed “hazardous” employment; 4 States give no
coverage to occupational diseases and 18 cover only certain listed diseases. Ex
cessive litigation is common, with both legal and medical chicanery * * *.
Eleven States have no factory inspector. Almost all compensation payments
are inadequate by present-day standards, particularly in providing for total
rehabUitation * * *.
When monetary benefits are awarded, they are usually inadequate. Theo
retically determined by a percentage of wages, cash payments are actually
restricted to a statutory maximum of $25 to $38 per week in most States, a
maximum which does not reflect current inflationary trends: medical-expense
payments are strictly limited in 17 States; only 19 States have provisions for
rehabilitation in workmen’s compensation.
LONG-TERM COM MITM ENTS
I shall discuss private pension plans later. But here I want to com
ment briefly on the violence done to contracts by long-term price and
income rises. It is easy to understand why pensions are in many ways
unsatisfactory. Where the plans are not funded, of course, there is
always the problem of whether adequate cash will be available when
the payments have to be made. Even where they are funded, there is
always the problem that prices and per capita income rise, and a pen
sion plan that is worked out on the assumption that wages will remain
where they are today causes large underpayments in the years when
the worker retires. In the face of past history, if, for example, the
average wage is $80 a week today, the pension plan should really be
worked out on the principle that when a young man entering the labor
market is given a right to a pension, the wage will be at least double
in a period of 25 years. This allows for a small amount of inflation as
well as the gains of productivity. In college experience over the
years, it has been discovered that if a plan is made out on the theory
that 50 percent of the wage at the end of the working period would
be available, what is actually available comes to about 25 percent.
In general, what we can conclude is that the adjustment to rising
prices has been achieved in some programs, for example, in unemploy
ment insurance where the tieup has been with wage rates, but in many
programs for long periods of time the benefit payments have not
matched the rise of prices, and a fortiori the rise of per capita income
or average wages. This is a problem which is likely to become of
much greater importance as the maintenance payments continue to
rise. For example, the Department of Health, Education, and Wel
fare and the U.S. Department of Labor have estimated that the in
come maintenance payments under public programs rise as follows:
1940, $2.1 billion; 1950, $6.3 billion; 1955, $11.1 billion; 1965, $18.5
billion; 1975, $24.3 billion; 1985, $29.6 billion.
THE
69
INCIDENCE OF INFLATION
The major factors by 1985 would be: 50
B illio n
Old-age and survivors insurance------------------------------------------------------------- $18.8
Unemployment insurance----------------------------------------------------------------------2. 0
Public assistance----------------------------------------------------------------------------------2. 0
Federal civilian retirement-------------------------------------------------------------------1. <S
Federal uniform services, retirement----------------------------------------------------1. 5
State and local government retirement---------------------------------------------------1. 0
Workmen’s compensation----------------------------------------------------------------------1.0
Railroad retirement------------------------------------------------------------------------------.5
VETERANS' BENEFITS
In 1940, veterans’ benefits amounted to $585 million; by 1947, $6,530
million; by 1957, $4,681 million; and in 1960, according to the Presi
dent’s budget estimate of January 1959, $5,088 million. The break
down in 1947 and 1957 was as follows:
[In millions]
1946-47
Pensions and compensations--------------------—..................................... ................. .
Readjustments____________________________________________________________
[fealth and medicine
- _
______________________________________
Education------------------------------------ --------------------------------------------------------------Welfare and others_____________________ _______________________ _______ __
Total
____ _ - _________________________________________________
1956-57
$1,831
1.512
572
2,251
365
$2,906
6, 530
i 4,681
770
811
194
i
1$159,000,000 not included here are from State and local sources.
Source: Social Security Bulletin, October 1955, pp. 6-7, and October 1958, p. 23.
The President’s budget gives the following figures for 1960:
[In millions]
A. Readjustments:
1960
Education and training--------------------------------------------------------------- $490
Loan guarantee and other benefits------------------------------------------------107
Unemployment compensation------------------------------------------------------8
B. Compensations and pensions:
Service-connected compensations_________________________________ 2,043
Non-service-connected pensions__________________________________ 1,203
Burial expenses, other___________________________________________
61
C. Hospital and medical care, except construction________________________
891
D. Hospital construction________________________________________________
55
E. Insurance and servicemen’s indemnities______________________________
49
F. Other services and administration____________________________________
181
Total______________________________________________________________5,088
Source : President’s budget 1960, p. M72.
50 See “Veterans* Benefits in the United! States,” report to the President by the Presi
dent’s Commission on Veterans’ Pensions, April 1956, “Findings and Rcommendations,”
p. 118; also see Social Security Bulletin, October 1958, p. 27.
48084— 59-------6
70
TH E INCIDENCE OF INFLATION
On the assumption that the present laws would continue, the Presi
dent’s Commission on Veterans’ Benefits made the following estimates
of costs in the future:
T a b le 7 -1 . —Veterans9 Administration
budget expenditures under present laws,
selected fiscal years, 1940-2000
[In billions]
1940
1955
1965
1975
1985
2000
Total................................................................................
$0.56
$4.4
$4.7
$4.8
$6.0
$5.8
Non-service-connected pensions..............................
Service-connected disability and death benefits...
All other....................................................................
.12
.33
.11
.8
1.9
1.7
1.7
2.0
1.0
2.0
1.8
1.0
3.4
1.6
1.0
3.6
1.2
1.0
Source: The President’s Commission on Veterans’ Pensions, “Veterans’ Benefits in the United States,”
a report to the President, “ Findings and Recommendations,” p. 106.
The estimated numbers of veterans is as follows:
T a b le 7 -2 .—
Estimated total living war veterans and veterans aged 65 and over,
selected dates, 1940-2000
[Millions of veterans]
1940
1955
1965
1975
1985
1995
2000
Total........... ........... ............... ..........................
4.3
21.9
21.7
18.7
14.6
9.4
6.8
Veterans under 65_____________________
Veterans 65 and over__________________
4.2
0)
21.2
.7
19.4
2.3
16.6
2.1
10.1
4.5
2.3
7.1
.2
6.6
>Negligible.
Source: Ibid., p. 70.
The Veterans’ Administration estimated that the per capita cost of
public income maintenance rose from $19 in 1940 to $83 in 1955, and
would rise to $132 in 1965, $144 in 1975, and $156 in 1985 (the last
three estimated). Veterans’ costs would rise from roughly one-fifth
of the total, that is, $16 out of the total of $83 in 1955 to 24 percent in
1975, $34 out of $144, and to 27 percent by 1985, $42 out of $156. The
assumption is the continuance of present legislation.53
In assessing the benefits to be paid to veterans, the President’s Com
mission suggested that needs should be taken into account. Variations
in the cost of living and in the income levels in different parts of the
country should also be considered. The benefits should be based on the
minimum needs as assistance programs are, and should in the long
run be, lower than benefits paid under old age insurance. The Com
mission believed it was important that payments be not so large as to
damage incentives.54
The Commission was also greatly concerned that the benefits should
not be put on such a level as to put an excessive burden on future
generations. Of course, in modern theory each generation pays its own
bill for the most part, although it is true that commitments for ex
cessive benefit payments might impose a serious financial problem on
later generations. In view of the tendency for prices and incomes per
capita to rise over the years, I do not believe that present proposals
71
TH E INCIDENCE OF INFLATION
on veterans’ benefits are likely to prove as much of a burden on future
generations as the Commission itself believes. The tendency on the
whole is to underestimate the rise of future income in current dollars.
In this connection, note the following table:
T able
7-3.— Public expenditures for income maintenance programs and the
national income (selected years, 1940-85)
Actual
Projected
Item
1940
1950
1955
1965
1975
1985
A. National income (billions of dollars)1............. ........
81.6
240.0
322.2
414.0
571.0
756.0
B. Expenditures for income maintenance programs
(billions of dollars):
1. Assuming no increase in benefit rates:1
General programs____________________
VA compensation and pensions............ .
2.2
.4
6.4
1.9
11.1
2.7
18.5
6.6
24.3
7.6
29.6
11.1
Total............... ................ ...................
2.6
8.3
13.8
25.1
31.9
40.7
2. Assuming future increase in benefit rates at
half the rate of increase in national pro
ductivity:
General programs______ ______ _______
VA compensation and pensions_______
2.2
.4
6.4
1.9
11.1
2.7
21.1
7.5
32.1
10.0
45.9
17.2
Total..................................... ..............
2.6
8.3
13.8
28.6
42.1
63.1
C. Adjusted income maintenance expenditures, as
percent of national income (percent):
General programs...............................................
VA compensation and pensions_____________
2.7
.5
2.6
.8
3.5
.8
5.1
1.8
5.6
1.8
6.1
2.3
Total......................... ........................... ...........
3.2
3.4
4.3
6.9
7.4
8.3
i Estimated figures assume present laws and benefit rates with 2 exceptions: OASI cash disability bene*
fits at age 50 are assumed along the lines of H.R. 7225, 84th Cong.; VA service pensions are assumed as de”
scribed in text.
Source: The President's Commission on Veterans’ Pensions, “ Veterans’ Benefits in the United States,”
a report to the President, “ Findings and Recommendations,” p. 124.
This table suggests that on the basis of present benefits the veterans’
compensation and pension program would increase from $2.7 billion
to $6.6, $7.6, and $11.1 billion in 1965, 1975, and 1985, respectively.
But if it is assumed that the future increase in benefits increases at
half the rate of the increase in national productivity, the amounts
would be $2.7 billion for 1955, and $7.5, $10, and $17.2 billion for 1965,
1975, and 1985. The latter, as a percentage of total national income,
would be 0.8 percent in 1955, and 1.8,1.8, and 2.3 for 1965, 1975, and
1985, respectively. And for all public income maintenance programs,
on the assumption of a rise of benefits of one-half the rise of produc
tivity, the percentage of income would be 4.3, 6.9, 7.4, and 8.3. This
seems like a fairly large rise in the costs of maintenance programs in re
lation to the national income. It should be noted that the estimate is
that the national income would be about $756 billion by 1985. Even
this seems rather to underestimate the possibilities. The Republican
policy committee has just issued a projection of $900 billion long be
fore 1985. In 30 years the per capita income should be at least double
what it is today and therefore the average nonfarm family income
might well be of the order of $12,000 or $14,000. Under these circum
stances, even if these maintenance programs took 8.3 percent as
against 4.3 percent in 1955, this would be no great catastrophe, all the
more so since a large part of the benefits would come from payments by
the contributors. Actually, in view of the slow adjustment of benefits,
72
THE
INCIDENCE
OF INFLATION
the percentage is likely to be less than 8.8 and especially since income
is projected rather conservatively. An increase of benefits at one-half
the rate of productivity increase might very well match the increase
in the cost of living and in the absence of new legislation and aside
from increased costs under current legislation leave benefits at a re
duced level in relation to real income. In other words, the assump
tion of an increase equal to half the productivity gains (in current
prices) might provide no real rise of benefits except insofar as the
aging of the population would bring increased numbers with benefits
and higher benefit as years of coverage are increased.
On the other hand, in view of the large numbers concerned and par
ticularly in the older age group, political pressures will be great and
these political pressures might very well result in the rise of benefits
more rapid than the increase in the cost of living. It is to be noted
that in the 1950’s this pressure has been reflected in much more rapid
adjustments in benefits under old-age, survivors, and disability insur
ance.
In this connection, it is worth while looking back to see what has
happened in the past.
T able
7-4.— Estimated average cost of veterans9 benefits per serviceman in each
war
[In thousands]
UNDER PRESENT LAWS
Conflict
Civil War_______________________________________________
Spanish-American War___________________________________
World War I. ..............................................................................
World War II __________________________ ____ ____ ______
Korean conflict____________________________________ _____ -
Cost per serviceman
Number in
Armed Forces
during war
Non-serviceconnected
Total
pensions
2,213
392
4,744
16,535
2 5,331
0)
0)
$6.5
6.6
9.9
$3.7
12.2
12.7
14.1
14.9
ASSUMING SERVICE PENSION FOR RECENT CONFLICTS
Civil War _____________________________ _______ _________
Spanish-American War___________________________________
World War I _________________________ ____ ______________
World War II. ........................................................... ................
Korean conflict__________________________________________
2,2 1 3
392
4 ,7 4 4
0)
(»)
3 .7
1 2 .2
2 2 .0
16, 5 3 5
1 6 .0
2 1 .4
2 8 .7
2 5 ,3 3 1
2 9 .7
3 4 .5
1 No breakdown is available, but expenditures were largely for non-service-connected pensions.
2 Does not include 1,476,000 veterans who served in World War II and Korean conflict.
Source: “Veterans’ Benefits in the United States,” p. 115.
I
have compared the cost per serviceman in the four major wars,
both on the assumption that present laws will continue and on the
assumption that service pensions for recent conflicts will be made
available as for earlier conflicts. Should the latter happen, of course,
the costs would increase greatly. The table does suggest that the cost
of veterans’ benefits vis-a-vis per capita income has tended to decline
under present laws substantially since World War I. Inflation has
had something to do with these losses. The large increase in World
War I vis-a-vis the Civil War can be explained in no small part by the
very small benefits made available under the Civil War. But should
THE
73
INCIDENCE OF INFLATION
the service pensions of wars before World War II and the Korean
war be voted for the later conflicts, then the increase in cost would be
substantially larger than the rise of per capita income for World War
II veterans, though not for those of the Korean war. Actually, the
differences are not large. In relation to World War I, however, the
rise of pensions would be much less than that of per capita income.
These results are based on comparisons with pensions in the Civil War.
I should make clear that I have measured the per capita incomes as an
average of the four wars, that is, the per capita incomes during the
wars and not the per capita incomes when the benefits are to be paid.
Obviously, per capita incomes on the average would be much higher
in periods of payment than during the course of the war. To this
extent, this table overestimates the costs of the pensions in relation to
per capita income.
Table 7-5.— Per capita income in 4 major wars, compared with veteran pensions
per serviceman in each war, tinder present laws and assuming service pensions
for recent conflicts
Wars
Civil War____
World War I....
World War II..
Korean war___
Per capita in
come (Civil
War=100)
100
200
Oil
957
Cost per service
man under
present laws
100
343
380
403
Cost per service
man assuming
service pensions
for recent
conflicts
100
505
776
032
Source: Calculations from “ President’s Economic Report, 1959” ; “ Historical Statistics of the United
States, 1789-1945” ; my computations; also ‘'Veterans’ Benefits in the United States,” p. 115.
This table does suggest at least that the benefits have more than
responded to the price level and they had not responded to the rise of
per capita incomes. But should pensions be voted for recent conflicts
as in the past conflicts, then the total cost per serviceman would rise
roughly in the same proportions as the per capita income at the time
of the wars. But the rise would probably be less than the increase in
per capita income at the time the benefits are being paid. Moreover,
this is in relation to the Civil War when benefits were surprisingly
small. In relation to World War I the rise of benefits even with
pensions as in earlier wars, would be only 15 percent as large as the rise
of per capita income for World War II and Korean war veterans.
Another indication of what has happened is given by the following:
World War I basic rate for 100-percent disability=$30 per month regardless
of rank:
1919__________________________________________________________________$80
1924__________________________________________________________________100
1954__________________________________________________________________181
It should be noted that those who had 100 percent disability were
only about one-eighth of the total in 1955.55
The benefits to those with 100 percent disability has increased sub
stantially more than the price level, and by 1955 had even exceeded
the rise of per capita income.
55 “Veterans’ Benefits in the United States,” pp. 148,149.
THE INCIDENCE OP INFLATION
74
T able
7-6.— Rise in veterans9 benefits for 100 percent disability and rise in the
cost of living, 1914-54, various years
Benefits,
total dis
ability
100
1914.
1919
1924
1954.
267
Cost of
living
100
157
170
Sources: Last table and also “ Veterans’ Benefits in the United States,” p. 148.
Benefits do not seem excessive for disabled veterans: $1.4 billion is
being paid out to 2 million disabled veterans, or an average of about
$700 a year, with monthly payments varying from $17 to $181. The
average amount was after all only $700 a year. Nor do the payments
to survivors of $400 million to 380,000, or a little more than $1,000 a
year, seem excessive.56
In one area, of course, there would be substantial loss, and that is
the insurance made available to servicemen. In 1955, there were
5,600,000 policies still in force or in waiver status, with a total of $37
billion of national service life insurance.57 The continued rise in
prices since this insurance went into effect does, of course, cut down
the benefits substantially.
Finally, there is a good deal said in the veterans’ report about the
high economic status of veterans, a status that is substantially higher
than that of nonveterans. It is also pointed out that the average
basic pay plus allowances has varied as follows:
[1918=100]
1865__________________________________________________________________ 58
1898__________________________________________________________________ 55
1945__________________________________________________________________ 187
1952__________________________________________________________________ 304
1955__________________________________________________________________ 333
The rise in real compensation, and this leaves out of account some
special benefits such as medical care, was 86 percent from 1918 to
1955. For enlisted men alone, it was 76 percent.58 The point made
here by the Commission is that the soldiers have been treated well
while enlisted and to that extent do not need special compensation.
The average basic pay rose from $797 in 1918 for all military person
nel to $1,805 in 1955, an increase of 126 percent. This rise is, how
ever, considerably less than the increase of per capita income in the
whole population.
PENSION FUNDS
Partly as a result of dissatisfaction with the benefits under old-age
survivors and disability insurance programs, workers and their repreby the workers. By the end of 1957, it has been estimated that the
number of workers covered by private pension programs total 17.7
THE INCIDENCE OF INFLATION
75
million. Estimates of the amount of assets in these funds are now
about $40 billion.
^In 1945 there were only 6,400,000 covered with employer contribu
tions of $830 million per year and employee contributions of $160
million. By 1957 the total increased to 17,700,000 workers covered,
employer contributions to $3.9 billion, and employees to $680 million.
The reserves in 1957 were estimated at $35 billion, the number of
beneficiaries 1,250,000, and the amount of benefit payments $1,150
million. A survey as of January 1,1957, of all pension plans in New
York State, which covered 1,500 workers or more, established the fol
lowing median benefits for workers with 30 years of service:
Median benefit as a percent
of compensation
Average annual compensation
$3,000.
$4,200.
$6,000.
Monthly
median
benefit
(a)
(b)
Excluding Including pri
social security mary social
benefit
security
benefit
$65
72
95
64
54
40-42
Source: R. Tilove, “ Pension Fund and Economic Freedom,” 1959, pp. 9-17.
The Fund for the Republic study estimates that 45 to 55 percent of
all wages and salary labor force, outside of Government and agricul
ture, would be covered by private pension plans in another 10 years.69
A private pension program is especially likely to suffer as a result
of inflation. As inflation proceeds, a public program may make ad
justments, or may put part of the burden on the taxpayer, but this
is more difficult under a private plan. Hence, the private plan tries
to protect workers increasingly 1by putting its assets into common
stocks. In a general way, it may be assumed that the price of com
mon stocks will rise more than prices. The reason for this is partly
that with rising prices, dollar value of profits will increase even more;
and also because the corporation is run primarily for the common
stockholder rather than for the bondholder. In 1951, corporate pen
sion funds had 11.8 percent of their investments in common stock; by
1957, 24.7 percent. Virtually all other assets are fixed in dollars and,
therefore, insofar as inflation proceeds, the pension funds lose part of
their real value.60 In 1957 corporate pension funds invested 37 per
cent of the net receipts in common stocks, a ratio considerably higher
than in any previous period.61 The Fund for the Republic estimates
that by 1965 self-insured pension funds may hold $17 to $20 billion
in common stock, although this may amount to no more than 6 per
cent of the total outstanding.62
The beneficiary of private pension funds will suffer to some extent
as a result of inflation. To some extent, the losses may be recouped
through larger investments in common stock, but there are institu
« Ibid., p. 19.
00 Ibid., p. 33.
« Ibid., pp. 35-36.
« Ibid., p. 84.
70
THE INCIDENCE OF INFLATION
tional and other legal obstacles to heavy investments in common stock.
In the past there also have been serious losses as the result of excessive
payments to banks and insurance companies, and some corruption by
trade union leaders. The result has been that a substantial part of
the payments made 011 behalf of workers has been wasted. The hear
ings before the Senate Labor and Welfare Committee in 1955 and
later years reveal large losses of this kind and resulted ultimately in
the Pension Disclosure Act, which at least provides some publicity in
the handling of the pension funds.
C h a p t e r 8. A
ssets a n d
I
n f l a t io n
in the long debate over the pros and cons of inflation, much lias
been made of the fact that inflation robs savers and in particular those
who have fixed claims in dollars—for example, of bonds or bank de
posits. It is, of course, clear that if one purchased a Government bond
yielding 3 percent in 1945 and the bond were to be redeemed in 1955,
and if prices have risen by 75 percent in the meanwhile, the bondholder
has suffered a serious loss. Again, in 1958, the American public held
$550 billion o f life insurance on 124 million individuals. Legal reserve
life insurance companies accounted for $494 billion, or an average of
$ 11,000 per insured family. In 1939, there had been $109 billion of
life insurance in force. Obviously any o f this insurance that was
realized in later years suffered from a rise in prices and a reduction
in the real value of the payment. Anyone, for example, receiving
benefits o f an insurance policy issued before 1939, in the year 1958
would experience a substantial loss in the real value o f his policy. The
same, o f course, would hold for the $152 billion o f life insurance in
force in 1945, for there has been a substantial inflation since 1945. But
it is of interest that since 1949, the total life insurance in force rose
from $214 to $494 billion in 1958, or a rise o f $280 billion, and during
this period, the inflation has not been large. O f course, the bene
ficiaries o f these policies are not likely to realize their benefits for many
years, and therefore the real issue is : How much will inflation develop
once their benefits are received ? It is an interesting fact that despite
the large inflation since 1940 life insurance per family has risen from
$2,700 to $8,800, and in relation to disposable personal income per
family, from about 160 percent in 1940 to a little less than 160 in 1958.
In view o f the very large advance o f public insurance, notably the old
age insurance, this is a remarkable record and suggests that the public
still is not aware or very much interested in the process o f inflation.
One o f the greatest blocks to inflation is the lag in the general realiza
tion o f its presence.
O f course, life insurance companies might conceivably invest in
equities, and therefore protect their policyholders by investing in
equities as an offset to inflation. This would not do the policyholder
any good unless he had an escalator clause in his policy. But they are
unable to invest in equities as a rule, in part because o f the legal re
strictions. O f $108 billion o f assets held in 1958, U.S. insurance com
panies had only $4 billion in stocks. They also held a little over $3
billion in real estate, an asset that rises with inflation. But most of
their assets are in types that do not respond to rising prices.63
68 All facts in these last few paragraphs are from “ Life Insurance Fact Book,” 1959, and
my calculations.
77
THE INCIDENCE OF INFLATION
But let us take a broader look at the problem of assets.
T able
8-1.— The share of national saving in ch anges in national wealth and
combined net worth, period totals: 1897 to 19^9
Change in
wealth
Change in
combined
net worth
Saving
(1)
(2)
(3)
Billiom
Billions
$8
45
87
112
-2 0
11
35
170
439
447
Period
1897-1900.......................................................
1901-12................................................ .........
1913-22...........................................................
1923-29................................ — .....................
1930-33................... — .............— ................
1934-39............................. ................... .........
1940-45................ -•............................. ..........
1946-49......... ......................... ................. .
1901-49......... ...................................... - ........
1897-1949
...............................................
Billions
$19
77
169
105
-109
65
175
328
810
830
$103
216
227
-206
69
264
357
1,030
Share of saving in
change in (percent)—
Wealth
Net worth
(4)
(5)
42
58
51
107
18
17
20
52
54
54
44
40
49
10
30
13
48
43
Source: R. W. Goldsmith. “ A Study of Savings in the United States,” vol. I, p. 194, 1955.
This table reveals the increases in national wealth and net worth from
1897 to 1949. Assets seem to rise in all periods except in 1980-33
when there was a substantial deflation, As a general rule, savings ac
count for roughly one-half o f the increase in wealth, but there are
important exceptions: 1930- 33, 1934- 39, 1940- 45. Wliere savings
are not the explanation, the major factor is of course the increase in
the price level. The explanation o f the greater rise o f national net
worth than wealth is given as follows by Goldsmith : 64
While national wealth changes essentially reflect three factors only—cumu
lated domestic saving, changes in the prices of tangible assets, and, though much
less importantly, the net foreign balance—combined national net worth shows
in addition the effects of realized capital gains and losses, and of unrealized
appreciation or depreciation due to changes in the prices of equity securities.
Combined national net worth is, furthermore, affected by certain duplications,
such as the inclusion of the market value of corporate stock and the balance sheet
value of participations in unincorporated business enterprises, first in the calcula
tion of the net worth of the owners and then again in that of the balance sheet
value of the equity of corporations and unincorporated businesses. Combined
national net worth may therefore be expected to show wider fluctuations, at least
in absolute terms, than national wealth, with the result that cumulated saving
accounts for a smaller proportion of changes in the former than in the latter.
64 Ibid., pp. 194-195.
78
THE INCIDENCE OF INFLATION
Table 8-2.— Comparison of personal saving and net worth changes for the period
1901-49
[In*billions]
Holdings
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Real estate.......................................
Producer durables...........................
Consumer durables............ ............
Inventories-.....................................
Currency................................... ......
Deposits.........................................
Life insurance reserves___________
Pension and retirement funds........
U.S. Government securities..........
Other bonds....................................
Stocks.............................................
Mortgage holdings...........................
Equity in unincorporated business.
Other assets........... ................... ......
Total assets.......... ..........................
Debt.............................................. .
Change in
net worth
Saving
1901-49
Difference
(3 -4 )
(2)
(3)
(4)
(5)
$349.4
28.8
99.3
31.9
23.3
$304.1
26.6
93.3
24.7
$108.7
End of
1900
End of
1949
(1)
$45. &
2.2
6.0
7.2
1.3
6.1
1.6
0
.6
3.9
10.7
4.2
6.3
4.1
99.4
14.9
120.2
58.8
45.7
61.6
17.7
117.8
25.3
68.8
24.2
1,072.8
105.3
22.0
114.1
57.2
45.7
61.0
13.8
107.1
21.1
62.5
20.1
973.4
90.4
20.1
81.2
3.0
22.3
118.5
58.1
46.2
67.4
25.3
50.2
17.8
1.1
619.9
94.7
$195.4
6.5
12.1
21.7
-.3
-4 .4
-.9
-.5
-6 .4
-1 1.5
56.9
3.3
62.5
19.0
353.5
-4 .3
Source: Ibid., p. 197. For notes, see original table.
This table shows that savings account for about two-thirds of the
rise o f net worth. The largest increase in an item over the amount of
savings toward the accumulation of this asset is in real estate.
Between 1900 and 1949 the current value of real estate held by individuals has
increased by about $200 billion more than owners’ saving, i.e., the amounts
which have been spent for construction and related outlays, after account is
taken of depreciation allowances on original cost basis and of increase in mort
gage debt. This represents well over one-half of individuals’ total unearned
increment. The valuation difference of $200 billion, of which about two-fifths
is accounted for by the increase in the value of land while nearly three-fifths
reflects the increase in the level of construction costs, is equal to over threefifths of the current value of all real estate held by individuals at the end
of 1949.65
It will be noted that the other two items that reveal large gains in
relation to the savings made available are inventories and stocks.
Another interesting item in relation to the inflation problem is the
growth o f liquid assets in relation to total assets. Where liquid assets
tend to rise to that extent the protection against inflation is reduced.
By liquid assets we mean monetary metals, currency, commercial bank
deposits, deposits in other financial institutions, U.S. Government and
State and local government securities. For example, for households
the percentage of liquid assets to total assets rose from 7.8 percent in
1900 to 10 percent in 1929, to 23.6 in 1945 and 19.7 percent in 1949.
For business enterprises, the percentages in these years were 11.4, 11.1,
47.4, and 35.5.66
This increase o f the share of liquid assets is an important factor in
the rise o f the share of intangible to total assets. Here are the per
centages for 1900, 1929, 1945, and 1949:
« Ibid., pp. 196-198.
w Ibid., p. 202.
79
THE INCIDENCE OF INFIjATION
T a b l e 8 -3 . — Ratio
of intangible assets to total assets
1900........................................................................
1929.......................................................................
1945........................................................................
1949........................................................................
Households
Business
enterprises
37.5
56.0
60.5
53.5
54.7
61.6
71.6
64.4
Govern
ments
Total
18.4
26.1
44.8
35.9
43.3
56.5
63.4
56.3
Source: Ibid., p. 203.
G o ld s m ith c o n c lu d e s as fo llo w s : 67
The structural changes in the first half of this century have thus reduced
the share of price-sensitive assets of households, banks, and governments, and
have increased the share in the case of nonfinancial business enterprises. This
means that, since debt-asset ratios have as a rule decreased substantially over
the same period, a change in asset prices of the same proportion has now a
relatively smaller effect on the net worth of farm and nonfarm households
than 50 years ago, but has still about the same repercussion on the equity of
nonfinancial business enterprises, and a larger effect on that of certain financial
institutions. In other words, the real net worth (i.e., the current net worth
divided by an index of the general price level) of individuals is now more
susceptible, at least for households in the aggregate, to dilution by inflation
than it was 20 or 50 years ago.
In other words, with the rise of prices households are much more
sensitive to rising prices, for the value o f their assets is not likely
to rise as much as in earlier periods. They hold a smaller proportion
of price-sensitive assets. Nonfinancial business enterprises, on the
other hand, have increased their share of price-sensitive assets as
against households, banks, and government which have reduced their
shares.68
T able 8 -4 . — Percent
price-sensitive to total assets, major saver groups, 1900 to
1949
1900
Households____ ___________________________________ _____ _
Business enterprises______________________________________
Nonfinancial corporations___________________________ ____ _
Governments. ...................... ............................ ..........................
81.4
50.2
67.1
71.3
1929
75.6
50.4
70.4
67.1
1949
65.4
40.5
74.7
50.3
Source: Ibid., p. 205.
A good indication of the effects of inflation on particular groups
is given by the debt ratio o f the major saver groups. Households
are in a less favorable position, business enterprises in a more favor
able position, and governments in a much more favorable position.69
80
THE in c id e n c e o f in f l a t io n
F ig u r e s f o r 1 9 0 0 ,1 9 2 9 ,1 9 4 9 a re a s fo llo w s :
T ab le 8 - 5 .— Debt
ratio of tnajor saver groups, 1900, 1929,194$
Households
Business
enterprises
Governments
10.1
12.7
8.6
54.0
55.0
57.7
48.3
63.6
156*2
1900— .........................................................................................
1929.........................................................................
1949.......................... .....................................................................
Monroes: Ibid., p. 210.
C h ap te r 9. M ortgages , I n s t a l l m e n t P a y m e n t s ,
I n f l a t io n a r y P rocess
and the
M o rtg a g e e s a re in a n e s p e c ia lly g o o d p o s itio n to p r o fit fr o m in fla
tio n . A s p ric e s ris e , th e y p a y b a c k in te re s t a n d a m o r tiz a tio n in d o l
la r s o f re d u c e d p u rc h a s in g p o w e r as in fla tio n p ro c e e d s . I n a d d itio n ,
th e ir b u rd e n is re d u c e d in s o fa r as in c o m e ris e s w it h in c re a s e d p r o
d u c tiv ity ; th e b u rd e n is re d u c e d in s o fa r as e a c h h o u r o f w o rk y ie ld s
a la r g e r n u m b e r o f d o lla r s .
N o n fa r m m o rtg a g e s in c re a s e d fr o m $ 3 1 b illio n in 1 9 2 9 to $ 4 5 b il
lio n in 1 94 8 a n d $ 1 4 4 b illio n in 1 9 5 8 . T h e m o rtg a g e e g a in s in s o fa r
as m o re d o lla rs a re a v a ila b le , e ith e r b ecau se o f in c re a s e d p r o d u c tiv ity
o r in fla tio n , in r e p a y in g a m o rtg a g e d e b t. B u t i t m u s t b e re m e m
b e re d th a t o th e rs lo s e . A c c o rd in g to th e F e d e r a l R e s e rv e B o a r d in
J u n e 1 9 5 9 , th e re w e re $ 1 2 5 b illio n o f n o n fa r m m o rtg a g e s o u ts ta n d in g
o n o n e - to fo u r -f a m ily h o u ses, a n d $ 1 6 9 b illio n i f w e in c lu d e m u lt i
fa m ily a n d c o m m e rc ia l p ro p e rtie s . F a r m m o rtg a g e s w e re $ 1 2 b illio n
a d d itio n a l in J u n e 1 9 5 9 .
A g a in s t th e g a in s o f th e m o rtg a g e e a re th e losses o f th e m o rtg a g o r.
O f a ll m o rtg a g e s o f $ 1 8 1 b illio n , fin a n c ia l in s titu tio n s a c c o u n te d f o r
$ 1 3 8 b illio n , r o u g h ly th re e -q u a rte rs . T h e h o ld in g s in J u n e 1 9 5 9 o f
a ll r e s id e n tia l m o rtg a g e s w e re c o m m e rc ia l b a n k s , $ 1 9 .6 b illio n ; m u
tu a l s a v in g s b a n k s , $ 2 1 .7 b illio n ; lif e in s u ra n c e c o m p a n ie s (n o n fa r m ),
$ 3 5 b illio n ; s a v in g a n d lo a n a s s o c ia tio n s , $ 4 6 b illio n ( in 1 9 5 8 ).70
T h is b re a k d o w n does s u g g e s t t h a t a t le a s t som e p a r t o f th e g a in s m a d e
b y th e m o rtg a g e s w ill b e o ffs e t b y losses s u ffe re d b y th e m a n d o th e rs
in s o fa r as th e y o w n s ta k e s in c o m m e rc ia l b a n k s , m u tu a l s a v in g s b a n k s ,
l i f e in s u ra n c e c o m p a n ie s , a n d th e lik e . I n g e n e ra l, th o u g h , th e re p ro b
a b ly is a tr a n s fe r o f in c o m e as p ric e s ris e fr o m h ig h in c o m e g ro u p s —
le n d e rs — to lo w in c o m e g ro u p s — b o rro w e rs — a n d e s p e c ia lly o n th e a ll
im p o r ta n t re s id e n tia l r e a l e s ta te .
S a v in g s a c c o u n ts , in c lu s iv e o f a c c o u n ts in s a v in g s a s s o c ia tio n s ,
m u tu a l s a v in g s b a n k s , c o m m e rc ia l b a n k s , p o s ta l s a v in g s a n d c r e d it
u n io n s , a n d lif e in s u ra n c e re s e rv e s , in c re a s e d as a p e rc e n ta g e o f n e w
re s id e n tia l c o n s tru c tio n fr o m 3 9 .7 p e rc e n t in 1 9 5 0 , th e y e a r o f th e o u t
b re a k o f th e K o re a n w a r , to as h ig h as 9 5 p e rc e n t in 1 9 5 7 , a n a v e ra g e
o f 7 4 p e rc e n t in 1 9 4 8 -5 7 , a n d a n e s tim a te d in c re a s e to 8 2 p e rc e n t in
1 9 6 1 -7 0 . T h e s e s a v in g s a c c o u n ts p ro v id e d $ 1 0 5 b illio n in 1 9 4 8 -5 7
a n d a n e s tim a te d $ 1 9 3 b illio n in 1 9 6 1 -7 0 .
T h e n o n fa r m re s id e n tia l m o rta g e d e b t, $ 1 2 0 .5 b illio n in 1 9 5 7 , is
e s tim a te d to ris e to $ 1 5 4 b illio n b y 1 9 6 0 , a n d $ 2 9 5 b illio n b y 1 9 7 0 .
70Federal Reserve Bulletin, September 1959, pp. 1179-1181.
81
THE INCIDENCE OF INFLATION
T h e in c re a s e o f th is m o rtg a g e d e b t as a p e rc e n ta g e o f in c re a s e in s a v
in g s a c c o u n ts a n d l i f e in s u ra n c e re s e rv e s v a r ie d fr o m 5 6 p e rc e n t in
1 9 5 7 to 1 5 0 p e rc e n t in 1 9 5 0 , a n d a v e ra g e d 81 p e rc e n t in 1 9 4 8 -5 7 , a n d
a n e s tim a te d a v e ra g e o f 7 3 p e rc e n t in 1 9 6 1 -7 0 . T h e in c re a s e o f n o n fa r m re s id e n tia l m o rtg a g e d e b t in r e la tio n to to ta l s a v in g s v a rie d
fr o m 3 6 p e rc e n t in 1 9 4 8 to 5 6 p e rc e n t in 1 9 5 7 , 58 p e rc e n t in 1 9 6 0 , e s ti
m a te d , a n d 6 5 , e s tim a te d , in 1 9 7 0 .71 I t is c le a r th a t m o rtg a g e s o n n o n r e s id e n tia l c o n s tru c tio n a b s o rb a la r g e p a r t o f to ta l s a v in g s .
B u t w h o a re th e p e o p le w h o g e t in to d e b t o n in s ta llm e n t p a y m e n ts
o f v a rio u s k in d s ? T h e v e ry lo w in c o m e , s a y , th o s e w it h in c o m e s o f
less th a n $ 1 ,0 0 0 in 1 95 7 a n d 1 9 5 8 in c u rre d l i t t l e in d e b te d
ness. W h e re a s 5 2 p e rc e n t o f a ll in c o m e s d id n o t h a v e in s ta llm e n t d e b t
p a y m e n ts to m a k e , 7 3 p e rc e n t o f th o s e w it h in c o m e s u n d e r $ 1 ,0 0 0 h a d
n o p a y m e n ts to m a k e . T h e p e rc e n ta g e d e c lin e d s te a d ily to 3 5 p e r
c e n t f o r th o s e w it h in c o m e s o f $ 6 ,0 0 0 to $ 7 ,4 9 9 a n d th e n ro s e to 58
p e rc e n t f o r th o s e w it h in c o m e s o f $ 1 0 ,0 0 0 o r m o re . T h e m o d a l r a te
o f in s ta llm e n t d e b t p a y m e n ts as p e rc e n ta g e o f d is p o s a b le in c o m e w a s
1 0 to 19 p e rc e iit f o r th o s e w h o h a d p a y m e n ts to m a k e . T h e p e rc e n ta g e
f o r th o s e m a k in g 1 0 to 19 p e rc e n t p a y m e n ts to th e ir d is p o s a b le in c o m e
w a s 1 7 f o r a ll in c o m e s , 7 p e rc e n t t o r th o s e w ith in c o m e s u n d e r $ 1 ,0 0 0
a n d th e n r is in g g r a d u a lly to 2 8 p e rc e n t f o r th o s e w it h in c o m e s o f
$ 6 ,0 0 0 to $ 7 ,4 9 9 a n d d e c lin in g to 1 2 p e rc e n t f o r th o s e w ith $ 1 0 ,0 0 0
o r m o re .
I n g e n e ra l, i t m a y b e s a id th a t th e v e r y lo w in c o m e s d o n o t p r o fit
as m u c h fr o m th e b e n e fic ia l e ffe c ts o f in fla tio n o n th e ir d e b t p a y m e n ts
as d o th e s o m e w h a t h ig h e r in c o m e s . F o r e x a m p le , th e m o d a l fig u re
fo r r a tio o f r e g u la r p a y m e n ts to d is p o s a b le in c o m e is 3 3 p e rc e n t," a n d
th e p a y m e n ts w e re 2 0 to 3 9 p e rc e n t o f d is p o s a b le in c o m e . T h e p e r
c e n ta g e s b e g in n in g a t $ 1 ,0 0 0 in c o m e a n d r is in g b y $ 1 ,0 0 0 s tep s to
$ 5 ,0 0 0 -$ 5 ,9 9 9 a re 1 0 , 2 4 , 2 4 , 3 2 , 3 9 , 4 3 . T h e p e rc e n ta g e f o r $ 6 ,0 0 0 $ 7 ,4 9 9 is 4 5 , f o r $ 7 ,5 0 0 -$ 9 ,9 9 9 is 4 4 , a n d f o r $ 1 0 ,0 0 0 a n d o v e r is 2 8 .
H e n c e , th o u g h o n th e w h o le lo w in c o m e s g a in , th e la rg e s t g a in s d o n o t
g o to th o s e w it h th e lo w e s t in c o m e s .72 T h e m id d le in c o m e g ro u p s g a in
m o s t r e la tiv e ly fr o m th e e ro s io n o f th e a d v a n c e d d o lla r .
I t is a ls o c le a r th a t o n th e w h o le th o s e th a t g a in p a r tic u la r ly fr o m
G o v e rn m e n t-g u a ra n te e d m o rtg a g e s a re n o t th e lo w e s t in c o m e g ro u p s .
F o r e x a m p le , th e ta b le b e lo w s u g g e sts t h is :
T able
0-1.— Characteristics of 1-family new home transactions, FHA, sec. 203,
1952 and19571
Median
Property value_____________________________ ______________ ________
Calculated area........................................................................... square feet..
Number of rooms__________________________________________ ___ ____
Annual income of mortgagor______ _____ ____________________________
Annual housing expense______________________ _____________________
Expense-income ratio....................................................................... percent ..
1962
$10,022
923
6.3
$4,811
$988
19.6
1957
$14,261
1,105
6.8
$6,632
$1,382
19.7
i Bata from Federal Housing Administration, preliminary draft, 24th annual report, table 37. The
expense-income ratio was computed from the averages, rather than the published medians.
Source: Subcommittee on Housing, Committee on Banking and Currency, U.S. Senate, “ Study of
Mortgage Credit," 1968, p. 76.
n Committee on Banking and Currency, Subcommittee on Housing, “ U.S. Senate Study
of Mortgage Credit,” 1958, p. 18S.
« Ibid., p. 69.
82
THE INCIDENCE OF INFLATION
In 1957, the average property value guaranteed under F H A , section
203, was $14,261. The rise in value from 1952 was 43 percent. In
this same period per capita income rose only 17.4 percent, and it will
be noted that the annual income of the mortgagee rose by 88 percent,
suggesting that these homes had to be built and mortgaged by higher
income groups. The expense-income ratio did not change, but this is
explained partly by the appeal to higher income groups. The annual
housing expense, however, increased by about 38 percent.73 The in
crease is for family and unattached individual’s income, average mean
after taxes.
How much does one gain from mortgaging a home? It should be
noted that a 30-year $14,000 mortgage of 5% percent interest costs
$157 per month in the Northeast. O f the $157, only $82, or roughly
one-half, are for interest and amortization. It is assumed that the
other items, for example, heat, insurance, will increase with the rise in
prices. It can also be assumed that one-half of these payments will be
made by the 15th year and that the average payment would be made
in the 15th year. By this time, we might expect inflation o f perhaps
about one-third, and hence a reduction of costs on that account of
about one-quarter, and also a corresponding rise o f real income, which
would also cut the burden substantially. I f we allow for the increase
in per capita income due to rising productivity and an inflation o f
about 33 percent in 15 years, and if we assume that the gains to the
mortgagee from rising income would be limited to one-half o f the
total payments—others respond to rising prices and incomes—then
roughly the net gain for a 30-year mortgage would be about 20 percent;
that is, the cost in terms of stable dollars would be roughly 80 percent
as much as when the contract was made.74
That the low-income groups are largely excluded from these Federal
guarantees is explained in part by the high rate o f interest as well
as the period of amortization. For example, it is estimated that a
$14,000 mortgage requires a gross annual income o f $7,536, if the loan
is for 30 years and the rate of interest is 5% percent. But with a 2percent rate and a 35-year mortgage, the required income would be
$5,808.75
Much concern has been expressed on the issue whether the large
numbers o f mortgagees with heavy mortgages on property will be able
to meet their obligations. Since prewar there has been a steady in
crease in the average term o f years of mortgages as well as an increase
in the loan value in relation to the value of the property. The ten
dency to give a large proportion of value in mortgage and to extend
the term has resulted in a payment in relation to income that has not
changed greatly. For example, here are some figures from 1948 to
1956.
78 “ Study o f Mortgage Credit/' p. 76; “ Survey of Current Business,” April 1959, p. 10:
“ Annual Economic Report of the President, 1959.”
83
THE INCIDENCE OP INFLATION
T ab le
9-2.— Characteristics of FHA-insured and VA-guaranteed mortgage loans
on new houses, 1948 and 1956
1948
Average term, years
_______________________________________________________
Average loan to value ratio, percent
_________________________________________
Average payment in relation to income, percent_________________________________
Mortgagees median effective income (1954 dollars) ______________________________
Percent of loans with no downpayment_________________________________________
Percent of loans with 30*year terms_____________________ - _____________________
21
81.0
16.1
4,469
19.0
12.6
1956
25.5
86.7
16.1
5,921
31.9
2 53.2
1 1950.
2 1958.
Source: Ibid., p. 247.
C h a p te r 10 . O t h e r S h a res, W a g es a n d P r ic e s
in t r o d u c t o r y : t h e cau ses op i n f l a t i o n
In the last few years there has been much discussion of a new kind
o f inflation, namely, what is generally called a cost-push inflation.
This means that costs rise, and the rise o f costs in turn brings about
an increase in prices. Much of the emphasis, of course, is put on the
great increase in wage rates which in turn induces higher prices. In
sofar as there is support to the general idea that the recent inflation
i.e., 1955 to 1958, has been a wage inflation or cost-push inflation, to
that extent it might be argued that labor does not lose much from rising
prices. It leads and, therefore, gains at the expense o f the rest of
the community until other shares rise with the increase in prices.
For example, Gardner Means, distinguishes between a classical type
o f inflation with too much money chasing around too few goods, the
kind o f inflation that monetary policy has generally treated, and the
inflation which comes with a rise of market-dominated prices in the
recovery from a recession or depression. He holds that the latter is
a healthy kind o f inflation and is a normal and a necessary part of
economic recovery.
The third type of inflation has been called an administrative inflation, which
involves a rise in prices in the more concentrated industries where there is a
considerable area of discretion within which price and policy can be made.76
It is in the period since 1953 that Gardner Means finds the evidence of
administrative inflation. This type o f inflation comes, of course,
where there is a heavy concentration of industries such as in steel and
automobiles. For example, Means finds that in the steel industry,
prices from 1953 to October 1958 rose by 36 percent, whereas all whole
sale prices rose by only 8 percent.77 According to one study from 1947
to the spring o f 1957, unit labor cost in the iron and steel industry
had increased 35 percent, but the steel industry had raised prices by
96 percent, or two to three times the increase in labor cost. The cost
per unit o f other materials and services purchased by United States
Steel, aside from some o f the direct labor costs, increased 37 percent
in a period o f 10 years. Obviously, prices rose considerably more than
was justified by the rise of wages or even of costs.78
w G. C. Means, “ Administrative Inflation and Public Policy,” 1959, pp. 4, 5.
w Ibid., p. 16.
78 See Senate Committee on Finance, “ Investigation of the Financial Condition of the
United States,” pt. 7,1959, p. 2140.
THE INCIDENCE OF INFLATION
Economists are not all in agreement that the recent inflation has
been a cost-push or a wage inflation. But, certainly, this view has
gained a considerable number of adherents in recent years. Just to
give some indication of varying views, I note the following. For
example, Prof. Gardner Ackley writes:
* * * Our postwar inflation is not basically the direct result of excessive de
mand. Bather, it represents a process that flourishes under conditions of high
demand—but that can and does continue to operate even in the face of some,
perhaps considerable, deficiency of demand.
On the analytical side, I criticize the usual view that inflation must result
from excess demands or from a cost-push. It seems to me that inflations in
our postwar economy can be understood primarily as a process of jockeying
relative positions between labor and capital. In fact, the two groups extend
claims that add up to more than the total GNP—inconsistent claims that can be
resolved only by inflation.
* * * Wage setting can be to some considerable extent independent of supply
and demand forces: Wage rates do not rise only when demand exceeds supply
and rapidly fall in a reverse situation * * *.TO
Prof. Albert Rees had this to say:
There is no firm evidence that unions are a cause of inflation, and there is a
good deal of evidence that in rapid inflations, wages set by collective bargaining
lag behind other wages. The view that gradual inflation results from a “wage
push” is based on casual observation, which can be highly misleading. Much
further study is needed before it can be accepted as a basis for public policy.80
Professor Weston has this to say : 81
T able
10-1.— Behavior of selected economic variables for manufacturing
industries, 1949-56
(1)
Industry
Lumber and wood products___ _____
Furniture and fixtures______________
Stone, clay, and glass_______________
Primary metals------------------------------Fabricated metals__________________
Machinery (not electrical)_________
Electrical machinery_______________
Transportation equipment_____
Miscellaneous manufacturing..............
Food and kindred products_________
Tobacco manufactures--------------------Textile mill products--------- ---------Apparel______ ___ _____ . . _______ Paper and allied products___________
Chemical and allied products _ ____
Products of petroleum and coal... ....
Rubber products___________________
Leather and leather products_______
(2)
(3)
(4)
(5)
(6)
Percent in
Percent in* Percent in
creases, net
Average
Percent in
crease
creases
profits to sales,
level of
crease in
in hourly
in prices
1949-56
concentra
production, earnings,
tion of em
(1947-49),
1949-56
1956
1949-56
ployment,
Before After
1950
taxes
taxes
35.2
28.4
56.4
25.5
39.2
51.5
111.2
89.5
46.9
15.3
4.9
6.1
12.0
62.2
73. 5
35. 6
38.5
8.3
41.0
35.5
43.4
47.2
40.4
43.1
36.8
38.0
38.1
44.2
49.5
21.2
26.5
43.3
44.8
41.9
39.9
31.6
25.4
24.8
26.9
53.4
32.7
41.7
37.6
29.1
-9 .0
6.1
15.5
-12.1
-.3
27.2
7.2
17.3
45.8
-.7
-27.2
11.9
12.2
28.8
-10.3
2.8
-16.1
-20.2
8.1
-9 .1
50.7
-24.3
43.2
13.2
15.4
17.5
50.9
10.3
-33.9
3.0
-4 .7
13.6
-21.6
-15.6
-33.3
-35.7
0
-27.3
16.3
-36.6
23.8
-6 .2
-2 .4
20.2
15.8
-4 .5
27.7
25.3
50.2
53.8
39.9
41.6
59.6
60.4
38.4
42.3
57.3
31.5
-.3
33.5
51.6
40.3
75.0
24.2
Sources: For detailed sources, see original, and footnote 81.
78 Hearings, “ Relationship of Prices to Economic Stability and Growth,” Joint Economic
Committee, 1958, p. 394.
80 Ibid., p. 401.
81!,The Joint Economic Committee, “ Relationship of Prices to Economic Stability and
Growth,” compendium of papers submitted by panelists, 19*58, p. 317.
THE INCIDENCE OF INFLATION
The above table illustrates some of the problems of the microeconomic ap
proach. With the data, matrices may be made of the influence on prices, wages,
or profits or changes in production at different levels of concentration. These
matrices show that, for increases in production, price rises would be larger, but
not uniformly so. For the same increases in production, larger increases in
hourly earnings result in larger price rises, but not in a consistent fashion.
Profit changes are most directly connected with increases in output, but nu
merous exceptions are found. Increases in production have a greater influence
on price increases than the average level of concentration in manufacturing
industries. Small production increases are associated with lower levels of con
centration and large production increases are associated with higher levels of
concentration. This explains why the incautious or determined will find with
selected data an association between levels of concentration and price level
changes.
Also impressive is the clustering of wage increases during the period at the
40- to 45-percent level, except for industries where production (sales) increases
have been small. Where production increases have been lower, wage rises have
also been smaller.
These data show that the many simple generalizations explaining recent price
changes are subject to considerable reservation. Cost push does not explain
price increases, because hourly wage increases and price increases have been
smaller where production increases have been smaller. Administered prices
do not explain price increases, since price increases are small even among con
centrated industries where production increases have been small—for example,
tobacco manufacturers.
In a study o f rate o f increase and straight time hourly earnings
and changes in employment in durable and nondurable manufactur
ing, for the years 1947- 57, Professor Eckstein finds a rise in employ
ment in durables by 27 percent and in nondurables virtually no change.
Yet the increase in wages was very similar. This suggests that the
increase in wages in one segment of the economy tends to spread to
other industries, even where the pressure on employers is not nearly
so great. As others have noted, also, the increase according to Eck
stein was larger in good times than in bad. Wages tend to rise more
when employment rises and rise less when employment increases little
or is shrinking. Eckstein also finds that wages tend to move up with
consumer prices. The explanation o f this, o f course, may be that the
rise o f wages brings about higher consumer prices, and higher profit
margins are also a stimulus to wage increases. Eckstein also notes a
higher association between increases in productivity and a rise of
wages, a rise associated with increasing investment in plant and equip
ment. Variations in productivity are very large, varying from 9.7
percent in 1950 to 0.4 m 1956 over a period of 10 years from 1948 to
1957. Obviously, wage movements do not move closely with pro
ductivity. Therefore, if there is a large rise of productivity, the net
effect, despite a wage increase, may very well be declining prices; but
where wages rise, despite small increases in productivity, prices tend
to rise. When productivity rises by 9 or 10 percent, wage rates will
not rise by a corresponding amount; but, when it increases by only
less than one-half o f 1 percent, wage rates will tend to continue to
increase, though not as much as in the earlier years.82
#According to Dr. Hickman of the Brookings Institution, the large
rise o f demand for durable goods was especially accompanied by large
increases in wages and prices. This held in the early postwar period
and also in the years 1956- 58. These wage increases were then
spread, with a lag, to other industries. Hickman also claims that
82 Compendium, op. cit., pp. 364-369.
48084— 59------- 7
THE INCIDENCE OF INFLATION
86
although the rise of wages was roughly similar in durable and non
durable goods industries, the increase in prices of materials was not.
This may account for most of the contrasting amounts o f average price
increase* Hickman also claims that since labor lost some of its real
income in the early postwar period when wages did not respond ade
quately to rising costs in the later period—that is, from 1945 to 1957—
labor not only tried to make up for the losses in the earlier period
but tried to increase its real income. He also claims that demand
pressures were not nearly so great in 1956-57 as in the early postwar.
And, rather than call the pressures a cost push, he would categorize
them as bottlenecks, for then the connation would be increased demand
as well as increased costs in sectors where prices were rising rapidly .83
RELATION OF PRICE MOVEMENTS AND OTHER VARIABLES
I examined the relevant variables to see if I could find any corre
lation, for example, between price rises and the distribution o f manu
facturing income. For the yearly aggregates I found no correlation,
nor could I find in manufacturing any correlation between ( 1) com
pensation o f employees as a percentage o f manufacturing income,
( 2 ) corporate profits after taxes and unincorporated income as a
percentage of income, and ( 3 ) corporate profits before taxes and
unincorporated income as a percent o f income in the years 1947-57 as
functions o f the manufacturing wholesale price level. Nor is there
any clear correlation between the percentage change from year to year
for the compensation o f employees and price changes. Profits seem
to be more clearly correlated with production level than they do
with price changes, nor do we find any clear relationship in manufac
turing between the excess of the percentage change in employee
compensation over that of production and the percentage change in
the wholesale price index for manufacturing. I also examined the
percentage change from year to year of the excess o f change in em
ployee compensation over that of production and the percentage
change o f the wholesale price index for the years 1947- 58. No close
relationship is revealed, although, as might be expected, there
is a very rough coincidence. W e can conclude that yearly data do
not indicate whether profits or wages have gained as a result o f price
increases. Perhaps one explanation of this fact is the cyclical nature
o f profits and the steady upward trend of wages and prices. In big
boom years like 1947, 1950, and 1955, prices rise and profits gain
relatively to wages. In more normal years like 1956, prices rise and
profits decline relatively to wages.
I have also tried to compare for eight major industrial sectors of
the economy the movements of compensation o f employees—that is,
wages, profits after taxes, and unincorporated income, profits before
taxes and unincorporated income, and net interest for various periods
from 1929 to 1958. This table is summarized below, and it suggests
variations in the experience for different parts o f the economy and
also in different periods. For example, wages seem to have" done
very well when one compares 1929 and 1955- 58, but profits seemed
to do relatively well from 1938 to 1941 and from 1942 to 1945, though
88 Hearings, Joint Economic Committee, “Relationship of Prices to Economic Stability
and Growth,” pp. 93-9i5.
87
THE INCIDENCE OF INFLATION
in general, as we know, wages gained more than profits, taking all
these periods together. But at least in one segment o f the economy—
for example, finance, insurance, and real estate—on the whole profits
and interest have gained more than wages during most periods. This
is clearly not true in manufacturing.
T a ble
10-2.— Major gains by factors, various periods
Manufacturing ___________________
Communications and public utilities—
Contract construction________________
Mining....................................................
Services ___________________________
Finance, insurance, and real estate.......
Transportation.......... . ...........................
Wholesale and retail trade
________
Manufacturing _____________________
Communications and public utilities. ..
Contract construction.............................
Mining_____________________________
Services____________________________
Finance, insurance, and real estate.......
Transportation........................................
Wholesale and retail trade____________
1938-41 to
1942-45
1942-45 to
1946-49
1946-49 to
1950-54
Wages_______
....... do..............
....... do_______
____do..............
Profits_______
____do_______
....... do_______
___ do...............
Profits_______
Wages.............
Profits_______
....... do_______
Interest1........
Wages.............
....... do_______
....... do..............
Wages_______
Profits............
Wages.............
____do..............
Interest1___
Profits............
------do_______
Wages_______
1938-41 to
1955-58
1942-45 to
1955-58
1946-49 to
1955-58
Wages_______
....... do.............
....... do..............
____do..............
Interest1____
Profits............
____do..............
Wages_______
Wages_______
....... do..............
------do..............
Profits............
Interest1_____
Wages.............
------ do..............
....... do..............
1950-54 to
1955-58
Wages.
Profits.
Wages.
Do.
Interest.!
Do.i
Wages.
Do.
1929 to
1955-58
Wages_______ Wages.
Profits............
Do.
Do.
Wages.............
____do..............
Do.
Interest1_____
Do.
Wages.............
Do.
....... do..............
Do.
....... do..............
Do.
i Although interest gained relative to the other 2 groups, wages gained relative to profits.
Source: Original figures, Department of Labor and Department of Commerce.
It is clear that wages, much more so than in the past, tend to lead in
periods o f rising prices. This is clear even, for example, by comparing
the early postwar and the period since 1955. I consider also what
happened in some of our recent major wars. In the Civil War, aver
age hourly earnings rose one-third as much as the cost of living; in
W orld War I, 1.4 times as much as the cost of living; and in World
War II, they rose almost twice as much. These figures do suggest a
trend toward a lead in wages over prices. But we must also allow for
the fact that in W orld W ar I I there was a considerable control of
prices and, therefore, with the shortage of goods and price control, the
large rise in wages vis-a-vis the increase in the cost of living may be
misleading. To some extent the small rise in prices against the large
increase in wages could suggest that the purchasing power of each
dollar o f wages had been cut because o f the unavailability o f supplies.
THE INCIDENCE OF INFLATION
88
SOME HISTORICAL EVIDENCE
I n h is im p o r ta n t s tu d y o f r e a l w a g e s 84 D o u g la s c o n s id e rs th e ris e
o f m o n e y a n d r e a l w a g e s fr o m 1 8 8 9 to 1 9 2 3 . F o r e x a m p le , h e fin d s :
[1890-99*100]
Relative
production
per wage
earner
1899................................................................................................
1909................................................................................................
1919................................................................................................
1923................................................................................................
Real average
Average
annual
annual
earnings per
earnings
wage earner
100
111
113
152
99
102
ill
128
100
123
276
299
I t w ill b e n o te d th a t a c c o rd in g to D o u g la s ’ fig u re s , r e a l w a g e s d id
n o t ris e as m u c h as th e r e la tiv e p ro d u c tio n p e r w a g e e a rn e r. D o u g la s
e x p la in e d th is in p a r t b y th e d e c lin e in th e p u rc h a s in g p o w e r o f m a n
u fa c tu re s .
A t th e tim e th a t S e n a to r (th e n P ro fe s s o r) D o u g la s w ro te , th e s e fig
u re s m a rk e d a g r e a t a d v a n c e in o u r k n o w le d g e o r m o v e m e n ts o f r e a l
w a g e ra te s . O f c o u rs e s in c e th e n w e h a v e h a d m o re s o p h is tic a te d
a p p ro a c h e s to th is p ro b le m t h a t ta k e in to a c c o u n t m a n -h o u r o u tp u t as
w e ll as th e in p u t o f la b o r a n d c a p ita l. A t a n y r a te , a c c o rd in g to
D o u g la s ’ fig u re s , w a g e s ro s e m o re th a n th e c o s t o f liv in g d u r in g th is
p e rio d w h ic h o n th e w h o le w a s a n in fla tio n a r y p e rio d , b u t n o t as m u c h
as m ig h t b e s u g g e s te d b y th e ris e o f p ro d u c tio n . H e r e a re th e m o v e
m e n ts in th e c o s t o f liv in g a n d r e a l w a g e s in a n u m b e r o f in d u s trie s
o v e r s e v e ra l y e a rs fr o m th e D o u g la s s tu d y :
T able
10-3.— Cost of living and real wages
[1890-99*100]
Cost of
living
1890.....................
1909.....................
1919.....................
1923.....................
104
121
247
234
Clerical and
lower salaried
Ministers Teachers Postal
Manufac Transpor workers:
employ
Transporta
turing
tation
ees
tion and man
ufacturing
101
102
111
128
98
97
109
122
88
99
80
95
99
88
64
89
87
137
118
199
95
95
68
86
All
groups
100
104
109
127
Source: Paul H. Douglas, op. cit., pp. 33,22,23,24.
I n g e n e ra l, th e s e s ta tis tic s d o s u g g e s t th e v a r y in g ra te s o f re s p o n s e
to r is in g p ric e s . C le r ic a l w o rk e rs , f o r e x a m p le , a c tu a lly e x p e rie n c e d
s u b s ta n tia l re d u c tio n s in t h e ir r e a l in c o m e w h ile th e p r o d u c tiv ity o f
th e N a tio n w a s r is in g . T h e s am e c a n b e s a id f o r m in is te rs a n d F e d
e r a l G o v e rn m e n t e m p lo y e e s w h o s u ffe re d s e v e re re d u c tio n s . T h e m o s t
s tr ik in g ris e w a s f o r te a c h e rs .
U n d o u b te d ly th e e x p la n a tio n in p a r t is th e v a r y in g ra te s o f d e m a n d
f o r th e p ro d u c ts o f e a c h g ro u p . T h is w a s a p e rio d o f la r g e in c re a s e s
** See Paul H. Douglas, “ The Movement of Real Wages and Its Economic Significance,”
American Economic Review, Supplement, March 1926.
THE INCIDENCE OF INFLATION
89
in e d u c a tio n , w it h th e re s u lt th a t th e re w a s c o n s id e ra b le b id d in g u p
o f th e p ric e s p a id to te a c h e rs . A t th is tim e S ta te a n d lo c a l g o v e rn
m e n ts d id n o t h a v e the. p ro b le m s o f fin a n c e t h a t th e y h a v e e x p e rie n c e d
in m o re re c e n t y e a rs . T h e u n a v a ila b ility o f re s o u rc e s f o r m in is te rs
e x p la in s in n o s m a ll p a r t th e r a th e r u n s a tis fa c to ry s ta te o f t h e ir r e a l
w a g e s . T h e F e d e r a l G o v e rn m e n t a ls o w a s s lo w in a d ju s tin g w a g e s
o f t h e ir e m p lo y e e s . I n g e n e ra l th e w h ite -c o lla r w o rk e rs , w it h th e e x
c e p tio n o f th e te a c h e rs , seem ed to s u ffe r d u r in g th e s e y e a rs o f in fla tio n .
A n o th e r in te r e s tin g p o in t is t h a t a la r g e p a r t o f th e im p ro v e m e n t
in r e a l in c o m e o c c u rre d fr o m 1 9 2 0 o n , a n d d u r in g W o r ld W a r I th e re
w a s a ls o a s u b s ta n tia l im p ro v e m e n t. I t w a s p a r tic u la r ly s in c e 1 9 1 3
th a t th e la r g e a d v a n c e s in r e a l e a rn in g s wre re m a d e b y th e g ro u p as
a w h o le , th o u g h e v e n d u r in g th is p e rio d som e g ro u p s a c tu a lly e x p e r i
e n c e d a d e c lin e in th e ir r e a l e a rn in g s .
TEACHERS
I n a m o re re c e n t p e rio d te a c h e rs h a v e te n d e d to e x p e rie n c e s e rio u s
losses in p e rio d s o f in fla tio n . T h e e x p la n a tio n o f th is in n o s m a ll
Ea r t is th e fa c t th a t lo c a l g o v e rn m e n ts a n d e v e n S ta te g o v e rn m e n ts
a v e e x p e rie n c e d s e rio u s fin a n c ia l p ro b le m s . I t is o f in te re s t t h a t
a ro u n d 1 9 0 0 c ity re v e n u e w a s a b o u t fiv e tim e s as la r g e a s S ta te r e v
e n u e . T h e c u r r e n t s itu a tio n is c o n s id e ra b ly d iffe r e n t w it h b o th S ta te
a n d lo c a l g o v e rn m e n ts o b ta in in g ro u g h ly e q u a l re v e n u e s . T h is te n d
e n c y o f lo c a l re v e n u e to b e co m e m u c h less im p o r ta n t r e la tiv e ly is
e x p la in e d in p a r t b y th e d e p e n d e n c e o n th e g e n e ra l p r o p e r ty ta x ,
w h ic h re s p o n d s t a r d ily to r is in g p ric e s a n d in c o m e . I n h is b o o k ,
P a u l D o u g la s s h o w e d t h a t fr o m 1 8 9 0 to 1 9 2 6 p u b lic s c h o o l te a c h e rs ’
r e a l w a g e s ro s e b y 7 2 p e rc e n t as a g a in s t a ris e o f 1 7 p e rc e n t f o r
fa c to r y w o rk e rs a n d a lo ss o f 1 p e rc e n t f o r m in is te rs . F o r a n n u a l
w a g e s ( r e a l) th e in c re a s e w a s 1 1 4 p e rc e n t f o r p u b lic s c h o o l te a c h e rs ;
a ll in d u s trie s , e x c lu d in g fa rm e rs , 31 p e rc e n t; m in is te rs , 1 1 p e rc e n t.85
B y 1 9 2 6 th e te a c h e r w a s a ls o a s o m e w h a t d iffe r e n t c o m m o d ity th a n
in 1 8 9 0 , h a v in g h a d m o re a n d b e tte r tr a in in g , a n d h e w a s p ro b a b ly
d o in g a b e tte r jo b .
I n th e in fla tio n d u r in g W o r ld W a r I I te a c h e rs ’ s a la rie s la g g e d
b e h in d th e g a in s o f th e re s t o f s o c ie ty . F o r e x a m p le , in 1 9 3 8 -3 9
s a la rie s b e fo re ta x e s w e re $ 1 ,4 0 8 f o r te a c h e rs . B y th e e n d o f th e w a r
th e y h a d ris e n to a b o u t $ 2 ,0 0 0 , b u t in 1 9 3 5 -3 9 d o lla r s th e d e c lin e h a d
b e e n fr o m $ 1 ,4 1 6 to $ 1 ,2 7 0 in 1 9 4 2 -4 3 , a n d a n e s tim a te d in c re a s e to
$ 1 ,4 5 7 b y 1 9 4 5 -4 6 .86 B y 1 9 4 6 -4 7 th e te a c h e rs in th e s m a lle r c itie s h a d
o b ta in e d a n in c re a s e o f 4 8 p e rc e n t as a g a in s t th e ris e in th e cost o f
liv in g o f 4 8 p e rc e n t. F r o m 1 9 3 6 -3 7 to 1 9 4 6 -4 7 s a la rie s o f te a c h e rs
in th e la r g e r c itie s s t ill la g g e d b e h in d th e co st o f liv in g .87
Teachers o f course were most unhappy even when their pay rose
as much as the rise in the cost o f living. Whereas other groups had
gained substantially, the average teacher had experienced a substantial
loss in real income at first, though by the end of the war had rough
ly regained their prewar position in real dollars.
85 P. H. Douglas, “Real Wages in the United States, 1890-1926,” 1030, pp. 130, 201, 203,
382, 386, 39il<.
80National Education Association Journal, December 1946.
MNEA Research Bulletin, February 1947.
90
THE INCIDENCE OF INFLATION
But there were large differences among different groups o f teachers
and particularly the high-priced teachers experienced large relative
losses. In general, the increase was largest in the smallest schools
where pay had been very low. For example, in the State of New York
in the common one-teacher school the rise from 1939 to 1946 was 45
percent; in cities of population o f 100,000 and over, the increase was
only 13.6.88 In New York, representing the richest State, the increase
in the war period was roughly about 20 percent; whereas for the whole
country, the increase was about 40 percent.89
Gradually, under the pressure of increased enrollments and relative
ly good organizations, the public school teachers were able to increase
tneir pay after the war and to recoup a good part of their relative
losses. For example, the Office of Education writes:
*
* * When expressed in terms of 1955-56 dollars, however, the increases
are not as large as they appear * * *. For example, the average annual salary
for the continental United States in unadjusted (current) dollars rose from
$1,441 in 1939-40 to $4,156 in 1955-56, an increase of 188 percent. When ex
pressed in terms of 1955-56 dollars, the increase between 1939-40 and 195556 was only 50 percent. During the same 16-year period, the personal income per
member of the labor force (in 1955-56 dollars) increased 76 percent * * *.®°
But this is not exactly a fair comparison because personal income in
cludes not only income earned but also all other kinds o f income and,
therefore, the relative position of teachers is better than is indicated
in this excerpt.
On the whole, the colleges have suffered more than the schools.
This is partly due to the fact that the college faculties are not as
well organized, and furthermore there is a degree of rapport between
the schoolteachers and their local finance and tax authorities and the
parents who want their children to get a good school education. One
could, for example, go back to the days o f President Eliot when
he astounded the world by announcing that hereafter the professors
at Harvard would be paid $4,000. But the fact is that by 1959, when
the average pay was $16,000 for a professor and, therefore, there had
been an increase of three times in the average pay, the situation was
not as good as it seemed. Prices had risen by two times and, there
fore, there was not a large gain in real income. Moreover, the per
capita real income of the population had risen several times as much
as the income o f the average Harvard professor.
Even as late as 1955- 56, a full professor at major private uni
versities was receiving from 10 to 20 percent less in real dollars than
he had received before the war. Since the average worker in our so
ciety has improved his position by about 50 percent, this suggests at
that time the relative deterioration of at least 40 percent in the real
position o f the professors in top private institutions. Indeed, the
losses to other members o f the teaching fraternity were not quite so
large, though in general there were very few who by 1955-56 were
receiving larger real incomes than they had before the war. Indeed,
to some extent, this was made up by an increase in fringe benefits and
88New York State Education, November 1946, pp. 137-139.
89 See S. Rept. 323, “Federal Assistance to States in More Adequately Financing Public
Education,” 1943, p. 5.
80 U.S. Department of Health, Education, and Welfare, Office of Education, “ Statistics
of State School Systems, 1955-56; Organization, Staff, Pupils, and Finances,” 1959,
p. 22.
THE INCIDENCE OF INFLATION
91
also to some extent there was more outside work done by the average
faculty member. Relative to the whole population, and the college
teacher had suffered a serious reduction in his real income and part
o f the explanation, o f course, was the inflation to which the response
o f revenue had not been adequate. The public universities tend to
adjust to inflation through increased tax receipts, but slowly; and
the private institutions more recently have tended to invest heavily
in common stock as a protection against the rise of prices. But since
endowment income only yields about 6 percent o f total college income,
this is not a very important contribution. What is required is, of
course, large increases in appropriations by State legislatures and
much higher tuition fees. But the colleges are frightened o f large
increases in tuition fees because they fear that the net result would
be a change in the type o f students.
PRODUCTIVITY, PRICES, AND WAGES
In general, over our history, the increase in real wages is associated
with the rise o f productivity, more so than with a rise of prices. In
fact, there were long periods when prices did not contribute anything
to the rise o f the GNP. The increase in the GNP reflects rising pro
ductivity and increases in the population more than a rise in prices.
The following, from Dr. Raymond Goldsmith, discusses this issue
w ell : 91
The largest variations in the average rate of change are shown by the price
level, more specifically by the gross national product deflator which is a weighted
average of the prices of all final goods and services produced. For the 120 years
as a whole, prices have increased on the average at the rate of iy8 percent per
year, a rate which probably now would be regarded as within the range of price
rise characterizing a “ creeping inflation.” Price trends in the first half of the
period, when the average rate of change was virtually zero, differed considerably
from those observed during the second half starting in 1899, during which the
rise in prices averaged 2y2 percent per year, probably near the upper boundary
of what is thought to be compatible with a creeping inflation. However, if
subperiods of 40 years’ duration are taken, prices advanced most rapidly from
1879 to 1919 when the rise averaged 1.9 percent per year rather than in the last
40 years, during which the average rise amounted to only 1.4 percent.
The result of these variations in rates of increase of total gross national
product in current prices, in the price level, and in population is that the residual,
the rate of growth in real national product per head, shows more stability within
the range of iy 2 to 1% percent than any of the other three series. The contri
bution of the three factors—real output per head, population, and prices—to
the average rate of growth of aggregate current output thus has differed greatly
in the different periods.
For the entire 120 years population growth has accounted for two-fifths of the
total increase in the monetary value of aggregate output; the rise in the price
level for one-fourth; and intensive growth, the rise in real output per head,
for one-third. In some of the subperiods the change in the price level has con
tributed nothing to the increase in aggregate gross national product at current
prices, as for instance from 1839 to 1899; or has even offset part of the increase
in population and intensive growth, e.g., from 1869 to 1899. There is no period
during which the rise in the price level accounted for as much as one-half of the
rate of growth in total current aggregate output. During the last 40 years the
rise in the price level has been responsible for fully one-third of the rate of growth
of current aggregate output, while population growth has contributed threetenths and intensive growth almost two-fifths.
However wages respond to rising prices, the history o f our economy
in the last 68 years shows that real hourly earnings have tended to
81 Joint Economic Committee, hearings on “Employment, Growth and Price Levels.
Pt. 2 : Historical and Comparative Rates of Production, Productivity and Prices/' p. 272.
THE INCIDENCE OF INFLATION
92
rise substantially more than output per unit of labor and capital com
bined in the private domestic economy. The rise of real hourly earn
ings tended to be much larger from 1919 to 1957 relative to the increase
o f output per unit o f labor and capital combined than from 1889 to
1919.
T able
10-4.— Average rates of increase in productivity, total input per man-hour,
and real hourly earnings, 1889-1957
Average annual percentage rate of change
1889-1957
Output per unit of labor and capital combined, private
domestic economy____________________ _________________
Total input per man-hour, private domestic economy..... .
Real hourly earnings, private domestic economy, all workers
(including proprietors and family workers)_______________
Real hourly earnings, manufacturing, wage earners____ _____
1889-1919
1919-57
1. 7
.6
1.3
.7
2.1
.5
2.4
2.3
1.7
1.9
3.0
2.6
Source: Joint Economic Committee, hearings on “ Employment, Growth, and Price Levels.”
“ Historical and Comparative Bates of Production, Productivity, and Prices," p. 322.
Pt. 2,
A table that indicates the trends in recent years, at least from 1948
to 1956, is presented below.
10-5.— Indexes of labor and nonlabor payments per dollar of real product,
prices real product per man-hour, employee compensation per hour in current
and constant dollars, private nonagricultural sector of the economy, 1947-56
T able
[1947=100]
1948
1. Private nonagricultural product (current
dollars)_______________________ _____ _
Employee compensation (current dollars)
Wages and salaries (current dollars)_____
Nonlabor payments (current dollars)____
Private nonagricultural real product (195(5
constant prices)___ ______ ____________
G. Employee compensation per dollars of real
product_____________________________
7. Wages and salaries per dollars of real prod
uct_________________________________
8. Nonlabor payments per dollars of real
product_____________________________
9. Implicit price change—private nonagricul*
ture_________________________________
10. Man-hours of employees________________
11. Real product per employee-hour...............
12. Average hourly compensation___________
13. Average hourly wages and salaries.............
14. Consumer Price Index___ ______ _______
15. Average hourly compensation in constant
dollars_____ _________________________
16. Average hourly wages and salaries in con
stant dollars______ __________________
2.
3.
4.
5.
1949
1950
1951
1953
1954
1955 19561
110.9 111.7 124.7 141. 149.6 159.2 158.1 173.0 182.9
110.3 108.6 119.7 137. 5 147.6 158.9 157.4 170.7 183.5
110. 5 108. 5 118.7 135.8 145.9 156.9 154.9 167.7 180.3
111.7 115.8 131.0 147.3 152.3 159.7 159.0 176.0 182.1
104.1 103.8 114.4 121.9 125.8 131.6 128.’
106. 0 104. 6 104. 6 112.8 117.3 120.
139.4 143.4
122.3 122.
128.0
106.1 104. 5 103.
111.4 116.0 119.2 120.
107.3 111. 0 114.
120.8 121.1 121.4 123.5 126.3 127.0
106.5 107.7 108.9
101.4 96.8 101.0
102. 107. 2 113.3
108.8 112.1 118.5
109.0 112.1 117.5
107.6 106.6 107.6
116.3
106.6
114.4
129.0
127.4
116.2
119.0
108.3
116.2
136.3
134.7
118.8
120.9
110.9
118.7
143.3
141.5
119.
101.1 105.3 110.1 111.0 114.7 119.
101.3 105.
120.3 125.7
122.8 124.1 127.6
106.3 111.1 113.7
121.1 125.5 126.1
148.1 153.6 161.4
145.7 150. 158.6
120.2 119. 121.7
123.2 128.1 132.6
109.2 109.6 113.4 118.1 121.2 125.9 130.3
i Preliminary.
Source: Joint Economic Committee, “ Productivity, Prices, and Incomes/’ 1957, p. 143; for the sources
and comments, see the original.
This table reveals the follow ing:
1.
That employee compensation roughly corresponds to the dollar
rise in private nonagricultural product. Nonlabor payments roughly
rose just as much as wages and salaries. This suggests that the con
tribution o f wages and salaries to rising prices is no greater relatively
than the nonlabor payment, except that wages and salaries are a larger
THE INCIDENCE OF INFLATION
93
part o f the total payments. The real nongricultural product rose from
1948 to 1956 by 43 percent, but employee compensation per dollar of
real product rose by 28 percent. This, of course, suggests an inflation
ary effect o f employee compensation. Nonlabor payments per dollar
o f real product rose a little more than wages and salaries per dollar
o f real product. Where the real product per employee hour rose by
26 percent from 1948 to 1956, the averag;e hourly compensation rose
by 61 percent— another indication o f the inflationary effects of rising
wages. Average hourly compensation was up 61 percent; the Con
sumer Price Index up 22 percent—again suggesting a greater in
crease of wages than of prices. To some extent, of course, this is
explained by the increase in productivity, though there also is involved
some transfers from other groups to labor.
By 1951, the Consumer Price Index was up 16 percent, and average
hourly wages and salaries 27 percent; but by 1956 the relative in
creases were 22 and 59 percent; suggesting the great degree o f wage
inflation from 1951 to 1956. By 1956, the rise o f average hourly wages
and salaries over 1948 was close to three times as large as that in the
Consumer Price Index. It was noted before that in the first few years
after the war, prices rose sharply and wages did not keep up with
the increase in prices. Apparently one explanation o f this rapid
increase in unit property costs as against labor costs was the large
consumption of capital during this period per unit o f output.92
It is clear by now that labor costs tend to rise faster than the price
o f output though there are great variations from period to period. For
example, the staff of the Joint Economic Committee shows that the
ratio o f compensation of employees—
per unit of the GNP deflator rose from 1909 to 1915, fell until 1919, rose again
to 1921, and then remained fairly stable or declined until 1929. The ratio
rose sharply as prices fell from 1930 to 1932 and then fluctuated between the
1929 and 1932 ratio until 1941. In the war years, the ratio rose to a new peak
which was followed by a decline until 1950. The next 3 years were marked by
another rapid rise in the ratio which by 1956 exceeded the 1944—45 levels. In
other words, wages tend to rise relatively as real wages tend to increase in
periods of deflation and also in some periods of inflation. But it will be noted
that this was not true from 1915 to 1919 nor in the early post-World War II
period.®8
It is not easy to explain the changes in real hourly earnings in
manufacturing in recent years.94 O f the years 1945-46 to 1956- 57,
real hourly earnings in manufacturing declined in 1945-46 and
1946- 47. This reduction could be explained both by the inflation of
the period with the failure o f wages to rise as rapidly as prices, as
well as the fact that productivity experienced a substantial decline.
The substantial rise in 1948- 49, a recession year, of 4.8 percent in real
hourly earnings in manufacturing as compared to an increase of pro
ductivity o f 2.3 percent may again be explained in terms of the re
sponse o f wages to falling activity. In 1949- 50, wages did not rise
as much as the productivity increase o f 6.3 percent. Here again is a
substantial inflation with resultant failure of wages to rise as much as
productivity. In 1950- 51, the continued inflation was reflected in a
very small increase in real earnings—namely 0.5 percent. In most of
<* Ibid., p. 46.
w Compare, for example, “Employment Growth and Price Levels,” hearings, pt. 2, p. 329.
94
th e
in c id e n c e
of
in f l a t io n
the other years, up until 1956- 57, the increase in real hourly earnings
greatly exceeded the rise in output per unit o f labor and capital com
bined. These on the whole were prosperous years, with the exception
o f 1953- 54, and though productivity varied considerably, the general
trend o f real hourly earnings in manufacturing was distinctly upward.
Here wages and inflationary policies undoubtedly played a part.
In a table published by Professor Fabricant based on work by
Mr. Kendrick, o f the National Bureau o f Economic Research, the
author presents some interesting trends from 1899 to 1953 on output,
input, output per unit o f total input, real hourly earnings, and price
o f product. I have analyzed this table and have brought together
some excerpts from it in an attempt to look further into the relation
ship between real hourly earnings and output, productivity, and prices.
First, it is important to note the large differences among different
industries—for example, manufacturing industries; and all these fig
ures relate to 1899= 100, and the figures are for 1953.
The extremes in manufacturing output were a rise in lumber, 128
percent, and electrical machinery, 6,264.
The extremes in productivity, that is, output per unit o f input, were
also large, though the differences were not nearly so large as in the
total output.
Rubber, with output o f 878, had the largest rise o f productivity;
namely, 778 percent; and lumber, with 177, a rise o f 77 percent, had
the smallest.
As might be expected, the prices o f the product might reflect the
varying increases o f productivity. Rubber had declined in price dur
ing this period by 42 percent, whereas the price o f lumber had increased
by 961 percent. Actually, rubber had experienced an increase o f out
put o f 4,853 percent and a rise in productivity o f 778 percent. A small
increase in both output and o f productivity in lumber is reflected in a
relatively large rise o f prices, just as the great increase o f output and
productivity in rubber is reflected in a very small increase in prices,
or rather a net decline in prices.
In contrast to the large differences between the lowest and highest
industry in the census classification for manufacturing, for price o f
product and productivity, the extremes o f real hourly earnings in
manufacturing were only from a 123-percent rise for miscellaneous
manufacturing to a 305-percent rise for paper. O f course, these
figures do suggest that wages in each industry are not primarily deter
mined by productivity, for if they were the differences in real hourly
earnings, or rather in the increase in real hourly earnings, would be
much greater than is suggested by this table.
This table does suggest very large rises in real hourly earnings in
54 years. That means, o f course, that hourly earnings rose much more
than the price level. For the whole economy, there is a moderately
close relationship between the rise o f real hourly earnings and the
output per unit o f total input. But this relationship is anything but
close among industries. I list a number o f industries in the order of
their increase in output from 1899 to 1953. The differences in the
rise o f real hourly earnings are substantial, but nowhere near as large
as those in output per unit o f input. Nor is there any evident relation
ship between the trends o f prices of products and those o f real hourly
earnings by industries. The electric light and power industry expen
THE INCIDENCE OF INFLATION
95
enced the largest rise o f output and the greatest gains of productivity
and, as might be expected, a decline in prices since 1899. But the
hourly earnings rose relatively little, namely, 189 percent. The gains
o f productivity are distributed among all industries, with wage trends
not closely tied to the productivity gains o f each industry.
T a b le
10-6.— Output, real hourly earnings, output per unit of input, and prices,
1899-1953
[1899=100]
Name of industry
Electric light and power___________________________________
Electrical machinery______________________________________
Rubber products__________________________________________
Leather products____________________________ _____________
Lumber products_________________________________________
Anthracite coal____________________________________________
Output
Real
hourly
earnings
Output
per unit
of input
24,550
6,264
4,943
185
128
51
289
332
371
306
334
362
1,764
338
878
198
177
147
Prices
62
276
58
432
1,061
436
Source: Joint Economic Committee, hearings, ‘‘Employment, Growth, and Price Levels,” pt. 2, pp.
336-337.
The anthracite coal industry, which actually experienced a reduc
tion o f output by about one-half in a period of 54 years, is an espe
cially interesting case. Yet, and even though its productivity, i.e.,
output per unit of input, increased only by 47 percent, real hourly
earnings were up by 262 percent, a rather large increase, and prices
were up by 336 percent, a substantial rise as might be expected. These
figures suggest, in part, that a strong trade union, despite a declining
market and a rather slow rise of productivity, can achieve for its de
clining number of workers who can hold on to their jobs, a large in
crease in hourly earnings.
The extremes in real hourly earnings index from 1899 to 1953 were
as follows ( 1899= 100) :
Petroleum, coal products, manufactured_______________________________ __ 577
Mining, manufactured gas (public utilities)___________________________ __ 431
Mining, oil and gas__________________________________________________ __ 409
Manufactured paper_________________________________________________ __ 405
Electric light and power_____________________________________________ __ 289
Tobacco____________________________________________________________ __ 276
Farming____________________________________________________________ __ 247
Mining, metals______________________________________________________ __ 239
Beverages__________________________________________________________ __ 224
Miscellaneous manufacturing________________________________________ __ 223
Mining, nonmetals__________________________________________________ __ 158
An indication o f the lack of association between the rise o f pro
ductivity and movement in real wages is given by the following table.
10-7.— Ratio, rise of real hourly earnings to the increase of the
output to the total input, 1899-1953
Anthracite coal___________________________________________________
Lumber__________________________________________________________
Petroleum coal products manufacturing____________________________
Transportation equipment_________________________________________
Manufactured gas________________________________________________
Tobacco__________________________________________________________
Electric light and power__________________________________________
T ab le
1.
2.
3.
4.
5.
6.
7.
Source: Ibid. (my calculations).
total
5.58
3.04
1 .94
.59
.34
,34
. 11
96
' ’
th e
in c id e n c e
o f in fla t io n
As might be expected, especially where there has been relatively
small rises o f productivity, the gains of real earnings have been espe
cially large in relation to that in productivity. Where the gains of
productivity have been very large, the rise of earnings in relation to
gains o f productivity have been low. The variations in rise o f hourly
real earnings vis-a-vis that in productivity were as high as 50 to 1.
There seemed to be, from this small sample, almost an inverse rela
tionship, the greater the gain of productivity, the smaller the rise of
real earnings versus that in productivity.
TH E SHARE GOING TO LABOR
It is clear, from all kinds of statistics, that the rise o f wage rates has
exceeded the increase in man-hour output, and, moreover, the rise of
wage rates has exceeded the increase in prices. Hence, if, on the
average, productivity rises by 2 percent and prices rise a little more
than 1 percent, then we may well expect an annual increase in hourly
money earnings of about 3 percent, or an increase, say, o f real earn
ings o f 2 percent. Insofar as labor's share increases, the wage rise
would exceed the gains suggested by rising productivity and prices.
A rise o f wage rates in excess of that in productivity suggests that
wages lead the rise o f prices and to that extent contributed to inflation.
I f wages rise more than man-hour output and more than the per
centage rise of man-hour output and prices, it might be expected that
other shares in the national income might be squeezed. The evidence
is that over the last few decades, this has actually happened. For
example, Dean Bach writes:
Thus overall wage costs have risen somewhat more rapidly than selling price
with the result that profits have been squeezed. Indeed, wages throughout the
western industrial world seem to be increasingly mobile upward in many instances
linked to rising prices through built-in escalator clauses • *
In general, Dean Bach seems to believe that this trend will persist
so long as the Government continues to guarantee fiscal and monetary
policy that will make it possible for each group in society to demand
higher income payments in dollars.96
Professor Kendrick comes to somewhat similar conclusions. He
finds that between 1919 and 1953 the gross private domestic product
rose by 3.3 percent a year on the average in current prices in relation
to the physical volume o f resource inputs. The general level o f prod
uct prices rose only 1.2 percent, however, and this is explained by an
annual rise in the rate of total productivity o f 2.1 percent. But it
should be noted that the average hourly labor compensation in this
period from 1919 to 1953 rose by 3.8 percent a year on the average, as
compared with 1.9 percent for the annual increase in compensation
per unit o f capital input. The 1.8 percent increase largely reflects the
rise in the price of capital goods, including land. Because the stock
o f capital per worker increased greatly during this period, it was
possible for the average total compensation per man-hour, including
all fringe benefits, to go up approximately twice as fast as the price
o f capital. Despite the declining input of labor, the large relative
86 Joint Economic Committee, “ The Relationship of Prices to Economic Stability and
Growth.” compendium of papers, 1958, p. 37.
Ibid., pp. 35-37.
THE INCIDENCE OF INFLATION
97
increase in the price of labor made it possible for labor's share in the
private domestic national economy to rise from 72 percent in 1919 to
79 percent in 1953.97
It might be assumed that because wages rise more than prices and
more than is justified by the increase in both prices and productivity
that wages lead in the rise o f prices and are responsible for the inflation.
But aside from the problem of prices, which has been discussed briefly
before, there are other and related considerations about which we
should say a word. Dr. Ruth Mack, of the National Bureau o f Eco
nomic Research, has put some of these issues very effectively. She
shows, for example, that between 1947 and 1956-57 spot market prices
fell by 10 percent o f the 1947-49 average. Prices of crude materials
rose by 16 percent, and prices of all manufactured goods rose by 34
percent. Labor costs increased about 15 percent of the 1947-49 aver
age, or about as much as prices o f crude materials. She seems to
find a very vigorous response in prosperous periods to rising raw
material prices and to the prices of the finished product, and some
resistance to any fall in these prices when material prices tend to go
downward.
What then is the explanation of the divergent trends? It does not lie pri
marily, of course, in bulging profits. Rather, must it be found in the increasing
amounts of fabrication to which materials are submitted, increased marketing
costs, increased administrative costs, costs of research, of insurance, of develop
ment. These shifts in productivity and cost structures thrive in the general
atmosphere of the times. Many of the emphasized costs are of the overhead or
burden type. There is a widespread belief that the strong output trend and
demand is truly durable. This weakens usual fears of saddling a business with
heavy overhead type costs * * *.as
Other studies also show a tendency for employees to obtain a larger
share o f the national income. For example, the Joint Economic Com
mittee, in one of its studies, shows that compensation for employees
rose from 58 percent o f national income in 1929 to almost 70 percent
by 1956. Once allowances are made for the shifts in the relative
importance o f industries and legal forms of organization at different
labor cost ratios, the rise o f labor’s share is greatly reduced.
*
* * If one excludes the effects of such shifts and limits the comparison to
changes in the relative share going to labor (and implicitly to property) within
each of the component industries of the economy and different legal forms of
organizations (corporate versus noncorporate), then the 12 percentage points
increase in the labor share is reduced to less than 3 percentage points.
The Committee also finds that unit property costs rose about threefourths between 1909 and 1955, as compared with a tripling o f prices
and an increase in unit labor costs o f about 3% times the 1909 level.
Within property costs (before taxes)—
capital consumption per unit rose about 69 percent between 1929 and 1955, com
pared to 27 percent for profits and other property income per unit, and about
36 percent for total property costs per unit * *
97 Hearings, “The Relationship of Prices to Economic Stability and Growth,” Joint Eco
nomic Committee, 1958, p. 99.
08 See hearings, “Relationship of Prices and Economic Stability and Growth,” pp. 130131, and “Relationship of Prices and Economic Stability and Growth,” compendium, pp.
269—284. Also see R.P. Mack, “Inflation and Quasi-Elective Changes in Costs.” Review of
Economics and Statistics, 1959, pp. 225-231. In his contribution to the same hearings,
Professor Ruggles presented a somewhat similar position, particularly on the large rise of
overhead and service costs. See hearings, ibid., pp. 134-136.
09 “Productivity, Prices and Income,” 1957, pp. 10-11.
98
THE INCIDENCE OF INFLATION
INFLATION AND DEPRECIATION
One aspect of the redistribution o f income resulting from price
movements should receive a little more attention. This is the prob
lem that arises because of the fact that as prices rise, the depreciation
fund set aside to replace capital and equipment, tends to be inade
quate, and the cost of replacement of inventories increase. The setaside are related to a price level at the time of purchase; but as prices
rise, the resources available to replace the necessary machinery and
equipment, are not to be had. In that sense businessmen perhaps
believe their profits are higher than they in fact are.
Dean Bach has even stated that the underestimation o f depreciation
and inventory replacement costs, and the corresponding overstatement
o f profit costs, may induce- business to produce more than they other
wise would. According to B ach:
Partial estimates suggest that this understatement of replacement oasts may
have approached one-third of corporate reported profits during the decade of the
1940’s.1
In a statement that I made for the Joint Committee 011 the Eco
nomic Report in hearings on corporate profits in 1948, I had this to
say on the present issue. A t that time, the problem was one of deciding
what were high or low profits and, o f course, the question of how profits
should be estimated for taxes was relevant.
*
* * Perhaps the most perplexing one (problem) relates to the valuation of
assets. In a period of rising prices, inventories, and capital generally rise in
value.
Higher values for inventories mean higher profits. But should inventories
be revalued at replacement costs, then profits would be substantially reduced.
Profits would have been $6 billion less in 1946, or about one-seventh of the
profits of this year prior to taxes.
In the same year, business depreciation charges were $8.7 billion. It is clear
that had depreciation been at replacement value, profits would have been less
by several billion dollars additional. But the tax collector does not generally
allow depreciation charges to cover replacement in periods of rising prices as
against acquisition or book value.
I cannot enter into the merits of this debate. The accountants under pres
sure from business, are reconsidering the whole problem. It is well to remem
ber also that with depreciation based on replacement value and with inventories
carried at replacement value, if profits would be lower in periods of rising prices,
they would be higher in periods of depressions and falling prices.
What business would gain now, they would lose in periods of depres
sion * * *.2
A t the same hearings, Professor Slichter took a somewhat different
position. He thought that the inventory replacement would absorb
a substantial part o f the book profits. Assume a gain o f $ 100,000
from the sale of inventories as a result o f higher selling prices. In
Slichter’s view, these $ 100,000 would not be available for release to
stockholders because they would have to be used to replace the inven
tories that had been sold. Similarly, he pointed out that with prices
100 percent above the 1940 level, the replacement o f plant equipment
would be at a much higher cost than was the price of acquisition. I f
we assume that the prices o f the postwar would be at least 60 percent
above prewar in the foreseeable future, then Professor Slichter would
1 “ Relationship of Prices to Economic Stability and Growth,” compendium, p. 37.
a Hearings, Joint Committee on the Economic Report “ Corporate Profits,” 1948, p. 43.
THE INCIDENCE OF INFLATION
99
have increased the depreciation charges by 60 percent above the 1940
level.3
Obviously, if prices are to continue to rise, as seems likely, and
certainly seems much more likely than was generally assumed in the
early postwar period, then to that extent, the profits of industry tend
to be overstated. The tendency to allow accelerated depreciation
since the above evidence was given, to some extent prevents the emerg
ence of this difficulty. A t any rate, it may be held in a general way
that profits are a smaller part of the national income, and that once
correction is made for the overstatement because of the inadequate
allowances for replacement of inventories and equipment, then profits
have suffered an additional loss.
This problem has also received attention in recent years, particu
larly by the Machine and Allied Products Institute. This organiza
tion shows that in 1955 the ratio of current prices to average prices
underlying historical cost depreciation was about 1.31 compared to 1.38
shown by the study of the Department of Commerce.
The MAPI study also shows profits adjusted by reducing them by the amount
of the additional depreciation required to shift from historical to current prices;
by adding accelerated amortization in excess of depreciation otherwise allow
able ; and by adjusting for the effects on profits of changes in inventories values
as estimated by the Department of Commerce. The effect of these shifts is to
reduce profits by less than 1 percent on the average during the period 1925-29
and by about 25 percent for the average of the years 1946-55 * * *.
In a period of rapidly rising prices such as has prevailed in the last 15 years,
the use of current replacement costs rather than original or historical cost in
calculating the cost of fixed assets will have the effect of reducing the ratio of
corporate profits to sale, to net worth, or to income originating * * *. Contrari
wise when prices are falling, the use of current price depreciation will result in
higher profits than if original cost depreciation were used. Which basis should
be used in calculating depreciation has been and still is a matter of widespread
debate * * *.4
In 1929 the ratio of current-year cost to original cost for structures
and equipment in manufacturing establishments, was 1.17; in 1983 it
was 0.92; by 1939, 1.09; for 1948, 1.58 and by 1955, 1.38.5
PRICES AND WAGES I N CYCLES
One o f the striking developments in the last generation or so, has
been an increased tendency tor prices to be relatively rigid in the
downward phase of the business cycle, and also for wages to be much
more rigid. In earlier cycles, as the business situation deteriorated,
wages would be cut sharply and prices would fall greatly. But now,
whatever the reason, there seems to be a greater degree of stability. In
fact, in periods of recession now, average hourfy earnings tend to rise
though not as much as in periods of prosperity.
Even before the war, this situation had become apparent. For ex
ample, here is a table based on Daniel Creamer’s study.6
a Ibid., pp. 5-7.
* Joint Economic Committee, “ Productivity, Prices and Incomes,” 1957, p. 30.
5 Ibid., p. 99.
6 “ Behavior of Wage Rates During Business Cycles/’ NBER occasional paper 34, 1950,
p. 37.
THE INCIDENCE OF INFLATION
100
T able
10-8.— Average hourly earnings, prices and production in manufacturing,
1921-88
Date of corresponding contractions
January 1920 to July 1921..................................... ........................... .
May 1923 to July 1924______________ _______ _____________ ____
October 1926 to November 1927_____________ _____________ ______
June 1929 to March 1933_____ ____________________________ _____ _
May 1937 to June 1938_______ ________________ _______ ____ ______
Percent
change
manufac
turing
average
hourly
earnings
-24.0
-.8
-.4
-25.0
-.8
Wholesale
prices of
Factory
finished production
goods
-15.1
-7 .6
-7 .4
-36.9
-10.7
-33.9
-13.8
-5 .0
-49.9
-27.7
In the table that follows, I summarize the results of my study
through three business cycles in the postwar period.
T a b le
10- 9.— Changes
in
average h o u rly ea rn in g s ( in clu d in g overtim e), ivholesale p r ic e s , a n d p ro d u c tio n {s e a s o n a lly a d ju sted ), f o r selected
in d u stria l g ro u p s in 3 p ostw a r rec e s sio n s
48084— 59--------------- 8
All manu
facturing
Processed
food
$0.07
-$0.02
$0.04
0
$0.04
0
-14.1
-3
-12.2
-26
-12.2
-10
-7 .4
-13
-13.9
-8
00
-25
Apparel1
Paper and
allied
products
Chemicals
and allied
products
Primary
metals
0
0
-3 .1
-16.82
$0.03
2.3
-11.0
-6
$0.03
1.7
-.8
-10.79
$0.03
1.6
.4
-14.65
$0.05
3.1
-1 .6
-7
$0.04
0
$0.07
-$ 0 .01
$0.01
$0.09
-8 .1
-6
-4 .8
-2 0
-4 .8
-1 4
-1.36
-7
-.7
-6
CO
-25.9
$0.06
2.9
2.4
-14.09
$0.07
3.2
-.4
-21.56
$0.05
2.7
-.6
-6
$0.05
$0.01
0
$0.02
$0.02
—7
-4
-2 .8
-13
-2 .8
-1 5
-.4
-9
-.9
-5
-$ 0 .03
-30.3
00
-$0.03
-12.0
00
$0.05
-12.5
CO
$0.05
-2 .3
00
$0.13
-25.9
00
$0.16
-14.7
00
Fabricated metal
products 6
Structural
Nonelectrical
machinery
1 The prices given are for textile products and apparel combined.
Transport
equipment
Nonstruc
tural
(
-8 .0 1
— 15
1
$0.03
-1 .7 1
— 16
1
$0.05
-1 .8 1
16
1
s Average hourly earnings figures are for primary smelting and refining of nonferrous
metals.
4 Average hourly earnings figures are for rolling, drawing, and alloying of nonferrous
metals.
* As noted in the table, 2 price figures are given, for “ structural” and for “ nonstructural”
abrieated metal products.
Electrical
machinery
$0.15
00
-42
Finished
steel
products2
-$ 0 .05
1.1
(8)
$0.13
-.6
00
$0.17
7.7
(6)
Automobiles,
trucks, and Bituminous
parts
coal
-1 .8
$0.07
(7)
-2 3
0
-5 .8
-1 3
$0.03
00
-2 0
0
-1 .9
-16
-$ 0 .07
-7 .7
-3 9
—1.5
$0.09
(7)
-1 8
$0.05
-1 .3
-2 0
$0.05
00
-15
$0.11
-1 .2
—35
-$ 0 .01
-12.4
-3 3
-1 .4
$0.11
(7)
-2 3
$0.09
4.4
-2 3
$0.10
00
-21
$0.13
9.9
-59
$0.07
-4 .2
-34
6 Not available.
7 BLS has no composite index for nonelectrical machinery prices. See table 11-2, for
the movements of prices of 5 subgroups of nonelectrical machinery prices. They indicate
smaller declines or bigger price rises in each succeeding recession. Prices for all 5 groups
rose, 1957-58.
Source of series: Monthly Labor Review and Federal Reserve Bulletin.
OF INFLATION
Nonferrous
mill shapes4
2 Average hourly earnings figures are for blast furnaces, steel works, and rolling mills.
Textile
products 1
-6 .4
-8.65
Primary
nonferrous
refinery
shapes 3
November 1948 to October 1949:
Average hourly earnings........ ........................
Wholesale prices (percent)____ ____________
Production (percent).................... ..................
July 1953 to August 1954:
Average hourly earnings..........—..................
Wholesale prices (percent)............. ................
Production (percent)..... ........... .............. ......
July 1957 to April 1958:
Average hourly earnings.................................
Wholesale prices (percent)........ .....................
Production (percent).......................................
Nondur
able goods
INCIDENCE
November 1948 to October 1949:
Average hourly earnings.......
Percent............................
Wholesale prices (percent)—
Production (percent).............
July 1953 to August 1954:
Average hourly earnings____
Percent________________
Wholesale prices (percent) —
Production (percent)_______
July 1957 to April 1958:
Average hourly earnings.......
Percent.................. ..........
Wholesale prices (percent)—
Production (percent).............
Durable
goods
102
THE INCIDENCE OF INFLATION
It will be noted that in the 1948-49 cycle, manufacturing prices did
fall by 6.4 percent for all manufacturing; but average hourly earn
ings did not change at all. In the 1953-54 recession, average hourly
earnings rose by 1.7 percent while wholesale prices declined by 0.8
percent; in the 1957-58 recession, average hourly earnings actually
rose by 6 cents or 2.9 percent, and wholesale prices also rose by 2.4
percent despite a decline in production of 14 percent. Variations on
this theme are to be found in different industries and where the indus
tries approach the raw material stage, there is a greater tendency for
prices to fall, though even in these cases wages tend to stay up and
actually increase in the more recent recessions. Obviously, if there
is little change in prices, or if prices rise somewhat, and wages rise
even more, it follows that even m periods of recession there is a sub
stantial gain o f real wages. Indeed, the rise is smaller than in periods
o f prosperity.
The most marked deviation from earlier cyclical behavior has taken
place in primary metals, machinery, automobiles, and transportation
groups; that is, in the heavy durable industries and in coal. In these
industries, the average hourly earnings have gone up more in each
successive decline beginning with a steady level in the 1949 recession
and increasing at most among these groups by about 5 to 10 cents in
the 1953-54 recession and 10 to 15 cents in the average hourly earnings
in the 1958 recession. Prices also tend to rise more and decline less
in these groups with the exception of nonferrous metals, prices o f
which are tied to world primary market price to a large extent. In
each succeeding postwar cycle, our production has tended to fall more.
Prices and wages in the metals and machinery group, excluding auto
mobiles, have risen more than most groups in the boom periods o f the
postwar cycles. Prices, especially in the metals, machinery, and oil
groups, have risen far more than average, but production has not.
In steel there seems to be a tendency for profits after taxes as a
percentage o f sales to rise since 1939. These, of course, include some
4 years when profits were kept under control through price fixing.
In 1950 with operating rate at 96.9 percent o f capacity, profits m
relation to sales were 8.1 percent. This high level, o f course, reflects
the speculative influence o f the breakout of the Korean war. What is
also o f some interest is that from 1952 on, there was a steady rise in
the profits after taxes as a percentage of sales. Beginning with 1952,
the figures were 5, 5.6, 6, 7.8, and 7.3 in successive years through
1956. In these 5 years, the operating rate was 86, 95, 71, 93, and 90
percent o f capacity. Even at 71 percent of capacity, profits had risen
from 5.6 to 6 percent even though in that year operating rate as a per
centage o f capacity had fallen from 95 in 1953 to 71 percent in 1954.
In these 5 years, the employment costs—percent of total—varied from
a maximum o f 36.7 in 1954 to a minimum of 32.3 in 1951/
For five nondurable groups for which I have statistics, there does
not seem to be any consistent tendency for hourly wages to rise more
and more in recent recessions since W orld W ar II, but in these indus
tries also prices have tended to fall less in each succeeding cycle, or
they have risen in each cycle.
In his study 8 Dr. Geoffrey Moore also shows a tendency for fluctua
1 See hearings, “Relation to Prices to Economic Stability and Growth,” p. 703.
8National Bureau of Economic Research, “Measuring Recessions,” 1958.
THE INCIDENCE OF INFLATION
103
tions to be less in more recent cycles than before the war. Not only
does industrial production tend to fall less, but personal income stays
up remarkably well. This is explained partly by the tendency o f
wage rates to stay up or even to increase and also by the fact that
transfer payments play a much larger part and these tend to increase
in the decline of the cycle .9 As might be expected, therefore, the
personal income falls much less than GNP .10
According to Hultgren’s studies, the following conclusions are
suggested—
* * * The most rapid rates of increase in output per man-hour appear during
that portion of the business cycle—the last stages of contraction and the early
stages of expansion—when replacement, increase of plant and equipment are
proceeding most slowly; and that during the initial stages of contraction, decline
in output per man-hour joins with increasing wage rates to push unit labor
costs up.11
CONCLUSION
One o f the most discussed problems in recent years has been what
causes inflation. A viewpoint that attracts increasing numbers of
adherents is that inflation is more a cost-push than the old classical
excess-demand type of inflation that the monetary authority is able
to treat with some effectiveness. I f this interpretation is correct, then
it follows that labor and other cost factors improve their position
under modern inflation: their rewards exceed and lead the rise of
prices. Others are squeezed. In one sense even this kind of inflation
may be called an excess demand inflation; for its progress can be halted
and output affected by curtailment of monetary supplies.
In inflationary periods experience o f industries varies greatly. In
some, wages seem to gain absolutely and relatively, but in others, for
example, services, capital seem to gain. It has also been noted that
prices rise and wages gain especially when production expands greatly.
In these industries profits may also greatly rise when prices are ad
ministered. In the long run this conclusion o f the relation o f produc
tion and prices is not so clear.
Much historical evidence exists to show that the trend is toward
wages leading in the wage-price inflation. W ar experiences in three
major wars are evidence of this point. But a long historical survey
such as revealed in Professor Paul H. Douglas’ classic study reveals
varying responses to rising prices. Over a period of 30-40 years, real
wages m manufacturing rose substantially, as they did for transport
workers and teachers, but Government workers, ministers, and clerical
and lower salaried workers experienced deterioration in their standard
o f living.
These figures do suggest that much depends upon the market for
the product and institutional factors. The large gains of teachers in
the generation before the great depression suggest great strides of
education and public financing which had not yet experienced the
impact o f inelasticities of real estate taxes. In contrast the difficulties
o f teachers in the more recent inflation period are to be associated,
despite the unusual growth o f demand, with weakness in financing
methods.
•Ibid., 268-269.
10 Cf. p. 287.
11 Hearings, Joint Economic Committee, “ Employment Growth and Price Levels,” pt. 2,
p. 306.
104
THE INCIDENCE OF INFLATION
In general, the rise of wages over the last 60 years or so has exceeded
that o f productivity, and the combined rise of productivity and of
prices, suggesting relative gains of wages against property income.
Part o f these gains are associated with the changing structure of in
dustry; that is, gains for those employments where labor especially
profits. These improvements have not been uniform, however. There
were periods when labor lost as well as gained. But more recent
movements suggest consistent gains for labor, exclusive of early post
W orld War II. In a depression period, wage rates increasingly tend
to fall less or even rise as prices fall. Hence gains may be substantial
in depression periods. But increased rigidity o f prices tends to rob
workers of gains in such periods in recent years and, o f course, they
lose jobs.
Over long periods of time one finds little association between wage
movements and productivity in individual industries. In fact, the
largest gains in real earnings seem often to come to those industries
which have experienced the smallest gains in output and productivity,
e.g., lumber and anthracite coal, and the smallest gains to workers
in those industries where output and productivity rose a maximum,
e.g., electric light and power. For example, the real hourly earnings
in anthracite coal rose 5.58 times as much as productivity from 1899 to
1953, whereas in electric light and power the increase was 0.11 percent
as much as that in productivity—a 50 to 1 relative advantage for an
thracite coal.
In general, the wage share in total income tends to rise. This is
often held to be part o f the explanation o f the inflation. But Dr. Ruth
Mack and Professor Ruggles have also stressed the rise of other
costs and the increased outlays on marketing, administration, research,
and the like.
The gains o f labor are also associated with the much larger increase
in supplies of capital than of labor and hence the depressing influence
o f rises in supply of capital on property income.
In periods o f rising prices, profits are lower than they seem to be,
for replacement costs of inventories and plant and equipment exceed
acquisition costs on the basis o f which depreciation is measured. One
estimate put replacement costs o f structure and equipment in manu
facturing industries at 1.58 o f acquisition costs in 1948 and 1.38 in 1955.
C h a p t e r 11. A
ttem pts
T o B e a t I n f l a t io n
Increasingly the public is aware of the dangers o f inflation. Gov
ernment officials and others proclaim the probability that creeping
inflation will result in galloping inflation, thus tending to encourage
the use o f protective devices against inflation. It would be well if the
Government instead of proclaiming the dangers o f inflation would
bring to the attention of the public that over a period o f 120 years
the rise o f prices averaged a little more than 1 percent. But what
ever the explanation, there is an increased tendency to take measures
to protect against inflation. This is evident, for example, in the large
rise in the stock market, and the desertion o f the Government bond
market in recent years. It is also evident in the doubling in the num
ber protected by cost-of-living escalator clauses in wage contracts
since 1955. According to the Department of Labor, they now cover
THE INCIDENCE OF INFLATION
105
more than 4 million workers. Long ago Irving Fisher suggested that
bonds be issued with an escalator clause to protect against inflation.
In fact, he headed a corporation which used this form o f financing.
I also earlier suggested the wisdom o f using an escalator clause in
payments o f benefits to the old.
Indeed, a general use of escalator clauses would be dangerous. The
result would be that instead o f a lag o f some prices behind the general
inflation inclusive o f income payments, these would immediately
react to an inflation and therefore tend to aggravate the inflation.
But a judicious use o f escalator clauses in limited markets, where the
cost o f inflation is the greatest, might well be justified.
It should also be noted that any attempt to escape the effects of
inflation through the purchase of equities is also subject to some res
ervations. The more people buy common stock in order to protect
against rising prices, the higher common stock prices will rise and,
therefore, the return will tend to be reduced. This was clearly evident
by 1959 when the yield on common stock declined substantially below
that on bonds. In fact, when the Government issues a 5-year note pay
ing 5 percent, as it did in 1959, the investor who wants to protect him
self against inflation achieves substantial security. It can be argued,
for example, that the investor seeking the maximum return would
expect to get a 2-percent additional return on fixed income yielding
assets to match the annual 2 percent inflation, say, anticipated. A
bondholder should receive 2 percent additional return to offset the
anticipated inflation and in addition be compensated for the greater
gains accruing to holders o f equities because corporations are run on
their behalf. In corporate finance the stockholder is the one who
profits most from corporate management, not the bondholder. With
rising prices and profits, the latter in part associated with rising prices,
the gains would probably go largely to those who hold equities. Hence
if stocks yield, say, 2 percent at present high values, then the bond
holder might well expect a return o f at least 5 percent in order to
make up for the advantages o f those who are especially protected
by the directors and management o f corporations and to offset the
effects o f inflation.
In an interesting stud}' the First National City Bank tries to show
what the effects of different kinds o f investments would be upon the
value in current dollars, and purchasing power o f varying kinds o f
investment12in 1948 dollars.
32 First National City Bank Monthly Letter, “Business and Economic Conditions,” April
1959, p. 46.
THE INCIDENCE OF INFLATION
106
T able
11-1.—Changes in the nominal value of various types of investments, 194858, in current dollars and in dollars of 1948 purchasing power
Nominal value
Cash................................................................................................
Bonds:
U.S. Treasury 2^s of 12/15/67-72..........................................
New York City 4s of 10/1/80................................. .................
State of New York 194s of 3/15/85...........................................
Preferred stock average, (Standard & Poor’s).............................
Common stocks, averages (Standard & Poor’s):
Railroads.................................................................................
Industrials...............................................................................
Public utilities.........................................................................
Fire insurance..........................................................................
New York City banks...........................................................
Real estate:
Selling price, typical 1-family residence1........ ......................
Farm real estate, Department of Agriculture index
(1947-49=100).......................................................................
Percent
change
in
nominal
value
Percent
change
in 1948
dollars
Dec. 31,
1948
Dec. 31,
1958
$100.00
$100.00
100.15
124.25
86.58
169.50
85.20
102.00
66.80
151.70
—14.9
-17.9
-22.8
-10.5
-29.1
-31.6
-35.7
-25.4
13.92
15.12
16.04
13.05
11.18
34.39
58.97
43.28
34.30
24.25
+147.1
+290.0
+169.8
+162.8
+116.9
+105.8
+224.9
+124.8
+118.9
+80.7
$14,231
$17,220
+21.0
+0.8
106
163
+53.8
+28.1
-16.7
i Computed from Roy C. Wenzlick & Co., “ The Real Estate Analyst.”
For example, in 1948 dollars these investments o f 1948 in 1958
would be worth as follows:
I f invested in—
Cash________________________________
U.S. Treasury 2y2s for 1967-72________
New York City 4s of October 1980______
State of New York l% s of March 1985—
Preferred stock---------------------------------Common stocks:
Railroads------------------------------------Industrials-----------------------------------Public utilities___________________
Private insurance companies______
New York City banks_____________
Typical one-family residences_________
Farm real estate_____________________
—16.7 percent.
—29 percent.
—32 percent.
—36 percent.
—25 percent.
+106 percent.
+225 percent.
+125 percent.
+19 percent.
+81 percent.
slight gain in real selling price.
+28.1 percent in selling price,
stable dollars.
The First National City Bank then goes on to explain the difficul
ties o f purchasing common stocks that yield these returns, for these
are only averages. Similarly with houses, which o f course may de
preciate in value as well as rise, even though the average may be
worth more after 10 years in dollars of stable purchasing power.
It took 16 of the 30 stocks presently in the Dow-Jones industrial average more
than 20 years to get back up to 1929 highs—and 4 of the stocks still haven’t
made it.
The bank also points out that some common stock declines have
occurred in periods of rising consumer prices.13
“ Ibid., p. 46.
107
THE INCIDENCE OF INFLATION
T able
11-2.— Some common stock declines since 1900 versus Consumer Price Index
Period
June 1901 to November 1903__________________________________________
January 1906 to November 1907______________________________________
November 1916 to December 1917_____________________________________
November 1919 to August 1921_______________________________________
September 1929 to November 1929________________________ - __________April 1930 to July 1932_____________ „_____________ ___________ _______
March 1937 to March 1938____________________________ _______________
October 1939 to April 1942______ _____________________________________
May 1946 to June 1949_______________________________________ *_______
April 1956 to October 1957___________ „________________ _______________
Stock
prices,
percent
change
-4 6
-4 9
-4 0
-4 7
-4 8
-8 6
-4 9
-4 0
-2 4
-1 9
Con
Interval
sumer to recover
stock
prices,
percent price high
change (months)
+6
+7
+19
-4
-0 .4
-2 0
-1
4*15
+29
+5
45
128
32
61
302
287
105
63
47
29
Source: November 1957 Journal of Insurance—except for the last 2 periods which are our own
computations.
What is more, if one invests in stock yielding 2y2 percent, it may take
a considerable period of time to make up for the return that might be
had in more normal periods on stocks or other investments.
I f, for example, one purchases stock yielding 2y2 percent and as
sumes the rate o f annual dividend to increase 3 percent, it would take
23^ years to achieve a return o f 5 percent. I f we assume the annual
dividend increase is 5 percent, it would take 18 years to achieve a return
equal to 6 percent. Considering the return on capital stock one must
also remember that there is a 25-percent capital gains tax i f the stocks
are sold.
The inflation is, of course, a serious matter for any organization or
for any business that cannot increase its receipts as fast as its expenses
increase. This is particularly true, for example, o f the average col
lege. Costs rise because the college has to pay increased prices for
material and services even though it does not profit from the increase
in productivity that, for example, business firms very often experience.
The service the college provides is often a matter o f individual atten
tion between the teacher and the student, or between the teacher and
the student in relatively small groups. Here automation does little
good, though sometimes college administrators underestimate the pos
sibilities. Because they are determined to keep tuition down in order
not to penalize poor students, college administrators experience finan
cial difficulties as they find their expenses rising 5 percent a year, say,
and without offsetting productivity, and their revenues do not respond
to the inflation. Government appropriations, tuition income, gifts,
and endowment income tend to lag, and specially when enrollment is
rising.
To some extent they try to deal with this problem by investing in
common stock insofar as they have funds to invest. It must be re
membered that a relatively small number of colleges have resources to
invest in common stock or any other investments. In fact, since the
beginning o f the century the return on investments has declined from
about 25 percent of the income o f institutions o f higher learning to
about 5 percent today. This reduction is the result not only o f infla
tion, though I estimated recently a loss o f $1 billion since the prewar
or about one-third as a result o f inflation. Endowment income plays
a smaller part also because incomes have risen from other sources as
enrollment has risen.
108
THE INCIDENCE OF INFLATION
Those colleges that depend upon endowment have tended to move
into common stock to protect against the inflation. O f course, one
result is that they tend to get lower returns on their investments,
especially when they move into growth stocks. There are certain dis
advantages in moving into growth stocks for colleges because they
tend, therefore, to favor later generations when presumably higher
incomes will be available rather than the present generation which is
having difficulties o f its own. The objective of purchasing growth
stocks is o f course to obtain higher returns on investments in the long
run. Unless bonds yield, say, 2 percent more than stocks to offset
inflationary trends and, say, 3 percent additional reflecting growth,
the case for bonds is weak indeed. Yet the colleges have been slow
to move into common stock. For example, in 1926 institutions of
higher learning had about 9 percent of their funds in common stock;
by 1933, partly to take advantage o f bargains that could be had in
common stock, the amount had risen to almost 13 percent. Even by
1947, according to one study, only 30 percent was in common stock and
the most recent figure suggests 50 to 60 percent in common stock. It
should be noted that the major part o f the increase in investments in
common stock has been not the result of going out to purchase more
common stock but the continued rise in their value.14
This tendency to protect against inflation is, of course, also evident
in the holdings o f stock by other institutions. For example, the Fund
for the Eepublic report shows that according to a stock exchange esti
mate, institutional holdings of stock in 1949 were $76 billion and in
1957, $196 billion, or an increase of 12.4 percent of the total outstand
ing to 15.3 percent. The largest relative increases were by invest
ment companies, open-end, with an increase o f '379 percent, and noninsured pension funds from $0.5 to $5.7 billion, or 1,140 percent. This
can, however, be only a rough estimate for the figure given for college
and university endowments o f $2.4 billion in 1957 seems excessive
on all counts. Financial intermediaries, that is commercial banks,
mutual savings, private life insurance companies, etc., do not seem
to have increased their holdings in recent years. In 1939 they held
$22.2 billion and in 1957, $20.5 billion .15
What is striking is the large proportion of net purchases of com
mon stock by institutions. Apparently in 1954 they purchased $ 1,520
million net as against $120 million by foreigners and $460 million by
domestic individuals. In other words, they purchased about Sy2 times
as much as domestic individuals though tlieir total holdings in 194957 was 12 to 15 percent.16 These figures do point to an increased ten
dency for institutions, so far as they are not restricted by legal pro
visions, to protect themselves against inflation by purchasing equities.
14. gee especially J. H. Cain, “ What Is Happening to College and University Investment
and Income ?” American Council on Education Study, June 1941, p. 3 0 ; J. I. Kirkpatrick.
“ A Study of University Endowment Funds,” 1947, p. 4 1 ; The Boston Fund, “ A Study of
the College and University Endowment Fund,” June 30, 1956; Vance, Sanders & Co.,
“ Brevits,” vol. N, No. 2 1 ; Barron’s, June 17, 1957; also my forthcoming book on “ The
Economics of Higher Education.”
_
15 R. Tilove, “ Pension Funds and Economic Freedom,” a report to the Fund for the
Republic, pp. 38 and 40.
™Ibid., p. 43.
109
THE INCIDENCE OF INFLATION
C h a p t e r 12.
S o m e I n t e r n a t io n a l A
spects
It is well known that in 1958-59 the United States experienced a
drastic reversal in its trade or its balance o f payments position. From
1950 to 1957 foreigners acquired short-term claims o f $ 10.3 billion
on the United States, but they only converted $2.6 billion into gold.
But in 1958 total accumulation o f foreigners was $2.3 billion or our
gold and $ 1.3 billion additional o f short-term dollar assets. In other
words, whereas in the preceding 7 years they had taken only about
one-quarter o f their dollar assets in g old ; in 1958 they took more than
two-thirds. This suggests that the large outflow o f gold and the
accumulation o f dollars by foreigners by 1958 reflects some doubts on
the dollar.
Actually, foreign countries have been building up their gold and
dollar balances since 1950. From 1950 to 1958 U.S. gold sales
amounted to $3,712 million and foreign banks and official institutions
increased their dollar holdings by $7.5 billion or total losses of $11
billion, an average o f $ 1.2 billion per year. Only in 1957 was this
movement reversed, and this resulted in part from the Suez crisis.17
It is a matter o f common knowledge that rising prices in one coun
try not offset by rising prices elsewhere would tend to result in re
duced exports and in increased imports for the country that experi
ences the larger rise o f prices. It is not, therefore, surprising that
Chairman Martin o f the Federal Reserve Board and others have
brought home the point that if our balance o f payments is not to de
teriorate further it is important to stop the inflation. In other words,
inflation in this country is held responsible for the adverse balance
o f payments o f the United States.
In an analysis of the situation by G A T T ,18 the conclusion was
drawn that rising manufactured prices in the United States had helped
contribute toward the decline in the exports of the United States. The
point was also made that the leveling off o f activity in other indus
trial countries resulted in a large fall o f exports of crude materials,
fuels, and metals from the United States. And when the recession
came, there was not the expected drop in prices that generally accom
panies a recession. In other words, an inflation may be more costly
in international position these days because prices seem to respond less
than they have in the past in recession periods. Hence, as exports
fall and imports rise and gold leaves the country, the usual price cor
rectives may not operate. Particularly where prices are largely ad
ministered, the effect is likely to be a greater stability o f prices and,
therefore, a failure to get the downward adjustments o f prices that
are expected from a loss o f gold and dollars to foreign interests.
The G A T T report concludes as follow s:
*
* * In view of such factors as the influence of commodity composition on
the unit-value index, the long-term structural changes, and the fact that some
markets were hit more severely by the recession than others, the evidence seems
to be insufficient to claim that the price development in the United States is the
single most important cause of the relative and absolute decline in the country’s
exports of manufactured goods. Nevertheless, the United States has lost much
of the superiority it had in productivity during the years immediately following
17 Joint Economics Committee, hearings on “Employment, Growth, and Price Levels,
Pt. 5 : International Influences on the American Economy,” 1959, p. 967.
18 “ International Trade, 1957-58,” July 19t59i, pp. 107-108.
110
THE INCIDENCE OF INFLATION
the Second World War, while, at the same time, Western Europe and Japan
have greatly improved their ability to deliver promptly all kinds of manufac
tured goods. Prices, therefore, play a more important role than hitherto, and
U.S. exports are, consequently, more sensitive to rises in labor costs than had
been the case earlier.
W ith 1953 as 100, the report concludes that unit-value indexes of
exports o f manufactured goods in the United States was 113 in 1958,
107 for Canada, 110 for United Kingdom, 103 for Germany, 99 for
France, 101 for Italy, and 94 for Japan .19 These figures do, indeed,
suggest a deterioration o f competitive position to some extent.
Bernstein has shown, however, that U.S. exports as a percentage o f
world exports stayed up well, with the figure fluctuating from 17.7
percent in 1950, a minimum o f 17 percent in 1953 and 1954 and a maxi
mum o f 19.9 percent in 1957, and a decline to 17.2 percent in 1958, a
reduction he explains by special circumstances.20
But this position may be carried too far also. Should one take the
whole postwar period, one would find that the price history o f the
United States is considerably better than that o f most o f its com
petitors. For example, here is a table which gives the prices in the
United States and other countries in 1948-57 and 1953- 57 :
Table 12-1.— Comparison of domestic inflations, autumn, 1&57
1948=100
Chile.............................................
Argentina____________________
Indonesia_____________________
Brazil________________________
Japan________________________
Peru_________________________
Colombia_____________________
Mexico_______________________
Australia_____________________
France_______________________
United Kingdom______________
Sweden_______________________
Netherlands__________________
West Germany________________
Canada_______________________
Belgium______________________
United States_________________
Tndiq,_________________________
Switzerland____ ______________
Philippines1__________________
Italy_________________________
Venezuela____________________
Cuba_________________________
Whole
sale
prices
Cost of
living
1,880
1,650
590
347
342
195
209
206
200
201
185
153
147
152
117
127
115
118
115
111
104
132
120
90
372
291
274
205
204
183
159
150
143
117
115
113
113
108
103
103
99
97
1953=100
Whole
sale
prices
Chile....................................
Brazil____________________
Argentina________________
Indonesia_________________
Colombia_________________
Peru_____________________
Mexico___________________
France___________________
Sweden___________________
Australia_________________
United Kingdom_____ ____
Netherlands______________
United States_____________
Belgium__________________
Denmark_________________
Japan____________________
West Germany____________
Switzerland_______________
Italy_____________________
Canada___________________
India. ....................................
Philippines_______________
Cuba......................................
Venezuela. _...........................
Cost of
living
696
195
160
136
135
119
108
108
107
107
107
106
105
105
105
105
103
101
101
101
100
99
636
212
182
163
140
127
143
111
113
113
118
117
106
109
113
111
108
107
114
107
106
106
97
102
11950—100.
Source: Joint Economic Committee, “ The Relationship of Prices to Economic Stability and Growth,”
compendium, 1958, p. 289.
This table does show in general that price history in the United
States has been more than satisfactory as compared to its major rivals.
Indeed, this table does not give all the facts because we have to take
into account what happened to exchange rates. I f a country doubles
Ibid., pp. 107-108.
20 Joint Economic Committee, hearings on “ Employment, Growth, and Price Levels.
5 : International Influences on the American Economy,” 1959, p. 966.
Pt.
THE INCIDENCE OF INFLATION
111
its prices and then cuts the value o f its currency by 50 percent, the
net export capacity should not be greatly influenced. But actually
the depreciation or devaluation of exchanges has tended to lag behind
the rise o f prices and particularly since 1953. What is more, where a
devaluation has tended to anticipate or come before a rise of prices
in the period o f high employment and elastic monetary supplies of
the postwar, it was not long before prices became quickly adjusted to
the fall in the exchange rate.21
Even from 1955 to 1958 the increase in wholesale price in the
United States was not greatly out o f line. The rise was 3 percent for
the United States, 1 percent for Europe, 8 percent for the United
Kingdom, 5 percent for Germany, 2 percent for Italy, and 3 percent
for France, and no change for Japan. In the export area the increase
from 1955 to 1958 was 6 percent for the United States, 1 percent for
Europe, 8 percent for the United Kingdom, 3 percent for Italy, and
2 percent for France.22
It is also well to remember that up until recently the United States
had a very important competitive advantage over other countries.
Most other countries had been seriously damaged by the war, whereas
our economy had become more productive. It would be expected that
gradually as these countries recovered, their competitive position
would improve vis-a-vis the United States. The major explanation
o f the reversal o f gold movements recently is not the inflation in the
United States but rather the fact that our competitors are catching up
in productivity, partly as the result o f help given by the United
Sates through various aid programs. W e would therefore expect that
the U.S. position in international trade would deteriorate to some
extent, but still would be considerably better than it had been, say in
1938. For example, here are some statistics on this point which
show that with 1938 as 100, the U.S. export trade had reached in
volume 200 by 1948, as against 77 for other countries. Then there
was a gradual deterioration, relatively speaking, to 207 for the
United States and 110 for other countries in 1951. By 1956 the figures
were 255 for the United States and 147 for other countries. By the
second quarter o f 1957, the figures were 261 and 156, respectively;
and by the second quarter of 1958, 230 and 161, respectively.23
The Common Market in Europe may well result in further losses of
competitive position for the United States, although how much will
depend upon the rise o f productivity in these countries and the extent
to which this improvement is taken in reduced prices and how much
through higher wages.24
21 Harris, S. E., “ International and Interregional Economics,” 1957, ch. 25.
22 Figures from IMF, “ International Financial Statistics,” April 1959.
23 GATT, “International Trade, 1957-58,” p. 6.
24 Joint Economic Committee, hearings on “Employment, Growth, and Price Levels. Pt.
5 : International Influences on the American Economy,” 1959, pp. 1033-1037 and 1 04 8 1051 for interesting remarks by Messrs. Despres and Scitovsky on the effects of the Com
mon Market.
112
THE INCIDENCE OP INFLATION
Here is the trend of trade from 1955 to 1958:
Table 32-2.— North America s trade in manufactures and total, 1955-58
[In billions]
Total manufactures
1955 ____________________________ ________ ________________
1956 .. ................................. ...................................................
1957_____________ ____ _______________________ _____ _______
1958___________ _______ _____ _______ _____________ _________
Grand total
Exports
Imports
Exports
$8.86
10.14
11.20
10.40
$5.77
6.88
7.07
7.14
$17.55
21.01
23.36
20.33
Imports
$15.64
17.75
18.18
17.49
Source: “ International Trade, 1957-58,” pp. 100-101.
Striking is the very large decline of about $3 billion in exports in
1958 and reduction of imports of only $0.7 billion.
The situation has continued to deteriorate in the first half o f 1959.
In the year 1958, the increase in foreign gold and liquid dollar assets
at the expense of the United States was roughly $3.4 billion, and
this figure reflects the adverse balance of payments of the United
States. In the first quarter of 1959 the loss was almost $900 million
and in the second quarter almost $1 billion .25
These large losses of gold and dollars are indeed serious, and if
they are not stopped or greatly reduced strong measures would have
to be taken. Though the total gold reserves of the United States are
large, the excess, once one allows for the dollar reserves of foreign
banks and the like, is not large. A t the end o f 1958, against $20.5
billion of gold, foreigners held claims in dollars of $12 billion, and
by August 1959 we had lost an additional billion of gold .26 Obviously,
under these conditions there would be some pressure to restrain price
rises. Unfortunately, any serious restrictive monetary policy would
also have a serious effect on the economy and may not bring about
the increase o f exports and the reduction of imports that the policy
is supposed to induce. Part of the explanation of this fact is, of
course, the relative rigidity of prices and costs.
The Government might use three different approaches to the prob
lem o f trying to solve the balance o f payments. Assume we are
losing $3 billion o f gold and dollars a year. I f we pay out $9 billion
a year for military outlays and civilian expenditures abroad, for Gov
ernment loans and grants and private capital expenditures, an
obvious way out is to reduce these claims in dollars. Another ap
proach is, of course, to cut down on our imports; and finally, a third
approach is to increase our exports. The last way requires antiinflationary policies and may also involve increases in efficiency and
the discovery of new products, and, through all these measures, the
greater penetration of foreign markets.
In conclusion, I am not inclined to argue that the small inflation
we have had in the last few years has been a decisive factor in our
deterioration. Rather it has been the reestablishment of the competi
tive positions of the rest of the world, that we must take into account.
25 See Survey of Current Business, September 1959, p. 90.
26 Federal Reserve Bulletin, September 1959 and IMF, International Financial Statistics,
April 1959.
THE INCIDENCE OF INFLATION
113
I f prices and costs are increasingly rigid, it becomes more difficult to
make the necessary adjustments through the usual therapy. Insofar
as we now have to take account o f adverse balances, in turn related
to some extent to inflationary forces, we may be inhibited as we have
not been for many years in the pursuit o f maximum growth and high
eniployment policies.
Presently not anti-inflationary policies but a reconsideration of our
Government foreign credit and grant policies is required.27 In view
o f trends in reserves and credits in recent years, our contribution
should be reduced. In this connection, a decision in October 1959 by
the U.S. Government to require purchases under advances through
the Development Fund in the United States is of interest. This sug
gests an approach through protectionism; that is, favoring our sellers,
rather than a cut in aid.
27 Cf. “Employment, Growth and Price Levels,” V, pp. 970-971 and 1036 for similar
conclusions by Messrs. Bernstein and Despres.
STU DY PAPER NO. 8
P R O T E C T IO N A G A IN S T I N F L A T I O N
(B y H . S . H
o uth ak ker)
115
STUDY PAPER NO. 8
PROTECTION AGAINST INFLATION
(H. S. Houthakker, Stanford University, Stanford, Calif.)1
T he Sum m ary
This paper deals with the ways in which individuals may protect
themselves against inflation. The possibility of further inflation is
taken for granted, but it is also argued that inflation will probably
come to an end within the next 5 years.
The burden of inflation can be divided into redistribution, which
appears to be quantitatively less important than is usually believed,
and the results o f uncertainty concerning the value of money. A proEosal is made for an extension of the national accounts and flow-ofmds statistics to permit better measurement of the redistributive im
pact o f inflation.
From then on the emphasis is on the demand for assets as it is
affected by inflation. The increased demand for equities on the part of
the public is seen as a response to full employment, wThich has reduced
corporate risks. The resulting rise in share prices is held to be a
transition phenomenon and a further fall in yields on stocks is con
sidered unlikely.
Under “ creeping” inflation the rise in the general price level tends
to be discounted in the yield of bonds, but not in that o f shares, because
profits are believed to be a more or less constant fraction o f national
income; also cash balances are reduced relative to other assets. These
conclusions agree with current experience.
The present inflation is transitional, being a consequence of the
increase in demand for risky assets and in the willingness to incur
debts. It is not due to Federal extravagance, and in practice cannot
be cured by a tight money policy. Barring unforeseen contingencies
the end o f inflation is foreseen within 5 years.
A t present there are no generally satisfactory arrangements whereby
people o f modest means can protect themselves against inflation. In
surance policies whose payout is linked to share prices have disad
vantages similar to those of shares themselves. Individuals need
assets primarily for three reasons: ( 1) as a reserve, ( 2) to smooth out
anticipated fluctuations in income and consumption, such as those due
to old age, and ( 3 ) as a source of revenue to build up an estate. O f
these three goals achievement o f the second is made most difficult by
inflation.
Two measures are proposed to facilitate saving for old age and
similar purposes. The first is the introduction of index bonds whose
1 The author is indebted to David E. Kaun for research assistance and to Edward Shaw
for useful comments. Neither is to be held responsible for the views expressed in this
paper.
48084—86----- 8
117
THE INCIDENCE OP INFLATION
118
interest and redemption value are linked to a retail price index. A
rate o f interest o f 2% percent is suggested to fit in with current yields
on existing bonds. Objections that this step might destroy confidence
in the dollar, or that it would itself be inflationary, are refuted.
The second measure is stimulation of the supply of shares by making
it less attractive for corporations to issue bonds. This would be
achieved by limiting the deductibility of interest paid for corporate
tax purposes to the amount of interest received. The net cost of in
terest and dividends to the corporation would thus be equalized. The
tax rate on corporate income would simultaneously be reduced to keep
the total tax yield unchanged. Transition rules would prevent sud
den changes in tax liability.
These two measures would also have a favorable effect on the market
in Federal bonds.
T
he
A
n a l y s is
:
P
r o t e c t io n
A
g a in s t
I
n f l a t io n
This paper is concerned with the ways in which individuals may
protect themselves against inflation. It does not deal, except by im
plication, with the prevention of inflation, nor does it heap more sin
on any o f the several scapegoats that have so far been singled out.
For the purpose of this paper inflation is regarded, if not quite as a
fact o f life, at least as a real possibility. I do not mean to suggest
that the prevention of inflation is impossible or necessarily inconsistent
with full employment and economic growth, but it appears that the
combined achievement of these goals in the short run would call for a
degree o f economic statesmanship that is not available in any country
at the present time.
It is quite conceivable that inflation comes to an end without any
deliberate policy to that effect; some reasons for expecting such a
course o f events are discussed below. Indeed, we still cannot com
pletely rule out the possibility that deflation may become a more serious
menace, though the postwar record has been encouraging. However
that may be, the fear of inflation now exercises so large an influence on
the behavior o f so many people that it is appropriate to give some
thought to methods o f protection.
THE BURDEN OF INFLATION
Let us first consider what it is that individuals and society need to
protect themselves against. The burden of inflation can be divided
into two parts: A redistribution of income and wealth among those
who lose and those who gain, and the unfavorable consequences of
uncertainty as to the future course of prices.
REDISTRIBUTIVE EFFECTS
O f these two groups o f factors, the first can be further divided into
two components: The gains or losses in terms o f real purchasing
power arising from the holding of assets (or the owing o f debts)
whose amount is fixed in money terms, and the gains and losses experi
enced by various individuals because their incomes rise (or fall) faster
(or more slowly) than the prices o f the goods which they consume.
THE INCIDENCE OF INFLATION
119
Our knowledge in this area is still very fragmentary, and a suggestion
for improved information is made below. Some preliminary calcula
tions concerning the magnitudes of these components have been made,
however, perhaps the best study being the one by Bach and Ando .2
The results of Bach and Ando are particularly remarkable in that
they throw much doubt on a popular preconception concerning the
effects o f inflation. They do not find any great difference between
the experiences of such groups as business and labor or creditors and
debtors. Although some groups may temporarily lag behind others
during periods of rapidly rising prices, on the whole the distribution
o f income appears to be determined by much stronger forces than
merely monetary ones. These conclusions are surprising, but I see
no reason for doubting their substantial accuracy. What Bach and
Ando do find—and here they agree with public opinion—is that in
flation has benefited working people at the expense of retired people,
the young at the expense of the old. I f such a transfer of wealth took
f)lace once and for all, it would certainly be a reason for concern, but
ess so if inflation were a continuous phenomenon, for them the chil
dren who have robbed their parents will one day be robbed by their
children. The war and postwar years during which this transfer
occurred also saw a considerable extension of the old-age and sur
vivor’s insurance program which wholly or partly offset the losses of
the aged.
A PROPOSAL FOR IMPROVED STATISTICS
Illuminating though the calculations just referred to are, they do
not provide a sufficiently comprehensive picture of the impact of in
flation. For that purpose a more systematic and continuous effort,
undertaken by a Government agency, is necessary. Such an effort
could best be conceived as an extension of the national income ac
counts, or o f what is now emerging as their financial tw in: the flowof-funds system of the Federal Reserve Board. Two things are espe
cially needed:
(a) To measure the gains and losses due to different rates of in
crease o f income (the second type o f redistribution discussed earlier)
the existing data on income flows should be calculated both at cur
rent and at constant wage rates, just as expenditures are calculated
both at current and at constant prices. Although this will not be
possible for all types of income, it should not be too difficult to do for
many important sectors; extensive data on wages and hours worked
are already in existence. A breakdown by industry would be espe
cially helpful. As a result, there would not only be new information
on differential trends in real and money wages, but also on the closely
related and even more important subject of labor productivity. This
extension of the national accounts would not call for any fundamental
changes in present practice, and could probably be introduced at
short notice.
( b) Measurement of gains and losses on account o f assets and claims
requires a rather more ambitious approach. It would mean the in
troduction o f a national balance sheet in addition to the national in
* Review of Economic Statistics, February 1957.
120
THE INCIDENCE OF INFLATION
come or flow-of-funds accounts. This balance sheet (or rather, sys
tem o f interlocking balance sheets) would show for a number o f
sectors (households, nonprofit institutions, corporate nonfinancial
business, unincorporated nonfarm business, agriculture, financial in
stitutions, Government, “ rest of the world” ) the types o f assets held
(both physical and financial, at market value, and also at constant
prices) and the claims outstanding (also at market value and constant
prices, and distinguished by types such as bonds, equities, loans, etc.).
Such a system would in any case be a valuable complement to the
present accounts and would improve their accuracy; in addition to
allowing the measurement of capital gains and losses it would afford
valuable insight into such problems as overcapacity, the productivity
o f investment, access to capital markets, and the adequacy of deprecia
tion allowances. Its introduction would require some preparatory
research; useful spadework has already been done, especially by Ray
mond W. Goldsmith .3
TH E FEAR IN
INFLATION
Although we do not yet have sufficient knowledge o f the matter, the
results of Bach and Ando indicate that the redistributive effects of
inflation are not as serious as is commonly held to be the case. The
other group of consequences—namely, those connected with the fear
o f inflation—may well be more important, though they do not receive
as much attention. Fear of inflation manifests itself in an increased
reluctance to possess assets whose value is determined entirely in
money terms, such as bonds, mortgages, bank balances, and savings
deposits; and, conversely, in an increased willingness to owe debts
whose amount is fixed in money terms. As I will discuss in more de
tail below, this change in preferences can to some extent be offset by
an increased rate o£ interest on these assets or debts, but for the
moment I want to concentrate on the shift in wealthholding from
money to other types of assets.
H
e d g in g
A
g a in s t
I
n f l a t io n
MISCELLANEOUS ITEMS
In the course of history very different kinds of assets have served as
a hedge against inflation. Gold, the classical favorite, has turned out
to be worse than useless in recent American experience. Because of
the usually high carrying charges, commodity inventories are more
suitable for protection against violent bursts of inflation, such as those
classically associated with the outbreak of war, than against the more
moderate and sustained kind that is now the main problem; in any
case, the prices of primary commodities have tended to fall behind.
On the other hand, real estate, another old favorite, has lost neither
its appeal nor (as far as we can see) its effectiveness. The continued
boom m construction and in urban and rural land has no doubt been
stimulated by the expectation of further rises in the general price
8 In his “ Study of Saving,” vol. I ll, p. I. An annual survey of public and private debt
is already being undertaken by the U.S. Department of Commerce. The tables on financial
assets and liabilities in the flow-of-funds system of the Federal Reserve Board go even
further in the direction here advocated. The introduction of a national balance sheet
might incidentally help in the much-needed reconciliation and integration of the national
accounts and the now-of-funds tables.
THE INCIDENCE OF INFLATION
121
level, though there are many other contributing factors. Foreign cur
rency, on the other hand, has become less attractive because after
W orld W ar I I inflation has become a worldwide phenomenon, varying
only in degree; moreover, the prevailing system of international pay
ments makes exchange rates almost completely rigid.
EQUITIES
Perhaps the most striking feature of the contemporary search for
protection against inflation has been the shift toward equities. Longregarded as close to the bottom of the ladder o f financial respect
ability, common stocks are now the foundation of even the most con
servatively managed portfolios. The resulting increase in demand
for shares has in many cases depressed yields on dividends to a point
below that o f virtually risk-free bonds, though the yield on earnings
is still frequently considerable. Most buyers of shares, however, look
nowadays not only at current earnings, but perhaps even more at
prospective capital appreciation resulting from “ growth.” The under
lying theory, reiterated in hundreds of brokers advertisements and
similar media, appears to be that returns to equity will, in the long
run, be a more or less constant fraction of national output, so that by
investing in those equities the investor can assure himself of a return
that, in real terms, is more or less independent of monetary factors.
Crude though it is, this theory is certainly not without its plausi
bility, nor does it contradict the historical evidence.
The spectacular rise in share prices during the last two decades or
so, however, can only partly be associated with the theory just men
tioned. As the increase in price-earnings ratios shows, an increase in
corporate profits cannot be the only explanation of the stock market
boom. What has happened, rather, is an increased awareness of the
attractions o f equity holdings in a period of full employment and in
flation, and a consequent increase in the number o f people who are
willing to bid for existing securities. Those who bought stocks be
fore the present boom are not only reaping the rewards of better
profits, but also of their early realization of the merits of equities as
a form o f holding wealth.
This explanation of the boom is not correct unless we also look at
the supply o f shares. The increased demand for shares on the part o f
the public has been matched by an increased reluctance on the part of
corporations (that is to say on the part of those large shareholders
who control corporate finances) to issue new shares. The corporate
profits tax, especially, makes it advantageous for corporations to fi
nance their capital expenditures by means of bonds and long-term loans
rather than equities.4
* During the years 1947-58 the total funds obtained by corporations (excluding banks
and insurance companies) were divided as follows:
Source o f funds
A m ount ( billions)
Stocks (including preferred)------------------------------------------------------------------------------$29
Bonds____________________________________________________________________________
52
Other external long-term sources-----------------------------------------------------------------------15
Internal sources (depreciation and retained profits)____________________________ 255
External short-term sources (bank loans, etc.)-------------------------------------------------61
412
Total________________________________________________________________________
Thus only 7 percent of investment was financed from equities.
Data from U.S. Income and Output (1959) and Survey of Current Business (July 1959),
table V -10.
122
THE INCIDENCE OP INFLATION
The rise in stock prices far beyond the rise of prices in general is,
therefore, partly a transitory phenomenon, resulting from a gradual
adjustment in the allocation of wealth among various types of assets in
response to full employment and inflation. What has happened, in
essence, is that share prices have already advanced more closely to
their ultimate level than the prices of other securities. Since, even
now, according to the best estimates, there are no more than some 12
million shareholders in the United States, it is possible that this transi
tion has not yet come to an end, and that the stock market boom will
consequently continue. Like any major price movement, this boom is
also to some extent feeding upon itself, though high margins require
ments and occasional shakeouts have probably prevented the worst
speculative excesses. A t the same time, it appears unlikely that yields
will fall much below their present, already depressed levels. Any
further increase in stock prices must therefore be a result o f improved
earnings.
T h e P attern
of
Y ields
IN THE LONG RUN
So far I have discussed, in very general terms, the changes in the
pattern o f wealth holding that are correlated with inflation. I shall
now try to be a little more specific about the way in which this pattern
is determined. For this purpose I shall assume, to begin with, that
inflation continues at a rate o f perhaps 2 percent or 3 percent per year,
but does not significantly increase its pace. Furthermore it will be
supposed that this rate of inflation is a matter o f general knowledge,
even though some people may be more concerned to protect themselves
against rising prices than others. I also assume that full employment
is maintained, and expected to be maintained, except possibly for oc
casional recessions of the kind which we have had in the postwar period.
The latter assumption, as I shall explain in more detail below, is of
considerable relevance to the present state of the economy.
Under those assumptions it is clear that the rise in the general price
level will be more or less accurately discounted in the returns obtained
from various kinds of assets. For instance, if prices are expected to
rise by 3 percent per year, and people would otherwise be satisfied
with a return of 2y2 percent per year on Government bonds, then the
new return on Government bonds must be 5y2 percent. To take an
other example, if in times of constant prices the yield on good quality
equities is 2 percent above the yield of good quality bonds, and prices
are again expected to rise by 3 percent per year, then under the latter
conditions the yield on equities will be 1 percent below that o f bonds.
Similar reasoning can be applied to almost any other kind of asset.
In the case of cash, for instance, it follows that under inflationary con
ditions there will be a tendency to hold relatively less cash, for its re
turn (apart from convenience) is zero so it becomes less attractive
123
THE INCIDENCE OF INFLATION
compared to other kinds of assets.5 The velocity of circulation in
creases, and this is a crucial link in the chain of events leading to a
rise in the general price level.
The theory just outlined, simple though it is, also seems to be con
sistent with the emerging pattern of yields on capital assets.6 To put
the matter briefly, a high enough rate of interest on bonds will offset
the consequences o f any steady rate of inflation.
From this point of view it might seem as if “ creeping” inflation is
nothing to worry about, since the yields on assets will adjust them
selves more or less automatically to the changes in the general price
level. True, those who favor large Government expenditures might
be concerned if the nominal rate of interest on Government bonds
reaches a level of 5 percent or 6 percent, but actually the burden is
not as large as this figure might suggest, for the real value o f the
taxes from which the bonds have to be repaid falls by a considerable
percentage every year, so that the net real cost of Government borrow
ing is much lower than it appears. There would consequently be no
need to cut down on desirable Government expenditures, as is often
believed. Indeed if inflation proceeds more or less as expected, which
is what I am assuming at the moment, there will not even be any undue
tendency to invest in real assets (such as houses, factories, etc.) rather
than in monetary assets (such as bonds) because the cost o f finance
(such as mortgages) will adjust itself so as to provide as much o f a
deterrent as in times o f unchanging prices.
« Recent experience supports this conclusion. The following table shows the percentage change in cash
(more exactly: currency and demand deposits), total financial assets (more exactly: claims on other sectors)
and in total expenditure, for the U.S. economy as a whole and for some important sectors. It refers to the
period from 1949, when cash was at the lowest point since World War II (presumably because wartime ex
cesses of liquidity had been worked off) through 1958.
Percent change i n Sector
Total
financial
assets
Cash
Consumers and nonprofit organizations____________________
Noncorporate nonfinancial business______________________
Corporate nonfinancial business____________________ ______
Nonbank financial (savings institutions, insurance, etc.)_____
State and local governments______________________________
Economy as a whole______________________________________
24
38
36
69
45
27
113
49
80
130
122
C)
Total
expendi
ture
62
(•)
(•)
(•)
126
71
• Not available.
Source: For “ cash” and “ total financial assets” : Federal Reserve Bulletin, August 1959, pp. 10561061. For “ total expenditure” : Survey of Current Business, July 1959, pp. 6-7.
• Following are the (nominal) yields on some important types of assets in recent years:
Yield in—
1950
U.S. Government long-term bonds___________________
Tax-exempt municipal bonds________________________
Corporate bonds____________________________________
Common stocks (yield on dividends)_________________
• Not entirely comparable with earlier years.
Source: Federal Reserve Bulletin and Statistical Abstract.
2.32
2.00
2.86
6.50
1955
2.84
2.57
3.25
3.93
1959 (July)
4.11
3.92
4.72
®3.11
124
THE INCIDENCE OF INFLATION
SOME RESERVATIONS
The state of affairs just described, in which the rate of inflation is
discounted in all interest rates and other rates o f return, is not com
pletely realized at the present, nor will it be in the immediate future.
It is useful to bear in mind that complete adjustment is conceivable,
for public opinion on inflation is perhaps too much influenced by
gloomy recollections o f a few spectacular inflationary episodes in the
past. Nevertheless we must consider in what respect this ideal differs
from reality.
In the first place complete adjustment can only be expected in the
long run, and especially in the case of inflation it is often doubted
whether there will be a long run. Thus many people believe that
inflation cannot go on at a steady rate year after year, but that it must
necessarily become more and more rapid. As evidence they point to
the spectacular inflations I just alluded to. I do not think this fear
is well founded. The many instances o f rises in the general price
level that did not lead to an explosion largely outnumber those few
episodes. Indeed the most famous of those episodes (the German
experience up to 1923) would probably not have occurred without
the active and deliberate intervention of the German Government.
It is true that in its initial stages, which may last several years, an
inflationary movement tends to gather momentum, especially if gov
ernment policy is ignorant or perverse, but this acceleration is only a
transitional phenomenon.
I would go further and argue that is unlikely that an inflation will
be able to maintain a steady rate; after the transitional phase the net
returns to different kinds of capital wil] once more have come into
mutual equilibrium and the momentum will gradually be spent.
WILL INFLATION CONTINUE?
A t the risk of departing from my main theme, I venture to add my
opinion that the present inflationary period is also mainly of a transi
tional nature. Its mainspring, I think, is the adjustment on the part
o f the public to a policy of full employment. The increasingly wide
spread conviction, reinforced after every shortlived recession, that a
major depression will not recur, calls for truly fundamental changes
in the entire financial sector. Quite apart from any inflationary con
sideration it reduces the relative attraction of cash and other liquid
assets. Liquidity is desired mainly as a precaution against disaster,
and economic disaster has been made less likely for nearly everybody
by Government policy. On the other hand, investment in productive
enterprises, even rather risky ones, has become correspondingly more
attractive. This means an upward shift in the demand for risky secu
rities, of which the present stock market boom is a symptom.
Risk can never disappear completely, but it certainly has become
less evident than it was up to the second W orld War. A policy o f
full employment, for instance, favors the capital goods industries,
traditionally the victims o f depression, relative to the consumer goods
industries, even though the consumer goods industries themselves
THE INCIDENCE OF INFLATION
125
benefit in an absolute sense.7 The heavy industries have always had
a tendency to be highly concentrated, but a full exercise o f monopoly
power was made difficult by the perennial risk of depression. This,
o f course? applied equally well to management as to the labor unions
in those industries. With the advent o f full employment monopoly
power has become a much more serious danger to the economy as a
whole, as witnessed by the record of the steel industry after the war.
The steel firms and unions are not more wicked now than they were
before W orld War II, but they have more scope to pursue their (not
especially immoral) aims.
The consequences o f a full employment policy do not stop here.
The consumer’s outlook, too, is still changing drastically. Consumer
debt has expanded more sharply than any other form o f debt, and this
may well be only the beginning. The recent introduction o f banksponsored charge account and overdraft plans points to further sharp
increases in the velocity of circulation o f money. Home ownership,
which can be more ox a burden than a blessing when hard times
threaten, has come to appear more desirable than ever.
Compared to these powerful forces in the private sector any posi
tive contribution which the Federal Government has made to the in
flationary process is definitely minor,8 though it is also true that by
making money really tight the general price level could have been
kept stable. Perhaps nothing short o f a full-scale depression scare
will stop the inflationary pressures that are now rampant in so many
countries. A tight money policy is desirable to the extent that it helps
to bring the real rate o f interest (that is, the nominal rate after ad
justment for price changes) closer to its equilibrium level. Unless
it is so severe, however, that employers lack the cash with which to
j)ay wage increases, and investors the cash with which to buy addi
tional securities, it will scarcely have much impact on prices in gen
eral. I f monetary policy reaches this degree of rigor it clearly be
comes a serious menace to growth and prosperity.
The above analysis may appear pessimistic to those who are con
cerned about inflation as such. This impression is unjustified, for my
argument is mainly intended to show that although the present infla
tion could probably not have been cured without injury to the patient,
it will sooner or later come to an end. How soon? According to my
guess the length o f time still necessary for general adjustment to a
full employment economy will not exceed 5 years, and may be as
7 From 1947 to 1958 a comprehensive price index for personal consumption expenditures
rose by 33 percent, while a similar index for producers’ durable equipment rose Dy 55 per
cent. (Data from U.S. Income and Output and Survey of Current Business (July 1959),
table VII—2.
8 The justification for this statement, which contradicts widely held convictions, is that
the net Federal debt has not shown any upward trendi during the postwar period; if any
thing it has declined, so the Federal Government can hardly be accused of living beyond its
means. The share of Federal debt held by the banking system (sometimes considered to be
more significant for monetary analysis) has not increased either. On the other hand, ac
cording to figures in the Survey of Current Business (May 1957 and May 1959) the net
debt of State and' local governments rose by 253 percent between 1947 and 1958, of corpora
tions by 127 percent, of individuals and unincorporated enterprises by 238 percent In
the last-mentioned category the rise of small residential mortgages! by 318 percent and of
consumer credit by 290 percent are noteworthy.
It must be admitted, however, that figures on net debt do not tell the whole story. Only
complete sectoral balance sheets could give the necssary information, but they are not
available. Data on financial assets and liabilities (from FRB, August 1959, p. 1057) show,
for instance, that consumers* financial assets rose by 113 percent between 1947 and 1958,
while their liabilities rose by 332 percent. Moreover, of the rise in financial assets at least
two-thirds is attributable to the rise in share prices and does not represent newly acquired
assets.
48084—59------10
126
THE INCIDENCE OF INFLATION
s h o rt as 2 o r 3 y e a rs . W h e n th a t p e rio d is o v e r, th e p a tte r n o f y ie ld s
a n d o f p ric e s w ill g r a d u a lly b eco m e m o re s ta b le (b a r r in g , o f c o u rs e ,
w a rs a n d o th e r u n fo re s e e n c a la m itie s ).
E c o n o m is ts w it h m o re k n o w le d g e o f th e m o n e ta ry s p h e re th a n I
c a n c la im w ill h a v e to d e c id e o n th e v a lid it y o f m y d ia g n o s is ; th e
v e r d ic t m a y e v e n h a v e to b e le f t to h is to ry . I n a sense th is d ia g n o s is
is o n ly in c id e n ta l to m y s u b je c t, th o u g h i t does h a v e a n in d ir e c t b e a r
in g o n th e v a lid it y o f som e o f m y p ro p o s a ls . L e t u s n o w r e tu r n to
th e m a in a rg u m e n t.
DIFFERENCES IN EXPECTATIONS
So far the discussion has proceeded mainly on the assumption that
inflation is generally recognized and more or less accurately expected.
This was only a simplified assumption, needed as a foundation for
statements about the long run. It is now time to remove the assump
tion.
A s soon as w e d o so w e h a v e to fa c e th e fa c t t h a t p e o p le m a y h a v e
d iffe r e n t e x p e c ta tio n s as to th e r a te o f in fla tio n th a t is g o in g to p r e
v a il in th e fu tu r e . M o re o v e r, th e m e re u n c e r ta in ty ox th e r a te o f
in fla tio n h a s im p o r ta n t consequences. T h e s e tw o c o m p lic a tio n s in te r
a c t a n d n e e d n o t b e d isc u s se d s e p a ra te ly .
A d iv e rg e n c e c o n c e rn in g th e e x p e c te d r a te o f in fla tio n is n o th in g
u n u s u a l in s e c u rity a n d c o m m o d ity m a rk e ts . In d e e d d iv e rg e n t o p in
io n s a re th e essence o f s u c h m a rk e ts . A m o n g th e m a n y m o tiv e s w h ic h
m a k e p e o p le b u y a n d s e ll s e c u ritie s , th e a n tic ip a te d r a te o f in fla tio n
is a lre a d y o n e . N e v e rth e le s s i t is im p o r ta n t to r e a liz e t h a t th e re is a t
th e m o m e n t n o m a r k e t w h ic h is e x c lu s iv e ly d e v o te d to re c o n c ilin g
d iffe r e n t o p in io n s c o n c e rn in g th e c o u rs e o f p ric e s in g e n e ra l. T h e re
a re m a rk e ts (th e c o m m o d ity fu tu re s m a r k e ts ) in w h ic h o p in io n s o n
th e c o u rs e o f in d iv id u a l p ric e s , e s p e c ia lly o f r a w m a te r ia ls a n d fo o d
s tu ffs , a re b ro u g h t in to b a la n c e w it h e a c h o th e r a n d w it h p ro d u c tio n ,
c o n s u m p tio n , a n d in v e n to rie s . T h e to ta l c o v e ra g e o f th e s e m a rk e ts ,
h o w e v e r, is to o s m a ll to p ro v id e a h e d g e a g a in s t th e m o v e m e n t o f th e
g e n e ra l p ric e le v e l; m o re o v e r, th e la r g e s iz e o f in d iv id u a l tra n s a c
tio n s , a n d th e ir r e s tr ic tio n to p e rio d s o f 1 2 o r 18 m o n th s a h e a d , p r e
c lu d e t h e ir u se b y th o s e w h o a re n o t s p e c ific a lly in te re s te d in th e c o m
m o d itie s tr a d e d . T h e b a s ic p u rp o s e o f th e s e m a rk e ts , fu r th e r m o r e , is
p ro te c tio n a g a in s t p ric e fa lls f o r h o ld e rs o f in v e n to rie s , r a th e r th a n
p ro te c tio n a g a in s t p ric e ris e s f o r p ro s p e c tiv e b u y e rs . T h e la t t e r k in d
o f p ro te c tio n c a n a ls o b e p ro v id e d b y fu tu re s c o n tra c ts , b u t o n ly a t
c o n s id e ra b le c o s t; i t is u s u a lly a c h ie v e d m o re e c o n o m ic a lly b y h o ld in g
in v e n to rie s . A fu tu re s m a r k e t c o n c e rn e d w it h th e g e n e ra l p r ic e le v e l
w o u ld th e r e fo r e m e e t in s u p e ra b le te c h n ic a l d iffic u ltie s . A s w e s h a ll
see b e lo w , a m o re w o rk a b le s u b s titu te c a n b e d e v is e d .
O n th e s to c k m a r k e t, w h e re fe a r o f in fla tio n is o n e o f th e fa c to rs
d e te rm in in g s to c k p ric e s , m a n y o th e r fa c to rs e x e rc is e a n e v e n g re a te r
in flu e n c e o n th e p ric e o f in d iv id u a l s e c u ritie s . I n fa c t, th is is o n e
o f th e re a s o n s w h y c o m m o n sto c ks d o n o t p ro v id e a n a lto g e th e r s a tis
fa c to r y p ro te c tio n a g a in s t in fla tio n , e v e n i f th e p re m is e m e n tio n e d
e a r lie r ( t h a t p ro fits w ill c o n tin u e to c la im a n a p p ro x im a te ly c o n s ta n t
s h a re in n a tio n a l in c o m e ) is g ra n te d . T h e r e is a w id e d is p e rs io n in
th e re s u lts o b ta in e d fr o m d iffe r e n t s e c u ritie s .9 I t is tr u e t h a t e v e n
®Cf. Append!*.
THE INCIDENCE OF INFLATION
127
people o f modest means can achieve diversification by buying shares
m investment trusts and similar institutions, but they have to pay
rather heavily for the management skills thus bought. The yields
on investment trusts are consequently quite low, often still lower than
those on the bluest o f the blue chips.10 Unless the investor is prepared
to pay this price for management, he will have to familiarize himself
with the merits o f a large number of individual companies. On the
average investors will probably do very well, but the risk o f choosing
a bad company remains.
The bond market situation is also unsatisfactory in this respect.
The heavy participation by financial institutions, especially banks,
means that the determination of bond prices reflects considerations
which are not of much importance to most individual investors. It
may be admitted that insurance companies, another important cate
gory of bondholders, ideally reflect only the preferences o f their
policyholders; on the other hand the long-term nature o f many insur
ance contracts, and the strict legal regulations to which the industry
is subject preclude a very close adjustment of the insurance com
panies’ policies to the needs of their customers.
INSURANCE POLICIES LIN K ED TO SHARE PRICES
In this connection it might be useful to comment briefly on the
plans now materializing to issue insurance policies whose proceeds
are linked to the value o f a portfolio o f selected stocks. One such
scheme was started some years ago by the Teachers Insurance and
Annuity Association, while a major life insurance company was re
cently authorized to go ahead with similar plans. The T IA A plan
has met favorable response among its special public, and it seems
likely that the newer and more general proposal will also find wide
acceptance. Despite opposition within the insurance industry, similar
plans will presumably be introduced by other companies. I f these
plans are, in fact, successful, they will offset a trend that has caused
considerable anxiety to the industry, namely the shift away from
equity-building insurance (such as endowment policies) toward term
insurance.11 Although an exact evaluation of this movement is made
difficult by the simultaneous shift from individual to group policies,
it would seem that fear o f inflation has been a major cause o f the
decline o f traditional insurance.
Those who devised the new insurance schemes deserve credit for
understanding the signs o f the times; nevertheless, it must be doubted
whether this solution will be satisfactory in the long run. So far
stock prices have risen much more than prices in general, but this
hardly suggests that they will continue to do so. On the contrary,
if my diagnosis is correct, the rise in stock prices, though partly based
on real factors and not merely on fear o f inflation, has brought equities
10 Of 24 investment funds for which the price on Dec. 31, 1958 was related to dividends
paid in 1958 the yield was less than 2 percent in 6 cases, from 2 to 3 percent in 10 cases,
from 3 to 4 percent in 7 cases, and! over 4 percent in one case. Many of these funds also
pay out capital gains from time to time.
11 From data in the “ Life Insurance Fact Book 1959,” it can be estimated that o f the
total life insurance in force in the United States 45 percent was term insurance in 1957,
against only 30 percent in 1950. If group and credit insurance are left out of account the
percentage of term policies in the remainder (ordinary and industrial) rose from 11 per
cent in 1950 to 17 percent in 1957.
128
THE INCIDENCE OF INFLATION
closer to their long-term equilibrium values than prices in general, and
may already have pushed them beyond this level. Consequently, the
stock market does not provide an ideal hedge for such inflation as is
still to come. The stock market, in other words, has already dis
counted a good deal o f future inflation. And, even apart from these
necessarily speculative considerations, it is not at all clear that those
who buy insurance policies to provide for their needs in old age are
well served by having their fortunes linked to the vagaries of the stock
exchange.
T
he
N
eed
fob
A
ssets
This brings us to the heart of the whole subject—protection against
inflation. What types o f assets do individuals really need ? This is
almost the same question as, W hy do people save ?
There are three main reasons why people save, and each o f them is
associated with particular kinds of assets:
1. To have a reserve against unforeseen contingencies.—This need
can be met by having highly liquid and comparatively risk-free re
sources, especially cash and bank deposits, though an overdraft agree
ment with a bank (o f the type that is now becoming popular) would
be almost as suitable. For this type of asset, which is probably of
minor quantitative importance, yield is not an important consideration.
2. To Smooth out anticipated fluctuations in income arnd consump
tion, especially those associated with old age, but also with such things
as a college education for children, or the building o f a house. Here
liquidity is o f less significance, but freedom from risk is important,
while yield is a close second. Now, “ freedom from risk” needs to be
carefully defined. It does not necessarily mean that a certain number
o f dollars will be available when anticipated, but rather that the indi
vidual is able to satisfy those needs for which he had saved. Some
times, especially in the case o f housing, this can be arranged by pur
chasing ahead o f time, as it were, but there exist no facilities for the
delivery o f food or clothing 20 or 30 years from now. Hence, saving
in some form is necessary. Here again we must recall that risk can
never be avoided completely. By buying Government or prime cor
porate bonds, for instance, one can virtually avoid the risk o f default,
but the risk o f changes in the price level remains. There is nothing
in the history of the United States or of any other country that war
rants the assumption o f a stable price level in the long run. By buy
ing equities one can probably achieve partial independence from
changes in the price level, but other kinds of risk are thereby incurred;
much the same applies to real estate as a form of wealth. The choice
between those various kinds o f risk depends on their evaluation by the
individual, and on his ability to cope with them. People who know
nothing about corporate finance may shy away from the stock market
and put their faith in assets with a fixed money value, thus exposing
themselves to the risk of inflation.12 A t the moment, there is no way
out from this dilemma, although I shall discuss one later in this paper.
The willingness to assume various types of risks also depends on the
income and wealth of the individual concerned; people with large
incomes or fortunes evidently can take greater risks than others. Until
12 This, in fact, is the basic virtue of a stable price level: that it permits investors
without special knowledge of any line business (rentiers) to earn a (necessarily modest)
return on assets with a fixed money value.
THE INCIDENCE OF INFLATION
129
recently shares were considered the most risky assets, so their owner
ship was largely confined to high-income groups. Now that risk
evaluation has cnanged drastically; their ownership tends to be much
less concentrated.
As a result some shares are now held by individuals who are not
financially able to shoulder the risk o f corporate ownership, though
the gradual realization of “ people’s capitalism” should perhaps oe
welcomed on other grounds. Finally it should be borne in mind that
to some extent risks can be offset by yield; thus if the rate of interest
on Government bonds were high enough the risk of inflation might be
amply offset, and if share prices were lower the ownership o f equities
could be recommended to people o f modest means who at present prices
do not stand to gain much considering the risks they would take.
3. To build up an estate.—Although not always recognized as such,
this motive appears to be more important in the aggregate than the
provision for old age. It is true that the active desire to build up an
estate is confined to a rather small section of the population, but this
section does seem to account for the bulk o f savings and assets. For
this purpose liquidity is o f small significance, and freedom from risk
is also less important. Yield, whether in the form o f current return
or o f capital appreciation, is the primary consideration.
T w o P rop osa ls
I have given most space to the second motive for saving, because
this is the one most affected by inflation. The usefulness o f cash and
bank deposits to satisfy the first need (that for a reserve against
emergencies) is little affected by creeping inflation, while the third
motive (estate building) is mostly pursued by means o f shares, owner
ship o f unincorporated business, and real estate, all o f which are more
or less immune. It is clear, however, that even a modest rate o f in
flation may interfere seriously with provision for old age and similar
long term needs. There are two steps that can be taken to improve
this situation.
INDEX BONDS
The first step is the introduction of Government bonds whose rate
o f interest and redemption value are linked to a retail price index
number; I shall refer to such bonds as “ index bonds.” This is not
by any means a new proposal; in fact, its history can be traced back
at least to Alfred Marshall in the late 19th century. Nevertheless
there have been only a few attempts to put it into practice.13
I should make it clear at once that I do not propose index bonds
as a means o f combating inflation. The forces that have caused prices
in general to rise are too strong, in my opinion, to be overcome by any
such device if introduced on the modest scale that I have in mind.
I am not suggesting that the whole Government debt, not to mention
private debt, be put on an index basis, as has apparently been at
i3 under the category of index bonds I do not count bonds that are linked to the value of
gold in some w ay; they would be of little help in present circumstances.
130
th e
in c id e n c e
of
INFLATION
tempted in Finland .14 What I am contemplating, rather, is the issue
of some new series of Government bonds, along with a continuation
o f present types.
In the field o f savings bonds especially, indexing might improve
matters considerably. There has been a fairly steady decline in the
amount outstanding o f savings bonds; one suspects that the decline
would be even larger if a considerable portion o f savings bonds were
not sold by payroll deduction plans. A large proportion of bonds are
cashed in after as little as two years.15 It is conceivable that the
rather small increase in interest rates recently authorized by Congress
will stimulate the savings bond program somewhat, but it would be
surprising if it made sales exceed redemptions.
I f savings bonds with an index clause were issued, the rate o f in
terest should not be more 2% percent. Judging from the dividend
yield on blue chip shares, many people would be happy with this
modest return if it were guaranteed in real terms; it also agrees with
the current real rate o f interest on Federal bonds (a nominal rate
between 4% and 4% percent less an annual rate o f inflation of some
what under 2 percent). The low rate of interest does not necessarily
mean that the Government will have a bargain, for the index clause
may turn out to be expensive if prices rise much. It is sometimes
held that index bonds will encourage the prevention of inflation be
cause of its impact on the budget when index bonds are outstanding,
but I would not put too much stress on this argument. The introduc
tion of index savings bonds has to be viewed primarily as the filling of
a gap in the range of financial instruments and as a service to the
many people who are providing for old age on an individual basis.
It may also have a favorable influence on the volume o f saving, but
here again it would be imprudent to claim much.
Fortunately, if my diagnosis of the present monetary situation is
correct, the Government will not be forced to pay out large sums on
account o f the index clause. I f inflation comes to an end within the
next decade^ the public would have benefited not only from greater
security in its old-age planning, but also from a lower cost o f the
Government debt. And if inflation does not come to an end, index
bonds would clearly be even more useful.
Savings bonds, which are held by individuals are an obvious choice
for the introduction of indexing on a modest scale.16 There is no
reason, however, to confine this device to savings bonds.17 The in
14 See the unsympathetic account in an article entitled “Creeping Inflation” in the June
1959 issue of the Monthly Review of the Federal Reserve Bank of New York. Unfortu
nately, the analysis in this article could hardly be more superficia'
For an intelligent,
though almost equally one-sided, discussion of index bonds, primarily from a theoretical
point of view, see Guy Arvidsson, “ Reflections on Index Loans,” Skandinaviska Banken
Quarterly Review (Stockholm), January 1959.
15 Of the series E bonds sold in 1957, for instance, more than half had already been
redeemed on March 31, 1959 (Source: mimeographed statement from the Fiscal Assistant
Secretary of the Treasury).
16 From an administrative point of view there is an advantage in discount-type bonds—
such as the present series E— with an index feature ,* it would then be unnecessary to have
interest coupons with variable money value and only the redemption value would have to be
readjusted. To discourage early redemption it might be advisable to apply indexing only
to the redemption value after a certain number of years. Furthermore the introduction of
special bonds (at a lower rate of interest than regular index bonds) with a minimum
redemption value might be considered, since some people might feel that the risk of price
falls makes index bonds too speculative otherwise.
17 Index savings bonds could be marketed through existing channels, including payroll
deduction plans. In general, however, I see no harm in having index bonds of larger
denominations traded over the counter or on the stock exchange; indeed such negotia
bility would enhance their attractiveness to investors.
THE INCIDENCE OF INFLATION
131
troduction o f index bonds with larger denominations and longer
maturities might enable insurance companies to offer policies whose
return is based on the Consumer Price Index .18 Such policies would
probably be preferable to the stock market linked policies that are
now coming into existence. (See the discussion above.)
It is sometimes argued that indexing cannot be introduced by any
government without destroying confidence in its own currency. This
view, I think, is based on an unrealistic conception o f contemporary
monetary management. As I have argued earlier, the role of the
Government in fostering inflation is relatively small, except to the
extent that the increased demand for certain assets by the public and
the increased circulation velocity o f money are due to a full employ
ment policy. In history there have been many instances in which the
Government budget was a prime contributor to inflation, but this does
not seem to be the case in the United States at the present time. No
doubt the Government and its instrumentalities have an obligation to
keep the value o f money as stable as possible, but this obligation has to
be reconciled with other objectives of Government policy. The intro
duction o f indexing would be no more than the recognition o f a fact.
This is not to deny that a certain amount of public education would be
necessary before such a step could be taken.
As far as the domestic economy is concerned it must be doubted,
therefore, whether confidence in the dollar would be further reduced
by the introduction of index bonds. The official attempts to exorcize
inflation by professions of abhorrence have failed, and at the moment
no policy with greater promise of results appears to be available; the
time for realism seems to have arrived. On the international scene
the situation is not basically different. It is true that public recog
nition o f the possibility of inflation might cause international finan
ciers, who are a nervous crowd, to withdraw some “ hot money” from
New York. But this would be only a temporary reaction; it is even
conceivable that index bonds, which at the moment are not available
in any major country, would attract foreign long-term capital to the
United States.
Another possible objection to index bonds is that they represent more
“ escalation,” and thereby promote price rises. This objection is based
on the now fashionable cost-push theory o f inflation which I need not
refute here.19 From the viewpoint adopted in this paper escalation
is not itself a cause o f inflation but at best one o f the factors determin
ing the speed at (rather than the degree to) which the underlying
causes o f inflation affect the different sectors o f the economy. By
slowing down the adjustment to these causes one only creates dispari
ties ; hence escalation need not be feared.
INCREASING T H E SUPPLY OF SHARES
The second measure to improve the possibility o f providing for old
age in an inflationary period which I would recommend is an increase
in the supply o f shares. Although equities are not a suitable invest
18 There are some technical problems concerning the choice of a suitable price index,
but they are not serious. Any reasonably comprehensive index number would do for
practical purposes.
19 See, for instance, the valuable article by Richard T. Selden in the Journal of Political
Economy of February 1959.
132
th e
in c id e n c e
of
in f l a t io n
ment medium for all individuals, it would probably be in the public
interest if they were more widely held than they are at present, and
they have considerable merit as a hedge against inflation. A t present,
however, share prices are too high to warrant investment by small
savers. The supply o f shares could be increased, and prices reduced,
by making corporations more willing to finance new investments from
shares rather than from bonds or bank loans. As I mentioned earlier,
a major deterrent to the expanded issue o f new shares is the fact that
the corporate income tax treats interest as deductible. Hence, the true
cost o f dividends to the corporation is relatively higher than that of
interest on bonds and loans. To take the figures of footnote 5, the cost
before tax o f dividends to the corporation, at a 52-percent tax rate, is
6.5 percent when the yield to the share owner is 3.1 percent, as com
pared to a pretax (and posttax) cost o f 4.7 percent of interest on bonds.
Even now substantial numbers o f new shares are being issued, but
probably not as many as would correspond to the high level o f stock
prices.20
The issue of new shares could be stimulated, if this argument is
correct, by changing the deduction rules for the corporate income
tax. Interest paid by corporations, according to this proposal, would
no longer be fully deductible; instead it would be deductible only to
the amount o f taxable interest received; in other words, net interest
paid would not be deductible.21 A t the same time the percentage rate of
the tax would be adjusted downward, since the purpose o f this change
would not be to increase the total tax burden on corporations. To
safeguard payment o f interest on senior obligations, it would be made
possible to obtain a tax deferment to the extent necessary to pay such
interest in case of insufficient income. An incidental consequence of
this change, and one which most economists would probably applaud,
is a modest reduction in the marginal tax rate on corporations, which
would probably lead to a greater effort to cut down unnecessary ex
penses and a greater willingness to bear risk.22 The principle that
taxation should not interfere with sound rules of corporate finance
will also find general support.
The introduction o f this change in the deductibility o f interest
would have to be gradual, for the proportion of junior to senior ob
ligations varies between firms so that the tax payable by some firms
might change suddenly even though the total yield o f the tax remains
unchanged; a sudden change in the tax base might therefore be dis
ruptive. One way of arranging a smooth transition would be to leave
the present deduction rules in force for interest on debts contracted
before a cutoff date and apply the proposed rule only to debts con
tracted or renewed after that date. The tax rate could then be reduced
^Footnote 6 also shows that the advantage of bonds over stocks has narrowed con
siderably in recent years, although it is still appreciable.
21The deduction of interest paid from interest received would be allowed since other
wise financial corporations, whose business it is to lend money obtained by borrowing,
would be placed in an impossible position. Only taxable interest received would be
considered, since tax-exempt interest already enjoys special consideration.
22 If in 1956, the last year for which corporate income tax data are available, the pro
posed change had been fully in effect, taxable income of $46.9 billion, would have been
increased by about $3.5 billion or 7.4 percent; the last-mentioned figure was estimated
by adding the interest paid and subtracting the taxable interest received in all industrial
groups for which there was a net outflow of taxable interest. The average tax rate of
45.6 percent could consequently have been reduced to 42.5 percent, and the marginal tax
rate of 52 percent correspondingly to 48 percents
THE INCIDENCE OF INFLATION
133
gradually over a period of years to reflect the proportion o f debt under
the old and the new rules. A detailed study o f corporate debt would
help in establishing the new schedule o f tax rates.
TH E GOVERNMENT BOND M ARKET
The two proposals just made clearly will have some influence on
the market in Federal bonds, about which there is much concern at
the moment. The introduction o f index bonds will presumably lead
to some amelioration of the Federal Government’s difficulties in
placing new savings and long-term bonds; unfortunately there is no
basis for forecasting the demand for such bonds. A questionnaire
survey could yield considerable information on public interest in them.
Recent events indicate the existence of a large demand for conven
tional bonds at an interest rate around 4% percent, but if my diagnosis
concerning the future course o f inflation is at all accurate, the Federal
Government would be ill-advised to attempt the placement o f long
term bonds at so high a rate, even if Congress could be persuaded to
permit this. I f inflation ends within the next 5 years bond interest
will fall from its present level, which includes an allowance for a
decline in the value of money. Index bonds would not reflect this
factor, and would consequently present a convenient solution to the
immediate problems in Federal finance.
A successful attempt to stimulate new share issues through changes
in the tax laws may have an even greater impact on the bond market.
It has often been pointed out that the Federal Government is at a
disadvantage there compared to; two other important classes o f bor
rowers. State and local government bonds are tax exempt and hence
carry a lower rate of interest; one curious consequence o f this is that
these authorities find it more and more profitable to own Federal
bonds.23
Whether anything should be done about this is a problem outside
the scope o f this paper.
The proposed change in the corporate income tax, on the other
hand, will clearly improve the competitive position o f the Federal
Government, even though that is not its primary purpose.24 The ten
dency on the part o f important investors, notably insurance companies
and mutual savings banks, to hold fewer Federal bonds and more
corporate bonds, could perhaps be halted or reversed. It should not
be forgotten, in this connection, that the rise o f corporate bond
financing stems only partly from fiscal factors, and reflects also the
general decrease in risk resulting from a policy o f full employment.
83 From 1947 to 19i58 their holdings increased from $7.3 billion to $17.3 billion, in
marked contrast to the holdings of most other classes of lenders which declined (source:
Federal Reserve Bulletin),.
24 There are no good grounds for giving special help to the Federal Government in the
bond market, but neither is there any reason why it should continue to be at a disad
vantage because of regulations that are harmful in themselves, such as the tax rules con
cerning corporate interest payments. Because of the many side issues involved I have
not gone into the deductibility of interest in the individual income ta x ; evidently roughly
similar considerations apply there, but they may be to some extent offset by the desirability
of homeownership from a social point of view.
APPENDIX
T h e R etu rn F rom C o m m o n S to cks
In order to provide insight into the net results of investment in common stocks
a sample of equities quoted on the New York Stock Exchange has been analyzed.
In principle the sample was randomly selected; the need for quotations extend
ing over several years, however, restricted the choice somewhat. This factor,
and also the criteria for quotation on the New York Stock Exchange, no doubt
biased the sample in the direction of better investment quality.
For each stock the total yield was calculated by averaging the total dividends
paid and reducing the capital gain (or loss) to an equivalent annual rate of
interest; thus a stock which 4 years ago cost $100, had paid a total of $10 in
dividends and now sells at $1041/16 had a yield of 2 percent and a capital gain
equivalent to a 1-percent rate of interest, giving a total return of 3 percent.
According to the usual convention stock dividends of less than 25 percent were
at cash values; larger stock dividends were treated as splits. Taxes were
not taken into account.
This calculation was made for examples of stocks bought in 1925, 1930, 1935,
etc., through 1955 and sold in 1959. The date in each year was close to January
15. Of the stocks bought in 1925 and 1930, some disappeared through bank
ruptcy.
The results are as follows:
Total return from stocks from year indicated to 1959
Year of purchase
Annual return
1925
Negative______ _______ ___________________ 21
22
0 to 5____ _____________________ _______ —
5 to 1 0-__________________________________ 35
10 to 15___________________________________ 24
15 to 20___________________________________ 13
20 to 30__________________________________
10
30 to 40_____ _____________________________
7
40 to 50______________________________ ____
4
50 and over_______________________________
6
Total number____________________________ 142
Mean return______________________ _______ i 7.43
Median return____________________________
8.95
1930
1935
1940
1945
1950
9
40
40
15
14
7
6
1
2
134
2 6.75
6.91
4
9
20
26
19
29
8
7
6
128
24.65
15.77
0
4
36
25
20
17
9
4
10
125
23.64
14.11
0
12
22
20
23
33
9
2
3
124
18.49
17.30
2
6
18
15
25
37
14
6
3
126
20.62
16.54
1956
8
14
20
23
25
21
9
5
1
126
15.81
14.27
1 The mean return would be 15.56 if the 10 stocks that became a total loss are ignored.
2 The mean return would be 13.58 if the 8 stocks that became a total loss are ignored.
This table shows that there is considerable dispersion in the results from
different stocks, and hence considerable risk, but that by and large, even over
long periods, stockholding has been rewarding enough. The median return is
nearly always below the mean return, indicating that a majority of stocks do
less well than the mean; even at the median the return is substantial.
It is of some interest to separate the two components of total return, which
was done for 2 purchase years as shown in the following table:
134
135
THE INCIDENCE OP INFLATION
Returns from dividends and from capital gains or losses on stocks "bought in
1925 and 1955
Equivalent capital gain
Dividend yield
Stocks bought in 1925 i
Neg
ative
0 ........................................
0 to 2 _ ..............................
2 to 4 ................................
4 to 6 ...............................
6 to 8..................................
8 to 10................ - ..............
10 and over__ __________
2
11
5
Total_____________
0-5
Stocks bought in 1955
Neg
5-10 10-20 Total ative
1
4
7
10
16
19
11
10
1
3
4
20
19
77
28
0-5
40
5-10- 10-20 20-40 and Total
over
2
2
4
8
4
1
1
3
12
5
1
3
1
1
3
10
7
2
8
6
18
15
17
22
15
39
8
132
20
26
24
2
1
1
4
17
8
3
3
7
4
3
1
37
16
3
1
1
7
6
15
54
29
9
6
126
* Excluding shares of corporations that subsequently failed.
The left-hand half of this table suggests that in the long run (or at least
from boom to boom) dividends are more important than capital gains; in fact
capital gains are positively correlated with dividends, which agrees with eco
nomic theory. In the short run capital gains are more important and their
correlation with dividends is weaker.
STU DY PAPER NO. 9
TH E
S H A R E O F W A G E S A N D S A L A R IE S I N
M A N U F A C T U R IN G IN C O M E S , 1 9 4 7 -5 6
( B y A l f r e d H . C on ra d )
137
STUDY
PAPER
NO. 9
THE SHARE OF WAGES AND SALARIES IN
MANUFACTURING INCOMES, 1947-56
(By Alfred H. Conrad, Harvard University, Cambridge, M ass.)1
1. This study is concerned with the relative shares of wages and
gross profits in manufacturing value added in the period 1947-56.
1 will first discuss the way in which the shares would be expected to
change under each of the main theories of the inflation process. Then
I will outline the changes that have occurred in the distribution be
tween wages and nonwage income (1) in the whole economy, (2) in
the manufacturing industries as a group, and (3) among manufac
turing industries taken singly. Finally, I will use the detailed in
dustry statistics to attempt to test some of the main hypotheses about
wage-setting in periods of high-level employment.
2. The relationship between income shares and inflation is not a
simple one. The share going to wages depends upon the wage rate,
the productivity of labor, and the response of finished product prices
to changes in wages and material costs. Each of these, in turn, may
depend upon the pressure of demand, the degree of monopoly control,
Government policies, and other influences in the economy. Since we
are dealing with shortrun changes, I will not consider the major
longrun or equilibrium determinants of income shares, such as the
marginal productivities of capital and labor inputs.
3. A few years after World War X, John Maynard Keynes pointed
out that the social effects of inflation are important—
only insofar as its incidence is unequal * ♦* [A] change in prices and reward^
as measured in money, generally affects different classes unequally, transfers
wealth from one to another, bestows affluence here and embarrassment there,
and redistributes Fortune’s favors so as to frustrate design and disappoint
expectation.2
His description of the process revolves about the relative abilities of
the investing class, the entrepreneurial class, and the wage earners in
ensuring that their respective incomes at least keep up with and,
possibly, move ahead of the prices of the things they buy.
4. There is another aspect of the keeping-up process which may be
quantitatively more significant than the social consequences. The
speed with which economic groups are able to recoup losses in their
respective shares of income is one of the most important determinants
1 This study was begun at the Harvard Economic Research Project as part of the proj
ect’s work on labor requirements and income generation. I have benefitted greatly from
discussions with Prof. John R. Meyer and from the patient assistance of Mrs. V. McK.
Nail, who prepared the raw data, and Miss Beverly Scott, of the Littauer Statistical
Laboratory, who carried out the final computations.
* J. M. Keynes, “ Social Consequences of Changes in the Value of Money (1923), in Essays
in Persuasion, London, 1957, pp. 80-104.
139
140
THE INCIDENCE OF INFLATION
of the speed of inflation. Putting this in “cost-push” terms, we should
saj that the speed with which workers are able to respond to cost-ofliving changes and entrepreneurs to changes in wages and the cost of
materials will determine the rate at which prices rise in the ecenomy.
The cost-push theories imply that once any major cost has gone up and
stayed up, other prices will respond, creating a second round of cost
increases, and so on round again. The novelty in such explanations
has been that they explicity do not require that demand for com
modities or labor be increasing or pressing upon capacity. Indeed,
we know that it is possible for prices to continue rising m a period
of falling output and increasing unemployment. The cost-push
theories have been developed to explain the experience of declining
industries and the rigidity of prices in recessession years in the post
war inflation. If it is true that trade unions are responsible for the
major cost-push, then we should expect the wage share in produced
income to increase. Only if prices were raised in proportion, not
simply in absolute amount, to compensate for the wage increase would
the wage share remain stable. That is, if gross profits are also going
up, the wage share will not increase significantly, and the trade union
cost-push explanation must be held in serious question. As a matter
of fact, this is precisely what the study shows: the share of wages in
manufacturing value added remained remarkably stable from 1947 to
1956-
5. In the orthodox demand-pull explanation, an increase in demand
more rapid than the increase in productive capacity causes prices to
rise. Then, especially if the money supply is allowed to increase
rapidly, or people are willing to hold money substitutes, incomes will
increase and, subsequently, so will expenditures. If wages do not
respond quickly to the consumer-price rise, profits and investment
demand will increase. If this explanation, taken alone, were true,
declines in employment and demand would be sufficient to stop the
rise in prices. The aggregate demand-pull theory implies that the
wage share declines as wages lag behind prices.
6. On the individual industry level, the demand hypothesis is less
simple. We may interpret it to mean that wages and prices in specific
industries responds to demand pressures upon the industries. Or, we
may interpret the wage-price behavior asymmetrically over the cycle;
that is, during economic expansion wages everywhere follow the in
crease in the most rapidly expanding sectors, while in the down
swing, only severe overall unemployment will make specific wages
responsible to specific industry demands.
7. Neither the simple cost-push nor the simple demand-pull ex
planation is wholly satisfactory. Once inflation starts, especially if
there are downward rigidities in important wages and prices, it is al
most impossible to say how the inflation got started or is maintained.
The discussion tends to break down into a chicken-or-egg controversy
or—and this is more important—into a struggle for public policy
aimed at one or another set of villains. Actually, both processes are
likely to be going on simultaneously, each feeding the other, in differ
ent parts of the economy. Operationally, the question of cause and
effect should not be one of relative villainies, but of what policy
weapons will be effective in stopping the steady rise in prices. Un
less we are willing to impose direct controls on wages and prices, the
way to stop inflation must be found by influencing those economic
THE INCIDENCE OF INFLATION
141
conditions which influences the short-run movements of wages and
prices. Finding cures, in this sense is equivalent operationally to
finding the cause of inflation.
8. The evidence on the distribution of national income in the post
war inflation does not show any large-scale shift among the major
income categories. The wage and salary share gained about 4.5 per
centage points between 1947 and 1956; corporate profits did not change
perceptibly; the relative loss fell heavily upon unincorporated busi
ness. The reduction in the income of unincorporate business is hard to
assess, for two reasons: (1) It is not at all clear how much of the
income of proprietors of unincorporate business ought to be counted
as wages and how much as entrepreneurial returns; (2) the sharp
decline in the farm income share is accounted for largely by the with
drawal of farm proprietors from agriculture to other occupations.
During World War II, the increase in farm proprietors’ income (per
sonal income per recipient) was more than twice as rapid as the in
crease for nonfarm proprietors and more than three times the em
ployees’ gain. The relative decline after the war years may, there
fore reflect some catching-up on the part of employees.3
9. Within the corporate profits share there are other ambiguities.
The most important possible source of error arises from the use of
historical values as the base of depreciation charges in a period of
rising replacement costs. In 1947, it is estimated by the Machinery
and Allied Products Institute, the ratio of current prices to average
prices underlying historical-cost depreciation of plant and equipment
was 1.44; in 1956, it fell to 1.31. In manufacturing, the ratio de
clined from 1.54 to 1.38, between 1947 and 1955. The difference be
tween the share of property income in manufacturing net income
under current value as opposed to book value depreciation is an
almost constant 2 percent. There are two observations to be made
on the basis of this data: (1) Profits net of depreciation charges will
be overstated in the light of historical changes in plant and equip
ment costs. Gross profits will contain a larger part necessary to
provide for replacement than would appear in the income accounts.
(2) To the extent that accelerated depreciation arrangements, such
as the sum-of-the-digits method, do not compensate for these changes
in replacement costs, manufacturing corporations are pressed to add
to their markups over direct costs an amount sufficient to provide for
the added cost of new equipment.4
10. When we turn to the manufacturing sector the appearance of
stability in the wage and salary share becomes even more striking.
Between 1947 and 1956 the share of total compensation of employees
in value added (which is gross of depreciation) increased from 53.4
percent to 55.2 percent. Within the period, however, the short-run
variation was much wider. In 1949 the wage and salary share rose to
56.7 percent and in the following year, fell back to 53.4. The share
rose again until in 1953 and 1954 it was 56.4 percent, after which a
decline of 1.6 percentage points occurred.
* Cf. G. L. Bach and A. Ando, “ The Redistributional Effects of Inflation,” Review of
Economics and Statistics, February 1957, especially, pp. 4, 5.
4 Department of Commerce, Survey of Current Business, November 19)56, pp. 11, 20, and
Machinery and Allied Products Institute, Capital Goods Review, No. 29, quoted in Joint
Economic Committee, “ Productivity, Prices, and Incomes,” Washington, 1957, pp. 99-101.
48084— 59-------11
142
THE INCIDENCE OF INFLATION
11. The proportion of value added going to production workers as
opposed to the share of all employees, declined markedly from 42.7
percent in 1957 to 37.3 percent m 1956. Part of this decline was due
to the rapid increase in the number of administrative and profes
sional workers in industry. In 1947, salaried overhead labor made up
16.6 percent of the manufacturing work force. The proportion was
up to 23.5 percent by 1956. The growth of salaried employment in
manufacturing was 55 percent between 1947 and 1957, more rapid
even than the increase in trade and service employment. The cor
responding relative decline in the employment of production workers
was almost sufficient to account for the fall in their share of value
added; in addition, the current average annual earnings of nonpro
duction workers grew about 5 percent more rapidly over the decade.
The increase in the share of value added that was paid out in com
pensation of employees over the period, then, was partly due to the
shift to higher salaried administrative personnel; another part was
due to the more rapid increase in the current earnings of salaried
workers.
12. From 1947 to 1953, production workers’ wages and administra
tive salaries increased at about the same rate. Then, in the 1953-54
recession, average wage earnings fell, while administrative salaries
continued to rise. Some of the drop in wage earnings was the result
of shorter hours and layoffs, of course, but even after the contraction
ended, salaries rose more rapidly than production wages. This dif
ference in behavior is symbolic of a number of changes that have
taken place as a result of the change in proportions between wage
and overhead employment. Total labor costs show much less flexi
bility over the cycle as the relatively fixed, overhead component in
creases. Similarly, the gap between unit labor costs and physical
productivity measures becomes greater as occupational proportions
shift. Unit payroll costs over the period increased to a significant
extent as a result of the shift to higher paid administrative workers.
One result is that year-to-year output/unit labor indexes become
much less significant for the explanation of inflationary wage pres
sures. A second result of these shifts is that the downward rigidity
of earnings levels is increased, adding to the general cost-push influ
ence. For these reasons, the data on production and nonproduction
workers have generally been kept separate in this study.
13. In the remaining paragraphs two questions will be answered.
First, what were the mechanics of the change in the employees’
share of value added in manufacturing? Second, what is the sig
nificance of these results, in terms of the pace of inflation? In the
next section, we will consider wage changes, in terms of specific
products.
14. In order to show how the wage and salary share moved in the
postwar inflation, we will break the value added into its components
and show mechanically how the different parts must move in order
that the wage and profits shares should change. Value added by
manufactures is measured by subtracting the cost of materials, sup
plies, containers, fuel, purchased electric energy, and contract work
from the value of shipments of manufacturing establishments. What
remains can be broken down roughly into compensation of employees
on one hand, and profits plus depreciation, on the other. Let—
THE INCIDENCE OP INFLATION
143
Pi= the price of the ith manufactured commodity.
q i= th e quantity of the ith manufactured commodity shipped during the year.
w *=the average annual earnings in the ith industry.
l*=the average number of employees in the ith industry during the year.
* = the ratio of number of employees to physical quantity of the ith commodity.
That is, h = a t q« and
1«
/q i= a i, the input of labor required per unit of output.
pm= the price of raw materials.
m = th e quantity of materials used in the production of one unit of the ith com
modity.
k i= th e markup over the cost of labor and materials in the price of the ith
commodity.
Then, the value of the commodity produced by the ith manufacturing industry is
p* q«=q* [l+k« (wi a«+pm m i)].
The unit value added is
P i—p*» m*=Wi a i+ k i (w i ai+p#» m i),
and the share of wages and salaries in value added in the industry is
0 4
4
Wi a*____ ^ ki (Wi ai+Pm mi)
Pi—Pm mi
Pi—Pm mi
The employees’ share is defined in terms of the wage rate, the labor/
output ratio, and the markup of prices over material costs. It moves
with the wage rate and the labor product ratio and against the markup
over direct costs. The indexes for each of these variables for all manu
facturing are shown in table 1.
T able 1.— Income shares— All manufacturing, 1947-56
A. ALL EMPLOYEES
[1947-49=100]
Year
Wages and salaries
Value added
Percent
1947...................................
1949...................................
1960...................................
1951...................................
1952...................................
1953...................................
1954...................................
1955...................................
1956...................................
53.4
56.7
53.4
55.1
56.0
56.4
56.4
54.8
55.2
Index
96.9
102.9
96.9
100.0
101.6
102.4
102.4
99.5
100.2
Annual earnings
Real
index
Current
index
99.3
103.3
107.7
110.7
114.5
119.8
121.7
127.9
131.8
Output
Unit wage
Employment plus salary
cost, index
Index
94.8
105.2
110.7
122.9
130.0
137.0
139.7
146.4
153.2
98.5
98.5
107.8
109.0
109.6
112.1
110.9
117.3
118.1
95.9
102.5
100.2
111.1
116.1
118.3
120.5
119.8
124.4
B. PRODUCTION WORKERS
Year
Wages
Value added
Percent
1947......................................
1949......................................
1950......................................
1951......................................
1952.....................................
1953.....................................
1954.....................................
1955......................................
1956.....................................
40.7
40.1
38.6
39.8
40.1
40.3
38.1
37.4
37.3
Index
100.7
99.3
95.3
98.5
99.1
99.6
94.3
92.5
92.2
Annual wage earnings
Real
index
100.6
102.1
108.2
110.8
114.9
120.0
118.8
125.6
129.1
Current
index
96.1
103.9
111.2
123.0
130.4
137.3
136.4
143.8
150.0
Output
Employ
ment
Unit wage
cost index
Index
96.2
100.9
110.0
110.9
112.8
115.5
317.7
123.9
125.8
97.7
100.2
98.8
107.3
109.2
111.3
108.4
108.9
112.0
144
THE INCIDENCE OF INFLATION
O. PROFIT MARGINS AND VALUE ADDED
Year
1947.................................................................................................................
1949.................................................................................................................
1950.................................................................................................................
1951............................................................................................................ .
1952.................................................................................................................
1953.................................................................................................................
1954.................................................................................................................
1955.................................................................................................................
1956.................................................................................................................
Value added
Value of product
Index
93.3
103.8
105.0
115.7
114.8
114.3
115.0
118.6
121.3
Profit
margin
on sales,
index
99.6
95.4
108.0
87.0
75.7
74.3
82.7
94.0
84.2
Sources: Wages and salaries, value added, output/employment indexes: Census of manufactures. Value
added/value of product, unit wage and salary costs: “Productivity, Prices, and Income,” table 51. Profit
margins: First National City Bank, in “Productivity, Prices, and Incomes,” table 22.
15. How have these variables moved during the 1947-56 period?
The share of wage earners and salaried employees increased from 1947
to 1949 and then fell sharply in the first year of the Korean war.
In 1951 the employees’ share began to improve and continued to rise
until in 1953 the level of 1949 had almost been recovered. There was
a decline in 1955 and some recouping in the following year. The gain
in the wage and salary share over the period was almost two per
centage points.
16. If we separate the production workers from the administrative
group, the pattern changes considerably. Between 19485 and 1950,
unit wage costs fell, since real output per production worker rose more
rapidly than production workers’ earnings. At the same time the
markup of prices over material costs and, therefore, the profits margin
continued to rise. The wage share fell by more than three percentage
oints between 1949 and 1950. In 1954 and 1955, a similar thing
appened: unit wage cost fell, this time under pressure of declining
employment, while output per man-year and the value added markup
continued to rise. Over the whole period, the production workers’
share of value added declined by about 3.4 percentage points.
17. As the economy recovered after the recessions of 1949 and 1954,
the profit margins made the first gains. Much of the shift in favor of
profits was due to the reduction m unit overhead costs as output in
creased from levels far below plant capacity. And some of the gain
represents the fact that unit production wage costs fell at the start of
both recessions and did not turn up again until the second year of
recovery. By contrast, wholesale prices responded very quickly to the
revival of demand in 1950, and the value added markup did not
decline at all. Between 1954 and 1955, total unit labor costs fell, while
unit value added rose.6 It seems clear from this evidence that in 1950
the increase in profit margins provided much of the inflationary
pressure. In between, from 1951 to 1954, the rapid rise in unit wage
and salary costs was achieved at the expense of profit margins; whole
sale prices and unit value added did not change noticeably. In 1955,
profit margins rose and the wage and salary shares fell. In 1956, the
E
* In 1948, the unit (production worker) wage cost index was 102.5.
6 Unit production labor costs fell sharply from 1953 to 1954 and stayed down until 1956.
145
THE INCIDENCE OF INFLATION
unit wage and salary costs rose and the employees’ share recovered
slightly, at the expense of profit margins.
18. Even in this short period between 1949 and 1956, there seems to
be a significant cyclical pattern: in the recession years, wage earnings
lost ground and then, after a lag of about 1 year, revived and pressed
on to gains clearly in excess of productivity increases. Over the
decade, current wages increased more rapidly than output per unit of
labor. This, given the relative stability of the profit ratio, would be
sufficient to explain the pressure upon commodity prices. But it is at
least doubtful that prices would have risen by as much had there not
been the cyclical pattern of loss in the wage share and accelerated
recoupment twice in the decade. This evidence supports an interpre
tation of the cost-push process in which wages and profits alternate
in capturing short-run gains (with wages lagging somewhat in the
recoveries), but move together when the changes are averaged over
the cycle.
19. When we consider the individual manufacturing industries, the
pattern appears to be more varied and it becomes possible to answer
more specific questions about the relationships among wages, profits,
and prices. The frequency distribution of first differences between the
share of compensation of employees in 1947 and 1956, computed from
the three-digit industry distributions, is shown in table 2A ; corre
sponding data for production workers is shown in table 2B. The
frequencies are given in numbers of employees and production workers,
respectively.
T able 2.— Frequency
distribution changes in share of wages and salaries in
manufacturing value added, 1946-57
A. ALL EMPLOYEES
[Thousands]
Earnings
Value added
Earnings (
Value added U
1
1
[
All manufacturing......................
Industries 20-29...........................
Industries 30-39...........................
Durable goods sector (24, 25,32,
33, 34, 35, 36, 37, 38, 39)............
Nondurables sector (20, 21, 22,
23, 26, 27, 28, 29, 30, 31)............
Consumer durables sector (227,
251,301, 358, 362, 366, 371)
(-.05) (-.03) (-.01) (+.01) (+.03) >+.05
< -.05 -(-.03)
-(-.01) -(+.01) -(+.03) -(+.05)
2,980
507
2,473
2,455
525
895
1,966
114
1,852
1,474
492
212
1,958
1,004
954
962
996
0
3,219
1,210
2,009
2,269
950
762
1,793
751
1,042
1,060
733
0
2,025
1,580
446
424
1,602
54
2,516
2,359
157
770
1,862
0
B. PRODUCTION WORKERS
[Thousands]
Earnings
earnings
<-.05
Value added(1956) value added (1957)
All manufacturing___________________
Industries 20-29_____________________
Industries 30-39_____________________
Durable goods sector (24, 25, 32, 33, 34, 35,
36, 37, 38, 39)............................ ................
Nondurable sector (20,21,22,23,26,27,28,
29,30,31)______________ __________
Consumer durables sector (227, 251, 301,
358, 362, 366, 371).....................................
4,371
513
3,858
3,714
657
893
(-.05)- (-.03)- (-.01)- (+.01)(—.03) (-.01) (+.01) (+.03)
2,319
519
1,800
1,849
470
226
2,431
1.414
1.017
1.017
1.414
371
1,430
1,189
241
327
1,103
417
Source: Census of manufactures, computed at Harvard economic research project.
561
435
126
98
463
0
>+.03
2,104
2,029
75
445
1,659
146
THE INCIDENCE OF INFLATION
20. Recall that the wages and salaries share increased by 2.8 percent,
from 0.534 to 0.552 of value added in manufacturing over the decade.
The distribution of changes does not show any concentration about
that figure. There is a mode in the interval of small changes—that is,
less than 1 percent difference in either direction. But there are two
additional peaks: one at the upper end of the scale—changes of more
than 5 percentage points, and another, somewhat larger, at the
negative end of the scale. When we split the distribution into two
industry groups—those producing nondurable goods and those pro
ducing durables—a more significant pattern emerges. The nondurable-goods group displays a definite peak at the high positive end
of the scale. The durables sector has a peak in the small-change
bracket and another, larger one in the negative open-end bracket, that
is, the bracket containing losses in share that were greater than 5
percentage points. There was evidently a real difference in the wageprofit relationships between the two major groups.
21. Within the durables sector, motor vehicles and iron and steel
accounted for three-quarters of the frequency at the open end of
the negative scale showing losses in wage share. The decline in em
ployees’ share in blast furnaces and steel mills (331) was almost 16
percentage points; 6 the decline in motor vehicles (371) was almost 10
points. In both cases there was a fairly steady downward trend over
time, with a minor revival in 1952 in the basic steel wage and earnings
share, and a somewhat stronger rise in 1953 in the employees’ part of
the motor vehicles distribution. There was less concentration in the
nondurables industry distributions, although broadwoven fabrics
(223) and men’s furnishings (232) are major industries that showed
large gains.
22. For production workers, the wage share in value added declined
from 40.7 percent to 37.3 percent. However, when the distribution
of changes is considered in more detail, we find a similar pattern to
that among all employees, but with a more decisive concentration at
the lower ends of the scale. There is a minor peak in the frequency
of gains of greater than 3 percent, contributed almost entirely by the
nondurables. There was one strong gain in the production workers’
share in the durable goods sector—by the lumber and basic products
wage earners.
23. From the definition of the wage share given above, it would
be expected that the decline in the production workers’ share might
be partly accounted for by some difference in the growth of salaried
workers between the two sectors. There is a slight difference, in the
expected direction; nonproduction workers employment increased a
little more in the durables industries. The gain was from 16.3 to
23.4 percent in durables, compared with a change 16.4 to 22.7 in the
nondurables (from 1947 to 1957). The gain in salaried employment
the durables, however, was not concentrated in the primary metals
or motor vehicles industries, where the reductions in the production
wage share were heaviest, but in ordnance (which is not included
in mis study), aircraft, and machinery.7 There does not seem to be
much of an explanation in this change.
6 If the employment costs as a percent of revenue given in the Eckstein-Fromm Study,
table 11, are translated into wage costs as a share of value added, the agreement of these
two estimates from different data sources becomes reassuringly clear.
7Cf. Murray Wernick, “ Occupational Shifts in Manufacturing Employment: Some
Implications for Productivity and Unit Labor Cost Measurements.” Speech before the
Cleveland chapter, American Statistical Association, Mar. 4, 1958, mimeographed, table 4.
THE INCIDENCE OF INFLATION
147
24. The major declines in the share of employees’ compensation
occurred among the durable goods producers, and in particular, in
the primary metals and automobiles sectors. Let us review the major
elements of change in the shares identity for these industries. Average
earnings of production workers increased about 10 percent faster
in the durables than in the nondurables. Earnings in basic steel, in
turn, went up about 10 percent faster than the average for all dura
bles ; earnings in the automobile industry went up a little less rapidly
than the durables average. Output per production worker man-hour
went up more rapidly in durables than in all manufacturing and
more rapidly in basic steel than in all durables. A crude unit labor
cost index snows an increase of at least 10 and probably 15 percent
greater for nondurables over the decade.8
25. The profit margins are somewhat more complicated. Between
1947 and 1956, the ratio of net profits before taxes to sales, for 200
large manufacturing corporations, went from 13.3 percent up to a
peak of 17.9 in 1950, and then down to 12.9 at the end of the period.
The durable goods manufacturers in the group earned profits at gen
erally lower rates and the nondurables rate was correspondingly
higher. The rates are quoted in table 3.
T able 3.— Profit margins, 200
large manufacturing corporations
[Profits before tax as percent of sales]
Year
All Durables Nondurables
1947......................
1948___________
1949___________
1950___________
1951___________
13.3
14.4
13.8
17.9
16.8
11.9
13.2
13.4
13.1
15.9
15.5
16.3
14.5
18.1
18.2
Year
1952___________
1953___________
1954___________
1955___________
1956___________
All Durables Nondurables
13.4
13.2
12.5
14.7
12.9
12.6
12.5
12.0
14.7
12.3
16.3
14.6
13.4
14.8
14.0
Source: Board of Governors, Federal Reserve System, in “Productivity, Prices, and Incomes,” tables
34, 35, 36.
26. In order to compare the basic steel and automobile margins to
those for all manufacturing, it is necessary to turn to yet another set
of figures. In this series, which is again net of depreciation and
therefore not directly comparable to the value added shares, net profits
as a proportion of sales m all manufacturing (except newspapers)
went from 11 to 9.7 percent. The ratios in primary iron and steel
went from 10.9 to 13.2 percent over the period and the ratio in motor
vehicles manufacturing from 10.7 to 10.8 percent. There were inter
mediate peaks in 1950 and 1955 in all three series, but the automobile
earnings in those years were significantly higher than the ratios for
basic steel and all manufacturing. The automobile profit ratio was
15.1 percent in 1955.9
27. It seems to me to be a fair interpretation of this data to say that
the decline in the share of employees compensation in durable goods
production is not the combined result of unusual productivity gains,
8 Sources: Production worker wage earnings. Bureau of Labor Statistics, tables 44, 45,
Productivity, Prices, and Incomes: Production indexes, Board of Governors of the Fed
eral Reserve System. Federal Reserve Bulletin. December 1957.
• Source: Federal Trade Commission— Securities and Exchange Commission, Quarterly
Financial Report for Manufacturing Corporations, quoted in Productivity, Prices, and
Incomes, tables 33, 149,189.
148
THE INCIDENCE OF INFLATION
which would have held down unit labor costs, but of increases in profit
rates in steel and automobiles that were out of line with those generally
reported in the rest of the group. The greater than average wage
gains in these industries were not made at the expense of profits, but
were accompanied by increases in the profit margins.
28. The causal link between the factor share distribution and the
pace of inflation in manufacturing will not be analyzed in detail here;
instead I shall simply outline some of the more apparent relationships.
In the nondurables, where price increases were relatively low, the
wage share gained at the expense of gross profits. The increase in
unit labor costs and the increase in profit margins, which together
would be expected to push prices upward, were balanced for most of
the period by declining material prices, especially in the food sector.
In both consumer durables and producer durables, the unit labor costs
appear to have lagged as productivity increases compared more favor
ably with wage increases. For most of the group, profit margins were
also lower than the manufacturing average. But in steel and auto
mobile production, the relatively large wage increases were matched
by very high profit margins. Under the double impact, prices in these
sectors rose more rapidly than output and, especially in the steel case,
reverberated throughout the economy.10
29. Finally, I will use the detailed wage data in order to answer
some questions about the determination of wages in the postwar infla
tion. What has been said thus far seems to indicate a cost-push
phenomenon in which profits and wages in succession, after the reces
sions of 1949 and 1954 pushed up costs and prices in manufacturing.
The push has not been uniform, especially from profits, but the best
explanation seems to revolve about the combined behavior of the two
income shares. A reasonable analytic scheme, suggested by the aggre
gate data, might be constructed along the following lines. Assume
that wage earners’ demands are responsive to changes in the cost of
living, the rate of profits, and the increase in output per unit labor.
They will resist any reduction in their real wages, but they will also
attempt to maintain their share of the total income produced; that is,
they will respond to real increases in output per unit labor or to profit
rises especially those that are due to price changes in their own prod
ucts. Assume similarly that entrepreneurs attempt to maintain stable
profit margins over direct costs.
30. Let us say that wage earners have a consumer price target to
ward which they aim. That is, wages are expected to stay in close
relationship to rises in the cost-of-living index. Wage changes will
respond, with a lag, to negative deviations from the desired relation
ship. The pace of inflation is determined by the deviations from
desired wage-price (or wage-profit) relationships which, in turn,
create pressures for greater increases in the succeeding period.11 But
10The relationship between steel wages and profits and the postwar inflation is analyzed
in detail in Study Paper No. 2, “ Steel and the Postwar Inflation,” by Otto Eckstein and
Gary Fromm. In a simple input-output computation carried out at the Harvard economic
research project and quoted by John Dunlop, in “ Policy Problems: Choices and Proposals,”
American Assembly. Wages, Prices, Profits, and Productivity, pp. 151, 153, I suggest
some effects of price changes in steel and other major products upon costs throughout
the economy. Using a diffusion index ratio of intermediate deliveries of a sector
to its total gross output), it was shown that fabricated structural metal products, iron
and steel and lumber and wood products, stood among the first 6 out of a total 45, industry
table, in their relative ability to spread cost changes through the economy. In terms of
their wage costs, in particular, fabricated metal products and iron and steel stood 7th
and 13th in relative effects ( potentially, not in terms of actual relative wage change) upon
the consumer index. The Eckstein-Fromm study paper contains a much more detailed
input-output analysis of the effects of steel costs upon inflation,
i* Cf. A. J. Brown. “ The Great Inflation,” 1959-1931.
149
THE INCIDENCE OF INFLATION
wages will not simply respond to consumer price changes; they may
overcompensate as a result of actual or threatened demand declines
due to the lagged price increase. The remainder of this study is an
attempt to find out to what extent the degree of over or undercompen
sation is related to demand pressures and productivity changes.
31.
From the complete list of three-digit Census of Manufacturing
industries, for the years 1949 to 1956, 61 industries provided con
tinuous data from which several linear multiple regression equations
attempting to explain specific wage changes were fitted by single
equation least-squares. (Every possible observation of year-to-year
change was used in the original correlation matrix.) There were,
therefore, seven cross-sections combined over approximately 61
industries.12
A selection of the correlation coefficients with (1) changes in average
wage earnings, all employees; (2) changes in average wage earnings,
production workers; and (3) the three-digit industry price index are
shown in tables 4 and 5.13
T able 4.— Simple
correlation coefficients: Annual changes, all 3-digit industries,
1949-56
Change,
wages and
salaries
Mean value_________________________________________
Standard deviation. ________________________________
Wholesale price index______ __________________________
Change: Deflated output______________________________
Change: All employees_______________________________
Change: Production workers___________________________
Change: Output per unit, all labor ________ ____________
Change: Output per unit, production worker...........................
Change: Wages and salaries_____________________ ______
Change: Production workers’ wages_____________________
104.83
4.41
.1262
.0447
.0382
.0719
.0596
.0209
1.0000
.7154
Change,
production Wholesale
workers’ price index
wage
104.94
5.57
.0800
.0398
.1230
. 1475
.0365
.0289
.7154
1.0000
122.95
18.39
1.0000
-.0 0 6 8
.0160
.0036
0025
- .0 1 4 4
.5967
.6131
Sources: Census of manufactures, computed at Littauer Statistical Laboratory. Wholesale price indexes,
Bureau of Labor Statistics.
12There is no reason to believe that the omissions were systematic among the variables
finally chosen. The industries are listed in table 5.
13The wage earnings are annual averages for all employees and production workers
derived from the census total salaries and wages series. It includes all forms of com
pensation (including paid holidays and vacations) except the employer social security and
unemployment insurance payments, and other nonpayroll costs such as pension plans, group
insurance premiums, and workmen’s compensation, and is therefore a closer approxima
tion to total labor input costs than the average hourly earnings or straight time hourly
earning series. The price indexes are derived from the input-output industry-base series
developed from the Wholesale Price Index by the Bureau of Labor Statistics, Division of
Productivity and Technological Developments, and made available to the Harvard eco
nomic research project.
150
THE INCIDENCE OP INFLATION
T able 5.— Simple correlation coefficients: Average annual changes, all 3-digit
industries, 1949-56
Average
annual
change,
wages and
salaries
Mean value
Standard deviation________________________________________
Wholesale price index______________________________________
Average change: Deflated output____________________________
Average change: Production workers_________________________
Average change: Output per unit, all labor____________________
Average change: Output per unit, production worker____________
Average change: Wages and salaries__________________________
Average change: Production workers' wages___________________
Average change, all employees_______________________________
________________________________________________________________________
1 0 4 .7 0
1 .8 6
.3 3 0 6
.0 4 2 0
-.0 5 4 8
— .0 9 3 5
Average
annual Average
change, wholesale
production price
workers’ index
wages
1 0 4 .8 2
2 .3 3
.2 5 2 7
.0 1 4 2
1 2 3 .1 1
1 6 .1 1
1 .0 0 0 0
.0 0 0 3
- . 0472
— .1 1 0 7
.1 0 2 5
.0 2 0 5
— . 1258
— . 1579
.8 1 7 0
-.0 7 5 7
-.0 6 6 6
.0 1 6 1
.3 3 0 6
.2 5 2 7
.1 2 4 8
1.0000
1.0000
.8 1 7 0
N ote—Industries included in sample: 205, 212, 213, 223, 226, 234, 241, 261, 271, 272, 275, 276, 277, 291, 301,
302, 303, 309, 311, 312, 313, 316, 321, 322, 323, 324, 327, 328, 331, 332, 334, 335, 336, 339, 341, 347, 352, 353, 355, 362,
363, 364, 365, 374, 375, 379, 381, 382, 383, 385, 386, 397, 211, 224, 235, 252, 254, 264, 265, 266, 267.
Source: Census of manufactures, computed at Littauer Statistical Laboratory. Wholesale price indexes,
Bureau of Labor Statistics.
32. One is immediately struck by the remarkably low degrees of
relationship displayed between changes in employee compensation and
price indexes, on the one hand, and the labor and commodity demand
variables, on the other. In an attempt to avoid cyclical effects, which
would be expected to distort the short-run demand relationships if
there were downward rigidities, changes in the same variables were
averaged over the 1949-56 span and the long-run average changes were
correlated over the same sample of industries. The same selection is
shown in table 5. The same remarkably low degree of relationship
is evident.
33. An explanation that must be considered, of course, is the possi
bility that significant relationships are hidden. Strong negative re
lationships with other independent variables may mask a relationship
between the dependent and a presumably obvious independent in
fluence. In such cases, significant correlations may appear when the
relation with the other variables have been allowed. Two multiple
regression equations for the whole array are reproduced below. They
were designed to be rough tests of the demand-pull hypothesis.
34. The first multiple regression equation, shown in table 6, was an
attempt to explain the average annual wage and salary increase in
terms of (1) the average annual real output change; (2) the average
annual change in real output per unit labor; (3) the average annual
change in employment; and (4) the mean price index, over the period
1949-56 for the specific three-digit industries. As with the simple
correlation coefficients, the results were remarkably insignificant.
Wage changes at the relatively detailed (i.e., more homogenous threedigit) industry level are not explained by demand pressures (inde
pendent variables 1 and 3) or by productivity changes within the
industry itself (independent variable 2) ,14
14 It is perhaps overdue at this point in the study to point out that the use of the word
“productivity0 to mean simple variations in output/labor ratios, while customary, is incor
rect. It would be better to restrict measures of productivity, so called, to changes in
the unit requirements of capital, labor, and materials, combined. It would be still better
to drop the common usage and restrict the use of “ productivity change” to those changes
which can be properly identified as shifts in the production function, as opposed to move
ments along the function. See, e.g., the attempt to derive measures of technological change,
in the strict sense, by Robert M. Solow, “ Technical Progress and the Aggregate Production
Function,” Review of Economics and Statistics, August 1957, pp. 312-320.
THE INCIDENCE OF INFLATION
151
T a b le 6. — Regression equation
[Average annual wage change, all employees, industry 1949-56-B i (average annual real output change),.*
+/?2 (average annual change output per unit l a b o r ) , ( a v e r a g e annual change, all employment),*
(average wholesale price index),*]
Partial
Regression correlation
Beta
coefficient coefficient coefficient
Independent variable
Average annual output change,**.......... ...........................
Average annual output/labor change, < _____ _________
Average annual employment change,*-................. ...........
Average wholesale price index, »•______________________
Regression constant=1_______ ____ _____________
Standard error of estimate_____
___ ____________
Multiple correlation coefficient________ ____ _________
Coefficient of multiple determination, E 2. . ___________
Degrees of freedom=?z—5=56.
0.0060
-.0128
-.0427
.0401
105.46
1.72449
.3695
.1365
0.0429
- . 1162
1322
.3486
0.0399
1090
- . 1252
.3484
Standard
error
of Beta
coefficient
0.1242
.1244
.1254
.1252
NOTES
Industries included in sample: 205, 21i, 2i2, 213, 224, 226, 234, 235, 241, 261, 264, 265,266,267,271,272,276,
277, 291, 301, 302, 303, 309, 311, 312, 313, 316, 321, 322,323, 324, 327, 328,331, 332, 334, 335, 336,339,341,347,352,
353, 355, 362, 363, 364, 374, 379, 381, 382, 383, 385, 386, 397, 375.
The coefficient of multiple determination for wage changes was
E 2= 0.1365, almost all of which was contributed by the addition of
the mean price index to the regression. The multiple correlation co
efficient was 5=0.3695. The wage changes are negatively, though
weakly related to changes in output per unit labor and employment.
The negative coefficient between wages and employment implies that
wages have continued to rise even when employment has fallen behind
in a given sector. The negative coefficient when changes in output
per unit labor is the independent variable means that specific indus
try unit labor improvements count for less in terms ox wage gains
than the maximum improvement or at least the average improvement.
This is not, by any stretch of the imagination or will to believe, a re
sult to support the demand-pull hypothesis as an explanation of wage
increases m inflation.
35.
The second equation (table 7) used the same variables, ordered
somewhat differently, in an attempt to relate the rate of specific price
increase in the three-digit industries, first, to the wage change and then
to the demand variables, taking productivity into account again. The
coefficient of determination for prices upon wage changes was R2=
0.1093. Addition of real output change and productivity change added
virtually nothing; the addition of annual employment change to the
independent variables brought the coefficient of multiple determina
tion up to B 2= 0.1359. Among the partial correlation coefficients, that
is, the measures of relationship between the dependent and single in
dependent variables when the other independent variables are fixed,
only the price-wage relationship (12 345= 0.3486) and the price-employment change relationship (15 234^ 0.1635) approach economic
significance.
152
THE INCIDENCE OF INFLATION
Table 7.— Regression
equation
[Average wholesale price index industry i, 1949 56=Bi (average annual wage change, all employees), < + # 2
(average annual real output change), »+ 2?3 (average annual change, output per unit labor), %+Bi (average
annual change, all employment), <]
Partial
Regression correlation
Beta
coefficient coefficient coefficient
Independent variable
Average annual wage change, *........................................
Average annual real output change, f............. ...............
Average annual change, output per unit labor, *............
Average annual change, all employment, *.....................
Regression constant________________________________
Standard error of estimate__________________________
Multiple correlation coefficient.......................................
Coefficient of multiple determination, R 2.......................
0.30279
-.00150
. 00647
. 04587
247.99
14. 9782
. 3695
. 1359
0.3486
0122
.0673
.1635
0.3487
-.0113
.0631
.1549
Standard
error of
Beta
coefficient
0.1252
.1243
.1250
.1249
N ote.—Degrees of freedom —-n —5=56.
36.
These are rough-and-ready tests. From the matrix of simple
correlation coefficients r, it did not appear that multicollinearity was a
problem in these data. But the likelihood of errors of observations
for averaged year-to-year change indexes is very serious. Still it is
difficult to interpret these results as showing anything but that manu
facturing wages and prices during the recent inflation have been essen
tially free of demand influence at the specific industry level, and that
prices appear to be significantly related to combined profit and wage
increases. The evidence presented earlier and the negative correla
tions between changes in wage share and changes in wage earnings
imply that price rises have not been simply related to wage costs, but
have been accompanied by gross-profit increases.
CONCLUSIONS
This study has examined the behavior of wages and gross-profit
margins among the 131 Census of Manufacturing three-digit industry
groups. Its purpose was, first, to find out whether wage shares had
gained or lost in the course of the postwar inflation. Second, the
specific industry patterns were to be used to investigate the way in
which different sectors, with different output and price behavior, were
struck by the inflation. Finally, some of the data was used to test a
very simple version of the demand pull hypothesis on wage and price
increases. A summary of the results follows:
(1) The share of production workers in value added declined from
1947 to 1956, largely as the result of the relative reduction in produc
tion worker employment. Salaried clerical, administrative, and tech
nical personnel experienced a gain in their share, in part as the result
of an increase in proportions, partly as a result of more rapid salary
increases.
(2) Among production workers and all employees, there was a
clear-cut pattern of gains in the wage and salary share in the non
durable sector and losses in the durables, with the relative wage loss
concentrated especially in the basic steel and automobile industries.
The declines in the durables were balanced, arithmetically, by the
gains in gross profits; in steel and auto production the net profit
margins showed unusual gains during the period.
THE INCIDENCE OF INFLATION
153
(3) The gross-profit increase in durables includes depreciation al
lowances. This avoids the problem of distortion in the #profitsdepreciation split which is inherent in the use of historical costs and
changing amortization procedures. It has been suggested, however,
that the rising gross-profit share in basic steel especially is due to the
necessity to finance replacement and expansion at continually higher
prices. Given the apparent tie between wages and profits and the
nature of cost-push, it is at least questionable that increasing profit
margins provide the most satisfactory, i.e., least inflationary, method
of financing steel changeover and expansion.
(4) The detailed industry data do not give any support to the simple
demand-pull hypothesis and limited support to the cost-push explana
tion. Combined with the overall stability of shares and the alternat
ing pattern of profit and wage gains, the inflationary cost push should
be interpreted as a profit-ana-wage push, at least in the manufacturing
sector.
(5) The results of this study, limited to manufacturing and to spe
cific industry demand pressures, do not bear upon the likelihood that
demand pressures elsewhere in the economy might spill over into wage
setting in manufacturing. Such a possibility, especially in the pricing
of services, could very well set off the inflationary profit-wage alterna
tion described here.
o
@FedFRASER