Economic growth and total capital formation


Material Information

Economic growth and total capital formation
Physical Description:
Kendrick, John W
United States -- Congress. -- Joint Economic Committee
U.S. Govt. Print. Off. ( Washington )
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 25739997
oclc - 2039463
System ID:

Table of Contents
    Front Cover
        Page i
        Page ii
    Letter of transmittal
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
    Economic growth and total capital formation
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
Full Text


2d Session J






,- ci
~ 1
b ~

\ :~' t'.AAR

'$ 6K~fpN1T~~







FEBRUARY 18, 1976

Printed for the use of the Joint Economic Committee



I -



(Created pursuant to sec. 5(a) of Public Law 304, 79th Cong.)

II UBF RT H. II UMPHII REY, Minnesota, Chairman
WRIGHIT PATMAN, Texas, Vice C'hairman

JOIN .I'A RKMAN. Alabama
1I.LIA.M P'RIXNMIRE. Wisconsin
ABR-AHAM RIBIC('FF, (CoiJirvit lit
F )WARD1 M. K E N N E DI)Y, Massachusetts
JA('(COB K. JAVITS. New York
(CHARI.ES H. PERCY, Illinois
PAU'L J. FANNIN, Arizuona

HENRY S. REUSS, Wisconsin
GILLIS W. LONG, Louisiana
GA RRY BROWN, Michigan
J01IN H. ROUSSELOT, California

JOiXN R. STARK, I'Eimctt( Director




RICELARD F. KAUFMAN, (;etineral Counsd









LLOYD M. BENTSEN, JR., Texas, Chairman

ABR AI A.AM RIBICI" (')FF, Connecticut
HIUBERT 11. 1 UN'MIPHREY, Mi.inesota
EDWARD M. KENNEDY, Massachiusetis

WILLIAM S. MOORHuEAD, Pennsylvania
MARGARET M. HECKLER, Massachusetts



FEBRUARY 16, 1976.
To the Members of the Joint Economic Co&,mitte:
Transmitted herewith is a study entitled "Economic Growth and
Total Capital Formation," which was performed for the Subcommittee
on Economic Growth by Professor John W. Kendrick of the George
Washington University. Since the findings of this study have a direct
bearing on the future economic growth of our country, I believe the
members of the Joint Economic Committee and other Members of
Congress will find it most useful.
The views expressed in the study are those of the author and do
not necessarily represent the views of the members of the Joint
Economic Committee or the committee staff.
Chairman, Joint Economic Cormmittee.

FEBRUARY 12, 1976.
Chairman, Joint Economic Committee,
U.S. Congress, Washington, D.C.
DEAR MR. CHAIRMAN: Transmitted herewith is a study by Professor
John W. Kendrick of the George Washington University and the
National Bureau of Economic Research, entitled "Economic Growth
and Total Capital Formation."
Professor Kendrick's study is a pathbreaking attempt to estimate
the total level of capital formation in the Americnm economy from
1929 to 1973. Going beyond the traditional measures of capital and
investment in the official national income accounts, Profe-sor Kendrick
defines total capital as "all tangible durable goods (structures and
equipment) plus inventory accumulation of all sectors, households and
governments as well as business, and all intangible investments
designed to enhance the efficiency of the tangible factors. The intailgi-
bles comprise outlays for re(eoarch and development (R. & D.),
education and training, health and safety, and mobility."
In the study, Professor Kendrick has made a serious effort to
construct a complete profile of how much total investment our nation
has made, both in tangible and intangible forms of capital. This is the
first time such an attempt has been made.
The analyses and conclusions Professor Kendrick makes on the
basis of his new data have serious implications for the continued
growth of the American economy. Between 1929 and 1966, our nation
continuously increased the proportion of our national output going to
tangible and intangible investment, but since then the proportion has
declined. Furthermore, there has been a major shift from tangible
investment to intangible investment, and from the business sector to


At the end of lis study, Professor Kendriek makes a number of
recommendations concerning how we can reallocate our investments
to get our economy growing again and avoid recreating the bottlenecks
which slowed down our growth two years ago and which helped
precipitate our current economic malaise.
The views expressed in Professor Kend rick's study are those of the
author and do not necessarily represent the views of the members of
the Subcommittee on Economic Growth.
Chairman, Subcoinmmithe on Economic Grou'th.


Letters of transmittal---------------------------------------------- III
Gross investments in relation to GNP------------------------------- 2
Composition of investment by type---------------------------------- 3
Sectoral composition ----------------------------------------------- 4
Gross capital stocks in relation to GNP ------------------------------ 6
Composition of gross wealth ---------------------------------------- 9
Rates of return on total capital------------------------------------- 11
Concluding comments --_------------------------------------------ 12

Digitized by the Internet Archive
in 2013

By John W. Kendrick*
Evaluation of the impact of the tax system, and changes in taxes, on
economic growth and progress requires an understanding of the chief
sources of economic growth. Broadly defined, capital formation, in all
its many forms, is by far the most important source of growth.
As traditionally measured in the official national income accounts,
capital formation or "investment" comprises only the tangible, non-
human categories of purchases of new structures, producers' durable
equipment, and business inventory accumulation, plus net foreign in-
vestments. This narrow definition was in line with Keynesian macro-
economic theory, which highlighted business tangible investment as
the chief independent variable involved in determining the level of
national income and product. Certainly it is the most volatile form of
investment over the business cycle, and is heavily influenced by condi-
tions in the financial markets.
However, from the viewpoint of economic growth analysis, it is use-
ful to define and measure net investment more broadly as comprising
all current outlays that augment income- and output-producing
capacity (capital) for future periods. Gross investment includes addi-
tionally the investment required to offset capital consumption (chiefly
depreciation) reflecting the gradual wearing out and/or obsolescence
of capital. Thus broadly defined, capital formation consists of outlays
for all tangible durable goods (structures and equipment) plus inven-
tory accumulation of all sectors, households and governments as well
as business, and of intangible investments designed to enhance the
efficiency of the tangible factors. The intangibles comprise outlays for
research and development (R. & D.), education and training, health
and safety, and mobility. Also, for the sake of logical consistency, one
may also include tangible human investment, defined as the cost of
rearing children to working age, which parallels the cost of the brick,
mortar, and machines that comprise the tangible nonhuman fixed
capital. The intangibles are generally embodied in the human and non-
human capital, increas-ing the productivity of the physical constituents.
The importance for analysis of looking at all forms of investment is
that they all compete for the finite savings of the community; and that
to promote the optimum allocation of resources, investments in each
type should be carried to the point where the expected rate of return
equals the marginal cost of funds. Even though we recognize that
human investments, in particular, are undertaken for noneconomic as
well as for economic reasons, in varying degree, it is ncverthele-s
iiueful to have estimates of all forms of invcl inent in devising policies
to influence economic growth. At least rough allowance can be made
for the nonpecuniary ;s well as the monetary returns to human
invest ment in attempting to formulate growth strategies.
*Professor economics, George Washinigton University, and senior research sI IT member, National Bureau
of Economic Research.

Since the U.S. national income and product accounts do not identify
or provide estimates for many of the categories of "total investment"
as defined above, I undertook the preparation of estimates of such
investments, together' with the associated capital stocks, in a study
soon to be publishLed by the National Bureau of Economic Research.1
For this paper, I have updated the gross investment estimates, by
type and sector, through 1973. In the following section, I summarize
briefly the chief findings of the study.

Gross Investments in Relation to GNP

In table 1 the ratio of gross saving and investment, as conventionally
defined, to GNP is shown for selected business cycle peak years,
1929-73. The ratio of the current dollar magnitudes has been relatively
stable at around one-sixth. In constant prices, the ratio was higher in
1929 and 1948 than in more recent years.

Gross private
domestic invest-
Gross ment plus net Ratio (2)+(1)
national product foreign investment (percent)
(1) (2)

Billions of current dollars:
1929------------------------------------------............................................. 103.1 17.0 16.5
1948------------------------------------------............................................. 257.6 47.9 18.6
1957------------------------------------------............................................. 441.1 71.2 16.1
1966------------------------------------------............................................. 749.9 123.9 16.5
1969------------------------------------------............................................. 930.3 137.9 14.8
1973-..--.- ---.-.....-.....-...........- .. 1,294.9 209.4 16.2
Billions of 1958 dollars:
1929................................------------------------------------------ 203.6 42.0 20.6
1948----------------.. --------------------------... 323.7 62.8 19.4
1957---------------..................----- ---------------------- 452.5 72.3 16.0
1966---.......---------.---..---------------------------.............. 658.1 111.4 16.9
1969.......------------------------------------------ 725.6 109.7 15.1
1973-..-..---------------------------------- ---...-. 839.2 138.2 16.5

Source: Bureau of Economic Analysis, U.S. Department of Commerce.

In comparing estimates of the saving and investment concept
expanded to include nonbusiness tangible capital outlays, and in-
tangible investments of all sectors, it is necessary to adjust the official
GNP estimates for comparability. We make no attempt here, however,
to enlarge the estimates generally to include all imputations for non-
market activities and other adjustments sometimes advocated to
produce a better welfare-oriented measure. Rather, we adjust the
official GNP estimates only for the several items necessary to obtain
consistency with the expanded investment estimates. As shown in
table 2, these comprise the imputed rentals on nonbusiness capital
stocks, investments charged to current expense by business, and
imputed compensation of students and the frictionally unemployed
which is included as part of human investment. It will be observed
that the adjustments to GNP rose gradually from 24 percent in 1929,

1 John W. Kondrick, The Formation and Stocks of Total Capital (New York: National Bureau of Economic
Research, in press). This volume and the subsequent research by the author at NBE R have been supported
by grants from the National Science Foundation.

and 27 percent in 1948, to around 34 percent in 1969 ai d 35,2 p1ercet
in 1973. In constant prices, the ul)war(td trend was less nilarked, alnd
levelled out at around 32 percent in both 1969 and 1973

[In billions of dollars]

1969 1973

Current dollars:
GNP, commerce concept-----....--------------------- ---- ---------.- ------- 1 929.1 1,294.9
Personal sector imputations:
Student compensation ------------------------------------------- 92.3 143. 1
Frictional unemployment------------. -------------------------- 16.0 24.1
Rentals on household capital--------------. --------------------- 100. 1 138.5
Rentals on institutional capital..--..--...------ ------------------- 5.7 8. 5
Business: Investments charged to current account:
Tangible---.. ---. ------. ----------. ----. ----------------------- 2.3 3.3
Intangible.-------..----------------------------. ------------- 35.4 45.6
General governments: Imputed rentals on public capital-----... --- 67.0 91.2
Equals: Adjusted GNP.................----- ......----------..-----. 1,247.9 1,754 3
Ratio to Commerce GNP.. ------------...----....---.---.-. -------- 1,343 1.355
Constant 1958 dollars:
Commerce GNP..............--------------------------------------------------- 724.7 ,839.2
Adjusted GNP- ---------------------------------------------------- 957.2 1,104.5
Ratio.------..----. ---------------------------------------- 1.321 1.316

I The Commerce Department estimate of GN P for 1S69 was subsequently revised slightly upward from the number shown
here which is consistent with the adjusted GNP series for the period 1929-69 presented in John W. Kendrick, "The I-or-
mation and Stocks of Total Capital." (New York: National Bureau of Economic Research, forthcoming 1976.)

Even in relation to a significantly larger GNP, as a(ijusted, total
gross investment was a much larger proportion than conventional
investment-around half by the latter 1960's, as compared with one-
sixth by the conventional measure,. And even though adjusted GNP
rose somewhat relative to the official numbers, the ratio of total gross
investment to adjusted GNP rose significantly between 1948 and the
latter 1960's. The proportion in 1948 (as in 1929) was (close to 4.3 per-
cent; by 1966 it reached 50.5 percent. Significantly, the ratio receded
thereafter to 49.5 percent in 1969 and 48.5 percent in 1973. (See
table 3.)
The ratios of total net investment to NNP were only about half the
gross ratios in 1929 and 1948, but thereafter showed an even sharper
relative increase-from 21.4 in 1948 to 29.7 in 1969. (We have not
extended these estimated to 1973.) In constant prices, the ris.e in the
total investment proportion was somewhat less marked, since the
implicit price deflators for gross investment rose more than those for
adjusted GNP. On the real basis, the proportion increa.:-ed from 44
percent in 1948 to 49j1 percent in 1966, and leveled out thereafter,
according to our preliminary estimates.2 As noted below, the leveling
out in the real investment ratio, and drop in the current dollar ratio,
was associated with a deceleration in the rate of growth in total capl)ital
after 1966, which was one factor in the subsequent slowdown in growth
of real product and productivity.

Composition of Investment by Type (Table 3)

All of the increase in the ratio of-gross investment to adjusted
GNP was (lue to the relative increase of intangible investments firom
about 12 percent to 21 percent in 1973. The inciea-e was stea(iy, \\with
2 See K, i ii k. ikl id.. Table 3-3.

ex''p)tion of the i la-,t -iibpl)eriod, when the intallgible in vesttment ratio
d ro p1 Il slightly. The t.angil)le, no iIhumalln investment ratio showed
no )rInoiunedl1 triIend, account iing for a b1it over 23 lericent of adj listed
GNP in 1929, 194S, and 1973. 'Tantgible humanin investinents in rear-
illg, arteCr ldrllpping Lorin) 7.7 percent in 1929 to about 514 percent in the
iiid 1 940's, 1ro-e r.tlatively to over 6 percent inll the latter 1950's. Since
then a steay rlvelative decline, reflecting the declining birth rate
bi)r niiht tlie r:atio down to 4.3 in 1973. Pre-1iin inbly, tlie decline is con-
tinuinii.., relcali n funds for other typ)e-, of inve-,tment as well as for
c()IIl-lmption (-illce reariii,. co-ts not only reduce family saving, but
al-o have an b:ll.tinence ,4t(ht).

(In billions of dollars and percentages, selected years]

Intangible investments
Education Health
Grand and and Tangibles
total Total training safety Mobility R. & D. lota,

Billions of current
1929.--------------- 55.0 15.7 11.0 1.9 2.5 0.3 39.2
1948--.....---------. 139.9 45.0 31.0 5.2 6.6 2.4 94.9
1951 ............... 272.0 92.7 61.4 10.6 10.5 10.3 179.2
1966 ............... 495.8 198.1 137.4 21.4 17.0 22.3 297.7
1969 ...... -..... 611.7 267.8 192.4 27.9 21.3 26.2 344.0
1973.---........... 851 0 369.6 262 6 45.9 31.0 30.1 481.4
Percent distribution of
total gross investment:
1929 --------------- 1 43.1 28.5 20.0 3. b 4.6 .5 71.5
1948 .......... ..... 142.7 32.1 22.0 3.7 4.7 1.6 67.9
1957 --------------- 147.6 34.0 22.5 4.0 3.8 3.8 66.0
1966 ...1.------- '150.5 40.0 27.7 4.3 3.4 4.5 60.0
1969--------------- 49.0 43.8 31.5 4.6 3.5 4.3 46.2
1973-..-..------------- 48.5 43.4 30.9 5.4 3.6 3.5 56.6

1 Percent of adjusted GNP.

Within the intangil)le category, tlhe sharpet proportionate increase
was in R. & D. outlays up to the mid-1960's, but then ,the ratio to
adjusted GNP dropped from 2.3 percent in 1966 to 1.7 percent, in
1973. Thi- decline has been cited as one explanation for the productivity
slowdown (and has- prompted proposal for a R. & D. tax credit).
Outlays for education and training, the largest type of intangible
investment, grew steadily from less than 9 percent of adjusted GNP
in 1929 to more than 15 percent in 1969, but then declined a bit be-
twecni 1969 and 1973. Health and safety outlays roe i relatively over
the entire period, from 1.5 percent in 1929 to 2.6 percent in 1973.
Mobility outlav- did not quite keep pace with GNP sagging from
2.0 percent in 1929 and 1948 to 1.8 percent in 1973.

Sectoral Composition

With regard to the -c(-tor comp)o-ition of gro-s invc,-tnient, all of the
net inci'ia-,e ini the ratio to adji-ted(l GNP ctilme in the public :ector.
"I'lie Government iinve-i.ii't ratio in 1953 ; 11.6 percent,
however, and th( n,:,ifte i--gred. paLiticularlyv in the 1969-73 -ubperiod.
After 1953, oth business iA personal inv-tient -1t'cmnghened :oine-
what relatively, although the bi-ii,,-, iv;' -tmient ratio fell from a
peak of 12.8 peir('nt in 1966 to 11.8 p('rcellt in 1969. The personal

investment ratio has held quite steady at around 261/2 'p(r'-(nt' in
the 1966-73 period, close to the 1929 ratio.
Changes in the ratio of gross invest meant to GNP fori each -..'tor
(';n be better understood by looking first at their ratio of the sectoral
disposable income to GNP, and then at the proportion of that income
devoted to investment. (See table 4.) Thus, it is apparent that tlie
increase in public investment as a proportion of GNP was due chiefly
to the relative increase in public revenues. While the government
take from GNP more than doubled, the proportion of public revenlle
channeled into investment rose modestly from 44.3 percent in 1929 to
46.5 percent in 1973. It will also be observed that tlhe Government
share of gross product (income) rounded out in the latter 1960's, and
the drop in the public investment ratio 1969-73 was due to the de-
clining government income share reinforced by a mild decline in the
ratio of public investment to revenues.
[Percentages, selected peak years, 1929-73]

1929 1948 1957 1966 1969 1973

DI/GNP .................................... --------- 78.8 70.4 68.5 63.0 G7.0 *13.8
Inv./DI----------- --------------------- 33.2 35.2 37.3 41.9 39.4 38.5
Inv./GNP----.------------------------- 26.1 24.8 25.5 26.4 26.5 26.5
DI/GNP------------------------ ------10.0 10.2 10.1 12.5 9.5 9.3
Inv./DI---------------------------------................................... 124.4 123.8 109.5 102.4 123.9 128.0
Inv./GNP-----------.... --- --------- ------------ 12.4 12.6 11.1 12.8 11.8 11.9
DI/GNP---------------------------------10.4 18.7 20.9 24.2 23.4 21.7
Inv./DI---------------------------------- 44.3 28.6 52.2 46.3 48.1 46.5
Inv./GNP..-------------..-------------------.......... .. 4.6 5.2 10.9 11.2 11.3 10.1

Notes: Dl =Disposable income of each sector equals gross income earned from current production plus transfers (includ-
ing taxes, in the case of governments) received from other sectors less transfer (and tax) payments. lnv.=Tctal gross
investment, both tangible and intangible, of each sector. GNP=Sum of disposable income of each sector includingg rest-of-
the-world, not shown here) plus the statistical discrepancy between income and product.

Conversely, the share of the personal sector in gross iiational
product declined over much of the period, as a result of increases in
personal tax rates. But the proportion of disposable personal income
devoted to investment rose, so that the ratio of gross investment to
gross product remained relatively stable. It should be noticed, however,
that these tendencies were reversed after 1966, when the share of
personal income in GNP rose, but the proportion of income devoted
to investment fell, resulting in continuing stability of the investment!
GNP ratio.
Business disposable income (cash flow) was 9.3 percent of GNP in
1973, a bit lower than the 10 percent of 1929. But the ratio of gross
investment to income was 128 percent, somewhat higher than in
1929. Consequently, the investment/prodluct ratio in 1973 was only
slightly below the 1929 figure. The business share of gross product
was highest in 1966. The sharp dl,')p!) between 1966 and 1969 was
partially offset by a -'itificant rise in thle in ve;met income ralio
but declining relative income was nevertheless re-ponsible for l'
percentage,' point drop in the ratio of gr o- busiluess investment to)

Alhioiigh busines, wavs a net borrower throughout thle period,
1u-.illy investing about oei-quarter more than its internally generated
funl(l-, veII in groo(d ear-, the other two dome-,tic sectors were net
-vers. Saving exceeded( investment, a-s defined(l here, by more than 2
1r(cenilt of di-powable income, on average, in the personal sector, and
b\ ;iii average of over 3 percent in the public -ecct(or. It will be reliem-
bredl that in our study we sharply -eparate current and capital
accounts, (so that saving represents the difference between a sector's
dispoable income (including imputations) and its current outlays.
In thi-. view, whereas the public sector has generally been a borrower,
on net balance, the borrowing has been les than public investment in
most years, indicating net saving on current account.
Before leaving the topic of investments, the importance of sectoral
shifts should be pointed up. Thus, given the often-overlooked fact that
gov;niinents invest a higher proportion of their disposable incomes
tli',,i pi.-onls, the relative shift of gross income and product from
per-o)ns to governments between 1929 and 1966 contributed to the
ri-sinp national total saving-investment ratio. It also contributed to the
fas-ter increase in intangible than in tangible investment, since a much
higher proportion of public than of private sector investment is devoted
to intangible outlays that enhance productive efficiency. Since 1966,
however, the reverse shift of income back to the per-onal sector played
a 'ol(, in the declining national saving-investmient ratio, accentuated
by a reduction in the investment/income ratio of persons. The relative
decline in total gross investment also reflected a drop in the business
income/product ratio, particularly between 1966 and 1969. Since
business has the highest investment/income ratio (greater than
unity), it is obvious that diversion of-disposable income from business
tends to reduce the national investment ratio.

Gross Capital Stocks in Relation to GNP
As a i result of the high and rising (until 1966) rate of saving and
investment, plus inflation, the current value of the total gross capital
stock increased from aboit $1.2 trillion in 1929 to 15.6 trillion in 1973,
an average annual rate of increase of 6.0 percent. The growth rate of
adjusted GNP was 6.1 percent, so the capital coefficient showed a
slight net decline from 9.4 in 1929 to 8.9 in 1973. (See table 5.) Con-
ver-ely, the ratio of product (income) to total capital increased on
balance from 10.6 percent in 1929 to 11.2 percent in 1973. With some
adjustments to convert product to factor income, this relation can be
used( to calculate the rates of return on capital, which we do in a later
section. Note, however, that after leveling out 1966-69, the capital
coefficient rose, implying a declining rate of return in recent years.

[Dollar amount in billions

Current 1958= 100) Constant

A-Absolute levels
Adjusted GNP:
1929------------------------------------------------......................................................... $127 50.5 $252
1948...-------------------------------------------------....................................... 328 77.9 421
1966-------------------------------------------------......................................................... 983 114.8 856
1969-----------------------------------------------......................................................... 1,248 130.4 957
1973.------------------------........ .... -----------------------........... 1,754 158.8 1,105
Total GNW:
1929...- --------------------....--------------------------................ 1,203 45.4 2,648
1948-----------------------------------------------................................................ 3,012 76.0 3,964
1966-----------------------------------------------......................................................... 8,518 118.5 7,187
1969----------------------------------------------.............................................. 10,907 135.2 8,070
1973-----------------------------------------------......................................................... 15,641 166.7 9,383
B-Average annual percentage rates of change
Adjusted GNP:
1929-73..-----------------------------------------------............................................... 6.1 2.7 3.4
1929-48.---------------......... ...-------------------------------. 5.1 2.3 2.7
1948-66.................................-----------------------------------------------.. 6.4 2.3 4.0
1966-69-----------------------------------------------...................................................... 8.3 4.3 3.8
1969-73.......-------------------------------------------.--- 8.9 5.1 3.6
Total GNW:
1929-73...------......----.------------------------------------- 6.0 3.0 2.9
1929-48-------------------------------- -- -------------- 4.9 2.7 2.1
1948-66 --------------------------------- -------------- 6.0 2.6 3.3
1966-69-------------------------------------------------- 8.6 4.5 4.0
1969-73.--------- -------------------------------------- 9.4 5.4 3.8
C-Ratios, GNW/GNP
Total GNW/GNP:
1929..----------------------------------- ---- ------------ 9.4 .90 10.5
1948----------------........ --.. ------------------------------- 9.2 .98 9.4
1966..-- --------------------------------------- ----.--- 8.7 1.03 8.4
1969......--.- .......------------ --------------- 8.7 1.04 8.4
1973.---------------------------------------------------- 8.9 1.05 8.5

In constant prices the relative growth of product was more pro-
nounced, since the implicit price deflator for GNP rose by 0.3 per-
centage point a year less than that for capital (which we call GNW,
for "gross national wealth"). As shown in table 5 the average annual
rate of growth of real GNP was 3.4 percent compared with 2.9 percent
for real GNW 1929-73. This is reflected in a net decline of the real
capital coefficient from 10.5 in 1929 to 8.5 in 1973. This means that
"total capital productivity" (the inverse of the capital coefficient) rose
an average annual rate of 0.5 percent. This is a measure of all the
noninvestment related forces affecting economic growth; notably,
changes in (1) values and institutions; (2) rates of utilization of
capacity; (3) actual efficiency in relation to potential efficiency with
a given technology; (4) economies of scale; (5) the degree of economic

(allocative) efficiency; and (6) changes in the inlerent quality of
natural and human resources. In addition, possible errors in the esti-
mates of capital and product, if not offsetting, could affect the pro-
ductivity estimates.
As was true of the current dollar capital coefficient, however, the
real coefficient also leveled out between 1966 and 1969 and rose
som lewhat to 1973-which means that total capital productivity
,dropped. Tlhis is consistent with other evidence on productivity.
Real product per man-hour, as measured by BLS (which does not
take account of nonhuman capital inputs) slowed a retardation in
growth after 1966; and so did Kendrick's total tangible factor pro-
ductivitv series, which takes account of tangible nonhuman as well as
huian ccapital inputs, but not inputs of real intangible capital.3
In order to quantify rates of change in productivity, it is desirable
to confine the measures to the private domestic business economy, for
which the real product and capital measures are independent. (In the
nonbusiness sectors, we impute a rental value to the capital stocks
and add it to labor compensation to estimate income and product
In table 6, line 6, it can be seen that the rate of growth of tangible
capital (human and nonhuman) productivity, which was 1.7 percent
a year 1948-66, slowed to 0.2 percent 1966-1973. (Tlhese numbers are
lower than the usual measures of total tangible factor productivity,
which adjust human capital to reflect changes in average hours
[Private domestic business economy, average annual percentage rates of change

1948-66 1966-73
1. Real adjusted gross product_ ---------------------------------------------- 4.1 3.5
2. Real gross capital stock-total------------------------------------------------- 3.1 4.1
3. Tangible capital------------..... ---..----..... ---.. ----------------.------------.. 2.4 3.3
4. Intangible capital------------------------------------.. ------------------ 4.1 5.2
5. Ratio: real total capital over real tangible capital (2-3)-------.---------------------- .7 .8
6. Tangible capital productivity (1-3)...---.--------.. --...--------------..-----.---------- 1.7 .2
7. Total capital productivity (1-2) -- ------------------------------------------ 1.0 -.6

Presumably, a major element helping to explain chlianges in tangible
capital productivity has been the growth of intangible capital per
unit of tangible capital, since the intangibles are designed to increase
the quality and efficiency of the human and nonhuman tangible factors
in which they are embodied. Line 5 shows the relative growth of
intangible capital, weighted to give effect to its smaller share of the
total capital stock than that of the tangibles. In the period 1948-66
the relative weighted growth of intangibles accounted for 0.7 per-
centage point, or over 40 percent, of the 1.7 percent increase in tangible
capital productivity. Total capital productivity (the relationship of
real product to total capital, intangible as well as tangible) grew at an
average rate of 1.0 percent a year. This reflects the net effect of the
half dozen major noninvestment forces noted earlier.
The contrast of the 1966-73 period is startling. Even though in-
tangible capital grew even faster relative to tangible capital than in
, See John W. Kendrick, Postuar ProJuteit fy Trends in the United States, 1948-1969 (New York: National
Bureau of ECOioiMii, Research, 1973).

the 1948-66 period, its 0.8 percent a year weighted relative increase
was associated with only a 0.2 pler'nt a year increase in Itanrible capital
productivity, and thus with a 0.6 percent annual decline in total
capital productivity!
The marked deceleration in productivity, based( on this and other
measures, appears to be due to a number of factors. Thlie slower rate of
growth after 1966 meant fewer opportunities for economies of scale,
of course. The bulge in labor force growth after 1965 reduced the
average experience of workers, and slowed the growth of re.:l product
per worker for the time being since compensation and value added of
you:m workers are below average.
The rate of utilization of the labor fore, was lower in 1973 than in
1966 (4.9 percent uneml)loyment vs. 3.6 percent); yet there were
capacity bottlenecks in many basic industries, e.g., steel, aluminum,
paper, and petroleum. This suggests that there had been inadequate
busine- ss tangible investment in the several earlier years, and po--ibly
some misallocation of investment. The inadequate amount, in view of
the rapid growth of the labor force, was related to a declining net rate
of return on investment, especially when adjustments to profits are
made for revaluation of book depreciation charges to replacement co-t.
The declining rate of return reflects the use of macroeconomic policies
to combat the accelerating inflation which, on balance, held increl(a-es
in the price level below increases in unit costs. Some misallocation of
investment probably resulted from the wage and price control pro-
grams from August 1971 to April 1974.
Further, the increasing amounts of inve-f ment required for environ-
mental protection and occupational health and safety reduced the
proportion available for direct productive purposes. Since the benefits
of these programs are not reflected in real product but the invest ients
are reflected in the real capital measures, the programs tend to reduce
increases in productivity as measured.
It also seems probable that the relative decline of R. & D. invest-
ments, and the leveling out of the relative R. & D. stock (see below)
tended to slow down productivity advance, since R. & D. is the foun-
tainhead of scientific and technological advance.
Finally, there were various negative social tendencies, particularly
in the latter 1960's, which probably reduced productivity growth.
Examples are increase ing drug use and crime, increased( anti-e4stablish-
ment and antibusiness sentiment, and a po-ible loo-cninii of the work
ethic. The development of social indicators has not yet re-.ched the
point of permitting the quantification of the economic impacts of these
and other social developments.

Composition of Gross Wealth
Just as intangible investments rose as a proportion of total invest-
ment-, up until the latter 1960's, so have intangible capital stocks
increa -ed as a proportion of total wealth-from about 2:{ )percc'it in
1929 to almost 40 percent in 1973. (See table 7.) But it is apparent that
the relative growth of intangible capital has been decelerating since
1966. This importantly reflects the leveling out of the ratio of R. & D.
stocks to total capital after 1966, following on the most rapid relative
growth of any form of capital. But the relative growth of both educa-
tional and health capital, which had exceeded 50 percent between 1929

and 1916, also slowed down. Yet thlcir prolportiois of the total con-
tine'd to expand slowly, as did that of mniobility costs, which lad
d lroppled SOiewhat prior to the miid-1960's.

JBy type and by sector; selected years]

Tangible capital Intangible capital
Year Total Human Nonhuman Total Human (R. & D.)

A. Percentage distribution by majcr
1929-------------------------........................... 76.8 24.5 52.3 23.2 23.0 0.2
1948 ------------------........................... 73.0 21.3 51.7 27.0 26.4 0.6
1957-------------------------........................... 68.9 18.1 50.8 31.S 29.7 1.4
1966-------------------------........................... 63.2 16.3 46.9 36.8 34.3 2.5
1969...............------------------------- 61.5 15.2 46.2 38.5 35.9 2.6
1973-------------------------........................... 60.2 15.1 45.1 39.8 37.2 2.6

net foreign
claims as
percent of
Year Personal Business Governments GDW

B. Percentage distribution, by major sectors:
1929--------------------------------------........................................... 58.0 30.6 11.4 1.4
1948............-------------------------------------- 56.2 22.4 21.4 1.3
1957--------------------------------------........................................... 55.9 22.6 21.4 .9
1966--------------------------------------........................................... 55.4 21.9 22.7 .8
1969 ........................................... 54.9 21.6 23.5 .6
1973--------------------------------------........................................... 55.5 21.3 23.2 .4

With respect to tangible capital, .human tangibles, while rising
absolutely as the adult population grew, hlad declined relatively until
the latter 1960's, when the proportion leveled off as a result of the
bulge in labor force growth. Nonhuman capital grew less rapidly than
the total throughout the period, due particularly to the relative
decline of land and other natural resources. Machinery, equipment,
and nonbusiness durables were the only type of tangibles whose
ratio to total capital expanded.
When the various types of capital are recombined in human and
nonhuman categories, it is seen that the human share rose steadily
from 47Y percent in 1929 to almost, 521 percent in 1973. The relative
growth of human intangibles more than offset the relative decline of
tangible human capital. The recombination of capital types into
human and nonhuman groupings is helpful in computing rates of
return, which we discuss in the next section.
With regard to the sectoral composition of capital (see panel B
of table 7), it is apparent that the share of general governments more
than doubled over the 40-year period 1929-69, before receding slightly
1969-73. The relative growth of public wealth was primarily at the
expense of the relative hliare of business, which declined steadily, and
to a lesser extent at the expense of the personal sector. But the personal
sector share did grow somewhat 1969-73, while the business sector
share continued to recede. Note also that net foreign claims (including
monetary metals), despite a substantial absolute growth, declined
in importance relative to domestic wealth throughout the period.

Rates of Return on Total Capital

Rates of return can be computed by dividinri factor compen-ation,
gross or net of depreciation, by thle value of gross or net total capital
stocks. In addition to overall rates of return, returns on human and
nonhuman capital may be calculat,,1 separately by split tini national
income between labor and property compensttion and (divi(ling by
the corresponding wealth estimate-.
As shown in table 8 gross and net rates of return on total capital
exhibit similar levels and movements. On a net basi, the 10.0 percent
return in 1973 was virtually the same as it was in 1929, on the eve of
the Great Depres.ion. Returns du'iig the early postwar period were
si,,'iificantly higher, reflecting the capital shortages u'ried over from
the depression and World War II, in conjunction with ,,nerally
high levels of aggregate demand. Rates of return in 1957 and 1906
were below the 1948-53 levels, reflecting more ample capital supply,
but still well above the 1973 rate. The 1960 rate was lower than
1957, reflecting the incomplete recovery from the 1958 contraction;
and the 1969 rate was lower than the 1966 rate, reflecting the restric-
tive monetary and fi-cal policies adopted to combat inflation. Rates
of return in 1973 were still lower, due to the net effect of wage and
price controls as well as continued restrictive macroeconomic policies.

[In percentages; selected peak years]

Total Human Io0nhuman

A. Cross rates of return:
1929--.......--------------------------------------------------- 10.2 11.7 9.2
1948------......-....-------------.--------------------------------- 12.1 12.2 12.0
1953..----------------------------------------------------- 12.1 13.5 10.8
1957..---------------------.. --------------------------.........-----.-.. 11.4 12.7 10.1
1960------.........-----------------------------------...---------- -- 10.0 12.3 9.7
1966----------------------------------------------------- 11.8 12.2 11.4
1969----------------------------------------------------- 10.8 11.7 9.9
19731---------------------------------------------------- 10.4 10.8 10.1
B. Net rates of return:
1929--......--------------------------------------..-.------------ 10.0 10.1 10.0
1948. ---.---------------------------------.--------------- 13.4 12.6 14.2
1953.----------------------------------------------------- 13.1 14.8 11.4
1957-----------------------------------------------------...................................................... 11.6 13.4 9.9
90-......................................................... 11.0 12.9 9.2
1966------------------------.......----------------------------- 11.4 12.8 10.7
1969.-----...............- ...... --..........-- .....--..- ....- 10.6 12.2 8.9
19731---------------------------------------------------- 10.0 11.2 8.8

1 Preliminary.

Some analysts prefer to look at gro-.s rates of return, which do not
require a necessarily somewhat arbitrary division of gro- income be-
tween depreciation and profit. But the gro- rates tell much the same
story, except that they were fractionally higher in the boundary years
than the net rates, and did not rise as much in the 1948-53 period.
The two sets of rates diverge somewhat with nr-pect to human and
nonhuman returns, however. On a net basis, the rates of return on the
two type' of capital were almost the same in 1929, but by 1953 the
human return siornficantly exceeded tIe rate of return on nonhuman
capital. Although both trended downwards after 1953, except for the
1960-66 improvement, human capital continued to enjoy a higher rate

of return than the nonhulmman. But between 1969 and 1973, the decline
in tlhe return on human capital was much sharper than that on
On a gross basis, the rate of return on human capital started out in
1929 significantly higher than that on nonhuman. Thereafter, the
pattern was similar to that for the net rates, except that by 1973 the
differential between the two sets of rates lhad narrowed substantially.
In fact, between 1969 and 1973 while thle gross rate of return on human
capital dropped markedly, that on nonhuman capital rose slightly.
This buoyancy apparently reflected the shortages of business produc-
tive capacity, despite the downward pressures on overall rates of
return on capital as a whole.

Concluding Comments
In order to promote the re-;umption of economic progress (defined
as increasinIg real income per capital which most. Americans seem to
want, the material in this paper trongly suggests that an acceleration
in the rate of capital formation is not required so much as an accelera-
tion of productivity advance. Here, we are not referring just to the
cyclical recovery of productivity which is currently underway and will
extend well into 1976. Rather we are referring to an acceleration of the
/rt:(ld-rate of productivity advance, at least. back to the 1948-66 rate.
Of course it would be helpful at least to halt and possibly to reverse
the drop in the national total saving-investinent ratio which began in
1966 after two decades of advance. But the rate of growth in real total
capital stocks accelerated to a record rate between 1966 and 1969, and
decelerated only modestly 1969-73. The problem is that tlhe rate of in-
('reat in real product lalred, indicating a drop in total capital pro-
ductivity, compared with significant advances 1948-66. This suggests,
in turn, that a major attack on the problem must come through some
reallocation-; of investment and capital, in order to increase capital
InII my view the most important policy objective which would both
reverse thie downward trend in the saving-investment ratio and im-
prove the allocation and productivity of capital would be to increase
the disposable (after tax) income of the business sector as a percentage
of GNP, reversing tlhe decline which began in 1966. Since the busines i
-ector con.,i.tently plows back all of its gross disposable income, and
more, into investments, an increase in its relative income would ob-
viously tend to increase the national saving-inve-tment rate. The
shortages of capacity encountered in 1973, the eventual capacity
requirements of the current exl)panion, the continuing )pressures for
cost-reducing innovations, the further capital requirements of social
programs (EPA and OSHA in particular) and domestic energy pro-
grams all point to the desirability of a faster relative increase in busi-
ne-s investment.
One way to accomplish this objective would be the pursuit of mone-
tarv and fiscal policies during this economic expansion that would per-
mit the restoration of higher rates of return than were permitted by
the restrictive, anti-inflationary policies followed in the recoveries of
1967-69 and 1970-73. Possibly the adoption of a stronger incomes
policy would be called for later in this expansion as an alternative to a.
profit squeeze leading to contraction.

An alternative or supplement to the above policy would be a reduc-
tion in business income taxe.s. This could take one or iore of severall
forms: An increase in the investment tax credit; a decrease in corpo-
ration income tax rat(,-; further reduction or elimination of the double
taxation of dividends; recognition of "inflation accounting," patirlic-
ularly the restatement of depreiation charge,, from historic book
costs to current replacement costs, in calculation of business income
subject to tax; and possibly the institution of a R. & D. tax credit of 10
to 20 percent, or possibly a larger credit on incremental R. & D.
The last proposal, although a new one, seems particularly appropri-
ate at this time to reverse the relative decline on R. & D. inve-tlent.
Studies by Terleckyj, Griliches, Mansfield and others all indicate a
high productivity effect and rate of return on R. & D. outlays, partic-
ularly those designed to improve pro(lucers' goods an(d proce.:es. In
my view, a most important element in rai in.: the productivity trend
is a renewed relative growth in the stock of knowledge and know-how
resulting from R. & D. embodied in men and machines.
This brings us to the public sector. There, too, I believe that a
stronger and steadier expansion of expenditures to support R. & D. is
desirable. It will also aid in maintaining and possibly increasing the
national saving-investment ratio if the public sector continue, its
relatively high ratio of investment to disposable income. As to alloca-
tion of investments among the alternative types discussed in this
study, and among specific projects, more work is needed to refine co-t-
benefit and prospective rate-of-return estimates, and to regularize
capital budgeting.
Generalized social and private rate-of-return studies would also
help in guiding personal sector investment decisions. For example,
recent studies indicate a sharp drop in rates of return to investing in
higher education between 1969 and 1975.4 As a result, the fraction
of young men choosing college has declined. Our numbers also suggest
that the incremental returns to human investment predominantlyv
education) had dropped faster than the incremental rate of return on
nonhuman investment 1969-73, but we did not include rate-of-return
estimates by category of investment, by sector. Support for develop-
ment of more such estimates could have a big payoff in making p1)os-
sible more rational investment deci-ions, and a more efficient capital
As far as the personal sector is concerned, it is signific,.tnt that the
ratio of investment to disposable income has drol)pp)ed from the peak
rate reached in 1966. In part, this may reflect a decline in expenditur.-
for rearing children as a fraction of DPI. But it does indicate that the
potential for higher saving and direct investment is there. Investifment
in self and in one's family could be stimulated byN more generous
deductions in the personal income tax for outlays for education,
training, medical purp)o-(-, and mobility. Recommendations for
liberalization should not be made, however, unless stu(lies indicate a
sufficiently high social rate of return on such exp)enditulnhrs.
In conclusion I mu.-t note th;it my ob-ervations on p)o-ible tax
l)olicie, to promote capital formation stem from mny personal interl)re-
tations of the material preu-cntedl here, plus my general rea(linlg of
4 Se" Richard R. lri-.inman, "Overinvestment in College Trainirng?", 7TheI Journtal of Human
Resources, X-2. 1975).

1111111l'iBtl IN I 1111111 Ut1111111
14 1 3 1262 09113 3693

reelit ,, economic history.5 Others may well come to different conclu-
sioli-. But regardless of differences among analysts, the broader view
provided by the new total investment and capital estimates should
ielp) in clarifying the issues and reaching a sounder policy consensus.
5 h.iving ,lr.Itvil this paper I was interested to find lmch the same viewpoint ably expressed by
.'i C T. Soimi r, "Six ial ,;i)als nnd EtL'onomic Growth-The Policy Problem in Capital Formation,"
The Conftrence Board Record, December, 1'.75.