PROPOSED PROJECT: DETERMINATION OF LABOR INPUT REQUIREMENTS FOR (1) INDUSTRIES COMPRISING THE 80-ORDER CLASSIFICATION END (2) SPECIFIC PRODUCTS INCLUDED IN THESE INDUSTRIES, BY BOTH PHYSICAL AND VALUE UNITS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78-04718A002600390047-2
• Release Decision: RIFPUB
• Original Classification: S
• Document Page Count: 10
• Document Creation Date: December 23, 2016
• Document Release Date: March 12, 2014
• Sequence Number: 47
• Publication Date: January 1, 1947
• Content Type: REPORT

Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
Tab A
S-E-C-R-E-T
S.* ISL., ???? *an.
Proposed Projects Determination of Labor Input
requirements for (1) industries comprising the
80.order classification and (2) specific products
included in these industries s by both Ifteical
and value units
Introduction
After exploratory investigation and discussions with technicians in
the manpower fields It has been ascertained that the basic data required
to conduct a project of thio nature are available from the 1947 Census
of Manufactures and individual studiee conducted in the Bureau of Labor
Statistics, particularly the Divisions of Interindustry Economicss Pro-
ductivity and Technological Developmentss and Manpower and Employment
Statistics, It has been suggested that a thorough review of the indivi.
dual BLS studies in the productivity and emplortant field be made to es-
tablish to what degree of detail it is possible to extend unit man-hour
requirernents on a product detail basis which would be consistent with the
overell primary product industry approach. For labor input requirements
based on the 80-order classifications procedures can be developed. from
1c47 Census data which will yield consistent results, As outlined belows
the asic problems apart from the choice of proper standards of measure.
ment relates to the transformation of Census data on employment and out-
put Aeom the establisheent basis to the primary product wherever produced
one. The recently established output control totals for the IO (O
sector) industries defined on a primary product wherever produced basiss
will enable properly weighted labor inputs to be obtained. Before pro.
ceeding to a discussion of the procedures proposed to obtain measures of
laboe input requirements on an 80-order industry primary product basis,
it ie worthwhile to discuss the three possible measures which can be em.
ployed to describe labor input requirements and the advantages and diee
advantages of each,
I, General Discussion of Labor Input Coefficients
Labor input eeacuremente relate to tll.. eeount of labor vequirede
Le dollar or plefeicel tarnes to produee A unit of output; they likewiss
:ay be expressed In dealer or physical torte. The labor compament included
in the labor input per mit of output may refrr to "production and related
yorkorei or to "all employeeW which inaludta production and related
vsorkore" and "all ono? eeployeeo," 1/ :it invo exception of the unit
17797477?ensua? of Naeaveturess Vol. lin pp. 12elhe
e 4 e
S.E.CReEer
sa. '1.? q1,2 34,40
npriacsified and Aooroved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
S-E-C-R-E-T
man-hour requirement approach, labor input per unit of output can be
readily expressed in terms of "all employees" or "production and re.
lated workers."
A. Labor C22fastentil Based on ilmo3j,_..a
The labor input coefficient which most closely approximates the usual
interindustry flow coefficient relates to the ratio of tctal annual wages
and salaries of an industry-i.e., payrolls or money cost of labor-to the
annual monetary output value of the industry. The resultant relationship
is described as the labor input in monetary terms per designated monetary
unit of output. This measure is inapplicable when the objective is to
compare relative labor input requirements for the same industries over
time or for different industries or economics. Payroll data used in this
particular index reflect not only the industrial techniqees, economic
activity levels, and labor force compositione.i.e.? full time, part time
employeeseebut also the accompanying institutional factors associated with
the indalstry. Thus payroll data include payments for such items as vaca-
tion,, sick and dismissal pay, and nonproduction bonuses.
B. Labor Coefficients Based on Numb.se_gzem!
A second index which can be employed is the ratio of either the average
yearly number of "all employees" or "production and related workers in an
industry" to the total yearly monetary output of the industry; this maybe
described as the labor input in physical terms per designated monetary unit
of output. This particular labor input is beset with definite limitations
which can give rise to serious instability in the coefficients. Although
the total number of workers in an industry may not change significantly,
yearly hours worked per worker may register wide variations from year to
year in an industry due to changing economic conditiona. Thus, if the
total number of workers in an industry remains constant for two periods of
time and productivity does not change for the two time periods, labor in-
put coefficients for the industry may register a decline for the latter
period because of increased output over the base year due to an increase
in yearly hours worked per worker. Itt Notwithstanding these aforementioned
limitations, this particular index of labor input requirements per unit of
output mays as indicated below, serve a useful analytic purpose and should
not be summarily discarded,
r."?,tation may result from changing proportions Lutween full and
part time employees and by *standing the work week at cull and part time
*splay-acts.
t
SeEeC.R.E.T
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
SE-C-R-E-T
C. Labor Coefficients Based on Maxi-hours
The unit manehours criterion used to measure labor inputs is defined
as the ratio of the total number of man-hours worked per year by "produce
tion and related workers" in an industry to the total yearly output of the
industry in value terms. Mille the physical labor input coefficient per
designated unit of monetary output is seriously hampered by changing cone
ditione? other than technological, over time, the unit manehours require.
ment index appears to imply straight-line proportionality in relation to
the scale of production and thus maybe inadequate, in itself, to indicate
the degree of increased output per worker per year, or conversely, the
physical labor input requirements per year associated with a stipulated
increase in output. The industry unit man-hour requirements which can be
computed from Census data represent an average for the whole year. Howe
ever, if unit nmnehour requirements are not constant over the entire range
of the working year, the ratios obtained may have built-in biases which
limit the validity of the proportionality assumption. 2/
D. Observations
For individual industries or for the more basic operating units, the
problem of unit man-hour requirements may be a more important considerae
tion than the physical labor input employed in producing the annual out-
put. The former has impact upon the financial status of the operating
units, while the latter, being conditioned by increased average hours
worked yearly by full or part time employees, may continue to fall and re-
sult in rising unit man-hour requirements and, consequently* higher cost
per unit of output, assuming wage rates to be constant. However, when
mandatory increases in output, irrespective of labor cost, are part of a
national program, physical labor input coefficients may SWUM considerable
importance* the underlying assumption being that the labor supply is more
inelastic over a period of time for all industries as a whole (i.e., the
nationel economy) than for specific industries. On an overall national
scale, increasing man-hour requirements may serve as indicators and regu-
lators of the limiting conditions imposed in the reduction of the labor
input coefficients for a given labor force, Thus national considerations
might Cemand that certain production be carried to the point where mar-
ginal unit man-hoer requirements approach infinity, and, therefore, physical
labor input coefficients become irreducible*
7-111S'Oruairiy man-hour statistics reported in the Census of Manufactures
and the Annual Surveys of tenufacturos refer only to production and related
works and do not include the category of "all other ebc4yees." For cer-
tain industries this understateeent of men-hours for "all employees" ele?Y be
considerable*
'.3 -
11?11. Oar ???? 0110 MVO
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
. Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
S -R-E7T
E. General Conclusions
Clearly, both unit man.hour requirements and physical labor input
coeffie.ents ("production and related writers") per unit of output are
desirable indices to have available for analytic purposes where labor
requirements are an essential part of an overall national economic
study. In utilizing unit man.hour requirements for inter-year compari-
sons in order to project manpower requirements, changes in average hours
worked or stipulated per year may be worthy of separate examination inde-
pendent of the study of change in unit man-hour requirements resulting
from the application of advances in teohnology. Productivity studies
which are not based on the division of total man-hours by annual output
but on individual performance studies may prove useful in discerning pro-
ductivity changes which do not contain the elements of a yearly average.
In respect to yearly hours worked per worker, if significant increases
are postulated, it may be appropriate to examine the implications in terms
of changing unit man-hour requirements and physical labor input require.
manta. An estimate of the hours worked per year per worker in an industry
can be obtained by dividing the total number of man-hours worked in the
industry by the total number of workers employed in the industry. Esti-
mates a? average hours worked per year are in themselves rather crude
tools welen used to appraise the possibilities and extent of expanding out.
put by utilizing the labor force more intensively. Further useful infor-
mation would include composition of labor force (i.e., full or part time
employees, sex, age) to determine relevant factors tending to restrict
expansion of average hours worked per year (i.e., family responsibilities,
physical limitations, etc.). Always lurking in the background in this
type of analysis is the assumption of sufficient plant capacity and the
maintenance of the necessary synchronized flow of materials. Thus, de-
pending upon an analysis of these various factors which affect labor re-
quirereets, different assumptions may be made as to limitations imposed
on marginal unit man-hour requirenents and physical labor input coef-
ficiente.
Summarizing, it would appear that estimates for both unit man-hour
requirements and physical labor input coefficients per designated monetary
unit or output should be attempted. I have not as yet had the opportunity
to conduct a careful analysis of labor input reqeirements for specific
products, in physical and value units, included in the 80 order industries
and therefore have refrained from =king any substantive coulents on this
subject. I-expect to prepare a nowt owthis tentative phase of the pro-
ject at 4 later date.
As a final note to the general discussion, it shou2d be .1Anderscored ?
that an interindustry and inter.osecnomy inferences to be drawn from allY
set of labor input coefficients computed according to tLe indices outlined
- 4 e
Sae.C.R.E.T
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
above merit the closest examination to ascertain their validity or, at
least, their appropriateness. Labor input coefficients should not be
used for value judgments among industry groups within an econogy; for
example, one should not infer comparative efficiency or inefficiency as
to the use of labor among industries. Labor and plant and equipment may
all be equal, but a variance in labor input coefficients may occur among
industries simply because of differences in cost of materials in rela-
tion to output resulting from the particular stages of production with
shich particular industries are concerned. Even if labor and plant and
equipment are not the same, variance in labor input coefficients among
industries might occur because of differing degrees of industrial inte-
gration among industries. It is conceivable that labor inputs, when comm
pared to value added for the industries in westion? might yield similar
ratios but would show extreme variations when compared to their respective
outputs, More appropriate inferences, and perhaps significant as indica-
tive of policy decisions, might be drawn from the relative dispersion of
labor input coefficients among industries and grouped by annual time
periods. This could be shown in graphical form by plotting labor input
coefficients along the ordinate and industries (i.e., from 1 to 80) along
the abscissa.
If any attempt were made to estimate laboi input coefficients for
similar industries located in different economies from information ob-
tained for industries in the United States economy, it would be necessary
to proceed with extreme caution and buttress the estimates with additional
information, particularly as to the relative industry capital structures in
both economies. y What would appear to be required is information cone
corning the scope of the manufacturing processes (degree of integration)
encompassed for comparable industries located in the domestic and foreign
economies and the capital associated with each manufacturing Step in the
integrated industries. Comparable capital structure bases could be ob.
tamed ty subtracting from the relevant industries the capital associated
with the integrated process tending to create the incorparability. In
additior to examining the degree of integration for similar industries
located in two different economies in order to neutralize these differences,
the amornt of capital plant and equipment per worker would undoubtedly be
an in: of the utmost significance for attempts at estimating labor input
coefficients for other economies based on estimates derived from the United
States ecenomr. If extremely questionable estimates are to be avoided, it
would speer that efforts should be directed to ascertaining, at a minimum,
the verying effects on the relevantlbor input doeffietients of differing
capital structures for comparable industries located in different economies.
y7-ps,as a general summary expression, it may be stated that Lebec
neat Coefficients for similar industries located in different economise
'my 'vary due to differences in capital structure, &wet., a integration,
ocepositton of labor force, hours worked, institutional factors, etc.
- 5 -
SeEe0eReEeT
Declassified and Approved For Release 2014/03/12: CIA-RDP78-04718A002600390047-2
*oeclasS-ifiedandApprovedForRelease2014/03/12:CIA-RDP78-04718A002600390047-2
S-E-C-R-E-T
???? SO. OM. 4?44
II. Operational Procedure to Derive Labor Input Coefficients
for Primary Product Industries
In order to derive "production and related workers" require-
ments for primary product industries based on physical labor input co-
efficients and unit man-hours per monetary unit of output, it is necese
sery to have yearly primary product industry data relating to the number
of "production and related workers," total man-hours worked by "production
and related workers," and value of output. These data are either already
available, as in the case of primary product industry output control
totals, or can be derived from the statistical compilations included in the
109147 Census or Annual Surveys of Manufactures (190, 1950, 1.951, l952)0
Census industry data for "production and related workers," arebased On
the establishment concept and hence require adjustment to accord with the
concept of primary product wherever produced before they can be employed
in the ?resent project. Below is an adjustment procedure intended to
transform the relevant Census data to the proper base and thus enable the'
computation of the desired labor input requirements for primary product
industries.
A- teroffeniety
In the Census of Manufactures industry statistics, the data caviled
refer to both primary and secondary products produced in an industry.
Consequently, the relationships between the labor employed in an industry
and the output of that industry which can be derived from Census data are
-inapplicable for this project, Schematically, the labor input coefficients
(and ure.t man-hours) which can be derived from Census data may be presented
ae folleeet
1. Labor input coefficient for Census industry i 5/
?
a0 labor input coefficient for primary products of
Census industry i (output of primary product of
Census industry 1) t
b, labor input coefficient for secondary products
of Census industry i (output of secondary product
of Census industry ?I).
c. Beth (a) and (b) are divided by total output of
Ceeeus industry i.
20 A allilar presentation may be eada for unit man-heu(te
Er7LT-Titfars P
111011.Akbee ) ?(i) io(i)0
tAD i PP 0 ilsp -1
or alternativay
) LIC(Opp LIC(i)sp =
6
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
. Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
=MP 11.1. IMO Ono ?????
;ontirrued
tilerefore
20.) LIC(1) ?= LIC.1(i)pp a 0(i)pp 4.. LIC(i)sp /0( 1),
N.rheret LIC(i) = 3.abor input coefficient for Census industry
P = "production and related writers" for :;ensus industry
o( i) = output. of Census industry 1.; and subscripts pp and sp
refer to piary products and secondary products, respectively?
,Obje.ctives of the lid ustment Procodure
Given primary product industryoutput control totals, which wore
derived in order to comp..te primary product irxut coefficients, the prob-
lem in this project is xeducod to detemininn for each Census industry
the distribution or the amber of f,production and related writers" and
their total man-hours between primary and secondary production and -then
transferrinc the estimates relatinc to secondary production to the proper
primary ;,..,1roducine industries? The rationale for this transfer procedure
is outlined belotn.
Total ic of "production and related workers" in Census
inalstry
less: number of "production and related workere attributable
to secondary production of Census industry
plus: nonbor of "production and related uorhors" of other
Census in?iustIles attributable to t o production outside
of 1Lr,...E3 industry A of its r: product,
?goals: I Its? :Lc f.er ot Census ". a.loi related uorkers'.
atirltutunle to the pxduction ol! products classified as
pr ry Census industry in 2'i
4, similar prowl-lath= is required for 7mn.heurso y
q--T-613s. -ffaav? Rafor Census industries were combined to ront T.0 indue-
.ries and thas the .3perstion described in Ott tine* can be palmed c*
mere .aggregativebie tilan indicated,
if he purpose of this project, estivate" have to be Dade of the
umber ot. proirlootima end ..clated worlmrs _ -1.-.;:i-ars) that wou.l. to-
r e.quircit prviucti ettepetitive inports anti inventory depletione for
letteyiejr,.
8' -4.4-
a
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
S-E-C-R-E-T
C. The Proposed Adustment Procedure
A preliminary operation in the adjustment procedure is the aggrega-
tion, where necessary, of Census industry data for production ane related
workers and their total manehours to the 1-0 industry definition. For
ease of operation, the next step is to determine the taunter of production
workers and their total manehours attributable to primary products of
Census industries produced in Census industries other than the producing
one (the soecalled "transfers-in"). 2/ If the assumption is mace that
the manehours required to produce a designated unit value of primary pro-
duct in industries other than the primary producing one are the an as
the manehours required to produce a designated unit of output in the
primary producing industry, total manehour estimates attributable to
"transfers-in" can be derived. The total value of "transfers-inv meItie
plied by unit man-hour requirements of the industry receiving the "trans-
fersein" can be utilized as an estimate of the total number of Ion-hours
attributable to the production of primary products outside of the primary
producing indmTMry. The concomitant estimate for the nueber of "produce
tion and related workers" attributable to "transfers-in" cant secured
as follows
For each industry transferring out secondary products, it is possible
to estimate the hours worked per year per worker by dividing the total
man-hours worked by "production and related workers et by the total number
of production workers. Assuming that for the traneferring-out industry
the hours worked per year per worker is the name for both primary and
secondary production, an estimate of the number of workers attributable
to each "transfer-in" can be obtained by Cividing the man-hour require-
ments for each "transfer-in" by the hours worked per yearper worler in
the transferring-out industry. .The increments for the "number of pro-
duction and related workers" and their total manehoure resulting from
industries receiving "transfers-in" need t? be offset, of course, by
appropriate decrements to the transferring-out industries.
Since Census industries usually ireaude secondary products in their
outputs, unit manehour requirements for industries which are computed
from Census data, in order to estimate total man-hours for "trensfers-
contain a bias. It is believed that where secondary products (trans-
fers-oul are not sieniricant, the resultant bias in the labor input
measurements will not be significant0 however, for industries in which
the production of oecondary products is important (e.g., more than lOper-
cent of direct .inauto)? the bias maybe sektantial. It is preponed,
.4.7.7".-reTe.aMtl information for egreonder, products is availible from the
;ark done for t'e turrent priecry product industry project.
SeE-Genene?
????? 41.
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
S-E-C-R-E-T
therefore, that the ?transfer-in" adjustments be performed first for
industries where the impact of secondary products is not considered to
be significant. Then intermediate estimates can be made to include all
relevant revisions resulting from recording "transfers-in" for the rela-
tively homogeneous industries and from partial revisions on the "trans.
fers-out" aide for both the relatively homogeneous and heterogeneous
industries. The objective of this intermediate computational stage is
to reduce as much as possible for heterogeneous industries the number of
"production and related workers," their total man-hours, and the industry
output attributable to eecondary products. After completion of this
stage, "transfer-in" revisions for the heterogeneous industries are to be
undertaken following the procedure outlined above.
After all relevant data are available in terms of primary product in-
dustries (i.e., number of "production and related workers," their total
man-hours, and monetary output), the data must be aggregated to the 80.
order industry classification before the desired physical labor input
coefficients and unit man.hour requirements are computed.
Do Coefficienta for
is .proposed that the data in the 1947 Census of Manufactures be
employed to obtain the basic estimates. If desirable, labor input re.
quirementsestimates for the years 1949, 1950, 1951 and 1952 can now be
constructed. A series such as indicated would require several adjust-
ments to the data in order to achieve comparability with the bane year,
1947. The "nueher of production and related workers," their total man.,,
hours, and Census industry output appearing in the Annual Census Survey
of Manufaetures would have to be adjusted for industry definition (i.e.,
from Census industry to I-0 industry and front the I-0 industry establish-
ment basis to a primary product one) and for changes in price level of
output in relation, to the base year.
The Bureau of Labor Statistics has developed interindustry wholesale
price indices (1947 8 100) for I-0 industries for each subsequent year
throsgh the first half of 1953. These price indices can be applied to
appropriately aggregated annual Census output data to achieve compara-
bility with the lease year. Scalar factors (coefficients) employed in
the primary product industry project tc achieve a transedrmation of out-
put control totals tram an establishment to a primary product industry
basis can be emeleeied to adjust the annual price corrected industry out-
put estimates derived 'rot Census data0
In order to *Move labor input relationships Which are consistent
with the prieere- !Product industry definitien, the following adjustment
should be *Wit
. 0 .
S.E.C.R.E1
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2
CONFIDENTIAL
1* The statistics total industry man-hours for "production and
related workers" (on an I-0 industry basis), can be adjusted
by applying a coefficient which will transform man-hours from
an establishment to a primary product industry base* This
coefficient maybe defined as the ratio of the 1947 primary
product total manehours estimate for "production and related
workers" to the 1947 total man-hours for "production and re-
lated workers" on an establishment basis.
2. Total nuMber of Census "production and related workers"
(adjusted to an I-0 industry basis) can be transformed to
the primary product industry concept by applying a factor
representing the ratio of the 1947 primary product industry
number of "production and related workers" to the 1.947
number of I-0 industry production and related workers on an
establishment basis*
It it were desired to adjust the physical labor input coefficients
to the 1947 bases an average hours worked per year per worker series
(properly weighted with 19)47 e 100) would be recleirede Unddjusted
estimates of physical_ labor input coefficients for years other than
1947 may prove useful, along with information available at the same
time concerning the relative change of average hours worked per year
per workers in analyzing the potential maximum utilization of a given
labor force* The difference between a stipulated potential maximum
utilization of labor and the utilization of the labor force for any
given period may indicate the possibilities of intensifying labor force
utilizations assuming the availability of necessary plant capacity and
materiels*
Declassified and Approved For Release 2014/03/12 : CIA-RDP78-04718A002600390047-2