POPULATION AND LABOR FORCE ESTIMATION PROCEDURES FOR SOVIET CITIES

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-01043R002000060008-1
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 203
• Document Creation Date: December 27, 2016
• Document Release Date: May 31, 2013
• Sequence Number: 8
• Publication Date: August 7, 1957
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
4
-SECRET--
U. S. DEPARTMENT OF COMMERCE
Bureau of the Census
.0
?00
International 'Population Reports
Series P-95, No. 49
August 7, 1957
50X1 -HUM
CD\c-c?
POPULATION AND LABOR FORCE ESTIMATION
PROCEDURES FOR SOVIET CITIES
by
Foreign Manpower Research Office
Bureau of the Census
Prepared under contract with the
Air Force Personnel and Training Research Center
Air Research and Development Comand
Department of the Air Force
50X1 -HUM
50X1 -HUM
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?
50X1 -HUM
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Department of Commerce
Bureau of the Census
SEEM'
50X1 -HUM
International Population Reports
Series P-95: No. 49
August 7: 1.957
POPULATION AND LABOR FORCE ESTIMATION PROCEDURES
FOR SOVIET CITIES
by
Foreign Manpower Research Office
U. S. Bureau of the Census
Under the Contract Monitorship of the
Intelligence Methods Branch
Office for Social Science Programs
Air Force Personnel and Training Research Center
Air Research and Development Cormsnd
SEOGIETT
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
50X1-HUM
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
50X1 -HUM
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
-
SENO
PREFACE
During the years 1953 to 1955, the Foreign Manpower Research
Office of the U.S. Bureau of the Census carried on a series of research
studies in demography directed towards the estimation of certain types of
data for cities and other urban agglomerates. The products of this research
are embodied in 47 reports covering 47 areas. For the most part, the areas
for which estimates were made were properly termed cities insofar as
boundaries could be defined. In some instances, however, the esthiates
pertained to larger urban units comprising both a central city and selected
suburban areas deemed to have demographic, economic, or military significance.
Each study featured four types of estimates, namely: (a) estimates of total
Population; (b) estimates of population distributed by age and sex; (c)
estimates of population by ethnic composition; and (d) estimates of labor
force size and industrial composition.
The principal purpose of this report is to present and discuas
analytically some of the problems that have been encountered in this
research, and to set forth and evaluate some of the techniques developed to
cope with these problems. Because of differences among cities with respect
to data available, and because of imPortant variations in city structure
and function, complete standardization of method has still not been realized.
Yet there are a number of methodological elements in common which have been
handled successfully, and there are also a number of elements which, although
not as yet susceptible to a standardized approach, offer hope that they will
become so after further research.
SEMET
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
SEMI
-
The entire research program, the methodology of which is considered
in this report, as conducted by the Foreign Manpower Research Office under
a series of transfer-of-funds arrangements with the Air Force Personnel and
Training Research Center, Air Research and Development Command, USAF, and
its predecessors-in-interest. The monitorship, guidance, assistance, and
sympathetic encouragement provided by these supporting agencies and their
staffs are gratefully acknowledged.
Comments and criticisms of the material covered in this report are
invited and should be addressed to the Foreign Manpower Research Office, U. S,
Bureau of the Census, Washington 25, D. C.
SEEM'
,?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
, 11-t:
'4
?
lii
SUMARY
PROBLEM: The principal purpose or this report is to present
to intelligence consumers some of the problems encountered and techniques
developed by this office for coping with these problems during the
course of making demographic estimates on Soviet and Satellite urban
areas on. which only fragmentary and obsolete census data were available.
The techniques described represent the lessons learned from several years
of research studies concerned with providing population estimates on
cities and larger urban areas as required. The products of this accomplished
research are contained in reports on 45 cities and urban areas (Appendix V),
17 of which were in the Soviet Union. They includc estimates on (a) total
population; (b) population distributed by age and sex; (c) population by
ethnic composition; and (d) labor force size and industrial composition.
METHOD: The method used is that of analytical evaluation of the
various demographic techniques found most effective by the research staff
in arriving at acceptable population and labor force estimates on each of
the 45 Soviet and Satellite cities. It was hoped that from the experience
gained in ranking these city estimates (1953-56) some standardized methods
would be found effective for use on foreign cities on which only very
inadequate and obsolete basic census data were available. It was found,
however, that because of differences among cities with respect to data
SENEr
? .
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
r
MET
iv
available and because of important variations in city structure and
function, complete standardization of method had not yet been realized
for any of the three types of demographic estimates made.
FINDINGS:
1. The methods developed for making estimates of total popula-
tions and breakdowns for Western cities were found to be or little use
for estimating the population of Soviet cities because of important
differences in the amount and kinds of data available. Official census
data with acceptable completeness were available for only two census years,
1926 and 1939. Even these figures are not entirely suitable as they may
have been based upon varying urban boundary delineations. Hence four
general and partially overlapping techniques could be applied: (1) the
use of ratio devices based primArily on election districts; (2) methods of
apportionment (or breakdowns) of a known aggregate Population for a group
of cities into that of individual cities for the same time period; (3)
extrapolation, or forward projection, of the trends of some period; and
(4) regression analysis, in which given certain data, the relationships
between certain variables can be statistically computed. Through combinations
of these techniques adapted to meet specific situations and conditions,
fairly reliable estimates and projections of total populations were made,
sometimes from the most fragmentary and incomplete basic data. These were
validated (confirmed) through convergence, closure, consistency and other
types of tests.
MR@MET
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
SEEM
2. The estimation of population distribution by age and sex
primarily involved the method of component projection. This method
consists of three parts: (1) estimation of the number of persons in the
populatian at some base date for which data are available who can be
expected to survive to the end of the projection period; adding to these
(2) the estimate of the net change of population due to migration and
(3) of the survivors of children born in the community between the base
and terminal dates. The method of component projection was found to be
workable, although somewhat laborious, except in cases where data were
lacking either with respect to the base population or one or more of the
components. An alternative approach called the functional estimation
technique was suggested. It inVolves investigation of the possibility of
ascertaining measurable relationships between patterns of economic activity
and the distribution of the population by age and sex. No satisfactory
method of measuring the ethnic composition of small areas in the U.S.S.R.,
short of an actual census, has as yet been devised.
3. Although estimates of the size and industrial composition of
the labor force for all cities were made, no satisfactory standard
'procedure -- that is, one which when mechanically applied, will produce
an estimate of total labor force distributed by branch of industry -- has
as yet been found. The core of a number of procedures described in the
report was, however, evolved. These procedures are generally applicable to
most problems of labor force estimation.
?ERR
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Ap roved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
-
goo,.
Preface
Summary
MET
C ONTENTS
Chapter
I. Estimates of Total Population for Cities of the U.S S R
Recent Soviet Estimates
Area Delineation
Evidence of area changes
Methods of Estimation
Ratio devices for estimating total population
Ratio devices based on electoral district statistics.
Types of electoral district data
Obsolescence of electoral district data
Legal ratio of population per election district
Acceptance standards for population estimates
based on electoral district data
Comprehensiveness of electoral district based
population estimates
Other ratio devices
Techniques of extrapolation
Apportionment techniques
Regression analysis
Validation of Estimates
Convergence tests
Closure tests
Consistency tests
Page
iii

1
1
5
6
11
13
13
20
24
26
27
33
37
39
41
45
50
51
51
52
II. Estimates of Age and Sex Distribution and of Ethnic Composition
for the Population of Cities of the U.S S R 61
Significance of Age and Sex Distributions and
Ethnic Composition 61
Estimation of Population Distribution by Age and Sex 66
Component projections 66
Survivors of the 1926 population 67
Evaluation of the 1926 census base 67
Replacement populations 68
Modification of survival rates 70
Survivors of births 73
Nature of the migration component......._ 76
Historical outline of the U.S.S.R. urban migration 79
Age and sex distribution of the migration component 84
Projection by stages...... 94
Evaluation of a component projection 94
An alternate procedure 99
Estimation of Ethnic Composition 101
The concept of ethnic composition............. ..... 102
Data and problems of definition. ocromoomovosi000etaoo 103
SEOGRET
Declassified in Part - Sanitized Copy Ap?roved for Release 2013/07/12 : CIA-RDP81-n1n4nPnn9nnnnarInno
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
8EMEIT
CONTENTS--Continued
Chapter
Page
III. Estimates of the Sise and Industrial Composition of the Labor
Force for Cities of the U.S.S.R. 112
Definition of terms 112
The scope of the problem. 116
Soviet labor force data 119
Procedural suggestions 123
Validation of city labor force estimates 128
APPENDICES
Appendix
T. Chronological List of Major SoUrces for U.S.S.R. City
Population Data X-,
II. Soviet Elections: 1937-1955 II-1
III. Illustrative Example of Certain Uses of Electoral Data
for Population Estimation III-1
IV. Illustrative Example of a Method of Estimating the Size
and Industrial Composition of the Labor Force of a
Soviet City--Tula: 1 January 1950 rv-1
V. Urban Area Reports of the Foreign Manpower Research Office,
U. S. Bureau of the Census V-1
FIGURES
Figure
1. Boundaries of the City of Tashkent; 1939 and 1950, and
Rayonal Boundaries, 1950 10
2. Hypothetical Relationship Between Percentage Change in
Population and Percentage Change in the Proportion of
the Labor Force in Large Scale Industry5 49
3. Relationship of Population Change to Change in Budget
Expenditures, Oblasts of the Kazakh S.S R 55
4. Collectivization and Out-Migration, for Selected Areas
in the U.S.S.R.: 1928-1931 83
TABLES
Table
1. Population of Selected Uzbek Cities: 1939 8
2. Electoral District Composition of Selected Soviet Cities:
All-Union Election of December 1937 17
3. Population Estimates for Moscow, Based on Electoral
Statistics: March 1950 to February 1951 20
27REF-7
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
,
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
SE@MYT
CONTENTS--Continued
Tables--Con. Page
4. Range of Population Estimates Based on Electoral District
Data for Selected Areas, by Type of Election 21
5. Legal Ratios of Population Per Electoral District, by City
Size Class: R.S.F.S.R. Local Elections of 1950 23
6. Percentage Deviation of Population Estimates Based on
Electoral District Data from Census Figures, by Tyne
of Election, for Selected Population Aggregates:
1937 to 1939 33
7. Discrepancies Between Population Estimates of the U.S.S.R.
Based on. Electoral District Data for All-Union and Republic
Elections: 1937/1938 to 1954/1955 35
8. Population Estimates for Chirchik, Uzbek S.S.R.? Based Upon
Selected Empirical Rates: 1935 to 1950 39
9. Estimated U.S.S.R. Male Military Losses in World War II 72
10. Mean Percentage Increase of Population, in Selected Types
of Soviet Cities: 1926-1939 81
U. Age Distribution of U.S.S.R. Internal Migrants, for Selected
Years: 1926 to 1934 90
12. Total Labor Force as a Percentage of Total Population for
Selected Cities and Dates 117
SEM
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
SERET
MSBMW:.
CHAPTER I
ESTIMATES OF TOTAL POPULATION FOR CITIES OF THE U.S.S.R.
The study of a city's current demographic and economic structure
requires an estimate of its current total population. Such an estimate has an
obvious descriptive value. It enables one to rank the city against other cities,
to compare it with cities in othar countries, and to form certain general notions
about the kind of place it must be. Equally if not more important are certain
analytic uses of a total population estimate:
1. It may be compared mith figures for earlier dates to indicate the
magnitude and rate of population growth.
2. If taken in conjunction mith vital statistics for the city, it permits
the estimation of net migratory changes in the population.
3. Such an estimate suggests limits for changes in -the composition of the
population by age and sex and by ethnic origin.
4. It permits the development of crude approximations of the dirnensions of
cbsnge in the labor force. An estimate of total population is often
useful for the analysis of other demographic and economic developments.
Recent Soviet Estimates
A recent important publication of the Central Statistical Administration
of the U.S entitled The Natiorn1 Economy of the U.S.S.R.: AStatistiss.2.,
Calalqa:;;ion and, n puhlica+.4(171 of thc. Statistical Administ:ation of the
R.S.F.S.R. entitled The National Economy of the R.S.F.S.R.: A Statistipal
pomgagoal have made available for the first time in many years a number of
important official populatirn statistics. Of particular interest at this point
8EOME
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R602600060008-1
N.
- 2 -
are tables of population estimates as of April 1956 for each city of 100,000
population or more in the U.S.S.R., for each city of 50:000 or more in the
R.S.F.S.R., and certain smaller places. During the years when the methods
described in this report were being developed such figures were, in general,
unavailable for dates after 1939.
Subject to a few reservations, the recently announced official Soviet
city estimRtes may be regarded as reasonably accurate, and as ordinarily to be
preferred to estimates computed by the methods described below. These reserva-
tions maybe snymArized as follows:
1. The scope of the official estimates is not certain. Although there
is a presumption that the figures relate to the entire population, it is
possible that certain classes, such as members of the armed forces or persons
in slave labor establishments, may not be included.
2. The geographic area to which the figures relate is not specified,
except in the case of Moscow, Leningrad, and Baku. It may be presumed,
therefore, that the estimates for other places of 100,000 or more inhabitants
relate to the city proper in each case. Nonetheless, since boundary
descriptions are not given, there is some uncertainty about the precise
area referred to. A discussion of the significance of boundary problems
is given below on pages 5 to 6.
3. It is not clear whether the estimates given are on a de facto or
de jure basis. In large cities there is often a considerable difference
between these two populations. The de facto population comprises all
persons actually present in the area at a specified time, regardless of their
usual residence. The de jure population comprises all persons who usually
reside in a place even though they may be temporarily absent at a specific date.
Declassified in Part-Sanitized CopyApprovedforRelease2013/07/12 ? CIA-RDP81-01043R00200006000A-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-3-
4, The estimates, although cited in an official source, are probably
of indifferent quality. Soviet authorities have indicated that they are
based on voter registration statistics, population registers, school
enrollment figures, and perhaps other series associated with specific
segments of the population. Under such circrimstances variations in the
Quality of the estimates are to be expected, but ordinarily there will be
no test readily available to reveal inadequacies in specific figures.
One possible approach to the evaluation of the adequacy of the official
estimates consists of comparing them with the population figures implicit in the
number of election districts set up for the 1955 Republic elections. These
districts were set up in December 1954 and the corresponding derived population
figures presmably refer to that date or to an earlier date. It is possible to
make this comparison for 128 of the 135 cities of 100,000 or more inhabitants
according to the official April 1956 estimates. In 1 case the 2 figures proved
to be identical, in 53 the 1956 estimate exceeded the 1954 figure implicit in the
voting district data, and in the remaining 74 cases the 1954 figure exceeded the
1956 official estimate. For the most part, the differences were well within the
error of estimate of the 1954 population. Moreover, even in the case of those
few cities, such as Voronezh, Tashkent, and Molotov, for which the 1956 estimste
exceeded the 1954 estimate by a figure amounting to a relatively large proportion
of the possible error inherent in the 1954 estimate, no inference as to the
adequacy of the 1956 official estimate may be drawn, because of the possibility
of growth during the intervening period. Similarly, the comparison is not
particularly useful in the case of those cities for which the 1954 estimstes
exceed those for 1956. In almost all of the 74 cases the difference found is
z
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 4 -
attributable to boundary differences or else amounts to only a minor fraction
of the possible error inherent in the 1954 estimates. The only exceptions are
Moscow, where the difference of 411,000 amounts to more than twice the maximum
error inherent in the 1954 estimate; Kirovabad, where the difference is almost
3 times the maximum error; Tallinn, where the difference is more than twice the
maximum error; and Kaliningrad, Saratov, Omsk, Chita, Vilnyus, and Kaunas, where
the difference In each case exceeded half of the maximum error. Unless there is
reason to postulate population losses in these 9 cities between late 1954 and
early 1956, the comparison indicates that the official figures should be used
with caution as they are probably understated. In the case of Vilnyus, Kaunas,
Tallinn, and Kirovabad, however, the understatement, although relatively large
in terms of the inherent error of the 1954 estimstes is small in absolute terms,
and maybe disregarded.
Another approach to the evaluation of the 1956 estimates involves the
use of official data on square meters of dwelling space in some of the larger
cities. These data were published for 32 cities in the same source as the
official population estimates. If the housing space data are expressed in
per capita terns on the basis of the 1956 official population estimates, it is
found that the per capita housing space for most of these cities varies very
little. The mean of the per capita housing space figures for the 32 cities
is 7.19 square meters, and the standard deviation is 1.55 square meters. One
per capita figure, namely 12.04 square meters for Riga, falls more than three
standard deviations from the mean. The very high value of the figure for Riga
brings into question both the population estimate and the housing space information
for that city. The housing data are not Particularly useful in the evaluation
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy A proved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 5 -
of the population estimates for the other cities, except in the negative sense
that they do not show anything unusual.
No doubt, with some ingenuity, there could be devised we.ys of evaluating
the adequacy of the official population estimates in of additional Tlieeea
of information; and this type of evaluation should be attempted wherever possible.
It does not appear, however, that a general quantitative expression of the acouracy
of all the city estimates given for 1956 can be had, and the user will usually
have to be content with the data as they are despite the areas of uncertainty
which have been touched on above. Xn any event, the official estimates are
likely to be superior in quality to the which can be derived by the methods
described below.
Arca Delineation
Before a city's population may ee estimated meaningfully it is necessary
that there be a clear underateacting of the territorial boundaries to which the
estimate must apply. It is not enough to manipulate statistical data which
appear to apply to a place having a speallIed name. To a large extent the
delineation of the city or urban area for which estimates are =Wit depends on
the purposes of the study. The object of the investigation may be the legal
city as it was constituted on a given date; or a subdivision of the legal city;
or the city plus its populated suburbs; or a somewhat larger territory, often
called the metropolitan area, in which the legal city and fairly extensive
outlying areas are joined by social and economic ties into a single community;
or even a target complex, comprising pertions of the legal city plus selected
contiguous and noncontiguous territory of strategic importance. Each of these
types of area mould, in general, have a different population. It is also
Declassified in Part - Sanitized Copy A .proved for Release 2013/07/12 ? CIA-RDP81-01043R007nnnnsnnnR
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 6 -
conceivable that an urban area Trith a large commuting labor force might have
a unique spatial distribution of daytime population and that the centers of
concentration of the daytime population would become the object of the research.
In such a case the area to be studied might properly be defined in terms of the
concentrations of daytime population, with the purely residential quarters
regarded as outside the scope of the study.
Although research requirements inevitably govern the delineation of
the area for which the population estimate is desired, the nature of the
statistical data available to make the estimate may require important
modifications. It is not always possible to attain an ideal if the data are
not quite in point. For example, when there have been changes in area over
time the statistics available for a city generally do not relate to a constant
area, and it is not always possible to extricate the changes in population size
resulting from boundary changes from those attributable to natural increase or
net migration.
At the very least it is necessary to determine the geographic area
associated with each Soviet population statistic to be employed in the research,
so that this may be compared with the area for which a current estimate is
required, and, if possible, adjusted to conform to the latter. A proper areawise
interpretation of Soviet data should be attained before further work on the
population estimate may be pursued profitably.
Evidence of area changes.
Direct evidence of area changes by comparison of numerical measures
of area at two dates can rarely be had for recent periods. As a rule the
statistics available on the area of Soviet cities are 20 to 25 years out of date0
In some instances area measurements for the late twenties and early thirties
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
7
are found in handbooks of that period, but this is infrequent. Large-scale maps
and town plans, usually for prewar dates, may be measured to determine the area
within administrative boundaries. Sometimes reports showing population density
are found. These, together with the corresponding population estimates, permit
the computation of the area as of the date to which the figures apply. Apart
from such help as maybe found in these sources, evidence of territorial
redelineation over time comes mainly from Soviet announcements of administrative
changes and from the comparison of statistics on population with the same date
of reference but different dates of publication.
The chief sources of published information on changes in the extent
of administrative-territorial units in the U.S.S.R. are: (1) a series of
administrative-territorial handbooks12 and (2) the Journal of the Supreme Soviet
of the U.S.S.R.3 These sources provide a qualitative account of territorial
changes. For example, with reference to the city of Tula, these sources indicate
that the metallurgical center of Kosaya Gora was administratively subordinate to
the city in 1938 but detached by 1950, and that the settlement of Novo Tuliskiy
was absorbed by Tula early in 1940 and became a district of the city by 1951.
Such information provides an interpretive framework for the analysis of population
statistics applying to intervening dates. This type of material, however, does
not deal with the number of people involved in each territorial change.
A quantitative evaluation of the effect of territorial changes on.
?
population can sometimes be obtained by comparison of figures for the same place
published before a territorial change and republished with revisions after the
change. For example, boundary changes between 1926 and 1939 for cities with
more than 50,000 inhabitants in 1939 will be indicated by a comparison of the
1926 city population figures released in 1939 with those given in the 1926 census.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 8 -
The ann.e.xation of Kosaya Gera by the city of Tula, to uhich reference has already
been rade, could have been discovered from such a compariscnt
POPULATION OF TULA 1926
As reported in 1939 ***** 155,005
As reported in 1926
Difference 2028
The difference between these figures is identical to the popnlAtion reported for
Kosaya Gera in the 1926 Census0 la this case a precise interpretation of the
difference UMB possible because of the detailed information provided for 1926.
When the information for the year of the comparison is more limiteda it may not
be possible to identify the parcels of territory involved. Nevertheless an
indication of the number of persons affected may still be gained. Thus,- we have
for certain cities of the Uzbek Republic:
Table 1.--POPULATION OF SELECTED UZBEK CITIES: 1939
City
Source A
Source B
Tashkent. ..........
SDS 00
585,005
134,346
567,000
138,100
84,665
86,300
Andizhan
83,691
86,300
Namangan.0........00000000400
77,351
79,800
dia.00.14?ROMMOCCE.M..
Source Ag Isvestiya, 2 Jane 1939.
Be Akademiya nal.* Uzbekskoy SSR. Institut
ekaaomiki. pApkistana Tashkent, 1950,
v. 72.
Ymm these data the inference may be drain that betmen January 1639, the date
of the census, and the date when the figures in column B were compiled by the
Institute of Economics of the Uzbek Academy of Science, a contraction of the
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
boundaries of Tashkent end an expansion of the boundaries of the four other cities
took place. Another possible inference is that the figures in column ft are
provisional figures released shortly after the enumeration, whereas those in _
column B are final tabulations. There is, however, other evidence for Tashkent
to support the hypothesis of a change in area.
The case of Tashkent illustrates still another means or detecting
boundary changes, i.e., by the use of election data. At a later point, the use
of election materials in making estimates of total population will be discussed
in detail, but their use in connection .with problems of area delineation should
also be noted!' This use of election data requires a list of election districts
with a description of the limits of each district and a large-scale nap Phasing
street names. With the use of such materials, for example, both the 1939 and
1950 city boundaries of Tashkent can be plotted.
The 1939 borandaries thus derived proved to be the same as those shown
in an official nap for 1938.3 The 1950 boundaries have not been verified
independently? but an internal check is possible, as follows:
Tashkent is known to have 7 city rayons. The same electoral
figures used to delineate the 1950 city boundaries may also be used
to delineate the 7 constituent city rayons. Figure 1 shows both the
1939 and 1950 boundaries of Tashkent as delineated in terms of the
electoral information. The difference in area is strv-in=. The
Tashkent estimate for 1939 cited in source B in table 1 could therefore
refer to 1939 population within 1950 boundaries.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
SECRET
- 10 -
Figure I.--BOUNDARIES OF THE CITY OF TASHKENT, 1939 AND 1950, AND
RAYONAL BOUNDARIES, 1950
CITY OF TASHKENT
LEGEND
- - - CITY BOUNDARIES IN 1939
CITY BOUNDARIES IN-I950
- - -- --CITY RAYON BOUNDARIES, 1950
1
SCALE
9 I
3
KILOMETERS
CITY RAYONS IN 1950
I. KIROV
2. OKTYABRI
3. STALIN
4. FRUNZE
5. LENIN
6. TSENTRAC- NYY
7. KUYBYSHEV
U.S. DEPARTMENT OF COMMERCE_ U.S. BUREAU OF THE CENSUS
? SOURCE: SEE TEXT
SECRET
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
????
11
Finally a word of caution is needed with regard to the interpretation
of disparate population statistics such as have been presented above. All
discrepancies between population figures for the same date are not to be taken
as indicative of changes in area. For example, there are certain intereeneal
figures reported for Tula which differ because one set of figures relates to the
total population and the other to the civilian. population..
The pcosibility that ostensibly comparable figures contain concealed
elements of incomparability unfortunately looms large. Soviet society offers
numerous illustrations of special population groups such as the armed forces,
security troops, slave laborers, members of ethpic minorities, and the like, which
may sometimes be omitted from statistical tabulations vithout so stating. Moreover,
especially for the years around 1926, there are numerous illustrations of figures
for the same place and date which differ because one was compiled on a de jure
basis and the other on a de facto basis.
Methods of estimation
Because of important differences in tue amount and kinds of data available,
the methods developed for making estimates of total population for Western cities
are usually of little use for estimating the population of Soviet cities. But
regardless of the specific character of the data at hand, there are four general
and partially overlapping approaches that may be followed. These may be classified
as (1) ratio devices, (2) methods of apportionment, (3) extrapolation, and (4.)
regression.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 12 -
To understend the reasons for adopting particular procedures under
any of the above headings, one should be familiar with the nature of available
population statistics arid with the strategy of solution of the problem dictated
by them. (lin outline; chronologically arranged, of the most accessible items is
given in eppendix I.) In general; in dealing with larger places one will have
total population figures for the two census years, 1926 and 19392 es well as
severe observed or officially estimated figures for the intercensal and postcensal
periods. Between 1939 and 1956 few figures on total population for particular
cities were released. This data deficiency is most serious because 1939 population
figures for urban areas are likely to be poor indicators of current population
size. The period 1939-1955 was one of relatively large growth of urban population.
During this period the urban, population increased by 30.9 million or about 55
percent. Part of this growth represents the annexation of urban territory and
the formation of some 2,000 new urban places, but much of it reflects growth in
cities which were in being in 1939. Thus, the absence of data for years after
/939 increases the possibility of errors of estimation in cases where a 1939 figure
is available, and even more so in cases 'where this information is lackink.
This, then, is the background against which estimates of city population
are to be made and evaluated. It may be taken as axiomatic that Soviet
demographers have more data at their disposal and are thus able to produce superior
estimates. Nevertheless; one may surmise that; due perhaps to the deficiencies
of their registration system and to the lack of a postwar census, Soviet
statisticians themselves are at times forced to rely upon materials similar to
those available to Western demographers who prepare estimates for cities in their
own coustries.6
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
13
Ratio devices for estimating total popaulatiaa.
Udder this heading are included .estimation procedures which convert
a brain or given value of one variable into an estimate of the value of a second
variable on the assumption that the ratio between the two variables is the same
as that during an earlier period, as that determined for some other place, or
as that given in a Soviet source for the same time =Apiece. Thus, if Ai
designates the value of one variable at date I, and Ni the value of a second
variable at the same date, yams of N for dates other than I may be estimated
e.S2
N4
The expression may be imolai for an earlier date, for another place, or for the
same date and place, even though Ni and As. are not known separate/y. Sometimes
N
A
the ratio () ? is given a value at date j by interpolation or extrapolation between
values of the ratio for two or more other dates.
Ni
N = A
iii
Ratio devices based on electoral district statistics.
The most commonly used ratios for making city population estimates are
those involving election statistics, especially the ratios of elected deputies'7 to
the number of their constituents. Data on the number of eligible voters are also
widely avei3able and may provide the basis for an estimate of population,
provided something is known or can, be estimated about the ratio of voters to
total population. This is the chief problem that one encounters in using data
from voter registration lists. It is a problem that, as we shall see, is handled
explicitly in the case of election district data. On the other hand, there is
greater assurance that eligible voter lists are current than is true of lists
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- ?
of election districts. The use of voter statistics is discussed further under
the heading of apportionment techniques.
According to Soviet statements/3 the number of election districts
established (deputies elected) for any election is determined by dividing the
population of the area concerned by the number of constituents per district as
N;
set by Soviet law0 ri general terms, Soviet law provides the value of , hero
AI
Ni designates population and At designates electoral districts. If the number of
districts in a specified place, AT is known, then the population of that place
may be estimated at:
Information on election districts generally appears in the press about
two months or so in advance of an election; for example, the lead time for the
announcement concerning the R.S.F.S.R. local elections of 1947 was 55 days.9
Sometimes the number of districts is announced directly, but the form of the
memouneement may vary from merely a detailed list of district names and territorial
descriptions to a list of names of candidates for office.10
There are a number of points about thA 4Ormation of electoral districts
which are not entirely clear, and these obscurities may Pefog the interpretation
of population estimates based upon electoral district data. The most important
problem is whether the electoral districts were defined In terms of de jure or
de facto population. 2he great flexibility that exists with respect to legal
residence reauirements for Soviet votersu permits the use of a de facto population
base for electoral district delineation, and it is probable that this base is used
more often than the de jure base. But one nevus knows which base was used in
particulnr instances. Another important problem is the determination of the date
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 15
of reference of the populaticn figures used by the Soviets in delineating
electoral districts. The practice at the time of the DeeMber, 1939 local
elections mas, generally, to form districts on the basis of the January 1939
census figures. Evidence of this is the fact that the aggregate endeof-the-year
population estimate derived from election data for 29 of the largest Soviet
cities differed from the first-of-the-year census count by less than two-tenths
of 1 percent. This situation is found despite the fact that the instructions
to the electoral commissions are to "establish the total number of election
districts on the basis of the population according to the Census of 1939 with
an account of changes occurring since the day of the census.H12
-There are, moreover, marked disparities between q1930 Census figures
and 1930 electoral imputations in certain areas in some cases theeemay be
interpreted in terms of territorial changes. In other cases such an interpretation
is not supportable and other hypotheses have to be considered. One explanation
for which there is some support -is that electoral commissions have taken account
of large-scale population movements such as the transfer of troops, large labor s
contingents, and so on, as they were directed to do. Adequate data to test this
notion are not available, but the following observations are offered in support
of this general thesis.
1. The greatest disparities are found in regions where such population
transfers are believed to have occurred; eg00 in the border areas of
the Northwest and Far East. Yet, in some instances where sizeable
disparities have been encountered there has been, no evidence of unusual
pepulation transfers. This, it seems, could be as much a natter of the
incompleteness of information regarding population movements as it is
a denial of the hypothesise
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
-16-
2 in a few instances a comparison of order of battle information believed
to reflect troop movements during the wear 1939 seams to provide a fairly
exact reconciliation of the differences between census figured and
electoral imputations. Thus, for example, in Arkhgel 'skews oblast the
available order of battle indicates the presence in 1940 of two divisions
that were not there at the time of the census. The difference between
census figures and electoral imputations in this case is around 30,000,
which is a reasonable figure for the troop concentration described.
Conversely, in Novosibirsk and Kuybyshev, differences of a similar
magnitude are associated in each case with the removal to border areas
of one division and several miscellaneous military formations. It should
be repeated, however, that there are numerous other cases where the
order of battle sheds little light upon differences of this type. This
should not be surprising since in most cases the imputations for December
1939 are greater than the census figure for qTanuary 1939, while at the
same time, the order of battle is more complete for the period before the
census than for the period following.
Districts established for nonlocal elections, i.e., for deputies to
national or republican bodies, frequently cut across the boundaries of lesser
civil divisions, thereby impairing to a degree their usefullness in connection
with estimates for small areas. This is an inevitable result of the Soviet aim
of equal representation for all electors; although an apparent compromise with
administrative convenience is made in the provision: that ablest, ]al, and republic
boundaries be .respected in the formation of election districts. The inaccuracies thus
produced are claimed to range between 3 and 5 percent14 of the prescribed represen.
tational norm in the case of AU-Union elections her maximum deviations are to.
be expected.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-
Another source of estimation error is, of course, that due to rounding.
For example, a city lath a single election district for the election of one deputy
to the Supreme Soviet of the U.S.S.R. (estimated population: 300,000) theoretically
could have a population anywhere between 150,000 and 450,000. However, as the
fo11or1ig tabulation (table 2) indicates, gross errors of this magnitude do not
appear to occur in actual practice. Taking all the cities which in 1939 feilvd-M2in
this theoretical range, it is clear that an attempt as rade to vary the structure
of election districts in accordance wah population differences. Cities having a
pcoulation close to 300,000 comprise one election district or its equivalent; larger
cities ecaprise more than one district; whereas smaller cities form only part of
one district.
Table 2.--ELECTORAL DISTRICT CaDOSITION OF SELECTED SOVIET CITIES:
ALL-VgIOM ELECTION OF DECEMBER 1937
1161....10.11armor..? .10.11?PCWSIS x..13.44o.
City
Population,
January 19391
Electoral district
composition2
Notes2
445,476
h. entire district and
part of another
000
Sverdlovsk.. .I.
495,544
000
Novosibirsk...
405,589
000
Kasen...
401,666
VI
CO 0
Voronesh......
326,836
1 district
000
YaroslavIg....
298,065
1 district
000
Zaparozh'ye...
289,188
Parts of 2 districts
000
Ivancvo.......
285,069
Part of 1 district
which also includes 2. rural rayon
Arkhangelisk. 0
2S1,091
Part of 1 district
vbich also includes 1 rural 7...ayon
0Mak0000000000
280,716
Part of 1 district
which also includes suburbs
Nisb-di Tagil..
159,864
Part of 1 district
which also includes suburbs and
Penza.........
157,145
Part of 1 district
1 rural rayon
which alsri,noludes 3 rural rayons
Smolensk....
156,677
Part of 1 district
which also includes 2 rural rayons
Shoalty. 0 0 00
155,081
Part of 1 district
which also include a smaller city
and 1 rural rayon
1 Tzvestiya? 2 June /939.
2 Ibid., 12 October 1937.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 18 -
Despite the care that may have been taken in delineating election districts
in conformity uith the population requirement, AU-Union electoral districts are
probably or limited usefulness for estimating city populations. This results from
the fact that there is =requirement that such electoral districts conform to city
boundaries. Thus, in only 2 of the 34 cities listed in table 2 does the electoral
district coincide with the city, In the other 12 cases a knowledge of the boundaries
of the electoral districts is .absolutely necessary, if more than a very rough guess
of the eityls also is desired. To illustrate, we know that Stalingrad contained
one electoral district plus a portion of a second. Henze its population could have
ranged from about 300,000 to 600,000, depending on the proportion of the second
electoral district in the city, and on the deviation of the population actually
in each electoral district from the norm of 300,000 prescribed by law. The same
observation may be made about Kazan, Novosibirsk, and Sverdlovsk. The information
provided by the electoral district data merely permits the conclusion that each
of the cities ktan a population of from 300,000 to 600,000,nere or less. Similarly,
each of the cities: Tvamovo? Arkhangel'sks, Cmsk, Nisbni Tagil, Pena, Smolensk,
and Shakhty? are included with a single electoral district, together with some
adjacent territory. 1111 that eaa be said from the data on electoral districts
is that each of these cities has a population which was probably less than 3002000.
The census takea about a year later places these in the range from 155,000 to
285,000. A rough indication of the extent to which the, population of any of these
cities differed from 300,000 may be had by Doting the number of extra pieces of
territory that have been included with the city in a single district (table 2,
column 4). There will rarely be enough additional information available to yield
estimates of the population of these extraneous satellite territories. Even
7c32iere such information may be found, there still r
Clue:
the problem of estimation
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 19 -
error; which for a single district may range up to ? 1-2-1 R being the legally
prescribed ratio of population per election district.
Obviously, fine discriminations and close estimates are not warranted
by the All-Union electoral data. However, such data are much more useful for
estimating oblast populations, because each oblast has an integral number of
electoral districts, all but one of which are defined so as to attain as closely
?
as possible the legal requirement of 300,000 inhabitants per district. It is
probable that one election district in each oblast has a population which is
appreciably different, unless it happens that the total population of the oblast
is close to being an integral multiple of 300,000.
There is no evidence that the system of delineation of All-Union electoral
districts has changed since 1937; it may be presumed that more recent electoral
district data suffer frcm the same limitations.
Electoral district data from local and Republic elections, when available,
should prove much better suited for deriving city population estimates; in these
instances it is believed that city boundaries are usually observed and not
overridden in the formation of electoral district boundaries.
This, then, is the general character of election statistics in terms
of their value as population indicators. When judiciously used for this purpose,
they are among the most valuable items released by the Soviets. Their value
depends upon:
1. The type of election (All-Union, Republic, or local),
20 the areal coverage of the data,
3. the degree of obsolescence.
These matters are considered belawa
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 20 -
Types of electoral district data.
Variability in the usefulness of-different types of electoral data
for population estimates may be illustrated by reference to four different types
of election held in Moscow between March 1950 .and February 1951.
Tpble 3.-40PULATI0N ESTIMATES FOR MOSCOW, BASED. ON ELEOTCRAL STATISTICS:
MARCH 1950 TO FEBRUARY 1951
0.1.??????,11110?????????????????????????????1.01.???????????=.1M
-
Type of election
Date of
election
Legal ratio of
population per
delegate
Range of populftion
estimate
To USSR Supreme Soviet
To RSFSR Supreme Soviet
To Moscow Oblast Soviet
To Moscow City Soviet
March 1950'
February 19512
December 19503
December 19504
300,000 : 1
150,000 : 1
30,000 : 1
30,000 : 1
4,950,000 - 52250,000
4,725,000 - 4,875,000
4,215,000 - 4,245,000
41275,000 - 4,305,000
izvestiya, 12 Tannery 1950.
2
W1E713 December 1950.
3 Moskovskaya Pravda, 18 October 1950.
4 Izvestiya, 19 December 1950.
In this illustration there is a range of more than 1 million persons between
estimates based on different types of electoral district data. In part, this
difference represents the effect of differences between the elections in the
legal ratio of population per delegate, and in part it reflects variation in
the area to which the electoral district data refer. There are, obviously,
practical limits on the size of deliberative bodies. It follows that the larger
the jurisdictional area, the greater the number of constituents per delegate.
Thus, each delegate to the Supreme Soviet of the U.S.S.R:. represents 300,000
Inhabitants, whereas each delegate to the Moscow City Soviet represents only
30,000. Hence, the ranges of error In the corresponding population estimates are
150,000 and i 15,000,.respectively. Moreover, although an attempt is made to
have the delegates represent recognized administrative units, equalization of
representation throughout the Union, the Republic, or the oblast often necessitates
?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 21 -
the crossing of boundaries of small civil divisions (e.g., cities, rayons)16 in
the delineation of electoral districts. In the case of the All-Union and Republic
elections, the districts created for the city of Moscow include the city of Tusbinoll
the villages of Stalinskiy? Lublinskiy, Ryblevos and Vidnoye,. as well as several
territorial sectors Voich penetrate deeply into the hinterland. The superiority
of local election statistics18 for population estimates does not, therefore, rest
entirely upon the greater precision resulting from the smaller legal ratio of
population per delegate; it also stems from the fact that local election data
refer to the resident population of a recognized administrative entity (e.g.) the
city soviet) and not to the administrative hybrids which may be embraced by the
election districts established for Republic and Union elections.
In addition to the factors discussed above, namely variation in the
legal ratio of inhabitants to delegates and variation in the area represented by
a delegate, there are other factors, more difficult to identify, which also cause
variation among population estimates for the same area based on electoral data
for different elections. Consider the following population estimates for Moscow
Oblast and for a Moscow Subarea, as derived from electoral statistics for different
elections:
Table 4.--RANGE OF POPULATIal ESTIMATES BASED OM ELECTORAL DISTRICT
DATA FOR SELECTED AREAS, BY TYPE OF ELECTION
d
of
Range of estimates
1?11?????10?IPIIIIMI.M40.11.10111111a.
Type
election
Moscow Oblast
Moscow City Subarea
A
To U.S.S.R. Supreme Soviet.... /10,650,000 - 10,950,000 12650,000 - 1,950,000
To :R.S.F.S.R. Supreme Soviet.. 9,975,000 - 1021252000 1,980,000 - 2,280,000
To Moscow Oblast Soviet.. ..... 9,1352000 - 9,165,000 1,695,000 - 1,725,000
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?22-.
In esch case, electoral district delineations respected the boundaries of the
civil division to which the estimates apply. A few observations may be made
about these data. First, for both Moscow Oblast and the Moscow City Subarea,
the central value of the population estimate based on the oblast data is smaller
'Ulan that based on either the Republic data or the All-Union data. Second, the
central value of the estimate based on the Republic data is smaller than that
based on the All-Union data in the case of Moscow Oblast, but larger in the case
of the Moscow City Subarea. No pattern of regularity may be discerned from these
data, but it is clear that there are variations, possibly random, for which no
suitable explwsetion is at band. The estimates based on the oblast data suggest
that when the oblast electoral districts were delineated some classes of the
population (possibly armed forces or slave labor) were not taken'into account,
but no proof of this has been found. It may be said, therefore, that the kind
of electoral data chosen (All-Union, Republic, or local) can have a serious
influence on the population estimate, and that in general, the local election
data have the advantage of a smaller range of error and the possible disadvantage
of a putative omission of some classes of the population.
These considerations are frequently academic, for local election
inforMation is much more difficult to find than other types. In addition, the
imperfections of local election data must be kept in mind. One of the more serious
of these is the indeterminancy which results from the practice of setting maxima
on the number of districts for each city size group. This practice results in
certain blind spots or ranges that are indeterminate with respect to population
size. These are illustrated by the system adopted for theR.S.F.S.R. local
electims of 1950.19
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Cop Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
-23-
Table 5.--LEGAL RATIOS OF POPULATION PER ELECTORAL DISTRICT,
BY CITY SIZE CLASS: R.S.F.S.R. LOCAL ELECTIONS OF 1950
-
City size
...-
Population
per
district
Maximum
number of
districts
permitted
.
Range of
nondiscrimination:
Under 12,000
(2)
35
0- 12,000
12,000-991999
350
250
1
870500-1005000
400
300
1200000-150,000
1501000-249,999....... .....
500
400
200,000-250,000
600
500
300,000-350,000
J 700
600
420,000-500,000
Over 500,006
900
700
Over 630,000
I The population of cities in these size ranges cannot be determined
more precisely from the number of electoral districts. See text for fur-
ther explanation.
2 Number of districts fixed at 35. Therefore population per district
Population of City
35
Thus, when the number of electoral districts or delegates for a city is 35
or below, or precisely 250, or 300, or any other number shown in the second
column above, the central value of the city population may fall anywhere
'within the range of nondiscrimination shown in the third column. But if
the number of districts is 365 or 2515 or 301, etc" the central value of
the population estimate is determinAte.
Except where such blind spots are encountered local election
statistics are ordirArily superior to other types of election data for the
derivation of city population estimates. Their scarcity, however, demands
constant attention to the development of substitute techniques. One such
procedure aimed at utilizing All-Union and Republic information for making
small area estimates -- this being a prime advantage of local election data
is presented in appendix III.
Declassified in Part - Sanitized Cop Approved for Release 2013/07/12 ? CIA-RDP81-01043R0n7nnnnRnnng
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-24-
Obsolescence of electoral district data.
One objection sometimes made to the use of election data for population
estimation is that the estimates reflect not the true population of the area,
but rather the population figure that happened to be used when the electoral
districts were defined. There is reason to suspect that the Soviets themselves
may not always have adequate, complete, and up-to-date population figures to use
in delineating electoral districts. Moreover, an examination of electoral
district statistics reveals an alarming degree of stability over time, supporting -
the hypothesis that population changes are not always reflected in the number of
electoral districts assigned to an area. It has also been maintained that new
electoral districts are formed only -where there have been relatively large
population increases in an area. These objections, if valid, indicate serious
limitations on the usefolriess of electoral district data for population estimation.
There are numerous references in the literature to the inadequacies of
the Soviet statistical system and population data.20 One Soviet source speaks
of local Population estimates being made on the basis of sell soviet registration
books and household books.21 Illustrations of methods of estimating population
for specific purposes are found in textbooks on city planning and educational
administration. The fact that officials must resort to a variety of estimation
techniques to obtain figures which are assertedly not to be had on a reliable
basis from censuses or other sources does not necessarily indicate that local
Population data are poor. A good current estimate is preferable to an obsolete
census figure. Thus, the alleged inadequacy of local population figures does
not necessarily mean that an adequate estimate is not available. It is not
clear, however, that the delineators of election districts always use good
current population estimates as a guide.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
7--
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
dfa $
"".m.NP ?
-25-
The evidence that adequate population estimates, even if available,
are not always used in the delineation of election districts stems mainly from
the stability over time of the number of election districts associated Nith an
area. For example, UD to 1955 there were 7 Republics which had a constant number
of electoral districts in each All-Union election. Similarly, in the R.S.F.S.R.,
42 of 69 oblasts each had a constant number of electoral districts in All-Union
elections held between 1938 and 1951. It may very uell be true that; at least
in connection with. All-Union elections, obsolete population data are often used
in delineating election districts.22
The case is less clear in Republic elections. Between 1938 and 1951,
15 out of 16 Republics did have changes in the number of electoral districts used
in Republic elections. And, in the R.S.F.S.R. Republic elections, there were 42
oblasts Which had changes in the number of electoral districts between 1938 and
1951, as compared with only 29 R.S.F.S.R. oblasts having such changes in the
AU-Union elections. In the case of local elections, there is even less evidence
of stability in the number of electoral districts, over time. Thus, there may
be some merit to the contention that obsolete data have been used in delineating
electoral districts; but the problem seems to be more serious in the case of the
All-Union elections than in the Republic elections or in the local elections.
It must be granted; however, that the suspected deficiencies in electoral
data by virtue of Obsolescence have not been demonstrated unequivocally. It can
be shown that there are a number of areas in which the number of election districts
bas changed over time, snet these include both areas having changes in geographic
boundaries and areas not having such dllorges. The only way for an area which
has not had a boundary change to have a change in the number of electoral
districts is by the use of changing population figures in electoral district
delineation.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- - . - an. -"""40111.1111111!
- 26 -
The remaining major criticism of electoral district data as the basis
for population estimation is that recognition is given to population changes in
delineating election districts only when there has been substantial growth, as
by migration. As a corollary, it is sometimes suggested that important population
losses are not taken into account in this process. Criticisms of this character
are consistent with the evidence of stability in the number of election districts
over time, particularly for the All-Union elections. No ready rebuttal of such
criticism can be made in terms of the information at hand. Further investigation
clearly seems indicated.
The inadequacies of election district data for population estimates
are painfully obvious. Nonetheless, such data are often the most useful kind
of information available upon which to base population estimates. The possibility
that election district data for particular places may be obsolete should be kept
in mind by both the makers and users of population figures based thereon, lest
there be an unwarranted degree of confidence in the estimates.
Legal ratio of population per election district.
Once a set of election district
adopted; it is necessary to determine the
per district or delegate; so that the two
Population estimate. It has already been
figures for a particular area has been
appropriate legal ratio of population
figures may be multiplied to yield a
noted that these legal ratios vary by
type of election. They may also vary over time, and for local elections, may
vary with the size of the population of the area.23 Values of the legal ratio
by type of election, year of election, and size of area are given in appendix II.
(See page II-I.) The only caution that need be observed is that the legal ratio
of population per district or delegate conform to the data on number of districts
which are being used.
Declassified in Part- Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R00200nonnnnR_i
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-27-
Acceptance standards for population estimates based on electoral district datti.
Because of the possibilities of error arising from the nature of electoral
district statistics, population estimates based on them should not be accepted
routinely or mechanically. Each such estimate should, if possible, be tested
to determine whether the population change it implies is consistent with
evaluations based on other types of data or with expectations based on general
knowledge -- perhaps nonquantitative -- about the city's development. Two
illustrations of this point are given here, in connection with estimates of
population for Samarkand and Kokand, prepared on the basis of electoral district
statistics.
Samarkand.--Development between 1939 and 1950 -- This is judged to be
a relatively stagnant city with moderate growth anticipated as the
result of a small area which contained about 4,000 inhabitants in 1939.
1939 population within 1950 boundaries 00 00 138,100
Estimated 1950 population.. 0000000000V000C,C.00 241502000
Percentage change, 1939-1950. ................. +90
The change in population indicated by these figures appears to be
moderate and in the right direction.25 It may be shown on the basis
of statistics on school enrollment and registered voters that the
population probably exceeded 145,000,26 thus lending some support to
the estimate cited here. Moreover; there is a suspicion that electoral
district data for native, as contrasted with Russian; communities
understate the true population.27 In the light of these relatively
vague impressions the estimate of 150,000 appears to be worthy of
acceptance. Some years after the estimate was made, the Soviet
government published an estimate of 170,000 for Samarkand as of
April 1956.28 This may be compared with an estimate of 160,000 for
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
41.
gm.
- 28 -
1 January 1955 derived by extrapolation from the 1 January 1950 estimate
of 150,000 which was based on the electoral data. Had the extrapolation
been carried forward to 1 April 1956, the result mild have been 165,000.
These figures indicate that the 1950 estimate of 1502000, that is the
estimate based an electoral district statistics, may have been a trifle
low, but not excessively so considering the possible errors inherent
in an estimate of this type. Hindsight, therefore, justifies the earlier
acceptance of the 1950 estimate of 150,000.
Kokand.--Development between 1939 and 1950--This is an ligbek city which
bad some industrial growth as well as an expansion of area after 1939.
Because of limited water supply, however: the growth of industry was
not very large. It appears to be a stable city in terms of population
size, baying changed very little between 1879 and 1939. It might be
expected that the population mould remain about the same in 1950 and
1939, despite some small growth of industry, because the city had a
relative abundance of underemployed labor and because of the mater
supply problem. A moderate increase in population would also be an
acceptable possibility, but substantial gains or losses appear to be
ruled out.
1939 population within 1950 boundaries....... 86,300
Estimated 1950 population. 00060000o00000004o0 2975,000
Percentage change, 1939-1950. ...............
- 13.0
Here, the direction and magnitude of the population Change implied by
the estimate for 1950, which was based on electoral district statistics,
are inconsistent with the expectations and should not be accepted.3?
Some indication of the correct nopulation may be had in other mays.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R00200noAnnnR_1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
..". '6.4".11171:1111?1111111.....
? 29 ?
For one thing; the city of Andizhan elected 248 delegates to its
city soviet in December 1950,31 indicating that its population was
?
about 87:000. Andizhan has been described as the largest city in
the Fergana valley." Hence; Kokand ,s population should be smaller
than 87,000. For another, the city of Kokand had 5 of the 48 election
districts in the Fergana Oblast for the elections to the Supreme Soviet
of the Uzbek S.M., in January 1951. It should, therefore; have
of the oblastts population, more or less. On the basis of oblast data
for the local elections of September or October 1950, Fergana Oblast
5 ,
had a population of about 812,000. Kokand, therefore, had wrhs of
this, or a population of about 85,000. This is virtually the same
as in 1939 and earlier years and is in accord with the expectation.
The decision to reject the estimate of 75,000, which was based on
electoral district data from a Republic election in favor of an
estimate of 85;000, -which was based on electoral district data from
a local election is in accord with the basic hypothesis that the local
election data give more accurate results.
The two illustrations above deal with situations in which the electoral
district data estimates of population indicate population growth or decline which
should be checked in the light of other information. The case of no indicated
growth should also be considered. This is the obverse of the Kokand illustration
In which no growtb, was expected, but a population decline was implied by the
electoral district data estimate. If ever a case is met in which the electoral
district data yield a populatiaa estimate which is the same as the number of
Inhabitants at an earlier date; a check should be made to determine whether, for
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized CO .y Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
....MINE111111111Mmi_.
- 30 -
the election in question, the electoral districts have in fact been adjusted
in areas of known or probable population change. There are grounds for believir
that there are sectional differences in the degree to which electoral districts
have been formed or altered to conform to real changes in population.
Consider, for example, the data on number of electoral districts
employed in the Republic elections in February and March of 1955. Leaving aside
for the moment the so-called special districts or military electoral districts
(of which there were 12 in the R.S.F.S.R., 3 in the Ukrainian S.S.R., 5 in the
Byelorussian S.S.R., and perhaps others elsewhere), it is found that the popu-
lation of the U.S.S.R. implied by the number of electoral districts amounts to
205.78 million. Not much is known about the so-called special districts. These
probably represent Soviet military forces outside the country -- as is known
to be the case for similar special districts employed in several All-Union
elections. In any event, it is not likely that they represent definable geographic
areas, for place names such as are customarily associated with each electoral
district in press accounts have not been found for these special districts. The
minimum estimate of 20 special districts cannot be converted into a precise
estimPte of population represented, for it is nct clear what legal ratio of
population to districts applies to these special districts. If each has the
legal ratio used for other electoral districts in the same Republic, the population
represented amounts to about 2.20 million, which is not an unreasonable figure
for Soviet armed forces abroad -- if that is what these constituents really are.
Hence, the population of the U.S.S.R. early in 1955 was about 205.78 to 207.98
million, according to the electoral data imputations. From data on total
population as of April 1956 and on rates of natural increase for 1955, recently
released by the Soviet Government," it would appear that the Population at the
Declassified in Part - Sanitized Co.y Approved for Release 2013/07/12 ? CIA RDP8i-niadqpnry,nnnr, r,r,,-,0
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
wee
- 31 -
beginning of 1955 was approximately 196 million, Over all, therefore, the
electoral district data seem to overstate the population.
This is not always the case for individual Republics. The estimates
for early 1955 based on electoral district data may be compared with the official
estimates for April 1956.34 In terns of the ratio of the electoral data based
estimate to the official estimate, the central Asiatic Republics milk as follows
(out of 16 Republics): 8, 111 14, 151 and 16. The rankings of the Transcaucasian
Republics are: 10, 12, and 13. The R.S.F.S.R. and the western Republics monopolize
the rsnks from 1 - 7; with Byelorussia rsntring ninth. Even if allowance is made
for differentials in the birth and death rates among these three sections of the
U.S.S.R. -- and this can be done only roughly -- it is clear that the electoral
district based estimates tend to overstate population in the R.S.F.S.R. and the
uestern. Republics, and to understate it in Transcaucasia and in Central Asia.
Consequently, a population estimate which is based on electoral district data
and which indicates no change in population since an earlier date, may suffer
from a failure to properly redistrict for electoral purposes; and may be
unacceptable for that reason. It is always worth testing to determine whether
this Is likely to be the case.
As a specific demonstration of this point, consider the following
data:
City
laShkent00?4.000000050.60
Margelan40006000OC OOOOOOO
pryoulation
19391 19502
585,005 600,000
44,700 45,000
1 1939 Census_
2 ImPuted from Republic election of 1950.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
....MN ? ???????MMINIEN.
-32-
As an initial conclusion it appears that neither city had much of an increase
between 1939 and 1950 indeed, since the deviation of both 1950 estimates from
their 1939 level is within the 5 percent margin of error claimed for electoral
imputations the conclusion of no change from 1939 is tenable. Before this
is accepted, however, the possibility of a failure to redistrict must be
considered. In this instance an examination of similar statistics for other
Uzbek cities indicates that redistricting was in fact carried out in preparation
for the 1950 elections. Therefore, unless Tashkent and Mhrgelan were treated
otherwise for some reason, the conclusion that their 1939 and 1950 populations
were approximately the same appears to be warranted.
Upon examination it turns out that there are reasons for accepting
this as roughly correct in the case of Marge1an.36 One would hardly expect,
however; that Tashkent, as the principal city of Uzbekistan, mould stagnate during
a period when the urban population of the republic increased by approximately
25 percent. Furthermore, as shown ?a -page 9 the case of Tashkent is complicated
by boundary changes. Therefore, it is not surprising that calculations based
on somewhat different data indicate the 1950 'population of Tashkent within the
1939 boundaries to be closer to seven than to six hundred thousand. The
imputation of 600,000 refers to the central city; the larger total to a larger
metropolitan area which approxinetes in area the city of 1939.
The illustrations given above do not, of course, exhaust the possibilities
for evaluation of the adequacy of population estimates based on electoral district
data. In other instances there will be other types of data available to facilitate
proper evaluation, and the ingenuity and skill of the evaluator will be an
important determinant of the types of analysis employed.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
447$
?????
?
?
- 33 -
Comprehensiveness of electoral district based population estimates.
Apart from the problem of how well a population estimate.baned on
electoral district data accords mith expected changes in population, there is
another type of problem to be considered before accepting such an estimate.
This is the problem of comprehensiveness or coverage. Are all the inhabitants
of an'area comprehended in the population estimate so derived, or are specified
or unspecified classes of persons excluded?
A comparison of populution estimates based on electoral data of various
types with population census figures is possible for the prewar period,
Table 6.--PERCENTAGE DEVIATION OF POPULATION ESTIMATES BASED
ON ELECTORAL DISTRICT DATA FROM CENSUS FIGURES, BY TYPE OF
ELECTION, FOR SELECTED POPULATION AGGREGATES! 1937 TO 1939
OmIIIIi????????????
Type of
election
1
Year.
1
Percentage deviation of
estimate from Census total
i
Population aggregate
?
All-Unionl......
1937
I 2+1.0
U.S.S.R.
Republic3?....
1938
z-0.4 '
Total of 11 Republics
Local:
City soviet,,
Oblast soviets
1939
1939
1 +0.2
+0.6
Total of 29 large citieu4
Total of 42 ?blasts,
autonomous oblastsland
krays in the R.S.F.S.R.v
------;,?
?
1 Election of December 1937.
2 Comparison with reported population of 169,000,000 for December 1937. See:
Pravda, 15 December 1937.
Election of lune 1938.
4 A similar comparison for certain Republics, A.S.S.R.'sland ()blasts can be
made against population estimates for 1938 which are given in Planovoye Khozyaystvo,
no. 5, 1941; pp. 75-89. The overall difference amounts to less than 0.2 percent. _
? Election of December 1939.
6 Aggregate population exceeding 15 million.
rj Aggregate population exceeding 97 million.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
mug ....1?1112.1
The very close agreement between the estimates of population derived
from electoral district data and the census figures is a very persuasive demonstration
of the thesis that; at least in the prewar period, the comprehensiveness of the
estimates was, for all practical purposes, identical mith that of the Soviet
census. Such close correspondence between electoral imputations and prewar census
figures does not resolve the question of comprehensiveness, for two questions
remain: (1) are census figures themselves comprehensive", and (2) does the
correspondence between estimates and census figures persist after 1939. An
affirmative answer to the first question appears to be indicated, although it
cannot be decided definitively? With respect to Soviet prewar census figures, the
following points suggest that they are relatively comprehensive:
1. The Soviets have striven for complete enumeration in an apparent effort
to conceal the large population losses of the thirties and in order Cso to
meet the informational needs of a planned economy?28
2. There are no marked abnormalities in the reported or estimated age-sex
structures of the U.S.S.R. population such as mould result from the
suppression of information on selected miloups.
3. There is evidence that slave laborers, the group perhaps most likely to
be omitted from the census, were actually enumerated."
4. The large gap (six or more million persons) in prewar and postwar estimates
between the number of eligible voters and the population of voting age
(18 years old and over) is suggestive of the comprehensiveness of the
census, since this gap is of a size that can readily accomodate the
estimated number of those unable to vote (slave laborers, inmates of
prisons, etc.).
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?????=??? ....,?????4111:=111.mmmo.
- 35 -
Inasmuch as there has been no postwar census, it is somewhat misleading
to speak of a persistence during the postwar years in the degree of correspondence
?
between estimates based on electoral data and census figures. There are varied
population figures for the postwar period, but none purports to be based on a
census, and differences between estimates and official figures may be due an much
to inadequacies in the official estimates as to inadeauacies in the basic electoral
district data. A summary of selected population estimates based on All-Union and
Republic electoral district data points up the problew
Table 7.--DISCREPANCIES BEIWEEN POPULATION ESTIMATES OF THE U.S.S.R. BASED
ON ELECTCRAL DISTRICT DATA FOR unal AND REPUBLIC ELECTIONS: 1937/1938
Po 1954/1955
???????????????????????????????.????????0/
Year
Type of election
JAll-Union -Republic
All-Union
I Republic
differences
1937/1938. ..... .......
170,700,000
169,630,000
1,070,000
1946/1947........ 0,0040
199,400,000
193,385,000
6,015,000
1950/1951........ 00000
202,0001000
198,610,000
3,390;000
1954/1955........
210,800,000
205,780,000
5,020,000
The official Soviet estimate of 200.2 million for April 1956, coupled with
official vital rates for the years 1950 to 1955, indicates that the population on
1 January 1951, was about 183.1 million and on 1 January 1955, was about 1959.
million.. It may readily be seen that estimates based on electoral district data
greatly exceed the official or officially implied population estimates. The
official figures imply a very great loss of population during World War II, and
it may be that this loss was not taken into account in the formulation of postwar
electoral districts. One hypothesis sometimes advanced in this connection is that
population gains in particular areas were taken into account, but that population
I??????
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 36 -
losses were ignored. Another possibility is that the official Soviet figures are
incomplete, perhaps excluding armed forces or other population groups. The
magnitude of the discrepancies between postwar estimates based on electoral
district data and official population figures is disquieting; and creates a
substantial doubt about the correctness of each set of figures.
Moreover, as may be seen in table 7; the level of population estimates
based on the All-Union electoral districts is much higher than that based on the
Republic electoral districts. Still another puzzling feature is the fact that
the population increase between 1951 and 1955, as implied by either set of estimates
in table 7, is considerably smaller than the increase implied by announced official
data on population and vital rates, that is some 7 - 9 million increase in the
estimates as compared with almost 13 million increase according to the official
data.
Lacking knowledge of the reasons for these discrepancies, the analyst
must use considerable caution in dealing with population estimates derived from
electoral district data in the postwar period, and particularly since 1950.
Unfortunately, other data resources are quite meager, and recourse must often be
had to electoral district data despite their deficiencies. The proper course to
follow calls for a selective use of electoral district data, judging results in
terms of their agreement with other information and in terms of their reasonableness
In a given situation. For example, the population of the Azerbaidzhan S.S.R. in
1949 as estimated by electoral district data" was 33.100,000. In view of the
fact that this republic is one of the Transcaucasian republics in which there
seems to be a systematic downward bias in population as estimated by electoral
district data, the estimate of 3,100,000 might be considered as an understatement,
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 37 -
Yet, information on the number of doctors in absolute and per capita terms41
In5icate2 a population of 3,099,000 for 1949. Apparently the same basic population
figure was used to prepare the figure on doctors per capita and to determine the
number of electoral districts in the republic. That number may have been too low,
but low or not, it is the number that was used,
Other ratio devices.
There is virtually no limit to the number of quantities that may be stated
in ratio to total population. The Problem is of course to discover items that;
in addition to being mensurable? can be shown to be reliably relatr_d to total
population. In the case of electoral district statistics this connection is
provided by decree, but for other items such relationships must be established
empirically. This becomes a problem in correlation and in the development of
estimating equations and is discussed briefly under the heading of Regression
analysis (p. 45 ). In this section, however, we shall give attention to (1)
reported ratios that yield relatively accurate approximgtions of census figures
and (2) the use of mixed sets of nonstandardized ratios where the test of validity
Is the convergence of results.
In connection with informational releases regarding the cultural and
welfare achievements of the regime, the Soviets occasionally publish certain useful
per capita ratios together with the associated numerator, e.g., the number of
doctors per 1,000 of the population and the total number of doctors. Such figures
are relatively rare for recent years and even when found are likely to involve a
useless combination of prewar ratio with a current absolute number. For the
immediate prewar period however, one may obtain for some 120 cities the ratio of
school children to total population as well as the number of school children.4'2
nprlaccifiPci in Part - Sanitized COPY Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
This is an especially valuable adjunct to the census since in this group of
cities are a number with less than 507000 inhabitants for which census results
have never been published. In general, however, the prospects for extensive
reliance on reported ratios of this type are poor.
Empirically standardized ratios on the other hand represent an unexplored
area of potentially great usefulness. 43 To date, however, relatively little work
along these lines has been carried out although Soviet arming literature -- to
cite one possible point of attack -- contains many suggested ratios.44 An
empirically standardized ratio is one which is not explicitly given in a Soviet .
source, but which is based on empirical information. For example, if there
should be available a series of data on the number of school teachers in a given
city, the ratio of school teachers to population at an earlier date or in a city
of similar attributes might be applied to the given number of teachers in the
city being studied to yield an estimate of population. Another illustration
might be data on cinemas, with no information available on the average number of -
persons served by one cinema. If, however: examination of cinema-population
ratios in other places: or in the Republic containing the city, indicates a
probable central value of a cinema-population ratio, that central value is what
is here called: for lack of a better term, an empirically standardized ratio.
In the population estimates program for Soviet cities carried on at the
Bureau of the Census, the use of such empirical ratios has been confined largely
to cases where there has been a complete absence of direct information on
population size and where the Objective of the work was merely the determination
of approximate magnitudes. Thus, in work done for the relatively new industrial - -
settlement of Cbirchik in the Uzbek SSR: the only basis for an estimate of
population size and for a quantitative expression of its development lay in the
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
-
- 39 -
use of such ezpirical ratios. For the most pert, the empirical mtios employed
were those which could be computed for the entire Uzbek Republic. The results
of this procedure are set forth in the following table.45 After the wol-k had
been completed, the 1939 Census figure for Chirchik was found." It proved to
be 14;700, and fits easily into the series of approximate magnitudes based on
empirical ratios.
Table 8, --POPULATION ESTIMATES FOR CHIRCHIK, UZBEK S.S.R.
BASED UPON SELECTED EMPIRICAL RATIOS: 1935 TO 1950
Year
Estimated population
Type of ratio
1935000000000
18,000
School children/Population
1936030000000
21:000
Labor force/Population
1938000003900
112,000-22,500
Republic electoral imputation
14;000-17,000
School children/population
1940.0?0000.
25,000
Newspapers/Population
27,000
Cinemas/population.
28,000
Workers' clubs/Population
33,000
School children/Population
1946....00000
37;500-52,500
Republic electoral imputation
101.
43;000
School children/Population
1950.........
45,000
School children/Population
37,500-52,500
Republic electoral imputation
3.
rfte990.993mmemmerGwer.00 (99310999090
Theoretical range is from 7,500-22,500. There were however,
a reported 7,000 workers and employees, making it unlikely that the
total population was as small as 7;500. The lower limit used allows
for less than 1 dependent per worker or employee.
Techniques of extrapolation.
Extrapolation procedures, which are merely the forward projection of
the trends of some known period, have frequently been used for postcensal estimates
of population. These techniques assume the continuation of existing trends and
are sound to the extent that the pertinent conditions affecting population growth
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 40 -
do not change radically. It follows then, that extrapolations are most trust-
worthy where short periods of time are involved. For these reasons, such methods
have not figured prominently in making population estimates for Soviet cities
since the projection periods involved ordinarily have exceeded 10 years. Furthermore,
the dynamics of the five-year plans and the crushing effect of World War IT
produced sharp and pervasive alterations in existing trends. If, for example,
either a simple arithmetic projection or a geometric projection had been used to
estimate the 1939 urban population of the U.S.S.R.; basing trends upon the period
1914-1926, the 1939 urban vottulation would have been underestimated by around
26 million persons, an amount equivalent to the total urban population in 1926.
On the other hand, these same procedures if used for the period 1939-1950, would
have produced an overestimation of some 10 million persons (without any allowance
for emexed urban areas and newly formed urban settlements). The accuracy of
arithmetic and geometric projections may often be improved by employing them in
connection with subgroups of the population or by combining the use of arithmetic
and geometric projections. For example, it has been advocated that arithmetic
projections be applied in making extrapolations of the number of expected migrants
and that births might be estimated on the basis of geometric projections. Variations
such as these generally have not been attempted for Soviet cities because of the
absence of data an subgroups, migration flows, and vital rates.
An investment of effort in the development of growth trend curves of
a higher order -- which quite possibly would conform to the data more closely
than either an arithmetic or geometric curve -- has been avoided because of a
preference for procedures with an explicit rationale.? At the same time,
component techniques of estimation, which give separate consideration to the
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
,
- 41 -
individual factors or components of growth, are usually not feasible because of
the almost complete absence of information on migration and vital rates."
Although these comments on the subject of extrapolation are generally
negative, it should not be concluded that such devices have no place in the making
of estimates of Soviet city populations. Extrapolation is a basic technique and
is frequently employed in a secondary role as, for example, the projection of
ratios or other relationships such as are encountered in connection with ratio
techniques or methods of apportionment.
Amortionment techniques.
Procedures included under this heading are those which break down or
apportion a larger whole. This might be the tctal population or perhaps a growth
increment of an area larger than a city. For example, the population of City A
may be estimated on the assumptinn that it bears some relationship to the total
urban population of its oblast or, again, the assumption might be made that the
city received a certain share of the growth estimated for a number of retties over
some period of time. Each problem ealls for an apportioning of shares of total
population on the one hand or growth on the other. Such procedures involve two
problems: (1) the estimate of the 'population of the larger whole, e.g.; either
the aggregate population of a group of cities or their growth increment and (2)
the estimation of what might be called an apportionment fraction. *
In the work with Soviet cities generally, two apportionment procedures
have been utilized. (1) Estimates of city population have been derived from an
expected share of oblast urban population. (2) In the case of cities of sufficient
size and importance to be organized into minor civil divisions (city rayons), the
problem has consisted of apportioning the estimated population increment, 1939-1950?
for cities of this .type. 'Where the problem involves the apportionment of urban
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 42 ?
aggregates larger than individual cities; the difficulties encountered in
deriving a figure for the aggregate may be as great and involve as much inaccuracy
as a direct estimate for a given city. Nevertheless, under special circumstances,
data may be available which permit such estimates to be made with an acceptable
degree of precision. The establishment of the apportionment fraction in such
problems is in most cases a difficult matter. The simple assumption of a
constant fraction derived from some earlier period is ordinarily not trustworthy
in view of the structural changes that have taken place in the urban population
of the U.S.S.R. during the Soviet period. Between 1939 and 1956 the number of
cities increased by 70 percent and the number of other urban settlements by 67
percent.' Under such dynamic conditions it is not likely that a city mould
represent a constant proportion of the urban universe over any appreciable span
of time.
Problems of a similar nature are met in the attempt to apportion an
estimated aggregate urban population increment, although these are somewhat
mitigated by working with subgroups of cities for which some homogeneity can be
assumed. Consider, for example; the universe of so-called Class I cities. These
are cities organized into city rayons. Included in this classification are most
of the cities of 100,000 or more inhabitants, as well as a few smaller cities.
As of 3anuary 12 1950, there were 64 cities of this class in the U.S.S.R. Designa-
tion of these cities as Class I is for convenience of reference; the terminology
is not found in Soviet usage. Among Class I cities there is a positive correlation
(r = .76) between the share of the aggregate increment received in 1926-1939 and
in 1939-1950;5? on the other hand correlation is negligible when computed for
all cities. The estimate of the growth increment of Class I cities is based
essentially upon data on the number of persons eligible to vote in local elections.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 43 -
The apportionment fraction customarily used is that computed for a city in the
earlier period, in this case, 1926-1939, although an examination of the relation-
ship suggests that some improvement could be obtained through the use of an
estimating equation which would take account of the tendency for smaller Glass I
cities to receive somewhat larger shares in the second than in the first period.
A6 mentioned above, data on eligible voters for the local elections
in 1939 and 1951 have been employed in estimating the growth of Class I cities.
Briefly, this procedure involves the assumption that the rate of growth of
Class I cities will have been similar to the computed rate of growth of the
number of voters, with an allowance being made for changes in age structure 51
Data on the number of voters, with the necessary classification by type of
administrative area, is most often available in the form of summary tabulations
by Republics. Typical of the format used is this tabulation published for the
R.S.F.S.R0 election of 1951:
Administrative division
Total number
of voters
Kray
7,262,140
Oblast
50,414,567
Okrug
428,683
Rayon
40,974,835
CitY0000119 OOOOOOOOO 0.0 OOOOO 0000?00
26,599,261
CityRayon.. repossee OOOOOOOO Coo000
15,129,863
Village and Settlement
'37,418,697
Source: p1.42yg.,p Moscow, December 22, 1950.
An easy and direct interpretation of such a tabulation is impossible because of
the overlapping character of the administrative categories that are used. Some
discussion of this problem is warranted by the usefulness these figures can have
when properly understood:
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Co Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
?
- 44 -
Voters shown as voting In elections for the soviets of krays, ()blasts,
and okr_ugs (total 582105,390) do not necessarily represent all voters in the
Republic. For example, in the R.S.F.S.R. in 1951, 64,800,000 voters were
listed as eligible to participate in the Republic elections held in February.
It is probable that the difference between this figure and the sum of voters
participating in kray: oblast? and okrug elections is represented by voters
residing in cities 'which are not subordinate to any of these administrative
units. The fact that such autonomous cities are sometimes districted in
connection with oblast elections tends to contradict this explanation: but
the internal evidence in the present tabulation suggests that even though
the population of autonomous cities may participate in oblast elections,
they are not so recorded in tabulations of voters. Another explanation of
this difference, for which there is considerably less evidence, is that
certain populations at large, having no identification with local administra-
tions and therefore not voting in their elections; do participate in
elections on a Republic level; e.g.: certain types of forced laborers, work
brigades, etc.
Rayon voters represent voters from all cities plus urban and rural
villages which are subordinate to rayon direction. This includes virtually
all voters except those residing in cities of oblast or Republic subordination.
The category of el.* voters includes all eligible voters from places
classified as cities, including Class I cities, but does not include urban
places otherwise classified, e.g., "workers settlements," "settlements of
city type."
Declassified in Part - Sanitized Co. Approved for Release 2013/07/1 : -R 1- "z
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-45-
Voters listed as city rayon voters are those from Class I cities.
Such cities are generally of either Republic or oblast subordination.
Village and settlement voters include those from either urban settle-
ments not classified as cities, as well as those from rural villages. The
sum of city voters and village and settlement voters (64,017,958) provides
a close approximation of the total number of voters in the Republic.
Finally full use of such tabulations regares that one consult the
official Soviet administrative handbooks:52 for the number of administrative
units of each type.
Regression analysis.
Ratio techniques, apportionment devices, and
all involve the acceptance of certain relationships, e
total population; of the share of the growth increment
share during 1939-1950; of the rate of increase in one
extrapolation procedures
.g., of school children to
during 1926-1939 to the
period
in the next; and so forth. In the case of electoral district
estimates of population derived from such data, the nature of
is adopted, often from some larger population or perhaps from
to rate of increase
imputations, i.e.,
the relationship
an analogical model;
sometimes it may be assumed on the basis of information for an earlier period or
for another place.
Still another, and intrinsically more promising approach is the msthod
of regression analysis, in which, given the data, the relationships between two
or more variables may be statistically ascertained. In essence, regression analyses
seek to formulate, in terms of suitable parameters: functional relationships between
related variables or, more precisely, between variables for which an a priori
interrelationship may reasonably be postulated.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 46 -
For example, it seems reasonable on the surfact that the crude birth
rate of a population is affected by such factors as the proportion of women of
childbearing age, the proportion of women who are married; the proportion of
married women who are employed, the number of marriages in the year preceding, the
average number of children per existing family, the availability of simple
inexpensive contraceptives, the proportion of the population holdinl; religious
views opposed to contraception, the average real income per family, and so forth,
Consider, therefore, the information that the crude birth rate of a country is
25 per thousand population, How helpful is this in determining an estimete of
the crude birth rate of a particular city or rural civil division? The national
rate may be a compound of very low rates in some oreas and very high rates in
others. If, however, it were pozsible to assemble data on crude birth rates and
on n of the other variables believed to affect the crude birth rate for a reasonably
large number of diversified areas, thes data could be expressed in a functional
relationship of the form
= ao aixi + a2x2 + 000 + anxn
where Y is the crude birth rate, and each x value is one of the other variables.
The coefficients, al:), a, coop
any are statistically derived parameters whose
magnitudes depend on the data assembled for the analysis, Then, knowing what
values xl, x2, xn have for some other area, these may be qubstituted in the
equation to obtain an esLAToat uf Y wiich takes .nto ;:eccorl the myriad
interrelationships of these variables. Incidentally, the coefficients themselves
have an analytic significance in that each is a measure of the variation in lc
attributable to the particular X with which that coefficient is associated; the
other factors being held constant.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?
Declassified in Part - Sanitized Cop A
?
proved for Release 2013/07/12 : CIA-R0P81-01043R002000060008-1
- 4? -
The particular formulatinn of a regression equation given above is in
linear form; that is, it assumes that Y and each X variable are linearly related.
This need not be assumed, however, if the pattern of emirically derived relation-
ships leads to another conclusion. The arithmetic grows more cumbersome, but the
nature of the estimating process is the same.
To return to the problem of using a multiple regression analysis for
the estimation of city population, consider the following illustration, which is
largely hypothetical.
1. Official Population estimates are available for the beginning of
1956 for 135 cities of 100,000 or more Population. Since population figures
are also available for 1939, the percentage of increase may be computed for
each city in this class. Let this be designated by yi, which is a particular
value of Y for city i.
2. Let it be assumed that data are also available for each city on
the percentage change in population during an earlier period, the percentage
change in the number of registered voters, the percentage change in the
number of children enrolled in school, the percentage change in the labor
force in large-scale industry, and the like. Let each of these variables
be designated by x., i indicating the city and j the specific independent
variable.Inthisnotatiaaxiiis a particular value of x. for city i.
3. A scatter diagram may then be prepared for Y andX.. For simplicity
of exposition it will be assumed that each scatter diagram indicates a
linear relation between Y and X..
4. Using well settled standard techniques for determining the parameters
of a multiple regression equation (see any elementary text on statistical
methods), an equation of the form
may be derived.
Y = ao aiKi a2X2?+ an Xn
Declassified in Part - Sanitized Cop Approved for Release 2013/07/12:
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 48 -
5. Given values of xij for city i, Y1, or the percentage of increase
in the population of city I way be estimated.
6. A graphical representation of the elements of this process is
given in figure 2, which, it must be emphasized again, is hypothetical.
The individual points represent pairs of values of yi and xv, x being one
of the independent variables. The line marked "mean is a representation
of the average value of yi for all the cities included in the diagram. The
solid regression line is a representation of the line obtained by fitting
a least squares straight line to the data, The small dash line (-----) is
a representation of the error inherent in estimating yi from the regression
line, and the large dash line (- a representation of the error
inherent in estimating y.1 in terms of the mean for all the cities in the
diagram. The illustration has, of course, been drawn to show the case of
a real gain in accuracy accruing from a regression analysis. There may be
eases in which there is no gain, however.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
0111..
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
SECRET
- 49 -
Figure 2.-- HYPOTHETICAL RELATIONSHIP BETWEEN PER ?
CENTAGE CHANGE IN POPULATION AND PERCENTAGE
CHANGE IN THE PROPORTION OF THE LABOR FORCE
IN LARGE SCALE INDUSTRY
PERCENTAGE CHANGE
IN POPULATION
90
MEAN
? ? .
? ? ? ?
? . ' ?
? ? . ? ?
?
30- ? ?? ? ? ?
? ?.
. ? ' ? ?
? ?
? .
? ?
? ? ?
? ?
20-
10-
?
?
? ?
? ?
?
0-I
o ;
I I I I I
1
10 15 20 25 30 35
I
410 4'5 510 5'5 60 65
PERCENTAGE CHANGE IN THE PROPORTION OF THE LABOR
FORCE IN LARGE SCALE INDUSTRY
U.S. DEPARTMENT OF COMMERCE U.S. BUREAU OF THE CENSUS
SECRET
.--
-
-
Ila.q.
-
-
-
-
-
_
-
_
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
-50-
? It will be recognized by those familiar with methods of regression
analysis that the prediction model discussed here is rather pat. There is little
hope, for elample, that significant improvements in prediction could be obtained
from the 1103 of a single variable such as upercent of change in the proportion -
of the labor force in large-scale industry. "53
Moreover, the low level of%conceptual and mensurative development in
the field of urban research will almost certainly require resort to statistical
devices 54 that can deal with nonquantitative attributes. Thera is little question
for example that a careful typology of cities and a sorting of cases by regions
would of itself narrow the range of estimation error as contrasted with the
practice of basing estimates on the mean value of an undifferentiated universe.
The real problem:: however, is not the adequacy of statistical devices but rather
the lack of both data and hypotheses in this area of research. Results of similAr
work in the United States? have shown a considerable part of the variation in
growth rates to be a function of (1) previous rates of growth and (2) regional
differences,. The nontheoretical nature of such findings, however, in the absence
of extensive comparative studies, leaves US without a basis for assessing the
likelihood that such a relationship will stand up from sample to sample. In such
instances we ere dealing merely with refined extrapolations. Hence improved
esti:maces of total population by vay of regression analysis must av7alt the necessary
prs!nt.rj of br,t;-1 copi y,d r:c7to
ValidaLion of Estinatos
The term validation is used here in the sense of arllie7ing confirmation
Three different types of procedures are proposed in addition to .such vAlidation
tests as may ordinarily be employed In adopting a7.:1 method of estimatim
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 ? CIA-RDP81-01043R0020000ennnR.1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 51 -
Therefore, this discussion refers only to the validation of results obtained
by the methods discussed in the earlier portions of this chapter. For convenience
these validating procedures may be referred to as tests of convergence, closure,
and consistency.
Convergence tests.
In view of the very great shortage of validating items it will
ordinarily be advisable to derive a city population estimate by more than one
of the methods that have been presented. An agreement or convergence of results
increases the acceptability of the estimate. For example, an estimate of
150,000 inhabitants for the city of Samarkand has been made on the basis of
the number of electoral districts formed for the Republic elections of 1951.56
In _support of this there is an alternative estimate of 145,000" based upon
Samarkand 's expected share of Kazakhstan's Class I city growth between the years
1939 and 1951. The agreement here is *within acceptable limits. As mentioned
in connection 'with the discussion of ratio techniques, each individual estimate
should be considered also in terms of the reasonableness of the changes it implies
and this should be done before attempts are made to validate it by way of convergence
with other estimates or by any other means.
Closure tests.
-
This method of validation refers to the requirement that population
estimates of individual city population when added together fit with known or
firmly estimataigroup or regional totals. Thus, for example; the population
estimates made for the cities of Fergana and Margelan were added to an estimate
previously made for the city of Kokand and the sum of these was compared with an
estimate of the total urban population of Fergana oblast in which all three cities.
were located. The totals involved were as follows:
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
-52-
Fergana-Margelan.. ......... e 120,000
Kokand 85,000
5 settlements of city type 35,000
Estimated total urban population of
Fergana oblast ..... 240,000
The aggregate population of the five settlements of city type in the example
above is a residual and represents a certain degree of looseness in the application
of this particular test. It does, however, yield a reasonable figure for the
average size of such settlements, which would not be true if there were gross
noncompensating errors in the estimates of any one of three cities. Ultimately,
tests of closure should be applied to estimates which have been made for large
numbers of cities by comparing their accumulated total with estimates independently
derived for larger urban aggregates.
Consistency tests.
The tests of validity dealt with heretofore have been primarily
demographic. The consistency of city population estimates with nondemographic
information is another approach to the question of validation. The statistic
that has been used in this way most often is the size of a city's total labor
force. Unfortunately not enough is known of the behavior of crude labor force
participation rates to establish criteria of acceptance or rejection ig terms of
given levels of participation. It is not improbable, however, that research in
the future will supply such yardsticks. An examination of labor force participation
rates in the United States in 1940-1950, in postwar Czechoslovakia, and in postwar
Poland, reveals rates averaging from 45 to 50 percent with standard deviations
of 2 to 4 percent. This suggests that under certain conditions the labor force
participation, rate is a highly stable statistic. However, the labor force
participation rates" of Soviet cities in 1936, the latest period for which
any-n-773g like comprehensive data are available, Show considerably greater
variation thqn that observed in the United States, Czechoslovakia, or Poland.
-
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
? ... - OPP ...M..
-53-
Cities of the Ukraine and of the Central Industrial Region, where the level of
crude labor force participation rates is comparable with that noted for the United
States and Czechoslovakia, showed over twice as much variation. This is due in
large part to the uneven economic development of Soviet cities up to 1936 as
well as to certain definitional problems of the labor force data in question
which tend to exaggerate veal differences. This does not mean that the labor
force estimate is not a useful statistic for validating city population estimates.
It does mean, lc.owever, that at the present time such validation must ordirnrily
be carried out on the basis of an examination over time of the labor force
participation rates of a given city. (Data for a chronological treatment of
the labor force are discussed in chapter 3.) In other words standardization must
be achieved through comparison with previous levels. Sudden changes in the level
of labor force participation rates2 especially declining labor force participation
rates in the case of cities favored by Soviet economic plans, challenge the
validity of a population estimate -- assuming the labor force estimate to have
been previously validated. On the other hand it does not always follow* that
one is dealing with an acceptable labor force participation rate if no violent
deviations from trends are observed. In such cases the level of the rate may
still be quest1oned. Some standardization may be obtained also by an inspection
of the known participation rates of cities of a similar industrial character.
Finally, all consistency tests which concern themselves with the labor force
are complicated by variations in the proportion of nonresident commuting workers
and by variations in the extent to which special tyues of temporary labor may
have been imported into an area.
Declassified in Part - Sanitized Copy 'Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-54-
Other types of demographic and labor force comparisons can be made
so as to provide secondary tests of the validity of estimates of total population
or of growth increments. For example, as is pointed out in chapter 2, the
estimation of the age-sex structure of a population depends to an important degree
upon the estimate of total population that has been made. Therefore, a comparison
of the labor force of a city and the population groups of working age available
to carry on the activities of the city will in turn affect the evaluation of the
total 'population estimate.
Data on budget expenditures which are available for some of the larger
Soviet cities provide another, apparently rather promising, approach to the
problem of validation. Experience with city budget data has been limited, but
as shown (figure 3) by the relationship between population growth and growth of
budget expenditures (one year lag) in the sixteen ()blasts of the Kazakh Republic,
these may not be unrelated events. Because of differential and compensatory
demographic movements within ?blasts, it is likely that the association between
over-all population change and over-all change in investment is somewhat
attenuated59 when the oblast is the unit of observation.
The tests that have been proposed do not exhaust the possible methods
of validation. Unique items of information are sometimes available and may
often provide tests that override any that have been discussed here. As work
advances in the field of human ecology as well as in other phases of the social
sciences, it is to be expected that many types of relationships will be uncovered
that will lead to a greater battery of available tests.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 55 -
SECRET
FIGURE 3. RELATIONSHIP OF POPULATION CHANGE TO CHANGE IN BUDGET
EXPENDITURES,OBLASTS OF THE KAZAKH SSR
PERCENT CHANGE IN
POPULATION, 1947-1955
70
60
50
40
30
20
10
0
-10
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
40 60 80 100 120 140 160
PERCENT CHANGE IN BUDGET EXPENDITURES. 1946-1954
180
SOURCE: ESTIMATES PREPARED BY THE U.S. BUREAU OF THE CENSUS ON THE BASIS OF INFOR-
MATION IN KAZAKHSTANSKAYA PRAVDA, 19 FEBRUARY 1947; II DECEMBER 1950; AND 30 DECEMBER
1954. ALSO BASED IN PART ON DATA IN TS.ENTRALI NOYE STATISTICHESKOYE UPRAVLENIYE, SOVIET-
SKAYA TORGOVLYA, MOSCOW, 1956.
U.S. DEPARTMENT OF COMMERCE
SECRET
U.S. BUREAU OF THE CENSUS
-
-
-
-
-
-
--
-
-
_
-
_
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
11 jT
- 56 -
FOOTNOTES
CHAPTER I
1 Tsentralinoye statisticheskoye upravleniye pri sovete ministrov SSSR. Narodnoye
khozyaystvo SSSR, statisticheskii sbornik (The National Economy of the U.S.S.R., A
Statistical Compilation): Moscow, 1956; and Statisticheskoye upravleniye RSFSR.
ficif66d-ElidigaTgiv-FR'SFSR, statisticheskii sbornik (The National Economy of the
Moscow,-f937:--- ------------------
-2 Verkhovnyy sovet SSSR- Informatsionno-statisticheskiy otdel. SSSR
Administrativno-territorial'noye deleniye soyuznykh respublik (U.S.S:g:-Administrative-
UFFIEaTal Divisions of the Union Republics), various years.
. Vedomosti verkhovnogo soveta SSSR (Journal of the Supreme Soviet of the
U.S.S.R.): intermittent.
TEre may be some disparity in areal coverage among various election statistics.
This problem is dealt with below.
5 Boundaries for 1936 shown on NKVD/ USSR. Kerte Uzbekskoy SSR (Map of the
Uzbek S.S.R.), Scale 1:1/500,000, Tashkent, 1938. For the published list of election
distriaardee Pravda Vostokal-24 December 1950.
6 In part this may be inferred from the fact that there has been no general
Soviet census since 1939. Moreover, Soviet delegate Pissarev to the World Population
Conference in 1954 admitted that "the polls of electors serve as a basis for calcu-
lation of the number of urban and rural (persons) 18 years of age and over....lists
(of school children) serve as a source for calculation of the contingent of children."
From a paper cireulated at the Conference entitled "Organization, notions: and
determinations of demographic statistics in the u.s.s.n.11
' Frequently expressed in term of election districts rather than elected deputies.
The former are preferred since such a list is not affected by the deaths of candidates,
and inconclusive contents both of which may reduce the number of elected deputies
reported as of a specific date.
8 Polozheniye o vyborakh v kraevyye, oblastnyye; okrugnyye, raionnyye? gorodskiye,
sel'sWes-i?p.oselkovyye sovety Te.D'utatov trudyaehchikhsya RSFSR (Regulations on
elections to krai: oblasts okrug/ rayon, city, village and settlement soviets of
workers deputies R.S.F.S.R.), Moscow/ 1939; also see Vyshirskiy? A. Ya. Izbiratel'nyy
zakon SSSR (Election Law of the U.S.S.R.): Moscow: 1950,
9 Nemev, N. Vybory v mestnye sovety d9utatov trudyashchikhem.(Elections to local
soviets of workers deputies); Moscow; 1947; see also Article 26 of the Decree on the
Elention Law of January 9, 1950.
10 Republic summaries of ulerstion results (deputies elected) by type of local
election, viz., to krais oblast, city: city won and village soviets, are available,
but because of the variation in the value of within each category these data have
little utility. Similar summaries of the numger of voters are useful and are
discussed under the heading of Apportionment technique;.
With respect to the election of deputies to the Sepreme Soviet of the U.S.S.R,
Artiele 135 of the 1936 coeutitution states that "all eitizens of the U.S.S.R. who
have reached the age of 18: irrespective or race or nationality, sex, religion,
education, domicile, social origin, property status or past activities have the right
to vote....with the exception of insane persons and persons who have been convicted
Declassified in Part- Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-0104:1Rnn9nnnnAMICIO
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
-57-
by a court of law and whose sentences include deprivation of electoral rights."
In the revision of the electoral decrees approved in 1950 the word residence was
subst.tuted for domicile. In addition, it is stated (Article 12 of 1950 revision)
that "All citizens who possess the right to vote, who reside (permanently or
temporarily) on the territory of a given soviet at the moment orgliTiR5HIThg of
voters' lists....are included in the voters' lists." In Article 13, it is specified
that no voter may be registered on more than one list. The mechanics for preventing
double counting on voters' lists are given in Article 20.
12 Polozheniye o vyporakh....op.cit.
13 This is not invariably true. For example, if election statistics for several
consecutive years are available and if during this period the area for which a
population estimate is desired appears in varying combinations with other adminis-
trative units; it may be possible to derive an acceptable estimate by way of a set
of simultaneous equations. In this connection, see U.S. Department of Commerce,
Bureau of the Census, CONFIDENTIAL, Estimated Population and Labor Force of the
Aleksin Industrial Area: 1 January ID50, Series P-951 No. 35, Washingten, D. C.,
10 August 1954, pp. 21-30, CONFIDENTIAL.
14 Zglngmlye o
15 This measure undoubtedly exaggerates the error inherent in an electoral district
ratio estimate of population in instances such as Voronezh and Yaroslavl' in table 2;
in such instances, there is no extraneous territory with which to contend,,,and
experience indicates that the error is apt to be considerably less than :`2!"- . In
general, however, without knowledge of the extent to which the extraneous territory
has been included in election district association with the city, this formula
applies.
If a city or oblast or other area has an integral number of electoral
districts, di, and these districts were delineated to conform ideally with a
prescribed population per district, RI the estimate of population by the ratio method
is:
:= d.
Pi
141-11 a range of error of ? -27 .
If each district had precisely the prescribed population, there would be
no error in this estimate. In general, however, an area will not be entitled to an
integral number of electoral districts, di., but rather to a fractional number,
P./ft., On the assumption that (D. - d.) is rectangularly distributed, i.e., it
Di
is equally likely for (D. - d.l) to have any value in the range t 1/2, its variance
willbe1/12.andthevarianceof.
PI will be R2/12.
This result permits the deternination of the range of error associated
with electoral district ratio estimates of population for larger areas which
are composed of a finite number of constituent areas for which electoral district
ratio estimates of population may be made. For example, consider the problem of
estimating the total population of a Union. Republic with N constituent oblasts each
having di electoral districts. The estimated population of each oblast would be:
P. = R d.
I.
and the estimated population of the Rnpublic would be:
P=Rtt.d
i?
The variance of P is equal to the sum of the variances of the estimates for each of
the N subareas, there being no correlation between these estimates; that is:
6p2 NR2
12
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-
- 58 -
and the standard deviation is:
It may thus be said with virtual certainty that P R d. 3R7", and that two
1 12
times out of three the estimate of P would have a range of error of ? RVIE
This
result
result conforms with intuition, for if each of the N subarea estimates has A range
of error of i R/2, it would be expected that some of the errors mould cancel each
other when the N subarea estimates were added together. At the worst, if each of
the subarea estimates had an error of + R/2, their sum mould have an error of + NR/2
which is 1.71/Pitimes as great as the standard error of P.
lc; In drawing un election districts, oblast boundaries are observed. Rayons and
cities may be split between several districts and combined with other administrative -
divisions.
17 Population approximately 25;000 in 1939.
18 The discrepancy between the imputations from the Oblast and city elections is
well within the tolerance limits for work of this kind.
19 American Russian Institute. Constitution of the RSFSR of January 21, 1937 as
Amended Through May 27, 1949 (a translation), April 1950, pp. 18-19. See especially
Article 145,
20 Foreign Broadcast Information Service, Daily Report, Foreign Radio Broadcasts,
No. 120, PO June 1956, op. DD-4, 5, 6. "S-67-Eilso Boyarskiy, A. Ya. and Shusnerin,
P. P., Demograficheskaya Statistika (Demographic Statistics), Moscow 1955, pp. 224,
225, and 235.
2i Voprosy Ekonomiki (Questions of Economics), january 1955, No. 1.
22 No Republic has suffered a decrease in the number of electoral districts for
elections to the Supreme Soviet of the U.S.S.R. Electoral districts are abolished,
however, in connection with Republic and local elections.
23 There are also differences between Republics in the norms employed in connection
with both Republic and local elections.
24 For elections of Uzbek S.S.R. Supreme Soviet the legal ratio of population per
delegate is 15,000. There were 10 deputies elected from Samarkand (see Pravda Vostoka)
December 24, 1950). The theoretical range for this estimate is 142;500 to 157,500.
25 This involves comparisons with other Uzbek cities which will not be made here.
For validation of this conclusion, see U. S. Department of Commerce, Bureau of the
Census, CONFIDENTIAL, Estimated Population and Labor Force of the City of Samarkand:
1 January 1950, Series P-95, No. 37, Washington, D. C., 20 January 1955, SECRET.
Ibid., p. 16.
27 Ibid., pp. 16-17.
28 Narodnoye khozyaystvo p. 25.
29 Five deputies from Kokand were elected to the Supreme Soviet of the Uzbek S.S.R.
in 1950 (see Pravda Vostoka;
3? See U. S. Department of Commerce, Bureau of the Census, gONFIDENTIALI Estimates
of the Population and Labor Force of the City of Kokand: 1 January 1950s Series P-95,
No. 46, Washington, D. Go, November 1955, SECRET.
31 Pravda Vostoka, 10 October 1950 and 4 November 1950.
32 Vitkovich? V. Puteshestviye po Sovetskomu Uzbekistann (Travels in Soviet
Uzbekistan), Moscow, 1st ed., 1951; 2nd ed., 19557 p. 131.
?3-37Pricidnoye khozyaystvo SSSR, on.cit., pp. 17, 243.
34 Ibid., P. 18.
33 Poiazheniye o vyborakh....on.cit.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R062060060008-1
.64
? 5-
-59-
36 See U. S. Department of Commerce, Bureau of the Census, CONFIDENTIAL,
Estimated Population and Labor Force of the Fergana-Margelan Metropolitan Area:
1 January 19507 Series P-95, No. 45, Wa6hington, D. O., November 1955, SECRET.
" After making due allowances for underenumeration resulting from technical
deficiencies.
38 For a somewhat one-sided but revealing account of Soviet Census methods
see Galin, P. Kak proizvodilis perepisi noseleniia v SSSR (How -Population
Censuses are Conducted-in the U.S.3.R.), Munich, 1951.
-----?--Petrov7-71affimir. soviet Gold, Farrar, Straus and Company, New York, -
1949t p. 369.
40 See Izvestiya, February 9 and 16, 1947 and Pravda, February 22-24 and
March 12 191..
41 According to data presented by Boyarskiy and Shusherin, there were 5,902
doctors in Azerbaydzhan in 1949 with a ratio of one per 525 inhabitants.
42 Tsentralinoye upravleniye narodno-khozyaystvennogo ucheta gosplana SSSR.
Kuliturnoye stroitellstvo SSSR (Cultural Construction of the U.S.S.R.) Moscow, 1940.
Not only is it possi le t at ratios can le found that wi -Taggess greater
reliability than those presently available, e.g., election data, but with the
development of batteries of such ratios based on a number of different kinds of
items, it should become possible to make estimates of total population more
quickly and with increased margins of confidence. This would be of special
importance under emergency conditions when a large volume of current estimates
would be required. It would also serve to sharpen the focus and increase the
return from our interrogations of POWs and refugees, as well as of our positive
intelligence.
44 See for example the volume published by Akademiya arkhitektury SSSR,
Yezhegodnik arkhitektora (The Architect's 'Yearbook), 1947, pp. 15ff. The list
includes the ratio to total population of such items as hotels, hospital beds,
various service groups, areas devoted to various types of land use, cultural
facilities and so on.
45 For fuller treatment of the Chirchik problem see U. S. Department of Commerce,
Bureau of the Census, CONFIDENTIAL, Estimated Population and Labor Force of the
City and Metropolitan Area of Chirchik: 1 January 1950: Series P-951 No. 24,
Washington, D.--C77-10--garch 1954, SECRET.
46 Akademiya nauk Uzbekskoy SSR. Institut ekonomiki. Uzbekistan, Ekonomiko-
geograficheskaya kharakteristika (Uzbekistan, Economic-geographic Characteristics),
Tashkent, 1950, p. 72.
47 Development of such curves is limited also by the small number and low
reliability of the available observations of city size.
48 In the United States, the migration component of population growth has been
estimated from data on school enrollment. Such data are sometimes available for
Soviet cities. See: U. S. Department of Commerce: Bureau of the Census,
Population, Special Reports, Series P-47, No. 4; U. S. Department of Commerce,
Bureau of the Census. Current Population Reports, Population Estimates, Series
P-25? No. 12; and Shryock? H. S., Jr. and Lawrence, N. "Current Status of State
and Local Population Estimates in the Census Bureau," Journal of the American
Statistical Association, v. 442 no. 246, June 1949, p. 157.
----4971-5-1?'odni:Fra-6F5a7stvo SSSR, op. cit., p. 10.
50 This correlation excludes Moscow, Leningrad, and three other cities that
were severely damaged by the war and represent, therefore,. marked deviations from
the general tendency.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
!:;
^
- 60 -
51 Data on eligible voters can be utilized in connection with ratios of the
type discussed in the preceding section. Where such data are reported for
specific localities the problem is merely one of estimeting the ratio of voters
to total population. Somewhat more reliable estimates can be achieved if, in
addition, data on school enrollment are utilized.
52 Verkhovnyy sovet SSSR. Informatsionno-statisticheskiy otdel. SSSR
Administrativno-territorialtnoye deleniye op.cit.
In fact, there may be virtually no association between the rate of growth
of cities and the rate of change in the percentage of the labor force in large-
scale industry. An attempt has been made to measure the association between
the rate of population growth during 1939-1950 and the percentage of the labor
force in large-scale industry in 1936, the last year for which data of this type
were available for Soviet cities. There was none. A moderate negative correla-
tion has been found by investigators working with data for the United States.
See: Bogue, Donald J. and Harris, Dorothy L. Comparative Population and Urban
Research via Multiple Regression and Covariance Analysis; Scripps Foundatii5E-for
Research in Population Problems, Miami University, and Population Research and
Training Center, University of Chicago, 1954, p. 21.
54 The joint consideration of quantitative and nonquantitative variables is
possible through the use of covariance analysis, see) for example, Bogue,
Donald J. and Harris, Dorothy L. op.cit., pp. 2-18.
55 See Bogue, Donald J. and HarYfg;-15orothy L.: op.cit., pp. 40-54.
56 U. S. Department of Commerce; Bureau of the Census, CONFIDENTIAL,
Estimated Population and Labor Force of the City of Samarkand: 1 January 1950,
STP.-95, No. 37, Washington, D. C., 20 January 1955, p. 18: SECRET.
57 Ibid., p. 16.
58 gtrictly speaking these are not labor force participation rates since only
members of the labor force carried on the roles of local enterprises and
classified as workers and employees are considered.
59 The coefficient of correlation in this case equals .74, and the coefficient
of determination is about .55; indicating that about 55 percent of the variation
in population growth is "explained by" changes in budget expenditures. Because
of the small number of cases involved in this illustration, all temptation to
generalize this relationship should be resisted-jii&E-fholidh the null hypothesis
can be technically rejected.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized CopyA proved forRelease2013/07/12 : CIA-RDP81-01043R-0-02-000060008-1
t
6
- 61 -
CHAPTER II.
ESTIMATES OF AGE AND SEX DISTRIBUTION AND OF ETHNIC COMPOSITION
FOR THE POPULATION OF orrIEs OF THE U.S.S.R.
The enigmatic character of Soviet demography is once more illustrated
by the nature and scope of data available for the preparation of estimates of
population by age and sex, or ethnic composition: for Soviet cities. Much of
the available material of potential utility is quite out-of-date: great gaps
in information must be filled in -- albeit often on a tentative basis, and
special adaptations of estimating techniques must be formulated to cope with
the severe limitations imposed by the nature of the data. A satisfactory
mastery of these problems is worth a considerable expenditure of time and effort,
for if it is possible to devise techniques which will adequately reveal at least
the main features of the age and sex profile of the population of Soviet cities,
a most valuable tool for social and economic analysis will thereby be created..
Significance of Age and Sex Distributions and Ethnic Composition
The distribution of the population of an area by age and sex is a unit
of information which has manifold potential analytical applications. It may be
used to compute significant age-specific rates; to compute standardized rates
by which cities may be meaningfully compared with each other; to establish
limits for the size of the labor force and the measurement of the nonresident
or commuting labor force; to measure the replacement prospects of crucial cohorts
such as males of military age or females of childbearing age; to measure depend-
ency ratios; to measure sex ratios by age; and for many other analytic purposes.
Recent history provides good illustrations of the practical importance of
demographic factors. Witness the demographic deficiencies of France and their
relation to its military defeat during World War II, or the impediments to
Declassified in Part - Sanitized Copy A. proved for Release 2013/07/12 ? CIA-RDP81 01(M1Pnn9nnnnprinno 4
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
France's postwar economic recovery stemming from the increase in the proportion
of superannuated workers. Consider also that throughout the industrialized
world: and especially in the Soviet Union, demographic influences in the form
of lowered sex ratios among adults have contributed to increased levels of
female labor force participation. The versatility of the age and sex distribution
as an analytic tool is very great, and the range of problems which may be the
more readily solved by means of this tool is equally great.
Ethnic characteristics may also be viewed as limiting conditions
although somewhat more susceptible to cultural elaboration and, therefore, of
variable significance as compared with age and sex. At any one time, however,
the interpretations given to ethnic differences will provide one of the central
principles of organization by which the success odds of a particular society
are set. Where ethnic groups are normatively solidified and coherently
organized, i.e., incompletely assimilated, ethnic differences will tend also
to be differences in training and education, in loyalties, motivation slin
morale -- in short, in all of the correlates of differential social status.
Stated in terms of patterns of communal action, the degree of ethnic crystalliza-
tion to a large extent conditions the general performance standards and energy
levels of a population and its resilience in the event of adversity. The
analysis of ethnic composition should not end, therefore, with the estimation
of a frequency distribution. It achieves full significance only when the ways
in which ethnic differences operate in'the social system have been revealed.
The implications to be gathered from a study of demographic charac-
teristics depend upon the extent to which the situation is one that cannot be
immediately altered. This is, in general, the case with large, self-contained
units, such as nations, where the influence of migration is limited. On the other
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-63-
hand; demographic descriptions of small; dependent areas have a high rate of
obsolescence due to the metamorphic effect of migration on age-sex and
ethnic profiles. Thus, for example; we cannot expect to attach the same
significance to a shortage of males or to a preponderance of a given ethnic
grove in a local situation as we would if this were found to be true for an
entire nation. It follows from this that a full evaluation of the demographic
structure of any area must make some reference to its degree of stability,
which in tarn involves a consideration of such matters as the extent to which
local demand for labor has been met and the demography of Potential areas of
population supply.
The utility of estimates of age-sex and ethnic structure is not
confined to what may be inferred from them directly. Going a step beyond the
view of population composition as a limiting factor in communal organizations
it may be hypothesized that for each set of organized activities there is an
appropriate population base, both in terms of size and composition. Thus,
for example, estimates of age and sex composition are useful in attempting to
validate estimates of other communal structures; such as the labor force,
which may have been derived independently. This is perhaps the most frequent
use that is made of age-sex estimates for small areas and it involves the
estimation of the labor force through the application of probable rates of
labor force participation to estimates of population by age and sex. Estimates
of ethnic composition cannot be so used because the relationship of ethnic
structure to the size of the labor force is less direct and less reliable. A
knowledge of ethnic structure will however have a bearing on the evaluation of
a city's regional or interregional role, and the extent of its Sovietization.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
:
- 64 -
Except for the Western border regions of the U.S.S.R. (including the Ukraine
and Byelorussia), important Soviet cities have been Russian cities, even when
located in non-Russian areas.1 Conversely, a city without t substantial
complement of Russians is likely to display a low degree of industrialization
and to be of local importance only. Even though ethnic data for cities are
generally lacking, the general truth of these propositions is indicated by
the patterns of discrimination in both industrial employment and education
revealed by Soviet data. Therefore, estimates of age and sex distribution
and ethnic structure provide both a quantitative and a qualitative framework
against -which the character and significance of a city can be evaluated.
The usefnlness of estimates of ethnic composition does not necessarily
increase with the number of distinct groups that are identified. The required
degree of precision for particular evaluations is frequently difficult to
specify, since this depends upon a knowledge of the important symbiotic
relationships, such as between Russians and Tatars, that operate in any
given community. For an analysis focused on manpower and the associated
problems of skill, training and education; it may be sufficient to identify
the numerically significant ethnic groups. The purposes of psychological
warfare may be better served by somewhat more detailed analysis2 for it is
often true that numerically insignificant groups have a significant influence
on the behavior of the larger community.
Closely allied to the question of significance are the problems of
accuracy and precision. It is obvious that dependable significant conclusions
cennot be derived from grossly inaccurate data, and where, as in studies of
the Soviet Union; measures of accuracy may be missing, it is often necessary
to state conclusions in the subjunctive. This is a familiar problem and it is
generally understood that analyses of Soviet society have many tentative
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
? 1-7,`
'4.11
- 65 -
aspects and are possible at all only through the substitution of assumptions
for unavailable items of information. The point to be rade here is that
this imposes limits upon the degree of precision that can be expected in the
final product. For this reason population estimates by age and sex for Soviet
cities have usually been presented in terms of very broad age groupings of
which the following is representative:
0 - 14
15 - 29
30 - 59
60 years and over
There is no doubt that the use of such broad gauge class intervals increases
the margin of error that can be tolerated in the various component steps of
the estimating procedure. This advantage is offset, however: by the fact that
the uses for age data presented with this low degree of Precision are greatly
restricted. Eliminated in this may is one of the most fruitful approaches to
a validation of labor force estimates, namely through the use of a carefully
derived, detailed set of age- and sex-specific labor force Participation rates,.
It will be noted also that this set of age intervals, which has been adopted
primarily in connection with manpower analysis, is ill suited for analysis of
other facets of Soviet society, e.g., the educational system, 3 Soviet fertility,
the electoral system and so on.
In summAry? the question of significance is a matter of the uses to
which the estimates are put. Although age and sex distributions and ethnic
composition estimates have a high potential for significant implications,
considerable caution is required in their interpretation for small areas where
the population structure may be undergoing a rapid transition. On the other
hand, the estimates of the composition of Soviet cities which our present
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
,
-66-
methods permit us to make with only a low degree of precision, meet certain
minimum specifications for labor force analysis. For the investigation of
other sectors and problems of Soviet society somewhat modified procedures
and results would bp demanded.
Estimation of Population Distribution by Age and Sex
Component projections.
One method very often used for deriving estimates of a city's
population by age and sex is the component projection. This method consists
of three parts, each with many associated minor problems which will be discussed.
The first task is that of estimating the number of persons in the population
at some base date for which data are available who can be expected to survive
to the end of the projection Period. To these survivors then are added
separate estimates of the net change in population due to migration and of
the survivors of children born in the comatunity between the base and terminal
dates.4 In the case of Soviet cities the base date is ordinarily 1926 which
is the last date for which detailed age and sex data are available for small
areas. Current estimates thus involve a projection period of about 30 years,
and unless more recent demographic information is made available by the
Soviet government it will probably become advisable for future work, because
of the length of time involved, to replace component projections based on
the year 1926 with some other method of estim-tion. An understanding of the
component method as it has been developed in connection with Soviet materials
is, however, a useful background in a search for new methods. Accordingly,
a portion of this chapter is devoted to an exposition of the three principal
estimation problems of the component method: (1) survivors of the base date
Population? (2) the survivors of births:, and (3) the net change attributable to
migration.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
67 -
Survivors of the 1926 population.
Of all of the steps involved in a component projection, the
estimate of survivors poses the fewest difficulties. Simply stated, the _
task is to estimate the population in the various age and sex groups of
the base population that can be expected to survive over a given period.
There are at least three subsidiary problems to be considered. These are:
(1) an evaluation of the 1926 Census base population, (2) the identification
and removal from the base population of special groups, such as military
garrisons, that are replaced by persons having similar characteristics, and
(3) selection or construction and application of survival factors from an
appropriate life table, and the calculation of mortality modification factors
where these are warranted by changes in mortality conditions subsequent to
the period to which the life table refers.
Evaluation of the 1926 Census base.
A, definitive evaluation of the 1926 population census has never been
made, although early steps in this direction were taken contemporaneously by
Soviet demographers and later by outside specialists. Further work on the
evaluation of the 1926 Census is now being carried on by the same U. S.
Bureau of the Census staff that has been concerned with U.S.S.R. city
population estimates. A report on this subject will appear in due course
in this same series of reports. Despite the absence of a definitive evaluation
there is clearly ample evidence of faulty enumeration. Most published
analysis has been devoted to revealing irregularities of profile, -- teaping?
extremeS sex ratios, and the like.5 These defects, coupled with the tendency
for variations in sex ratios by age to be associated with the differences
in ethnic character, strongly suggest underenumeration. Therefore, it is
advisable to begin a projection with a careful examination of the base
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 68 -
population as Presented in the 1926 Census and to incorporate necessary
manipulations of the data to correct whatever unacceptable features may
have been revealed. The principal kinds of manipulation that are likely to
be required are allowances for differential underenumeratiou of i'emPle
infants and children and graduation of the distribution by age to eliminate
irregularities likely to have been caused by inadequacies of enumeration.
It is likely that an adjustment to allow for differential underenumeration
of female infants and children will not fully allow for all infant and child
underenumeration and it will often be useful further to allow for under-
enumeration in these age classes on the hypothesis thab completeness of
enumeration is no better than in some other area for which data are available,
such as the United States.
Abnormalities in the trend of the sex ratio of the population by age
? for the adult Population do not necessarily imply faulty enumeration; they
may be unique results of the demographic history of Particular places. Yet,
if it is found that such abnormalities occur only among certain ethnic groups
it is a suspicious sign of incompleteness of enumeration. In such cases it
would be preferable to adjust the data so that they conform to a more plausible
function of the sex ratio by age, rather than to accept the abnormalities.
Other types of manipulation may be called for in specific instances;
the recognition of the need for adjustment and the formulation of an appropriate
device for adjustment will depend on the virtuosity of the analyst.
Replacement Populations.
This term is used, for want of a better exT,ression, to designate
special population groups for which it would be in-Droner to compute survivors
because of special circumstances. For example, a uriversity may contribute
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
:10
- 69 -
1,000 persons to the population of a city at the base date and at a date many
years later. Xt would be incorrect to assume that the survivors of the base
date population would include the survivors of the 1,000 university students.
The latter would have been replaced by others of equally young age, and there
would have been no aging of this base date 'emulation of university students
to take into account. Other illustrations of so-called replacement populations
include military garrisons; the inmates of certain types of institutions such
as orphan asylums and homes for the aged, and other nonpermanent population
elements which are continually being replaced by others of like age composition
rather than growing older as residents of the area.
Because of the urban location of Soviet military formations in 1926
and the underdeveloped state of Soviet higher education at this time, military
Personnel and their dependents are apt to be the most conspicuous of these
replacement groups. For large cities, the distribution by age and sex of
military personnel and their dependents is given in the 1926 Census volumes
which deal with occupations. In the case of smaller cities, for which detailed
occupationpl distributions are not available, the military population may some-
tines be estimated satisfactorily by Ri) ott4ng to the city in question BOMB
part of the known military formations reported for a larger admjnistrative area. --
This has sometimes been accomplished through an analysis of sex ratios leading
to an estimate of excess males who are assumed to be military personne1.6 The
military population that has been removed must be replaced as a final step in
the projection, and in this connection one is forced to depend almost completely --
upon order of battle materials assembled by military intelligence. Since order
of battle information refers only to troops without their dependents; an
estimate of the combined number of military and dependents must be made by
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 CIA-RDP81-01043R002000060008-1
Jt,
- 70 -
assuming some relationship between these two groups. This relationship has
often been computed from the 1926 Census data for the area in question,
since better figures were not at hand.
Modification of survival rates.
Life tables of recent date are not available for the Soviet Union.
For the period around 1926 complete life tables are available for the European
part of the R.S.F.S.R., for the F.epdblics of Byelorussia and the Ukraine, and
for certain other subregions of the U.S.S.R. Moreover, from handbooks published
by local statistical agencies throughout the Soviet Union in this period, age-
specific death rates are provided from which either abridged or complete life
tables can be constructed. In general, great refinements in life table construction
are not warranted, since variations in survival rates are among the less signifi-
cant sources of error with which one must contend. However, in view of the
very long projection period, and in view of the calamitous events of World War
II, certain modifications of the survival rates computed from life tables for
the period around 1926 must be considered. Accordingly, depending upon the
region involved, survival rates have sometimes been altered to allow for
improved mortality conditions after 1926. This has been done by assuming a
constant annual percentage decline in the mortality rate of each age-sex
group throughout the 23-year projection period. To implement this assumption
a set of multipliers were developed which were applied to the complements of
the age- and sex-specific survival rates. These multipliers represent average
improvement in age-sex specific survival rates as revealed by Polish life
tables for 1931-32 and 1948, by the slopes of curves representing trends in
age-specific mortality presented by Notestein, et.a1.7 and by U. S. experience
from 1910 to 1930.8 Whether or not improvement in mortality is to be assumed
will depend upon an individual evaluation of conditions in the area for which
the projection is to be made.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
?
,78r
- 71 -
Survival rates have been modified also to take account of war
losses? especially among males serving in the armed forces in World War XI.
The procedure adopted here requires the estimation of the age pattern of
war losses among males in the Soviet Union and incorporates these directly
into the survival rates. Two assumptions are involved in this method:
Cl) that the age pattern of war losses for the U.S.S.R. is applicable to any
subarea and (2) that the losses to urban areas are not erased by compensatory
rural to urban migration. The reasons for the latter assumption will become
clear as the entire projection procedure is reviewed. It represents, however,
one of the prinary defects of the method in its present foruL
In addition to problems raised by these assumptions, the data from
which estimates of war losses are made are far from ideal. Questions exist
both as to total number of casualties and to their distribution by age. Seven
million war casualties have been admitted by the Soviets2 but it is generally
felt that this figure does not include all military deaths that are directly
attributable to the mar. It is likely that due to the invasion of the Soviet
Union and the resulting administrative confusion, the Soviets themselves do
not have completely adequate records of their war losses .9 Several estimated
age distributions of war losses are available and are presented below expressed
in absolute numbers and a percentage of each cohort affected (age as of 190).
-V
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R00200006000s_1
Declassified in Part- Sanitized CopyAp roved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
-72-
Table 9.--ESTIMATED U.S.S.R. MALE MILITARY LOSSES' IN WORLD WAR II
(Figures in thousands)
Age in 1950
Pattern 12
4
Pattern 113 Pattern III.
Total
9,619
Losses
9 619
9 619
20 to 24 ye-rs
502
276
514
25 to 29 years
1,579
1,982
2,143
30 to 34 years
2,776
2,583
2:830
35 to 39 years
2:289
2,797
2:118
40 to 44 years....
1:515
1,483
1,304
45 to 49 years
701
256
489
50 to 54 years.. 4000000001,0
218
166
221
55 to 59 years.... ........ .
39
61
...
60 to 64 years.. .....
004
13
040
Percent of age group5
20 to 24 years
4.4
2.4
4.5
25 to 29 years
17.9
22.5
24.3
30 to 34 years ..... 00400000
35.9
33.4
36.6
35 to 39 years
26.8
32.7
24.8
40 to 44 years
21.7
21.2
18.7
45 to 49 years
12.6
4.6
8.8
50 to 54 years.... ? ? 00
5.7
4.4
5.8
55 to 59 years. 00.400 ..... 0
1.4
2.1
...
60 to 64 years...
...
0.5
...
1 Each distribution has been adjusted to total 9,619,000. See -
Intelligence Research Project no. 2898: Military Intelligence Service -
WDGS? The Effect of War on Soviet Population and Manpower, 4 March 1946.
2 BUFF-OF?Iestimeyar ReseaFEEMEElon: Library of
Congress Social Relations Technical Paper No. I-3: An Estimate of the
Developments in USSR Population Structure from JanuUrr.":777.75?-76?
aTEX-1.,p 192.?
Intelligence Research Project no. 2898: op. cit.
4 Lorimer: Frank) The Population of the Soviet Eion, League of
Nations, Geneva, 1945, pp.
5 Base population estimated by Lorimer, op. cit., DA. 254-255, without
adjustment of age cohorts for wax losses, ba-IEFieased uniformly by the
ratio of the postannexation to the preannexation population.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-
,-
-73-
The differences between these war loss patterns are slight, especially in
view of the widely different bases from which they are derived. Pattern I
was developed from data on German war losses in World War X with some
consideration of fragmentary data of French and German losses in World
War II. Pattern III is admittedly an arbitrary approximation. Pattern II
on the other liana was based upon a careful analysis of World War II data
relating to the Soviet cohort mobilization schedule; ages of Soviet PW's
captured by the Germans and passbooks taken from Soviet casualties. Inasmuch
as there was universal conscription and no significant deferment policy, it
was possible to estimate the period of exposure for each mobilized age class,
the latter, taken together with estimates of the chances of survival; provided
estimates of war losses for each age (mobilization) class. The chance of
survival factor is not a linear function of time, but is based upon reported
rates of kill (more or less constant) and capture (highly variable).
Survivors of births.
The second important component of a population distribution by age
and sex, derived by the component Projection method, is the survivors of
births which occurred between the base date and the estimate date. If data
were available, the straightforward solution would be to subtract actual
mortality by age or to apply suitable life table survival rates to reported
or estimated annual numbers of births during the estimation period. In actual
Practice this solution of the problem is rarely possible, principally because
-reliable data on births are rarely available for Soviet cities. There are
scattered crude birth rates for the U.S.S.R. and some of its major cities for
dates Prior to 1939? and of COUTE3, the U.S.S.R. series for 1950-1955 recently
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 74 -
published in Narodnoye khozyaystvo bue, in gelteral: sufficient data on Yrthe
to be used in this approach are simply not to De found. Moreover, there
are legitimate doubts about the complete:le:so of vital reEistvation which
render suspicious even the few published figure,i on births.
There aver ho,Tevers other types ce' Ja;a that may be used, One of
these is information for the late 'twentier; oe. ae-specific fertiljty rates,
which is available for a small number of CiViaS. Xt is possible to pToject
these rates by analogy to trends observed in edmilar rates for othef countries,
and to apply the projected rates to estimates of the female populations by
ages during the years following the base date. The results would be estimates
of the annual number of births4, which could then be used together with
appropriate life table survival rates to obtain estimates of survivors of
births at the end of the estimation period. There are several objections to
this procedure? however:
1. Age-specific fertility rates are available for relatively few cities.
2, The few rates which are 'Available me quite on of date and must be
projected in accordance with trends observed elseltheres whi?th may or
may not adequately represent the situation in particular Soviet eities.-
3. The method requires the computation of annual female age distributions
a very laborious computation process.
4, There is some circularity of ressouillg ingolved in the methods for the .
ermual female age distr4butions required depend to some extent on the
voss number of births. The cumulative effect of this circularity can
be very damaging to estimates for cities which have grown considerably
by migratiaa. The reason for this is that the annual female age
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
4.
'tf27,1a,it47.,
MM.
? distributions must include in-migrant females and exclude out-migrant
females. Yet, as will be seen below, the estimate of net migration
for the entire period of estimation depends in some measure on the
number of births during the period.
5. The component projection method calls for the estimation of the net
number of survivors of migrants over the entire period. This number
would have to be distributed into year of migration cohorts; and each
such cohort revived," that is augmented by an estiate of deaths to
the cohort) to each year for which a female age distribution is
required. This is a very time-consuming Process; and is fraught with
many risks of error that are entirely beyond control.
? Still another approach to the estimation of survivors of births is
by the use of child-woman ratios. These can be computed for every city from
1926 Census data and can be projected without excessive labor. Although not
an ideal solutionrit has the advantages of simplicity of application;
emoloyment of basic data which reflect local conditions, availability of
basic data for all cities, and avoidance of circular reasoning.11
The steps followed in the use of the child-woman ratio technique
are
1. The ratio of children to women aged 20-54 years is computed from
data for the very city being studied) using 1926 Census information.
The term children refers to persons whose age is less than the
number of years between the base date and the estimate date.
2. The corresponding' ratio is then computed for the U.S.S.R. as a whole,
for 1926, 1939; and 1950.12 As will be Seen below, it is desirable
? to accomplish the projection by stages, namely from 1926 to 1939,
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
, ?Ik
?
and from 1939 to 195). Xi; happens that 1950 was the terraimal date
for the estimates of city population by age and sex prepared by tho
U.S. Bureau of the Census. The terminal date could, however; have bem
selected as some other year; but this is immaterial to the description
of the method.
3. It is assumed that the child-woman ratio for the city being studied
changed with time in proportion to the change in the corresponding
ratio for the U.S.S.R. as a whole. The projected child-woman ratio
is then applied to the number of survivors of the base date population
to yield an estimate of surviving offspring of this Population on the
terminal date. It should be noted that this estimate does not include
all children of the age-group in question. The surviving offspring
or migrants must be separately calculated.
Nature of the migration component.
Once estimates of the numbers of survivors of the base date Population
and of births have been computed, they may be used together with estimates of
total population to obtain a measure of migration. Thus, if Po represents
.
the total population at the base date, S the number included in Po who are
expected to survive to the estimate date, PI the total Population on the
estimate date? B the survivors of offspring of the base date population; and
M a residual, we find:
Pi 2 S B M.
The residual, MI, may be readily computed from the values of the other terms
of the equation; which have already been computed. What, precisely does M
represent?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 77 -
In computing the S component of the above equation, it has been
assumed -- in effect -- that no individual included in the base population
moved away. They either died or remained alive in the area to be included
in S. In actual fact, this is a most unlikely situation. As a result, S
overstates the number of survivors of the base population adli in the area
on the estimate date by an amount equal to the number of out-migrants
minus the expected number of deaths among out-migrants, assuming the same
age- and sex-specific death rates and the same percentage distribution of
population by age and sex for both nonmigrants and out-migrants. Similarly,
the B component of the above equation is an overstatement of births among
the base population, because it allows for births among out-migrants. The
residual M, therefore; is an understatement of the net change in population
attributable to migration. The amount of this understatement is precisely
the same as the overstatement in the S and B components, namely: out-migrants
minus expected deaths among out-migrants Plus expected births among out-
migrants, or more simply, the expected net survivors and surviving offspring
of out-migrants.
It appears, therefore, that the component projection method
designedly overestimates two components and underestimates the third. The
residual M does not represent the true net change in population attributable
to migration; its true meaning is the net survivors and surviving offspring
of in-migrants minus the net survivors and surviving offspring of out-migrants.
To take a simple illustration, suppose that a city had a total
Population of 100,000; a constant annual death rate of 20 per thousand, and
???????,.,1,
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
-- 78 -
a constant annual birth rate of 30 per thousand. Assume also that 10,000
people per year moved away, that 12,000 per year in-mij,ratedv and that
these had the same crude birth and death rates as the nonmigrant population.
Tn general; if P re-Presents the population at the beginning of a year, B
the births, b the birth rate, D the deaths, d the death rate, and M the
migration, we have:
13 =
and
2b
d b
2d
- 1/2 M)
(P - 1/2 M).
2 4- d b
Applying these formulae to the data of this illustration, we find that after
5 years the original Population of 100;000 has shrunk to 53,800, including
the surviving offspring of these inhabjtants. On the other hand, the annual
input of 12000 in-migrants results in a total of 61;500 persons who moved
into the area since the base date and their surviving offspring. Thus, in a
total population of 115,300; only 47 percent are survivors or descendents
of the base population.
Had the population change in this illustration been measured by a
component Projection technique, with the assumptions of no out-migration
from the base population and an annual net in-migration of 2;000 per year,
there would be after 5 years about 105,000 survivors or descendents of the
base population and 10;300 survivors or descendents of net in-migrants. Again;
the total population is 115,300,, but in this case 91 percent represents
survivors or descendents of the base population.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 79 -
This illustration exaggerates somewhat the nature of the difficulty
created by the use of a component projection technique. If it had been
assumed that a portion of the annual out-migration of 10;000 bad come from
among in-migrants in years following the base date; the discrepancy noted
above would have been smaller) but still appreciable.
The significance of this characteristic of the component projection
method is that, although it does not affect adversely estimates of net
survivors and surviving offspring of migrants, it does have a profound effect
on the estimated composition of the population by age and sex, Particularly
when the distribution by age and sex of nonmigrants is much different from
that of in-migrants. Unfortunately, experience indicates that this condition
usually prevails.
Inasmuch as the characteristics of migrants have so important an
effect on the composition of the population by age and sex, and, inasmuch
as the data employed are for years past, a brief historical outline of urban
migration in the U.S.S.R. provides a desirable background against which to
assess the quality of such estimates.
Historical outline of U.S.S.R. urban migration.
For all Practical purposes internal migration and particularly urban-
ward migration in the U.S.S.R. assumes greater importance as the date of
migration apProaches the Present for tha proportion of surviving migrants
declines as the duration of their residence at their destination grows longer.
Yet even the migratory movements of the Pre-revolutionary era have some interest
because these are reflected in the data of the 1926 Census and are useful in
analysis for subsequent Periods. This brief surVey begins therefore with 18972
the date of the first Russian census.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Ap roved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 80 .
Phase 1: 1897-1914. Between the time of the first Russian
census in 1897 and the levolution there had been a steady increase
in the urban population, compounded almost equally of natural
increase and rural to urban migration. During this 16-year period
approximately four million persons settled in urban areas. Virtu-
ally nothing is known about the characteristics of this migration
since at this time the government kept no migration statistics.
Furthermore, the civil disruptions that occurred toward the end of
this phase precluded a satisfactory count of the population from
which something of the characteristics of migrants could have been
learned.
Phase 2: 1914-1917. During the war years the urban population
increased by some 1.4 million, to reach a total of 26.3 million,
or about the same as in 1926. Again our knowledge of the charac-
teristics of the migrants is lted.
Phase 3: 1917-1920. This period was characterized by an urban
exodus of large proportions. It is the losses of these years that
account for the apparent stagnation of the urban population from
the Revolution to 1926, the year of the first major census of the
Soviet Union. The magnitude of the over-all decline is not known
but its significance is suggested by the case of Leningrad, the
center of much revolutionary activity, in which the population
declined by 580 thousand in 1917, 837 thousand in 1918 and 320 thousand
in 1919; Leningrad changed from a city of over two million to one
of slightly more than one-half million during these few years.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy A proved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 81 -
Phase 4: 1921-1926. The restoration of order augmented by
rural distress brought about a renewal of urban growth based, as ,
in the past, largely upon rural to urban migration. This was the
period of the New Economic Policy (N.E.P.); it was a thee of consolida-
tion and recuperation for the U.S.S.R. and for the most part the result
of the net urban migration was to bring the urban population back to
its pre-revolutionary level.
Phase 5: 1926-1939. This period; bracketed by the two census
years, is one of greatly accelerated urban migration resulting from
the industrial developments of the early Five-Year Plans and the
collectivization drive in agriculture. Urban growth during this
time coincided closely with the development of the industrial and
extra-active centers of the country. The gradation in the rates of
increases by type of city 15 confirms the hypothesis of a general
association between industrialization and urban growth:
Table 10.--MEAN PERCENTAGE INCREASE OF POPULATION, IN
'SELECTED TYPES OF SOVIET CITIES21 1926-1939
Type of city
Number of
cities
Mean percentage
increase
Industrial........... ********
34
350
Transportation..... ...
0
20
189
Trade........................
17
153
Diversified....,.............
9
86
1 Cities of 100,000 or more inhabitants in. 1939.
Declassified in Part - Sanitized Copy A ?proved for Release 2013/07/12 ? CIA-RDP81-01043R007onnnsnnnR
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 82 -
The expellent force of collectivization is shown in figure 4,
which, although based on a limited number of cases, leaves
little doubt as to one effect of collectivization upon the rural
population. Altogether, more than 18 million persons moved citydard
during this period, with low oscillations in frequency at the
beginning of each five-year plan and peaks occurring midway in each
planning period.
Phase 6: 1939-1950. In general pattern, but not in amnlitude,
urban migration during this period resembles that of phases 2, 3,
and 4 taken together. The war-time evacuation of urban centers was
less pronounced than in phase 3, and in addition; the immediate
postwar period was characterized by the Fourth Five-Year Plan, which
was oriented toward national reconstruction and characterized by
economic expansion rather than by the uncertainties and economic
Paralysis of the N.E.F. of phase 4. During phase 6 the urban popula-
tion grew by approximately 10 million Persons largely as the result
of in-migration.
Phase 7: 1950-1955. National reconstruction under Plan IN,
was followed by Plan Vs which was adopted in 1952 but which covered
the years 1951-1956. The industrial emphasis of Plan V resembles
that of the early plans. Replacing the collectivization drive of
the thirties was the continued mechanization of agriculture and a
drsmatic program of kolkhoz consolidation. In this five-year span
the urban population grew by 17 million persons, 9 million of whom
came from rural areas)4
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?
.?
SECRET
FIGURE 4. COLLECTIVIZATION AND OUT-MIGRATION, FOR SELECTED AREAS IN THE U.S.S.R.: 1928-1931
PERCENT
PERCENT
COLLEGTIVIZED4-'
COLLECTIVIZED.'
100100
.
?NORTHERN
CAUCASUS REGION
?LOWER VOLGA
REGION
!
?MIDOLE
?URALS
VOLGA REGION
OBLAST
? BASHKIR
?CENTRAL
A. O.
BLACK
? TATAR ASSR
SOIL REGION
KARELIA?
?WESTERN
OBLAST
?MOSCOW
OBLAST
?LENINGRAD
OBLAST
90 90
80 80
70 70
60 60
50 50
40
30
40
30
20 20
10 10
100 200 300 400 500 600 700 800 900 1000
PERCENT INCREASE IN TOMER OF OUT-MIGRANT5,1925 -1931
1100
SOURCE: 5.5. XLIVANSKIT, 'SEASONAL MIGRATION IN THE U.S.S.R. IN 1928 1929-19317 VOPROST TRUDA, VOL. 10, NO.10, OCT. 1932, pp. 66-73.
CITED THUS IN SOURCE. PRESUMABLE REFERS TO PERCENTAGE OF PEASANT HOUSEHOLDS THAT WERE COLLECTIVIZEO.
U.S. AAAAAAA T OF 60501806
SECRET
1200
1300
1400
1500
u.S.IlusCAu Or THE Maul;
CO.
?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Iihje swt
-84-
A major development in Soviet urban migration has been the gradual
extension of governmental control over the process of industrial migration,
that is migration to meet the labor requirements of the country's expanding
industrial system. Contrary to popular belief the movement of population
under the Soviet regime has been and still is, to an important extent, a
voluntary movement. 1/ Before the advent of the Five-Year Plans it was
IrIplanned as well. The recruiting of industrial labor during the First and
Second Five-Year Plans was left largely to individual enterprises with the
national, regional or local governments having no immediate role. Mobiliza-
tion of the economy and the advent of the war forced the national government
to assume more direct control, which it did by identifying areas of surplus
population, 16 extending financial and other aid to migrants, carrying out
planned evacuations, and in a few instances, by resort to forced deportation.
One gets the impression, however, that urban migration during the decade
prior to World War 11 was less organized than the planned resettlement of
agricultural workers to new agricultural areas. Following the war the 8oviet
government continued to facilitate the flow of population to areas of labor
shortage both through organized recruitment of volunteers and the provision
of incentives. The role of the central government in this area was emphasized
and implemented in 1947 by the establishment of republic administrations for
labor recruitment.
Age and sex distribution of the migration component.
The term "migration component" is used here so as to make it clear
that the component projection method deals with an artificial entity correspond-
ing to the needs of the method rather than with actual migrants. This distinction
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R007nonnRnnnR
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 85 -
has been developed and expounded above in the discussion of "The Nature
of the Migration Component." As explained there, the migration component
is a residual obtained by subtracting an estimate of survivors of the
base population and survivors of their offspring from the total population
at the estimate date. It is not a measure of the number of surviving
in-migrants and surviving offspring of in-migrants. It is a legitimate
criticism of the component projection method that the residual must be
used in lieu of a figure representing surviving in-migrants and their
offspring, because the two figures are generally not the same and because
the age and sex distribution of nonmigrants is usually different from that
of in-migrants. But the component projection method requires that this be
done because too many people have been accounted for in the estimates of
the survivors of the base population and their surviving offspring, In
general, migrants include a higher proportion of males than do nonmigrants,
and also a higher Proportion of young adults. The component Projection
method, therefore; tends to yield a population distribution by age and sex
which understates these two population categories.
Although other methods of estimating migration have been devised,
they are not appropriate for use with the component projection methcd unless
they have incorporated an understatement of the net change in population
attributable to migration which exactly balances the overstatement inherent
in the other two components, i.e.; s vivors of the base Population and
survivors of offspring of the base population. For example, a method has
been devised for use with data for states and cities of the United States
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 86 -
which involves the use of school enrollment data." By a comparison
between expected and actual school enrollment a measure of the percentage
of the school-age population cohort comprising migrants may be obtained
and converted into a measure of the rate of migration for the total popula-
tion. This migration rate is then applied to the base Population plus
half the births to yield an estimate of net migration at all ages. Even
:if the data were available so that this method could be employed for
U.S.S.R. cities, it would be conceptually incorrect to employ the result
in a component projection method, because the result does not have precisely
the inadequacies called for by the method. It would, however, be possible
to modify the procedures used to compute survivors of the base date popula-
tion and of their offspring so that the three components would be conceptually
in accord with one another.
The problem, therefore, is to devise a method for distributing the
residual migration component by age and sex, keeping in mind that its
composition by age and sex must reflect not only the patterns of age and
sex composition peculiar to migrants, and prevailing mortality trends, but
also the cumulative effect of annual increments of migration which more
likely than not resemble each other closely. The meaning of the latter
stipulation may be illustrated simply by considering the ultimate age
distribution of in-migrants to a population which receives a constant annual
input for 10 years. To simplify the ease, let it be assumed that the
distribution by age of each annual cohort of migrants is rectangular between
ages 20 and 29 at time of arrival, and that there is no mortality. The
migrants at the close of the period will be distributed as follows:
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
'
87 -
Percent distribution
Total....
Age at arrival
Ultimate age
100
100
20-24 years....
50
10
25-29 years....... 6o4
50
35
30-34 years....... Q*0
"..
40
35-39 years.......
-
15
Thus; even though migrants may be relatively young on arrival in an area
they do grow olders and successive annual layers of migration produce a
distribution with a markedly different contour than is found in any annual
cohort of migrants. Xf mortality is taken into accounts as it should be; the
ultimate distribution varies somewhat differently, since the distribution
becomes skewed towards the younger ages.
Fortunately, the 3.926 Census of the Soviet Union provides data
on the age and sex distribution of in-migrants by period of arrival, 18 These
incorporate the specific migration experience of each area prior to 19262
the mortality rates then prevailing and the cumulative effect of successive
migratory inputs. Xt is possible, therefore, to compute the 1926 age and sex
distribution of in-migrants during the preceding 12 years, so that this might
be used for migrants arriving during the 12 years between the 1926 and 1939
Censuses. The same can be done for any period of years; as required. In
fact; this was done in the city population studies for the U.S.S.R. conducted
by the Bureau of the Census. Another method, which was used for Rumanian
cities, began at the other extreme. The distribution by age and sex at time
of migration for internal migrants in Czechoslovakia was cumulated over a
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
., ,?.?.
..r
. . -
19-year Period. adjusted for mortality in accordance with age- and sex-
specific death rates applicable to Rumania and the results converted
into a percentage distribution which was applied to the residual migration
component:19
Either of these two approaches will yield distributions appropriate
to survivors of migrants over vary:ng periods of time, but it must be under-
stood that neither affords a Precise representation of the age and sex
distribution of survivors of migrants ieLparticular cities. The first approach
imparts to the migrants of the estimation period the characteristics represent-
ing the net effect of: (1) annual numbers of arriving migrants propoctional
to the migration cohorts of years prior to 1926 for the same city; (2)
distribution of annual migration increments by age and sex proportional to
the distribution for years prior to 1926 fop the same city; and (3) age- and
sex-specific mox-taliV rates among migrants equal to those for years prior
t 6 for the same city. The second approach, that is the one developed in
connec-,ion with Rumanian city estimates, embodies a somewhat different set of
postulates, name),': (1) there are equal annual increments of migrants; (2)
the age and sex distribution of migrants at time of arrival is held to be
constant from year to years and equivalent to a distribution for a relatively
recent date determined for another area; and (3) age- and sex-specific
mortality rates for migrani;$ applicable to tbe estimation period rather than
to a prior period are employed. Each approach has theortical justifications
and each :?rovides a result in keeping with the hypothesis involved, but
neither assures literal reproduction of the true age and sex distribution
involved.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
?
rx,e
-89-
In the application of either of these approaches it is necessary
first to isolate the surviving offspring of migrants. This may be done
most readily by employing the same child-woman ratio used for survivors
of the base population. Some bias will be introduced, for the child-womn
ratio among the migrants is nicely to be higher than among nonmigrants
simply becausc the migrants generally comprise a younger population with
a far higher proportion of its members in the child-bearing ages. A. bias
in the other direction is also introduced because of the tendency of the
residual migration component to overstate the female migrants. The extent
to which these biases counteract each other cannot be determined.
Once the estimated surviving offspring of migrants have been
subtracted from the residual migration component the remainder represents
survivors of in-migrants minus survivors of out-migrants, as explained
above. The distribution by age and sex of this remainder is the next goal
in this process. Two approaches to this goal have been discussed above.
A third approach, based also on the 1926 Census data on surviving in-migrants
by age) calls for the computation of migrant to nonmigrant ratios for each
age and sex group in 1926) and the application of these ratios to the final
estimates of survivors of the base population on the estimate date. The
results are estimetes of surviving in-migrants by age and sex which may be
adjusted linearly to equal the predetermined remainder representing survivors
of in-migrants minus mvvivors of out-migrants. The effect of this procedure
is equivalent to the effect of the assumption of identical distributions by
age and sex for in-migrants and out-migrants. This relationship has not been
adequately explored, but it is likely not to be an identity, at least in the
circumstances operative in Soviet cities.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 90 -
Although the first approach mentioned/ that is the method of
applying to the residual migration component the age and sex distribution
of 1926 surviving in-migrants over a period of equal length; was used in
the Bureau of the Census studies for Soviet cities; it was not used without
a full understanding of the conceptual difficulties involved. An attempt
was made to determine pragmatically the importance of a fundamental caveat
to the methods namely the frailty of the assumption that the pattern of
distribution of migrants by age and sex was the same for pre-1926 migrants
and migrants of the 'thirties and 'forties. It was found that there was
relatively little variation in the composition by age and sex of surviving
migrants to urban areas during years prior to 1926, except for migrants,
arriving during the period of the war and the Revolution." For the period
after 1926 however, there is evidence of some change in the age structure
of U.S.S.R. urban migrants as the table below shows:
Table 11.--AGE DISTRIBUTION OF U.S.S.R. INTERNAL MIGRANTS
FOR SELECTED YEARS: 1926 TO 1934
Age
Year of migration
1926
1932
1933
1934
0-15 yearS?000????000???0?40?00?000
24.7
13.2
15.1
12.9
16-59 years... 20090000000?00?900000
71.7
83.0
80.9
83.7
60 years an.d oyer0?00000..0000000.
3.6
3.8
.1 I-%
?
????le?-0.010.11.
????????????.????????tAki.....
Source: 1926: All-Union Census/ vol. LI; Table IV-A.
1932, 1933, 1934: Besher Ias Problemy naseleniia v narodno-
khoziaystvennomplanirovaniims:_oliOaTion in
NaTIEFL Economic Planning), Moskva, 1937; p. 38.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
,
- 91 -
? Whether these differences are real or due to differences betveen census
and registration procedures is not clear. The greater concentration in
the prime adult ages and the somewhat smaller proportion of persons
0 to 15 years among the migrants of the Ithirties are changes of the type
that one would anticipate from what is known of the declining birth rate
and the nature of the demand for migrants at that time.21
Whether or not there was a real change in the characteristics
of migrants after 1926, there is little question that 1926 migration
patterns are quite inappropriate in many instances since this was a period
of significant change for many Soviet cities. Consideration might be given:
therefore, to generating families of migration patterns appropriate to cities
of a given type. The selection of an appropriate migration pattern then
mould be made on the basis of typological considerations rather than in
terms of the character of the cityls previous migration: which very likely
will have changed with changes in the economic structure of the city.
Another shortcoming of the body of 1926 migration data is its
unavailability for small cities. When dealing with such places, often the
only guide to an appropriate migration pattern is that of the larger adminis-
trative area in which the city is located,: For example: in estimating the
age and sex distribution of migrants to cities of the Uzbek Republic, the
distribution of persons migrating to all urban places within the Uzbek S.S.R.
was used. This procedure creates certain problems in connection with the
treatment of military personnel: since in calculating a migration. pattern
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
g
r 4*.
T
4..
military personnel and their dependents are excluded. Military personnel
and their dependents. however, cannot be divided into migrant and nonmigrant
,
subgroups on the basis of information in the census and therefore, out of
convenience: military personnel and their dependents have been treated as 11:
they were all migrants. Removing all of the military personnel from among
the migrants actually reduces the niimber of bona fide migrants since some
of the military are nonmigrant. This does not create any serious problems
in cases where equal or greater numbers of military dependents are restored
to the population on the terminal date. On the other hand, in cases where
the military have been removed in toto from the migrant distribution and not
restored, deflation in certain age groups will result. This problem is most
likely to be encountered in the case of cities having no military garrison
but for which the migration pattern of a larger administrative area, from
which military forces will generally have been deleted, has been assumed.
In situations of this kind it is recommended that an attempt be made to
distinguish between migrant and nonmigrant military personnel in 1926 and
to eliminate only the former from the civilian migrants.
To summarize, the procedures adopted for estimating the age and
sex characteristics of migrants have been dictated primarily by the character
of available data. In view of Soviet criticisms of their own migration data
available from population registers; it is unlikely that a completely satis-
factory set of statistics could be made available even with a free exchange
of information. Foremost among the problems with which we are forced to
deal is the assumption that the age and sex characteristics of migrants
have not changed substantially since 1926. This assumption is suspects
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R00200006onnR_1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
? ?
- 93 -
particularly for the postwar period, when both the supply as well as the
demand for migrants was affected by heavy, war produced mortality. The
problem is complicated further by the allowances which are made for war
losses in computing survivors of the base population. The Soviet emphasis
upon the development of heavy industry and the concomitant urbanization
of ithe population makes it appear likely that heavy urban male war losses
would be compensated for, to some extent; by higher sex ratios among the
migrants. The present projection procedure, however, takes note of excess
war mortality but not of any compensatory migration of males to cities.
Still another difficulty with the present method for handling migrants
is the assumption that in- and out-migrants have similar age and sex
distributions. There is, in fact, some evidence of differences elsewhere
in the composition of in:- and out-migration flows by age and sex, although
the degree of disparity by age does not appear to be serious.22
Improvement of the procedure for estimating characteristics of
migrants would seem to require a greater understanding, obtained through a
program of comparative research; of the relationship between the character-
istics of migrants and the characteristics of the sending and receiving
communities. More should be learned also of the range of differences in
the age and sex distribution of migrants and of corresponding differences
between in- and out-mi ants. Such research would probably have to be
conducted for areas other than the Soviet Union; where data of high quality
mould be available either from continuous population registers or from
successive enumerations from which the character of migration could be inferred.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
4
.a.v.
..-
'-94-
Projection by stages.
Component projection procedures for the U.S.S.R. and for Soviet
cities will be improved if conducted in two stages; namely 1926-1939 and
1939-x3 where x is the postwar year for which the estimates are desired.
It has already been noted that relatively great weight must be placed on
data from the 1926 Census because detailed figures for subsequent dates
are rarely available for cities. The necessarily lengthy time-span
involved in such a component projection procedure creates risks of estima-
tion which should be cut down where possible. One way to do this is to
introduce data from the 1939 Census of population. In general, the only
useful 1939 figure to he found is the Total population of the city. The
use of this figure provides a measure of control over variation in the
several components of population change, namely mortality, births, and net
migration. Thus, it has been found advantageous to esiimate the 'ageand
sex distribution for particular cities in 1939s and to use that distribution
as a base for estirptes relating to 1950 or other postwar years.
Evaluation of a component projection.
The component projection method used for selected Soviet cities
by the Bureau of the Census is an imperfect instrument. Man3, of the
considerations which lead to this conclusion have been set forth and dis-
cussed above. Yet, the method was adopted, despite potential weaknesses
because of the absence of current information from which a distribution of
the population by age and sex could be fashioned effectively. This very
absence of current information prevents an effective pragmatic validation
2:-
a
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
of the method. It has, however, been possible to employ the component
projection technique to the estimation of the distribution of the =ban
Population of the U.S.S.R., by age and sex, as of the date of the 1939
Census. The fact that this technique led to acceptable results in this
instance is not to be taken as a general validation of the method. It
represents merely one instance in which the method proved useful in actual
practice.
The reports of the 1939 Census indicated a total urban population
of 55,910,000. This was based on returns which were 99.44 percent complete.
On the assumption that this was the true urban population of the U.S.S.R.
in 1939, the same component projection method described above was invoked
to yield, among other estimates," the following figures which were
susceptible to checking against data from other sources:
Age
Estimate
0-7 years
9,911,000
8-17 years
10,479,000
18 years and over
35,520,000
Consider first the estimate of the population 0-7 years. Accord-
ing to a Soviet source124 18 percent of the urban population was 0-7 years
in 1939. Since the percentage figure cited was a rounded figure, this
must be taken to mean more than 17.5 percent but not as much as 18.5 percent.
If the total urban population. was 55,910,000, the population aged 0-7 years
was in the range 9,784,000-10?343,000. The estimate of 9,911,000 produced
by the component projection method falls squarely within this range.
^
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R00200006onnR_1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
- 96 -
Consider next the estimate of the population 18 years old and
over. The number of city voters in 1939 was reported as 30,312,000.25
This figure excluded voters in other urban places. At this time the
Population in cities numbered 47.3 million." Thus, in cities, between
64.02 and 64.15 percent of the population comprised registered voters.
In the country as a whole, however; the ratio of registered voters to
total population averaged only 55 percent. Thus, it is clear that in
rural areas the proportion of registered voters in the population was even
lower than 55 percent. Perhaps the best assumption that may be made for
the proportion of registered voters in the population of other urban
places is that it was the same in other urban places of 10;000 or more
as in the cities; and an intermediate value -- let US say the same as the
national average of 54.85 percent -- in other urban Places of less than
10,000 inhabitants. On this basis, the total number of registered voters
in other urban places may be estimated at from 5;052,000 to 51175?000.
The number of urban registered voters is, accordingly, in the range from
35,364,000 to 35,487:000. Before these figures may be compared with
estimetes derived by the component projection method, they must be put on
a comparable basis. The component projection method employed an estimate
of 55,910;000 urban population of all ages, whereas the estimated number
of registered voters was based on an urban population figure of 562100,000, .
which was not available until 1956. By a proportionate adjustment, the
estimpted number of urban registered voters comparable to the preliminary urban
populationtotal falls in the range from 35,244,000 to 35,367,0000 all of
whom are 18 years old and over. The estimated number of urban registered
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 97 -
voters accounts for from 99.2 to 99.6 percent of the population 18 years
old and over as estimated by the component project5on method. Since
there must have been at least a small number of nonvoters aged 18 years
and over in the urban population, the component projection method appears
to have scored another direct hit, or at ierorst, a very close near miss,
Other assumptions about the number of registered voters in other urban
areas in 1939 would yield substantially the same conclusion.
Finally, consider the estimate of the population aged 8-17 years.
Inasmuch as the total urban population is fixed and the estimates for the
population under 8 and over 17 years accord so well with other evidence;
the estimate for ages 8-17 years must be equally close to reality. Another
view of the matter may be had in terms of the relationship between the
population estimate and the 1939 school enrollment in grades 1-10 in urban
areas, reported officially at 9,298,000.27 Not all of these children fall
into the age group 8-17 years, however. It is reported that some 5 percent
of the entering class in the school year 1939/1940 were only 7 years of
age," and there were undoubtedly some few persons over 17 years who were
still in the 10-year school system. In the urban areas there are likely to
have been even more below-age and above-age children in the schools. Thus,
even though a precise figure cannot be obtained, it is not unreasonable to
interpret these figures as meaning that 9;000,000 or slightly fewer children
aged 8-17 years were enrolled in grades 1-10 in urban areas in 1939. This
would amount to from 85 to 86 Percent of the urban population aged 8-17
years; as estimsted by the component projection method. Is this figure
reasonable? Data for the whole of the U.S.S.R. in 1939 reveal the following
enrollment rates: 29
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
? 98 ?
Age
Rate
Age 0_
Rate
8
......_
76%
12
,.???...
96%
9
93%
13
91%
10
98%
14
80%
11
99%
15
69%
It is a very good assumotioi that the rates for urban areas were even
higher. Under the circumstances, the estir-te of population aged 8-17
years, as computed by the component projection method, seems to be a
good estimate.
One swallow does not make a summer, however; and one successful
application of the comnonent projection method does not unequivocally
demonstrate its general applicability. It may be said that experience
with the method has led to the conviction that under certain circumstances
it yields acceptable results. There are, however, a number of reasons for
developing a substitute procedure. These include:
(1) The length of the Projection interval. There is an inverse
and no doubt accelerating relationship between the accuracy
of a projection and the length of a Projection period. There
is no real hope at the present iame for breaking away from
1926 as a base period.
(2) Certain inherent difficulties in the method. These have
already been touched upon and include especially the Problem
encountered in dealing with war losses and migration.
(3) The inapplicability of the method in eascs ythere data are
lacking either with respect to the composition of the base
population or one or more of the components,
(4) The laboriousness of the method.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
-99-
xi alternate alternate Procedure.
The component projection procedure described above is one device
which may be used to estima;;e the current ago and sex distribution
the population of a city.3? Concern over ilc potential shortcomings has
stimulated considerable thought on the feasibility of other approaches
One such alternative which appears to have promise involves iav-stigation
of the Possibility of ascertainable measurable relationships between Patterns
of economic activity and the distribution of the Population by age and sex
This might be termed a functional as opposed to a longitudinal estimation
_
technique.
In one study) devoted to the estimation and analysis of the
'emulation of Chirchik,31 Uzbek S.S.R.0 the estimated distribution of the
Population by age and sex was derived by the use or an analogy to the Popula-
tion in the industrial centers of the Ku:foes and Oak Ridge) Tennessee.
Implicit in this Procedure was the hypotheois that areas with similar salient
industrial features would have populations with similar profiles.
The functional approach here presented as an alternative to the
component Projection method contemplates a program of systematic research
in the field of urban morphology- rather than the use of hand-picked analogies.
The envisioned end-product of this research would most likely consist of an
urban typology based upon selected economic; social; industrial, end
geographic variables which are found to have the characteristic of being
able to discrim;,,,Ite successfully between differea'c, wes of population
distributions by age and sex. The development of useful typologies of urban
_
neclassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?
- 100
areas would rely on data for cities in all Parts of the world, and would
feature extensive use of correlation and regression analysis.
successful: such research would 112A to the formulation of quantitatively
expressed ranges measuring the association between the variables incor-
porated into the typology and 'eee distibution of population by age anc:
sex. These could then be used for predietiom as -:?211. as for current
estimation.
Although this research appronch is presented here as a potential
alternative to existing demographic eetimatien devices; it may have a
wider usefulness particularly to t'pose users of demographic information
'nose concern is not with demograithy parse, but rather with problems in
the fields of manpower and industrial structure, for the resplutiou of which
demographic data are often used. A typology of urban areas such as that
suggested here nay; if so designed; shcd light directly on these problems.
Research in =tan morphology involves a number of collateral problems such
as auestions of comparability of raw data used to express the significant
variables and questions of the extant to wl-ich findings may be genalealized.
As a long-range investment of research energies; however; the field is
promising. It cannot, however, provide an immediately useful substitute
for more direct methods of analysis of population distribution such as that
described above.
Consideration should be given; of course, to Improving the
component procedure itself. tiorls along these lines will be advanced most
effectively if it is accompanied by research in urban morphology and by
serious inauiries into the problem of the distribution and movement of
population in the U.S.S.R. and its subregions. Studies of this nature would
supply sets of aggregate totals, ranges of variation; and other benchmarks
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 101 -
which will help to gauge the effect and validity of variations in
projection Procedure. Cross fertilization of techniques rather than
intensive concentration on a single approach should be the guiding
Policy of future research and logical inquiry in this area.
Estimation of Ethnic Composition
No thoroughly satisfactory method of measuring the ethnic
composition of small areas in the U.S.S.R., short of a census; has as
yet been devised. This is due in part to the scarcity of timely data on
the subject; in part to conceptual problems in the definition of ethnic
classifications, and in a very practical sense, to the lack of knowledge
of the ethnic characteristics of Soviet internal migrants. This bleak
generalization about the subject of measurement of ethnic composition is
frequently gainsaid by the fact that currently useful information on ethnic
composition is often at hand and need not be derived by estimation. This
seeming paradox merely reflects the fac'i', that ethnic composition changes
very slowly, barring large-scale migratory movements; and in such cases
data for earlier periods -- such as data from the 1926 Census -- reflect the
situation so well that more timely estimates are not required. Whether data
on ethnic composition from the past are used to depict the situation in the
present; or whether crude current estimates are derived, it is still important
to understand what the figures mean and the limitations that must be placed
on interpretations based on such figures.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
ei? 141:;? ?
,-- I
0.1
- 102 -
The concept of ethnic composition.
Ethnic differences within a population are important data for
demographic intelligence since they frequently represent fault or cleavage
lines in the social structure. An ethnic groin, defined as a population
having a common set of customs and traditions (an ethos) as well as some
degree of communal integration, is difficult to identify and cannot always
be distinguished from unorganized strata comprised of individuals having
a linguistic similarity or claiming a common ethnic descent. Yet for an
accurate appraisal of the action potentials of a population; such dis-
criminations are vital. When dealing with Soviet data, however, even the
correct labeling of a stratum may be a difficult task. Nevertheless, two
questions should be raised about any set of figures which purport to
represent ethnic composition: (1) to what extent do the figures depict
an ethnic group rather than a mere collection of individuals homogeneous
with respect to some single attribute? (2) by what criteria -- language;
nationality, religion; descent, residence, etc. -- has the group or stratum
been identified? In isolated, undisturbed; areas, the application of a
single criterion may suffice to identify both a stratum and an ethnic group.
In urban areas, however, the degree of internal integration (or conversely,
assimilation) of an ethnic group cannot ordinarily be measured meaningfully
in terms or a single criterion. Here some consideration must be given to
the degree of consistency afforded by various criteria. For example, a
group homogeneous with respect to both language and religion, and segregated
spatially, has a quite different connotation than does a group which is
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
- 103 -
linguistically similar, of varied religious affiliations and spatially
diffuse. These significant differences mould, of course, be obscured
by a procedure which described ethnic composition -solely in terms of a
language criterion. Similarly, unrevealing in this instance mould be a
treatment of ethnic composition solely in terms of single criteria such
as descent, nationality, or religion. In short, the statistical identi-
fication of an ethnic group -- as apposed to a stratum -- requires a
composite index capable of measuring the extent to which the gr cup has
a unitary organization.
While it is important to recognize the scope of a full analysis
of ethnic composition, a description of the major strata found in a population
is both useful and feasible. One expects, for example, that a city in Central
Asia made up predominantly of indigenous groups will have a different role
in a larger social and economic structure than one, similar in other
respects, but adulterated by Russians and Ukrainians. This is true even
though the identity of these nonindigenous persons, what they do, and how
they regard the indigenous population, may be known only by conjecture.
Applying a minimum standard of value, one is better off with this information
than without it. For a more complete analysis of ethnic composition; however,
an attempt should be made to discover what criterion was used in defining a
stratum. Are these Russians, for example, persons who call themselves
"Russians, n perhaps without legal or genetic support for their claims, are
they persons who habitually speak Russian, or are they persons who consider
themselves to be of Russian descent? And, regardless of how they are defined,
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R0020000ennnR_1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
?Aid r?;.,
-104.-
are these persons culturally assimilated or do they form a solidified
out-group with a variant subculture? It may not be possible to answer
any of these or similar relevant questions. They should be raised,
nevertheless, since this is the range of problems implied by the concept
ethnic composition. Otherwise there is some danger that a stratum
analysis may merely provide a springboard for jumping to conclusions.
Data and problems of definition.
Soviet data available for stratum analysis are severely limited,
although relevant items have been regular features of their census
schedules since 1897. Data from the Censuses of 1926 and 1939 may be
summarized as follows:
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
TERMINOLOGY AND AVAILABLE TABULATION REFERRING TO ETHNIC COMPOSITION:
SOVIET CENSUSES OF 1926 AND 1939
Year 1
Definition
Available
tabulations .
Where available
Remarks
English
Russian
equivalent
19261
?t,,
II
narodnost
r descent
For major administra-
tive divisions by
urban and rural; for
most urban areas.
1926 Soviet Census;
Part I.
,
Determination of narodnost
made by respondents prefer-
ence being given to that of
mother in case of doubt.
Instructions explicitly
exclude consideration of
HITTIEn, residence, or
citizenship.
rodnoy native
yazik language
ditto
ditto
Defined as language habit-
ually used or one which
respondent speaks best.
1939?
Patsional- nationality
nost
i
Distribution for total
population of USSR.
Fragmentary releases
for constituent
republics.
1 Izvestiya April 29;
1940. See also
publications of
, republic aftinis-
, trative and
scientific bodies.
Determination of natsional-
nost a matter of respondentlis
own self-identification.
Data not comparable with
1926 classification by
narodnost.
rodnoy - native
iyazik language
Data not released
..._....4_,e,
1 Not available
1Concept same as in 1926.
1 -.3,-414mmair,u.1.?-CICMI
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 106 -
In addition; there are for 8 republics and 16 ASSR's in 1938/39; data on
the distribution of students according to the language in which they wore
instructed..32 Also for this period and for later years: cartographic
materials" are available which give a nonnuantitative impression of.
changes in the distribution of the peoples (aarody) of the u.s.s.n.
For the purposes of a stratum analysis for any year later than
1926; the data described above are of insufficient detail when the focus
is a given city or urban area. In the face of such a problem; analyses
suggestive of possible developments may be made by adopting -- for a given
city -- certain assumptions about the ethnic composition of its in- and
out-migrants and the ethnic composition of its natural increase since the
base period (1926). Note again that this approach; which is essentially an
abridged component projection pnocedure applied to ethnic rather than age
groups; can only suggest the limits of Possible changes and does not provide
valid estimates of the number; size; and type of ethnic stratum of a given
population aggregate. Standardized Procedures for this purpose do not exist
at present.
It has sometimes been suggested that estimates of ethnic composition
can be made from an analysis of surnames taken from published lists of
political candidates and award winners. The validity of this device depends
upon the representativeness of the process whereby persons on these lists
are selected. Indeed; the a6Fumptioil of unbiased selection is often =warranted
because of the characteristically disproportionate numbers of Party members
and government workers; and the underrepresentation of peasants and industrial
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
-
- 107 -
morkers, in many of these lists, especially lists of political candidates.
There is at present no thoroughly satisfactory method for dealing with
the ethnic composition of small areas. It would appear desirable as a
first step in attempting to break through this problem to learn as much
as possible in Quantitative terms about migration in the Soviet Union
since 1926. Such knowledge, gained possibly from a detailed analysis of
election data for small areas;, would provide a means for approximating
the orders of magnitude of the ethnic shifts which up till now have been
revealed only in nonauantitative terms by comparisons of Soviet carto-
graphic materials.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 10C -
FOOTNOTES
CHAPTER II
1 For example in the Uzbek S.S.R. in 1939, Russians comprised less
than 12 percent of the population of the Republic but more than one-third
of the urban population. This disproportion is much more pronounced if
only the industrialized cities of Uzbekistan are considered.
2 Certainences are trivial such as those between Slays
and Greek Orthodox Finno-Ugrians in North Russia. The suggestion should
be resisted therefore that the best ethnic analysis is necessarily the
most detailed.
3 For planning and administrative purposes the following age gromirgs
have been advocated by some Soviet writers:
0 - 3 ("infants")
4 - 7 ("preschool")
8 - 17 ("school age")
18 - 54 - Female ("adults")
18 - 59 - Male ("adults")
55 and over - Female ("elderly")
60 and over - Male ("elderly")
4 Reference is to the All-Union Census of 1926. A city census conducted
in 1931 provides age and sex data for the largest cities (Moscow and
Leningrad) and for the urban population of the U.S.S.R.
5 For analysis of this type see Babynin: B., "Perspektivy rosta naseleniya
SSSR v 1927/28 - 1932/33" ("Prospects for the Growth of the Population of
the U.S.S.R. 1927/28 - 1932/23"), Planovoye khozyaystvo, v. 9-10, 1928,
pp. 320-338; Ptoukha, M. "La Population de l'Ukrainnifsqu'eh 1960" ("Popu-
lation of the Ukraine until 1960"), Bulletin de 1' Institut International de
Statistique, V. In': part 32 1930, Appendix X, up. 7749; ana LOTIEler; Frank,
The Population of the Soviet Union; League of Natiohs, Geneva, 19452 Appendix 6,
p77.117037-These authors-ai7g-ia agreement as to the fact of defects of
enumeration, but differ as to their diagnosis and recommendations. In general,
the disclosures of Babynin and Ptoukha relative to the enumeration of the
youngest cohorts are in line with investigations of the quality of enumeration
carried out in connection with the U. S. .flsuses of 1940 and 1950.
." For an illustration of this procedure, see: U. S. Bureau of the Census, _
CONFIDENTIAL Estimated Population and Labor Force of the City and Metropol-
itan Area of Andizhan: 1 ,)anuary 19 Series -9.) No. Augii.R-7937-1-p
SECRET.
7 Notestein, Frank et. al., The Future Population of Europe and the
Soviet Union, Geneva:, Columbia University Press: 1544.
6-F6Y-ftlfther discussion of the modification procedure, see U. S. Bureau -
of the Census:, CONFIDENTIAL, Estimates of the Population and Labor Force of
Selected Cities of the U.S.S.R.TOFIFET:9-57-R317-187-232 10 December 1953,
pp. 60-63. SRTO
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 109 -
9 The problem is intensified by the relative decentralization of record
keeping in the Soviet armed forces.
10 It would appear in the light of recent cohort fertility analysis
that even schedules of age-specific fertility rates adjusted for marital
status, for fecundity, and parity are less than ideal for such purposes.
The deficiencies of such rates will depend upon the extent to which con-
ditions influencing fertility change and the extent to which efficient
techniques of family limitation are available to the population. The more
underdeveloped the area, therefore, the more likely that simple age-specific
rates will be adequate as indicators of fertility.
11 The Bureau of the Census in its work on Soviet cities has experimented
with more complicated methods of estimating births. The procedure in brief
involves (1) an estimate of the annual number of births to survivors of the
base population obtained by the use of age-specific fertility rates reported
for Moscow in 1926 and (2) an estimate of the annual number of births to
migrants based upon a crude birth rate for migrants developed from the
application of age-specific birth rates to female migrants arriving in a
given city between 1921 and 1926. Both series of annual estimates of births
are adjusted by the relationship in 1926 between the crude birth rate of
Moscow and the crude birth rate of the city in question. A full discussion
of this procedure is contained in U. S. Bureau of the Census, CONFIDENTIAL,
Estimates of the Population and Labor Force of Selected Cities of the U.S.S.R.,
1 January 106, Serigg-P-95,-Nos. ]2-2, 10 December 1953, pp. 65:0Ta2a1ZT:
This methods abandoned because it was excessively laborious, because the
adoption of age-specific birth rates of another city, in this case Moscow,
failed to take account adequately of local influences on fertility, and because
s' ler methods appear to be equally valid.
2 To compute the child-woman ratio for the U.S.S.R. in 1939 and 1950
it is necessary to have a distribution of the population of the U.S.S.R. by
age and sex for these dates. One such distribution for 1939 is given in
Lorimer, Frank, The Population of the Soviet Union: History and Prospects,
League of Nations, Geneva, 1 p. lp . A siml az' istribu ion for 1310;
which has been used extensively in the Bureau of the Census work on estimates
of Soviet city population by age and sex is that of Dr. Eugene Kuliecher. See
Social Relations Section, Air Research Division; Library of Congress, An Estimate
of the Developments in U.S.S.R, Population Structure from January 17, Tto
January 1, 19-527755FIg: elations lechnicai Paper o. un ak; 2 p. 43,
SECRET. This distribution was prepared before the release of current Soviet
demographic data in Narodnoyt khozyaystvo, and suffers from not having had the
advantage of these data. Nonetheless it represents an excellent imaginative
reconstruction of the demographic profile of the U.S.S.R. on the basis of
very seenV raw data: It is not recommended for future use, however. The
Bureau of the Census has prepared new population distributions by age and sex
for 1955 and 1956, and projected distributions for 1960; 1961, 1965; 1966;
1970, 1971s 1975 and 1976, which embody information not available Prior to
1956. These distributions and projections will be published in the near future.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R0020onrAnnnR
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 110
13 The typology employed here is that developed by Chauncey Harris in
his article on "The Cities of the Soviet Union," Geographical Review;
v. XXXV, no. 1, January 1945: pp. 107-121.
14 Voprosy ekonamiki, v. 1, January 1955.
15 Throughout the period of Soviet power there have been, of course,
planned deportations of dissident and suspect groups, but such deportations
generally removed the unfortunate victims from urban areas rather than the -
reverse.
16 The Third Five-Year Plan listed 10 areas of the U.S.S.R. with nearly
two million "able-bodied collective farmers who can be utilized for work in
nonagricultural branches of the national economy and for migration" from
which it was planned to call up yearly more than a half million workers in
"an organized recruitment." See Sonin, M.?"Voprosy Pereseleniia v tretei
piatiletke" ("Migration Problems of the Third Five-Year Plan"), Problemy
ekonamiki, no. 3, 1940.
?1171.?awrence, N., and Greenberg, B., fl Illustrative :examples of Two Methods
of Estimating the Current Population of Counties," in U. S. Department of
Commerce: Bureau of the Census: Current Population ReporL,22 Population
Estimates: Series P-25, No. 20; REY-671M, Wasningten: D. C.
on migrants to Soviet cities, classified by age: sex, resident
status and time of arrival are available in part III, table 4 of the All-Union
Census of 1926.
19 U. S. Department of Commerce: Bureau of the Census, CONFIDENTIAL,
Estimated Population and Labor Force of Rumania, Urban and Rural: 1 January
1950: Series P-95, No. .39, December 1955, Pp. I-3 to 1-381 SECRET. Released
as Technical Memorandum 0ERL-TM-55-21: Officer Education Research Laboratory,
Air Force Personnel and Training Research Center, Air Research and Development
Command, Maxwell Air Force Base: Alabama.
20 A test of this involved the application of revival factors obtained
from a 1926 life table to the survivors of migrants reported in the 1926 Census,_
The validity of this procedure rests upon the untested and possibly incorrect
assumption of no differential age-sex selectivity of out-migrants as compared
with in-migrants.
21 Sonia: M., "Voprosy pereseleniia n tretei piatiletke," Problemy ekonomiki,
no. 3: 1940, pp. 80-90. "Migration of the first and second Fie-Year Plans was
primarily an industrial migration." In connection with such migration a demand.
for unattached adults might be expected.
22 See Thomas, Dorothy Swaine, Research Memorandum on Migration Differentials,
Social Science Research Council: New York, pp. 11-6().
23 The derivation of this estimated distribution will be presented in a
sUbseauent report, along with an analysis of certain implications it has for _
such problems as the character of urban migrations: the territorial distribution
of military personnel and disenfranchised elements of the population.
24 Marzeyev, A. N., Kommunal'maya gigiyena (Communal Hygiene): Moscow, 1951, _
p.42.
25 Izvestiya, 5 January 1940.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
26 Narodnoye khozyeystvo, op. cit., p. 26.
27 Walrafnoye uprOI046-ilifEffno khozyeystvennogo ucheta gosplana SSSR,
Kul'turnoye stroitelAstvo SSSR (Cultural Construction of the U.S Moscow,
1q494 P.
'u Bogdanov, X. M., Ocherki pc, statistike vsoobshchego shkol,nogo
obucheniya (Outline of StatielTeToriera?lo-c", hoscow, 1948,
p.8.
29 Bogdanov, op. cit., p. 84.
3? A modifiedure? which might be called a "controlled component
projection" has been used experimentally. See: U. S. Department of Commerce,
Bureau of the Census, CONFIDENTIAL, Estimated Population and Economically
Active Population of the City and MeIrTegartf.ZtaT'f-ay.f950,
Series 1.)-9, 2s4 Washington, D. C77-10-K57.75547-ir-T"ppeffa-UTUORTIOENTIAL;
and U. S. Department of Commerce, Bureau of the Census: CONFIDENTIAL, Estimated
Population and Labor Force of the Ordzhonikidze Metropolitan Area: 1 January
Ig51:578eries io. ?: as on, . .? epfa60-1554? pp. 116,MVIDENTIAL. The modified method uses information on population subgroups
such as school children and voters to improve the alignment of cohorts derived
by a component projection techniaue. Data on other population elements such
as *workers and employees," pioneer and komsomol members, etc., may also be
heWul in such an approach.
U. S. Department of Commerce, Bureau of the Census, ,CONFIDENTIAL,
Estimated Population and Labor Force of the City and Metropolitan Area of
EhTiTailuau-IotTeewarizg3711F.-N;luaglieoz; D.
? p. EMIET. See also: U. S. Department of Commerce,
Bureau of the Census, CONFIDENTIAL, Estimated Population and Economically
Active Population of the City and Mer-i7Fifia-377373W.an.y-1950?
Series 1)-9? No. 221,1ashington, D. 0., 10 May 193T7,07.--4g- p CO )
where the assumption is made that the age and sex distribution of Tula in
1950 would resemble that of 1939 within the limits set by the general
demographic situation in 1950. This Procedure was selected because of the
similarity of the economic base in 1939 and 1950, and because the estimated
age and sex distribution for 1939 was based upon a greater amount of
information than was available for any subsequent year.
32 Tsentral,noye upravleniye narodno-khozyaystvennogo ucheta gosplana SSSR,
Rul'turnoye stroiteltstvo SSSR. Statisticheskiy sbornik (Cultural Construction
lommilauion s ?scow, a e
oiaJy vaiva e in s regsTa-EFe the plates prepared under the
auspices of the MVD by Glavnoye upravleniye geodezii ? kortegrafii entitled
Kerte narodov SSSR (Maps of the Peoples of the U.S.S.R.), Moscow, 1953.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 112 -
MAYER XII
ESTIMATES OF THE SIZE AND INDUSTRIAL COMPOSITION OF THE IABOR FORCE FOR
CITIES or EHE Uc&S.R.
This chapter is devoted to problems of labor force definition insofar
as they relate to the development of efficient estimation techniques, to an
exposition of the types of data used in such techniques, and to the description
of the important details of the methodology that has been used by the Bureau of
the Census to estimate the labor force of Soviet cities and its industrial
composition. Little attention 1113.1 be given to alternative methods or to the
evaluation of the quality of the basic data used in this type of study. It is
planned, however; to incorporate a detailed evaluation of the weaknesses and
potentialities of all available Soviet data on this subject into a subsequent
report in this series (tentative title: Soviet Practise in Collection and Publi-
cation of Data on the Labor Force). An additional reasen for the relative brevity
of this chapter is that much of the work done in this area has been of a custom
character? rather than in accordance with standardized procedures; as a result
the emphasis has been placed on the smaller body of topics more or less common
to a nnmber of different procedures and not to any great extent on the custom
variations to vhic.b. resort was very Eminently taken.
Definition of terms_
Broadly defined; the term ulabor force' refers to that part of the
population engaged in economic pursuits. It comprises all persons at work or
bolding a job ns well as the unemployed who are seeking work.1 This is the rela-
tively unrestrictive definition that should be used in connection with the estimates
of the labor force of Soviet cities prepared by the Bureau of the Census and
presented in this series of reports. The term; or any near equivalent; is rarely
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
...,,????? ""-oncexpemp......???.
found in current Soviet literature, perhaps because of an unwillingness to concede,
on ideological grounds, that the Soviet system can breed unemployment other than
fictional unemployment. Most often, Soviet sources refer merely to the employed
population.
If there were truly no unemployment, the terms -employment" and
"labor force" T,7ould be close in meaning. But, it must also be noted that the term
"employed" refers to information collected at no place of mark, as in a census
of industrial establishments, and is a measure of economic activity; whereas the
term "labor force" has the connotation or a personal characteristic so that a person
in the labor force has this as an attribute in much the same way that he has the
attribute of being married, or being of a certain age, This is an operational
distinction. It follows from the different methods of collecting information; and
does not necessarily imply a practical difference in oveall estimates, although
it is usually true that the two typos of measures of the same population yield
different results.
In the Past, the Soviet literature has used different terms to identify
the population engaged in economic pursuits. The 1926 Census, which is of particular
interest as the source of much useful information for estimation purposes, featured
the term "active population." The definition used was: "... persons who have an
independent source of the means of existence (mages, income). Also included as
actives are dependents of State and public institutions, unemployed, and military
Personnel."2 Except for the inclusion of certain elements which mould be more
properly classified as other than active, this definition is close in meaning to
the customary meaning of the term "gainful worker ." Both relate to persons who
follow an occupation in which they earn money or a money equivalent, or in which
they assist in the Production or goods and services. Both automatically exclude
very young children, regardless of their actual situation.3 Both have the conno-
tation of habitual or usual status, rather than status at a particular point in
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 114 -
tIrio or during a specified time interval, Both terms, when employed for 1.1e
collection of inforimtion, re:;u1.06 in op:ratio:ial p-cceGes Thci tended to
include retired or permanently disabled persons who had not worked or sought work
?01 iont: periods of time,
Cr the other hand, there were some meaningful differences between the
Soviet term "active pooulationn an 1-Lc wee 0T.1eiyur;ed 'tgeThful vorker.'t For
eyample, the Soviet teen included all unee.ployed end 7t iF noted Teeifically
that "students who have finished a course and =actives who have reached a working
age" were deemed to be une:Lployed unless aetuall. at eorL.
Suet' persons were treated
as gainful workers only if they so regarded thense1ve:1 and so reported themselves
in a census featuring the gainful worker concept. The "labor force" concept also
covers DW workers, but only if they seek work. Similarly, the Soviet "active
populaticDP concept included as active such groups as ndependents of State and
public institutions whom the 8tate keeps either in compensation for previous labor
and servies for the State -Invalje.i.s of max and labor), or in expectation of services
from them in the future (sc)olars, pre-school Persons), or as a necessity (prisoners).
,
.... /Also counted as active are 7...0 plivings living on their own nonlabor income
from the payment of rent or Iron interest on securities and deposits the
declaesed groups of the population (sorcerers, thieves, prostitutes) rand/
verz:ons living on private charities and ,... on a1inony."3 None of these
groups, except perhaps those engaged in illegal occupations: are included in the
term "gainful workers" nor in the term "labor force", either in the broad sense in
which it is used in thr-se reports or in the less comprehensive sense in which it
is applied to data for the United States.
The term "labor force," as it is now used in connection with data for
the United States has an operational definition, that is, a definition arising
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 115 -
from the way information on the subject is collected. Xt comprehends both the -
employed and the unemployed, but with the following particular menflings of these
terms:
h loyedg persons at work, that is, persons who did any work for
pay or profit or who worked without pay for 15 hours or more during a
specified meek on a family farm or business; and persons with a job but
not at work, that is; those who did not work or look for work during a
specified meek; but had a job or business from mhich they were temporaril
absent because of vacation; illness, industrial dispute; had weather, or
layoff with definite instructions to return to work within 30 days of
layoff; and persons having new jobs to which they were scheduled to
report within 30 days0
gasoployed: persons who did no work at all during a specified week
and who were seeking work; and persons not working who would have been
looking for work except that they were temporarily ill, or expected to
return to a job from which they had been laid off for an indefinite
pericd: or believed no work was available in their line of work or in
their community.
Xt should be clear from what has been said that the labor force estimates
for Soviet cities prepared by the Bureau of the Census are not necessarily comparable
with I! labor force a figures for cities of the United States, nor with nemployment
data from Soviet sources, DOI' even with data from the 1926 Census of the U.S.S.R.
on the economically active population. The estimates whose preparation is described
belay are related conceptually to the data on which they are based and may not
be entirely comparable even among themselves because of variations in method end.
types of basic data employed. In general they come closest to the meaning memployedm
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
as used in the Soviet sources, and they approach the meaning mlabor in
the unrestrictive sense in which the term was first used above; when unemployment
is virtually absent.
The scope of the problem.
In broadest terms the subject of this chapter is the preparation of
estimates of total labor force and of the distribution of this total by branch of
industry for any Soviet city. This job is not of uniform difficulty due to the
fact that employment in certain types of activities is more variable than in
others .7 The greatest absolute variation is encountered in manufacturing, making
this significant category the most troublesome to estimate. Service and maintenance
activities on the other hand are subject to much smaller absolute variations in
size. To pick an extreme example: in Stalingrad in 1926 over 40 percent of the
male labor force was in manufacturing as compared with less than 18 percent in
the case of Orenburg. Yet the variation between the percentages employed in
specific major irdustry groups in these two cities was only 2.1 for trade and
credit, 0.6 percent for government, and 1.8 percent for construction. Compared
either with the percentage Point variation in. manufacturing (more than 22.0) or
with the expected estimation error commonly met by the analyst of Soviet labor
force data, these differences are small; if not negligible.
Variation in the relative size of various labor force categories
occurs in spite of the fact that overall participation ratios appear to be fairly
stable. The following table shows this stability for selected cities in a number
of countries and for a variety of dates:
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R0020onrAnnnR
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
mow --.11egaWanimmsiliss....
- 117 -
Table 12.--TOTAL LABOR FORCE AS A PERCENTAGE OF TOTAL POPULATION FOR
? SELECTED CITIES AND DATES
?
Country
Year
1
Number
cities
Mean
value
(percent)
Standard
deviation
(percent)
? ?
.
United States
1 1940
105
45
2.4
United States
i 1950
105
44
2.8
Austria
1939
4
49
1.7
Bulgaria
1934
3
39
1.7
Czechoslovakia
1930
5
51
1.6
Denmark
1930
2
48
0.8
Ireland
1936
2
44
0.8
England and Wales
1931
51
48
4.6
Finland
1930
4
59
2.1
Germany
1933
i 17
47
3.6
Greece
1928
/ 3
40
1.3
Hungary
1930
I 3
47
1.4
Scotland
1931
4
49
3.0
Spain
1940
I 4
40
0.8
U.S.S.R. (Central Industrial
Region and Ukraine)
1936
I 30
45
8.8
Differences in mean values in the table are to some extent a matter of
differences in labor force definition among the countries listed. The
significant fact is the stability of labor force participation ratios within
countries as indicated by the low standard deviations in the final column.8
What this means is that, to a considerable degree, variations in the
relative size of the manufacturing component are taken up by or squeezed
out of other labor force categories, rather than being a source of
variation in the size of total labor force.
This general approach to labor force structure indicates in a rough
way.s the strategy to be followed in making city labor force estimates: .(1)
concentration upon manufacturing as the unique and variable element in urban
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized CO .y Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 118 -
labor force structures; (2) minimal concern with elaborate refinements in estimating --
nonManufacturing components except where manufacturing has been found to be insig-
nifiaant, and (3) evaluation of (1) and (2) in terms of their approximation, when
combined; to a reasonable overall labor force participation ratio. Since the
latter., in a sense, stands in judgement of the component estimates, it is important
to have an independent estimate of the size of the total labor force. One suggested
possibility is the use of age- and sex-specific labor r.'orce participation ratios
applied to an appropriate age and sex distribution of the population.9 The
decision to approach the pToblem in this m nner depends of course upon the quality
of the estimated age and sex distribution and upon the extent to which data exist
which will permit meaningful sets of rates to be computed. Such rates must obviously
be specific for locality at the very least for urban-rural differences, as well
as for age and sex.
One final comment an technique is called for. Because of the very great
pitfalls and hazards in work of this kind, it has become standard procedure to
work forward from statistical bases twenty to thirty years in the past. Only in
this way can the reasonableness of current claims be safely assessed, Without a
knowledge of antecedent industrial structure; it is most difficult to know how
much confidence to place in figures derived through mechanical manipulation of
fragments or elusive Soviet data. This amolants to saying that where all data
must be treated as partial or distorted disclosures of actual fact, and where
estimating techniques are frequently improvised with insufficient opportunity
for establishing their power and aptness, the reasonableness of the implied leap
from the past is crucialL. Beyond this, for certain sectors of the labor force,
the magnitudes and relationships Observable in earlier structures are the stuff
Declassified in Part - Sanitized Co.y Approved for Release 2013/07/12 ? CIA RDP8i-ninaqpnry,nrInn r,r,r,,
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
? ? ? ? ? ? ? I ?
- 119 -
out of which current estimates are made. For this reason we begin this account
of estimating procedures with a brief review of Soviet labor force data beginning
with the most inclusive of all Soviet compilations, the 1926 Census.
Soviet labor force data.
Reference has already been made to the practical and conceptual
distinctions between the terms "labor force," "gainful workers," and "economically
active population." Some attention should be give). also to an exposition of the
occupational classifications employed in the 1926 Census, since these are so often
used in urban labor force estimation." Both the active and nonactive populations
were classified by occupation. The principal divisions or levels of classification
are as follows:I?
Class or "position in occupation." This is the principal division of
the classification system, and contains 10 categories
1. Workers (roughly, wage earners).
2. Employees (roughly, salaried persons),
3. Persons of free professions (doctors, teachers, etc.).
4. Proprietors with hired labor.
5. Proprietors working with members of families and members of
artels,
6. Individual proprietors.
7. Members of families who assist in the family occupation.
8, Persons not having, or not stating, an occupation.
9. Unemployed.
10. Armed forces.
Branch of the national economy. Within each of the 10 categories
above there is a secondary classification by branch of industry. A
particular principal division may have all, some, or none of thc
categories in this secondary division.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?120-
1. Agriculture.
2. Large-scale industry:,*
3. Handicraft industry.
4. Construction.
5. Railroad transport.
6. Other transport.
7. Trade and credit.
8. Institutions.*
9. Other.
* "Large-scale industry" was defined as Hall industrial
establishments (factories; plants, mines, pits, peat bogs,
etc.) which usually have not less than 16 workers if the
establishment has mechanical power, or not less than 30
workers in the absence of mechanical noweron In general,
the category Hhandicraft industry" contained all establish-
ments which did not meet the qualification for large-scale
industry. The category Hinstitutionsi(uehrezhdeniya)
included persons engaged in governmental activities, as well
as in education, public health, communications, commerce, etc.
Further subclassification of the secondary categories ere available
from the 1926 Census by groups of occupation (metalworkers, wooaworkers? etc.),
by individual occupations (blacksmith, locksmith, patternmaker, etc.), and, for
the worker occupations, by ski11.10
Although the 1926 Census is the source of much more detailed labor
force information than any that have been published since, major problems of
comparability are encountered in using these data serially with later information.
Thus; for example, to compare employment in government, education, health,
communications, or commerce, as reported in the employment survey of 1936, with
the numbers in these categories in 1926 involves a very laborious regrouping
of the occupational data in the 1926 Census. Persons so employed in 1926 were
included without differentiatl ea within the category HinsUtution300 The only
way in. which 1926 uerployment in government, for example, can be estimated, is
by gathering together all .occupations which by their descriptive titles appear
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
-
- 121 -
to belong in that category. Similar problems are met with in attempting
comparisons over time of categories such as steel or textile manufacturing,
food processing, etc. These are the problems which arise inevitably whenever
census data by occupation are compared with figures which have been classified
not by occupation but by industrial affiliation.
One further difficulty to be encountered in using 1926 Census materials
stems from the lack of tabulations for small areas. Detailed distributions of
the labor force are given only for the U.S.S.R., the republics; large economic
regions, each guberniya, aro cities with a population of 100,000 and over.
Abridged distributions are presented for the above areas and for nyezds (rayons),
with data for urban era rural subclassifications. Thus, no distribution is
available for cities which had a-population of less than 100,000 in 1926. An
estimated distribution for cities of this size may sometimes be derived from
the data reported for the urban part of the uyezd to -',fnich the city belonged.
This is done ordinarily by direct proportional allocations, unless there are
items of specific information which indicate a modified procedure.11
the decade following 1926, a great amount of labor force data for
small areas was reported in statistical handbooks published by local statistical
offices and nlanning commissions. These volumes were generally issued in connection
with territorial and administrative reorganisations; or sometimes as information
relevent to the implementation of the Five-Year Plans. For the most Dart,
relevant data were concerned with the size and operation of specific industrial
and cooperative establishments rather than with comprehensive coverage of al/
branches of the economy. For this reason, the figures on individual establishments
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 122 -
published during the early thirties, are useful chiefly in an auxiliary analytic
way, e.g., in thn development or adjustment of industrial productivity ratios.
It was only with the employment survey of 1936 that fairly inclusive figures
on the labor force of a large group of cities became available.12
The 1936 survey provides, for most cities and for all ?blasts of the
U.S.S,R.? the total number of workers and employees (raboebieve and sluzhashchilei
i.e., wage earners and salaried employees) plus a distribution by industrial
category. This survey was based on reports from establishments; agencies,
offices, etc., of the number of persons leen the rollson By official admission;
It is deficient in certain sectors and completely ignores otbfers. The deficiencies
occur in the omission of 'Isom enterprises and agencies of the People's
Commissariat of Defense and the People's Commissariat of Internal Affairs?11
The sectors which were omitted entirely included collective farms and all
Communist Party and Komsomol organizations. Other categories not covered by
the survey comprised the armed forces; the unemployed, individual farmers,
independent artisans and handicmftsmen (neklerirovannakultare), servants, and
peddlers, i.e., persons not on the rolls of regular establishments. By far the
most significant of these excluded groups are the members of the armed forces
and, persons employed by the Commissariat of Defense. There is no standard
procedure for the estimation of this group. Some of the considerations involved
in their approximation are discussed in the appended illustration. (See
appendix
One major difficulty in the USO of data from the 1936 survey is the
lack of tabulations of total employment in manufacturing by branch of industry.
Such a distribution an be achieved for most cities, however, by use of
'
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 123 -
industrial production data (in 1926/1927 rubles) reported in various atlases,
handbooks? etc." Uhen used in conjunction with appropriate reported er
estimated productivity rates, these production figures yield estimates of
employment by bre.nch of industry.
Information relative to the labor force of Soviet cities after 1936
is extremely scarce. It consists largely of a miscellany of newspaper and radio
reports, fragments from encyclopedias, local histories, specialized periodicals,
and the like. -Xn addition, the eyellitneVaa accounts of prisoners of war no
someuhat obsolete) and of visitors to the Soviet Union, official and otherwise,
are available. These are generally difficult to evaluate, and are useful
primarily as background material.
Out of this melnrge it is sometimes possible to gain fairly satisfactory
notions concern the most significant of recent developments. A demonstration
of suchen informational break-through is given with respect to the metallurgidas
industry at Tula in the extended illustration in appendix /V. It is extremely
unlikely, however, that more than a limited number of industrial sectors can
be successfully estimated on the basis of current information in any given case.
The approximation of other sectors - especially such nonmanufacturing activities
as government, health, transportation, and commerce - must depend to a considerable
degree upon some kind of a projection of the labor force for a prewar year. As
indicated above, this most often means the period 1935-1936. These problems are
treated in more detail in the section mbidn follows.
Procedural suggestions.
Nothing has yet been developed that can be called a standardized
procedure -- that is one which mechanically applied, will produce an estimate
of total labor force distributed by branch of industry. As previously Observed,
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R002nnnnAnnnA
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 124 -
each area constitutes a special problem with respect to both available information
and analytical peculiarities. A high proportion of custom mork with sore than
a little artistic fitting is inevitable. Nevertheless, a core or procedurea
has evolved, which: although reauiring adjustrent and aupplementation in
specifie cases, is generally applicable to most problems. An outline of an
estimating nTouedure which would be applicable under ideal conditions of avail-
ability of data is given below. An illustration of its application to a particular
city, Tula, is given in appendix IV.
Selection of base.--In view of the data available the base period
will generally be 1935-1936. A detailed labor force estimate as of that
date may be prepared on the basis of the following types of information:
A. Estimates of the total labor force, azd or the distribution
thereof by major branch of industry may be bad from the 1936 emnloyment
survey212 subject to such incompleteness of coverage as exists in that
source,
B. The distribution by specific industry branch of emplogment
in large-scale industry, which is a close equivalent to manufacturing,
may be accomplished by relating data cn ruble value of output per
worker (available often for major branches of industry fer 1931-1941)
to production in 1935, expressed in rubles. .
C. The estimates derived by stepsil and B frequently are
incomplete or subject to biases which must be taken into account. For
example, the labor force in agriculture, the self-employed population,
and the armed forces are not adequately covered in step A. Data for
these groups must be sought in other places, such as the reports of
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
the 1926 Census and the handbooks and statistical compilations of
the early and mid-thirties.15 The needed figures will rarely appGar
in these sources, but there will usually be additional data on mhich
to base an estimate.
D. The technique noted under step B for the estimation of labor
force in specific industrial branches needs amplification. As set
forth there, it calls for dividing data on production (in rubles) by
data on productivity per worker (in 'rubles). Fihe result of this
computation has one dimension -- workers (pebechiye). This must be
converted to total employment. A reasonable value of the ratio of
total employment to rabechiM is required for each industrial enter-
prise; such figures are obtainable by manipulation of certain published
data."
E. There is also an element of custom fitting that will often
need to be invoked. For example, the estimate of the labor force of
a particular branch of industry may have been computed by the use of
data on production and productivity, with the production data specific
for the time and place, but the productivity data based OR the industry
branch generally rather than the specific plant in the city being
studied. In such cases the local literature must be searched for
indications of the product mix in the local pleat or any other
information that will indicate whether the industry-wide Productivity
figure applies or must be modified.
II. Estimation of changes since the base date.--There should be separate
consideration of manufacturing and nonmanufacturing industries. It cannot
be too often asserted that the estimating technique to be followed Will
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 ? CIA-RDP81-01043R0020onrAnnnR
of.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 126 -
depend on the data available. In the case of certain manufacturing
industries," for which some data on output, production methods, plant
size, installed power, and the like, may be found for different dates?
there are a few general lines of approach which may be stated:
A. If data on output of a particular plant or set of plants
comprising the industry in the city are available over time, they
may be converted to estimates of exployment of rabochim? if productivity
rates are available or can be estimated for the same dates. If no
better measure of productivity is at hand, recourse can always be
had to extrapolations of prewar productivity rates consistent with
national trends. Estimates of rabos, so comouted would have to
be converted to estimates of the total labor force, as described
above.
B. If there are significant technicel and dimensional similarities
between the city being studied and some other city for which more data
are available, estiretes may be made by analogy. The acceptance of
an analogical model should be limited to cases for which reliable
data on production csnnot be procured or for uhich suitable produc-
tivity rates cannot be estimated. The most legitimate use of analogy
Is as a device for validation of unofficial data on exoloyment.18
C. In a fell instances19 data may be found on the gross value
of output in largescale industry for a specific city. Such figures
are a mixed blessing since all the problems of the interpretation of
gross output data are complicated many times over by the additional
problem of ascribing a relevant prOductivity rate where the output
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
- 127 -
mix is likely to be uncertain. Extrapolation of a known aggregate
productivity from some earlier date is likewise precarious because
of the differential movement of productivity rates among various
industries. In the case of Samarkand, for example, gross -Industrial
output for 1950 was estimated with fair confidence. All that was
knewn of productivity trends for large scale industry, however,
referred to the Uzbek Republic as a whole where the production of
producers, durables was considerably more imnortant than in Samarkand.
Guidance from this source, therefore, had to be rejected and an
alternative extrapolation model devised.
D. Postwar estimates of employment iv:various nonmanufacturip.g
activities we been based almost exclusively upon extrapolations of
prewar figures with recourse occasionally to simple ratio techniques.
Where the activity appears to be a ft/notion of population size, as
or example service activities, the extrapolation model generally
adopted has been:based upon change in total population. This procedure
has been followed for allnernn_nennfacinn.'ing satagories except transpor-
tation and communication in the case of Tula. (See appendix TV.) A
more appropriate extrapolation model for transportation and communication
appears to be the changes in employment or volume of production in
manufacturing. Where structural changes have occurred outside of
manufacturing; such as the changes resulting from greater investment
of human resources in health or educational activities, extraoolations
based on population changes have been abandoned in favor of devices
which can reflect these changes. VOX.I.OUB types or per capita ratios
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
? 128 ?
are sometimes helpful in these cases, e.g., ratios of students to
the number employed in education. Xn general, however, much greater
attention should be given to the development and validation of these
ratios. Soviet city planning literature contains other suggestions
as to ratios which likewise deserve greater investigation than they
have received heretofore.
E. The analogic apnroadh has sometimes been used to obtain an
approximate distalibution, of the =manufacturing labor force) by
major industry group. Thus, in the case of Chirchik?" a small but
significant city of the Vzbek, S.S.R.0 the estimated =manufacturing
components of the labor force were modeled upon those of the city of
Cheremkovo? which as the result of comparative study had been found
to resemble Chirchik with l'espect to: (a) general industrial
similarity (chemical industry, machine building), (b) population
size class, (c) sine of labor force in large scale industry relative
to total employment, (d) employment in education, (e) employment in
construction. Other adaptations of the analogical method were employed
In the studies made of Te.la (ace appendix I% Noluga21, and Ryszan,22
Validation of city labor force estimates.
Comments on the subject of validation can be brief. Aside from the
inherent reasonableness of the steps which lead to the final estimates, there
are no infallible tests of validity at the present time except the requirement
that all the pieces should fit together reasonably well. This means not only
the component parts of the labor force, but population estimates and general
interpretative context as well. At this stage nonquantitative judgements
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : _____
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
vow
- 129 -
inevitably 731ai a luze part, The substitution of ouantitative measures of
veliclity depends mon Droo:ess in systematic irmut-out,-put analysis; a Ereator
understanding o the structure or. SOViet inCtustry end fina1ly2 but most
irmortant; a areater flog of Info=ation from the 7J.S.Soila
'17
4.411.CS
Declassified in Part- Sanitized Copy Approved for Release 2013/07/12 :CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
- 13C -
FOOTNOTES
CHAPTER XXX
See also: Ducoff, L. er., and !agood, M. X. Labor Force Definition and
Measurement, Social Scienca heseareh Council Bul/etin tRiXr3Z7171WjL
2--Tsentra:Pneye ete.tisticheskoye upravleniye SSSR. Itspecguznaya oere0.sv
naseldniya 1926 goda (All-Unics C=SUB of Population 1.1/4516T-7-2.711t-Trosctig*
1929. 1. -
3 Children under 10 years uere not considered active in the 1926 Census of
the U.S.S.R.* and we not considered gainful workers in the 1930 Census of the
United States.
All-Union Census of Poalliation, l926? on oil-, ;WIVE; p. 519.
N-67 for exemple: U. S. Denartment of Commerce, Bureau of the Census.
nAnnual ReporL on the Labor Force, 1955n? Current ropulation Reports, Labor Force,
Series P-50, No. 67, March 1956, WashingtoE7W--127-
7 For the United States this has been demonstrated in an article entitled
nA Service Classification of American Cities n by Howard J. Nelson. Economic
Geogephy, v. 312 no.31 July 1955, DD0 189-21o.
fne medium sized standard deviations for England and Wales, Germany, and
Scotland are attributable largely to the extreme effect of a gmall number of
textile centers and one (jitirmberg) toy center where extraordinarily large numbers
of women are to be found in the labor force. The large standard deviation for
the Soviet cities is partly an artifact since the labor force data were subject
to arbitrary exclusions which were of greater significance in some cities than
in others.
?9 This technique has been used in estimating the labor force of a cluster of
Russian textile tcuns in the Ivanovo region, The chief -problem in this approach
is to give adequate quantitative recognition to the imoressive increases in female
labor force participation, Since these had traditionally been na-7-0M-The old
textile centers, data from the Census of 1926 provided a useable base for the
necessary computations.
For definitions and a description of the 1926 c/assificaticn system see
TsentraloncTe statisticheskoye upravienlye aSSR, OD nit?o, pp. TTTT-xiii, 1, and
52/-540.
11
Modified procedures are often required in connection with the estimate of
military personnel. For a discussion of this problem see pp. 69, 70.
12 Tsetrei'noye traravleniye narodno-khozyaystvennogo ucheta gosplana WM.
Otdel ucheta truda. Chislennost'_i zerabotna a plata rabochikb 3. sluzha_shcbilch
V0 sssa. (The Nue
nber and. Wages o ' War s anrap7oTaii=i tgrr8-1711,10scou,
1936. -
13 Oencit., D. 297.
14 hese-production data usually were reported only for large-scale industry:.
which was defined very similarly to the large-scale industry category in the
1926 Census (see above). The 1926 category, more -precisely cePed ncensus
(tsenzovaye) industry," did not include fisheries, lumbering, and railway repair
shops. For an informative discussion of these categories see Hodgman, Donald H.,
Soviet Industrial Production, 1928-1951, Cambridge, Harvard University Press, 1954,
pp 7-5;-3.75.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
-131-
15 These are cited in the Tula illustration.
16 Indications of the value of this ratio for major industrial categories
are given in a variety of sources. For specific, ie-CEE616EiddifTHEE6gEEEEds
enterprises the best source has proven to be a 1935 handbook, Tsentral'noye
?I narodno-khozyaystvernogo ucheta gosplana SSSR. Trud v SSSR.
Statisticheskiy spravochnik (Labor in the U.S.S.R., A StatisHaTT6nc115ook),
Moscow; 1936, pp. 64-67.
Manufacturing enterprises which have been identified unambiguously but
for which no other information is available (except sometimes a prewar employ-
ment figure) present special difficulties for which there is no general remedy.
To state the problem in its most formidable terms, there are in addition (1)
enterprises whose separate identity is in doubt, i.e., they may represent merely
an alternate designation for a previously identified plant; (2) enterprises
whose existence is a near certainty in view of locational factors or expectable
input-output relationships, but which cannot be positively identified. Our
techniques for handling these cases, whether it be the carry-over or modifica-
tion of earlier figures, or de novo estimation; are no better than the general
state of the arts today in the fd of input-output analysis. As our knowledge
of interindustrial relationships develops, more exact resolutions of these
problems will emerge. At the present time the estimation of industrial matrices
is little more than a theoretical possibility. In the work done to date; we
have been forced to rely upon a few selected extrapolation models and non-
standardized ratio devices.
1? In this connection see U. S. Department of Commerce, Bureau of the Census,
CONFIDENTIAL; Estimated Population and Labor Force of the Aleksin Industrial
Area: 1 January 19502 Series 15=95, No, 35, Washington, D. C.2 10 August 1954;
p. 50 IT, CONFIDENTIAL. The problem faced here was the validation of a
reported figure for employment at the Aleksin Chemical Combine; the area's chief
industry. Data were available relative to production, technology, plant area,
work schedule and power supply. With these criteria; a study was made of German
chemical plants for which there were, in addition to data of the type mentioned,
data on total employment.
19 See for example, U. S. Department of Commerce, Bureau of the Census,
CONFIDENTIAL: Estimates of the Population and Labor Force of Selected Cities of
the U.S.S.R.: 1 January lg50, Series P-95; Nos. 18-23, Washington; D. C.
10 December 1553; D. 55, SECRET; also ibid., CONFIDENTIAL; Estimated Population
and Labor Force of the City of Samarkand: 1 January 1950; Series P-g5, No. 37;
Washington; D. C.2 20 January 1955, pp. 44-45, SECRET.
20 U. S. Department of Commerce; Bureau of the Census, CONFIDENTIAL, Estimited
Population and Labor Force of the City and Metropolitan Area of Chirchik:
1 January 1950, Series P-95; No. 24, Washington, D. C., 10 March 19547-5. 49,
SECRET.
21 U. S. Department of Commerce, Bureau of the Census? CONFIDENTIAL; Estimated
Population and Labor Force of the City of Kaluga: 1 January 1950; Series P-952
No, 3, Washington, D. CTO June 1955, SECRET.
22 U. S. Department of Commerce, Bureau of the Census, CONFIDENTIAL, Estimated
Population and Labor Force of the City and Metropolitan Area of Ryazan':
1.-Sanliarri:55-0;-S-eiTei-P":55, No, 472 Washington, D. Ca, December 19552 SECRET.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
a
APPMEK
CHRONOLOGICAL LIST OF MAJOR SOURCES FOR U.S.S.R. CITY POPULATION DATAI
rear
Source
Remarks
1926 Tsentralsnoye statisticheskoye unrav-
leniye SSSR. Vsesoyuznaya perepis?
naseleniya 192rgoda (Ali-Unien Casus
aosTl
FIEZCZT5W02-97
1931 Tsentral'noye upravleniye narodno-
khozyaystvennogo ucheta gosplana
SSSR. Trud v SSSR (Labor in the
U.S.S.11757YETERiairrairi77737--
. Narodnoye khozyaystvo SSSR
(Natio.nal EcoliaTa-in77787Kff.),
MO736-61771:0- 2? Pe 4-03747/:
Last complete published census covering the
entire U.S.S.R. For many small cities this source
provides the most recent published population
figure. Special tabulations based on the 1926
Census but presenting data not found in the
regular census publications may sometimes be
found for local areas. These are compiled by
local statistical agencies.
This source presents population totals for 19262
1931 and 1935 for all cities with a population
1002000 and over in 1935. The 1931 figures are
based on city enumerations.
Presents data based on same 1931 city enumeration
as above2 along with totals for five earlier
dates going back to census of 1897. In some
cases the figure differ from those in Trud v SSSR.
This may be due to variations in the hadIing of.
certain population groups.
As with the 1926 Census, local statistical agencies
have sometimes published special tabulations for
their areas. These will frequently contain
information on cities below 1002000 in population.
(Continued)
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
CHRONOLOGICAL LIST OF mom SOURCES FOR CITY POPULATiON DATA---Con.
?
Date
Source
Remarks
1933 Tsentralgnyy ispolnitellnyy komitet
soyuza SSSR. Adminiotrativno-
territorialnoYrNENERraEVUZa
nAdininiatrative Divisions of
Yffe-1,75-78.
1935 Trud v SSSR be cit. also
Tsentral'noyeirdniye narodno-
khozyaystvennogo ucheta gosplana
SSSR. Sotsialisticheskoye
stroitA7F0761041X-06156rEist
ConswucIlon of tg-U7STSTET,
FiF6-c-iirrT9M-
1936 Besher; R. Ya. Problemynaseleniya v
?
narodnokhozyaysvennoanarovaTa:
(Problepe orropulatioconora-c
Planning); Moscow; 19'7.
A comprehensive list of official Soviet
estimates for cities and smaller settlements.
These estimates frequently appear to be
invalid; possibly they are the result of an
unrefined Projection of earlier census
materials. This source also contains
aggregate data on rural and urban population.
The definitions of rural and urban territory
employed here (those of the Central Adminis-
tration for Economic Accounting of
Tsentral'nyy ispolnitenw komitet soyuza
SSSR) differ from those of the Central Admin-
istration of Economic and Social Statistics;
Tsentralinoye upravleniye narodno-
khozyaystvennogo ucheta gosplana SSSR; the
agency Charged with conducting the census.
Official Soviet estilentes for cities of
100;000 or more inhabitants plus estimates
for a number of selected neu cities.
An apparent continuation of the series
presented in Sotsialisticheskoye stroitel'stvo
SSSR.r ? OD cit.
(Continued)
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Year
CHRONOLOGICAL LIST OF MAJOR SOURCES FOR U.S.S.R. CITY POPULATION DATAL-Con.
==.
Source
Rei -narks
1939 Isvestiya? June 2: 1939.
Tsentral'noye upravleniye nerodno-
hhozyaystvennogo uchetn gosplana SSSR.
Kullturnoye stroitellstvo SSSR
cEiiiai Oonstruction
Moscow: 19740770772-672-7.
? 1956 . Narodnoye khozyaystvo SSSR
(Naticari3eoriala-TS6-0:5:M%);
Moscow, 1956.
1956 Statisticheskoye upravleniye RSFSR.
Narodnoye khozyaystvo PaSFSR.
onomygiaMicheglirThs orniirctTa-tionai.
.-1Eg R.S.P=EaTistica].
TrailanaTTFIFECZ:7719317.
Results of the Census of 1939 for cities
with 50,000 or more inhabitants. Same source
provides 1926 populations for 1939 boundaries.
Educational statistics for a number of cities
are presented in percent of the 1939 Census
popul-,tion. This source is especially
useful for cities of less than 50,000
population.
Population figures for Republics, and cities
of 100:000 or more population for 1956;
selected vital statistics; statistics of
population distribution for urban and rural
residence.
Population statistics by ?blasts, urban sn
rural, for cities of 50,000 or more population,
and for smaller places which are oblast: kray,
okrug? and Republic centers; selected vital
statistics.
/ Soviet encyclopedias often include population figures in articles dealing with particular cities.
See Bollshaya sovetskaya entsiklopediya (Great Soviet Encyclopedia), Moscow, 1st and 2nd editions; also
Ma2igrOAYEERTrair=i5CRITR-CIBair-55a5T-EfiTFraggirETT-ffigEow:
let and 2nd editions.
tt.)
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
APPENDIX IX
SOVIET ELECTIONS: 1937-1955
Date of election Type of election/
SOURCES
FOR DATA ON:
Election results2
Election law
Number of election districts
December 1937
June 1938
December 1939
February 1946
February 1947
December 1947
All-Union
Republic
Local
All-Union
Republic
Local
Pravda, 10 July 1937
Pravda, 12 October 1937
Izvestiya: 15 Dec.1937
11 Feb.19383
WERT, 3 Aug.19393
Pravda, 22 April 1938'
Pravda: 15,27-29 June21938
IEV6Raya, 5 Jan.19404
(4)
Pravda, 17 Oct.1945
Pravda, 30 Nov.19463
Pravda, 12 Oct.1945
PFEVE', 27 Nov.19463
lzvestiyal 14 Feb.1946
PRivda: 13-19 Feb.19473
PRViarl Ukrayiny, 28 Dec.
1948-5
(4)
March 1950
All-Union
Pravda; 11 Jan.1950
Pravda, 12 Jan.1950
Pravda: 15 Mar.1950
December 1950
Local
REVa'g, 30 Oct.19503
(4)
22 Dec.19503
February 1951
Republic
?55EVEE, 12 Dec.19503
Pravda: 15 Dec.1950s
TT751E, 22 Feb.-1 March
1931
Feb.-March 1953
Local
Pravda, 3 Oct.19503
(4)
(4)
March 1954
All-Union
11 Jan.1950
Pravda, 13 Jan.1954
Pravda, 18 March 1954
Feb.-March 1955
Republic
12 Dec.19503
Xzvestiya? 25 Dec.
TEVigiiya, 3,4:8;10 March
Feb.-March 1955
Local
Pravda, 3 Oct.19503
(4)
(4)
????????
????????????????????????????=
1 All-Union - Election of deputies to the Supreme Soviet of the U.S.S.R. and to the Soviet of Nationalities
of the U.S.S.R.
Republic - Election of deputies to the legislative bodies (Supreme Soviet) of Republics of the U.S.S.R.
E5E1-7.- Election of deputies to the soviets of krays; oblast; rayons, cities and villages.
2 Names of elected deputies, number of ballots cast, percent of votes cast for party candidates.
The sources listed give information primarily for the republic and local elections of the R.S.F.S.R. Since
elections of this type are held throughout the Union on approximately the same dates: an examination of the
provincial press for a Period within several days of the dates given above will provide data for the areas not
covered in the sources indicated here.
4 Comprehensive releases of local election data are seldom found except for republic tabulations of delegates
and voters classified and aggregated by major administrative division (kray, oblast, city, etc.). Scattered
information may be found in the local press on or about election time.
0
CD
CD
CD
-0
CD
(I)
CD
CD
CD
6,
7:1
CD
CD
CD
CD
0
0
. .
0
7:1
0
-0
CO
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
APPENDEK III
This anpendiz is devoted to on illustration of the use
that may be made of All-Union and Republic electoral district data
to yield population information about areas so small that they
comprise less than a single electoral district. The electoral
district data by themselves shed little light on the problem, but
if they are used judiciously together with fragments of information
from other sources, they begin to offer useful possibilities.
The case selected for this illustration is Aleksin Rayon,
which lies in Tula Oblast of the R.S.P.S.R. The problem is to
determine; if possible, the population of the rayon and how it has
changed from 1937 to 1954. This rayon has never been large enough
to have an entire electoral district assigned to it; it has always
fallen into electoral districts which contain considerable other
territory as well A detailed analysis of the Aleksin area is
available in: U. S. Bureau of the Census, CONFIDENTIAL. Estimated
Po ulation and Labor Force of the Aleksin Industrial Area:
1 January 1950, Series P-95, No. 35, Washington, D. C., 10 August
1954, CONFIDENTIAL. It should be noted that the process of
-17
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
estimation described below deals with estimates of population of the
rayon as a whole, but that this was a matter of secondary importance,
being subsidiary to the main problem of estimating the urban popula-
tion of the rayon.
The electoral district information pertinent to the problem
is summarized in table II1-1. To employ this information successfully
it was necessary to use also information for 1936 on the population of
cities and rayons in Moscow Oblast, as it was then constituted. This
information is found in Rayony Moskovskoy Oblasti (Rayons of Npscow Oblast) _
Moscow, 1934, pp. 4-19. It was also necessarY to use a number of
relationships which had been estimated earlier for Tula Oblast, a
neighboring oblast with a rural population quite similar to that in the
rayons which, together with Aleksin Rayon, formed a single electoral
district. A detailed discussion of the derivation of the estimates for
Tula Oblast is given in: U. S. Bureau of the Census, CONFIDENTIAL
Estimated Population and Economically Active Po ulation of the Cit and
Metropolitan Area of Tula: 1 January 1950, Series P-95, No. 25,
Washington, D. C., 10 May 1954, CONFIDENTIAL.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Table III-1.--ELECTORAL DISTRICT INFORMAT/ON RELATING TO
ALEXSIN RAYON: 1937-1954
..???1???????????2?.?? ? YAW.. 11111.1?????????????ar ??????.,
Date of pliblica- tType of election and
tion of electoral legal ''atio of popu-
district lists lation per raprisen-
atative
Composition of electoral dis-
trict to which A/eksin
rayon belongs
1937(October).. ."All
-Union - 300:000
Aleksin rayon
Dugna it
Babynin "
Kaluga " (including Kaluga
city)
Tarusa II
Detchino "
1938 (April)... .
Republil: - 150,000
Aleksin rayon
Dugna I:
Kaluga " (excluding Kaluga
city)
1945 (October) .
All-Union - 300,000
Aleksin rayon
I Kosaya Gore rayon
1 city
--Tula
1946 (November)..
Republic - 150:000
?
Aleksin rayon (These two area.
Tula city comprise two
districts
1950 (January). .
All-Union - 300,000
Same as in 1945.
1950 (December)..
Republic - 150,000
Same as in 1946.
1954 (January). .
All-Union - 300,000
1
Aleksin rayon
Tula it
Zaokskoye "
Ivan'kovo "
Laptev? "
Venev It
Odoyevo "
Leninskiy "
Mordves "
Dnbna It
Khanino "
Sources:
1937
- Pravda, Moscow, 12
October 1937.
1938
- apmda, Moscow, 22
April 1938.
1945
- alavda, Moscow, 17
October 1945.
1946
- FEEL: Moscow, 30
November 1946.
1950
- anyAn, Moscow, 12
january 1950.
1950
- Lz_vs_slim, Moscow,
15 December 1950.
1954
- Pravda, Moscow, 13
January 1954.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Co .y A proved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
In general, the method used as to strip away from the
electoral district estimates or the population of each component area
other than Aleksin rayon, leaving the population of the latter as a
residual. The details of how this was done for different dates are
given below;
1937
1. The population of the electoral district containing
Aleksin rayon was estimated at 300,000. The electoral
district contained 5 other rayons which were in Moscow oblast.
2. The rural population of these 5 rayons in 1936 was
obtained from the work cited above and reduced by 2.7 percent,
which was the estimated percentage decline in the rural popu-
lation of nearby Tula Oblast during the same period.
3. Minor adjustments were then made for presumed
territorial changes in these 5 rayons, on the basis of the
number of seltsoviets.
Declassified in Part- Sanitized Co?yA proved forRelease2013/07/12 ? CIA R P I 1?4Pnno nnn nnnn 4
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
4. The resulting estimate of the rural population of
the 5 rayons in 1937 was 169,000.
5. The only urban population in these 5 rayons was that
in the cities of Kaluga and Tarusa, for which 1936 data were
available. The population of Kaluga was also known for 19392
the data of the census, and that of Tarusa was known for
1938. Estimates of the population of each city in 1937 were
obtained by linear interpolation. Their combined population
was estimated at 74,000.
6. By subtraction of these results (169,000 and 74,000)
from the 300,000 total population estimated for the entire
electoral district, there was derived an estimate of 57,000
for Aleksin rayon:
Estimated Population of Electoral District ? 300,000
Less- Rural population of Dugna, Babyninp
Kaluga, Tarusa, E.nd ;.atchino rayons- 169,000
Less. Urban population of these rayons,
Kaluga and Tarusa cities : 74,000
Estimated Population of Aleksin rayon
(residual) : 57,000
--+
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
1938
1945
1. The population of the electoral district containing.
Aleksin rayon was estimated at 150,000.
2. In addition to Aleksin rayon, this electoral district
contained the rayons of Dugna and Kaluga, but not Kaluga city.
3. It was assumed that Dugna and Kaluga rayons, ex-
cluding Kaluga city, had a decline in population of 5.3 percent
between 1936 and 1938. This is the percentage of decline
estimated for the rural part of neighboring Tula Oblast. The
.*
combincd 1938 population of these 'areas was thus estimated at
86,000, leaving an estimate of 64;000 for Aleksin rayon:
Estimated Population of Electoral District: 150,000
Less: Population of Dugna and Kaluga
rayons, excluding Kaluga city : 86,000
Estimated Population of Aleksin rayon
(residual) : 64,000
1. The population of the electoral district containing
Aleksin rayon was estimated at 300,000.
2. In addition to Aleksin rayon, the electoral district
contained the city of Tula and Kosaya Gora rayon.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
1946
In - 7
3. The population of Tule city in 1945 had been estimated
at 220,000 in the Bureau of the Census report on that city.
4. The rural population of Kosaya Gora rayon was esti-
mated on the assumption that its 10 selfsoviets had the same
average size as did all seAoviets in Tula oblast in 1945. The
urban population. of the rayon; which was in fact the popula-
tion of Kosaya Gora settlement, the rayon center, was estimated
on the assumption that its rate of population change since 1939
was the same as the rate estimated for the urban part of Tula
oblast. The total population of the rayon was thus estimated
at 28,000.
5. The population of Aleksin rayon was estimated by sub-
traction, as 52,000:
Estimated Population of the Electoral District: 300,000
Less: Population of Tula city 220,000
Less: Population of Kosaya .fora rayon 28,000
Estimated Population cf Aleksin rayon (resi-
dual) : 52,000
1. The population of 2 electoral districts containing
Tula city and Aleksin rayon was estimated at 300,000.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
e
1954
2. The population of Tula city had been estimated at 236,000
in the Bureau of the Census report on that city.
3. The population of Aleksin rayon was estimated by sub-
traction, as 64,000:
Estimated Population of 2 Electoral Districts- 300,000
Less: Population of Tula city
Estimated Population of Aleksin rayon
(residual)
236,000
64,000
The method of arriving at the population of Aleksin
rayon as a residual, which was applied with some success for
1937, 1938, 1945, and 1946, was not particularly useful in
connection with the data available for 1954. As is shown below,
it was possible to "peel off" figures representing the rural
population and the urban population outside of Aleksin rayon,
but the residual for the rayon was too small to be creditable.
The analysis proceeded as follows:
1. The total population of the electoral district
containing Aleksin rayon was estimated at 300,000. There were
10 other rayons in the election district. See table 111-1.
2. The rural population of the 10 rayons combined was
obtained for 1936 from the work entitled "Rayons of the Moscow
? 4-?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Oblast", and adjusted to 1939 on the assumption that it
changed to the same degree as did the entire rural popula-
tion of neighboring Tula oblast during the same period. There
was also a small adjustment for presumed territorial changes on
the basis of the number of sel;oviets. The resulting estimate
was 300,000.
3. The 1939 estimate of rural population for the 10 rayons
was reduced by about 30 percent, to yield an estimate of 210,000
for 1954. This reduction was made on the assumption that the
rate of population decline in the rural part of the 10 oblasts
between 1939 and 1954 bore the same relationship to the rate of
decline of the rural population of the R.S.F.S.R. during the
same interval, as did the rate of decline of rural population in
Tula oblast to the rate of decline of the R.S.F.S.R. rural popu-
lation during the period 1926-1939.
4. The urban population of this electoral district, apart
from the urban population in Aleksin rayon, consisted of nine
small urban settlements. One of these, Venev, had a population
of 6,827 in 1936. No data are available for the other eight
settlements. If they had the same average size in 1954 as did
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
workers settlements in Tula oblast in 1936, their aggregate
population plus that of Venev, would amount to 53,500. This
is clearly too high, for it implies that Aleksin rayon would
have, as a residual, a population of only 36,500 in 1954. It
is very unlikely that the population of the rayon could have
dropped so drastically.
5. It follows that the rural population must have de-
clined more rapidly than was assumed, that the average size of
a worker's settlement was lower in 1954 than in 1936, or
perhaps both to some extent.
6. It is a proper inference from the data that the popu-
lation of Aleksin rayon was no greater in 1954 than in 1948,
but the extent to which it might have declined is not a fact
that can be elicited from the electoral district population
estimate.
An alternative method of estimation placed the popu-
lation of Aleksin rayon in 1954 at about 50,000. If to this is
added the estimates for the urban and rural parts of the electoral
district apart from Aleksin rayon, the total of 313,500 is
attained. This is clearly within the range of error of the
estimated population of 300,000 for the entire electoral district.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
It appears, therefore, that the method discussed here for
estimating the population of a minor component of a7.3
electoral district by reference to other data, may fail when
the degree of error in the estimate of the electoral district
is a considerable fraction of the population of the minor
component. Inasmuch as the error inherent in any population
estimate for an electoral district is undetermined, there is
always the risk that the estimate for the minor component is
wrong. It is only by reference to estimates for other dates
that any feeling of confidence in the results may be justified;
and even if the figures for a series of dates appear to be
reasonable in the light of other information, they may be no
more than crude approximations to the true population.
Declassified in Part - Sanitized Copy Ap?roved for Release 2013/07/12: CIA-RDP81-n1 n2VIRnn9nrinnarinno
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
"
APPENDIX IV
ILLUSTRATIVE EXAMPLE OF A METHOD OF ESTIMATING THE SIZE AND
INDUSTRIAL COMPOSITION OF A SOVIET CITY -- TULA:
1 JANUARY 1950
I. General Method
Because of the fragmentary nature of the data available on
Tula, the general method used in estimating the economically active
Population was an analysis of its development between the census
year, 1926, and the final estimate year, 1950. By this means, the
total economically active population in the terminal year reflected
the expansion (in relation to industrial and population growth) of
base estimates for earlier years on which fuller data could be
developed. Finally, several checks were applied to verify approxi-
mately the results obtained.
Essential data underlying the estimates included the
following:
A. Direct employment data.
1. The 1926 economically active population, taken from returns
of the All-Union Census of 17 December 1926, is shown in
table III-A. This table gives the civilian economically
active population for the city of Tula only.
2. The 1931 employment for certain industries in Tula is
presented in table III-B.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
IV -
3. The economically active population reported as of 1 April 1936 for
clula was 115,243./ On the basis of the percentage of total popula-
tion represented by this figure, and of thc general practice in
other sources of treating the three areas as an economic unit, it
is believed that this figure includes Tula, Kosaya Gora, and
Novo-Tul'skiy. The economically active comprised 51.5 percent of
the torts1 population of 224,000, estimated as of 1936, for the
aggregate of these communities.
Table III-A.--CIVILIAN ECONCMICAILY ACTIVE POPULATION OF TULA:
17 DECRIER 1926
Category
1 Civilian economically active
3Ota1100 ???????00 000.0000 ?Oc? ? ??
I. Agriculture ......... ....... ..... .......
II. Factory Industry.. .....................
III. Handicraft industry ..... ... OW00000004
IV. Construction........ ...................
V. Railroad transport.. ...................
VI. Other transport..... ..... .. e000000.0600
VII. Trade and credit... ....... . *00.004.0.000
VIII. Institutions. oepo.,100[40"0?000.00000000
IX. Other. 0000000000000000000.00 ...... 00000
Percent
67,950
100.0
2,062
3.0
32,428
47.7
7,046
10.4
1,557
2.3
5,612
8.3
972
1.4
6,009
8.8
9,832
14.5
2,432
3.6
1 Branches of the economy "No occupation or unstated occupation," "Unemployed,"
and "Free professions," mhich are presented separately by the census volume,
have been distributed proportionately among branches
Source: Tsentrallnoye statisticheskoye upravleniye SSSR. Otdel perepisi.
Vsesoyuznaya pergis' naselenVa 1926 oda (All-Union Census of
Population, 1926), Moscow, v. 19, 1929, p. 128.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Iv - 3
Table II/-B.--WORKERS (RABOCHIYE) IN SELECTMMANNTACTURING
INDUSIBIES, TULA: 1931
Industry
Number of workers
Appare11?00?0.?....?.....00?0000??00????0?000?0?0
Food... 0 00 0 0000" ?0000 0.0 *0 0 ? 00000000 000000000 ?000
Sugar refining... ..00 V 00 000 00 000 000 0 0 ? 0 0 ? ?0 0
Disti11ing.?00000000000000000000 0000 00 000 0?0
3,598
2,237
1,620
178
Bakeries and factory kitchens. 0 0 ? *0400000 000
439
Wood and construction materials ....... ........
1,564
331
Sawmills. 0?000 0 00000 ?00000000 D?4000000000000
233
Furniture and
000000
1,000
Printing? ? 0 ? 00'? 0 ? 0 0000 0 0000000000 0000000000
0'000011
525
1 The apparel industry includes employment in clothing, shoes;
leather, and furs.
Sources: Ves' SSSR (All of the U.S.S.R.), Moscow, 1931, pp. 351-774,
and Tsentrallnoye statisticheskoye upravleniye SSSR.
Otdel perepisi. Vsesoyuznava perepis naseleniya 1926 goda
(All-Union Census of Population; 1926), Moscow, v. 19, 1929,
p. 128. Where direct information for 1931 was lacking,
detailed occupational statistics from the 1926 census were
used. An increase of 25 percent (the same as for population
growth, 1926-1931) was made for those persons employed in
occupations whose products were for use of the local
population.
B. Data on industrial production and facilities.
1. Table III-C presents a list -- a partial list, it should be
noted -- of the increase in productive facilities in the Tula area
between 1931 and 1950. Data from this table provide the basis for
estimates of change, by industry, of the economically active in
Tronufacturing.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
IV - 4
Table III-C.?INCREASE IN PRODUCTION FACILITIES, 1931-1950, IN
TULA; KOSAYA GORA, AND N0V0-T1JL1SKIY
Industry
Increase
in production facilitic21
1936-1939 i
1940-1950
1
Postwar reconstruc-
tion including
installation of a
iturbo-blower.
Postwar reconstruc-
tion, addition of
blast furnace No. 3.
Wartime employment or
Kalinin plants reports
at 20,000 or above.3
Plant No. 176, Imeni
1931-1935 I
Ferrous metallurgy
1. Kosaya Gora
2. Novo-Tul1skiy
Metal-working and
machine building2
1
Blast furnace No. 3
put into operation
in 1932.
Blast furnace No. 1
put into operation
in 1935.
?
?
?30,000,000
Blast furnace No. 2,
put into operation
in 1936.
Expansion of the
,armaments industry,
'especially Kalinin
1535 and 536, and
Plant No. 176,
Imeni Lenin.
Lenin, had an employ-
ment of similar magni-
tude.
Novaya Tula Plant
No. 187 has become a
major producer of
grain combines in the
postwar period.4
The Kirov Machine Tool
Plant No. 60 has been
reconverted since the
war and has greater
production than
prewar.5
Zars Motor Vehicle
Repair Plant newly
reported during the
war.
Novaya Tula Arma-
ments and Machinery
Plant No. 187
believed to have
begun production in
this period,
New construction at
Krylenko Hardware
Plant scheduled to
begin in 1936 and
to be completed in
1937. New annual
capacity planned at
rubles.6
?
1
I
(continued)
r_
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
IV-5
Table III-0.--INCREASE IN PRODUCTION FACILITIES, 1931-1950, IN
TO1A, KOSAYA GORA, AND NOVOAUL'SKIY--Con.
Industry
Increase in production facilitieal
1931-1935 1936-1939
1940-1950
Metal-working anti
machine bui1ding4
(continued)
Food
1. Meat
2. Flour milling
Apparel
? Meat kombinatl?
Flour mill"
1. Clothing
2. Rubber footwear!
A railroad repair
plant, with a
capacity to perforn
capital repair on
4,800 passenger
cars per year was
scheduled for com-
pletion in 1936.8
Addition to flour
mill, 11 macaroni
factory, 12 and
refrigerator 13
The Krasnyy Oktyabra
Plant was reconstruc-
ted in the postwar
period.?
Postwar plans for ?
the machinery plant -
Imeni Batischev
called for increased
production over the
prewar 1eve1.9
Sewing factorics,./4
Rubber plant."
1
This table presumably does not present a complete list of the expansion of
production facilities in the city between 1931 and 1950.
1= Information on the metal-working and machine-building industry is very
limited. Data presented here on changes in facilities at various plants cannot
be considered as comple4.
3 Reichsministerium fur die Besetzten Ostgebiete. Der Generalbozirk Tula?
Entwurf, abgeschlossen am 25 November 1941, Reichskommissariat Russland, 1941.
4 Izvestiya, Moscow, 13 July 1951; Selikhozmashina (Agricultural Machinery),
Moscow, October 1951, no. 10? p. 1.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
11.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
%
?
Tv - 6
Footnotes to table III-C.--(continued)
5 Izvestiya, Moscow, 8 July 1945, 17 January 1948, and 20 january 1950.
6 Gosplan. Narodnokhozyaystvennyy plan na 1936 god (National ECOMDMiC Plan
for 1936), Moscow, v. 2, 1936, pp. 324-340.
7 Promyshlennost' stroitelinykh mater ialov (The Building Materials Industry),
Moscow, 12 March 1950, p. 2.
8 Gosplan. Narodnokhowyaystvennyy plan na 1936 god, on. Cit.
9
Izvestiya, Moscow, 22 May 1945.
10 The meat kombinat and flour mill were not mentioned in the 1931 source,
but did appear in a 1934 listing of the city's industries. Vysshiy sovet
narodnogo khozyaystva SSSR. Spisok promyshlennykh pred riyatiy k atlasu
"Promyshlennost' SSSR na nachalo 2-y pyatiletki," (List of industrial Enter-
-arises for the Atlas "Industry of the U.S.S.R. at the Beginning of the 2nd
Five-Year Plan"), Moscow, 1934.
11 Listed in Gosudarstvennaya planovaya komissiya, Vtoroy pyatiletniy plan
razvitiya narodnogo khozyaystva SSSR, 1933-1937gg (Second Five-Year Plan for
the Development of the National Economy of the U.S.S.R., 1933-1937), Moscow,
V. 2, 1934: p. 368. Scheduled for completion in 1936 with a planned capacity
of 100 tons per 24 hours.
12 Listed in the same source as footnote 11 above. The factory was to be
completed in 1937 and have a planned capacity of 11,300 tons.
13 Listed in the same source as footnote 11 above. Completion date was 1937
and planned capacity 2000, tons.
14 Construction of the sewing factories was noted in two sources: Legkaya
promyshlennost' (Light Industry), Moscow, v. 7, December 1947, no. 12; and
U. S. Army, G-2 Report No. 9008913. The shops were to be completed in 1949,
with the following annual production: 75,000 men's suits, 80,000 pairs of
trousers, 30,000 women's suits, and 60,000 men's short overcoats.
15 Mentioned in many sources, this plant was to begin production by the end
of 1946, and to produce 1,200,000 pairs of miner's boots per year, as well as
other articles such as women's galoshes, insulating tape, rubber hose, shoe
soles, oxygen tubes, and fan belts. (Izvestiya, 1 June 1947; Turkmenskaya
Iskra, Ashkhabad, 1 June 1947, and Ministerstvo rezinovaya promyshlennost',
Zavody rezinovoy promyshlennosti (Plants of the Rubber Industry), Moscow,
1948.)
Declassified in Part- Sanitized CopyApprovedforRelease2013/07/12 ? CIA-RDP81-01043R00200006000A-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
- 7
2. Tho gross industriP.1 production of Tula, Kosaya Gora, and
Novo-'.oul'skiy in 1935 was shown on a Soviet map to betwecn
400 and 530 million 19'-6-19'7 ruble'.;; The dis,ribuaon of
'his production vas 7.53 perceni, in ferroY2 mesallurgy, 88 percent
in metal-working and ma'thine-building, and 4.2 percent in R11
other, Since the physical output of ferrout: metallurgy 5n
in 1935 is known, an4 its ldalue in 1926-1927 rubles easily
determined (table II1-D), it could h, entalllished, by applying
the calculated value of metallurgy and percentage of tilo
total, that the percentages given rq?er to a universe of soixe 317
million rather than 400 to 500 mfliion ruble L, The gLp in
coverage appears to fall in "other" catecory, .he calculated
value of which comes to 3.3 millior 1926-1927 rubles, as opposed
to an estimate (buil-6 up -:ndut...try by indusCry) of 85.8 million
rubles (table III-G).
3. For 1937, a figure of 'otal production "above 400 million rubles,"
broken up into 15.1 percent ferrous metallurgy, 47.2 percent
metal-working, and 37.R percen,, food has ben publishPd. Here
again, the universe represented Oy .11e percentage is incomplete.
The values represented 'oy ferrous metallurgy, metal-working, and
food are 4.9.8 156, and 125 million (1926-1927) rubles, respectively,
with a total of 330 million I should be noted that the figure
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
iv - 8
for metal-working is 135 million rubles loss than cited for 1935.
Obviously, this represents a difference in coverage; most likely,
the first is a comprehensive figure, while the e_cond covers
selected, presumably exclusively civilian, plants. The figure of
125 million for "food" is far too high, and must be considered as
covering food, clothing, and all other industries except ferrous
metallurgy and metal-working. If this assumption be granted, the
value of output for these residual industries increased by 47
percent between 1935 and 1937. The increase in employment within
these industries is estimated at 19.5 percent. The greater increase
in value of output is accounted for by increased productivity,
4. Production data on the ferrous metallurgy industry are relatively
complete, and have been used as a basis for obtaining the
distribution of the economically active in manufacturing by
industry in all of the estimates (table
C Annual productivity rates.
1. The conversion of indirect production data (output in 1926-1927
ruble values) into employment estimates has necessitated the
compilation of an annual productivity series, 1928-1950; and total
economically active to worker ratios. The latter, based on the
1941 Plan, appear to have been fairly stable since the early
1930's (table
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Iv-9
2. The great industrial and population growth of the Tula area between
1935 and 1950 had to be accompanied by a large increase in the
quanta of transportation required, and is Pstimated to have called
forth a significant rise in the employment in transportation and
communications. The calculations by which this increased labor was
estimated are shown in table III-F.
D. Analogical data.
1. In the absence of direct avidence for distributing the economically
active in nonmanufacturing by category, the percentage distribution
of this group found in the urban parts of the 1936 Moscow Oblast
(excluding Moscow city) has been used. Tula city comprised a
substantial part of this aggregate. The pattern is as follows:4
Category
1
Percent
Total.......
Construction.. oco000000000C00000,0000000 ... O
100.0
17.9
Transportation and communication........... ..
26.3
000000CCOcoo0000CC C
27.3
Health. 000000000oopooe008000C0c0 ....... Cepoo
8.3
Education. CooC000C0000000Oopc00oo0Coc ..... Co
9.4
Government administration and party
apparatus. ..............................,
7.5
Other (including agriculture)..... 0006000.00 C
103
This pattern was adjusted somewhat in the 1935 and 1939 estimates
to account for special circumstances in Tula, e.g., the increase
in transportation and communication, and an increase in the
government administration category reflecting Tula's becoming an
oblast center in 1937. The adjustments are aiscussed more fully below.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Table III-D,--FERROUS METALLURGY INDUSTRY; TULA METROPOLITAN AREA; 1932-1950
(Values given in 1926-1927 prices)
Year
IPmductio-
1 Blast
furnace voluw
albic m(ters)
1 (In metric tons of pig iron)
!Value (in
;thousands
'thousands
? of rubles)
Unit value
(in rubles
per ton)
(in
1
i Iotal
'
Kosaya 1
Core.
Novo-
iTurskiy
Total
1
iKosaya Gora
F
N.3"-
r
,
1
1932..1
104,600
-`104,600
(2)
(3)
(3)
1 1,410.5
41,410.5
(2)
1933..
i 166,200
5166,200
(2)
615.1
90.85
(1)
(3)
(2)
1934..1
220,400v
7220,400
(2)
619.9
90.29
j
(3)
()
(2%
. )
1935..
t 9285,400
i 1?180,000
11105,400
(3)
1290.57
(')
(3)
1930
1936..
9536,000
1 14171,000
15365,000
(3)
(3)
(')
(3)
131,860
1937.
16 550,0001
(1)
(3)
(3)
(3)
(3)
(3)
(3)
1950.
171,120,000
(3.)
(3)
(3)
3,616.0
1,426.0
2,190
---i
1 Gosplan, Otdel gornoy metallurgiy i promyshlennosti. Chernaya metqlt1:1.:c0a zhelezorudnaya,
margentsevaya ? koksovava pllomyshlennost' SSSR. Statisticheskiy spravochnik 1928-1934 gg
(Ferrous Metallurgy and the Iron Ore, Manganese; and Coke Industry of iale U.S.S.R. A Statistical
Handbook, 1928-1934); Moscow, 1935, p. 67.
2 The Novo-Tul'skiy plant was under construction between 1931 and 1935.
3 Data not available.
4 Gosplan, Otdel gornoy metallurgiy ? promyshlennosti. Chernaya p. 22.
5 Ibid., p. 67.
6 ibid., p. 59.
7 Xbid., p. 67.
8 Ibid., p. 59.
Gosplan, Narodnokhozyaystvennyy plan na 1936 god
.National Economic Plan for 1936, M0.930W,
v. 2, 1936, p. 324.
10 Bol'shaya sovetskava entsiklopediya (Great Soviet Encyclopedia), 1st edition, Moscow,
v. 34, 1937, col. 405.
11 Computed by subtracting Kosaya Gora from the total.
12 Average of prices for 1933 and 1934.
13 Malaya sovetskva entsiklopediya (Small Soviet En:*-clopedia, 2nd edition, Moscow:
V. 10; 1940: cols. 874-878.
0
CD
0
CD
CD
-0
CD
(1)
CD
CD
0
E
CD
7:1
CD
CD
CD
CD
0
0
. .
0
7:1
0
-0
CO
Footnotes to table III-D.--(continued)
496
14 g-?:raf a v sh (Geo Moscow, 1947, no.4
15 -vomputed by subtracting Kosaya Gore from the total.
16 Malam_sares3elsed'va.(n_eyalor_ETIdia), 2nd edition.
Nbscow, v. 10, 1940, cols. 874-878.
17 Estimate. Total U.S. production of pig iron and ferroalloys, 1950,
was 19.4 million metric tons, of which Central Russia and the Ukraine produced
58 percent, or 11.25 million metric tons. (Shimkin D.E. Min-es...2,1..?...:;_j_Ka
to Soviet Power, Cambridge, 1953, pp. 36 and 39.) The Tula metallurgical
plants produced an estimated 10 percent of this latter figure, or 1.12 million
metric tons. This estimate is based upon the percentage of total volumetric
capacity of blast furnRces in Central Russia and the Ukraine in 1950 represented
by Kosaya Gora and Novo-Tul'skiy.
0
CD
CD
CD
5'
-o
CD
(/)
CD
CD
0
CD
7:1
CD
CT
CD
CD
0
_10)
0
. .
C)
7:1
0
-0
CO
0
0
0
0
0
0
0
0
0
d
Table III-E.--ANNUAL PRODUCTIVITY IN RUBLES PER WORKER (RAHOOHIY)
FOR SELEOTEO INDUSTRIES OF THE U.S.S.R.
Year
Ferrous
metallurgy
Light
industry
Rubber
articles(general)
Brickyards
Sawmills
Furniture
Woodworking

Printing
1928
1929 1
1930. .....
1931
1932 .....
1933
1934. ....
1935
1937 .....
1938
1940..........
194116
1950
Ratio of
economically
active
worker21
4:144
24,770
35,340
45,092
5,235
5,623
7,214
59,140
1211,410
1412,545
1513,580
14,938
j 1916,147
1.53
8,344
(')
(1)
(1)
6,584
7,089
7,278
68,395
1310,629
(1)
1613,981
14,638
(1)
1.56
(1)
(1)
(1)
(1)
12,801
13,221
16,391,
(1)
(1)
(1)
(1)
(1)
2025,088
221.55
(1)
(1)
(1)
(1)
1,469
1,438
1,900
72,116
(1N(1)
(1)
(1)
(1)
(1)
231,60
(1)
(1)
`cc
(1)
4,994
4,639
5,321
65,743
(1)
(1)1
(1)
(1)
241.57
(1)
(1)
(1)
(1)
5,714
6,139
7,542
98,523
(1)
(1)
(1)
(1)
251.57
4,087
(1)
(1)
(1)
4,358
4,531
5,267
106,000
137,370
(1)
179,05
10,613
(1)
1.57
(1)
(1)
(1)
(1)
4;518
4,955
5,594
116,132
(1)
(1)
(1)
(1)
(1)
261.5
1 Data not available.
2 Tsentral'noye upravleniye narodno-khozyaystva ucheta gosplana SSSR, Sotsialisticheskove
stroitel'stvo soyuza SSR z 1933-1938gg (Socialist Construction of the U.S.S.R., 1933-1938),
Moscow, 1939, p. 36, and TsUlEhU, Sotsialisticheskoye stroite1'stvO1Socia11st Construction),
Moscow, 1935, p. 169.
3 TsUNKhU, Sotsialisticheskoye stroitel'stvo (Socialist Construction), Moscow, 1935, p. 169.
Ibid.
5 Computed as follows:
(a) Increase in employment in ferrous metallurgy, 1934-1935 = 373,600 to 386:200 . 1.0337
(Gosplan, Otdel gornoy metallurgiy ? promyshlennosti, Chernaya metallurgiya
zhelezorudnaya, margentsevaya I koksovaya promyphlennost SSSR. Statisticheskiy
s ravochnik 1928-1934 ? (Ferrous Metallur v and the Iron Ore Man anese, and Cote
industr Statistical Handbook 1928-1934 Moscow, 1935, p. 117.)
of the U.S.S.R. A
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Footnotes to table III-E.--(continued)
Footnote 5 cont'd.
(b) Increase in output: 1934-1935 - 31 percent (Gosplan, Narodnokhozyaystvennyy plan na
1936 god (National Economic Plan for 1936), 2nd edition, Moscow, 1936, p. 107.
(c) rhus, index of productivity, 1934-1935 - 1.31 =
1.2673.
(d) 1935 productivity . 7,214 x 1.2673 . 9,140.1.0337
6 Linear interpolation, 1934-1940.
7 Arithmetical extrapola-ton, 1932-1934.
6 An average of two estimates: an arithmetical extrapolation, 1932-1934; and fitting a
curve to rates for general woodworking productivity.
') An average of two estimaies: an ,rithmetical extrapolation; 1932-1934; and an increase
of 13.9 percent; 1935 over- 1934, for wooduorking general.
1? Interpolated by fitting curve to data for 1932-1934 and 1940.
1 Arithmetical extrapolation, 1932-1934.
12 Linear interpolation, 1935-1938.
13 Interpolated between 1935 and 1940.
14 Computed. Equal to 223.1 percent of the 1933 rate (TsUNKhU, Sotsialisticheskoye
stroitel'stvo soyuza SSR, 1933-1938gg (,S'oc,.ialist Construction of the U.S.S.R., 1933-1938),
Moscow,- 1939, p. 387:
15 Computed. 1941 . 110.2 percent of 1940 (Gosudarstvennyy plan razvitiya narodnoao
khoz-raystvo SSSR nE.2.1_.941_12d ;State Flan for the Development of the National Economy or the
U.S.S.R. in 1941), Reprint No. 30 by the American Council of Learned Societies, Universal
Lithographers, Inc,, Baltimore, Md., n.d., p. 516).
16 Computed. 1941 . 104.7 percent of 1940. (Ibid., p. 523.)
17 Computed. 1941 = 112.6 percent of 1940. (Ibid., p. 521.)
18 1941 rates are from Gosudarstvennyy plan......na 1941 god, pp. 512-566.
19 Computed. Index of productivity for Group "A" industries, 1940-1950, excluding machine-
building, is estimated at 1.189. Applying this to 1940 prcductivity in ferrous metallurgy
gives the 1950 estimate.
2? Computed. Index of productivity for Group "A" industries, 1932-1950, is estimated at
1.96. Applying this to 1932 productivity gives the 1950 estimate.
CD
0
Sr.)
CD
-0
m-
CD
0
E
CD
6,
7:1
CD
CD
CD
0
0
0
7:1
co
cio
o
0
0
0
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Footnotes to table ITI-E.--(continued)
21 These ratios are based on tables in Gosudarstvennyy plan......na 1941 god,
pp. 512-566. The -uotal economically active includes all persons (ralochiye,
sluzhashchiye, and engineering-technical personnel engaged both in production and
ancillary functions) employed by a specific Commissariat. The term worker refers
here to rabochiye engaged solely in production carried on by the Commissaria..
22 An average of the ratios for the Commissariats of Chemicals and of Light
Industry.
Figure is for the Commiscariat of Construction Materials; Republic Industry.
24 Figure is for Commissariat of the Lumber Industry.
25 ro d
26 An arbitrary ratio, assumed in the absence of other data.
Source: Unless otherwise noted, rates have been computed from Tsentralinoye
upravleniye narodno-khozyaystva ucheta gosplana SSSR, Sotsialistiches-
koye stroitellstvo SSSR (Socialist Construction of the U.S.S.R.),
Moscow, 1936, table 2, it737-57ig).
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Table III-E-1.--ANNUAL PRODUCTIVITY IN RUBLES PER WORKER (RABOCHIY) FOR SELECTED
METAL-WORKING AND MACHINE-BUILDING INDUSTRIES OF THE U.S.S.R.
Year
I Total Miscellaneous machine-building and metal-working
machine- 1
building 1 ';:extile
(including machinery
repair)
L--
Transport
machinery
Agriculture
machinery
Machine
tools
Other
t:al
me -
working
Average
1928....
1932 eeeeee erne"
1933
1934
1937....,CCOO4D"
1938.............
Ratio of econom-
ically active
worker ...... ....
.), c- 1
,61a 0
k )
5;126 I 5;534
6,070 6,101
7,085 i 7,262
28289 / ,i,
1
210.698 ;1
',-)
411,909 1 (1)
51.56 (1)
(1(\
5,653
6,494
7,63e.,
(1)
(1)
(I)
(i)
5,260
6,482
6,625
(1)
(1)
(1)
(1)
.14
k ;
5,250
6,082
7,793
(1)
1(1)
(1)
(1)
(1)
4,736
5.798
6.810
(1)
(1)
(1)
(1)
5,287
6.191
_7;25
8,453
(1)
(1)
61.56
1Data no available..
2interpolated between 1934 and 1938.
3Rate of increase, 1934-1935, assumed to be the same as that calculated for .ortal
machine-building.
4Reported as 196.7'2 percent of 1933. Tsentral'noye upravleniye narodno-khoayaystva
ucheta gosplana EiSSR, Sotsialisticheskoye s-t,roitel'stvo soyuza SSR, 1933-12;321ria
(Socialist Construction of the U.S.S.R., 1933-19307 Moscow, 1939. p. 38
5Computed from Gosudarstvennyy plan razvitiya narodnogo khozyaystvo SS3R na 1941 god
(State Plan for the Development of the National Economy of the U.S.S.R. in 1941),
Reprint No. 30 by the American Council of Learned Societies: Universal Lithographers, Inc.,
Baltimore: Md., n.d., ID. 520. The rate is for the total economically active in the
Commissariat of General Machine-Building to the number of workers employed in the industry.
6This ratio is assumed to be the same as the ratio for total machine-building.
Source: Unless otherwise noted, rates have been computed from Tsentralinoye upravleniye
narodno-khozyaystva ucheta gosplana SSSR; Sotsialisticheskoye stroiteltstvo
SSSR (Socialist Construction of the U.S.S.R.): Moscow: 1936, table 2; pp. 3-18.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Table III-E-2.--ANNUAL PRODUCTIVITY IN RUBLES PER. WORKER (RABOCHIY) FOR
SELECTED FOOD INDUSTRIES OF THE U.S.S.R. --------
Year
Total
industry
Meat
Flour
Milling
Bakeries
Sugar
refining
Distilleries
Macaroni
Factory
kitchens
1928
13,082
(1)
(1)
(1)
(1)
(1)
(1)1
(1)
1931
212,100
(1)
, (1)
(1)
(1)
(1)
(1)
(1)
1932
11,773
15,412
28,831
1 14,306
6,450
41,240
11,148
8,443
1933
11,973
20,730
H 33,241
13,974
6,726
41,081
12,245
8,715
1934 ......
13,057
24,349
32,059
15,730
7,533
38,000
12,000
10,549
1935
314,364
428,196
(1)
(1)
(1)
(1)
(1)
(1)
1937 ......
316,978
(1)
(1)
(1)
(1)
(1)515,600
(1)
1938
618,283
(1)
(1)
(1)
(1)
(1)
(1)
(1)
1940
722,130
h
(1)
(1)
(1)
(1)
(1)
(1)
1941 ......
824,033
(1)
(1)
(1)
(1)
(1)
(1)
(1)
Ratio of
economically
active/worker.
92.19
(1
(1)
(1)
(1)
(1)
(1)
(1)
1 Data not available.
2 Interpolated between 1928 and 1932.
5 Interpolated between 1934 and 1938.
4 Reported as 15.8 percent increase over 1934 Gosplan, Narodnokhozyaystvennyy plan na 1936
god (National Economic Plan for 1936), 2nd edition, Moscow, 1936, p. 184.
' Rate of increase, 1934-1937, assumed to be the same as that calculated for total food
industry.
6 Reported as 52.7 percent over 1933. Tsentral'noye upravleniye narodno-khozyaystva ucheta
gosplana SSSR, Sotsialistieheskoye stroitel'stvo soyuza SSSR, 1933-1938gg. (Socialist Construc-
tion of the U.S.S.R.; 1933-1938), Moscow, 1939, p. 38.
7 Computed. 1941 rate reported as 108.6 percent of 1940. Gosudarstvennyy plan razvitiya
narodnogo khozyaystvo SSSR na 1941 god (State plan for the Development of the National Economy
of the U.S.S.R. in Reprint No. 30 by the American Council of Learned Societies,
Universal Lithographers, Inc., Baltimore, Md., n.d. p. 524.
8 Gosudarstvennyy plan.....na 1941 god; p. 524.
9 Ibid. Ratio is that of total economically active of the Commissariat of Food to the number
of workers employed in the food industry.
Source: Unless otherwise noted, rates have been computed from Tsentralinoye upravleniye
narodno-khozyaystva ucheta gosplana SSSR, Sotsialisticheskoye stroiteltstvo SSSR
(Socialist Construction of the U.S.S.R.), Moscow, 1936, table 2, pp.3-18.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043k6000060008-1
Table III-F.--CTVILIAN ECONWICAtTY ACTIVE POPULATION IN TRANSPORTATION
IN TU1A, KASAYA GORA, AND NOV0-TUL2SKIY: 1935-1950
Year
1935...
1937...
1939...
1950...
IFeonorni,cqlly !
population
Shipir.snts
index 2--t Volume- indexNumberl
(1935 = 100) kin meLric tons) (1935 100)
(1)
11,000
14,000
15;000
21,000
(2) (3) (4)
100 11927;000
126 3;318;500
135 1 31690,100 1
192 6,327,000 1
100
172
191
328
'Productivity
1 (U.S.S.R.)
Index
(1935(5)100)
100
4137
5142
6171
3.
The figure for 1935 was estimated by analoy from the Moscow urban
Oblast pattern (ab).e III-G). Figures for the other years were computed
by using the indexes in column 2.
2 These indexes were derived as the quotient between the indexes in
columns 4 and 5.
Estimates of the volume of shipments for each year are comprised of
two elements: (1) an estimated shipment of five tons of materials for
each ton of pig iron produced in the area; and (2) a total of all other
shipments for the area based on an estimated 500,000 tons in 1935. This
latter figure was increased in direct proportion to the population
increase for each of the later periods.
4 This figure is based on a productivity series computed from employment
data (see Lorimer, Frank, The Population of the Soviet Union, Geneva; 1946,
pp. 219-221) and production data -576-e Grossman, G., and Shimkin, D. L.,
Mineral .Consumption and Economic Development in the United States and the
Soviet Union (a supplement to Shimkin, D.B., Minerals: A Key to Soviet
Power, Harvard University Press, Cambridge, 1953), Russian Research Center,
Harvard University, April 1952, p. 58).
Tzterpolated between 1937 and 1950.
6 Computed from production data (domestic ton-miles) in Grossman, G.
and Shimkin, D. B., loc. cit. and from employment data in Fortune, v. 471 no. 2;
February 19531 p. 119; and Shimkin, D. B.; "Russia's Strength Today,"
Automotive Industries, 1 October 1953 and 15 October 1953.
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 ? CIA-RDP81-01043R0020000e000R1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002066060008-1
Table III-G.--CIVILIAN ECONOMICALLY ACTIVE POPULNTION AND VALUE OF PRODUCTION
BY INDUSTRY FOR TULA7 KOSAYA GORA AND NOVO-TU;SKIY:
10 41TNE 1.935
(Numbers are rounded to the nearest thoul:and ;:nd percentages, computed froia
unroundsd ''sta, to thc m-nrost tanth, .-_juctm,-,nt to g*.cup totals
which are independently rounded)
Category
Economically active Value of production
Number Percent (in million rubles)
Total 113,000 100.0
Manufacturing ...... pocregoalsof40 73,000 64.9
Ferrous metallurgy............ 4,000 3.8
Metal-working and machine-
building .......... ......... 54,000 47.9
Food ...... 0000G0000OLosoaCC000 6,000 5.1
Wood, construction materials.. 2,000 2.2
Apparel ..... 000000000?eopop.4 5,000 4.6
Miscellaneous...............,. 1,000 1.3
Ccmmerce........................... 11,000 9.3
Transportation and communication... 11,000 9.7
Construction ..... . coawJeeo0OG000000 7,000 6.1
Health .. ........................... 3,000 2.8
Education ....... ............... . ... 4,000 3.2
Government administration and
party apparatus .... ... U4000.0000 4,000 3.5
Other ..... 0.0.6,1?0?00000?00e.oblik apro 1,000 O.
Source: See sectl.on II-A of this appendix.
402.8
25.8
291.1
37.9
10.5
30.2
7.3
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
IV - 19
II. Detailed Procedure
Estimates of the economically active population; with distribution
by industry and category, were made for 1935, 1939, and 1950. Each of these
is discussed below.
The area included in the primary estimates consisted of Tula,
Kosaya Gora, and Novo-Tulfskiy. This choice was determined by the greater
availability of data for the area as an economic unit. These primary
estimates were then adjusted to obtain estimates for Tula city and the Tula
metropolitan area as a whole.
Throughout this entire analysis it has been assumed that no changes
took place in those economic sectors for which evidence of changes, direct or
indirect; is lacking.
A. 1935 economically active population (table III-G).
1. Total.
The total economically active for 1935 was obtained by applying the
1 April 1936 ratio of economically active to total population, or
51.5 percent; to the estimated 1935 mid-year population. This gave
a total of 112,000, which is increased to 113;000 by the addition
of an estimated 1,000 as an approximation of Party and NOD personnel
excluded from the 1936 survey.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
4
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
IV - 20
2, Mhnufacturing.
The number of persons employed in manufacturing was estimated by
using direct employment and production data. The procedure, by
industry, was as follows-
E. Ferrous metallurgy. This represents a simple division of
value of output (table III-D) by annual productivity per
rabochiy (table III-E), multiplied by the ratio of total
economically active to rabochiy (table TTI-E).
b. Metal-working and machine-building. The value of output in
1935 has been calculated in section B-2 above, while data on
productivity and total economically active to rabochiy ratios
are given in table TIT-P.
C. Food. Changes in the economically active in this industry,
1931-1935, were fairly complicated. Employment in sugar
refining and distilling was left unchanged in the absence of
evidence to the contrary. Employment in the meat kombinat
and flour mill; which were added to the industry during the
period, was calculated on the basis of national urban consumption
levels of the two products,5 and the 1935 population of the area.
The number of persons employed in bakeries and factory kitchens
was computed by Obtaining a rate of increase over erployment in
1931 which was based on increases in total population and in
Ap?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002600060008-1
Iv-21
total .wp:.oymer.r. in manufacturing; respectively, divided by
estimated increases in annual productivity.
d. Apparel, wood and construction materials, and nrintinF.
No change in employment in these industries has been allowed
for the period 1931-1935, in the absence of concrete evidence.
e. Miscellaneous. An arbitrary factor was used to cover employ-
ment in "Miscellaneous" industries. This constituted a
number equalling five percent of the aggregate estimated
employment in food, apparel, wood and construction materials,
and printing.
The total value of industrial production, obtained by summing the
various industries, was 402.8 million rubles. This total was
within the lower limit of the area's production as given in the
Soviet source ,6 or 400-500 million rubles. A value at the lover
limit seemed reasonable, as another source listed the total value
of industrial production in 1937, two years later; as "above 400
million rubles."7
3. Nonmanufacturing.
The economically active population in nonmanufacturing categories
was obtained by subtracting employment in manufacturing from the
total. As mentioned previously, this residual was distributed in
accordance with the pattern of employment in the urban part of
Moscow Oblast (excluding Moscow city) in 1936.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
-
Iv - 22
B. 1939 economically active population (table III-4)-
1. Total.
The total economically active population for 1 January 1939
was estimated as shown in table TII-H. This part of the
population grew by some 291000 persons, 1935 to 1939.
2. Manufacturing.
a. Ferrous metallurgy. Direct information on prewar ferrous
me-callurgical output is available only through 1937; for
this year, the estimated employment in ferrous metallurgy
(using the productivity and economically active to worker
ratios of table III-E) is 7,000. Since the output of pig
iron in the U.S.S.R. as a whole increased only 3.5 percent
during the period 1937-1940,8 and there were no facilities
known to be added at Tula, it is assumed that neither
output nor employment showed any measurable increase at
Tula between 1937 and 1939.
b. Metal-working and machine-building. The approach used in
estimating this component was to gauge the size of the
residual to the economically active population of Tula, 1935
to 1919, after all other increases (e.g., ferrous metallurgy
and transportation) had been subtracted from the total
increase of the economically active population. This
resulted in a figure of 64,000, an increase of 10,000 over
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
"V;
1, 7
A
? IV - 2,
1935 (tables ITT-G and III-4). An apprc,Yimate check of this
estimate of 64,000 was than made. It .ras assumed that the
value
in it.ecal-working culcilatcil for 1937 (156
million ],6-1927 ruble::) covered nonmilitary 1,1ants only.
At 193 productivity rates, this would correspond an
ezploymen of some 22,000 persons. (The 1937 productivity
of 10,700 rubles per rabochiy, given in table II1-E-1, was
multiplied by 1.05 to correspond to the product-mix at Tula.)
A strong indication of the size of employment in
military metal-wcrking is provided by data on the workers'
club at the Machine-bui1dia,7; Plant- (sic) at Tula in 1939.
This club had 45 rooms plus an auditorium with a capacity of
1050, almost exactly the same as in the workers' club in the
ZIS plant at Moscow .9 The latter had an emplow..ent of
42,000 persons in 1937, and msobably sor:.ehat mo...'e in 1939.1?
Thus, it is very likely that at leasr, 40,000 were employed
in the integrated complex of ordnance factories denoted by
the Soviet term "Machine-building Plant,? and that the
residual figure for all employment in metal-working, 64,000,
appears to be reasonable.
^
C
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Table IIIeH.--CIVILIAN ECONOMICALLY ACTIVE POPULATION OF TULAp
KOSAYA GORAp AND NOVO-TULISKIY: 1935-1939
Date
(1)
Civilian
popu lationl
1 (2)
Increase
(3)
Natural
increase2
015 x (2)1
(4)
Migrants
1
Econom-
ically
active !
ila-
poioptn
1.
I
(7)
Economically
active as
percent of
civilian
popu lation
1 (8)
%
/ i
Total ;
(3 (
i ) - 4)4
1 (5)
Eeonom-
icelly
aeLive=.
/
1 (6)
30 June 1935....I
30 June 1936....,
30 June 1937....
30 dune 1938....
1 January 1939..
1
217,125
228,474
246,924
'4651..574 j
274,600
1
(4)
11,349
18,450
18,450
9,226
1 (4)
e 3,257
I 3,427
:11704
1 1,990
0)
8,092
. 15,023
J 1,746
7;236
1 (4)
1 5;227
9;705
i 9,526
4674
4112,819
118,0461
127,751
137,277
141,951
1 52.0
51.7
51.7
/ 51.7
,
I 51,7
1 Mid-year estimates arc interpolated from data in table I-A. The estimate for 1 January 1939
is based directly on the census total for 17 January 1939, without adjustment to 1 January.
2 The natural increase rate of 1.5 percent per year is assumed. In 1927 the rate for urban
areas of the Central Industrial Region was 1.72 percent, (Lorimer, Frank, The Population of the
Soviet Union, Geneva, 1946, p. 82) but some decline after that date is assumed.
3 It is estimated that 64.6 percent of the migrants entered the econethicelly ac hive population
each year during this period. The corresponding rate for migrants entering Tula in 1926, or
54.4 percent, was obtained by applying 1926 age-specific participation rates for the civilian
nonagricultural economically active population of Tula gubernia to the migrants who entered
Tula in 1926. Tsentral,noye statisticheskoye upravleniye SSSR. Otdel perepisi. Vsesovuznaya
perepis nasaleniya 1926 goda (All-Union Census of Population, 1925), Moscow; v. 19, 1929,
pp. 256-257, and v. 36, 1930, pp. 208-209.
4 This rate for 1926 wae increased to the 1935-1939 period by assuming the same rate of
increase as for the increase in the ratio of the U.S.S.R. urban economically active to U.S.S.R.
urban population between 1936 and 1939 (Section II-0 below). The rate of 64.6 percent is that
estimated for 1937, the mid-year of the 1935-1939 period.
I
I
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
IV - 25
Table ECONOMICALLY ACTIVE P0PULA1I0N OF TULA,
KOS= GOA, AND NOVO-TUL'SKIY! 1 aANUARY 1939
(Numbers are rounded to the nearest thousand and percentages, computed from
unrounded data; to the neares-:, tenth, without adjustment to group totals
which are indepenaeutly u.d)
Category
Economically active
Number 1 Percent
Total 142,000 ' 100.0
Manufacturing ...... oo0000o.?00000uobe.o.40o..0? Ov 90)000 63.5
I
g
o
Metal-working and machine-building 64,000 7,000 4.7
J 45.4
Food...... ..... ............... ...... i 8,000
1 5.5
Wood, construction materials
Apparel......... 0000000.4.04000000464000000
.4.00600%.0000000 ..... 006.4 2,000 t:70
2,000
Miscellaneous... 7,000
1.2
Commerce ....... ..........???....?...?.......???? 13,000 1 9.3
Transportation and communication ... .. 0.00600%04 15,000
110.8
Construction ..... 0000004.0.00006000 06040000600ew 8,000 5.5
Education.. 490000"0060000440000000000...00004.00 i 1
5,000 3.2
Health .. .... 00000.446.0000G000000000460000000400. 4,000 2.8
Government administration and party apparatus 6;000 4.4
Other .......... 044.0.000?60000G00...
1 , 1 000
-
_____J
0.5
Source: See section II-B of this appendix.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
0.?
IV - 26
4.4 percent for the category, in order to adjust for a
probable increase in employment.
c. All other nonmanufacturing categories were held at the same
percentage of the economically active as in 1935. Thus, in
effect, their estimated size increased at the same rate as
the population.
C. Validation of the 1939 estimate.
Three lines of evidence may be advanced to validate the estimate
detailed above.
An hypothetical economically active population for Tula,
Kosaya Gora, and Novo-Tultskiy may be constructed by taking the data
for age- and sex-specific civilian labor force participation rates
for the urban part of Tula gubernia in 1926 (data on the city
directly are lacking) and projecting them to 1939. This projection
is based on the following considerations. in 1926, the urban
population of the U.S.S.R. (26,314,114 persons) included 571,102
military personnel or 90.5 percent of the total armed forces strength
of the U.S.S.R.11 Of the civilian urban population, 17,976,832 were
aged 15 or above; 9,401,099, or 52.3 percent of the civilian urban
population over 15, were economically active. By 1939, the urban
population of the U.S.S.R. had increased to 55,909,908.12 Of this
number, at least 3 million (or 85 percent of the total estimated
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
7.:c. ? :7'4i 771-
-
IV - 27
armed forces strength) were military, leaving 52,900,000 civi1ians.1?
The age-structure of the U.S.S.R. changed relatively little between
1926 and 1939, with the percentage 15 and above increasing only from
62.8 to 63.8 percent; Or a relative increase of 1.6 percent. Applying
this rate of increase to those above 15 gives an estimated urban
civilian population 15 years of age and older of 37,513,000. The
civilian urban economically active population in 1939 is estimated
to have totalled 23,115,000 (Lorimer's national total of 28,539,000
workers and employees, less a deduction of 7;524,000 for rural persons
in agriculture, handicrafts, trade, education, health and administra-
tion, plus an addition of 10 percent for the probable under-counting
of domestics, casual labor, urban forced labor, etc.).14 This figure
comprises 61.6 percent of the estimated urban civilian population
15 years of age and older, which represents an increase of 17.8
percent (61.6 52.3 c 1.178) in civilian participation in the
economically active population.
In 1926, 61.5 percent of the civilian population 15 years
of age and over in the urban portions of Tula.gubernia was economi-
cally active.15 Adjusting this factor by the coefficient of increase
in participation in the economically active population as calculated
above gives 72.4 percent. The estimated population of Tula, Kosaya
Gora, and Novo-Tul'skiy aged 15 and over in 1939 is 199;700.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
e-
'
IV -23
This gives an estimatPd economically active population of 145,000,
or 2.0 percent above the figure estimated directly (table III-4).
A second check may be made by adjusting the 1939 economically
active participation rates by sex. For the civilian males aged 15
and over, two exclusions maybe rade: a deduction of 6.6 percent of
the total, or 6,460 for disability and old age, and one of 5;847 or
6.0 percent for students.16 The resulting estimate is a male civilian
economically active population of 85,500; or 87.4 percent of all
civilian males aged 15 and over. If this total is subtracted from the
aggregate of 142,000 (table III-J), the estimate for employed females
becomes 56,500, or 55.6 percent of all females aged 15 and over.
Against these figures we may set estimates based on the proportions
of female employment in each industrial category in the U.S in
1938-1940 (table IIT-K): 85,000 males and 57;000 females -- virtually
identical.
A final check may be made by comparison with detailed data
on the economically active population of Izhevsk in 1936, which also
is a city primarily devoted to the manufacture of munitions. The
comparative structures as percentages of total employment are the
following.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
IV - 29
PERCENTAGE OF TCTAL EMPLOYMENT
iTula, Kosaya Gore, i
Category and Novo-Tul'skiy Izhevsk
(1939)1 1 (1936)2
Total. 100.0 1 100.0
;
6.'
1
Manufacturing ........ ...00p?? i 63.5 63.5
Commerce............. .. ..0
.. 1 9.3 10.7

Transportation and
communication 10.8 1 6.2
Construction ........ 1 5.5
1 4.9
Health............ ........ 1 2.8 3.4
Education 3.2 1 6.0
Government administration 1 0
and party appa7atus ..... 1 4.4
1 3.9
Other....................,
/ 0.5
1.3
See table
2 Data from Tsentral'noye upravleniye narodno-khozyayst-
vennogo ucheta gosplana SSSR. Otdel ucheta trudo.
Chislennost 3. zarabotnaya plata rabochikh I sluzbashchikh
v SSSR (Numbers and Wages of Workers and Employees in the
U.S.S.R.), Moscow, 1936, pp. 170-171.
The most notable element in the comparison, of course,
is the high percentage of the economically active population in
both cities which is occupied in manufacturing. The greater
significance of Tula as a transportation center is reflected in
the higher percentage in that occupation. Likewise, the
relatively small percentage of Tula's economically active popula-
tion which is engaged in education is shown by the comparison.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
.?"
?
IV-30
Table III-K.-CIVILIAN EOONOADMLIY ACTIVE POPULATION OF
TULA, KOSAYA GORA: AND NOVO-TULISKIY, BY SEX:
1 JANUARY 1939
Category
.........
ve.onomically active
i Female 3
1 Percent
/
female'
Total
Male2
Total
Manufacturing .......
Ferrous metallurgy...... ow
Metal-working and machine-
building .....
Food ... ............
Wood, construction materials
Apparel.... ..... 0000*a..oece
Miscellaneous ......... ,
Commerce .. ......... 40?CooDeder0?00i
Transportation and communication
Construction
Health ......... Go0Opeo00090060060
Education...... *O000.0poycp?eopv.
Government administration and
party apparatus
Nher .. ....... p00040u0?40?0.**"
142,000
85,000
t
1
.1_57,000
33,000
2,000
i 19,000
I 4,000
1,000
6,000
1,000
7,000
6,000
3,000
3,000
3,000
2,000
(11)
1 5(39.2)
5(35.6)
23.9
29.7
47.2
643.8
82.4
42.1
752.0
836.5
938.0
1076.0
3.058.0
1035.0
1038.0
90,000
7,000
64,000
8,000
2,000
7,000
_
2 nO0
,
5
13,000
15,000
, 8,000
4,000
5,000
6,000
1,000
57,000
5,000
45,000
4,000
1 000
_,
1,000
1 000
- ,
6,000
9;000
5,000
1,000
2,000
4,000
1,000
1 Total economically active taken from table III-J.
2 Male economically active Obtained by subtracting female from the total.
3 Female economically active Obtained by applying the percentages from
column 4 to unrounded figures for the total.
4 Figures in this column are from Syssoeff, B.; Women's Participation in
the Labor Force of the U.S.S.R. (MSS), Russian Research Center, Harvard
University, 1950, pp. 13-16. Unless otherwise noted the percentages are for
1 January 1938.
5 These percentages were computed from =sounded totals obtained by
summing the female economically active calculated for the various categories.
6 An average of silicates and ceramics (43.4); and wood and sawmill (44.2).
7 This percentage is for 1940, and is an average of municipal services
(67.0), and trade (37.0).
8 This percentage is for 1940, and is an average of railroad transport
(25.0), and communication (48.0).
9 This percentage is for 1940. No rate for construction is given.
10 These percentages are all for 1940.
11 Less than 500. ?
^a.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
?
a?q.
ASV
Iv - 31.
D. 1950 economically active population (table III.-L).
The total economically active population for 1950 was obtained by
summing the estimates for all the separate categories. It is partially
dependent upon the estimate for the total population, for several of
the industries and categories were increased in the same ratio with
the population, but three-quarters of the total increment was computed
from production data independent of population growth.
1. Manufacturing.
a. Ferrous metallurgy. Once again information on production in
this industry served as a basis for estimating a large part
of the industrial economically active population. Using the
volume of production and productivity rates from tables
III-D and -E, and assuming a rate of increase in value of
production, 1935-1950; equal to the rate of increase in volume
of production for the same period (i.e., a constant price);
employment in this industry was computed for 1950.
b. Metal-working and machine-building. Employment in this industry
was estimated by means of the following equation:
B1935-1939
where
=
*1939-1950
A1939.-1950 A1935-1939
A = increase in employment in ferrous metallurgy
B G increase in employment in metal-working and
machine-building.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12: CIA-RDP81-01043R002000060008-1
_J
- '37
Table ECONOMICALLY ACTIVE POPULATION OF TULL
KOSA1A GERAs AND N0V0-TUL7SKIY: 1 jANUARY 1950
(Numbers are rounded to the nearest Ihouc_aL:7. r7'' percentages, cocputed from
unrounded data, to the rteare3t tenth, without adjustment to group *otals
which are independently rounded)
Category
Total..... ..............
? ? ? 0 0
Manufacturing .......... ...........,...
Ferrous
Metal-working and machine-building
a ? ? ?
.......
? ? ? 0 0 ? ?
Food.....................................
Wood, construction materials.
. .
Apparel ? .. ???4040?0GC 00?000*???
Miscellaneous................
Comerce?........ ..... 0....
Transportation and communication..
Construction ......... ...........c.
V..
??0
?4C
? 0 c ? ? 0 C ? ?
.... ?
Ot4????00
1.000004.4
???00?00c
^C ..... CC
Health ...... 4.0.1..004000000?0000.0.1400?00600.
Education ..... ........... ........... .....
Government administration and party apparatus.
Other?. ........ 000C0000000900u0eso.00000000.
Economically active
Number Percent
1 170,000 100.0
1
1
106,000
10,000
74,000
8,000
3,000
9,000
2,000
16,000
21,000
9,000
5,000
5,000 1
7,000
1,000
62.5
5.6
43.8
4.8
1.7
5.5
1.1
9.1
12.3
5.4
2.8
3.1
4.3
0.5
Source: See section II-D of this appendix.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
rP
-
Table III-L0-4IVILIAN ECONOMICALLY ACTIVE POPULATION OF TM
KOSAY.A GORA1, AND NOVO-TUI1SKIY! 1 ;JANUARY 1950
(Numbers are rounded to the nearest IhoutaL.:1 n7,4 pErTcQutages, cocputed from
unro,paded data, to the noarelt tenth, without adjustment to group -xtals
which are independently rounded)
Category
Economically active
Number
percent
Total
170,000
100.0
Manufacturing.. ......... ..... ....
106,000
62.5
Ferrous
10,000
5.6
Metal-working and machine-building
74,000
43.8
Food......... ....,..........
8,000
4.8
Wood, construction materials
3,000
1.7
Apparel........ UVe*C00040000d ....... Oa"
9,000
5.5
Miscellaneous.. C????006.4.0000OCCO See.00?00
2,000
1.1
Cu-mere? ..... v??? 0000?0??
16,000
9.1
Transportation and communication.. ??It ? ??0?Ibt Pr
21,000
12.3
Construction ......... ?dt it ? ..... 00
9,000
5.4
Health ...... 0?0?00?.00000t00?0?0?000,1?0?CO?00.
5,000
2.8
Education ......... ? ?.? .....................
5,000
3.1
Government administration and party apparatus.
7,000
4.3
Other.... ......... 00000000?0?4040????G0C .....
1,000
0.5
Source: See section II-D of this appendix.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
i.?
c. All other manufactures. As noted in section I-B-3, it has
been estimated that t.:2T--y:2- in all categories other than
ferrous metallurgy and machine-building increased 19.5 percent
beteen a-f 1937. 7hib grwtJ. ta.3 iistribued
by eategor. 'Ala. the period, 1937-1939. a period marked by
increasing concentration on ordnance production, it has been
assumed that increases in the consumers goods sector were
limited to a minimum in food (bakeries and factory kitchens)
and miscellaneous. The figures for 1939 are, therefore, those of
1937, increased by 1,000 each in those two categories.
3. Nonmanufacturing.
Several categories of the nonmanufacturing group underwent
significant changes between 1935 and 1939.
a.
Transportation and communication. The greatest change in
the economically active in this category occurred between
1935 and 1937, in connection with the rapid growth of
metallurgical production. The increase in employment
between these years was computed in table III-F. The growth
between 1937 and 1939 was much less because of a leveling
off in metallurgical production (table III-F).
b. Government administration and party apparatus. On 26 September
1937; Tula became the center of the newly-formed Tula Oblast.
The share of the category government administration and party
apparatus in the total economically active population was
arbitrarily increased by one percent of the total, or to
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Iv - 34
This gLves fiEure of 74,C00. veri Ication
this figure, i.e., an estimated increase of 10,000 in metal- _
workirg and machine-building employment, is given in many
postwar references to increased production in the Kirov Machine
Tool slant, the Novaya Tula Machinery Plant No. 1872 and the
Zars Motor VehirAe, Repair Plant, as well as the large munitions
factories.
c., volood and construction materials. While direct daa on an
expaision in rh s category Pre lacking, the size.P..ble growih
?
inc cement of 1,000 persons.
d, ApTarel, P.Pcdacdon -In this industry has expandPd markedly
in the postwcx period. Estima.:e:3 of the number of persons
employed lrse. IR factort=s, ar; on r-porled pro6uoion
(af0e froL. U. ;'.? '.4x1 a
f'n o'
enployment ir :,he now rubber rToducts plant (which has been
'erouped with anpLrel since the main product is rubber footwear)
aas made by use of reported production (table Prices
cbtained from an early Soviet source-3 were applied to obtain
value of production, and the number of persons employed was
then computed by use of annual productivity rates (table III-E).
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
?,PM
l? ?
?
IV - 35
e. Food and miscellaneous industries. In the absence or data
indicating expansion, no change was made in the 1939 estimate
for employment in food and miscellaneous industries.
2. Nonmanufacturing.
The number of persons occupied in all of these categories, except
transportation and communication, was increased in direct proportion
to the growth of total population for the period 1939-1950.
Transportation and communication was given an increment for this
period based on the growth in both population and production.
Data for this calculation are shown in 'Lane
E. Validation of the 1950 estimate.
In view of the absence of official population figures Or
other firm demographic landmarks for 1950, no absolute verification
of the estimates is possible. Nevertheless, it is possible to check
the consistency of the estimate of the economically active population
with the estimates of total population and age-sex structure for Tula,
Kcsaya Cora, and Novo-Tuliskiy as an aggregate. For 1950, the
estimated number of civilian males 15 and above in this group of
communities was 111,960, and the number of females 131,317. The
economically active participation rate for civilian males in the Tula
area in 1939 was already close to the maximum possible without a
curtailment in education. Consequently, this rate of 87.4 percent of
civilian males 15 and above may be applied in 1950.
?
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
r50
17,
IV - 36
For the female population the 1939 rate of 55.6 percent was
also high, although not quite maximum. However, since the character
of the city has remained strongly oriented toward industries both
requiring considerable physical strength, e.g., metallurgy, ,and
providing high wage rates -- and hence incentives -- it appears
unlikely that the proportion of women in employment would have increased
markedly. For this reason, the 1939 economically active participation
rate has been used for the female population also. The results are the
following: the estimated civilian economically active population is
171,000, or 0.6 percent higher than that calculated directly. The
proportion of female employment in tha total shows a rise from 39.8
percent in 1939 to 42.7 percent in 1950.
F. Expansion and redistribution of the 1950 economically active population
for the city and metropolitan area of Tula.
1. Kosaya Gora and Novo-Tultskiy.
The estimated population of these settlements is 49,000. Assuming
conservatively that the percentage of the population which was
economically active here was the same as for Tula, Kosaya Gora,
and Novo-Tulcskiy combined; gives an estimated economically active
population of 26:000. Of this; 10,000 is known to be in ferrous
metallurgy. A rough estimate of 500 in metal-working and 500 in
all consumers goods industries combined may be added to give a
total of 11,000 in manufacturing.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part- Sanitized Copy Approved forRelease2013/07/12 : CIA-RDP81-01043R002000060008-1
;:itor
W-37
It was assumed that manufacturing occupations comprised
64.9 percent of the economically active population, as was the case
in 1935 (table III-0). Accordingly, the nonmanufacturing
occupations were assumed to comprise 35.1 pe Tent of the economi-
cally active population, or about 5,000 persJns. Use of this
relatively high proportion of the economically active population
in manufacturing for the workers settlements appears justified in
view of the certainty that educational, administrative, and
simillr service facilities would be less developed there than in
the central city. This calculation leaves a residual of 10,000,
which is the estimate of workers and employees living in Kosaya
Gore and Novo-Tullskiy, but working in Tula city.
2. Tula city.
This estimate iu a simple arithmetic residual from the combined
Tula, Kosaya Gore" and Novo-Tultskiy figures after the removal of
the estimste for employment in the two workers settlements.
3. Barsuki, Ocrelki, Khrushchevo, and Plekhanovo.
The estimated population of these communities is 15,000 (6,000 of
which is at Plekhanovo, with 3,000 each for the other three
communities) and the economically active population is 8,000.
Of this latter; 55 percent may be ascribed to wage work, the
remainder being collective farmers.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
41, .
?
Iv - 37
It was assumed that manufacturing occupations comprised
64.9 percent of the economically active population, as was the case
in 1935 (table III-G). Accordingly, the nonnanufacturing
occupations were assumed to comprise 35.1 pel:ent of the economi-
cally active population, or about 5,000 perams. Use of this
relatively high proportion of the economically active population
in manufacturing for the workers settlements appears justified in
view of the certainty that educational, administrative, and
simillr service facilities would be less developed there than in
the central city. This calculation leaves a residu. of 10,000,
which is the estimate of workers and employee living in Kosaya
Gore and Novo-Tuluskiy, but working in Tula city.
2. Tula city.
This estimate iu a simple arithmetic residual from the combined
Kosaya Gore., and Novo-Tul'skiy figures after the removal of
the estimate for employment in the two workers settlements.
3. Barsuki, Gcrelki, Khrushchevo, and Plekhanovo.
The estimated population of these communities is 15,000 (6,000 of
which is at Plekhanovo, with 3,000 each for the other three
communities) and the economically active population is 8,000.
Of this latter, 55 percent may be ascriboa to wage work, the
remainder being collective farmers.
=Z.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
17r""
.?
&V,
Iv - 38
These settlements have had the following history:
Plekhenovo has been a workers settlement since 1939; Gorelki became
a city-type settlement in 1948-1949; Barsuki was not a city-type
settlement in January 1950, the date of the estimate, but became
one at the end of the year; Khrushchevo is still classified as
rural. On the basis of the legal definition of a workers settlement
(a minimum of 400 adults; of whom 65 percent are wage earners),19
we estimate that in Plekhanovo and Gorelki 65 percent of the labor
force were wage earners. The comparable proportions for Barsuki
and Khrushchevo were estimated at 50 and 30 percent respectively.
Applying the percentages from table LI-G gives an
estimate of 1,1000 persons employed in manufacturing. As of 1935,
43 percent of the economically active population in manufacturing
was engaged in metal-working.20 In the absence of later data,
this same percentage has been used for 1950 with the remainder
being allocated to the category "miscellaneous." For nonmanufac-
turing, the percentages in table III-G were used without change.
4. The final step in allocating the economically active population
was to add the estimates for Barsuki, Gorelki, Kbrushchevo, and
Plekhanovo to the totals previously estimated (table III-L).
-V
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
IV 39
NOTES
1 Tsentral'noye upravleniye narodno-khozyaystvennogo ucheta gosplana SSSR.
Otdel ucheta truda. Chislennost' I zarabotnaya .lata rabochikh I sluzhashchikh
v SSSR (Numbers and Wages of Workers and Employees in the U.S.S.R.), Moscow,
1936, p. 184.
2 Bollshoy sovetskiy atlas mira (Great Soviet Atlas of the World), Moscow,
v. 2, 1939, plate 45.
3 nla7asoveallsovIecveloedia), 2nd edition,
Moscow, v. 10, 1939, cols. 875-876.
4 TeUNKhU, Chislennost'...op.cit., pp. 60-65.
5 The 1935 national urban consumption level of meat was calculated at 14.9
rubles (1926-1927 prices) per capita. This was based on data from TsUNKhU,
Sotsialisticheskoye stroitelfstvo SSSR (Socialist Construction of the U.S.S.R.),
Moscow, 1936, table 2; Gosplan, Narodnokhozyaystvennyy plan na 1936 god
(National Economic Plan for 1936), Moscow, 1936, p. 430; and an estimated urban
population of the U.S.S.R. for 1935 of 45 million. The consumption level of
flour was estimated at 200 1:g. per capita for 1935, and the productivity of
flour at 360 tons per rabochiy annually. These estimates were based on data
from the three volumes of TsUNKhU, Sotsialisticheskoye stroitelistvo (1935
volume, p. 279; 1936 volume, table 2; and the 1933-33 volume, p. 77).
6 Bol'shoy sovetskiv atlas miral loc. cit.
7 Malaya sovetskaya entsiklopediya, be. cit.
8 Schwartz, Harry, Russia's Soviet Egonomy, New York, 1950, p. 226.
9 Facilities of the two clubs were. as follows:
Facility
ZIS plant Machine-building
Moscow b plant, Tula
Number of paid workers
Number of "circles" . .......... 0 ? 0 ? ? ? ?
Number of participants in them
Number of rooms for club work
Number of auditorium seats......
53 30
55 20
1,671 946
41 45
1,100 1,050
Source: TsUNKhU, r12111=tstroitelistvo SSSR. Statisticheskiy
abornik (Cultural Construction of the U.S.S.R. A Statistical
Compilation), Moscow, 1940, p. 162.
1? Employment in the auto industry of the U.S.S.R. was 87/000 in 1937. The
ZIS plant in Moscow pro-Weed 48.6 percent of the total value of output.
Assuming the same productivity in the three major plants, GAZ, ZIS, and YaAZ,,
employment in the ZIS plant would-thus be 48.6 percent of 87,000. (Sources:
Shimkin, D. Bey "Labor Productivity in the Soviet Automobile Industry,"
Memorandum on Labor Re uirements for Munitions Production in the U.S.S.R.,
Russian Research center, Harvard University, Cambridge, Mass.,19 April 1950,
(CONFIDENTIAL).
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Iv - 40
NOTES--(C)nt'd)
11 Tsentral'noye statisticheskoye upravleniye SSSR. Otdel perepisi.
Vsesoyuznaya perepis' naseleniya 1926 goda (All-Union Census of Population,
1226), Moscow, v. 34, 1930, p. 3.
72 Izvestiya, Moscow, 2 June 1939.
13 The estimate of 85 percent of the military strength enumerated among
the urban population in 1939 is based on the percentage of 90.5 in 1926.
Total strength of the armed forces was estimated at 3.5 million (appendix IX).
14 Lnrimer, Frank. The Population of the Soviet Union, Geneva, 1946,
p. 100. Deductions from this total are based on Lorimer's estimates, applying
the following percentages to obtain rural employment:
iPercent rural! Source
Occupation
Agriculture and
relat:d occupations?.I 100.0
Handicrafts
Trade000.0?0.0.GOdOe000000
Education44,40004.4, OOOOO 00
Health
Administration .....
50.0 Estimate
Equal to percent of total trade
turnover in cooperatives.
(Gosudarstvennyy plan razvitiya
narodnogo khozvaystvo SSSR na
1941 god (State Plan for the
Development of the National
Economy of the U.S.S.R. it 1941),
pp. 591-592.
63.2 TsUNKhU, Kullturnoye stroitel' _
45.8
25.0
/ stvo
1 pp. 37-38; 107.
Gosudarstvennyy Rlan...1941 god,
p. 608.
Estimate.
15 Tsentral'noye statisticheskoye upravlenlye SSSR. vsesoyuznaya perepis'
goda. v. 19; 1929, pp. 256-257.
- The deduction of 6.6 percent for disability and old age is based on
U. S. ;tate- It is equal to the percentage of males, 14 years and over in 1950,
who were not in the labor force for these reasons. (U. S. Bureau of the Census.
U. S. Census of Population: 1950, v. II, Characteristics of the Population,
Part I, U. S. Summary, Washington, 1953, p. 1-247.)
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Iv - 41
NOTES--(Cont'd)
Footnote 16 cont 'd.
The deduction for students is based on data from TsUNKhU,
Kul'turnoye stroitel'stvo SSSR , pp. 26, 65, and 104.
17 U. S. Bureau of the Census. Census of Manufactures: 1947,
Washington, D. C., v. 2, 1949) pp. 207-221.
16 Universallnyy spravochnik tsen (Universal Handbook of Prices),
Moscow, 1928.
19 Tsentral'noye statisticheskoy upravleniye gosplana SSSR. Slovar'-
spravochnik pc, sotsialeno-ekonomichoskoy statistike (Dictionary-Handbook
of Social-Economic Statistics), 2nd edition, Moscow, 1948, p. 438.
all Sources: Bollshoy sovetskiy atlas mira, be. cit.; Rayony moskoygitra
dblasti (Rayons of Moscow Oblast , Moscow, 1936, pp. 4-19; TsUNKhU,
Chislennost'....op. cit, pp. 60-65; Tsentral'nyy ispolnitelenyy komitet,
Administrativno-territorial'noye deleniye soyulia SSR na 15 iyulya 1934 roda
Taministrative-Territorial Divisions of the U.S.S.R. on 15 July 1934
Moscow, 1934, pp. 26-35.
,
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
APPENDIX
URBAN AREA REPORTS OF THE FOREIGN MANPOWER RESEARCH OFFICE
U. S. BUREAU OF THE CENSUS
U. S. Department of Commerce, Bureau of the Census. UNCLASSIFIED, Estimates -
of the Population and Labor Force of Major Polish Cities: 1 January 1950,1
Tes.---T--93. 1-17s VAshington, D7-d77-ib August 1953, UNCLASSIFIED.
. CONFIDENTIAL, Estimates of the Population and Labor Force of Selected
Cities in the U.S.S.R.:--1--YanuaTiTr.9757--OeFies?P--55, Nos. la:23,
WEHIngTon, D. O., 10 December 1953, SECRET.
. CONFIDENTIAL, Estimated Population and Labor Force of the City and
Metropolitan Area of ChircEIET -17-January 195077gEFIJE-F.:55, No. 24;
Raington, D. C., 10 MarEE71954, .1EUREar.
-----. CONFIDENTIAL, Estimated Population and Economically Active Population
of the City and M-e-criranA)----Thiria7F-1 ;January 19502 Serria-7.7957
1137 23,-Raaington, D. C.9-1-0-Wrig57;;TORDTERTIAL.
. CONFIDENTIAL, Population and Labor Force Estimates for Czechoslovakia:
1 January 1950, S-e=s17:97.3710.-2-6,7771371iTitW?
ilWRET.
? CONFIDENTIAL, Population and Labor Force Estimates for Prague City
and the Prague Metrorolitan Area: 1 3anuarrig302 Series P=53; No, 27,
Wasningzon, T. G.771"--ii75-1932;778E010T.
, CONFIDENTIAL, Population and Labor Force Estimates for the City of
Plzen: 1 January _lies P-95771707-7877aMielE77157.7777------
1 juE-6-1954; SECRET.
CONFIDENTIAL, Population and Labor Force Estimates for the City of
Pardubice: 1 jam:I-EFT-1M's Series P-95, No. 29; WgEHTFiffEBr, D. C.,
June 10Z7-gEOREI.
. CONFIDENTIAL, Population and Labor Force Estimates for the City of
Brno: 1 January Ig5077gareg-P:g3? No. as ina on, . C.
:Faure 1954, SECRET:
1 (Warsaw, Lodz, Wroclaw, Szczecin, Bydgoszcz, Chorzow, Sosnowiec, Gliwice,
Poznan, Lublin, Katowice, Krakow, Gdansk, Gdynia, Bytom, Zabrzes and
Czestochowa).
2 (Ivanovo, Shuya: Kovrov, KineshmaA Kirov, and Kostroma).
50X1-HUM
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
 . CONFIDENTIAL, Estimated Population and Labor Force of the City of
Iasi: 1 January 1950, Series P-937-767-4737-gshington: D. C.: --
October 1955, Sc.
50X1 -HUM
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
V-2
. CONFIDENTIAL: Population and Labor Force Estimates for Ostrava
City and the Ostrava tiztropoKTiErkvea: 1 YEITEUTTOUTgEFIETT-95,
Ra775fTgalhamgton,-17. c.7-1-"June-19549 SECRET. ---
. CONFIDENTIAL: Population and Labor Force Estimates for the City
of Bratislava: 1 J--aiivax0 Sfi .)952-1'767-32-71iMiTiWn, i5. c
1 -TEhri 1:954;-SEMY:
-----. CONFIDENTIAL, Estimated Population a,id Labor Force of the City
and Metropolitan AFEE7317Eaiznan: 1 janu-Tnai3O-7sel937,"
ITE7.732717arsTii'nEton? D. C., 3CF August:704; STORET.
? CONFIDENTIAL, Estimated PoPulation and Labor Force of the Aleksin
Industrial Area: 1 January 1530, Series13:(75c-Bio7-35? Washington, 5. C.0
10 August NFIDENI.
-----. CONFIDENTIAL, Estimated Population and Labor Force of the Ordzhonikidze
Metropolitan Area: 1 january r975-Mi;les P-0:771F776777-EBIEgton,
137-077373 SePteni1?FPED-11,
. CONFIDENTIAL, EstilToted Population and Labor Force of the City of
Samarkand: 1 January E93T,=-Tci-EI-PT:g3T-Fk7-'37Tit.rEiFaz'-/gIor17-b-.--GT:T
2073anuary 1935, SECRET.
. CONFIDENTIAL, Estimated Population and Labor Force of the City of
Kaluga: 1 January1-9313,--Nr-Tes)---77:971,72-0--3875Eforgton; D. 0.2
207-5377055TigENE1.
o CONFIDENTIAL y Estimated Population and Labor Force of Rumania:
1 January 1950, Sair5"-Pc7571E7759, rai-arrz"-v c.on,77:77713ft or 1955 ,
gETTRET.
-----. CONFIDENTIAL, Estimated Population and Labor Force of the Metropolitan
Area and City of Bucharest: 1 Janua.vy:1795.0.7Series P:797177:7-07?
Washington., D. C.: October 1557,-SECRET.
. CONFIDENTIAL, Estimated Population and Labor Force of the City of
Cluj: 1 January 1950, Series P-07357T TEWEEEIFirtony D. C.,
OWE-65E-105p SECRET.
. CONFIDENTIAL, Estimated Population and Labor Force of the City of
Constanta: 1 January 1950: Series v-95, No. 4d: wasnington, D. 0.9
73.175b"eM53; SECRET.
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
50X1 -HUM
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1
SEM
. CONFIDENTIAL, Estimated Population and Labor Force of the City of
Timisoara: 1JanuaryT)---X7.7.17-g5TITE. 477WEahington2 D. O.
TY67655F-14537-82cREY7-----
. CONFIDENTIAL, Estimsted Po.pulation and Labor Force of the Fergana-
Nargelan Metropo1ITHE7FET-I January:7E15u: Series P-59 No
Washington,. li77-710ovemer
CONMENTIAL, Estimates of the Population and Labor Force of the
City of Kokand: 1 January 050, Series .1?-9, No7-467 Washingtofi, iI. COI
N6eE5W-19337-gETTET---
. CONFIDENTIAL, Estimated Population and Labor Force of the City and
Metropolitan Area of Ryazan' 7-1 January 03O, Series 1517932 g6:
Washington, D. Cop December 1057-SECRET,
, CONFIDENTIAL, Estimated Population and Labor Force of Aktyubinsk:
1 January 1950s Series P79), No, 4&? WaliiiigTE2D7-5:7-DecemdeT-177555s
SECREq7-------
son
Declassified in Part - Sanitized Copy Approved for Release 2013/07/12 : CIA-RDP81-01043R002000060008-1