RUBLE-DOLLAR RATIOS FOR CONSTRUCTION

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP08S01350R000602050003-2
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 68
• Document Creation Date: December 22, 2016
• Document Release Date: March 15, 2012
• Sequence Number: 3
• Publication Date: February 1, 1976
• Content Type: REPORT

Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Research Aid
Ruble Dollar Ratios for Construction
ER 76-10068
February 1976
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
This publication is prepared for the use of U.S. Government
officials. The format, coverage and contents of the publication are
designed to meet the specific requirements of those users. U.S.
Government officials may obtain additional copies of this document
directly or through liaison channels from the Central Intelligence
Agency.
Non-U.S. Government users may obtain this along with similar
CIA publications on a subscription basis by addressing inquiries to:
Document Expediting (DOCEX) Project
Exchange and Gifts Division
Library of Congress
Washington, D.C. 20540
Non-U.S. Government users not interested in the DOCEX
Project subscription service may purchase reproductions of specific
publications on an individual basis from:
Photoduplication Service
Library of Congress
Washington, D.C. 20540
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Ruble-Dollar Ratios for Construction
February 1976
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Page
Summary and Conclusions ........................................... 1
Discussion ......................................................... 1
I. Introduction ................................................. I
II. US and Soviet Data on Construction Costs ...................... 2
III. Representativeness of the Sample .............................. 4
IV. Comparability of US and Soviet Projects ........................ 8
V. Empirical Results ............................................ 10
A. The Algebra of Ruble-Dollar Ratios ............................... 21
B. Adjustments for Location and Climate ............................. 23
C. Hospitals ....................................................... 27
D. Housing ........................................................ 33
E. Office Buildings ................................................. 41
F. Schools ......................................................... 49
G. Industry ....................................................... 57
H. Transportation .................................................. 63
1. Weighted Construction Ruble-Dollar Ratios ........................ 69
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
RUBLE-DOLLAR RATIOS FOR CONSTRUCTION
Summary and Conclusions
The ruble-dollar ratios computed in this study are based on cost estimates
of construction projects in the United States and the Soviet Union. These
estimates are fairly indicative of actual US costs (except military construction),
but tend to understate costs in the Soviet construction, industry, where over-
runs-averaging 20% of estimate costs-occur in more than 90% of all
projects.
The 1970 ratios for aggregate construction are summarized below:
Based on
Estimate Costs
Adjusted for
Probable
Actual Costs
US-weighted ...................
0.621
0.746
Soviet-weighted ................
0.539
0.647
Geometric average ..............
0.579
0.695
These ratios permit a rough comparison of US and Soviet construction in 1970
as measured in rubles, dollars, and the geometric mean of the two price systems.
Apart from possible definitional differences in the construction data, the results
indicate that the value of Soviet construction was between 71% and 99% of
US construction in 1970.
A comparison of construction projects by type reflects higher ruble-dollar
ratios for industry and single-family housing than for service-oriented structures
because the Soviet advantages of using standard plans and methods are not as
readily applicable to industry and single-family housing. The relatively high
ratios obtained for transportation construction projects are caused, in part, by
the large degree of earthwork in which Soviet builders are less efficient.
Every effort has been made to guarantee that both the US and Soviet
project samples are representative of construction in the respective country
and that the project samples are truly comparable. Unfortunately, the
conflicting nature of the criteria of representativeness and comparability and
the limitations of the data probably have caused some understatement of the
true ruble-dollar ratio.
DISCUSSION
US-Soviet comparisons require the use of purchasing power parity (ruble-
dollar) ratios. Construction price ratios are needed to compare output and
productivity in a sector that is an important part of national product in both
the United States and the USSR, is a key element of economic growth, and
contributes a great deal to military programs in both countries. The ruble-dollar
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
ratios currently used in US Government studies are based on a 1964 CIA report,
ER 64-26, 1955 Ruble-Dollar Ratios for Construction in the USSR and the US,
August 1964. These ratios have been moved forward each year with price
indexes and are badly out of date.
The purpose of this study is to compute new ruble-dollar ratios based on
costs in the two countries of a sample of construction projects representing as
many fields of construction as the data permit. The definition of cost and the
nature of the cost data are discussed in section II. Section III presents the sample
of projects and discusses its representativeness in both countries. Section IV
considers the comparability of the respective construction projects in the two
countries. The final section presents the results and evaluates their reliability
and probable bias in the light of deficiencies in the estimate cost data which
are used. Appendix A presents the algebraic bases of these ratios, and Appendix
B describes the techniques for deriving the average location in both countries.
Appendixes C-H go into considerable detail about the methodology and results
for each type of construction. Finally, Appendix I shows the computation of
the US- and Soviet-weighted ruble-dollar ratios for all construction, both with
and without an adjustment for Soviet cost overruns.
For purposes of this study, the cost of construction is defined as "the
agreed selling price at which a completed project is (or would be) turned over
to the investor." 1 The costs include the contractor's overhead and profits.
Several kinds of expenditure that the investor generally considers to be a cost
are excluded, however-for example, outlays for architectural and engineering
fees, land, paving, walks, landscaping, caissons, piling, and moveable furnishings
and equipment.
The first decision to make in a study such as this is how to define a unit of
construction. Three approaches have been used in previous studies: construction
inputs, construction components, and construction projects. We have selected
the project approach as the unit of measure primarily because of the greater
availability of data .2 The costs of these projects in both the United States and
Soviet Union, are estimate costs. Theoretically, actual costs are preferable
to estimates, but such data are seldom accessible in either country.
The US data are obtained from three primary sources: The Dodge Building
Cost and Specification' Digest (hereafter referred to as Dodge Digest); The Dodge
Building Cost Calculator and Valuation Guide (hereafter referred to as Dodge
Guide); and The 1972 Building Cost File/Eastern Edition (hereafter referred to
as Building Cost File). The Dodge Digest reports specific projects and winning
bids. The building costs in the Dodge Guide are not actual ones, but rather
estimates formed on the basis of construction parameters. The Building Cost
File is used for transportation construction because the other two sources report
1 CIA ER 64-26, 1955 Ruble-Dollar Ratios for Construction in the USSR and the US, August 1964,
p. 7.
2 For a discussion of these three alternative approaches, see Norman M. Kaplan, "Some
Methodological Notes on the Deflation of Construction," Journal of the American Statistical
Association, September 1959, Vol 54, pp. 535-555.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
nothing on this type of construction. Greater detail is provided about these
sources in the appropriate appendixes.
Most of the US prices used in this study are winning bid or contractual
prices from the Dodge Digest. In cases where bid or contract prices are
unavailable, an engineering estimate is derived from various US construction
costing handbooks. (These estimates also include allowances for profit and
overhead.) All dollar amounts are adjusted to mid-1970 prices by construction
cost indexes; the bid date is considered to be the starting time for the construction
project for purposes of deflating costs to a 1970 base.
The proper price basis for comparing Soviet and US construction is the
actual cost of completed projects, but these costs are hard to find for either
country. In the United States, however, the contract bid price is not
substantially different from the final price in private (non-governmental)
purchases of construction.
For the USSR the data are far less accessible. The typical development of a
construction project, according to both Soviet technical literature and private
sources, is as follows. After a project is proposed, a design organization prepares
a design and an estimated cost 3 If the project is then approved and included in
the economic plan, a construction organization is assigned to the project. In
more than 90% of all projects, the construction organization finds that the initial
cost estimate is too low. The design and estimate cost must then be revised.
Typically, this is the first of many cost revisions in a sizable construction project.
The revisions reflect both real design changes and cost changes arising from
overexpenditure of funds as compared with construction norms. These over-
expenditures may occur as a result of higher-than-expected prices of materials,
greater-than-expected difficulties, managerial inefficiency, inordinate delays,
and the like. By and large, almost all of these costs are covered by successive
revisions of cost estimates, so that at completion of the project the final cost
estimate is near to the actual cost. If the final estimate is less than the actual
cost, the difference is absorbed by the construction enterprise (from profits of
other projects) or if necessary by the construction ministry involved.
The construction enterprises report the cost of work completed in final
estimate prices in each reporting period. These reports, when summed by the
Central Statistical Administration, are reported as the total value of construction
in the USSR. Thus, final estimate costs apparently would be the proper and
consistent prices for comparison with the announced total of construction
activity. These costs are detailed initial engineering estimates amended by the
actual construction cost experience.
Neither the initial cost estimates nor the final revised estimates with over-
runs included are available. The alternative of making detailed cost estimates
from cost handbooks for a substantial number of projects is far beyond the
resources available in this study. As a substitute, use was made of a set of
handbooks (sborniki) that the Central Statistical Administration used in estimat-
ing the replacement cost of Soviet capital stock in the capital stock census of
1 January 1972 and 1973. The handbooks give simplified formulas for cost per
Estimated cost in the Soviet context is defined as the cost of construction if the project were to be
carried out strictly according to state norms and at established prices.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
cubic meter for structures (a) designed for various purposes (for example, offices,
hospitals, and warehouses); (b) of various specified sizes; (c) built with various
construction methods and materials (for example, brick and block, steel frame,
or reinforced concrete); and (d) situated in various locations (for example,
remoteness of material supplies and climate).
The editions of these handbooks have been published sporadically since
1970, and the costs, which are based on the prices introduced on 1 January 1969,
are assumed to be 1970 prices. These series are published in multivolume editions
which bear the formal titles: K sbornikam ukrupnennykh pokazateley vosstano-
vitel'noy stoimosti zdaniy i sooruzheniy dlya pereotsenki osnovnykh fondov (Hand-
books of Consolidated Indices of the Replacement Cost of Buildings and
Structures for the Revaluation of the Fixed Capital Stock) and K sbornikam
ukrupnennykh pokazateley vosstanovitel'noy stoimosti zdaniy i sooruzheniy,
imeyushchikhsya v uchrezhdeniyakh i organizatsiyakh, sostoyashchikh na gosudar-
stvennom byudzhete (Handbooks of Consolidated Indices of the Replacement
Costs of Buildings and Structures in the Institutions and Organizations In-
cluded in the State Budget). When individual citations are needed, these books
are referred to as Sborniki, Investment Series, and Sborniki, Budget Series,
respectively, but are collectively called Sborniki.
III. Representativeness of the Sample
The projects to be costed must be selected so that the samples represent
each country's construction patterns. The problem is to establish criteria for
deciding whether the samples collected are representative. Four criteria of
representativeness are: project function, location, size, and construction charac-
teristics. The construction projects used in this study do not satisfy all of these
criteria, but we have tried to minimize the problems involved.
The ruble-dollar ratios for construction are based on a sample of 277
projects representing a broad cross section of construction. These observations
have been grouped according to project function and ratios computed for each
group (see the appendixes for a detailed description of the sample and results,
category by category). This sample is described in the following manner:
Type of Project
Number of
Observations
Percent of Total
Observations
Hospitals ......................
51
18.4
Housing .......................
69
24.9
Office buildings ................
58
20.9
Schools ........................
53
19.1
Industrial facilities .............
32
11.6
Roads .........................
4
1.5
Airfields .......................
5
1.8
Railroads ......................
5
1.8
Functional Representativeness
To represent construction from the standpoint of function, the categories
of construction for which ruble-dollar ratios are computed in this study should
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
mesh with the categories used in each country's published aggregate data. More-
over, within each category the construction projects selected for the sample
should be representative of that category. There are major deficiencies arising
from unavailability of data. The importance of the deficiencies is discussed below.
The US Department of Commerce publishes statistics of new construction
put in place by type of project-residential, industrial, commercial, religious,
hospital, educational, military, road, railroad, communications, power, and sun-
dry types of water resource projects. The ratios computed in this study can be
matched to construction categories representing 74% of US construction. The
neglected categories include the following4:
Share of Total Value
of Construction in
1970
(Percent)
Total ......................................
25.7
Religious buildings .........................
1.0
Nonresidential agricultural construction .......
0.9
Telephone and telegraph construction.........
3.2
Electric light and power construction .........
6.2
Gas construction ...........................
1.7
Petroleum pipelines .........................
0.2
Military facilities ...........................
0.8
Conservation and development ...............
2.0
Sewer systems .............................
1.6
Water supply facilities ......................
1.2
Other public buildings ......................
2.8
Miscellaneous public construction ............
2.2
Miscellaneous private nonresidential building...
0.9
Residual private construction ................
1.0
The ruble-dollar ratios for the omitted US construction categories probably
are not far different from the ratios computed for the categories represented in
the sample. Religious buildings are excluded because churches are not currently
being built in the Soviet Union. Telephone and telegraph construction contains
items such as exchange buildings, office buildings, maintenance buildings,
repeater stations for microwave, microwave towers, telephone line construction,
etc. We believe that the ratios for this construction, except perhaps microwave
facilities and telephone line construction, lie within the range of the industrial
construction ratios. Considering the relatively greater US sophistication in com-
munications technology, the ruble-dollar ratio for the above exceptions would
probably be higher than for the rest of the sample. Fortunately, this type of
project has a small weight in US construction.
Electric light and power construction contains items such as all types of
powerplants, transmission lines, dams, cooling towers, and reservoirs.5 Most
of the projects included in this sector are similar to industrial or highway
construction. Petroleum-related construction contains industrial-type work
such as pipelines, wells, pumping stations, and gas production, distribution, and
storage facilities.
4 US Department of Commerce, Domestic and International Business Administration, Bureau
of Competitive Assessment and Business Policy, Construction Review, May 1973, pp. 10-11.
5 A few thermal powerplants were compared in a rough manner. The tentative results suggest
the ratios for thermal powerplants lie within the range of the industrial ratios.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Construction of military facilities contains a potpourri-airfields, roads,
housing, missile silos, and industrial facilities. Conservation and development
construction contains projects such as erosion control schemes and irrigation
facilities. Costing this type of construction is difficult, but it is unlikely that the
resulting ruble-dollar ratios would be grossly atypical of other construction.
We omit sewer systems and water supply facilities from the sample because of
the problems involved in identifying comparable Soviet and US items.
How representative the samples are of Soviet construction is difficult to
determine. Aggregative Soviet construction data arranged by function are pub-
lished only when combined with other types of investment-e.g., investment
in machinery. Thus, functional weights for construction and installation work,6
which in 1970 amounted to approximately 61% of Soviet investment, must be
independently estimated as is described in Section V.
We estimate that the sample categories provide good coverage of Soviet
construction except for agricultural construction and construction within the
construction industry itself. These omitted sectors, however, constituted about
19% of Soviet construction in 1970.
Nonetheless, their ratios probably are not atypical. Construction performed
for the construction industry is similar to some industrial construction in that it
includes structures such as warehouses, garages, sheds, and the like. Hence,
ratios for this construction should lie within the range of the ratios already
computed. The USSR publishes data on four categories of agricultural invest-
ment, although not solely for construction. The livestock raising category, which
constituted approximately 4% of Soviet investment in 1970, contains construc-
tion of barns, silos, and grain storage facilities. These ratios probably are similar
to the industrial ratios because they reflect similar types of structures. The
water resources category of investment (3.7% of 1970 investment) includes
projects such as dams, canals, and pumping stations. These ratios are likely to
be akin to the ruble-dollar ratios for road construction because they both require
large quantities of earthmoving work and concrete laying activity. A third
category of farm investment-electrification-represents less than 1% of Soviet
investment and probably is even less significant in Soviet construction activity
because such investment has an especially high equipment content. The fourth
category of agricultural construction includes the establishment of gardens,
vineyards, and other perennial plantings; it was only 0.5% of 1970 investment
and is not taken into account.
Representativeness of Location
Location cannot be ignored in ruble-dollar ratio studies because it helps to
determine both the utility and cost of a project. There is no national market for
construction as there is for many manufactured products; instead, a nation's
construction industry is a composite of many local and regional markets.
These decentralized markets are a result of imperfection and immobility
in both input and product markets. For example, construction materials are
frequently produced and consumed in local markets because transportation
costs represent a large percentage of value. Labor costs vary regionally by
6 Soviet investment data almost always lump the cost of construction and installation work
together. Henceforth, when we speak of official Soviet statistics, the word construction is our
abbreviation for construction and installation work. It is important to stress, however, that the
estimate costs derived from the Sborniki exclude the value of installation work.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08S01350R000602050003-2
substantial amounts because of the degree of unionization and because of the
refusal of workers to move from their homes to a location where wages are more
lucrative. Climate is another variable explaining regional differences. Finally,
when a building is constructed, it usually cannot be transported to another
location where it provides greater utility.
Ideally, the way to obtain a locationally representative ratio is to place the
whole construction sample for each country at a place where the costs are average
for that country, and this is what we have tried to do. (For a description of the
procedure, see Appendix B.) As long as observed cost variations within each
country are purely the result of different input prices, there is no problem in
comparing average prices. The use of average building costs in each country
may cause the projects to lose some comparability in ways undetected by the
coarse parameters of the data. For example, the average climate in the USSR is
harsher than our own, and buildings erected in Siberia might contain more
insulation and embody other minor structural changes to better withstand the
climatic rigors in ways which would influence costs. It is difficult to discriminate
between the regional variations explained by physical differences in construction
and those that are merely a result of different input prices. However, since
other sources of bias in this study-e.g., differences in quality between US and
Soviet construction-run in the opposite direction (that is, they lead to an
underestimate of the ratio), it would represent an offsetting error.
Representativeness of Project Sizes
Within the sample, projects should be representative in terms of size to the
extent that the ruble-dollar ratio is a function of project size. This study does not
ensure that samples containing representative project sizes are collected. Two
approaches are used to deal with project size within many of the functional
categories of construction. One approach employed in defining a US-weighted
sample uses a quasi-random sample of projects from US construction. This
results in a distribution fairly characteristic of US sizes, although some larger
projects have had to be excluded because of the absence of a Soviet counterpart.
In contrast, the second approach, which attempts to use a sample representa-
tive of Soviet construction for computing dollar-ruble ratios based on Soviet
weights, considers the size phenomenon by creating a sample stratified by size
as measured by the number of stories. Purchasing power parity ratios are com-
puted for both the largest and smallest possible physical construction volumes
for a structure with a given number of stories. Subsequently, the average of
these two ratios is accepted as the ratio for that type of construction. Although
this procedure is imperfect, the amount of error should be small if the ruble-
dollar ratio varies with size in an approximately linear fashion. Not much more
can be done to eliminate this methodological deficiency in the Soviet sample
because the size distribution of construction projects is unknown.
Representativeness with Respect to Physical Aspects
To be representative, projects in the sample should reflect typical physical
characteristics in terms of inputs, components, and technical specifications. In
this respect, for example, a typical Soviet building would probably have five
stories, a concrete exterior, a concrete frame system with a minimum of steel,
and no elevators or air-conditioning. A typical US building would probably be
taller, with a steel frame structural system, brick walls, elevators, and air-
Declassified and Approved For Release 2012/03/15: CIA-RDP08S01350R000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08S01350R000602050003-2
conditioning. Obviously, such extreme differences make it impossible to compare
a completely typical Soviet building with a typical US building and yet maintain
complete comparability. As a result, the need arises to trade-off between these
dual requirements-i.e., comparability and representativeness in studies of
construction ruble-dollar ratios.
In the approach based on a quasi-random sample of US construction, a
sample fairly representative of US construction has been derived. Another
approach forms a stratified sample of either Soviet or US construction by com-
puting ruble-dollar ratios for various structural categories-for example, grouped
by number of stories or the presence or absence of elevators-and then averages
these ratios by the use of weights based on estimated construction values.
The sampling procedure in this study excludes US construction projects
that have no Soviet counterpart and Soviet construction projects that do not
have a US counterpart. The US construction excluded by this procedure is gen-
erally ornate and luxurious; the omitted Soviet projects are usually primitive,
lacking in social amenities, or made from materials unique to the USSR. The
value of Soviet construction unrepresented in the samples probably has dimin-
ished in importance over time, but the excluded US construction probably is
growing more important.
In summary, the criteria for a representative construction sample are dif-
ficult to satisfy precisely. In selecting a sample, we have concentrated on function,
location, size, and physical characteristics. The procedures we have used are
shaped in important ways by the data that are available. The sample appears to
represent US construction quite well and Soviet construction less well.
IV. Comparability of US and Soviet Projects
In addition to being reasonably representative of construction in each
country, the construction sample should be selected so as to ensure comparability
between the US and Soviet projects included in the sample. The need for com-
parability covers six criteria: function, size, structural type, inclusion of ameni-
ties, quality of work, and aesthetics. Although this study has tried to deal with
the first four of these aspects, the remaining two-quality and aesthetics-are
much more difficult to handle.
Functional Comparability
To be comparable functionally, the projects compared should be used for
the same purpose. For example, American schools are compared with their
Soviet counterparts. Occasionally, a problem may arise about the appropriate
breadth of function. For example, should the function of schools be defined as
"a building in which educational activity occurs," or should a distinction be
made between primary and secondary educational facilities? Questions like this
have had to be resolved on an ad hoc basis.
Size Comparability
A second aspect of project comparability pertains to the size of the projects
being compared. Size could be measured by the project's output capacity or by
its physical size. In other words, we could compare projects that provide the same
volume of goods or services, regardless of construction differences, or we could
compare projects that occupy the same physical space as measured by volume,
area, or length.
Declassified and Approved For Release 2012/03/15: CIA-RDP08S01350R000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Physical dimensions rather than output capacities are the standard of
comparability in this study for both pragmatic and theoretical reasons. Data
on construction volume are much more readily available than capacity data, and
the concept of plant capacity is often too nebulous to define precisely in any
event. Moreover, capacity differences between two buildings identical in con-
struction often result from different equipment and methods of operation. For
example, a Soviet plant generally has less productive potential than its US
counterpart of the same size because of inept management, less productive
labor, and inferior machinery.
Structural Comparability
According to the third criterion of project comparability, the types of
construction should be comparable where possible. Hence, major structural
components, such as foundations, frames, roofs, and the like, should be equiva-
lent. In most cases the data are too limited to permit complete adherence to this
principle, but where possible we sought to achieve equivalency. For example, a
Soviet school made with bricks is compared with a like US school. In cases
where complete material comparability is lacking, the closest substitute was
selected. In some cases the construction method costing the least may be dif-
ferent in the two countries (i.e., brick in the United States versus precast con-
crete in the USSR). If no suitable substitute exists, the project was not used in
the sample. As two buildings are examined in more detail, many physical dif-
ferences become readily apparent, but the data on which the cost estimates
are based, both in the United States and the USSR, are not sufficiently detailed
to permit precise comparison.
The fourth criterion of project comparability stipulates that the two
projects should possess the same utilities and conveniences, such as heating, air-
conditioning, electricity, plumbing, and built-in equipment. These items may
collectively be referred to as amenities. A problem arises because Soviet and
American practices differ with respect to amenities. For example, few Soviet
buildings have air-conditioning, and few new American buildings lack it. Con-
struction specifically intended for human occupancy-hospitals, apartments,
offices, and schools-contains more amenities than facilities designed for indus-
trial and transportation purposes. Therefore, the differential for amenities
between the United States and the USSR is probably largest for the service-
oriented structures and lowest for transportation and industrial construction.
Many amenities are small items, but taken together they may add greatly
to a project's final cost. For example, Soviet apartments have fewer electrical
outlets, kitchen facilities, and miscellaneous amenities than their US counter-
parts. These minor differences cannot be quantified in the cost comparisons,
because data disaggregated to the necessary extent are seldom available.
Two types of amenities, built-in equipment and air-conditioning, have been
singled out for special treatment. In most of the US data, expenditures for
built-in equipment are separable and are removed to enhance the comparability
with Soviet construction. US data on the percentage of total cost represented by
air-conditioning for a certain type of building are available in a limited number
of cases. The average percentage for each building type is assumed as the stand-
ard air-conditioning component in all similar buildings and the costs are reduced
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
accordingly to remove the air-conditioning. This correction is discussed individ-
ually by construction sector in the appendixes. In theory, this approach only
goes halfway in the treatment of the problem. The Soviet buildings should also be
costed with air-conditioning and the built-in equipment in order to make them
comparable to the more elaborate US counterparts, but this procedure is
impossible with existing data.
Quality Comparability
The fifth aspect of comparability focuses on the quality of work and what
quality implies for subsequent maintenance outlays. There can be no doubt that
the quality of the average Soviet structure is inferior to that of the average US
structure. Not all of these qualitative differences imply, however, that the life-
span of a structure is necessarily shorter or that the utility provided by the
project is diminished. Many of the apparent incomparabilities affect only
appearance and reflect what each culture is willing to tolerate.
Not all these quality differences are even directly attributable to the con-
struction process itself, but instead reflect a different attitude toward main-
tenance expenditures. There exists a trade-off between construction and main-
tenance costs. For example, an inferior paint job during construction means the
walls must be repainted sooner than if the original job had been thorough.
Shoddy electrical wiring during construction means the wiring must eventually
be ripped out and re-installed later at a greater cost. The poor quality of Soviet
construction suggests that a portion of maintenance costs should really be
allocated to construction costs to permit valid comparisons.
Qualitative problems of comparability are difficult to handle empirically.
The best we could do was to arrange the comparisons so that the highest quality
Soviet construction was compared with average and poor quality US construc-
tion. Although this procedure is quite arbitrary and does not really eliminate
the problems, at least the correction moves in the proper direction.
Aesthetic Comparability
Aesthetic appeal also affects comparability. An attractive building provides
more social utility than an ugly one; an attractive factory may even increase
worker productivity by improving morale. Unfortunately, this factor cannot
be considered in a study of this nature, because aesthetics cannot be quantified.
V. Empirical Results
Value Weights for the Aggregate Construction Ratio
As mentioned earlier, US value weights for different types of construction
rely on the reported value of construction put in place as published by the Depart-
ment of Commerce. Corresponding ruble-dollar ratios could be matched with
74% of the value of construction reported in 1970, although some problems were
encountered. For example, the computed ratio for hospital construction is
probably distorted and unacceptable, as discussed later, so it has been replaced
by the geometric mean of the ratios for housing, office building, and school
construction. Also, we were unable to determine an independent weight for the
ruble-dollar ratio for airfield construction. Yet another problem was that the
commercial category reported by the Department of Commerce includes other
commercial buildings in addition to office buildings-e.g., stores and restau-
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
rants-but we have assumed that the ruble-dollar ratio for this other commercial
construction is equivalent to the ratio for office buildings.
Considering these difficulties, the US weights are presented below:
Category of Construction Weight
Hospitals ............................................ 0.048
Housing ............................................. 0.471
Commercial .......................................... 0.140
Schools .............................................. 0.093
Industry ............................................ 0.101
Highways and streets ................................. 0.143
Railroads ............................................ 0.004
The Soviet value weights must be derived in an indirect manner because of
the general absence of construction data based on a functional distribution.
Even the category closest to construction-i.e., the capital investment data-is
disaggregated into only seven rather gross classifications: (1) industry, (2) agri-
culture, (3) transportation and communications except railroads, (4) railroad
transport, (5) construction industry, (6) housing construction, and (7) con-
struction of trade and communal enterprises, forestry enterprises, and institu-
tions of science, culture, art, education, and health. Additional sources permit an
estimate of the amount of construction and installation work for all of these
categories except categories 6 and 7. They are estimated by computing the
residual of total construction minus the estimated value of construction for the
other five sectors displayed in the capital investment data. By using the residual
and the reported capital investment for these two sectors, an estimated
proportion of construction to capital investment is derived by assuming that
this proportion is equal for both residual categories. The derivation of these
weights is given in Table 1.
Unfortunately, these categories fail to mesh exactly with the ruble-dollar
ratios computed in this study, so allowances must be made. The industry, rail-
road transport, and housing construction sectors match up well with the ratios.
The composition of the transportation and communications other than railroads,
however, is rather murky, so we have arbitrarily assigned the geometric mean
of the ratios for construction of roads and airfields to this category. The trade
and communal enterprises category is equally imprecise. Presumably, construc-
tion in that category is largely schools, office buildings, hospitals, and stores.
Arbitrarily, we have assumed the appropriate ratio for this category is the geo-
metric mean of the ratios for office buildings and schools. We would have included
hospitals too, but as discussed later, we reject the hospital ratios because of
problems in attaining true comparability.
Unadjusted Computations
Ruble-dollar ratios calculated for eight major categories of construction
and the weighted ratio for all construction appear in Table 2. (For specific
findings relating to each category of construction, see the appropriate appendix,
and for derivation of the aggregate ratio, see Appendix I.) For some types of
construction, the methodology and available data permit estimating both a US
and Soviet value-weighted ratio. By a weighted ratio, we mean that a sample of
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Capital
Investment
(Billion
Rubles)'
Share of
Investment
by Construction
and Installation
(Percent) 2
Construction
and
Installation
(Billion
Rubles) 3
Share of
Construction
and Installation
by Function
(Percent)4
Weight of
Construction
and Installation
by Function
(Percent) 5
Total .................................
82.053
61
50.298
100.0
100.0
industry ...........................
29.567
50
14.784
29.4
36.2
Agriculture ........` .................
14.152
60
8.491
16.9
....
Transportation and communications
except railroad ....................
5.494
45
2.472
4.9
6.0
Railroad transport ..................
2.314
45
1.041
2.1
2.6
Construction industry ...............
3.008
31.3
0.942
1.9
Flousing construction ................
13.439
82
11.022
21.9
27.0
Construction of trade and communal
enterprises, forestry enterprises, and
institutions of science, culture, art,
education, and health ..............
14.079
82
11.545
22.9
28.2
I Narodnoye khozyaystvo SSSR 1922-1972, pp. 326-327.
2 The share of investment allocated to construction and installation work for industry, agriculture, and transport and communications
is obtained from Statisticheskiy yezhegodnik stran-chlenov soveta ekonomicheskoy vzaimopomoshchi 1973, pp. 172-179. It is assumed that
the shares for railroad transport and all other transport and communication are identical so an additional category may be used.
The estimated share for investment in the construction industry is based on data for the 1966-70 period from V. D. Belkin, Ekono-
micheskiye izmereniya i planirovaniye, p. 205. The share for housing and the enterprises and institutions category is obtained as a
residual of unallocated construction in column (4) and by assuming that both categories have the same share devoted to construction.
3 For all categories except the last two, the entry represents the product of column (2) and column (3). The last two items are
residuals to allocate remaining construction within the constraints of the control total derived from Narodnoye khozyaystvo SSSR,
1922-1972.
4 Column (4) = 50.298.
5 Column (5) _ 81.2. This reallocates the weights so they sum to 100.
Summary of Ruble-Dollar Ratios by Type of Construction
Type of Construction
US-
Weighted
Ratio
Soviet-
Weighted
Ratio
Geometric
Mean
Ratio
Hospitals .....................................
0.347
0.276
0.309
Housing ......................................
0.620
0.494
0.553
Single-family ................................
N.A.
N. A.
0.646
Multi-family ................................
0.586
0.454
0.516
Office buildings ................................
0.508
0.451
0.479
Schools .......................................
0.583
0.501
0.540
Elementary .................................
0.552
0.485
0.517
Intermediate ................................
0.617
0.518
0.565
Industry .....................................
N.A.
N.A.
0.601
Light ......................................
N.A.
N.A.
0.628
Heavy .....................................
N.A.
N.A.
0.576
Roads ........................................
N.A.
N.A.
0.790
Airfields ......................................
N.A.
N.A.
0.695
Railroads .....................................
N.A.
N.A.
0.836
Aggregate construction .......................
0.621
0.539
0.579
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
individual project ratios is combined in such a way that it is representative of
the construction of that type in the country and serves as the source of weights.
The most obvious observation from Table 2 is that, in every case where a
distinction is possible, the Soviet-weighted ratio is less than the equivalent US-
weighted one? This accords with the theory that a country has a relative
advantage in producing its own mix of output. As a result, in international
comparisons there is a frequently observed negative correlation between relative
prices and quantities.
Of the ratios presented in Table 2, the railroad ratio is the highest and the
hospital ratio the lowest. The railroad ratio makes sense because railroad
construction requires a great deal of earthwork in which the Soviets are at a
relative disadvantage. Furthermore, the Soviet construction industry uses steel
much less intensively than its US counterpart. On the other hand, we do not use
the hospital ratio for what we believe are compelling reasons.
Discussions -with architects suggest the comparability problem is particu-
larly severe in hospital construction and that the comparisons in this study suffer
from it. The United States has experienced a technological revolution in hospital
design in the past 10 to 20 years. The USSR has failed to keep pace with this
and now lags behind the United States by more than 10 years.
For example, US hospitals have sophisticated equipment and delivery
systems-food, medicines, laundry, oxygen, etc.-that are conceptually insepa-
rable from the structure itself. These systems are frequently stored in what is
nearly a full floor between floors called interstitial space, so that a five-story
hospital may really be comparable to a ten-story building. Interstitial space
affects both construction costs and maintenance costs, raising the prior and
lowering the latter. Meanwhile, this concept is hardly known in the USSR,
where their hospitals often lack even crawl space.
Part of these differences in US and Soviet hospitals is explained by different
levels of knowledge; another part reflects different rates of substitution between
labor and machinery in operating a hospital, with the Soviets utilizing a lower
capital-labor ratio. Nevertheless, these technological differences are apparently
not avoided by removing the cost of built-in equipment from the US hospitals
as we have done.
Thus, in computing the US-weighted ruble-dollar ratio for construction,
we used the geometric mean of the US-weighted ratios for housing, office build-
ings, and schools in place of the discredited hospital ratio. We suspect that once a
sample of truly comparable Soviet and US hospitals stripped of equipment is
formed, the ratio would probably not vary much from other service-oriented
structures. Using the actually computed hospital ratio-0.347-instead of the
substitute-0.568-would have lowered the aggregate ratio from 0.621 to 0.611.
Ruble-dollar ratios for industrial facilities are higher than the ratios for
service-oriented structures other than single-family housing. Service-oriented
buildings are more homogeneous in the USSR than in the United States because
Soviet plans are more standardized. The construction of standardized facilities
in turn permits realization of substantial cost economies. For example, the
standard plans for service-oriented structures permit a high degree of prefabrica-
7 This relationship, called the Engel-Gerschenkron effect, is not inevitable, as Samuelson points
out in the "Analytical Note on International Real-Income Measures," The Economic Journal,
September 1974, pp. 595-608.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
tion of construction components, a practice the Soviets frequently refer to as the
"industrialization of construction." This industrialization is practical only as
long as construction output is in standardized components.
Industrial construction, on the other hand, does not adapt as well to the
use of standardized forms and components, because factories are by nature more
heterogeneous than service-oriented construction. Also, the Soviet construction
industry is fraught with many difficulties and inefficiencies in the planning and
actual execution of construction. To a certain extent, these problems which
impact on cost are less pervasive when the same type or a similar project is built
over and over. Therefore, these potential cost savings are not garnered in indus-
trial construction as they are in the more repetitive service-oriented construction;
as a result the ruble-dollar ratio for industrial construction is larger than for
service-oriented structures.
Ratios for transportation construction also exceed the ratios for service-
oriented facilities. Construction of transportation projects and of buildings is
fundamentally different. Typically, US construction firms specialize in either
category, but not both. Transportation construction requires fundamentally
different machinery and technology. Also, it is potentially more capital-intensive
than other construction, so greater economies of scale can be realized. A US
construction. organization typically has more equipment at its disposal than its
Soviet counterpart. Moreover, Soviet construction machinery is typically smaller
and less productive than US counterparts. All these factors permit the US
construction industry relatively greater efficiency in building transportation
projects than in service-oriented structures.
Adjusting the 1955 ruble-dollar ratios computed in the previously cited
CIA study to a 1970 base by deflating for both Soviet and US price changes
permits a comparison of the relative productivity changes in construction during
the period 1955-70 (see Table 3). The aggregate construction ratio increased by
nearly 14% over this period, suggesting that, despite the rampant inflation in
the US construction industry, its productivity has increased more rapidly than
that of its Soviet counterpart. The number of observations in the earlier study
Hospitals ...................
1970 Ruble-
Dollar Ratios
0.31
1955 Ruble-
Dollar Ratios
....
1955 Ratios
Updated to 1970
Prices via
Price Indexes 2
....
Housing ....................
0.55
0.66
0.48
Office buildings .............
0.48
0.60
0.44
Schools .....................
0.54
....
....
Industry ...................
0.60
0.74
0.54
Roads ......................
0.79
1.00
0.73
Airfields ....................
0.70
1.02
0.75
Railroad ...................
0.84
0.78
0.57
Aggregate ..................
0.58
0.70
0.51
------- -------------
1 CIA ER 64-26, 1955 Ruble-Dollar Ratios for Construction in the USSR and the US, August 1964.
2 In 1967, Soviet construction and installation work equaled 33.6 billion rubles in 1955 prices and
41.5 billion rubles in 1969 prices, or an increase of 23.5%. The Department of Commerce Composite
Cost Index grew 68.9% during the period 1955-70.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
is too limited, however, to permit meaningful conclusions to be drawn regarding
specific types of construction, except that in all categories except airfields the
adjusted 1955 ratio is less than the corresponding current one.
Several explanations can be offered for this decrease in Soviet efficiency
relative to the US. Rapidly escalating wages for US construction labor have
stimulated contractors to substitute capital (i.e., construction machinery) for
labor. Meanwhile, the Soviet construction labor force remains substantially
less productive than its US counterpart and suffers from a chronic shortage of
skilled labor. Soviet labor's cheapness relative to capital, however, precludes
much substitution of capital for labor. Even when the substitution is made, the
results are frequently unsatisfactory. Unless the construction machinery is
imported, which has been the case for use on important projects, the machinery
has less capacity than its Western counterpart and is unreliable. Shortages of
both spare parts and the mechanics necessary to repair and maintain the equip-
ment also add to the difficulties of capital-labor substitution.
Adjustment of the Ratios
The ruble costs used in this study are estimated costs derived from the
Sborniki. Two questions must be answered in determining whether use of these
costs is permissible. First, the relationship between these Sborniki costs and the
detailed estimate costs, the first revised estimate that construction enterprises
provide themselves at the beginning of a project, must be determined. Second, the
relationship between the first revised estimates and actual costs must be con-
sidered. Unfortunately, the evidence for these two comparisons is fragmentary
at best.
Moreover, the Soviet distinction between standard (tipovoy) and nonstand-
ard (individual'nyy) project designs is important. Standard designs are those
already completed and approved by the state and are recommended for use
where possible. They represent minimum cost and quality and are designed to
capture the maximum economies of scale. Standard designs are available for
almost any kind of structure. According to one Soviet investment specialist,
they were used in recent years in structures representing 81% of the overall
volume of construction, including 97% in housing construction, 68% in industrial
construction, 87% in cultural and service construction, 88% in transport con-
struction, and 95% in agricultural constructions However, other sources say
that standard construction is used predominantly only in apartment construction
(but even in apartment buildings, that may not apply for enterprise-sponsored
housing or most cooperative housing). In industry, these other sources state that
standard designs are seldom used, although the use of standard design compo-
nents for many structures perhaps explains the allegation that 68% of industrial
construction is of standard design.
The difference in cost that the use of standard and nonstandard construc-
tion can make is impressive. A source familiar with urban construction per-
formance states that nonstandard construction of apartments or other urban
8 V. P. Krasovskiy, "The Investment Process and How to Improve It," Ekonomika i
organizatsiya promyshlennogo proizvodstva, No. 1, 1975. Also in JPRS 65240, Economic Affairs,
16 July 1975.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
cultural or commercial buildings is frequently two and one-half to three times
more expensive than construction based on standard designs. In part, this is
due to real design changes, which would also be more expensive in the United
States. But the higher cost also is due partly to higher prices and lower
productivity. Industrial construction may be worse in this respect. The implica-
tion is that a comparison of the costs of nonstandard design for a given
structure in the USSR with costs of a comparable structure in the United
States would be substantially less favorable to the USSR than a comparison
based on costs of standard designs.
The question of the validity of Sborniki costs as a measure of estimated
costs seems to have been answered in an article by Yu. I. Malimanov of the State
Committee of the Council of Ministries of the USSR for Construction Matters
(Gosstroy).9 The Sborniki supposedly are based on consolidated estimating
norms. These consolidated norms were developed to determine the construction
cost of buildings and structures that are individually designed but with "standard
design solutions and standard structure and members." During the development
of these consolidated estimating norms, Gosstroy compared the consolidated
estimates with estimates derived from the standard price handbooks. Gosstroy
approved only the consolidated estimates as norms if the deviations from the
price handbook estimate were less than 2%. For this reason, the Sborniki costs
probably lie fairly close to the detailed estimate costs.
As for the relation between the detailed estimate cost and actual costs, the
ubiquity and size of construction cost overruns are the subject of endless tirades
by Soviet economists and' planners. For example, Krasovskiy says:
. . . making use of the current system of estimated funds, capital
construction in essence has what may be termed an open bank account
and, without any great difficulties, can exceed the initial project costs
by 1.5 to two or even four times. No similar situation exists in any
other sector of material production in the Soviet Union io
But remembering that an initial revision of estimate cost frequently is required
to correct a deliberately and unrealistically low cost design, we should discount
the upper end of the range of overruns given above. Krasovskiy 11 cites an esti-
mate by the USSR Construction Bank that estimate costs rise 10% every year
(for given projects). Thus a large-scale project that takes eight years to complete
may have cost 43% more than the initial working estimate 12 This slippage is
9 Yu. I. Malimanov, "Improve Estimating and Price Setting in Construction," Ekonomika
stroitel'stva, No. 4, 1975, pp. 3-10.
1O Krasovskiy, op. cit.
11 Ibid.
12 Assume equal annual amounts of construction as measured by physical volume of work-i.e.,
12.5% per year.
Let: P = the total cost of the project;
I= the annual rate of increase in estimates; and
t=the number of the year.
Thus, the true cost of the project will be:
7
P=E 0.125 It
t=0
Where there is no "inflation"-i.e., 1= 1.0-the cost of the project equals unity (the original
estimate cost). Where estimates creep up yearly, the total cost is 1.429, or (0.125+0.138+
0.151 +0.166+0.183+0.201 + 0.221 + 0.244).
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
all the more critical because of the notoriously slow pace of Soviet construction.
According to Krasovskiy,
In the Ninth Five-Year Plan the construction period took about
twice as long as the normative called for. . . . In turn, our norma-
tives call for about twice as long as the actual timespan abroad.
Cost overruns are more prevalent and larger in nonstandard design con-
struction. Since the Sborniki are intended for standard construction, they are
especially deficient in handling cost overruns. A private source has estimated
that nonstandard office buildings typically experience cost overruns of about
20%-30%.
We conclude from this testimony that the costs estimated from the Sborniki
must be rather arbitrarily adjusted upward. Some fragmentary evidence con-
cerning the approximate magnitude of the upward adjustment suggests some
upper and lower bounds. On the low side, a recent article states, "The formula-
tion of the draft 1973 plan revealed that cost overruns for leftover construction
projects based on the USSR Gosplan totaled 8 billion rubles." 13 There are at
least three plausible interpretations to this rather cryptic statement. One, cost
overruns explain 8 billion rubles or 12% of unfinished construction.14 An alternate
interpretation is that the 8 billion rubles should be compared with that year's
planned unfinished construction-51 billion rubles 15 This would suggest that
actual costs exceed estimate costs by nearly 16%. Third, the 8 billion rubles
could be interpreted as the cost overrun thus far, which means that either the
12% or 16% is the probable minimum average deviation above estimated cost.
The high boundary of this range is provided in an article by I. N. Dmitriyev.
In it he states, "In 1973 ministries and agencies reviewed the estimated costs
for 2,400 construction projects and the increases averaged 29%." 16 This propor-
tion is on the high side because it includes deviations from estimates that were
drawn up before the completion of the designs. These preliminary estimates are
obviously coarse and inaccurate and deviate from the first revised estimates
that the Sborniki are supposed to gauge.
Two intermediate estimates are provided in an article by V. P. Krasovskiy.
First, he states that in 1,638 leading construction projects the rise in estimated
cost during the first two years of the Ninth Five-Year Plan equaled 19 billion
rubles or 18%.17 Presumably this sample includes a large part of Soviet construc-
tion because the imputed value of Krasovskiy's sample is 105.6 billion rubles and
total planned construction in 1971-72 was 110.4 billion rubles. Second, the Cen-
tral Statistical Administration in a study of 1971 construction found that for
782 new construction projects and 1,034 construction projects at existing enter-
prises the rise in estimated costs was 13.5% and 30%, respectively 18 If the
average values projects in each of the above categories were equal, the average
13 V. L. Perlamutrov and L. V. Braginskiy, "Cost Accounting Credit and Finance Levers,"
Ekonomika i organizatsiya promyshlennogo proisvodstva, No. 1, 1975, pp. 61-70. Also in
JPRS 65219, 14 July 1975.
14 Narodnoyc khozyaystvo SSSR v 1973 g., p. 558.
15 I Khizhnyak, "Ways to Decrease Unfinished Production and Construction," Planovoye
khozyaystvo, No. 7, 1975, pp. 106-109. The article states that at the end of 1973 unfinished
construction exceeded normative limits by 16 billion rubles.
16 1. N. Dmitriyev, "The Most Important Construction Tasks for the Concluding Year of the
Five-Year Plan," Ekonomika slroitel'stva, No. 3, 1975, pp. 5-11. Also in JPRS 64695, 5 May 1975.
17 V. P. Krasovskiy, op. cit.
Is Ibid.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
increase in estimated costs would equal 22.9%.19 In view of the accumulated
evidence, we suggest that the ruble costs should on the average be adjusted
upward by at least 20%.
Given the conclusion that Soviet construction costs are understated by an
average of 20%, two alternatives can be followed in adjusting the ruble-dollar
ratios for individual construction categories such as housing or industrial con-
struction. Ruble-dollar ratios for all types of construction can be adjusted by
20%, or the categories can be adjusted differently with the proviso that the
weighted average adjustment be 20%. The simplifying assumption that the
average proportion of cost overrun is identical for all types of construction would
be fulfilled only by some quirk of fate.
Differences in construction relating to the degree of standardization,
technology of construction utilized, the priority of the projects to Soviet policy-
makers, and the type of project involved are relevant variables affecting the
degree of cost overrun. The Soviets have tacitly recognized this in a 1972 decree
numbered 560, "On Measures for Ensuring the Lowering of Estimated Construc-
tion Costs." In it are established acceptable capital reserves for unforeseen
contingencies and expenditures with the reserve expressed as a percentage of
estimate cost. The amount of the permitted reserve depends on the type of
projects-e.g., industry or housing-and whether the construction is experimen-
tal, nonstandard, or standard20 Although these reserves are too low to realistically
compensate for subsequent cost overruns, they yield a relative indication of
the effects of different variables in construction on the size of the probable
overrun.
The available evidence supports the idea of using a differential adjustment',
but the data are too fragmentary to derive a reliable series of coefficients. On
the basis of the 1972 decree and the constraint that the average overrun be 20%,
a set of estimated coefficients are derived for purely illustrative purposes in
Table 4. Although these coefficients tend to move in the expected direction, they
have far too tenuous a foundation to use.
Even though using equal adjustment coefficients represents a departure from
reality, we adopt them in this study as the best approach when faced by such
skimpy evidence. The resulting adjusted ratios are displayed in Table 5. Com-
putation of the US- and Soviet-weighted ratios appears in Appendix Tables 1-3
and 1-4. We stress that the 20% upward adjustment made to the ruble-dollar
ratios represents purely our subjective assessment to be used or discarded
depending on the proclivities of the researcher using these findings.
Comparison of Value of US and Soviet Construction
The unadjusted and adjusted aggregate construction ruble-dollar ratios are
compared in the following tabulation:
Unadjusted
Adjusted
US-weighted ratio ..................
0.621
0.746
Soviet-weighted ratio ...............
0.539
0.647
Geometric mean ratio ...............
0.579
0.695
19 This is probably a conservative estimate because the cost of new construction projects
probably averages more than projects at existing enterprises.
20 "Lowering of Estimated Construction Costs," Planovoye khozyaystvo, No. 10, October 1972,
pp. 155-157. Also in JPRS 57413, 2 November 1972.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Percent
-------- -------- ----- -----
(1) (2) (3) (4) (5) (6) (7) (8)
Share of
Construction
of That
Type'
Relative
Overrun
Coefficient2
Share of
Construction
of That
Type'
Relative
Overrun
Coefficient2
Share of
Construction
of That
Type
Relative
Overrun
Coefficient3
Weight in
Soviet Cost
Construction Overrun'
Industry ..................
25.0
3.00
75.0
6.95
100.0
5.96
36.2 31.4
Transportation and com-
munications except rail-
road ....................
87.0
3.00
13.0
5.00
100.0
3.26
6.0 17.2
Railroad transport .........
87.0
3.00
13.0
5.00
100.0
3.26
2.6 17.2
Housing construction .......
93.5
2.00
6.5
3.50
100.0
2.10
27.0 11.1
Construction of trade and
communal enterprises, for-
estry enterprises, and in-
stitutions of science, cul-
ture, art, education, and
health ...................
85.7
2.00
14.3
7.50
100.0
2.79
28.2 14.7
-----
--
-----
---
----------
All categories except industry use the reported figures in Narodnoye khozyaystoo SSSR u 1970 g., p. 497. As mentioned earlier, the
claim that 69% of industrial construction is standardized is exaggerated. A better measure of the degree of industrial standardization
comes from the statement that consolidated estimating norms can be used for 20%-30% of industrial construction, Malimanov op. cit.
2 These coefficients are adapted from coefficients for the planned capital reserves for unforeseen contingencies. The industry
coefficient is a crude weighted average of capital investment in each branch of industry. "Lowering of Estimated Construction Costs,"
Planovoye khozyaystvo, No. 10, 1972, pp. 155-157. Also in JPRS 57413, 2 November 1972.
3 This is a weighted average of standard and nonstandard construction.
4 Given the relative magnitude in column 6, these are the overruns that fulfill the requirement that the average overrun using Soviet
weights equal 20%.
US
Weighted
Ratio
USSR
Weighted
Ratio
Geometric
Mean
Ratio
Hospitals ...............................
0.416
0.331
0.371
Housing ................................
0.744
0.593
0.664
Single-family ..........................
N. A.
N.A.
0.775
Multi-family ..........................
0.703
0.545
0.619
Office buildings ..........................
0.610
0.541
0.574
Schools .................................
0.700
0.601
0.649
Elementary ...........................
0.662
0.582
0.621
Intermediate ..........................
0.740
0.622
0.678
Industry ...............................
N.A.
N.A.
0.721
Light ................................
N.A.
N.A.
0.754
Heavy ...............................
N.A.
N.A.
0.691
Roads ..................................
N.A.
N.A.
0.948
Airfields ................................
N.A.
N.A.
0.834
Railroads ...............................
N.A.
N.A.
1.003
Aggregate construction ...................
0.746
0.649
0.696
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
These ratios permit an estimate of the magnitude of construction in the
United States and USSR in 1970. Soviet construction may be valued in dollars
and US construction in rubles?1
Value of US
Construction
Value of Soviet
Construction
Using unadjusted ruble-dollar ratios
Billion 1970 dollars ...............
94.2
93.3
Billion 1970 rubles ................
58.5
50.3
Using adjusted ruble-dollar ratios
Billion 1970 dollars ...............
94.2
77.7
Billion 1970 rubles ................
70.3
50.3
The comparisons resulting from these values are as follows (value of USSR
construction as a percent of the value of US construction):
Using
Unadjusted
Ruble-Dollar
Ratios
Using
Adjusted
Ruble-Dollar
Ratios
In 1970 dollar prices ................
99.0
82.5
In 1970 ruble prices .................
86.0
71.6
Geometric mean of dollar-ruble com-
parisons .........................
92.3
76.9
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08S01350R000602050003-2
APPENDIX A
THE ALGEBRA OF RUBLE-DOLLAR RATIOS
International comparisons of economic activity always encounter the index
number problem unless the mix of economic activity or the relative prices
attached to that activity are the same in the countries being compared.* Thus,
when the relative volume of Soviet and US construction is examined, the
comparison could be made in Soviet or US prices:
n
Fj
i=1
PsiQai
(1)
V8=
n
Fj
i=1
PaiQui
n
i=I
PuiQui
(2)
Vu=
z
PuiQai
Where:
Va= the ratio of the value of Soviet construction to US construction where both are
measured in rubles
Vu= the ratio of the value of US construction to Soviet construction where both are
measured in dollars
Qai= the quantity of Soviet construction of type i
Pai=the Soviet price of construction of type i
Qui=the quantity of US construction of type i
P,ii= the US price of construction of type i.
To compare construction rigorously in these countries in either the US or
Soviet price system would require complete recosting of each project in one of the
countries using the other country's prices. Specifically, a comparison using
formula (1) demands the revaluation of all US construction in rubles; equation
(2) demands the conversion of all Soviet construction into dollars. The effort
involved would be far beyond reason, and in any event, the requisite price and
quantity data simply are not available.
International price relatives provide a short cut solution to the comparison
problem. In this study the relevant price relatives are either ruble-dollar or dollar-
ruble ratios. These ratios represent averages of US and Soviet price ratios in
which each ratio is given a proportionate weight based on the importance of the
given type of construction in the total value of US or Soviet construction. Alge-
*The index number problem reflects international differences in tastes, levels of income, natural
resources, technology, and state of development.
Declassified and Approved For Release 2012/03/15: CIA-RDP08S01350R000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
braically, sample dollar-ruble ratios derived for a category are used with Soviet
value weights to obtain an average dollar-ruble ratio (1/R,), and sample ruble-
dollar ratios derived for a category are used with US value weights to obtain an
average ruble-dollar ratio (R?):
(3) 1/R8= Pu/Pei PatQ8i
n
P8iQ8i
i=1
E P, QBi.
= i=1
n
2: P8iQ8i
(4) R?=
P,,i/Pui PUPO = i=1
i=1 n
PuiQut PviQui
i=1 i=1
The results of equations (3) and (4) can then be used to estimate V. and V.
in equations (1) and (2), above-the comparisons of US and Soviet construction
volume in ruble and in dollar prices.
~)n
Z P8iQ8i
(6) Vu=
V8=
R. Fj PuiQui
i=l
1/R8 P8iQ8i
i=1
Thus, to make comparisons of US and Soviet construction, one needs the
value of each country's construction in domestic prices, a US-weighted ruble-
dollar ratio, and a Soviet-weighted dollar-ruble ratio. The ratios can be computed
from a sample of representative construction projects in each country in the
manner described in this study, and the values of construction in domestic
prices are reported in national statistics.
PuiQui
i=1
E PnAi
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
APPENDIX B
To find the location representing the average cost in the United States, a
sample of one hundred city price indexes from Building Construction Cost Data
1972 was collected. These indexes provide the relative cost of construction
between all these cities in 1971. Since New York City is the most expensive
construction location in the contiguous 48 states, all these city indexes were
adjusted to a base where New York City equaled 100. By weighting each of
these city indexes equally, an arithmetic mean of all the indexes was computed.
On the basis of these results, the average construction costs are approximately
83% of those in New York City.
The US data used in this study present costs in one of three ways: (1) average
cost for the country, (2) the cost for a major metropolitan area, or (3) the cost at
the actual project site. In the first case, there is no need for an adjustment to
have average US costs. In the second case, the costs are adjusted to those of
New York City and then multiplied by 83% to convert them to the average.
In the third case, we located the nearest major metropolitan area for which
there is a price index and then used that city as a base to proceed as in the
second case.
The determination of the location representing the average cost in the
USSR is different. When costing construction with the Sborniki, two location
identifiers must be assigned to the project under consideration. Each of these
identifiers divides the Soviet Union according to a different scheme of regional
cost variation. One system, which divides the country into ten territorial zones,
allows for variations in material costs resulting from the construction site's
remoteness from the source of materials production, and also considers labor costs
unique to each area. The other system of locational variation in cost divides the
country into four regions to account for cost differences contingent on climatic
conditions. (For maps indicating the boundaries of these regional zones, see
Figures 1 and 2.)
The precise distribution of construction by territorial zones cannot be
determined, because the Soviets do not publish such statistics. The territorial
zones sometimes overlap republics, economic regions, oblasts, national okrugs,
and krays. Construction data are usually available solely for independent
republics and economic regions. Only investment figures (including the cost of
machinery installed) are available for lower organizational units.
When the value of construction for one of these lesser organizational units
is estimated, the geographical pattern of construction is taken to be similar to
that of the economic region or republic. For example, all of the Ural economic
region is located in territorial Zone 2 except for Udmurtskaya ASSR, which falls
in territorial Zone 1. To distribute construction for this region between Zones 1
and 2, construction in Udmurtskaya ASSR must be isolated from the rest of the
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Figure 2 Climatic Regions of the Soviet Union
24
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
region. Although a distribution by oblasts and ASSRs is available for capital
investment, only the total construction and installation for the Urals region is
known, i.e., 2,973 million rubles in 1970. Udmurtskaya ASSR contains 9.0% of
the region's capital investment, so its estimated construction is 0.09 x 2,973
million rubles=267.6 million rubles, and construction for the rest of the region
is 2,705.4 million rubles.
Sometimes the allocation of construction to a specific territorial zone is
impossible. For example, Yakutskaya ASSR lies in both territorial Zones 7 and
10, and this delineation does not coincide with any administrative boundaries.
In such a case, the calculation is performed by alternatively alloting all con-
struction activity to the lowest and then to the highest cost zone in that area.
These two extremes provide a range of the possible average-weighted territorial
zone. The first calculation is the minimum distribution and the latter the
maximum distribution.
The derived distribution is shown in Table B-1. By reviewing the effects of
the most extreme assumptions, one can see that the distribution is fairly insensi-
tive to these assumptions, which permits a reasonably accurate estimate of the
precise distribution. Because most construction activity is located in the lower
zones, the true distribution is undoubtedly closer to the minimum than the
maximum. According to this table, 85% of construction activity occurs in terri-
torial Zones 1, 2, and 3. The value of the Soviet median territorial zone
Distribution of Construction
by Zone (Percent)3
1 ..........................
100
33.22
33.22
2 ..........................
104
47.02
45.33
3 ..........................
112
6.62
6.62
4 ..........................
120
8.36
4.82
5 ..........................
127
2.58
1.65
6 ..........................
172
0
0.93
7 ..........................
191
0.70
2.61
8 ..........................
220
0.91
0.59
9 ..........................
250
0.58
0.90
10 .........................
N.A.
0
3.30
Data compiled from Narodnoye khozyaystvo SSSR 1922-1972, yubileynyy statisticheskiy yezhegod-
nik; Narodnoye khozyaystvo RSFSR v 1970; and Narodnoye khozyaystvo RSFSR v 1971. All of these
statistical handbooks are compiled by the Central Statistical Administration.
2 This index is derived from a random sample of 25 Sborniki tables. Zone 10 is omitted because
its costs are presented in two ways: either a fixed cost as in the other zones, or some percentage
of the zone south of it. Zone 10 areas have various zones south of them, so there is no uniform zone
price. In addition, the volume of construction in Zone 10 is statistically insignificant. Source:
Gosstroy USSR, Sborniki, Investment Series.
3 In this table, 7.2% of construction could not be definitely allocated to any one territorial zone.
This residual was allocated in terms of the most extreme assumptions to ascertain the limits of the
particular oblast or kray; the percentage of construction and installation work in its economic
region is the same as its percentage of capital investment. Because of rounding, components may not
add to 100%.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
provides a measure of the representative Soviet construction location. The
median is 2 and, consequently, Zone 2 is used as the average zone for the
Sborniki costing estimates.
Assumptions analogous to those employed to find the average Soviet terri-
torial zone can be made to obtain patterns of construction by climatic region.
The results of these calculations are given in Table B-2. The mean weighted
climatic region and the median are approximately equal to Level II, so this
regional identifier is used as the average location.
Cost Index
(Region II = 100) 2
Distribution of
Construction by Region
(Percent) 3
I ......................................
107.0
30.87
II .....................................
100.0
46.09
III ....................................
95.8
13.61
IV .....................................
94.0
9.40
1 Data compiled from Narodnoye khozyaystvo SSSR 1922-1972; Narodnoye khozyaystvo RSFSR v
1970; Narodnoye khozyaystvo Kazakhstana v 1968; and Narodnoye gospodarstvo Ukrayins'koyi RSR v
1970 (in Ukrainian). All of these statistical handbooks are compiled by the Central Statistical
Administration.
2 This index is computed from a sample of 43 Sborniki climatic correction tables.
3 Because of rounding, components do not add to 100%. For purposes of this table and because
of scarcity of data, it is assumed that construction in Kazahkstan SSR is allocated by oblast
proportional to 1968 investment.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
APPENDIX C
Two approaches are used in deriving the ruble-dollar ratio for hospitals.
First, a US-representative ruble-dollar ratio is derived by selecting a sample of
US hospitals and costing the sample in rubles and dollars. In the second approach,
a Soviet-representative ratio is derived by selecting a sample of Soviet hospitals
and costing the sample in rubles and dollars.
The US-Representative Approach
The sample of 28 US general hospital buildings used in the US-representative
approach was drawn from the Dodge Digest and costed in rubles using the Soviet
Sborniki and in dollars using the Dodge Digest (see Table C-1).
The ruble costing of these 28 hospitals was fairly direct. The appropriate
Sbornik contains two types of general hospital structures comparable to US
construction. The type selected for each comparison with US hospitals depended
on the structure with the same construction volume. An adjustment was made to
give the Soviet hospital the same number of stories as the US hospital. The
hospitals were assumed to have been built in a locale climatically similar to the
USSR north of 50 degrees latitude and west of 50 degrees longitude. Also, these
buildings are located where regional construction costs apart from climate are
average, i.e., most of the western Soviet Union.
For dollar costing, the specifications for the 28 US hospitals were altered to
conform as nearly as possible with Soviet practice regarding air-conditioning,
built-in equipment, bid date, and regional cost differences. Since most US
hospitals are air-conditioned and most Soviet hospitals are not, the cost of air-
conditioning had to be deducted from the dollar cost. Although the exact cost of
air-conditioning is usually unknown, the maximum is known, so the deduction,
which averages 6%, is estimated for each observation. Any built-in equipment
included in the Dodge Digest was deleted because such equipment is not con-
sidered part of Soviet construction costs. The bid prices for US hospital construc-
tion, which were scattered over several years in the Dodge Digest, were adjusted
to June 1970 dollars by the Dodge index of construction costs for the major city
nearest each hospital. Finally, 83% of the average construction cost of hospitals
in New York City was used to eliminate regional cost differences.
The Soviet-Representative Approach
The sample of 3 Soviet hospital campuses encompassing 23 buildings was
selected from Spravochnik ukrupnennikh pokazateley smetnoy stoimosti i raskhoda
resursov zdaniya i sooruzheniya lechebno-profilakticheskikh uchrezhdeniy (Manual
of the Consolidated Indexes of the Estimated Costs and Expenditures of Re-
sources: Buildings and Structures of Medical and Preventive Medicine Institu-
tions), Moscow, 1968, hereafter referred to as Spravochnik. The three campuses
consisted of: a 1,000-bed hospital with a polyclinic of 1,200 places per day; a
600-bed hospital with a polyclinic of 1,200 places per day; and a 240-bed hospital
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
6) 'O
O m O y
v ~_ O O Of
,y U -r 7
F v OC
_ I
O ~' M O O 00 00 O'0 00 N m 00 O N N o0 c0 N O -. O OD 10 T O O K
h O -. M O 10 M .?+ O o0 .. .M N O P. o0 O 10 00 N N
N N M M N M M N N N M M N N V~ M M H' M M M m N d' M
O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O
N V ???~ .M c0 00 M P. t0 t0 00 t0 M t0 .0 N M, V? O O 00 t0 C' M C' 'Ir .. N m a' h N M M (= N N C) 0 0 0 (31 0 0 0 0 0 0 00 0 0 0 0 0 0 0 O) O O O C1
o -? O o O o- o o O o 0 0 0 0 0 0 0 0 0 o O o o O o o O
10 C) ?-? M -' O Cl t0 .M h m N 10 -?. m 00 O m O m 00 m m C) P. d' O
O P. V' .0 d' O 00 P. P. .1- O O 7~ M O C' C) O (m ?-. O 00 .t0 t0 O .)J
N 0 ?-+ -. 0 O -+ O O O -? N N O O N O O O O -? -? O '-?
t0 00 Cl N N 00 N M V~ eM 00 t0 Cl .0 h P. h 00 h '0 n' M t0 N O t0 -.
00 M -. O M N .0 .0 t0 t0 h 00 C) ~' M M M 00 M t0 ?0 h C) O h .0
-. 0 0 0 0 0 0 0 ?-? O O O ?-. O ??-~ -. O ?-+ O C) 01 O) C) O O O O
'O V n ?-. O O O N M M O N O -? M M .0 -. a' N O O O h 00 h .- - V' t0
.V+ E9 O O 00 "' M M eM O O ~' O N. . .M h O [~. O M O ?0 .. N M
. . . . . . . . . . . .
C o
.7 U] 7
O O CO M -,' t0 M ??. O O O 00 t0 h t0 M It C 00 .0 M P. .0 V O N .0
O O h O h 00 O h O O o0 t0 h h 00 h 00 t0 h h C) 00 M -. tD t0 O O
c x
o C)
1.
` c
C N
E o
O v O O
m O C y
U
0 0
V 0 0
> V N
C cd O
O U
pyj O Oh
00
V O y t w
O V L 0 U
s
CL
00)) V Cd
s ?fl 'O a) m
y Cl
c:c..,
o > ep
:D N C O
'O
O c. O .~ C
.0 y O -O ?C
c a -o
5 C
a 'a > 0
E 3 y C
= V
0 5 z O C =
^ - C .~ O cQJ
O .0 00 m 0 X
0
L E c
C O O n ?X E
y O 0. V
tp C 'O cd x
C)C 00 000
_O = O h
W X U 'D ~ d E
^ O C -. O 'O O
m +' c. O C ax) C cd N C v O
m Cl P. M .0 M O O 00 C> Cl M .0 m ~' 00 m .M O CO M Cl C O 'D O = C
O C 00 .M I- m 1' O O -. 00 CC .M P. C CO m O -? M 00 O C> O O h .M O ` '. .~ y ,D C) 0
V; d .M O M O -. M .0 h O -. ~' .fJ N O .0 O M M O N M O M h O M M O LV V C > E E _
LO r- C~
C p qi -. .-. .-?. .-. -. .-. ~. -. -. N N N N M M M t0 h h 0
O [-a U C O O V
V O _ C _
'O a cd O cd
o a
Leo q F a F o c
Leo 3 d ;ti .ci 'a 0 0 'v 0
. . . . . . . . . . . . . . . a q 8 ? ti P U ti U 0
. . . . . . . . . ... . . . . . . . B =
N M ' t0 h 00 C) O -. N M ~' .0 tD h 00 O O N M h t0 00 C cd ro m ?i '"
.. .. .. .. .. .. .. .. .. .. Cl N N N N N -C4 N
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
O q, w0 . _ -C U .0 .v
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
with a polyclinic of 500 places per day. This sample of buildings contains main
buildings, kitchen and dining facilities, pathology buildings, and a transformer
substation (see Table C-2).
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Volume of
Structure 1
(Cubic
US Cost
per Cubic
Meter 2
Air-
Conditioning
Adjustment 3
Adjusted
Total
US Cost 4
(Thousand
Unadjusted
Soviet Cost
per Cubic
Meter 5
Adjustment
for Number
of Stories 6
Total
Soviet Cost 7
(Thousand
Ruble-
Dollar
Meters)
(US $)
(Index)
1970 US 8)
(Rubles)
(Index)
1970 Rubles)
Ratio 8
1..........
108,307
111.95
0.940
11,397
30.9
0.97
3,246
0.285
2 ..........
16,409
109.15
0.940
1,684
24.2
0.98
389
0.231
3 ..........
21,099
109.82
0.940
2,178
23.8
0.98
492
0.226
4 ..........
2,547
99.33
0.940
238
39.4
1.00
100
0.420
5 ..........
3,507
99.23
0.940
327
32.1
1.00
113
0.346
6 ..........
8,194
99.22
0.940
764
27.9
1.00
229
0.300
7 ..........
2,400
99.17
0.940
224
39.5
1.00
95
0.424
8 ..........
262
99.24
0.940
24
34.6
1.00
9
0.375
9 ..........
78,792
111.95
0.940
8,292
28.4
1.03
2,305
0.278
10.........
12,125
102.02
0.940
1,163
24.2
1.00
293
0.252
11 .........
21,099
109.82
0.940
2,178
23.8
0.98
492
0.226
12.........
2,547
99.33
0.940
238
39.4
1.00
100
0.420
13 .........
2,645
99.06
0.940
246
32.1
1.00
85
0.346
14 .........
6,640
99.25
0.940
619
27.9
1.00
185
0.299
15 .........
2,031
99.46
0.940
190
39.5
1.00
80
0.421
16.........
210
100.00
0.940
20
34.6
1.00
7
0.350
17 .........
19,680
109.81
0.940
2,031
26.1
0.98
503
0.248
18 .........
4,542
99.30
0.940
424
24.4
1.00
111
0.262
19 .........
6,500
99.23
0.940
606
27.2
1.00
177
0.292
20.........
1,137
99.38
0.940
106
32.1
1.00
36
0.340
21 .........
3,115
99.20
0.940
290
35.1
1.00
109
0.376
22.........
825
99.39
0.940
77
33.9
1.00
28
0.364
23.........
235
97.87
0.940
22
34.6
1.00
8
0.364
Unweighted ruble-dollar ratio for buildings
Value-weighted ruble-dollar ratio
0.324
- - - - -
- - - - - -
- - - -
- - - - - -
1 Spravochnik ukrupnennikh pokazateley smetnoy stoimosti i raskhoda resursov, zdaniya i sooruzheniya
kticheskikh uchrezhdeniy, pp. 11-46, 65-74. The structures identified by number in the table are:
1.
Main hospital building with 810 beds
13.
Food services building
2.
Infectious ward with 120 beds
14.
Support services building
3.
Outpatient (polyclinic) facilities for 1,200 patients per day
15.
Pathology building
4.
Radiology building with hospital of six beds
16.
Transformer substation
5.
Food service building
17.
Main hospital building with 200 beds
6.
Support services building
18.
Infectious ward with 40 beds
7.
Pathology building
19.
Outpatient facilities for 500 patients per day
8.
Transformer substation
20.
Food service building
9.
Main hospital building with 510 beds
21.
Support services building
10.
Infectious ward with 84 beds
22.
Pathology building
11.
Outpatient facilities for 1,200 patients per day
23.
Transformer substation
12.
Radiology building with hospital of six beds
2 Dodge Guide, p. 4-3. US costs given on a square foot basis are converted to cubic feet and subsequently to cubic meters with the
assumption that the height of a story averages 11.6 feet.
3 6% is assumed throughout.
4 Column (1) x column (2) x column (3).
5 Sborniki, Budget Series, no. 5, pp. 7-24, and Sborniki, Investment Series, no. 3, p. 175.
6 Sborniki, Budget Series, no. 5, op. cit., p. 4.
7 Column (1) x column (5) x column (6).
8 Column (7) = column (4).
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
The ruble costs for these three campuses were taken from the Sborniki and
applied to the physical facilities assumed to be built in the same location as
described in the US-representative approach.
The Soviet hospitals are costed in dollars using the Dodge Guide, which
gives US dollar costs as of mid-1970, and are averaged to allow for geographic
differentials. Average quality US hospitals are used for comparison with the
Soviet hospitals, but the US costs are reduced by 6% to eliminate the cost of
air-conditioning.
The estimates are summarized in Table C-3. The US-representative sample
has a larger average building size than the Soviet-representative sample. The
unweighted ruble-dollar ratios for US and Soviet methodologies are 0.33 and
Range
Average Minimum Maximum
1. US-Representative Methodology (Sample size: 28 buildings)
Volume (cubic meters) ........................... 26,184 4,594 143,392
Soviet cost (thousand rubles) ...................... 782 112 4,564
US cost (thousand US 8) ......................... 2,252 401 10,645
Soviet weighted cost (rubles per cubic meter)....... 29.9 24.4 33.1
US weighted cost (dollars per cubic meter) ......... 86.0 71.76 119.67
Unweighted ruble-dollar ratios .................... 0.329 0.204 0.468
Value-weighted ruble-dollar ratio .................. 0.347
II. Soviet-Representative Methodology (Sample size: 3 campuses, 23 buildings)
Volume (cubic meters)
Campus ...................................... 108,283 36,034 162,725
Building ...................................... 14,124 210 108,307
Soviet cost (thousand rubles)
Campus ...................................... 3,068 973 4,684
Building ........:............................. 400 7 3,246
US cost (thousand US 8)
Campus ...................................... 11,822 3,784 13,772
Building ...................................... 1,449 20 11,397
Soviet weighted cost (building) (rubles per cubic
meter) ....................................... 28.3 23.3 39.5
US weighted cost (building) (dollars per cubic meter) . 108.4 91.60 105.24
Unweighted ruble-dollar ratios
Campus ...................................... 0.276 0.273 0.279
Building ...................................... 0.324 0.226 0.424
Value-weighted ruble-dollar ratio (building) ........ 0.276
0.32, respectively, and the coefficient of variation-i.e., the ratio of the stand-
ard deviation to the mean-is about 20% in each case. A 95% confidence interval
around the mean gives an interval estimate of 0.30 to 0.35 in both cases. A test
of the dispersion of the project ruble-dollar ratios about the mean suggests a
strong central tendency.*
*This test was conducted by separately dividing the sample of ruble-dollar ratios obtained by
each method into a frequency distribution of six classes which are: from two to three, one to two, and
zero to one standard deviations both above and below the average ruble-dollar ratio. The expected
frequency of occurrence in each class of the sample of ratios, if the sample were normally distributed,
was then computed. A chi-square test of the relationship between the actual sample frequency and
the frequency predicted by the normal distribution resulted in chi-square values of 6.38 and 2.78
for the US-representative and Soviet-representative samples, respectively. These values are not
sufficiently large to refute the assertion that there is a 95% probability that the sample of ratios in
each case is normally distributed and that, therefore, there is a strong central tendency.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
On the other hand, the value-weighted ruble-dollar ratio is lower in the
Soviet-representative sample. This accords with expectations because Soviet
builders should do relatively better at the construction of their own types of
buildings than they should at duplicating American projects. The value-weighted
ruble-dollar ratios are used for computation of the aggregate hospital ratio by
taking the geometric mean of each approach's ratio; this results in a hospital
ratio equal to 0.309.
The relationship between the unweighted ratio and the value-weighted ratio
should provide a clue of the effect of scale, if any, on the ruble-dollar ratio as
was discussed earlier. The results in this case are ambiguous because the US-
representative sample suggests the ruble-dollar ratio increases with project size,
but the Soviet-representative sample indicates the opposite. With such results,
the effect of scale must be inconclusive.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08S01350R000602050003-2
APPENDIX D
The ruble-dollar ratio for housing construction is computed by developing
a sample of ratios for specific types of housing. These ratios are then aggregated
into two weighted averages: one average is based on US construction weights
and the other is based on Soviet construction weights. The geometric mean of the
two ratios provides the aggregate housing construction ruble-dollar ratio.
The Approach
A sample of 69 different types of housing construction is used to determine
the aggregate ratio. This sample is selected so that unique features of buildings
pertinent to construction in either country are represented. For example, there
are ruble-dollar ratios for both four- and five-story walk-up apartments, which
are so common in the USSR and rare in the United States, and the four- and
five-story buildings with elevators, which are typical of US practices and
atypical in the USSR. Other features that distinguish the categories in the
sample include number of family units (single-family versus multi-family),
number of stories, and the inclusion or exclusion of basements. The buildings
are then costed in rubles from the Sborniki and in dollars from the Dodge Guide
as shown in Table D-1. Again, the buildings are costed for average locations in
each country to eliminate regional cost differentials. The actual costing of the
sample varies according to the unique features of the buildings described below.
The sample of single-family dwellings contains seven buildings based on
Soviet designs contained in Al'bom proyektov dlya sel'skogo stroitel'stva: zhilyye
zdaniya dlya individual'nogo stroitel'stva (Album of Plans for Rural Construction:
Housing for Individual Construction). The floorspace in these structures is com-
puted from the floor plans and cross-sections in the book. These areas are ad-
justed by subtracting one-half of the area of any verandas included to be con-
sistent with the American Institute of Architect's standards for computing floor
area. Although several of the buildings' construction volumes are published,
close examination reveals that many of the volumes have errors and differ from
Soviet standards for measuring volume, so independent volume estimates are
made. Since Soviet single-family housing is usually of low quality, these Soviet
buildings are compared with US construction of single-family, masonry wall,
low-quality structures.
The best quality one-, two-, and three-story Soviet apartment buildings are
compared with average US quality, masonry wall garden apartments with the
same number of floors. For each of these building heights, construction costs are
a function of both unit area and building volume. Since the Dodge Guide costs
are presented on the basis of average unit size, both the minimum and maximum
Declassified and Approved For Release 2012/03/15: CIA-RDP08S01350R000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Table D-1
(1)
(2)
Floor
Space'
(Square
(3)
US Cost
per Square
Meter2
(4)
Total
US Cost 3
(Thousand
(5)
Structure
by Volume 4
(Cubic
(6)
Soviet
Cost per
Cubic
Meters
(7)
Total
Soviet
Cost 6
(Thousand
(8)
Ruble-
Dollar
Type of Structure
Meters)
(US $)
1970 US $)
Meters)
(Rubles)
1970 Rubles)
Ratio
Single family housing ..............
131.2
151.75
19.9
438.0
24.5
10.7
0.538
68.7
171.08
11.8
252.0
26.4
6.7
0.568
97.5
165.15
16.1
409.0
25.1
10.3
0.640
97.5
165.15
16.1
372.0
25.0
9.3
0.578
191.4
108.40
20.7
564.0
23.0
13.0
0.628
189.6
99.22
18.8
538.0
27.7
14.9
0.793
140.5
121.65
17.1
552.0
24.1
13.3
0.778
Average ......................
One-story apartment
130.9
140.33
17.2
446.0
25.1
11.2
0.646
Small building ..................
104.5
114.89
12.01
312.1
27.2
8.49
0.708
139.4
104.51
14.57
416.2
26.6
11.07
0.760
Large building ..................
104.5
114.89
12.00
312.1
22.7
7.08
0.590
139.4
104.51
14.57
416.2
22.2
9.24
0.634
Average ......................
Two-story apartment
122.0
109.70
13.29
364.2
24.7
8.97
0.673
Small building ..................
104.5
107.31
11.21
312.1
23.3
7.27
0.649
139.4
97.96
13.66
416.2
22.8
9.49
0.695
Large building ..................
104.5
107.31
11.21
312.1
20.6
6.43
0.574
139.4
97.96
13.66
416.2
20.1
8.37
0.613
Average ......................
Three-story apartment
122.0
102.64
12.44
364.2
21.7
7.89
0.633
Small building ..................
104.5
103.52
10.82
312.1
24.2
7.55
0.698
139.4
94.35
13.15
416.2
23.7
9.86
0.750
Large building ..................
104.5
103.52
10.82
312.1
23.0
7.18
0.664
139.4
94.35
13.15
416.2
22.5
9.36
0.712
Average ......................
Four-story apartment
122.0
98.94
11.98
364.2
23.4
8.49
0.706
With elevators ..................
3,716
216.27
804
11,359
27.9
317
0.394
13,935
169.65
2,364
42,527
26.5
1,127
0.477
Without elevators ...............
3,716
207.20
770
11,359
25.1
285
0.370
13,935
162.53
2,265
42,527
23.8
1,012
0.447
Average ......................
Five-story apartment
8,826
188.91
1,551
26,943
25.8
685
0.422
With elevators ..................
3,716
216.27
804
11,359
27.5
312
0.388
13,935
169.65
2,364
42,527
26.1
1,110
0.470
Without elevators ...............
3,716
207.20
770
11,359
24.0
273
0.355
13,935
162.53
2,265
42,527
22.8
970
0.428
Average ......................
8,826
188.91
1,551
26,943
25.1
666
0.410
Six-story apartment ...............
3,716
190.29
707
11,359
27.7
315
0.446
13,935
157.54
2,195
42,572
25.6
1,090
0.497
Average ......................
Seven-story apartment
8,826
173.92
1,451
26,966
26.6
702
0.472
With basement .................
3,716
195.37
726
11,338
29.9
339
0.467
23,225
156.33
3,631
70,878
26.4
1,871
0.515
Without basement ...............
3,716
216.88
806
11,338
27.3
310
0.385
23,225
169.65
3,940
70,878
25.5
1,807
0.459
Average ......................
Eight-story apartment
13,470
184.56
2,276
41,108
27.3
1,082
0.457
With basement ..................
3,716
197.79
735
11,338
29.1
330
0.449
23,225
158.16
3,673
70,878
27.3
1,935
0.527
Without basement ...............
3,716
216.88
806
11,338
26.9
305
0.378
23,225
169.65
3,940
70,878
25.1
1,779
0.452
Average ......................
Nine-story apartment
13,470
185.62
2,288
41,108
27.1
1,087
0.452
With basement ..................
3,716
199.71
742
11,338
31.2
354
0.477
23,225.
159.34
3,701
70,878
32.3
2,289
0.618
Without basement ...............
3,716
216.88
806
11,338
26.7
303
0.376
23,225
169.65
3,940.
70,878.
24.2
1,715
0.435
Average ......................
13,470
186.40
2,297
41,108
28.6
1,165
0.477
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
The Housing Sample (Continued)
- - - - - - - - - - - - - - - - - - --- - - -- - - --
(2) (3) (4) (5) (6) (7) (8)
Soviet Total
Floor US Cost Total Structure Cost per Soviet
Space' per Square US Cost 3 by Volume 4 Cubic Cost 6 Ruble-
(Square Meter 2 (Thousand (Cubic Meter 5 (Thousand Dollar
Type of Structure Meters) (US $) 1970 US $) Meters) (Rubles) 1970 Rubles) Ratio 7
Ten-story apartment
With basement .................. 3,716 201.30 748 11,338 29.5 334 0.447
23,225 160.24 3,722 70,878 29.5 2,091 0.562
Without basement ............... 3,716 216.88 806 11,338 36.3 412 0.511
23,225 169.65 3,940 70,878 36.3 2,573 0.653
Average.. ......... ..13,470 187.02 2,304 41,108 32.9 1,352 0.543
Eleven-story apartment
With basement .................. 3,716 202.60 753 11,338 31.3 355 0.471
23,225 161.04 3,740 70,878 31.3 2,218 0.593
Without basement ............... 3,716 216.88 806 11,338 34.2 388 0.481
23,225 169.65 3,940 70,878 34.2 2,424 0.615
Average ...................... 13,470 187.54 2,310 41,108 32.8 1,346 0.540
Twelve-story apartment
With basement .................. 3,716 203.72 757 11,338 33.1 375 0.495
23,225 161.79 3,758 70,878 33.1 2,346 0.624
Without basement ............... 3,716 216.88 806 11,338 31.6 358 0.444
23,225 169.65 3,940 70,878 31.6 2,240 0.569
Average ...................... 13,470 188.01 2,315 41,108 32.4 1,330 0.533
Thirteen-story apartment
With basement .................. 3,716 204.66 761 11,338 32.0 363 0.477
23,225 162.32 3,770 70,878 32.0 2,268 0.602
Without basement ............... 3,716 216.88 806 11,338 32.0 363 0.450
23,225 169.65 3,940 70,878 32.0 2,268 0.576
Average ...................... 13,470 188.38 2,319 41,108 32.0 1,316 0.526
Fourteen-story apartment
With basement .................. 3,716 205.48 764 11,338 30.9 350 0.458
23,225 162.85 3,782 70,878 30.9 2,190 0.579
Without basement ............... 3,716 216.88 806 11,338 30.9 350 0.434
23,225 169.65 3,940 70,878 30.9 2,190 0.556
Average ...................... 13,470 188.72 2,323 41,108 30.9 1,270 0.507
Fifteen-story apartment
With basement .................. 3,716 206.20 766 11,338 33.8 383 0.500
23,225 163.30 3,793 70,878 33.8 2,396 0.632
Without basement ............... 3,716 216.88 806 11,338 33.8 383 0.475
23,225 169.65 3,940 70,878 33.8 2,396 0.608
Average ...................... 13,470 189.01 2,326 41,108 33.8 1,390 0.554
Sixteen-story apartment
With basement .................. 3,716 205.86 765 11,338 32.8 372 0.486
23,225 163.69 3,802 70,878 32.8 2,325 0.612
Without basement ............... 3,716 216.88 806 11,338 32.8 372 0.462
23,225 169.65 3,940 70,878 32.8 2,325 0.590
Average ...................... 13,470 189.02 2,328 41,100 32.8 1,348 0.538
' The floorspace areas for one- to three-story apartments are based on the average size of one unit, but all other areas are based
on total floorspace of the building. The areas for the single family housing are adjusted from Al'bom proyektov dlya sel'skogo
stroitel'stva: zhilyye zdaniya dlya individual'nogo stroitel'stva by halving the given area of any verandas to conform with American
Institute of Architects' guidelines. The floorspace areas of multifamily housing are derived from the Dodge Guide, pp. 1-52, 1-56, 1-60.
2 Dodge Guide, pp. 1-29, .1-52, 1-56, 1-60. For buildings above three stories, the costs of air-conditioning are excluded by
multiplying the given costs by a factor of 0.954. To determine the US cost of four- and five-story walk-up apartments, the
construction costs are reduced an additional 4% to remove the cost of elevators. This coefficient is derived from Building
Construction Cost Data 1972, by R. S. Means Company, Inc.
3 Column (2) x column (3).
4 Construction volumes for one- to three-story multifamily housing are the sizes of average units. The volumes for the single-
family housing are computed from the floor plans in Al'bom proyektov, op. cit., because the stated volumes in several cases seem
erroneous and not in conformity with Soviet methods for computing construction volume. The volumes for multifamily units are
from Dodge Guide, pp. 1-52, 1-56, 1-60, 1-74.
5 Sborniki, Budget Series, Ifo. 4, pp. 3-81. Sborniki, Investment Series, No. 28, pp. 3-38.
6 Column (5) x column (6).
7 Column (7) = column (4).
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
unit sizes are costed. These alternative unit sizes are in turn costed for both the
smallest and largest buildings with the same number of stories that are listed in
the Sborniki. The four ratios that are computed for apartment buildings with a
given number of stories can be arranged in the following matrix:
Smallest Largest
Soviet Soviet
Building in Building in
the Sborniki the Sborniki
Smallest average apartment in the
Dodge Guide .....................
Largest average apartment in the
Dodge Guide .....................
The major difficulty in estimating ruble-dollar ratios for apartment build-
ings lies in matching Soviet costs per cubic meter of building with US costs per
square meter of apartment space (excluding non-apartment areas such as common
halls). The size of an apartment is equivalent to what the Soviets call useful
space. Soviet urban apartments are constructed so that living space averages
appears
approximately two-thirds of useful space. A descriptive statistic, K2,
frequently in Soviet housing literature and is defined as the ratio of a structure's
construction volume to living space. In a survey of published housing plans, K2
averages 5.6 in the Soviet Union. The above relationships provide a direct link
between the US and Soviet ways of expressing cost.*
High-quality Soviet four-, five-, and six-story apartments are compared
with average US quality, masonry wall, medium-rise apartments. In these cate-
gories, total costs are based primarily on total building size instead of the average
size of apartment units. Accordingly, both the smallest and largest buildings listed
in the Dodge Guide are costed for a given number of stories. The four- and five-
story buildings are priced in both rubles and dollars, with and without elevators,
in order to represent differences in construction practices in the two countries.
Six-story buildings, however, possess elevators in both countries. Since all the
medium-rise apartments in the Dodge Guide contain elevators, the cost of eleva-
tors is excluded by reducing total construction costs by 4%-the average
proportion of construction costs allocated to elevators in US apartment con-
struction. Moreover, the US buildings are air-conditioned, so this cost is also
*Useful space is the total floor area of a building minus the space of external halls, stair-
ways, and external elevator shafts common to more than one unit and the space occupied by
walls. Living space is the area devoted to living rooms and bedrooms only.
Let: V = volume of a building
L = living space
U = useful space
K2 = V/L
Since: K2 = 5.6, and
L = (2/3) ? U
Therefore: V = 5.6L = 5.6(2/3)U = 3.73U
See Willard S. Smith, "Housing in the Soviet Union-Big Plans, Little Action," in Soviet
Economic Prospects for the Seventies, p. 406.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
removed. For 14 apartment buildings in the Dodge Digest, the cost of air-
conditioning can be isolated from total cost. The cost of air-conditioning in this
small sample averages 4.6% of total cost, with a range from 1.7% to 9.4%. This
average is used as the coefficient to reduce the Dodge Guide costs.
The highest quality Soviet 7- to 16-story apartments are compared with
average US-quality, concrete-frame, high-rise buildings. These buildings are
costed in both basement and non-basement versions for the smallest and largest
buildings in the Dodge Guide. The cost of air-conditioning is eliminated with
the same coefficient used for medium-rise apartments.
The ruble-dollar ratio for housing is a specially constructed average of the
individual building ratios. The ruble-dollar ratio for each particular type of
housing is the computed arithmetic mean of all the ratios of that type of con-
struction (e.g., single-family houses and five-story apartments). Unfortunately,
the Soviet data lack sufficient detail to permit the computation of value weights
for each category of housing construction. For example, although it is possible
to estimate the proportion of Soviet housing construction represented by both
three- and four-story apartments, the percentage that is three-story as opposed
to four-story is not estimable. Whenever this problem is encountered, the
categories are combined by geometric means up to the level of aggregation for
which a value weight can be estimated. In the above example, the geometric
mean of the three- and four-story ratios represents the three- and four-story
group. These geometric means are further combined by value weights into US-
and Soviet-weighted ruble-dollar ratios for housing construction.
Value weights are derived for US housing construction in 1970. The relative
weights of single-family and multi-family dwellings correspond to the value of
new construction put in place during 1970 for one-unit and two or more unit
structures. These relative values are directly available only for private construc-
tion; however, only the gross value total is available for public housing. Of
course, the preponderance of public housing is multi-family housing, but a small
single-family portion exists. This relationship is estimated by using a breakdown
of the number of construction starts of new units for public housing. This
represents a departure from strict value weights, however, because it assumes
that the units costs for public housing are totally unrelated to the number of
units per structure.
The weights within the multi-family category are based on a sample of 74
apartment buildings from the Dodge Digest. These buildings are classified by
number of stories, and the total value of construction in each classification is
computed. On the basis of this sample, the percentage distribution of the total
provides the weights within the multi-family sector. In order to improve this
sample by making it larger and more representative, the sample is not confined
solely to buildings erected in 1970. Although some were built either a few years
before or after 1970, the difference is negligible because distribution of number
of stories changes slowly over time. Another problem of bias in the weights
exists, however, because very tall buildings, for which we lack a meaningful
ruble-dollar ratio, are excluded. Correction of this omission would probably
raise the ruble-dollar ratio for housing.
Estimation of the Soviet weights proceeds from the tenuous assumption
that the sample's average ruble cost for each category of housing varies in the
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
same proportion as the actual average for all housing construction of that type.*
The ratio of single-family to multi-family housing, as measured by the quantity
of useful space constructed, is estimated by assuming that single-family housing
and private housing are identical and then by applying published 1970 data
arranged on this basis. Value weights are derived by multiplying these quantity
proportions by the average cost per cubic meter and adjusting the results to
add to unity. The weights within the multi-family housing category are obtained
by examining the distribution of state housing construction, by number of stories
as measured by useful space in 1970. These weights are shifted to a value basis
by a procedure analogous to that described above.
The derivation of the housing ruble-dollar ratio is given in Table D-2. The
results are summarized in the following tabulation:
US-
weighted
ratio
USSR-
weighted
ratio
Geometric
Average
Single-family housing .......
0.646
0.646
0.646
Multi-family housing .......
0.586
0.454
0.516
All housing ................
0.620
0.494
0.553
The ratios by type of structure and number of stories suggest that Soviet
costs are highest relative to US costs in the construction of single-family housing,
low-rise apartments (1-3 stories), and high-rise buildings. The primary Soviet
advantage lies in the construction of medium-rise buildings, especially the five-
story apartments prevalent in the USSR. As expected in international compari-
sons of this kind, however, the Soviet-weighted ratios are less than the
US-weighted ratios.
The ratios in Table D-1 provide further insight concerning factors influenc-
ing the ratio. The ruble-dollar ratio is higher for buildings with elevators than
for those without them. Elevators in the USSR are expensive-many are im-
ported-and frequently are a large component of total cost. Furthermore, the
ruble-dollar ratio is directly related to construction volume for low-rise apart-
ments, but the ratio is inversely related to size for buildings above three stories.
This suggests that the USSR has relatively greater scale economies in construc-
tion than the United States, but the inclusion of elevators and other costs
incurred by making buildings taller negates this advantage. Finally, structures
with a basement, except for the 10- and 11-story buildings, have higher ruble-
dollar ratios than those without one. Perhaps this is the consequence of greater
US efficiency in performing excavation work.
*This assumption may be mathematically expressed as follows:
X1 X2 X?
_ ...-=k
Al ?2 ?o
where,
Xi = average ruble cost per cubic meter in the sample of housing of type i
?i = average ruble cost per cubic meter in all housing of type i
k = a scalar constant
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Composi-
Composi-
tion of
tion of US-
Soviet-
Group
Weighted
Group
Weighted
Type of Structure
Ruble- Ruble-
Dollar Dollar
Ratio Ratio'
US
Group
Weights 2
Ruble-
Dollar
Ratio 3
Dollar-
Ruble
Ratio 4
Soviet
Group
Weights 5
Dollar-
Ruble
Ratio B
Single-family housing..
0.646 0.646
0.584
0.377
1.548
0.275
0.426
Multi-family housing
One-story ............
0.6731 "0.653
0.202
0.132
1.531
0.091
0.139
Two-stories...........
0.633)
Three-stories..........
0.706 0
546
0.044
0.024
1.832
0.061
0.112
Four-stories ...........
.
0.422
Five-stories...........
0.410 0.410
0.006
0.002
2.439
0.421
1.027
Six-stories ............
0.472
Seven-stories..........
0.457
0.464
0.034
0.016
2.155
0.127
0.274
Eight-stories ..........
0.452
Nine-stories...........
0.477
Ten-stories ...........
0.543
Eleven-stories.........
0.540
Twelve-stories ........
0.533
Thirteen-stories .......
0.526
0.534
0.130
0.069
1.873
0.025
0.047
Fourteen-stories.......
0.507
Fifteen-stories.........
0.554
Sixteen-stories ........
0.538
1 1
Geometric mean of US- and Soviet-weighted ruble-dollar ratios = (0.620 x 2.025 ) 2 = 0.553.
1 Geometric mean of ruble-dollar ratios for each story group.
2 Dodge Digest, September 1972, pp. K101-K109, and Construction Review, September 1974, p. 25.
3 Column (3) x column (4).
4 Reciprocal of column (3).
5 The weight for single-family structures versus multi-family structures is obtained by the assump-
tion that all private construction is single-family and no state or co-op housing is. All the Soviet
weights are derived from data in Willard S. Smith, "Housing in the Soviet Union-Big Plans, Little
Action," Soviet Economic Prospects for the Seventies, June 1973, pp. 412, 423.
6 Column (6) x column (7). The Soviet-weighted ruble-dollar ratio equals 0.494.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
APPENDIX E
Two approaches are used in deriving a ruble-dollar ratio for the construction
of office buildings. The first approach uses a quasi-random sample from the
Dodge Digest and computes the ratio with a methodology analogous to the
US-representative approach for hospitals (see Appendix C). Alternatively, a
stratified sample of office buildings is formed by using the Dodge Guide for US
costs and the Sborniki for Soviet costs.
The Dodge Digest Approach
The best Soviet office buildings are compared with US construction to
minimize problems of comparability. Thus, the most durable structure in any
office building category in the Sborniki is always used. These categories are
classified by the number of stories in the building. For most buildings, the cost
estimator has an option between a simple (prostaya) and an enhanced
(povishennaya) interior finish, the latter being 7%-11% more expensive than
the simple one. The enhanced finish is more comparable to the quality of US
construction and is used in this study. Adjustments are made to the cost of
Soviet buildings, when appropriate, for including items such as hot-water supply,
elevators, cloth-backed linoleum, and precast reinforced concrete roofing sheets
with interior waterproofing. Data for the US and Soviet cost estimates are given
in Table E-1.
The US costs of the 32 office buildings are adjusted to conform to Soviet
practice. The cost of built-in equipment is deducted from the US cost. The cost
of air-conditioning in each element of the sample is unknown, so building costs
are reduced by 6.8%-the average estimated from an admittedly limited sample
of 13 office buildings where the cost of air-conditioning was ascertained. As with
hospitals, the bid prices are adjusted to June 1970 dollars by the Dodge index of
construction costs for the nearest major city and converted to 83% of the
construction costs of New York City to eliminate regional disparities.
The Dodge Guide Approach
This approach uses two weighted averages to compute the office building
ratio. The methodology for estimating the weighted averages is analogous to the
procedure used in Appendix B for housing construction. A ruble-dollar ratio for
a building with a given number of stories is derived by using the arithmetic mean
of the smallest and largest office buildings in the Dodge Guide. The ratios for
each building height are then grouped to conform with estimable Soviet weights,
with the geometric mean of the elements in each group serving as the applicable
ratio for that group. These groups are further combined into Soviet- and US-
weighted ruble-dollar ratios, with the geometric mean providing the ultimate
result.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
M N M a M CD 00 00 CD -- 'n N 00 " C C M M N ~ -- n 10 N n
n C7 N O O CA Cn M O C 00 C0 V' M 00 C0 N M N M N
M C' V' a' d' Cn O CCJ M C' C M O M N ~' Cn C M Cn ~'
O C 0 0 C C 0 C C 0 0 C 0 C C O O C O 0 0 0 0 C 0
00 00 00 00 CD Ul 00 M GC1 00 00 O 00 00 O Ul 00 00 c0 O M 00 O
00 00 0 0 00 N 00 OCI 00 V' 00 01 m = 00 00 = 00 00 00 N = 'n 00 CP --
0 0 0 C C C 0 0 0 0 0 0 0 0 0 0 0 0 0 o o -- O O C
-- -- .-- -- -- -- -- -- -- -- M -- M -- -- M -- -- -- -- M M -- M M
N N N N N N N N N N n t- u0 N N 'n N N N N u0 00 N n CC0
N N N N N N N N N N N N N N N N N N N N N N N N N
C7> N N Cn N N M M N N CD N M M 00 N O N N N. O a' N - N IC
M O N N N O Cn -- O 00 O C0 00 M O O= N .- C C - CA CD
M "I O Cn CA N -- O ?--' ?-?' CA -- ?-- O --- O N CA O O O O 0n CA
V' O M N CP O 00 O N O CC's ?-- N V' Cn Cn N C 00 -?' N a_0 O O CP
-- O M Lt 00 CA C7> Cn N M -- Cn 00 Cn c Cn C N Cn M N 00
O O N M M a' ?C' ~' Cn CD CD N N N O a' Cn CD N O N O -- -- M
--' ?--'--' -- -- -- -- -- --' -- ?-' -- ~. .-+ N N N N c q- M C V' " C
A O
O O
OC
to O N 00 N N O O N O 00 O O M O N M Cn C' O N m -- CD
U Cn N ob M N c0 CD 00 O O m m N O N N M N Cn N M cD
6> y V1 N O 00 r CA ' M N G V' N N M N O M O -- O 00 . N -- O N
d ty '~ CD CD M ~n CD CD M Cn Cc] N N u0 c0 C' N a' Cn C+0 Cn CD N Cc]
Cn -- C I' N N M CC. -- O Cn 00 ~' Cn CC] ' 00 N 00 C0 N CD CD
00 O N 00 O a0 M m s m a' N -- M m O O N M 00 N v' Cn CD
?- -- -- .-- .~ -- ^ ^ -- N N N N N
O O CP N N L- N V' ~' W O O 00 CD CO Cn 00 N CD 00 -- O -- N
O O N N Cn O MC'i CD O O m CA M M CO D O O M CC] Cn ?-- N
MM M- - M MM O O O CA O O MI=1=01 O O O C IM CA O O O
00 CA O
N M e' Cn O N IXI Q~ O -- N M C'
C
N N N N
?-- -- -- ?--' -- ?-- ?-- ?-- N N
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
M N 00 O . t- C)
M uJ d~ era V~ 00 cD
0 0 0 0 0 0 0
.. )C N c0 c0 O N
N u~ 1- M l~ 00 r
cD t`
N C) t- h 00 N-
O O C) O O O O
--i -r O --~ O O --~
.,J t` on O O C
N N N N M M M
00 O N M I- M cC
O 00 O )0 00 N I
M N (y0
00 00 C) 00 O .0 O
V )0 O O h O
O 0 0 0 0 0 0
N N N )0 N N t-
?-~ O ?--~ t- N N C)
O 00 C) O C) C) N
-? O O O O O -~
t` 00 O V' O O O
M O ~ 00 l~ ~' N
N 00 t- O --~ O C
c0 .0 c0 N ~' M
N 00 O t.- O n N
CC to w n ' ? Q)
LID ti N 00 ~
C cd
G V
O G '9
U a)
y a U
V
w
C
v a o X
y o cU. .~
V )
C y V
L C C N
a) C
E O
O C 0
O C d
C O O
>
U O M
a
'n L V w
V V X
M H G
O O of
C
w U
75
II > L...L
V O 00
? C
> . II 0
r -w U
O
O V O
O O O
y'7 'O t`
F
L
C m O E
U U C a
V O
O L G O V
N '"' 'O X t`
ti Gp >) .
C7 M
: D v . O ... N
7 L C o0 C
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
The US weights are derived from a sample of approximately 200 office
buildings in the Dodge Digest. This sample includes more than $136 million worth
of contract bids. Each structure is assigned to a group, depending on the number
of stories, and the bid costs for each group are summed. This grouping provides
the data for deriving a percentage distribution to use as weights.
The Soviet Union has not published data for office buildings distributed by
number of stories as has been done for housing. Since the economics of construc-
tion in the USSR are similar for office buildings and apartments, the housing
weights as measured by quantity are applied to office buildings. These weights
are then adjusted to a value basis on the assumption that the sample average
cost per cubic meter for each story varies in the same proportion as does actual
average cost.
Ruble cost estimates are based on the same methodology as the Dodge
Digest approach, but differences exist for the dollar methodology. The costs in the
Dodge Digest are already adjusted to mid-1970 prices, with regional differences
eliminated. US office buildings of average quality are always used. These
buildings are grouped into three categories: walk-ups (1-3 stories), low-rise
(1-4 stories), and high-rise (more than four stories). Only the low-rise are without
air-conditioning, so the standard 6.8% correction is used in the other cases. Since
two building categories overlap for one to three stories, the ratio for these levels
represents the average of four ratios: two for walk-ups and two for low-rise. The
basic costs of high-rise office buildings do not include interior partitions, so they
must be added. Since finished gypsum board is a common Soviet wall material,
it is used for the US costs at a rate of $1.80 per square foot of partition. With
the quantity of interior walls assumed to be equal to one-third of total floorspace,
the cost per square foot of floorspace is 60 cents; this amount, which is equivalent
to $1.77 per cubic meter, is added to the basic cost. The data for individual
observations are given in Table E-2.
The results of the Dodge Digest approach appear in Table E-3. The
unweighted and value-weighted ratios are 0.491 and 0.525, respectively. The
coefficient of variation for the unweighted ratio is 24%; a 95% confidence
interval encloses the mean in a range from 0.448 to 0.534. Testing the dispersion
of the individual ruble-dollar ratios suggests they are distributed with a strong
central tendency.* Since the unweighted ratio is less than the value-weighted
one, this indicates the existence of a positive scale effect with the ruble-dollar
ratio increasing with project size.
The ratio for the Dodge Guide approach is derived in Table E-4. The
result-0.471-is substantially less than the result from the Dodge Digest
approach. The Soviet-weighted ruble-dollar ratio is less than the US-weighted
ratio because the Soviets perform better at their own construction mix than at
the US mix. Thus, the lowest ruble-dollar ratio is for five-story buildings, which
is their most frequent structural type. The ratios for three- and four-story struc-
tures are higher probably because they are five-story buildings redesigned for
fewer stories, and the buildings are thus apparently stronger than necessary.
The ruble-dollar ratio increases for buildings above five-stories because the United
*The chi-square test for central tendency results in a test value of 6.00. A value larger than 14.07
is needed to refute the assertion of a central tendency at 95% confidence.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Number of
Volume of
Structure'
(Cubic
US Cost
per Cubic
Meter 2
Total US
Cost 3
(Thousand
Soviet Cost
per Cubic
Meter 4
Total Soviet
Cost 5
(Thousand
Ruble-
Dollar
Stories
Meters)
(US $)
1970 US $)
(Rubles)
1970 Rubles)
Ratio 6
1 ..............
657
71.40
46.9
29.5
19.4
0.414
1 ..............
3,285
79.76
262
29.5
97
0.370
1 ..............
6,570
51.48
338
29.5
194
0.574
1 ..............
22,993
46.40
1,067
28.0
644
0.604
Average........
8,376
62.26
428
29.1
239
0.490
2 ..............
657
71.40
46.9
27.6
18.1
0.386
2 ..............
3,285
79.76
262
27.6
91
0.347
2 ..............
6,570
51.48
338
27.6
181
0.536
2 ..............
22,993
46.40
1,067
26.2
602
0.564
Average........
8,376
62.26
428
27.2
223
0.458
3 ..............
657
71.40
46.9
30.9
20.3
0.433
3 ..............
3,285
79.76
262
32.6
107
0.408
3 ..............
6,570
51.48
338
30.9
203
0.601
3 ..............
22,993
46.40
1,067
31.0
713
0.668
Average ........
8,376
62.26
428
31.4
261
0.528
4 ..............
3,285
79.76
262
32.7
107
0.408
4 ..............
22,993
46.40
1,067
31.0
713
0.668
Average........
13,139
63.08
664
31.8
410
0.538
5 ..............
26,280
81.09
2,131
31.0
815
0.382
5 ..............
45,985
67.62
3,110
31.0
1,426
0.459
Average ........
36,132
74.36
2,620
31.0
1,120
0.420
6 ..............
31,536
81.09
2,557
32.7
1,031
0.403
6 ..............
55,182
67.62
3,731
32.7
1,804
0.484
Average ........
43,359
74.36
3,144
32.7
1,418
0.444
7 ..............
36,792
81.09
2,983
32.7
1,203
0.403
7 ..............
64,379
67.62
4,353
32.7
2,105
0.484
Average........
50,586
74.36
3,668
32.7
1,654
0.444
8 ..............
42,048
81.09
3,410
39.3
1,652
0.484
8 ..............
73,579
67.62
4,975
39.3
2,892
0.581
Average ........
57,814
74.36
4,193
39.3
2,272
0.532
9 ..............
47,304
81.09
3,836
39.3
1,859
0.485
9 ..............
82,773
67.62
5,597
39.3
3,253
0.581
Average........
65,038
74.36
4,716
39.3
2,556
0.533
10-14 ..........
n5,2567
81.09
n4267
44.2
n2327
0.545
10-14 ..........
n9,1977
67.62
n6227
44.2
n4077
0.654
Average ........
n7,2267
74.36
n5247
44.2
n3207
0.600
' Dodge Guide, pp. (2-3)-(2-6). US costs given on a square foot basis are converted to cubic feet
and subsequently to cubic meters with the assumption that the height of a story averages 11.6 feet.
2 Ibid. Costs for buildings with air-conditioning receive a standard 6.8% deduction to eliminate
the air-conditioning and conform to Soviet practice.
3 Column (2) x column (3).
4 Sborniki, Budget Series, no. 4, pp. 82-97.
5 Column (2) x column (5).
6 Column (6) - column (4).
7 The symbol "n" represents the number of stories where n has a range from 10 to 14.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Range
Volume (cubic meters) ...............................
4,705
1,014
23,912
Soviet cost (thousand rubles) .........................
138
30
742
US cost (thousand US $) .............................
263
61
1,487
Soviet-weighted cost (rubles per cubic meter) ...........
29.3
25.7
32.6
US-weighted cost (dollars per cubic meter) .............
55.8
35.6
86.2
Unweighted ruble-dollar ratio .........................
0.491.
0.305
0.837
Value-weighted ruble-dollar ratio ......................
0.525
....
....
Composition
Composition
of US-
of Soviet-
Group
Weighted
Group
Weighted
Ruble-
Dollar
Ruble-
Dollar
US
Group
Ruble-
Dollar
Dollar-
Ruble
Soviet
Group
Dollar-
Ruble
Number of Stories
Ratio
Ratio 1
Weights2
Ratio 3
Ratio 4 Weights5 Ratio 8
1 .....................
0.4901
0.474
0.341
0.162
2.110
0.121
0.255
2 .....................
0.458
3 .....................
0.5281
0.533
0.125
0.067
1.876
0.086
0.161
4 .....................
0.538)
5 .....................
0.420
0.420
0.305
0.128
2.381
0.572
1.362
6 .....................
0.4441
7 .....................
0.4441
1
0.486
0.032
0.016
2.058
0.183
0.377
8 .....................
0.5321
9 .....................
0.5331
10 .....................
0.600
11 .....................
0.600
12 .....................
0.600
0.600
0.197
0.118
1.667
0.038
0.063
13 .....................
0.600
14 .....................
0.600
Geometric mean of US- and Soviet-weighted ruble-dollar ratios 0.491 x 2 2181/2=0.471
1 Geometric mean of ruble-dollar ratio for each story group.
2 Dodge Digest, pp. G40-G420.
3 Column (3) x column (4).
4 Reciprocal of column (3).
5 Willard S. Smith, "Housing in the Soviet Union, Big Plans, Little Action," Soviet Economic
Prospects for the Seventies, June 1973, p. 412.
6 Column (6) x column (7). The Soviet-weighted ruble-dollar ratio equals 0.451.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
States has a comparative advantage in tall buildings. Data in Table E-2 confirm
the presence of the positive scale effect in that the ratio changes while the
number of stories is held constant. This indicates that US industry obtains
greater economies of scale for the construction of large office buildings than
the USSR.
The reason for the difference between the ratios derived by the Dodge
Digest and Dodge Guide approaches is easily explained, although it is unclear
what to do about it. Table E-5 disaggregates the ratios of each approach by
number of stories; the table stops at five stories because that is the tallest building
in the Dodge Digest sample. The only gross discrepancy between the results of
the two methodologies is for four-story buildings. Close examination reveals
that the four-story buildings in the Dodge Digest sample are larger than the
Dodge Guide sample. The already demonstrated positive scale effect explains
why the ratios based on four-story Dodge Digest are larger; moreover, these
ratios exert a large weight in the computation of the value-weighted ratio-i.e.,
more than 20% of the value of the entire sample.
The geometric mean of the value-weighted ratio in the Dodge Digest ap-
proach (0.525) and the US-weighted ratio in the Dodge Guide approach (0.491)
provides the US-weighted ruble-dollar ratio (0.508). Likewise, the geometric
mean of the US-weighted ratio and the Soviet-weighted ratio (0.451) equals
0.479 and is used for the office building ratio.
Number of
Number of
Number of Stories
Observations
Ratio
Observations
Ratio
1 ..........................
19
0.494
4
0.490
2 ..........................
8
0.415
4
0.458
3 ..........................
2
0.523
4
0.528
4 ..........................
2
0.748
2
0.538
5 ..........................
1
0.455
2
0.420
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
APPENDIX F
The computation of a ruble-dollar ratio for school construction is similar to
that used for office buildings. The Dodge Digest approach, which uses a sample of
23 elementary schools and 10 intermediate schools, results in a ratio representa-
tive of US construction. The Dodge Guide approach determines a ruble-dollar
ratio by number of stories for each type of school and then combines them into
an aggregate ratio based on estimated Soviet weights. In each approach, high
school and college buildings are not considered because true international
physical comparisons are difficult. Moreover, nonquantifiable socially determined
esthetic tastes probably have a larger impact on the cost of these buildings than
on elementary and intermediate school buildings.
The Dodge Digest Approach
The best quality Soviet schools are compared with the US sample. The
building categories in the Sborniki are classified by number of stories. The
7%-9% more expensive enhanced (povishennaya) interior finish is used in
preference to the simple (prostaya) finish. The basic Soviet costs are adjusted
to incorporate the cost of installed boiler equipment, hot-water supply, and a
ventilation system. The data for the US and Soviet estimates appear in
Table F-1.
The US construction costs of these schools are adjusted to conform to Soviet
practice. The cost of built-in equipment except boiler equipment is removed
from the US costs. Only one of the elementary and eight of the intermediate
schools are air-conditioned; in these cases the cost of air-conditioning is estimated
and deducted from the total cost. The bid costs are adjusted to June 1970 dollars
by the Dodge index of construction costs for the nearest major city and altered
to 83% of New York City construction costs to eliminate regional variation.
The Dodge Guide Approach
The methodology of the Dodge Guide approach which uses Soviet weights
results in a ratio more nearly representative of Soviet construction. The pre-
ponderance of US school construction is of one-story or at most two-story
structures. On the other hand, a large proportion of Soviet school buildings are
taller, mainly five-stories. Generally, US schools are built out and Soviet schools
are built up. Since the ruble-dollar ratio depends partly on the number of stories,
the Dodge Digest sample excludes construction that the Soviets perform best;
this imparts an upward bias to the ratio.
The Dodge Guide approach partly avoids this problem. First, a ruble-dollar
ratio for a given number of stories is calculated by using the arithmetic mean of
the ruble-dollar ratio for the smallest and largest building in each category.
These ratios are grouped by geometric means into a series of ratios compatible
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
N N M N O(7 O O m m a' 00 N h M M 00 M CD M " N
N CfJ N 0 0 00 O 0) N OCJ OCI ?J CO ,j CO M C C h O N O CO CO
Cq C1j Ct~ l~ Cry M Cn h Cn O OC) V' O 00 h M t' Cn C!7
0 0 0 0 0 0 0 O O O O O O O O O O O O O 0 0 0 O O
r O
O 0~
U ?- ,~
,O % O O .--i W O ti M ?-i m O.0 Cf~ O M M 00 C9 W M
l'- O ?--~ O O N M C] N CO CO CO h ti V' U 0 N N N O M CO
N N N N N N N M M M M ,1' a' V C[J Cry Cry Ctj 'V' O N
0
o a]
F i?. O 00 l~ N N N O O 00 M Cn OS) 00 O O on CD --~ O -' l~
eM l- t'.
O cD a CO O N ? J V' 00 N V' Co N ti 0 a> 0
U RS C/~ '~ N N a' M M 0 0 0 0 0 0 u') QC 00 0 T O M.
M M 00 0 N 00 7 0 w CO N w a' Ca C] 1- 0 an n -' Cn
N C" d' O M d' M V' M ~-' N 00 IX) n CO ?-' M M C[) 00 CO
T m 01 N T ?-+ O O .--~ C O ?? ?? O ?-? O 01
V] V
'fl O N c- N O N M 00 O a' O- a' O O n M O O ~' N M V'
~?` V U M r+ u'7 M M O N M Cry .--' O ~N O n V' ?C!' O T l~ T M Cry C' d'
C4 O d y O 00 . O Ct. 0 a' O M. N N N O 00 ~' M N N C-
a. ~r CO V' M ~J 'C'J 'CJ M M '' h C' C'7 CO a' M M C!) '' M a' CO a' d'
O O / v
O O
401,
O
O CS
I. h ' 7
O O CO CI! 1O h O O O O O O V' h O0 M a' ^-* CO ,~ CD
w e n u O w O O n n tti ti a) a CD l'- h t'-
N N N N N N N N N N N N N N N N N N N N N N N
O N O Cn ' 00 F- C* d' M V' N d' On 0 N W r- N
O O O CO CO CO M CO 00 C] CO F- M CO N l- 00 0 N 00 O> N
O O O 0 00 O O O O O N O> oo O O T O0 O O O T N O>
.Oa rai -r - -r .??i O -- -- -y -? O --+ O O O ?--~ O O O .??~ O O .--i O
o'fl
W ' L
o y e O o0 CC C] O~ O N N O M O O0 cc M M O0 N
O N 0 N
d w 0> N N O O N h M t- O 00 CD C] t` 0 N
O V' CCU N ~' on 0 O l- V' CD M g w 00 a0 ti 00 h O 0~ M O V' O Cry O ti
a ` V M ~' V' 00 IXI Oft 0+ O O O M M M V' uJ Cry CD h 00 0- O M
U
W ci ri w I* Cc ti ai a, o .; C~ m~ n c ~ a o o ci m
C11 C11 04
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
O N
-
V
1111
C
CO
t-
C) C)
1
CO
m
E
O 0
a)
w
a) C)
00
OO
t`
t-
y
CD C)
CO CO
~J
M
CD
D
h
M
0
0
0
0
0
0
0
0
0
0
O
O
O
O
C
O
~.e
U
U
0
M
M
)0
C)
0
O()
t-
CO
O
1-
ti
CO
U
E
t?
CO
M
O
O
M
CD
CD
C)
M
N
CO
O
V~
y
t`
CD
C)
M
t.
CD
M
CO
uJ
O
?
N
tom..
rn w 1, m o N. OO CO cO M
)0 )rn CO CO 0 Cc t- "r C) )r
N N N N M N N N N N
M
t-
~'
M
N
OO t.
C)
M
N
O
C)
)0
CO
t- 0
M C)
O
O
T CO
O
l~
C~
CO
C U
.-?i
--i
N
m ?--~
O
)0
N
t`
C-
C)
OO
M V~
O
O
CC
T
_
02
t..
N
.0
O
M
M
N-
O
'CN CO
C) O
-y
O
-r -r
-?i
O
-?i
-r
O -?~
C)
O
O O
))7
.0
C
1"'
CO CO
CO
C)
M
O
CD
C) C)
N
C)
?)
CO
W
O
C)
C)
M
O
0-.000000--
O
N
i
M
V~
CC
11:
CD
117
O
N
t~
)t
OO
O
n
n
C)
N
.
N
M
M
N
C+9
M
CM
M
.0
0
o
O
~
" Y
M
M
.0
CD
O
OO
t`
M
t`
C)
CO
C)
CD
M
CD
CD
C)
P.
V~
N
t`
.0
O
C)
C
-
Y O
UI
00
00
~
M
C
m
M
O
M
.
-i
N
N
M
O
CD
N
O
U
Z
d
w
_N
O O
C) o o
U C+
O U
O U N U
.E e
m 0 'S
CCO^ ., O p
d II
Y E L
.0 .a o ?D
E c
?~ a) a m
E Z o
a)
D
?7
C
a O
d C r
0
U O O y C aV
a) m ? n
_ M
C O U O
_O C ~7
V L '? C
d o o.
N Y ~ O
Q p .2 C O
- U x ;p x p
a`) C Cu C C) C 0
O O o N y v C
CY C C
a ?? a a o
d C) ~+ p O O U
CO C O U
x ?? x ?~?
y O O C
n ~ .y v
Q O pa,'~ C .y C C
F F
ro ro roro 8 ? 7
Q )- U CO U U
., ~, M e ,~ m o m
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
with the estimated Soviet weights. The weights are derived by assuming that the
distribution by number of stories is identical with that for apartment buildings of
up to five-stories, although there are a negligible number of schools above
five-stories. This quantity distribution is then adjusted to value weights by
using average cost per cubic meter as is described in Appendix D.
The ruble costs are estimated the same way as in the Dodge Guide approach,
but there is a difference in the dollar costing. When using the Dodge Guide for
other forms of construction, average quality construction was always used, but
the source reports only construction costs for good and excellent quality schools.
A sample of the other categories of construction, however, reveals that average
quality structures cost approximately 83% of good quality, so the US school
costs are reduced to that percentage. As before, the US costs are automatically
adjusted for regional and temporal differences. Moreover, the expense of air-
conditioning has been eliminated from these costs. The data for this methodology
are in Table F-2.
Table F-3 contains the results of the Dodge Digest approach. The ruble-
dollar ratios for elementary schools are less than the corresponding ratios for
intermediate schools. The value-weighted ratio is less than the unweighted ratio
for both types of school, although combining the two samples virtually eliminates
the apparent scale effect. The coefficient of variation of these three samples
ranges between 21% and 24%. Confidence intervals of 95% established around
the unweighted ratios provide estimate ranges of the unweighted ratios of 0.506-
0.622 for elementary schools, 0.547-0.741 for intermediate schools, and 0.539-
0.637 for the combined sample. Testing the dispersion of the individual ratios
suggests an existing central tendency for the samples of both the elementary
schools and combined data, although the size of the sample for intermediate
schools is too limited for meaningful conclusions.*
The Dodge Guide ratios are computed with Soviet weights in Table F-4.
The geometric mean of the Dodge Guide ratios for elementary and intermediate
schools provides that approach's overall school ratio. These ratios are substan-
tially lower than the results of the other methodology which accords with the
expectation that the Soviet-weighted ratio would be lower than one weighted
by US construction. A test was performed to ascertain that the difference be-
tween the results of the two approaches is really a product of the weighting
systems. This test used the Dodge Guide ratios for number of stories and weighted
them with US weights for the same number of stories as derived from a large
Dodge Digest sample. These US-weighted ratios-0.588 and 0.606 for elementary
and intermediate schools, respectively-are nearly equal to the value-weighted
ratios from the Dodge Digest approach. Additional corroboration is provided by
the decline of the ruble-dollar ratio as the number of stories increases for both
school types.
Table F-2 shows that even though the ruble-dollar ratio diminishes as the
number of stories are increased, the ratio increases with the size of the structure.
*The chi-square test for central tendency gives value of 6.83 and 4.75 for elementary schools and
the combined sample, respectively; a test value above 14.07 would have required serious
examination of the central tendency assertion. A test value for intermediate schools would have
dubious validity because of the limited sample size.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Table F-2
School Sample: Dodge Guide Approach
- - - - - - - - - - - - - - -
(2) (3) (4) (5) (6) (7)
Total
Volume of Total US Soviet Soviet
Struc- US Cost Cost 3 Cost per Cost 5
lure 1 per Cubic (Thousand Cubic (Thousand Ruble-
(Cubic Meter 2 1970 Meter 4 1970 Dollar
Elementary schools
1 ........................ 3,681 54.16 199 24.6 91 0.457
1 ........................ 25,768 44.01 1,134 30.4 783 0.690
Average .................. 14,724 49.08 666 27.5 437 0.573
2 ........................ 3,681 54.16 199 24.0 88 0.442
2 ........................ 25,768 44.01 1,134 25.5 657 0.579
Average .................. 14,724 49.08 666 24.8 372 0.510
3 ........................ 3,681 54.16 199 23.4 86 0.432
3 ........................ 25,768 44.01 1,134 23.9 616 0.543
Average .................. 14,724 49.08 666 23.6 351 0.487
4 ........................ 3,681 54.16 199 23.0 85 0.427
4 ........................ 25,768 44.01 1,134 23.2 598 0.527
Average .................. 14,724 49.08 666 23.1 342 0.477
5 ........................ 3,681 54.16 199 22.7 84 0.422
5 ........................ 25,768 44.01 1,134 23.0 593 0.523
Average .................. 14,724 49.08 666 22.8 338 0.472
Intermediate schools
1 ........................ 22,087 54.70 1,208 26.4 583 0.483
I ........................ 58,899 40.72 2,398 30.8 1,814 0.756
Average .................. 40,493 47.71 1,803 28.6 1,198 0.620
2 ........................ 22,087 54.70 1,208 25.1 554 0.459
2 ........................ 58,899 40.72 2,398 29.4 1,732 0.722
Average .................. 40,493 47.71 1,803 27.2 1,143 0.590
3 ........................ 22,087 54.70 1,208 23.8 526 0.435
3 ........................ 58,899 40.72 2,398 24.7 1,455 0.607
Average .................. 40,493 47.71 1,803 24.2 990 0.521
4 ........................ 22,087 54.70 1,208 23.2 512 0.424
4 ........................ 58,899 40.72 2,398 23.7 1,396 0.582
Average .................. 40,493 47.71 1,803 23.4 954 0.503
5 ........................ 22,087 54.70 1,208 23.0 508 0.421
5 ........................ 58,899 40.72 2,398 23.5 1,384 0.577
Average .................. 40,493 47.71 1,803 23.2 946 0.499
1 Dodge Guide, pp. 5-4. US costs given on a square foot basis are converted to cubic feet and
subsequently cubic meters with the assumption that the story height averages 13 feet.
2 Ibid. Costs are presented for only good or better quality schools, so the quality is reduced to
average by a multiplicative correction factor of 0.83.
3 Column (2) x column (3).
4 Sborniki, Budget Series, no. 7, pp. 41-50, 63-72.
5 Column (2) x column (5).
6 Column (6) = column (4).
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
This partly conflicts with the result of the other approach, which indicates,
although ambiguously, the opposite scale effect. It is more plausible, however,
for the ruble-dollar ratio to increase with school size when the number of stories
is held constant. Larger buildings cannot derive as much from the Soviets'
advantages of the industrialization of construction and the standardization of
plans because they are less homogeneous. This view is even partly supported by
the Dodge Digest (in addition to the Dodge Guide) results where the intermediate
schools have a higher ratio than the elementary ones. In both cases the inter-
mediate schools represented by each sample are larger structures than the average
elementary school.
The geometric mean of the value-weighted Dodge Digest ratio and the
Dodge Guide ratio is used for the final school ratios. These aggregate ratios are:
Elementary schools-0.517
Intermediate schools-0.565
All schools-0.540
A.
Elementary (sample size= 23)
Volume (cubic meters) ............................
13,390
3,094
33,726
Soviet cost (thousand rubles) ......................
359
76
931
US cost (thousand US $) ..........................
651
170
1,571
Soviet-weighted cost (rubles per cubic meter) ........
26.8
24.4
28.4
US-weighted cost (dollars per cubic meter) ..........
48.6
31.8
70.2
Unweighted ruble-dollar ratios .....................
0.564
0.393
0.893
Value-weighted ruble-dollar ratio ...................
0.552
....
....
B.
Intermediate (sample size = 10)
Volume (cubic meters) ............................
30,557
17,975
63,244
Soviet cost (thousand rubles) ......................
833
475
1,790
US cost (thousand US $) ..........................
1,350
634
3,037
Soviet-weighted cost (rubles per cubic meter) ........
27.3
24.8
30.0
US-weighted cost (dollars per cubic meter) ..........
44.2
29.3
63.3
Unweighted ruble-dollar ratios .....................
0.644
0.399
0.910
Value-weighted ruble-dollar ratio ...................
0.617
....
....
C.
Combined sample (sample size = 33)
Volume (cubic meters) ............................
18,592
3,094
63,244
Soviet cost (thousand rubles) ......................
503
76
1,790
US cost (thousand US $) ..........................
863
170
3,037
Soviet-weighted cost (rubles per cubic meter) ........
27.0
24.4
30.0
US-weighted cost (dollars per cubic meter) ..........
46.4
31.8
70.2
Unweighted ruble-dollar ratios .....................
0.588
0.393
0.910
Value-weighted ruble-dollar ratio ...................
0.583
....
....
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Table F-4
School Results: Dodge Guide Approach
-
- - -
- - - - -
- - - -
(2)
(3)
(4)
Number of Stories
Ruble-
Dollar
Ratio
Group
Ruble-
Dollar
Ratio 1
Soviet
Group
Weights 2
Group
Dollar-
Ruble
Ratio 3
Composition
of Soviet-
Weighted
Dollar-Ruble
Ratio 4
Elementary
1 ........................
0.5731
0.541
0.188
1.848
0.347
2 ........................
0.510
3 ................
0
4871
........
.
0.482
0.107
2.075
0.222
4 ........................
0.477
5 ........................
0.472
0.472
0.705
2.119
1.494
2.063
Intermediate
1 ........................
0.620 j
0
605
196
0
1
653
324
0
2 ........................
0.590 J
.
.
.
.
3
0
5211
........................
.
0.512
0.106
1.953
0.207
4 ........................
0.503
5 ........................
0.499
0.499
0.698
2.004
1.399
1.930
1 Geometric mean of ruble-dollar ratio for each group.
2 Willard S. Smith, "Housing in the Soviet Union: Big Plans, Little Action," Soviet Economic
Prospects for the Seventies, June 1973, p. 412.
3 Reciprocal of column (3).
4 Column (4) x column (5). The ruble-dollar ratios are 0.485, 0.518, and 0.501 for elementary,
intermediate, and all schools, respectively.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
APPENDIX G
The Approach
Two different samples were collected for derivation of a ruble-dollar ratio
for construction projects in industry. The first sample, which was taken from the
Dodge Digest, contains 27 observations. Since this sample is composed primarily
of light industrial projects, a second sample was collected from private industry
sources to represent construction in heavy industry. Neither sample is particu-
larly representative of its segment of industry, but most of the differences in
industrial investment projects are in terms of equipment and not construction.
Thus, a valid ratio for industrial construction need not be totally representative
of all industrial activity.
The data for the two samples, which are denoted as the "light industry"
and the "heavy industry" samples, are presented in Tables G-1 and G-2. The
dollar costs are adjusted to eliminate any air-conditioning and built-in equip-
ment, where required to parallel Soviet construction practices. Also, the dollar
costs are converted to mid-1970 dollars and to the costs of an average US location.
For the heavy industry sample, only the dates for the completion of construction
were available; this does not correspond to the practice elsewhere, because
bid-date dollars were adjusted to 1970 prices. To reconcile the data, an arbitrary
bid date of two years prior to completion of construction was assumed and the
dollars adjusted accordingly. The two-year lag is conservative because a longer
lag would result in a lower ruble-dollar ratio.
The Soviet and US projects are compared on the basis of size and similarity
of construction. This matching does not imply, however, that the Soviet plant
has the same production capacity as its US counterpart. Ruble-dollar ratios
depending on the productivity of labor, equipment, and other inputs which are
beyond the scope of this study would be needed for that kind of comparison.
Table G-3 summarizes the raw data and the resultant ratios. The unweighted
ruble-dollar ratios for light industry and heavy industry are 0.584 and 0.519,
respectively, and the coefficients of variation are 23% and 19%. The establish-
ment of a 95% confidence interval around the mean produces a wide-interval
estimate because of the large dispersion of ratios for both samples and the very
small sample size for heavy industry. The interval of the estimate suggests a
ruble-dollar ratio between 0.533 and 0.637 for light industry and between 0.423
and 0.619 for heavy industry. The large degree of overlap between the two
confidence intervals and the small sample for heavy industry negate any conclu-
sion that the light industry ratio differs from the heavy industry ratio. Testing
the dispersion of the individual ruble-dollar ratios indicates the existence of a
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
A
O
0
h
T'
00
a0
v~
.0.0
O 10
M M
N
O
t`
n CO
01 CO
M T
CO
N.
o0
CO
0
O
It O
N C.
h t`
61
O
0
0
0 0
0
0 0
0
0
0 0
O l' 1CO N 0
C- O l~ N
00 M
C- N
CO
n
M
..
M 00
CO CO
m U o
O .F.i Q1
N O 0 a' 00 N 0 00 n N N CO
V' CO N 01 n .. CO 00 00 N 0 4 O
CO t` ~ M M O CC0 ~ ~
0 0 0 0 0 0 0 0 0 0 0 0
O M O O CO h 00 M M 00
N CO M O.0 O .0 N. l-
.. X O 1- X CO C` C` m try O CO 01! N N m O W CD ..
CO 00 CO 'W N.1 O T CO V' 7~ C~ 00 00 a' 00 .2 00 M M CO 01 O
V~ CO N --~ .. .. .. .. .-~ .. .. .. N N N .. .. .... - ...
O "' N
1 CO N O N l~ 00 N N M ' 0 00 l~ t` .. CO M C~7 O -+ 00 t'. O
,O O 00 Q1 CO vJ O 01 O M 1' uJ M
00 00
CO .0 CO C'. N. O l'. O
N 01 C CO Vl 01 CO m 01 N N N O '0 .. CO
V O O .. N
O
- 0n
O OS
O
1. O q
9
m Q
O 8 61
d'
.Q
03 F
a
,fl
rQ
~. G
V 1.
O 6
a0 O yp
O V1
U v .7
M 00 N O O --~ o0 00 h O CO 1[J ?J M M CO N h N M
n N t` 0 o0 0o -: . O M O O V' O t` M 01 .-. O
O O `
. ~ . y O t
O N. N.
.~ .-~ O .~. .-. .~ ~N O
- N N CO N l~ N CO M O oD O M oo n ao M M l~ N M M
C O .. M .. CO a' M t` 00 00 N O 00 'C] ~' O M M
M O O 01 O -. 01 O 01 M O- 0 0 0 -. 01 .-' 01 --~ O 01 CP
O 01 N O CO 01 00 O N CC CO 00 CO 0 O 00 N N 10 00 CO N
M y N m .. N O O t` O .. M .. M 00 M O O CO--' C`
' d'
O M M CO M 01 N . N N NM 00 M N CO [~ N O T O CNN
O CO C - .. .. .. .. N .-. .. .-. 1' a' .. N ..
O O v p~ T N O
0 h h N CO N CO 00 N N M O O O -+ N O o0 O t` 01 d' CO O
M l~ N CO O DD ~ CO a0 M O 1f) 1 O .. m 00 l~ 1ry hCO CO oN0 Oc' 41 .. N M .-' N CD '0 M O N M CD l~ 00 M M M ..
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
O O
bi -r.
C. eo E o N m m
C G. G .G G d 0 eJ ,G G 0 . C Oo
o m o a a G
It m
.03 Co Co '? t1o
v m as E c m a a c a n ?` 'a .~
C p
Q: 11W
w C>
m w ti w
E ." c G? G E 6 c It It
.1 a1 w ..a .1 i1
'C G O v v" 'b 3 +~.~ O d" "' v, m v, y G G 7 G
CO N 0 d O ?q 0 0 0 .G .G N y N 0 O 0 '0 ad
U FW WU RC FUr U') ,P.ZZ~ ".?
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
.,I- r tc N O00
I'D 00 h M 00 N
h 07 10 O
0000 O O
N
h
h
cD
h
N
~'--~ M
.?. O
-7 c0 t
O
119
-*
M
O
O
0
-?
C
M
M
O
N
h
m
N
00
?
O
M
O
C ~
N
00
O
CO
N
O
O
N
".
O
M
1
O~
uJ
N M 00 M
N 00
M a h
N -+
CD O O 0
N O , --~
h O O~ N
N N M ~J
O
C m
O ?~
=
.~-i
O
W
O
O
N
N
N
M
N
CN
N
N
O
N
h
N
-O O
y h
~
m O
-
O ~ G
C 0 0
'O x
a.~
N
N-
h
h
cD
O
O
N
O
h
11
h?
b v
U C1 ~
O y O
x y
FV.
E
=
c0
M
to
M ..
a'
-
N
h
N
00
x 00
axi m N
CQ
O
M
M
h
v C
L d E
F o 5
`
= o
U
0
4. U q
=
V
? N X a
V
.??
N M
~
~
O [~
00
O
O
^
~
M
ma
X .~?
ti
a F, ^ -sz
w
.
i
.
y
E E
o a o o a
cu m v v~ ?o
U U 0
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
p m ?-? N O --N - cO
A.? ? n? ,nom I
a s 0 0 0 C
CD t
n N .n N p I
o "" N
:?~~ N
E.
~ U o d v oo N o -- h y
V ;fl y C~ d ~D 00 ~n h
;fix N
~L
^-
-~- F' V1 h .--, 00 w M 00 O
A ? x
U
o
.-? h N M M 00
p
`
C V ?
V
Q N N N h h
h h h ,~. L OU
N N N
.~. .A. '. .. .-. -. y o
i "O a~ h
0 CO
u~ 00
O .0
e0 C U X
1E G ^ U
N N N N N hO C "O
N >, ^ ? E V Oo Oo QNi C> 'C L C
C7 '"
C
c o 0 0 o
.c 'fl E o " U
00
d
.fir G V^ [~ O .n M N
C>0 00 C ^o>
V ^ + o boo
N q ~O o
D
?3
I c .. o a w o h ? U
n D o- o c L o
v 8 o :: c~ v n = 0
'~ = U Y h N o 0 C -C
M ao
c .. n h
o L c
. U Q d
c co .n ? oo
GCO : : .U v a cn cO :O .n
? O x '~ 7 ..
V C
b1l
a ~b.= o~ Fso o?
C C U 2 C U
C'I
Z ' ao a ; ? ci o0
a a
C, 0 CL -a -Z
c 00 a o o? o0
N M I.
y N 7
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
A. Light Industry (sample size = 27)
Range
Volume (cubic meters) ............................
23,355
1 , 509
138,581
Soviet cost (thousand rubles) ......................
360
22
2,176
US cost (thousand US $) ..........................
573
57
2,492
Soviet-weighted cost (rubles per cubic meter) ........
15.4
9.7
46.4
US-weighted cost (dollars per cubic meter) ...........
24.5
17.7
96.8
Unweighted ruble-dollar ratios .....................
0.584
0.332
0.873
Value-weighted ruble-dollar ratio ...................
0.628
....
....
B.
Heavy Industry (sample size = 5)
Volume (cubic meters) ............................
358,034
45,307
844,130
Soviet cost (thousand rubles) ......................
4,400
520
14,435
US cost (thousand US $) ..........................
7,634
1,037
23,259
Soviet-weighted cost (rubles per cubic meter) ........
12.3
5.0
27.7
US-weighted cost (dollars per cubic meter) ..........
21.3
9.4
68.9
Unweighted ruble-dollar ratios .....................
0.519
0.402
0.621
Value-weighted ruble-dollar ratio ...................
0.576
....
....
fairly strong central tendency for light industry. The heavy industry sample is
too small to conduct any tests.*
The value-weighted ratios are 0.628 for light industry and 0.576 for heavy
industry. In each sample the value-weighted ratio exceeds the unweighted ratio
by more than 7% ?/,. This accords with the expectation that the ruble-dollar
ratio tends to increase as the project size increases. The amount of customization
required in the larger projects prevents the USSR, from taking advantage of
standardization and other cost-saving practices in construction. The geometric
mean of the two value-weighted ratios-0.601-is used as the aggregate
industrial construction ratio.
*The chi-square test for central tendency results in a test value of 1.17. A value in excess of
12.83 would have been needed to refute the assertion of a central tendency.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
APPENDIX H
Three types of transportation construction are selected for this study:
roads, airfields, and railroads. A somewhat different methodology from that
previously described for other types of construction must be employed because
both the Dodge Digest and Dodge Guide are devoid of US cost data for trans-
portation construction. Two alternative sources-the 1972 Building Cost File:
Eastern Edition and the Building Construction Cost Data, 1972-are used instead
to cost various roads, airfields, and railroads described in the Sborniki. A major
drawback of this approach is that these alternative sources present the cost data
solely on the basis of a unit of construction-cost per kilometer of railroad or
cost per square meter of airfield-instead of the cost for a total project. Thus, it
is impossible to determine the effect of project scale on ruble-dollar ratios; for
example, differences cannot be discerned between the ruble-dollar ratio for a short
stretch of road and that for a major interstate highway system.
The Approach
Four types of road construction are included in the transportation sample-
concrete, asphaltic-concrete, bitumen-paved, and gravel. The road costs used for
the ruble-dollar ratio include expenditures on the road surface, normal excavation,
drainage, and grading. Excluded from the costs are site preparation, removal of
existing structures, relocation of utilities and other roads, and other indirect
construction costs.
The technical specifications for the four roads derived from the Sborniki
are presented in Table H-1. The ruble values published in these Sborniki tables
contain items in addition to the cost of earthwork and pavement such as small
bridges. Fortunately, an available table yields the percentage distribution of cost
by construction element. By using this information, it is possible to eliminate
the cost of these extraneous elements.
The dollar expenditure for these four roads is estimated from data contained
in the Building Cost File. The dollar costs in this source are given in 1972 prices
and represent construction costs in New York City. To render these dollar
amounts compatible to the values of other types of construction in this study,
the costs are converted to 1970 dollars by a multiplicative coefficient of 0.884
and to an average US regional cost by a coefficient of 0.83.*
The amount of earthwork per kilometer-in terms of both physical volume
and US costs-is equal for all of the roads except gravel. The total cost of earth-
work for the three paved roads comprises the sum of expenditures for excavation,
*The time coefficient is the ratio of the Department of Commerce composite cost index for June
1970 and June 1972. See page 6 in the main text for discussion of the regional adjustment.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Road Description 1
1. Concrete Road (7 meters wide,
18 cm thick concrete, 20,600
cubic meters ofearthwork) .
2. Asphaltic Concrete Road (7
meters wide, 9 cm thick
asphaltic concrete, 25 cm
gravel base, 20,600 cubic
meters of earthwork) ......
3. Bitumen Paved Road (7
meters wide, 6 cm thick
bitumen coat, bitumen shot
6 cm deep, 25 cm gravel
base, 20,600 cubic meters of
earthwork) ...............
4. Gravel Road (4.5 meters wide,
20 cm thick of gravel, 7,500
cubic meters of earthwork).
Average............'...
US Cost per Kilometer of Road
Soviet Cost per Kilometer of
Road
(1970 US $)
(1970 Rubles)
Ruble-Dollar Ratio
Earth-
Pave-
Earth-
Pave-
Earth-
Pave-
work 2
ment 3
Total4
work 5
went 6
Total 7
work 8
ment 9
Total 111
28,579
100,213
128,792
38,280
58,080
96,360
1.339
0.580
0.748
28,579
72,313
100,892
35,432
53,457
88,889
1.240
0.739
0.881
28,579
77,468
106,047
35,344
44,556
79,900
1.237
0.575
0.753
9,433
7,037
16,470
5,976
6,806
12,782
0.634
0.967
0.776
23,792
64,258
88,050
28,758
40,725
69,483
1.112
0.715
0.790
------------- -------------- ------- -
1 Sborniki, Budget Series, no. 23, pp. 6, 9.
2 1972 Building Cost File: Eastern Edition, Construction Publishing Company, N.Y., pp. 17-18.
3 Ibid., pp. 328-329.
4 Column (2) + column (3).
5 Sborniki, Budget Series, no. 23, pp. 6, 9, 15.
6 I bid.
7 Column (5) + column (6).
8 Column (5) = column (2).
9 Column (6) = column (3).
10 Column (7) 4- column (4).
95% compaction by a sheepsfoot roller, and the dumping and spreading of fill
and gravel materials in 15.24-centimeter layers. The quantity of excavation equals
the stated volume of earthwork, i.e., 20,600 cubic meters; a medium clay composi-
tion is assumed for the excavated material. Moreover, one-half of this volume is
actually dumped, spread, and compacted. The quantity of earthwork for the
gravel road is much smaller than the others; this is largely attributable to the
comparative narrowness of that road.
Costing the pavement portion of the construction in dollars depends on the
nature of the surface. All the road shoulders are considered as gravel or crushed
stone except that dirt shoulders are assumed for the gravel road. The cost of the
asphaltic-concrete road represents the sum of expenditures for crushed stone for
the road base and shoulders, a nine-centimeter-thick leveling course of asphaltic-
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
concrete, and a six-centimeter-deep shot of bitumen paving below the surface.
The bitumen paved road has crushed stone like the asphaltic-concrete, a six-
centimeter-deep penetration by the bitumen paving, and a'six-centimeter-thick
layer of bitumen paving on top. The gravel road has a 14-centimeter-thick loose
course of gravel and a six-centimeter top layer.
Estimation of the cost per square meter of runways on airfields-takeoff
and landing strips-is quite direct. Various combinations of runway thickness are
arbitrarily selected and then costed in both rubles and dollars. Since the physical
volume of earthwork represented in the Sborniki costs is unknown, the construc-
tion estimates contain only the cost of the concrete surface and a sand and gravel
base. Again, the dollar costs are adjusted to an average location, 1970 base,
using the same coefficients as for roads.
The estimated costs of railroad construction per kilometer of track is the
sum of expenditures for ties, ballast, and the rails themselves. Since there are
several different possible rail sizes, depending on the proposed traffic density of
the railroad, ruble-dollar ratios are derived for various sizes. Moreover, ratios
are computed for both lines with wooden ties and concrete ties.
The quantity of ballast and sand base that should be used in one kilometer
of single-track Soviet railroad is reported in Spravochnik po zheleznodorozhnomu
stroitel'stvu (Handbook for Railroad Construction). The costs of these two items
are readily obtained from the appropriate Sborniki and the Building Cost File.
The dollar amounts are adjusted to a 1970 average cost basis as elsewhere.
Direct comparisons between three types of rails are feasible. These are
90-pound, 100-pound, and 130-pound rails for the United States and R-43, R-50,
and R-65 rails for the USSR.* A ruble cost for R-43 rail with concrete ties cannot
be determined. Apparently the Soviets deem R-43 rail as too light for heavy
traffic and, therefore, not worth the additional expenditure for concrete ties.
In costing the rails, 25-meter lengths are assumed. Frequently, for any given rail
weight, a choice exists as to how the rails are connected; when this occurs, the
average of the options for that rail size is used.
The Building Cost File includes the cost of wooden ties in the cost of the rail.
This differs from the Sborniki in which the cost of the ties is estimated separately.
The Soviet source presents an option of three alternative frequencies of ties-
i.e., 1,600, 1,840, and 2,000 ties per kilometer. These three frequencies have been
matched with R-43, R-50, and R-65 rails, respectively, on the judgment that the
heavier rail requires more ties.
Costing the concrete tie railroads in rubles uses the identical procedure as
for wooden ties. Since the Building Cost File incorporates wooden ties in the
cost of the rail, their cost must be deducted and the cost of the concrete ties added.
This is accomplished by using Building Construction Cost Data, 1972, which
compares the cost of individual concrete and wood ties. The difference between
these two unit prices represents the additional cost of concrete ties. This dif-
ferential is augmented by 25%, which is that source's suggested markup for the
overhead and profit already included in the Building Cost File, and is added to
the cost for the appropriate wooden-tie railroad.
*The Soviet rail number is approximately the weight of one meter of rail as measured in
kilograms.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
The results of the transportation construction study are presented in Table
H-1 for roads, in Table H-2 for airfields, and in Table H-3 for railroads. The
sample ruble-dollar ratios for roads range between 0.748 and 0.881 and average
0.790. As expected, the ratio for concrete roads is lowest-perhaps reflecting the
wide use of concrete in construction in the USSR. Moreover, the average ratio
for earthwork is substantially higher than the ratio for pavement, which supports
the contention that Soviet excavation and other earthwork is relatively ineffi-
cient because of the low capacity and poor quality of Soviet construction
machinery.
The airfield ratios range from 0.670 to 0.729 and average 0.695. These ratios
are lower than the ratios for roads-even those for concrete roads. In part, the
ratios are lower because earthwork is excluded from the airfield construction
estimates, eliminating from consideration a kind of construction in which the
USSR is at a relative disadvantage. Finally, no relation is evident between air-
field runway thickness and the size of the ruble-dollar ratio.
The ruble-dollar ratios for railroad construction vary between 0.761 and
0.891 and average 0.836. The ratio rises as heavier rail is used. The effect of
the use of concrete ties on the ratio is uncertain, however, because the ratio is
lower for heavy-duty R-65 rail with concrete ties and higher for medium duty
R-50 rail with concrete ties. Perhaps the railroad ruble-dollar ratio is higher than
other forms of transportation construction because roads and airfields require
less sophisticated construction techniques than railroads, for which precision
and quality are essential.
Description of Pavement
US Cost per
Square Meter 1
(1970 US $)
Soviet Cost per
Square Meter 2
(1970 Rubles)
Ruble-Dollar
Ratio3
15.24 cm thick concrete ............
12.78
8.56
0.670
20.00 cm thick concrete ............
15.88
10.88
0.685
25.40 cm thick concrete ............
19.07
13.58
0.712
30.48 cm thick concrete............
22.32
16.28
0.729
38.10 cm thick concrete ............
29.29
19.93
0.680
Average ........................
19.87
13.85
0.695
1972 Building Cost File: Eastern Edition, pp. 328-329.
2 Sborniki, Investment Series, no. 24, p. 12.
3 Column (2) = column (1).
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Track Description'
Wood ties
90 lb US rail vs R-43 Soviet
rail ......................
100 lb US rail vs R-50 Soviet
rail ......................
130 lb US rail vs R-65 Soviet
rail ......................
Average ................
Concrete ties
100 lb US rail vs R-50 Soviet
rail ......................
130 lb US rail vs R-65 Soviet
rail .......................
US Cost per Kilometer of Track
Soviet Cost per Kilometer of Track
(1970 US $)
(1970 Rubles)
---
Ruble-
Stone
Sand
Rail and
Stone
Sand
Rail and
Dollar
Ballast 2
Base 3
Ties 4
Total 5
Ballasts
Base 7
Ties 8
Total 9
Ratio 10
14,431
4,623
58,280
77,334
16,698
4,418
37,760
58,876
0.761
14,431
4,623
65,360
84,414
16,698
4,418
48,942
70,058
0.830
14,431
4,623
74,367
93,421
16,698
4,418
62,100
83,216
0.891
14,431
4,623
66,002
85,056
16,698
4,418
49,601
70,717
0.827
14,431
4,623
89,765
108,819
17,549
5,015
68,562
91,126
0.837
14,431
4,623
97,576
116,630
17,549
5,015
77,750
100,314
0.860
1 The Building Cost File contains construction costs for US 90-, 100-, and 130-pound rail. The numbers refer to the weight of one
yard of rail. These weights converted to a metric basis are 44.65, 49.61, and 64.49 kilograms per meter, respectively. The Soviet.
rails closest to these magnitudes are R-43, R-50, and R-65 which weigh 44.65, 51.64, and 64.64 kilograms per meter, respectively.
Sources: Spravochnik po zheleznodbrozhnomu stroitel'stvu (Handbook for Railroad Construction), p. 154, and Zheleznyye dorogi
obshchiy kurs (Railroads: A General Course), p. 46.
2 The quantity of crushed stone ballast used for a kilometer of track is 1,518 cubic meters. Source: Spravochnik, op. cit., p. 159,
and Cost File, op. cit., p. 339.
3 The quantity of sand used as base for the ballast prism per kilometer is 796 cubic meters. Sources: Spravochnik, op. cit., p. 159,
and Cost File, op. cit., p. 339.
4 Ibid.
5 Column (2) + column (3) + column (4).
6 Sborniki, Investment Series, no. 21, p. 175. Also see note 2.
7 Ibid. Also see note 3.
8 Ibid., pp. 176-182, and Building Construction Cost Data, 1972, Robert Snow Mearis Company, Inc., pp. 29, 185.
9 Column (6) + column (7) + column (8).
10 Column (9) = column (5).
11 This is the unweighted average of each specific track description, i.e., the mean of IA, 113, IC, IIA, and 1113.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
APPENDIX I
WEIGHTED CONSTRUCTION RUBLE-DOLLAR RATIOS
Weighted
Ruble-Dollar Ruble-Dollar
Weights Ratio Ratio
Hospitals 1 ....................................... 0.048 0.568 0.027
Housing ......................................... 0.471 0.620 0.292
Commercial 2 .................................... 0.140 0.508 0.071
Schools .......................................... 0.093 0.583 0.054
Industrial ....................................... 0.101 0.601 0.061
Highways and streets ............................. 0.143 0.790 0.113
Railroad ........................................ 0.004 0.836 0.003
US-weighted ratio ................................ .... .... 0.621
1 The geometric mean of the US-weighted housing, office building, and school ratio is used
instead of the computed but discredited hospital ratio.
2 The office building ratio is used for the commercial category.
Weighted
Ruble-Dollar Dollar-Ruble Dollar-Ruble
Weights Ratio Ratio Ratio
Industry ........................... 0.362 0.601 1.664 0.602
Transportation and communications
except railroads ................... 0.060 0.741 1.350 0.081
Railroad ........................... 0.026 0.836 1.196 0.031
Housing .......................... 0.270 0.494 2.024 0.546
Trade and communal enterprises, for-
estry enterprises, and institutions of
science, culture, art, education, and
health2 .......................... 0.282 0.475 2.105 0.594
Soviet-weighted ratio ................ .... .... .... 1.854
1 The geometric mean of road and airfield ratios is used for this class of construction.
2 Presumably forestry construction is a negligible portion of this category. The geometric mean of
the office building and school ratios is arbitrarily assigned to this group.
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
 Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2
Weighted
Ruble-Dollar Ruble-Dollar
Weights Ratio Ratio
Hospitals ........................................
Housing .........................................
Commercial ..................................... .
Schools ..........................................
Industrial ......................................
Highways and streets .............................
Railroad ............ ............................
Adjusted US-weighted ratio ........................
0.048 0.682 0.033
0.471 0.744 0.350
0.140 0.610 0.085
0.093 0.700 0.065
0.101 0.721 0.073
0.143 0.948 0.136
0.004 1.003 0.004
???? .... 0.746
Category Weights
Weighted
Ruble-Dollar Dollar-Ruble Dollar-Ruble
Ratio Ratio Ratio
Industry ........................... 0.362
Transportation and communications
except railroad .................... 0.060
Railroad ........................... 0.026
Housing ............................ 0.270
Trade and communal enterprises, for-
estry enterprises, and institutions of
science, culture, art, education, and
health ............................ 0.282
Adjusted Soviet-weighted ratio........ ....
0.889 1.125 0.068
1.003 0.997 0.026
0.593 1.686 0.455
0.570 1.754 0.495
???? .... 1.546
Declassified and Approved For Release 2012/03/15: CIA-RDP08SO135OR000602050003-2