USSR: Measures of Economic Growth and Development, 1950-80

Approved for Release: 2019/07/19 C05210421
The most common description is a statistical produc-
tion series expressed in physical terms: tons, square
meters, units of production, or some other convention-
al measure. Examples of this type of data are tons of
coal, square meters of tile, or number of subway cars.
The product series are then multiplied by actual or
estimated 1967 enterprise wholesale prices to derive
constant price series. In some cases the raw product
data are adjusted by indexes of quality change in the
average or standard product over time. For example,
not only has Portland cement production increased
over time, but the output mix has steadily shifted in
favor of higher “marks” (greater strength) of cement.
Raw figures for Portland cement production are
therefore weighted by both the 1967 wholesale price
and an index of average mark to account for the
improved quality of the product over time.
Official ruble value series reported by the Central
Statistical Administration of the USSR (CSA) are
also used as sample elements. A few examples are the
value of production of agricultural machinery, metal-
cutting machine tools, and forge presses. Until recent-
ly these series were reported in official 1967 wholesale
prices, although earlier years were reported in 1955
wholesale prices. In 1976 the Soviets began to publish
these value series in prices of 1 January 1975. The
official statistics have overlap in the different prices
for a given year to permit an approximate linkage
between the earlier and later periods, so that a
common price basis can be maintained. Since the
physical output samples are weighted by 1 July 1967
prices, the new value indexes were linked backward to
estimate them in 1967 prices to conform with the
remainder of the sample.
These rubles series, however, must be used with
caution because many of them contain a large degree
of concealed inflation. The products in this category
are usually machinery items subject to the problems
of new-product pricing. Some share of the reported
growth arises from the use of first temporary prices
and then permanent prices that are substantially
higher than increased performance would justify.’ In
addition, since these series include both intermediate
and final goods, they contain a substantial amount of
double-counting. To the extent that the magnitude of
double-counting has fluctuated over time, these series
could be biased up or down.
The third kind of production measure consists of gross
value of output (GVO) indexes prepared by the
Soviets for various product groups such as mineral
chemicals, repair of machinery, and metal structurals.
They are supposed to represent the aggregation of all
output in a given branch—sometimes including work
in process and major repairs. These indexes are
subject to even greater limitations than the CSA
value series.‘ Therefore, they are used only to fill out
the sample in some crucial areas where better indica-
tors are lacking.
Although each line item in the SPIOER sample
ostensibly consists of a production entry for every year
since 1950, the quality of coverage is not uniform.
Some gaps in coverage exist for the earlier years,
especially during the 1950s. These gaps fit three
categories: missing intervening observations at irregu-
lar intervals, series that begin several years after 1950
despite production during earlier years, and items not
produced in earlier years. The extent of the resulting
problems and procedures to circumvent them vary by
type of gap.
’
See the section “Biases in the Basic Data” for a fuller treatment of
the new-product pricing phenomenon. It has been widely discussed
in the literature. In particular, see Abraham S. Becker, Ruble Price
Levels and Dollar-Ruble Ratios of Soviet Machinery in the I 960s,
Report R-1063-DDRE (Santa Monica, California, The Rand Cor-
poration, 1973); Joseph S. Berliner, The Innovation Decision in
Soviet Industry (Cambridge, The MIT Press, 1976); Padma Desai,
“On Reconstructing Price, Output and Value-Added Indexes in
Postwar Soviet Industry and Its Branches,” Oxford Bulletin of
Economics and Statistics (February 1978): pp. 55-77; Central
Intelligence Agency, An Analysis of the Behavior afSoviet Ma-
chinery Prices, 1960-73, ER 79-10631 (December 1979); and James
E. Steiner, Inflation in Soviet Industry and Machine Building and
Metalworking (MBMW), I960-I975, SRM 78-1042 (Working pa-
pezr, July 1978).
'
As an index, the output data are expressed only as a proportion of
another year. The publication of indexes with Qitferent base years
and in extremely rounded form limits accuracy. In addition, an
index cannot be scaled to compare its relative importance with
activity elsewhere in the economy.
181
93-89? O - 82 — 13
Approved for Release: 2019/07/19 C05210421