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

Approved for Release: 2019/07/19 C05210421
materials permits a smaller value of materials to be
used per unit of construction, then the ratio of
material inputs to gross output would decline. On the
other hand, if the use of improved materials permits
the substitution of materials for labor and capital,
then the ratio of material inputs to gross output would
rise. The Soviet Union has published values for the
gross output and net material product (NMP) of the
construction sector for 1964-78 in current prices. The
ratio of material inputs (gross output less NMP) to
gross output declined slowly in this period (from 57
percent in 1964 to 54 percent in 1978), suggesting that
the material-input approach is reasonably valid. This
conclusion depends on the relative price stability
between gross output and material inputs. The avail-
able evidence does not suggest a divergence in the
prices of construction output and its inputs. In addi-
tion, US data show a stable ratio of material inputs to
gross output in constant prices, ranging between 52
percent and 55 percent, since 1950. International
comparisons are hazardous, but this stability also
lends support to the use of a material-input index.
Weights
The weights used to combine the purchase indexes are
derived from the 1972 Soviet input-output table. The
construction column in that table enumerates the
ruble value of purchases from each of 75 input-output
sectors. The 1972 input-output table is available with
88 sectors. In order to make comparisons with the
1959 and 1966 input-output tables, however, tables in
comparable 75-sector formats were used. The con-
struction sector did not purchase any materials from
l l of these sectors in 1972. Production indexes are not
available for an additional 10 sectors. The purchases
by the construction sector from the remaining 54
sectors were summed and converted into percentage
weights (table B-1). The construction sector purchased
over half of its inputs from just two sectors—construc-
tion materials and transportation and communica-
tions. A desirable development would be to disaggre-
gate the construction materials sector. The data
needed to compute the purchase indexes for the
separate construction materials sectors for other
years, however, are not available.
Description of the Index
The construction index is a weighted average of 54
purchase indexes, each showing the purchases by
construction from the input-output sectors listed in
table B-1. Each purchase index is computed in a two-
step procedure. First, the shares of the gross output of
each of the 54 sectors sold to construction were
computed from the 1959, 1966, and 1972 Soviet
input-output tables. It is assumed that the rate of
change in these shares between 1959 and 1966 and
between 1966 and 1972 was linear. It is assumed also
that there was no change in the shares before 1959 or
after 1972. The resulting time series shows the share
of each sector’s gross output sold to construction.
Second, the shares so derived were multiplied by the
production index for that sector to arrive at an index
of purchases by construction from each sector. Each
purchase index was rebased so that 1972 = 100. A
sample computation of a purchase index is shown in
table B-2.
To form the construction index, the 54 purchase
indexes are multiplied by their corresponding weights
from table B-1, and the products are added. The
resulting index is shown in column 2 of table B-3.
All but four of the 54 production indexes are for
industrial sectors and were obtained from the index of
industrial production described by Ray Converse in
J EC, Industry. The indexes for the crops and animal
husbandry sectors were derived from the data con-
tained in the index of agricultural output as described
by Margaret Hughes and Barbara Severin in JEC,
Agriculture. The index for the transportation and
communications sector is the index for freight trans-
portation derived below in this appendix. The index
for the trade sector is the index for wholesale trade
derived below in this appendix.
Although the material-input method is theoretically
less desirable than either the gross output or double
deflation method, it appears to produce reasonable
results and to be the best available method given the
data constraints. The material-input index can be
84
Approved for Release: 2019/07/19 C05210421