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

Approved for Release: 2019/07/19 C05210421
Description and Evaluation
of the Consumption Index
The Index in Established Prices
General
The index of total consumptionwshown in summary
form in Table 3 and in detail in Appendix A—is a
weighted aggregate of separately constructed indexes
for three broad categories of goods—food, soft goods,
and durables and miscellaneous goods—six categories
of personal services—housing, utilities, transporta-
tion, communications, repair and personal care, and
recreation—and two communal services education
and health. The indexes for major categories, in turn,
are weighted aggregates of varying numbers of com-
ponent indexes.
The aggregate index is a sample index based to a
considerable extent on time series in physical units.
The sample is large and in one way or another »
represents more than 90 percent of all consumption
expenditures. 'All important foods, soft goods, and
durables are included, as are most personal services
and all education and health services. Both the sample
categories and the time series selected to represent
them were severely constrained by the availability of
published Soviet data. Ideally, one would like to have
suitable and detailed expenditure time series and
corresponding reliable price deflators. In general, the
Soviet government publishes neither. Data on retail
sales, regularly published in current prices with a
disaggregation of several dozen groups, are flawed by
the presence of sizeable residual groups of obscure
content and by inclusion of sales to enterprises and
sales of intermediate and investment goods that do not
reflect household consumption. Retail sales data also
include sales of so-called “productive” services, such
as shoe repair. The data with which to remove or
reclassify these various items are not available over
time. Official retail price indexes are seriously defec-
tive, as already noted, since they reflect official price
changes only on a fixed sample of basic products; the
index is, in essence, an index of prices on an official
price list and not an index of prices paid by consumers
for goods actually available for purchase. The Soviet
government publishes neither quantities nor values for
consumption in kind, a major element in total food
consumption. Neither expenditures nor price indexes
are available for most personal services. Official data
on government outlays for education and health are in
current prices, and include transfer payments and
some expenditures that belong to investment rather
than to consumption.
In general, the Soviet Union publishes only a small
fraction of the data that are required, ideally, to
construct a reliable consumption index; moreover the
definitions and methodologies underlying the pub-
lished data are seldom available in more than cursory
detail. As a result of these problems with the data, the
quality of the components of the CIA indexes is
uneven. Most satisfactory, perhaps, are those for food,
soft goods, transportation, and communications: least
satisfactory are those for durables and miscellaneous
goods and for recreation. The indexes for housing and
for communal services suffer from an inability to
reflect improvements in quality; this shortcoming is
not unique to the indexes constructed here for the
USSR, however. Each of the major component index-
es is evaluated below with respect to coverage and
quality of the time series data and internal weights
within major components.
Several sources of possible bias exist. The choice of
1970 as the base year may -impart some bias to the
index. In general, however, 1970 was a fairly
“normal” year for consumers, in the sense that no
major crop failure or other unusual event occurred.
The matter of the degree of overall equilibrium in
consumer goods markets is considered below. To test
the effect of the choice of base year, alternative
weights for major categories on a comparable basis
were constructed for 1960 and 1976. Rates of growth
shown by the three sets of weights do not differ much.
The use of 1960 weights speeds up the index some-
what, while the use of l976 weights instead of 1970
weights has a negligible impact on the index. Table 4
summarizes the growth rates given by alternative sets
of weights for major categories. i
The use of time series expressed in physical units may
bias the index downward because of the inability to
allow fully for improvements in quality. This source of
bias could be significant, probably, only for footwear,
hosiery, and knitwear and perhaps for education and
health services. On the other hand, those series that
328
Approved for Release: 2019/07/19 C05210421