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

Approved for Release: 2019/07/19 C05210421
a permanent price. Because innovation is risky given
the incentives confronting the Soviet manager, specifi-
cations of old products often change only slightly.
This is enough, however, to allow him to charge a
substantially -higher price for this “new” product and
to lead him to halt production of the older and
cheaper model. Purchasers have little choice but to
accept the more expensive item. As a consequence,
value of production has increased more than could be
justified by the degree of product change.
An inflationary bias is also introduced into the value
series by the way constant or, in Soviet terminology,
comparable prices (sopostavimye tseni) are set. For
items in production before the base year for the
comparable prices, the prices in effect in the base year
are used in the series. For products introduced subse-
quently, the first permanent price is the comparable
price. The output series therefore reflects a mixture of
constant and current prices; the presence of price
inflation will then bias the output index upward.
New-product pricing represents the serious measure-
ment problem in industries where products are hetero-
geneous, where value measures would provide the only
meaningful measure of output, and where many op-
portunities for innovation exist and product complex-
ity makes it difflcult to distinguish between real and
cosmetic improvements. Although these features exist
to some degree in all industrial branches, they are
most characteristic of machinery and chemicals.
The study already referred to (An Analysis of the
Behavior of Soviet Machinery Prices, 1960-73), can
be used to gain some notion of the possible impact of
disguised inflation on the value series. One type of
machinery analyzed in that study is machine tools, for
which SPIOER relies on the offlcial value series in
the Narkhoz. According to the value series, machine
tool output in 1970 was 2.33 times the 1960 level—an
average annual increase of 8.8 percent. The machin-
ery price study concludes, however, that machine tool
prices in 1970 were 28.7 percent above the 1960 level
after allowance for changes in product characteristics.
Deflating the value series by this price increase
suggests that output in 1970 was only 181 percent
above 1960 or that machine tool output increased at
an average rate of only 6.1 percent per year, 2.7
percentage points per year less than the value series
used by SPIOER. If the wholesale price index derived
in that paper based on a sample of several types of
machinery is representative of all of the value series
incorporated in SPIOER, the hidden price changes
incorporated in the value series and hence wrongly
counted as real growth, would be significant:
Implied Price Inflation Average Annual Percent
in Machinery
1961-66 2.7
1967-70 5.7
1971-73 — 1.6
This would suggest that in the 1960s the value series
are upward biased and that at least in the early 1970s
the value series are downward biased. Over the entire
13-year period, the implicit rate of inflation in the
value series would average 2.6 percent per year.
Not all value series are subject to a large upward bias
arising from price inflation. In some of the rapidly
developing industries like precision instruments and
electronics, prices have been reduced repeatedly to
prevent the excessive accumulation of profits. Because
those industries are in a relatively early phase of
development, expansion of output generates substan-
tial economies of scale and cost savings. In such cases
the amount of disguised inflation would be much
smaller than in more developed branches that are also
represented by value series.
The relative importance of this kind of sample bias in
the branch indexes can be determined from the
composition of the sample in each branch. Six of the
branch indexes are completely based upon quantity
series. Distortions in the value series reported by the
CSA could carry over to the synthetic indexes for the
wood, pulp, and paper industry, light industry, chemi-
cals, and machinery. The tabulation below indicates
214
Approved for Release: 2019/07/19 C05210421