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

Approved for Release: 2019/07/19 C05210421
If this expression is deflated by a Paasche (current-
year weighted) price index, then the desired Laspeyres
output index is obtained:
N
2 Q(t,i)p(t,i)
i=1
N N
2 Q(0,i)p(0,i) 2 Q(t,i)p(O,i)
i=1 _ i=1
N T N
2 Q(t,i)p(t,i) 2 Q(0,i)p(0,i)
1=l i=l
N
E Q(t,i)p(0,i)
i= 1
Rarely is information available on production and
prices for all goods and services in both the base and
the given year. If only a sample of observations is
available, then the two sides of the previous equation
may not be the same. For example, suppose there are
N products, but price and quantity data are available
only for n (less than N) products. The estimated
growth of output from year O to year t using the
deflated current-price index would be:
N N n
Z Q(t.i)p(0,i) 2 Q(t,i)p(t,i) Z Q(t,i)p(0,i)
i=1 _ i=1 i=1
N T N X n
Z‘ Q(0,i)p(O,i) Z Q(0,i)p(0,i) Z Q(t,i)p(t,i)
i=1 i=l i=l
and the estimate using the quantity index would be:
_.,,Z
-Q
t,i)p(0,i) Z Q(t,i)p(0,i)
N = n
21 Q(0,i)p(0,i) 2 Q(0,i)p(O,i)
i= l i= 1
If n=N, then the two equations are identical. lf n is
less than N, the reliability of the deflated current-
price index depends on the accuracy of the price index
43
computed from the sample data, and the reliability of
the quantity index depends on the representativeness
of the quantity sample. Hill argues that the price
index is likely to be more accurate, especially when
large numbers of new products are being introduced
and old products withdrawn.” Unfortunately, Soviet
price data are so scarce and unreliable that the
deflation method is not practical either for GNP as a
wlhole or for most individual sector indexes. The
quantity index method is relied upon almost
exclusively.
Specific Index Number Problems
Several specific index number problems are also
important in the measurement of Soviet GNP. They
arise in the treatment of regional production, seasonal
production, price discrimination. quality changes, and
unique products.
Regional Production. The resources required to pro-
duce a given product may be quite different depend-
ing on where a product is made. For example, a school
building of exactly the same dimensions may cost
much more to construct in Siberia than in the
Ukraine because of greater transportation expenses,
more stringent structural requirements, a larger heat-
ing unit, and more difficult construction conditions.
To the extent that the added cost reflects real differ-
ences in the use of resources, a shift in production to a
more expensive region represents an increase in pro-
duction rather than a price increase. Theoretically, we
should construct weighted regional production index-
es, but the scarcity of data does not permit doing so.
The influence of regional production on the growth of
GNP is probably most important in the construction,
agriculture, and fuel sectors.
Seasonal Production. Changes in seasonal production
patterns present a similar problem. A vegetable grown
near Moscow in a hothouse during the winter requires
more resources than the same vegetable grown out-
doors during the summer. Theoretically, the two
vegetables should be treated as different products,
and a production or price index containing them
” Hill, Measurement, p. 25.
Approved for Release: 2019/07/19 C05210421