THE IMPACT OF GORBACHEV'S POLICIES ON SOVIET ECONOMIC STATISTICS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP89T01451R000400480001-9
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 93
• Document Creation Date: December 27, 2016
• Document Release Date: December 19, 2012
• Sequence Number: 1
• Publication Date: July 1, 1988
• Content Type: REPORT

STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
/a~"-~"~t~\ LIIGCIVIii IG Vl ?
Intelligence `"
The Impact of Gorbachev's
Policies on Soviet
Economic Statistics
00~~59'~ -^---
==~ :~~~
.~ 2 ~ ~~-,I ~.~
Gt'...:/IFC/~Cd,_h3~1G~ f4kirt~f4t~ {,t},_~.f~-.~.'~.:5+!OR,.
( ~w5 ft~~ ~t.d~~~` ~~~it;I~R~lw:,5~v~"~3
f
IWMJK ?~
PROl7DCT N[)MBEFt _ .`~ /~- ~ ~ o ~--' g ~ -
s.
PAGE NC>~S ~ t d ~~
TOTAL NIIrKRER OF aOPIES ~ ? b~
DISSEM DATE ~`j ~
E~~2A OOPIFS
RE7UC)RD CII~EFt Q
JOB NCIIMBER
SOV 88-! 0049
July 1988
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
The Impact of Gorbachev's
Policies on Soviet
Economic Statistics
SOV 88-10049
July 1988
i
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
This report includes a summary of a conference on the impact of General
Secretary Mikhail Gorbachev's policies on Soviet economic statistics
sponsored by CIA's Office of Soviet Analysis and the texts of the papers
presented at the conference. The papers have been lightly edited to
standardize their formats and eliminate instances of overlapping coverage.
The views expressed by the non-CIA participants reflect the speakers' own
judgments and do not necessarily reflect the views of the CIA or those of
other US Government agencies.
SOV 88-10049
July 1988
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Overview 1
Conference Papers
New Challenges to Soviet Official Statistics: 11
A Methodological Survey
STAT
Changes in the Availability of Economic 27
Data Under Gorbachev
STAT
Perestroyka and Soviet Statistics 51
STAT
The New Look at Soviet Statistics: Implications for CIA Measures 69
of the USSR's Economic Growth
STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
The Impact of Gorbachev's
Policies on Soviet
Economic Statistics
Overview On 11 December 1987, the CIA's Office of Soviet Analysis hosted an
unclassified conference in Langley, Virginia on the impact of General
Secretary Mikhail Gorbachev's policies on Soviet economic statistics. The
participants in the conference-CIA analysts and distinguished experts on
the Soviet economy from US universities and private research institutes
(see list of participants on riext pagerfocused on four questions:
? Have the revelations occasioned. by Gorbachev's glasnost policy increased
our understanding of the nature or the severity of the failings of official
Soviet economic statistics, particularly the degree to which these statis-
tics overstate real economic growth?
? Has glasnost substantially increased the availability of the types of
statistics required for better assessments of Soviet economic
performance?
? Have glasnost and the anticorruption and discipline campaigns improved
the quality of recent Soviet economic statistics or corrected past falsifica-
tions or distortions of economic data?
? What are the implications of glasnost for Western-especially CIA-
estimates of Soviet economic growth and trends in resource allocation?
Glasnost and the Failings of Official Statistics
began the conference STAT
wit a paper on the criticisms that Soviet economists have recently made of
the statistics on economic growth published by their central statistical
authorities. In particular, concentrated on the criti- STAT
cisms made by Vasiliy Selyunin and Grigoriy Khanin in an article that has
aroused great interest among Western Sovietologists because of its muck-
raking tone and its assertion that Soviet economic growth has been much
lower than claimed by the authorities.'
concurred with Selyunin and Khanin's contention that the STAT
overpricing of new products-a technique employed by enterprise manag-
ers and economic ministers to ease the task of plan fulfillment-has
historically resulted in the inflation of official claims of real growth.
Western students of the Soviet economy, ~ noted, long have STAT
identified such hidden inflation as a major flaw of official Soviet data, and
other Soviet economists have expressed similar skepticism about the
reliability of the official growth statistics.
STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
however, also criticized some aspects of the Selyunin-Khanin
study. He noted, for example, that when using the methods and sources de-
scribed in earlier articles by Khanin, he obtained results that, for the most
part, were about the same or higher than the officially reported industrial
growth figures. (Khanin was unable to publish specific estimates of
industrial growth, but indicated that application of his methods yielded
much lower estimates of growth than those of the Soviet statistical
authorities.) While acknowledging that his inability to reproduce the
estimates in Selyunin and Khanin's study need not imply that they were in-
correct,Omaintained that "reproducibility" is a reasonable
criterion by which to judge their reliability. In addition, he noted that the
methods described by Khanin generally rested on assumptions that would
be difficult to defend-for example, that the ratio of industrial growth to
electricity consumption in the Soviet Union would be the same as in the
United States. Finally, he argued that the methods of Selyunin and Khanin
and other Soviet critics of official statistics, which often use physical
measures such as tons and kilowatt hours in place of official ruble value se-
ries, completely ignore qualitative change and that this is a mistaken
approach. As a result, while he is broadly in agreement with Selyunin and
Khanin,Oconcluded that their estimates, rather than being
accepted as accurate measures of Soviet growth, should be viewed as a
STAT
STAT
STAT
STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
lower bound. He believes real growth probably lies somewhere between
them and the official figures released by the Soviet statistical authorities.
In the group discussion that followed) conference STAT
participants applauded analysis of the Selyunin-Khanin study, but STAT
there was some disagreement with conclusions. STATT
disputed the suggestion that the Selyunin-Khanin esti- STAT
mates should be accepted as lower bounds for Soviet economic growth,
arguing that, until their underlying methods and evidence had been
clarified and evaluated, the estimates could not be regarded as more than
offered a different opinion, STAT
inability to reproduce Selyunin and Khanin's STAT
problem than theirs, and that until STAT
more is known about their underlying evidence it would be inappropriate to
dismiss them. Implicit in point was the belief that the STAT
Selyunin-Khanin estimates, as statements of apro-Gorbachev position,
enjoyed at least a semiofficial status that makes them deserving of
consideration as alternatives to the official figures.
also took issue with ~~ point that Selyunin and STAT
Khanin failed to take improvements in product qualit into account.
Specifically, argued that, although criticism was STAT
valid for the 1960s and most of the 1970s, even official Soviet data
indicated that the degree of product improvement was negligible in the late
1970s and early 1980s. opined that the discussion of whether STAT
product quality was improving or deteriorating might be beside the point.
Specifically, he argued that, regardless of whether a new product is better
or worse than its predecessor, it is proper to give the new product a higher
price if its manufacture requires more inputs than the product it replaces.
Viewed in this light,0 argued, new product pricing may be less a STAT
source of inflation than many have argued.
Glasnost and the Availability of Economic Statistics
the Office of Soviet Analysis began the second STAT
session of the conference with a paper on the impact of Gorbachev's
glasnost policy on the availability of economic statistics. After reviewing
the history of Soviet policy on the disclosure of economic data since the
1950s, she described the changes in the quantity of economic statistics
included in the economic yearbooks published since Gorbachev came to
power and, more briefly, discussed the changes that have occurred in other
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Soviet statistical compendia. Her conclusion was that, on balance, Gorba-
chev's policies have had a positive impact on the availability of Soviet
statistics, as the Soviets have resumed publication of data series previously
withdrawn from their economic yearbooks and provided some types of
information never published previously. The Soviets' announcement of
plans to publish several major statistical compendia on economic sectors of
critical importance to Gorbachev's program and their unprecedented
efforts to market their data in the West, in her view, also bode well for the
future availability of data of interest to Western Sovietologists.
At the same time,~~cautioned that significant gaps in Soviet STAT
reporting remain-for example, in the area of defense expenditures, in
which the Soviets have pledged to publish comprehensive spending figures
but only in a few years following the completion of a major price reform. In
addition, she noted that Gorbachev is not the first general secretary to
increase the availability of economic data in the early years of his tenure.
During the Khrushchev and Brezhnev periods, the more "liberal" informa-
tion disclosure policies of the general secretaries' first years in power gave
way to greater secretiveness as their economic programs faltered.
In the group discussion that followed ~~ presentation, most STAT
participants in the conference echoed her modest optimism about the
prospects for Soviet publication of additional types of data. Some noted,
however, that troubling gaps and discontinuities in the statistics on
monthly, quarterly, and annual economic performance were occurring that
hampered the construction of production indexes based on physical mea-
sures such as Khanin and the CIA use. also STAT
argued that it may be a mistake to regard newly released data series as the
"tip of an iceberg" because the Soviets themselves may lack much of the
data that Western economists fault them for withholding.
Changes in the Quality of Economic Statistics
opened the third session of the conference with a paper ad-
dressing whether Gorbachev's policies are resulting in the release of
more-or less-reliable statistics. This question is especially important to
an assessment of Soviet economic performance under Gorbachev. If, for
example, the crackdown on the overreporting of output and the overpricing
of new products are reducing the degree of inflation in official growth
statistics, then, in comparison with earlier years, the economy's perfor-
mance would be better than the official data suggest. If, however, the
STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
. crackdown on overreporting and overpricing is offset by the crackdown on
underperforming-that is, if enterprise managers "cheat" more to realize
their demanding plan targets-then the growth statistics released under
Gorbachev would exaggerate the success of his economic policies.
presentation placed heavy emphasis on the political roots of
glasnost, appropriately so in view of Gorbachev's use of this policy to
discredit his predecessors and opponents and cast himself and his support-
ers in a more favorable light. While not dismissing the Selyunin-Khanin es-
timates of past economic growth or Abel Aganbegyan's statements that
growth in the late Brezhnev years was lower than officially claimed,
described these estimates as political actions designed as much to stimulate
public support for Gorbachev's programs as to set the historical record
straight.
STAT
STAT
acknowledged that there has been at least one major STAT
instance-the correction of historical statistics on cotton procurement in
the Soviet economic yearbook for 1986-in which glasnost has resulted in
"setting the record straight" in the area of economic statistics. Although
the correction was belated and, in some respects, inevitable given Soviet
press reports of gross falsification of cotton production and procurement
statistics in Uzbekistan during the Brezhnev period,~-egarded it as a STAT
welcome move. He also suggested that a change in reported values in the
national income section of the economic yearbook for 1986 (which is
related to the acknowledgement of subsidies) represents a potentially
important correction of a methodological nature.
On the other hand, noted that some Western analysts have seen a STAT
"dark side" to glasnost. This is illustrated most clearly by the official
Soviet data on retail sales and national income for 1985 and 1986. In both
years legal sales of alcohol were reduced substantially as a result of
Gorbachev's antialcohol campaign. Because alcohol sales constitute a large
share of retail trade, these reductions should have resulted in a drop in con-
sumption and national income, but, in the official figures, an increase was
reported. expressed the belief that the Soviets did not simply STAT
"doctor" t e books and that methodological errors probably were at fault.
Still, he regarded the Soviets' failure to explain what had been done as
grossly inconsistent with their professed commitment to glasnost.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
More important,~~assessment was that "evidence of improved STAT
methodology or fundamentally revised time series" has been obviously and
unfortunately absent from the economic yearbooks and statistical press
releases issued under Gorbachev. Thus, while qualifying his judgment by
observing that it may simply be too early to expect the Gorbachev regime
to have made fundamental improvements in Soviet economic statistics,
overall conclusion was that the General Secretary's policies have STAT
not substantially increased the reliability of Moscow's official economic
statistics.
In the group discussion that ensued, STAT
who was in the audience, noted that there had been additional STAT
moves to correct the historical record in the area of demographic statistics.
He also noted, as hadOthat part of the credit for this development STAT
should go to Western Sovietologists whose analyses had embarrassed the
Soviets into making changes. Other conference participants seconded
point that clarification of the underlying methodologies is essential STAT
if official Soviet statistics are to establish their claim to greater credibility.
Despite Soviet press reports of arrests and prosecutions for falsifying
statistics, conference participants were skeptical that fear of punishment
for such offenses would outweigh the Soviet manager's fear of the penalties
associated with failure to fulfill production plans. As a result, they
anticipated continued overreporting and overpricing, with some arguing
that these sources of inflation would grow even stronger under Gorbachev
because of the demanding goals he has set.
Implications for Western Estimates of Soviet Economic Performance
In the last of the papers presented at the conference, f the STAT
Office of Soviet Analysis assessed the implications of what is being learned
from glasnost for Western, especially CIA, estimates of the size and
-growth of the Soviet economy and the allocation of Soviet resources.~~ STAT
explained that, in general, the revelations occasioned by glasnost have
reinforced CIA's long-held views about the deficiencies of official Soviet
economic statistics. Since the 1950s, for example, CIA has recognized that
official statistics on aggregate industrial growth were inflated by the
overpricing of new products. It has always constructed its own index of
Soviet industrial production, based, for the most part, on disaggregated
Soviet data on production measured in terms of physical units rather than
rubles. In addition, CIA recognizes that the relative prices of Soviet
products, which are set by the state rather than being determined by
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Comparison of Soviet Statistics
on and CIA Estimates of
Soviet Economic Growth
National Industrial Machine-Building
Income Production Output
Produced
CIA Measures of Soviet Economic Performance b
a Data were obtained from Narodnoye khozyaystvo (NarkhozJ
SSSR, various issues; they are expressed in so-called "comparable"
prices.
b 1982 prices at factor cost.
market forces, often fail to reflect the different amounts of inputs used in
producing them. Therefore, CIA uses the adjusted factor cost standard
developed by to value Soviet output in terms that better re- STAT
flect its resource costs. In applying these methods, CIA historically has
produced estimates of Soviet economic growth substantially lower than
those reported by the Soviet authorities (see table above).
Nonetheless, as~acknowledged, in constructing its index of Soviet STAT
industrial production, CIA has been unable to completely dispense with
officially published ruble measures of the growth of output or the
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
allocation of resources to end uses. In the case of the machinery sector of
industry, for example, about 40 percent of the sample used in CIA's index
of production consists of ruble data, the growth of which is probably
overstated as a result of the overpricing of new products. Similarly, CIA's
index of the growth in the machinery and equipment portion of new fixed
investment is based on official ruble value measures, which, although
generally believed to be less inflated than the data on aggregate machinery
production, probably do overstate the growth of this end use.
acknowledged that the inclusion of these ruble measures in CIA's STAT
sector-of-origin and end-use indexes would, by itself, result in some
overstatement of growth in the sectors in question. He noted, however, that
the impact of such overstatement decreases as one proceeds to higher and
higher aggregations of CIA's estimates. In the case of the ruble data used
in the machinery sample, for example, the upwardly biased data account
for almost 40 percent of estimated machinery output but only about 10 per-
cent of estimated industrial output and only about 3 percent of estimated
GNP. Moreover, Omaintained, this upward bias is probably at least STAT
partially offset by downward bias in other parts of the sample.0 STAT
argued, for example, that the portions of the machinery index constructed
on the basis of physical measures probably understate growth because they
fail to reflect qualitative improvements in output. Similarly, referring to a
recent Ph.D. dissertation done at the Massachusetts Institute of Technol-
ogy by Professor Mark Prell, he stated that CIA estimates of growth in the
Soviet service sector probably were on the low side.
(then presented the results of what he described as "a primitive STAT
round of sensitivity testing" which used available alternative estimates of
the growth of those components of GNP that have been criticized for being
biased up or down in place of current CIA estimates for the entire period,
1951-86. On the sector-of-origin side, the alternative estimates of the
growth of GNP's components had a negligible impact on the growth of
overall GNP. On the end-use side, however, the impact of using alternative
indexes of the growth of key components was substantial by the period's
end. In particular, the estimated share of consumer goods in GNP
decreased by about 4 percentage points; the share of consumer services
increased by about 6 percentage points; and the share of investment in
GNP decreased by about 8 points. The estimated share of defense in GNP,
however, was about the same as the current CIA estimate.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
STAT
In closing,~~predicted that the assessment of Soviet economic growth
is likely to become increasingly difficult in the next few years as a result of
the often contradictory effects that Gorbachev's policies are likely to have
on Soviet economic statistics. On the one hand, the crackdown on
overreporting and the leadership's demands for greater methodological
rigor on the part of the statistical authorities may lead to a reduction in the
inflationary component of the official statistics. The simple pressure of
having to publish more statistics may have a similar impact, if additional
data actually are published and the statistical authorities are thereby
forced to assure that the statistics disseminated are mutually consistent.
On the other hand, the heightened pressure on Soviet managers to meet de-
manding plan targets and the decentralization of pricing authority may
increase the managers' incentive and opportunity to overstate real growth.
The encouragement of private economic activity may also prove to be a
new source of inflation in the official statistics if it leads to the surfacing of
activities that once went totally unreported because they were illegal or
frowned upon. STAT
Group discussion of ~~presentation centered much more on his
predictions of future difficulties in assessing Soviet economic performance
than on his analysis of the likely sources of bias in CIA's estimates, which
was generally accepted as accurate. Most of the participants who voiced
opinions predicted that pressure to meet plan targets would offset the
crackdown on overreporting and that, as a result, inflation of the official
statistics would not only remain a problem but might even increase. Several
participants concluded that the need for independent estimates of Soviet
economic growth would increase correspondingly.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Conference Papers
New Challenges to
Soviet Official Statistics:
A Methodological Survey
STAT
There are certain aspects of Soviet economic life that
are not known to the Soviets themselves. Inflation is
the most prominent. Denounced as a capitalist phe-
nomenon, it has simply not been measured for the
Soviet economy. Consequently, real Soviet economic
growth remains a mystery. Fortunately, however,
Gorbachev's policy of glasnost is removing a long-
standing taboo from the subject. Some Soviet authors
have already joined Western analysts in the quest for,
as they put it, "the true figure." (The writings of some
Soviet authors may even seem too radical to some of
their Western colleagues.) Sovietology will definitely
gain from the new alliance, and the field will become
more exciting.
Novyy mir, a Soviet literary and journalistic maga-
zine that has been famous for its liberal traditions
since the 1960s, recently entered the new field of the
Soviet economy. Among several economic articles
published in the magazine in 1987 was one by Se-
lyunin and Khanin' which, while not challenging the
Soviet socioeconomic order, presented findings on
Soviet economic growth that made it an interesting
subject for both domestic and overseas debate. The
most sensational assertion made by Selyunin and
Khanin was that Soviet national income did not grow
90 times in the 1928-85 period, as officially claimed,
but only six to seven times.2 In other words, according
to the Selyunin-Khanin estimates, the Soviet growth
rates, if measured by national income, were on aver-
age only about 7 percent of the officially reported
growth rates. Without taking sides in the dispute,
one's immediate response could be: How is such a
difference possible? This is the chief question I will
address in this paper.
The Soviet authorities' willingness to allow the Se-
lyunin-Khanin estimates to appear is unusual. Khan-
in's earlier publications reflected different official
treatment.' Ina 1981 article Khanin discussed his
alternative indices of industrial growth in the 1961-70
period. The discussion, however, took a strange form:
percentage differences among six sets of estimates
were shown without the estimates themselves. Pro-
gress was made in Khanin's 1984 article. Although he
did not show his estimates there either, he at least
mentioned his findings concerning the nature of
growth. He said there was a relatively slow growth of
Soviet national income in the 1928-41 period. Fortu-
nately, the new Selyunin-Khanin article appears to be
written at the right place and at the right time.
In revising the whole picture of Soviet economic
growth, Selyunin and Khanin have touched an espe-
cially sensitive period-the first three five-year plans.
This period was for a long time a "sacred cow" for the
Soviets. To be proud of Soviet socialism, one should
believe that the period of building its foundation was
not just full of Stalin's excesses. Rather, one must
believe that it also was the time of great achievements
when the country leaped from a backward to an
industrialized society. The terror was the price for the
gigantic progress. But if the Soviet economy grew not
by the gigantic five times, as officially claimed, but
only by a modest 1.5 times, as Selyunin and Khanin
claim, how could the ideology justify the sacrifices?'
Selyunin and Khanin are not the first ones who openly
challenged the success of Stalin's industrialization.
But the previous attempts were only publicized
through samizdat channels.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Some other studies which I will discuss along with the
Selyunin-Khanin article also suggest alternative esti-
mates of Soviet real economic growth. Unlike Se-
lyunin and Khanin, their authors are not concerned
with the entire Soviet period, but only with the last
several decades. From the methodological standpoint,
this time is more interesting than the Stalin period. In
analyzing the views of the authors, I will stress both
the pros and the cons of their approaches. The
comparisons with the prevailing Soviet methodology
will be made, and the problems of the latter will also
be given attention.
The second section of this paper describes the method-
ology of an alternative approach to estimating Soviet
economic growth suggested by Khanin. In the cases
when aggregate calculations for Soviet industry are
possible, the estimates that I obtained on the basis of
available information are also given. In the third
section the estimates of the critics of the official
Soviet statistics are analyzed. The arguments of the
defenders of the Soviet statistics and the views of the
officials are discussed in the fourth section. The
conclusions are drawn in the last section.
The Methodology and Its Application
There is a significant difference between the elaborate
discussion in the Selyunin-Khanin article and the
description of methodology in Khanin's earlier publi-
cations. To explain his methods in detail, Khanin
would have needed numerical examples which were
not permissible at the time of his earlier publications.
For this or other reasons, the explanations of his
methods are not clear, and my analysis will be based
in part on Khanin's words and in part on my interpre-
tation of them.
Central to Khanin's approach are six alternative
methods for estimating industrial growths (Not only
is industry the most important sector of the Soviet
economy, creating half of its national product, but
most of price inflation takes place there.) Since
Khanin realizes that each of the methods would have
shortcomings, the thrust of his approach is to use as
many methods as possible, check the consistency of
the resultant estimates, and average them. Whenever
building alternative indices, Khanin tries to avoid
using prices or the value indicators that, he believes,
are significantly affected by price increases.
In the first method, a representative sample of physi-
cal outputs is selected, and a quantity index of a
conventional Laspeyres type is constructed. However,
instead of base-year prices, coefficients of labor time
required for the production of each good in the base
period are used. As usual, the chief problem concerns
new goods which were not manufactured in the base
period. Khanin apparently manipulates the list of the
new goods, and he takes into consideration only those
for which he can find old goods that they replaced.
The old good's labor coefficient in the base period is
then used as an equal weight for each such pair. Even
if an increase in the labor time for the new good can
be justified, it is probably not taken into account.
Hence, this Khanin index is equivalent to the conven-
tional Laspeyres index in which the base-year prices
of new goods are always the same as the base-year
prices of old goods.
In the second method, the growth index is measured
as the product of a labor productivity index and an
employment index. The estimation of labor productiv-
ity is the key part of the method. Measured by TsSU
(Central Statistical Administration) as an average
product per worker, labor productivity can grow as a
result of the. increase of both real output and prices.
As I understand it, in the absence of data on real
output, Khanin substitutes industry's material expen-
diture (purchases from other firms) in the numerator
of the labor productivity index. Actually, real, not
total material, expenditure should have been related
to real output, but data on real expenditures are
harder to find. In making this substitution, two
underlying assumptions are required: first, that the
material expenditure-to-output ratio is constant over
time, and, second, that there is no significant price
inflation of material inputs. Despite the fact that
material expenditure per ruble of industrial output
officially fell by 4 percent in the 1971-84 period,
however, the second assumption does not appear
plausible.6 Probably Khanin excluded certain unspeci-
fied components of material expenditure (e.g., pur-
chases of finished parts) in order to reduce the
influence of price increases.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Given the two assumptions, the computation of the
index of labor productivity is based on the fact that
material expenditure is the difference between pro-
duction costs and wages. Using indices of cost, aver-
age wage per worker, and the proportion of wages in
cost, Khanin finds the index of labor productivity.
Since the formula explaining this computation is the
only one in Khanin's article, it deserves special atten-
tion. The formula is erroneous. A correct formula
consonant with Khanin's intent and in his notation is
the following form:
(1-KZ~) IZ~ X KZ~_,
(1)
where IPf and IZ~ =index of labor productivity and
index of average wage per worker in the jth period,
respectively; KZ~ and KZ~_, =the proportion of
wages in production cost in the jth and (j-1)th period,
respectively.
In this case Khanin obtained estimates consistent with
those from other methods, so it is unlikely he actually
used the wrong formula: I believe that his computa-
tion was correct from the standpoint of his methodolo-
gy, but he failed to formalize it adequately in the text.
Using the index of labor productivity IPA from formu-
la (1) and the index of employment ILA, the index of
production IQ~ is determined:
IQ~ = IPf X ILA (2)
Since the computation in this method is performed at
the aggregate level and the information is available
for Soviet industry, I did my own computation for the
1971-75, 1976-80, and 1981-85 periods. The data are
in table 1 and the computation is in table 2.
As the computation in table 2 illustrates, the idea of
formula (1) is that: the greater the level of wages per
worker and the lower the share of wages in production
cost in the current compared to the base period, the
greater is the level of material expenditure. This, in
turn, means that, under the assumption of material
expenditure rising along with real output, average
product per worker also grows. The indices from table
2 are compared with the estimates from the two other
methods, obtained below, and with the official Soviet
indices in table 7. The results are apparently close to
the official indices. This could be expected, given that
cost is affected by the same forces as the value of
production. Khanin assumed that material expendi-
ture is less subject to the inflationary pressure than
total output.
If the assumption is correct, then total output has to
grow more slowly than its net value component, that
is, the sum of wages and profit. But the indices of
average wages from table 2 were growing much less
than the indices of total industrial outputs illustrated
below in table 3. As for the profit component, only in
the 1981-85 period did it grow more than industrial
output. To generalize, the correlation between the
variables used in the above example, that is, costs,
wages, and material expenditure, on the one hand,
and economic growth in Soviet industry, on the other,
may not necessarily be strong. This implies that the
data used by Khanin must differ from those in table 1,
but we do not know what discounting procedure, if
any, he applied.
If the material expenditure component is imputed in
the second method, it is directly accounted for in the
third method. The purpose of this third method is to
estimate an index of material expenditure and to use
it as a proxy for the index of production. One
recognizes here the same assumptions as in the second
method, that is, that real output grows at the same
rate as material expenditure and that material inputs
are much less subject to price increases than the goods
they help to produce. To justify the use of the index of
material expenditure as an indicator of economic
growth, Khanin points to the experience of developed
capitalist countries and asserts that the difference
between their growth rates for outputs and material
inputs, as a rule, does not exceed 0.2 to 0.3 percentage
points.'
However, if one looks at US statistics, one will find a
variety of different relationships between the two
variables. For instance, in the 1972-80 period the
average growth rate of the cost of material inputs in
US manufacturing was 13.2 percent, and that of
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 1
Information for the Computation of the
Production Index for Soviet Industry (Second Method)
Proportion of wages in production cost (percent) 16.1
Average monthly wage per worker (rubles) 133.3
Number of workers in industry (thousands) 31,593
Sources: Narkhoz SSSR 1970, p. 174; Narkhoz SSSR 1975,
p. 230; Narkhoz SSSR 1980, p. 153; Narkhoz SSSR 1985,
pp. 126, 391, and 397.
Table 2
Computation of the Production Index
for Soviet Industry (Second Method)
Index of average wage per worker 162.2 _ 1 22
133.3
Index of labor productivity (1-.146)1.22 (.161) = 1.37
(1 - .161) (.146)
Index of employment 34054 = ] 08
31593
Index of production ].37 (1.08) = 1.48
Source: Table 1.
185.4= 1
14
210.6= 1
14
.
162.2
.
185.4
(1-.148)1.14(.146) _
1 12
(1-.141)1.14(.148) =
1
21
(1 -.146) (.148)
.
(1-.148) (.141)
36891 = 1
08
38103 = 1
03
.
34054
.
36891
1.12 (1.08) = 1.21
1.21 (1.03) =
1.25
output (the sum of value added and the cost of
material inputs) was 11.9 percent, with the consequent
difference of 1.3 percentage points.8 In the 1980-84
period these rates were 4.2 and 5 percent, respective-
ly, with output this time surpassing the cost of inputs
by an average 0.8 points. Since, in the two consecutive
periods, the difference between these growth rates
changed direction, it narrowed for the entire 1972-84
period but still remained 0.6 percentage points. Yet
even if the difference between the two rates for the
United States was within the narrow margin indicated
by Khanin, this fact may be only remotely related to
the Soviet economy.
As in the second method, the computation in the third
method is performed at the aggregate level. I did it for
the 1971-75, 1976-80, and 1981-85 periods. Because
the data on industrial material expenditure are not
available, I estimated an implicit index of material
expenditure in the following form:
14.6
14.8
14.1
162.2
185.4
210.6
34,054
36,891
38,103
HMS X IQ~
' HM~_,
(3)
where HMJ and HM~_, =the proportion of material
expenditure in the output of the jth and (j-1)th
period, respectively; IMF and IQ~ =index of material
expenditure and output in the jth period, respectively.
One can verify that index (3) is equivalent to the
explicit index IMF = M~/M~_,, for HMS = M~/Q~ and
IQ~ = Q~/Q~_t. In essence, formula (3) uses the index
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 3
Information for the Computation of the Index
of Material Expenditure for Soviet Industry
(Third Method)
Proportion of material
expenditure in industrial
output (percent)
Index of industrial produc-
tion by five-year period
(percent)
Table 4
Computation of the Index of Material
Expenditure for Soviet Industry
(Third Method)
64.1
65.2
63.2
61.8
1975
.652 (1.43)
- 1.45
.641
143
124
120
1980
.634 (1.24)
- 1.21
.652
1985
618 (1
20)
.
.
= 1.17
i. 99, 12
7.
.634
of production to restore the index of material expendi-
ture; the production index is then ignored, and the
material expenditure index is identified with the index
of real production. The data used in formula (3) are in
table 3 and the computation is illustrated in table 4.
In table 7, the estimated indices are compared with
the official growth indices. As follows from the table,
the estimates from the third method obtained in table
4 are almost exactly the same as the official indices.
Hence, this method, as well as the second one, does
not dramatically reduce the official growth figures as
it should have according to Khanin. This means that
he used other data on material expenditure. In partic-
ular, as he indicates in the case of the third method,
he excluded from material expenditure the cost of
purchased finished parts. It is not clear whether other
discounting procedures were also applied to the data.
If the second and the third methods implement the
same idea but different computational schemes, the
fourth method is yet another version in which the
index of labor productivity" is used". An assumption is
made that there is a functional relationship between
the growth of output per worker and productive
consumption of electricity. Moreover, another as-
sumption is that the ratio between the growth of
output per worker and productive consumption of
electricity must be the same for the US and Soviet
economies. If so, then the data on the United States
can be used to derive conclusions for Soviet industry.
More specifically, the index of labor productivity for
Soviet industry can be found on the basis of the ratio
of the average product per worker to productive
consumption of electricity in US industry and the
information on electricity consumption per worker in
Soviet industry.
For example, if, in a given period, the average product
per worker in US industry rose fourfold and electric-
ity consumption per worker doubled, the ratio be-
tween the two indices equals two. Therefore, if in the
Soviet case electricity consumption per worker grew
1.5 times in that period, the labor productivity index
must equal three. Actually, since Khanin's procedure
is equivalent to solving a proportion in which output in
the numerator and electricity consumption in the
denominator are divided on a per worker basis for
both the Soviet and the US economies, such a division
cancels out the number of workers. Probably Khanin
performed such a division since there was readily
available information on electricity consumption per
Soviet worker in statistical yearbooks. To illustrate
the fourth method, I will use the simplified version in
which there is no division by the number of workers: "
IQSU = IQ X IEsu?
IE~S
where IQ and IE =index of industrial production
and electricity consumption for Soviet and US
(4)
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 5
Information for the Computation of the
Production Index for Soviet Industry
(Fourth Method)
US industrial production index for the five-year
period
US electric utility sales to industry 571
(billion kilowatt-hours)
Soviet index of electricity consumption per
industrial worker
Sources: US Bureau of the Census, Statistical Abstract o/ the
United States: 1987 (Washington, DC, 1986), pp. 553 and 734;
Narkhoz SSSR 1985, p. 78; table 2.
industries, respectively. The information and the com-
putation according to formula (4) are in tables 5 and
6. (Since the information on Soviet electricity con-
sumption is given per worker, the index of electricity
consumption can be found by multiplying the index of
electricity consumption per worker and the index of
employment in industry.)
In table 7 the estimates from table 6 are compared
with the estimates from the two other methods consid-
ered above and the official growth indices. As one can
see from table 7, if the official and the two other
indices demonstrate a similar pattern of decline, the
fourth method, on the contrary, points to a picture of
slightly accelerating growth. In particular, this
growth is significantly higher than that officially
reported for the 1981-85 period. One of the reasons is
that the relationship between the ratio of US industri-
al output to productive consumption of electricity and
Soviet labor productivity is far from evident. But,
while this explains why the resultant estimates may be
unreliable, it does not explain why Khanin obtained
different estimates. He probably should have used the
same data on the US economy. Of the three indices in
table 6, the only one pertaining to Soviet industry
refers to the productive consumption of electricity.
Yet, even for this index, it is difficult to expect big
discrepancies since it is verifiable by comparison with
the growth of Soviet electricity production in physical
terms.
discrepancy between meeting plan targets in money
terms and meeting the targets in physical terms and
to use that discrepancy in building the index of
industrial production. For that purpose, an index of
meeting the physical production quota is computed for
each good from a selected sample. Using the individ-
ual goods indices, a weighted average index is found,
with labor time spent on the production of each good
as a weight. If it then follows from the average index
that the plan for physical output is surpassed by, say,
2 percent, but the plan for the value of output is
surpassed by 5 percent, the difference is attributed to
inflationary growth. Consequently, the difference is
not counted as real growth. Thus, if in this example
the planned growth rate for the value of production is
4 percent, the estimate of the achieved real growth
rate equals 6 percent (104 X 1.02-100). The method,
is, therefore, based on the distinction made between
the planned production values that do not foresee
hidden price increases and the production values
reported afterward that do incorporate such increases.
The purpose of the index of meeting the plan targets
in physical terms is to permit a correction of the ex
post estimates.
The fifth method can be considered as a version of the
first method, even though the two methods may seem
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 6
Computation of the Production Index
for Soviet Industry (Fourth Method)
US index of electricity consumption in industry 688 _ 1 20
571
Soviet index of electricity consumption in 1.27(1.08)=1.37
industry
Table 7
Comparison of Different Indices
of the Growth of Soviet Industry
(Year - 5 = 100)
Year Estimated Index
Second Method Third Method Fourth Method
~1) (2) (3)
1975 148 145 123
1_08X1.37 =1.23
1.20
815=1.18 8?7= 1.01
688 815
L28X1.18 =1.28 115X1.16 =1.32
1.18 1.01
Sources: Tables 2, 4, and 6; Narkhoz SSSR 1985, p. 99.
to be unrelated. The only difference is that, in the first
method, the physical outputs of the current and base
years are directly compared, whereas, in the fifth
method, an intermediate indicator of plan targets in
value terms is inserted. It is reasonable to expect that
in both methods the same set of physical goods is used
and the same weighting procedure is applied. Under
these conditions, if no price increases or decreases are
planned, the two methods should result in the same
estimates.
In the sixth method, Khanin again returns to the
index of labor productivity used in the second and
fourth methods. This time, the index of the cost of
industrial exports is estimated. Probably, the cost of
industrial exports is related to total cost, which, in
turn, is related to output. What is clear from the
method's description is that a ratio of the change in
the export cost to the change in export proceeds is
calculated, and then, given this ratio and the index of
export proceeds, the iridex of the cost of exports is
found. Export proceeds are estimated in constant
world prices, and the reliance on the true "constant"
nature of those prices seems to be the rationale for the
method. I view this method as the least justifiable
from an analytical standpoint, not to mention the
number of assumptions required.
The Criticism and Its Analysis
The methods described above give an idea of Khanin's
approach to deriving "alternative estimates" of Soviet
economic growth by sector and as a whole. Khanin
does not demonstrate his estimates for industry and
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 8
Comparison of the Growth
of Soviet National Income
capacity to characterize the efficiency of the Soviet
economy and to eliminate the effect of the "price
factor." For estimating economic growth, they also
use physical output. Calculations were performed for
industry and its three sectors-electric power, chemi-
cal, and machine building and metalworking
(MBMW?-by five-year periods since 1950. What is
interesting about the Val'tukh-Lavrovskiy estimates is
that, until the 1966-70 period, they exceeded the
official growth indices for industry, and since then
they have sharply decelerated.' Even though it is
impossible to compare Khanin's and the Val'tukh-
Lavrovskiy estimates for industry, we learn that they
deviate from the official indices in opposite directions
until the late 1960s and in the same direction ever
since.
Period
Selyunin-Khanin
Official
Total
Growth
Growth
Rate
Total
Growth
Growth
Rate
1961-65
24
4.4
37
6.5
1966-70
22
4.1
45
7.7
1971-75
17
3.2
32
5.7
1976-80
5
1.0
23
4.2
1981-85
3
0.6
19
3.5
Sources: Vasiliy Selyunin and Grigoriy Khanin, "Lukavaya tsifra,"
Novyy mir 2 (1987), pp. 196-197; Narkhoz SSSR 1975, p. 563; and
Narkhoz SSSR 1985, p. 39.
other sectors of the Soviet economy, but Selyunin and
Khanin do so for national income. The growth per-
centages from the Selyunin-Khanin article are com-
pared to the official data in table 8. As one can see
from the table, Selyunin and Khanin especially dis-
count the official growth rates for the period from
1976 to 1985. Although Selyunin and Khanin pay
attention to distortions in growth indices caused by a
variety of factors including faulty or fraudulent statis-
tical reports (pripiski), concealed price inflation is
stressed in Khanin's methodology.
Is a higher pace of price inflation in the Soviet
economy in the 1976-85 period responsible for such a
difference between the official 'and the Selyunin-
Khanin estimates? Unfortunately, there are no
grounds for conclusive answers. As indicated above,
Khanin's methods raise questions of justification,
plausibility of assumptions, and the validity of data
used. An even greater problem in this respect is that
Selyunin and Khanin's estimates do not ailow for - -
product quality improvements. They admit this short-
coming but, apparently, do not believe it is serious.
Since such a belief is typical for the critics of the
official Soviet growth figures, it is crucial we under-
stand the motives for it.
The critics in general prefer physical output as an
indicator of economic growth. Thus, Val'tukh and
Lavrovskiy use the indicator of physical production
It is, however, possible to compare the Val'tukh-
Lavrovskiy and the Selyunin-Khanin estimates for the
MBMW sector. Although Selyunin and Khanin's
article does not contain their indices for this sector,
one can restore their approximate range from the
information they give. Thus, according to their calcu-
lations, in each of the five-year periods since 1965
hidden inflation in the MBMW sector floated be-
tween 27 and 34 percent. The growth in the 12th
Five-Year Plan (1986-90) is foreseen to be 43 percent
of which, according to Selyunin and Khanin, 30
percent will constitute the inflationary component,
that is, growth "on paper."10 (Apparently, they consid-
er the real growth rate as the difference between the
nominal and inflation rates, neglecting the crossrate
term.) The Selyunin-Khanin estimates obtained by
discounting the official Soviet growth indices are
compared with the Val'tukh-Lavrovskiy estimates in
table 9. As one can see from table 9, the 1961-85
deceleration rate in the Selyunin-Khanin estimates,
from an average 8.2 to 0.9 percent, is much lower
than in-the Val'tukh-Lavrovskiy estimates; from-10:7-
to 0.3 percent.
Fal'tsman uses the maximum power of equipment as
an indicator of growth for the MBMW sector. In
particular he says that, while from 1970 to 1982 the
average rate of output growth for 11 civilian MBMW
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 9
Growth of the Soviet MBMW Sector
by Five-Year Period
Year
- - ----
Selyunin-Khanin
Val'tukh-Lavrovskiy
Official
Growth
Growth Rate
Growth
Growth Rate
Growth
Growth Rate
1965
45-52
7.7-8.7
66
10.7
79
12.3
1970
40-47
7.0-8.0
33
5.9
74
11.7
1975
39-46
6.8-7.9
26
4.7
73
] 1.6
1980
14-21
2.7-3.9
6
1.2
48
8.2
1985
1-8
0.2-1.6
1 a
0.3
35
6.2
1990
13 b
2.5
NA
NA
43 b
7.4
Sources: Narkhoz SSSR 1970, p. 137; Narkhoz SSSR 1975,
p. 197; Narkhoz SSSR 1985, p. 99; Vasiliy Selyunin and Grigoriy
Khanin, "L.ukavaya tsifra," Novyy mir 2 (1987), p. 187;
K. K. Val'tukh and B. L. Lavrovskiy, "Proizvodstvennyy potentsial
strany," Ekonomika i organizatsiya promyshlennogo proizvodstva
2 (1986), pp. 24 and 29.
a 1983.
b Projection.
ministries was about 8 percent in money terms, it was
only 3 percent in equipment maximum power." The
difference of 5 percent is interpreted as an inflation-
ary component of growth. Fal'tsman's average growth
rate of 3 percent turns out to lie in between the
Val'tukh-Lavrovskiy and the Selyunin-Khanin aver-
age estimates of 2.5 and 4.5 percent, respectively, for
the same 1971-82 period. Fal'tsman's estimates may
be more convincing than those of Selyunin and
Khanin or Val'tukh and Lavrovskiy, since he takes
into consideration not only the physical output of
machines and equipment but also their power. His
approach, however, is only justified when power could
be singled out as the most important characteristic of
machines.
These examples demonstrate that, in estimating Sovi-
et economic growth, Soviet economists pay little
attention to quality change. In a similar situation, a
Western analyst would never ignore such change, at
least verbally. People in the Soviet Union generally
give quality less consideration than quantity. This
should not be understood to mean that quality is not
important to them; on the contrary, seeking goods of
an adequate quality consumes a substantial part of
their lives. But they are hardly used to improvements
in domestic manufactures and, for that reason, often
disregard the factor of quality change. Persistent
shortages play their role, too. As for the producer, he
takes whatever the supply system is able to provide.
Since there is no shopping around, the problem of
quality becomes too hypothetical in the producer's
case.
In my work on estimating real growth in the Soviet
MBMW sector, I used an approach similar to Fal'ts-
man's as well as a more general one. The estimates of
real growth for the Soviet electrotechnical industry
are an example of the first approach. According to my
calculations, the output of the electrotechnical indus-
try rose by 138 percent in the 1961-75 period, whereas
the Soviet statistics report a 356-percent growth.
Hence, my estimates command an average 6-percent
growth rate, and the official, 10.6 percent.'Z It there-
fore follows from these estimates that economic
growth for the electrotechnical industry may be over-
stated by 1.8 times (10.6 - 6); this is lower than the
2.7 times (8 _ 3) implied by Fal'tsman's estimates for
the entire MBMW sector."
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
As an example of a more general approach, the
estimates for Soviet passenger car production in the
1970-82 period could be noted. In this project, it
would be impossible to use indicators such as produc-
tion capacity or any more reasonable one which
clearly dominates the others. For that reason, I built
an index of car quality based on the following charac-
teristics: the number of occupants, size, weight, en-
gine volume, horsepower, and fuel efficiency. Esti-
mates were obtained for both the wholesale and retail
levels of sales revenue. Although an inflationary
component was found to be present at both levels, at
the wholesale level it turned out to be insignificant
when compared to resultant real growth rates: The
average nominal growth rate was discounted by only
1.8 percentage points, from 14.2 to 12.4 percent. At
the retail level, this rate was discounted by 4.3
percentage points, from 16.7 to 12.4 percent." In the
1976-82 period, there were even years of disinflation-
ary growth in wholesale prices. Such a result is
somewhat unexpected, in light of a widespread belief
that the Soviet automotive industry has been notori-
ous for its price inflation.
Why then do we see quality improvements in the
Soviet economy where Soviet economists do not? As
outside observers, we are perhaps ready to consider as
improvements such changes in product characteristics
which, under normal conditions of consumer sover-
eignty, should not have been identified so. The deter-
mination on quality is made in the USSR by state
certification committees, and regulations set up stan-
dards for "objective" characteristics of quality. When
these characteristics are the same for a large product
group, they only reflect capacity, power, durability,
and the like. This is not the place to discuss the well-
known issues of Soviet product quality. However, if
agriculture needs small and versatile tractors and the
industry, instead, continues to raise their horsepower,
a Soviet economist does not want to accept such a
change as an improvement. That is the only way I can
interpret the position of Selyunin and Khanin or
Val'tukh and Lavrovskiy.
On the other hand, there is a problem of measuring
inflation in this case. An increase in a tractor's
horsepower likely causes a rise in production cost and,
consequently, the price. The approach of Soviet econ-
omists discussed above means that, since such an
increase could not be viewed as a quality improve-
ment, the price increase should be treated as inflation-
ary. While the issue is indeed complicated, one should
take into account the reality of planning and evaluat-
ing success in the Soviet economy. The Soviet produc-
er has to play by the rules established by the state
certification committees and pricing authorities for
the evaluation of product quality and price setting.
Accepting the reality and the criteria established by
the authorities, we consequently look at the quality
problem from the producer's, rather than the consu-
mer's, standpoint: If the producer has to manufacture
a good with new characteristics improved in accor-
dance with the established criteria, a proportionate
price increase may be justified. In essence, we say: If
the growth is not inflationary, it is real. The Soviet
critics of the official statistics do not accept such a
compromise.
They are not the only critics of Soviet statistics. The
planners themselves always prefer the indicators in
physical goods to those in money terms. The national
economic plan is primarily built on balances of pro-
duction and distribution of physical goods, and in
many instances monetary aggregates are used for
accounting purposes. Thousands of material balances
drawn in planning, to a large extent, insulate plan
targets from price distortions. Although fraudulent
reports on physical outputs are not exceptional, it is
much easier to control them than the value indicators.
Not surprisingly, people more or less accept Soviet
statistical data on the production of oil, machines, or
shoes. Serious imbalances do, of course, occur in the
Soviet economy. But to consider them a result of poor
planning, as many critics do, would imply that better
planning would cure Soviet economic problems.
The primacy of planning in physical terms led to a
"neglect" of monetary indicators. They did not play
an active role in planning before 1965, when their
important political function was to demonstrate high
rates of Soviet economic growth. Since the 1965
economic reform, inflating value indicators became
much less harmless from the planners' perspective.
The reason is that profit and average labor productivi-
ty were chosen, among others, as the new success
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
indicators. Since the rise of these monetary indicators
could affect wages and bonuses, this created a new
reality in the game between plant management and
planners; managers intensified their effort to raise
prices, and planners attempted to create barriers for
unjustified price increases. Continuing price rises
indicate that the planners' preventive barriers did not
succeed. As a matter of fact, along with the prewar
years, the 1976-85 period is one for which the Se-
lyunin-Khanin estimates reduce the official growth
rates the most drastically.
Soviet statistics do not appear to respond to these
dramatic developments at all. There are in general
many unanswered questions with respect to Soviet
statistics, and the chief one is: If there is inflation in
the Soviet economy, why is it hidden and why is it
reported as economic growth? The reason is not to be
found among statisticians themselves or in their prac-
tices but in Soviet methodology for the computation of
growth and the procedures for pricing new goods.
The Soviet methodology of building growth and price
indices is based on using comparable prices (soposta-
vimyye tseny). In industry, 1982 prices are now used
as comparable ones; in agriculture, 1983 prices are
used, and in construction, 1984 prices. Only at the
moment of setting these prices do they resemble
conventional constant prices. For a good introduced
later, say, in 1987, the first approved price becomes its
1982 comparable price. Technically, this solves the
problem of estimating the base-period price for the
new goods, and the Soviets do not have to worry about
linking the old and the new specifications when
building an index. But what it actually means is that
the whole array of goods introduced in between the
setting of comparable prices (in industry, the base
years were 1952, 1955, 1967, 1975, and 1982) does
not affect the price index one way or another: Their
base- and current-year prices are one and the same.
Such a price index, therefore, only reflects the revi-
sions of prices for the goods produced in the base
years.
As Selyunin and Khanin suggest, the Soviets should
have used representative samples of goods in each
product category instead of entire outputs, and built
indices for the samples. The rationale is simple. Data
verification on the entire output is hardly possible.
Yet, as I explained elsewhere, the Soviets apply the
same comparable price principle to sample indices,
too.15 Therefore, the mere change of the sample size
will not ease the problem of hidden inflation. Why
then not change the methodology and switch to a
conventional Western procedure that involves the
computation of base-year or current-year weighted
indices? This, I am afraid, is not a panacea either.
At a closer glance, the Soviet methodology is not bad
at all. The linking procedure usually involves a com-
parison, either direct or indirect, of a new good with
the existing ones in the same product category. One
would hardly comprehend exactly how the Bureau of
Labor Statistics in the United States makes a judge-
ment for goods other than automobiles. The Soviets
possess an advantage in this respect, since such a
comparison is required and is performed in the process
of approving a new good's price. It is therefore logical
to set the new good's price with the account of
improvements in its characteristics. Hence, theoreti-
cally, price increases for new goods are allowed to the
extent of their projected quality improvement. Yet,
due to a number of well-known systematic forces,
these improvements mostly remain on paper, while
price growth is quite real.
Suppose that the Soviet methodology of constructing
indices is revised, and more conventional procedures
for linking newly introduced goods to those in the
market basket are adopted. The same considerations
that are now used in pricing new goods would have to
be used in quality and price comparisons by statisti-
cians. Then, in the cases when the Goskomtsen (State
Committee on Prices) experts find price increases
justifiable, there is little chance that the statisticians
constructing the price index would be able to add any
new insights to the issue. In other words, the change
in the Soviet methodology per se would not raise the
reliability of Soviet growth and price indices. In the
absence of consumer sovereignty, any procedures for
evaluating goods quality would turn into something
similar to what Goskomtsen does at present; if so, the
results would not be much different.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
The Analysis of Orthodox Views
The Selyunin-Khanin estimates have been challenged
by Adamov.16 He believes that a growth index based
on the entire output, as in the Soviet Union, is more
accurate than one based on a sample of goods, as in
the United States. Adamov demonstrates that, by
manipulating a sample of goods, it is possible to
obtain for the 1976-85 Soviet industrial output either
a decline of 29 percent or a rise of more than 300
percent. This is done to prove that, by purposeful
selection of an "appropriate" sample of industrial
goods, Selyunin and Khanin intended to discredit
official growth figures. Two comments are appropri-
ate. First, the Selyunin-Khanin estimates are much
less vulnerable to Adamov's criticism than, for exam-
ple, are the Val'tukh-Lavrovskiy estimates. The rea-
son is that Val'tukh and Lavrovskiy's estimates were
built on the basis of a sample in physical terms; it is
quite possible that the growth characterized by the
sample for which they were able to obtain data is
biased with respect to the entire output. The Se-
lyunin-Khanin estimates, on the other hand, are aver-
aged over the indices computed by several different
methods, only two of which depend on sample estima-
tion. Second, whereas there are both cons and pros in
increasing the size of a sample, the chief problem is in
the method applied, not in the fact of using a sample.
Thus, if the methodology of the US Bureau of Labor
Statistics were applied to exactly the same set of
goods the Soviets use, the results would undoubtedly
differ from the Soviet official indices. Conversely, if
the Soviet methodology, the idea of which was dis-
cussed in the previous section, were applied to the
entire industrial output and separately to a represen-
tative Soviet sample, the two results would probably
not differ significantly.
Adamov asserts that when a switch to new compara-
ble prices takes place, the last year's value of output is
recalculated in new prices, to allow for the computa-
tion of a chain growth index. This is true; but the
recalculation is done only for the so-called compara-
ble products (sopostavimaya produktsiya), that is,
those manufactured in both current and previous
years. The prices of these goods seldom rise, and the
washing out of cheap items and the introduction of
new goods make the difference. Therefore, despite
Adamov's assertion, the Soviet experience has proved
that the mere fact of using the chain index does not
ensure the smooth continuity of indices. While Khan-
in's criticism is chiefly directed toward the Soviet
methodology, Adamov believes that the prime sources
of erroneous results in statistics are the deliberate
distortions in the initial information (pripiski) or the
use of samples rather than the entire sets of goods.
Adamov makes several points on Selyunin and Khan-
in's estimates. He notes that if Soviet national income
rose six to seven times in the 1928-85 period as they
say, then in 1985 the ratio of Soviet national income
to that of the United States would remain at the 1928
level, that is, 10 percent on average. Then how could
the Soviets afford parity in military spending and still
produce consumer goods? The argument clearly
makes sense. However, Adamov believes in both the
official growth of 90 times from 1928 to 1985 and in
the official ratio of Soviet-to-US national income of
66 percent in 1985, the combination of which does not
make sense. Indeed, if, according to Adamov, a 10-
percent ratio of Soviet-to-US national income follows
from asix- to seven-times growth in Soviet national
income, then a 66-percent ratio would command a
6.6-times greater growth, that is, one in the 40- to 46-
times range. The latter still falls short of the official
growth of 90 times.
Several other of Adamov's numerical examples were
aimed to disprove the Selyunin-Khanin estimates and
can also be challenged. For example, following up the
Selyunin-Khanin approach to look at the proportions
between technologically related industries in the US
and Soviet economies, Adamov uses metallurgy and
the MBMW sectors. He states that in the 1961-85
period the ratio of MBMW-to-metallurgy growth for
the United States was equal to 3.3, with the indices of
3.07 and 0.93, respectively. From that standpoint, he
indicates as reasonable a similar 2.7 ratio for the
Soviet economy. Yet, to find the growth index for the
Soviet MBMW sector, Selyunin and Khanin might
use the metallurgy growth index in physical, not in
value, terms. Using steel production as an indicator,
the growth index for Soviet metallurgy in the 1961-85
period would be 2.4. Since a similar index for the
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
MBMW sector equals 10.9, one would obtain a 4.5
ratio (10.9 - 2.4), rather than 2.7." (The ratio is
important in the sense that, the lower its level, the
more "reasonable" are the official growth rates for
the MBMW sector.)
Adamov's views on the Selyunin-Khanin findings
would probably coincide with the official reaction of
TsSU. But TsSU, recently transformed into the State
Committee on Statistics (Goskomstat), has so far
ignored them. In his article outlining the program for
perestroyka of statistics, the head of Goskomtstat
Korolev only in passing remarked that all Selyunin-
Khanin calculations and conclusions are "deeply er-
roneous."'B Among the targets for improvement out-
lined by Korolev are the composition of statistical
indicators, the reliability of the initial data, and the
methodology of the measurement of economic growth
and the investment process. The directions of opera-
tional changes will include the elimination of all the
channels of collecting statistical data outside of the
Goskomstat system and the reduction of the amount
of data required from industrial firms and
organizations.
When it comes to the composition of statistical indica-
tors, the Soviet press reports a lot of criticism and
vague suggestions for improvements. The existing
indicators are called separate (razroznennyye) or
piece-wise (kusochnyye), while they are supposed to
provide a composite picture of social and economic
development. But that is what Soviet planners at-
tempted to achieve for decades! Take, for example,
the plan for technological change. Many research
projects were devoted to development of an indicator
characterizing an overall state of technology. The
projects failed because there is no such indicator. The
current campaign can only result in the introduction
of a broader range of indicators of economic and
social statistics. If, in addition, they are published,
such a development can only be welcomed.
The official Soviet perception of the reliability of
statistical data is that whatever statisticians process is
true to the extent the initial information collected
from firms and organizations is true. This, in particu-
lar, implies that the methodology per se challenged by
Selyunin and Khanin is also true. Every Soviet leader
beginning with Stalin tried to prosecute statistical
fraud (pripiski) and eyewash (ochkovtiratel'stvo). At
the same time, the leaders themselves initiated and
encouraged the fraudulent Stakhanovite movement
and other forms of socialist competition. But this type
of fraud was considered "innocent" in the sense of,
first, being authorized and, second, not affecting the
overall production statistics. Thinking of the new
Soviet campaign against statistical fraud, Gorbachev
may be sending a signal that there will not be a
discriminatory treatment of pripiski and that none of
them will be tolerated any more.
There has already been some evidence of Gorbachev's
new broom sweeping pripiski. For example, according
to the head of the Moldavian Goskomstat, Vorotilo,
after the 1986 resolution of the Central Committee on
eyewash and fraud in Moldavia, Kirovograd province,
and the Ministry of the Automotive Industry, more
than 2,000 Communists were charged with misde-
meanors in Moldavia." Disciplinary punishments
were given to 1,260 people, one-third of whom were
managers, and 111 of them were fired. For fraudulent
data, 29 cities and districts were stripped of their
positions as the winners in the socialist competition.
But, as the official admits, cheating is still alive.
Further, the head of the Uzbek Goskomstat Sadukov
indicates that, despite all the effort, one out of six
enterprises in the republic still engages in pripiski.20
This is not the place to analyze the distortions of
statistical information at industrial firms, where the
bulk of it originates. Even though the recent disclo-
sures illustrate the serious nature of the problem, I
think it is premature to conclude that the level of
proportions is critical. This statement is in part based
on my experience of working with plan and statistical
data in the 1960s and 1970s. There have always been
indicators that are more trustworthy, such as physical
outputs, and those that are untrustworthy, such as
economic benefits from modernization projects. Plan-
ners are well aware of exaggerated volumes of excava-
tion or freight traffic. But in the past those false
volumes were quietly accepted as a means of paying
workers decent wages and thus keeping them on the
job.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
But what about pripiski at the macro level, in situa-
tions when aggregate statistics do not look impressive
or even move in undesirable directions? Whenever
Soviet statistical reports are unbelievable, Western
analysts naturally suspect deliberate distortions. I
personally did not come across the cases of "primi-
tive" distortion, in the sense of cooking up good
numbers. The distortions practiced are more of an a
priori nature that may have nothing to do with
falsification. They follow from the Soviet planning
and statistical methodology or the changes in the
methodology or accounting procedures which remain
undisclosed. For example, the production of milk was
sharply increased in the early 1970s by lowering its
fat and increasing its water content; the meat con-
sumption per capita was "raised" by including the
estimated consumption of lard by collective farm
households, which was not included before. One could
find numerous examples of deliberate lowering of
product quality in order to beef up outputs. But what
is important in all these cases is that the production,
of whatever quality, actually grows if it is reported so.
Otherwise, the numbers would disappear from statis-
tical yearbooks.
History does not necessarily repeat itself, and things
might have changed in Soviet statistics. Western
analysts, however, have recently pointed to new dis-
crepancies in Soviet statistics, in particular to conflict-
ing measures for certain crucial indicators appearing
in different and even in the same reports. There were
also accusations of deliberate distortions. For exam-
ple, Vanous, analyzing the 1985 and 1986 data on
Soviet retail trade turnover, came to the following
conclusion: "It can now be proven beyond any doubts
that some of the official statistics were deliberately
distorted over the past two years."21 I, in contrast,
believe there are many other factors that could con-
tribute to the difference in estimates for that specific
period. I do not, of course, rule out distortions and
errors. However, since deliberate distortions apparent-
ly were not practiced in the past, and since Gorbachev
required a greater scrutiny from statisticians, I find it
hard to believe that someone in Goskomstat would
take a chance of "beautifying" the statistics.
As for the improvement of statistical methodology, we
cannot expect significant changes until the relevant
changes are accomplished in planning methodology.
For example, statisticians have been criticized severe-
ly for using gross value (valovyye) indicators of eco-
nomic growth. According to the rules of the game, the
Goskomstat officials have engaged in self-criticism
calling for the introduction of "scientifically justified"
(nauchno-obosnovannyye) success indicators.ZZ How-
ever, little could be done since a major shift from
value to net indicators in planning has still not been
completed. (And even if it is completed, the results
may turn out to be far from scientifically justified.)
The changes that can be expected will probably focus
on the revision of the indicators used in statistics and
the introduction of new ones, rather than on the
revision of the methodology.
Conclusion
This paper analyzes the recent developments in the
debate over Soviet statistics, in particular the debate
over Soviet economic growth rates. What is encourag-
ing about these developments is that some of the
Soviet economists have joined their Western col-
leagues in an attempt to estimate real Soviet growth.
Although research in this area had been going on in
the Soviet Union for years, only under Gorbachev's
policy of glasnost could specific estimates be pub-
lished. In this respect, the most interesting are the
estimates for Soviet industry and its sectors, including
MBMW, by Val'tukh and Lavrovskiy and for the
Soviet economy as a whole by Selyunin and Khanin.Z'
Their results drastically contradict the official Soviet
growth figures. For example, Selyunin and Khanin
assert that from 1928 to 1985 Soviet national income
rose six to seven times, instead of 90 times as officially
claimed.24
Such polar differences in estimates are a result of
problems in both the official methodology and the
methodology applied by its critics. Along with specific
shortcomings, the methods of the critics have some-
thing in common: their failure to consider change of
product quality. This sends the important message
that the critics do not believe that quality change is
indicative of the Soviet economy. If Soviet machines
become bigger, or heavier, or more powerful, they
probably do not view this as an improvement, since in
many instances the consumer does not need all of
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
these changes. Yet, even accepting the motives of the
critics, we have to realize that the quality standards
are imposed on the Soviet producer by certification
committees, Goskomtsen and Gosstandart, to name a
few, not the consumer. If, under these conditions, the
producer is required to manufacture a more powerful
machine, a consequent price increase, with all of the
usual reservations, may be justified.
Even though the Selyunin-Khanin indices do not close
the chapter in estimating Soviet economic growth,
they have a great advantage of consistency. For
industry, for example, six methods are used and the
resultant indices are compared and averaged. Taking
into account the problem of quality change, these
estimates should be viewed as a lower bound for real
Soviet growth rates, with the official indices making a
logical upper bound. One can wonder what would be
the level of significance for such an interval.
The chief questions with respect to the official Soviet
methodology are: Why does it encourage concealed
inflation and what could be done to cure the problem?
When the Soviets build their growth indices, their
linking procedures work so that any price increases
for newly introduced goods would be ignored, and
only price changes for goods that have already been in
the market basket would be counted. This is done
because price and quality comparisons are performed
in the process of price setting, and, theoretically, there
should not be unjustified price increases. In reality,
however, many well-known factors impede the process
of quality improvements, but prices, especially for
producer goods, grow. From the methodological
standpoint, it does not even matter whether the
producer does or does not raise prices. What matters
is the pressure on the producer to do anything to
justify price increases. For this reason, there is little
hope that any changes in statistical methodology
could reduce the effects of hidden inflation before
major revisions in Soviet pricing policies and in the
general economic mechanism are undertaken.
In the meantime, there is much room for partial
improvements in the Soviet statistics. Among those
could be the establishment of a single base year in
measuring growth and price indices. Not only do the
base years periodically change, but it is always one
year for industry, another for agriculture, a third one
for construction, and a fourth for the balance of the
national economy that usually coincides with the one
for agriculture. Altering the procedure for estimating
the base-year comparable prices for goods introduced
afterward might also be helpful. While it is difficult to
foresee a consistent implementation of such a mea-
sure, at least the prices set for new goods should not
be traced back to the base year automatically.
Some organizational and substantive changes have
already been going on along the lines of perestroyka
of Soviet statistics. For example, new economic and
social statistics are to be introduced. Yet an even
greater problem in this respect is the accessibility of
the existing statistics. It is the Soviets' obsession with
secrecy that makes the data unavailable even to their
own scholars, much less to the general population or
foreigners. The resultant omissions, cryptic style, or
the lack of comments on methodological changes
frequently puzzle Western analysts. The controversy
over the growth of Soviet retail trade turnover and
some other aggregate statistics is a good example.
Analyzing some of the factors responsible, I have
excluded deliberate distortions on the part of Soviet
statisticians. In the Khrushchev period, Soviet statis-
tics were made much more open. Will that happen in
Gorbachev's period? I believe that only the success of
his economic programs may pave the way to glasnost
in Soviet statistics. It should be easier to reveal the
truth when there is good news.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
1. Vasiliy Selyunin and Grigoriy Khanin, "Lukavaya
tsifra," Novyy mir 2 (1987), pp. 181-201. See also: A.
Popkova, "Gde pyshnyye pirogi?" Novyy mir 5
(1987), pp. 239-41; and Nikolay Shmelev, "Avansy i
dolgi," Novyy mir 6 (1987), pp. 142-58.
2. Selyunin and Khanin, op. cit., p. 192.
3. Grigoriy Khanin, "Al'ternativnyye otsenki rezul'ta-
tov khozyaystvennoy deyatel'nosti proizvodstvennykh
yacheyek promyshlennosti," Izvestiya AN SSSR, ser-
iya ekonomicheskaya 6 (1981), pp. 62-73; Grigoriy
Khanin, "Puti sovershenstvovaniya informatsionnogo
obespecheniya svodnykh planovykh narodnokho-
zyaystvennykh raschetov," Izvestiya AN SSSR, ser-
iya ekonomicheskaya 3 (1984), pp. 58-67.
4. V. Adamov, "Chto stoit za indeksami," Ekonomi-
cheskaya gazeta 29 (1987), p. 14 and Selyunin and
Khanin, op. cit., p. 193.
5. Khanin (1981), pp. 62-73.
6. Narkhoz SSSR 1985, pp. 99 and 127.
7. Khanin (1981), p. 67.
8. US Bureau of the Census, Statistical Abstract oJ'
the United States (Washington, D.C., 1986), p. 723.
9. K. K. Val'tukh and B. L. Lavrovskiy, "Proizvodst-
vennyy potentsial strany," Ekonomika i organizatsiya
promyshlennogo proizvodstva 2m (1986L), pp. 24 and
29.
10. Selyunin and Khanin, op. cit., p. 187.
11. V. K. Fal'tsman, Proizvodstvennyy potensial
SSSR: voprosy prognozirovaniya (Moscow: Ekono-
mika, 1987), p. 72.
12. Fyodor I. Kushnirsky, Estimation o.J'Real
Growth and Productivity in the Soviet Machine-
Building and Metalworking Sector: The F~ects on
Economic and Military Capabilities (Falls Church,
VA: Delphic Associates, 1986), p. 186; Narkhoz
SSSR 1970, pp. 205-06 and Narkhoz SSSR 1975,
pp. 255-56.
14. Fyodor I. Kushnirsky, "Growth and Productivity
in the Soviet MBMW Sector" (in progress).
15. Fyodor I. Kushnirsky, "Methodological Aspects
In Building Soviet Price Indices," Soviet Studies 4
(1985), pp. 505-19.
16. Adamov, op. cit., p. 14.
17. Narkhoz SSSR 1985, pp. 98 and 140.
18. Mikhail Korolev, "Zadachi perestroyki statistiki,"
Vestnik statistiki 4 (1987), pp. 3-12.
19. "Uluchshat' delo statistiki," (Excerpts from the
Meeting of the Collegium of TsSU) Vestnik statistiki
5 (1987), p. 45.
21. Jan Vanous, The Dark Side of'Glasnost :? Unbe-
lievable National Income Statistics in the Gorbachev
Era, P1anEcon Report 6, 1987, p. 1.
22. Nikolay Belov, "Povyshat' uroven' analiticheskoy
raboty," Vestnik statistiki 11 (1986), pp. 3-7.
23. Val'tukh and Lavrovskiy, op. cit., pp. 17-32;
Selyunin and Khanin, op. cit., pp. 181-201.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Changes in the Availability
of Economic Data Under Gorbachev
STAT
Introduction
On balance, the Gorbachev regime has had a positive
impact on the availability of Soviet statistics. The
Soviet Government is publishing more statistics than
were released in the late 1970s and early 1980s, re-
releasing some data series that had been withdrawn
previously from publication, and providing some new
information for the first time. Moscow also has begun
marketing statistical data, both at home and abroad,
and is developing new data series for the use of
planners and managers.
At the same time, however, much less economic
information is released in the USSR than in Western
countries and much of the available Soviet data are
poorly documented and methodologically flawed. For
example, some of the published data on national
income appear inconsistent. In addition, information
on defense-related matters continues to be almost
nonexistent.
This paper focuses first on past changes in the content
of the Soviet statistical yearbook (Narkhoz). Changes
in the most recent handbook, Narkhoz SSSR za 70
let (hereafter referred to as Narkhoz 198v7, are then
briefly analyzed, followed by a discussion of new data
available this year. The paper then speculates about
possible future changes in data availability.
The Soviet Statistical Handbook-A History of
Change
Narkhoz is the single most important source of
economic data openly published by the Soviet Govern-
ment. It is published annually by the State Committee
on Statistics, Goskomstat.' The availability of data in
' Other Soviet sources of economic data include an abbreviated
statistical handbook, SSSR v tsifrakh, published each spring; a
foreign trade handbook; statistical compendia for each of the 15
republics; and a CEMA handbook titled Statisticheskiy yezhegod-
nik strap-chlenov soveta ekonomicheskoy vzaimopomoshchi. Eco-
nomic data also appear in newspapers such as Ekonomicheskaya
gazeta and economic journals such as Vestnik statistiki.
this publication during the Gorbachev period can be
examined by comparing the current edition of the
statistical handbook with those published in previous
years.
According to a recent study of the informational
content of the Narkhoz from 1970 to 1985, the
availability of Soviet economic statistics has varied
markedly over time. Although the first volumes pub-
lished in the 1950s were small,z the initiation of the
Narkhoz series signified a change in the policy of
extreme economic secrecy that characterized most of
the Stalin era. That thaw continued into the late
1950s when the size of the handbook reached a peak.'
Volume size in the early 1960s declined somewhat,
indicating a mild reversal of Khrushchev's policy
toward openness.
Larger volumes appeared again in the mid-1960s at
the beginning of the Brezhnev era-a trend which
persisted until the early 1970s. During this period the
volumes were all of similar size-from a low of 752
pages in 1970 to a high of 821 in 1971. The largest
volume of the Brezhnev era was the 908-page 1967
Narkhoz, although the inclusion of the 1966 input-
output table in that volume exaggerated its
appearance.
During the later years of the Brezhnev era an atmo-
sphere of secrecy returned. Indeed, a precipitous and
protracted decline in the size of the handbook began
in the mid-1970s. The coverage of many tables was
reduced, and entire tables were deleted. In addition,
the sections that contained explanations of the meth-
odologies employed to derive the statistical data were
dropped. ?
The 1955 yearbook-published in 1956-had only 246 pages of
statistical material.
' The 1958 volume had 924 pages.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Changes in the contents of the Narkhoz on a sector-
by-sector basis over the past 15 years have varied.
While the majority of losses occurred during the mid-
70s, the coverage of specific sectors was changed at
different times. Some of the categories affected and
the volumes from which important information was
removed follow:
Data Availability Under Gorbachev
Greater statistical openness seems to be occurring
under Gorbachev. A major turning point in the
availability of statistics occurred with the publication
of the 1985 Narkhoz (published in 1986). Statistics on
grain production, infant mortality, life expectancy,
and alcohol sales and consumption were all returned
to Narkhoz 1985; trade volume data reappeared in
Category
Transportation 1971
Population 1973, 1974, 1977, 1978
Narkhoz 1986. Other statistics have appeared for the
first time-such as mechanization of labor by sector,
production of precision machine tools, and automobile
ownership by union republic.
Science 1974 The reappearance of such important information as
1975 grain production and infant mortality statistics is
si
nific
t
Th
d
i
i
t
bli
h
h
d
b
g
an
.
e
ec
s
on
o pu
s
t
ese
ata pro
a-
Labor 1976
bly was made at the highest levels of the party and
Capital 1981 government. There are several possible reasons for
Grain 1981, 1982 statistical openness:
Trucks 1982
Buses 1984
Rail equipment 1984
? An initial period of openness has generally charac-
terized leadership changes in the Soviet era.
Alcohol 1984 ? The Kremlin is using the release of statistics to help
At the same time, some new data-on computer
education, production of industrial robots, and ma-
chining-center production-were added during this
period. Other data series were added as new topics of
concern to the Soviet Government arose, such as
information on labor brigades and on self-financing
enterprises.
The primary reason for the data reduction policies in
the late 1970s seems to have been the worsening
performance of the Soviet economy. Another apparent
factor has been the utilization of these statistics in
Western studies of the Soviet economy. Following the
publication of the CIA's negative appraisal of Soviet
oil prospects in April 1977,' for example, Narkhoz
1976, the next statistical yearbook to appear, had a
large portion of energy data removed (on regional
production of oil, gas, coke, and coal, plus information
on fixed capital in the energy sectors).
gain public support for the removal of political
adversaries by showing the failings of previous
regimes.
? The release of some uncomplimentary statistics is
being used to educate the Soviet public about the
severity of social problems, to dispel public apathy,
and to justify the kind of "radical" reform the
regime is trying to implement.
? The release of more statistics may be a precursor to
seeking entrance into Western economic organiza-
tions (e.g., the IMF or GATT).
? An appearance of more openness is being used to
influence Western public opinion and policies to-
ward the Soviet Union.
? Many of the statistics being re-released are from
sectors of the economy that have experienced some
improvement in performance, thereby placing the
regime in a favorable light.
STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
A Look at Narkltoz 1986
Compared to Narkhoz 1985, the 1986 handbook
contains 83 more pages of data, 106 more tables, and
a 21-page methodological appendix-the first such
appendix published since Narkhoz 1978. In addition,
many tables in Narkhoz 1986 have been streamlined
and combined, but with little loss of important data.
The 1986 Narkhoz-like many before it-has under-
gone extensive reorganization and restructuring.
Some of the major changes include:
? The foreign relations section has been doubled in
size and divided into subsections on "foreign trade,"
"CEMA cooperation," and "economic assistance."
? A major new section has been created titled "Social
Development and Raising the Living Standards of
the People." This section combines material in the
following sections of Narkhoz 1985-"Area and
Population," "Growth of the Well-Being of the
People," "Labor," "Trade," "Services," "Educa-
tion," "Culture," and "Health."
? Much of the material in the old "Area and Popula-
tion" section is combined with some old "Labor"
section material in a new subsection, "Population
and Labor Resources."
? A new subsection, "Wages and Incomes of the
Population," combines data from "Labor" and
"Growth of the Well-Being of the People" in Nark-
hoz 1985.
? A new subsection, "Supplies of Goods and Servi-
ces," combines data from Narkhoz 1985 sections on
"Trade," "Services," and "Growth in the Well-
Being of the People."
? Data on National Income have been split. Data on
National Income Produced (NIP) are in the section
"Development of Material Production." Data on
National Income Used (NIU) are in a subsection of
"Social Development and Raising the Living Stan-
dards of the People" called "Use of National
Income."
? A table comparing the Soviet and US economies has
been moved from the section titled "International
Comparisons" to the first section, called "Basic
Indicators of the Economic and Social Development
of the USSR During 70 Years of Soviet Power."
The reorganization is clearly related to Gorbachev's
restructuring policies. Some of the changes in Nark-
hoz 1986, for example, are called for under the
guidelines for "radical improvement of statistical
work" included in the decrees on reforming the
management of the Soviet economy adopted in July
1987. One of the tasks laid down in that document is
for Goskomstat to improve the quality of statistical
information used for social planning and everyday
management of the economy. Other goals which seem
to be reflected at least partially in the reorganization
and expansion of data in Narkhoz 1986 include
information on reserves for increasing output, mea-
sures of the use of productive potential, measures of
savings of labor and material resources, and indica-
tors of the efficiency of production and the use of
science and technology in the economy.
Trade Data. Trade data in physical units missing
since Narkhoz 1975 have reappeared (Narkhoz 1986,
p. 641). Also restored is a table missing since Narkhoz
1974 showing the share of imports in total Soviet
consumption of 29 products (p. 644). A new table has
been added giving the share of exports in total Soviet
production for 31 products (p. 642).
Plan I~tformation. A considerable amount of new
detail on plan goals for 1987 is given in summary
tables of basic indicators (Narkhoz 1986, pp. 5, 7, 8).
Tables on the production of individual industrial
goods and transport services also include plan data-
roughly 430 products and eight transport services (pp.
125, 132, 161-195, 340). This is more data than was
provided in the last anniversary Narkhoz, for 1976,
which included substantial amounts of data on the
1977 plan.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Investment Data. A new table has been added which
gives capital investment in the "social-cultural" com-
plex-as distinct from nonproductive investment. Also
included is a definition for this complex-housing, the
communal economy, health, education, culture, art,
trade and public catering, urban passenger transport,
and everyday services for the population (Narkhoz
1986, p. 330).
Another new table gives the structure of "productive"
state capital investment by industrial complex (p.
105). In addition, it is now possible to back out a
number for investment in nonferrous metals using old
groupings of investment data for sectors of industry
given in Narkhoz 1985 and new groupings given in
Narkhoz 1986-roughly 2.6 billion rubles in 1984
prices (Narkhoz 1985, p. 368; Narkhoz 1986, p. 330).
Collective Farm Markets. Narkhoz 1986 includes
several new tables with information on collective farm
markets (CFMs~the number of such markets, their
amenities, the volume of their sales by product,
indexes of prices by product, and average CFM prices
relative to state retail prices. The material confirms
that, despite moves to improve CFM facilities, CFMs
are still little more than allotted space, often in open
areas (Narkhoz 1986, pp. 484, 485).
Regional Data. A considerable amount of new data is
provided on regions of the USSR. Of particular
interest are demographic data by union republic. New
tables of regional data include the following: distribu-
tion of urban centers by number of inhabitants (Nark-
hoz 1986, p. 376); population figures for the economic
regions of the USSR (p. 377); growth of industrial
GVO (gross value of output) by union republic and
economic region (p. 135); and life expectancy and
infant mortality by union republic (pp. 408, 409).
They also include: average money wages of workers
and employees by union republic (p. 434); average
wage of kolkhozniks by union republic (p. 435);
growth of real per-capita incomes by union republic
(p. 442); sales to the population of construction mate-
rials by union republic (p. 475); the number of CFMs
by union republic (p. 484); the number of state and
cooperative bookstores, drug stores, and stores selling
construction materials per 100,000 people by union
republic (pp. 490, 491); the number of self-service
stores by union republic (p. 492); and the volume of
paid services per inhabitant in 1986 by union republic
(p. 499).
Industrial Statistics. New information is included in
Narkhoz 1986 on overall production, the pace of
modernization, and the use of labor and capital in
industry. Some new material is available as well on
the production of consumer goods. The following new
tables are included: transition to two- and three- shift
work schedules (p. 137); shift-work coefficients for
industrial sectors (p. 138); the average use of calendar
time of industrial workers (p. 140); loss of work time
and turnover of workers in industry (p. 140); machin-
ery commissionings by industrial sector (p. 148); and
retirements of productive fixed industrial capital by
industrial sector (p. 149). Also wear and tear of
industrial productive fixed capital by sector of indus-
try (p. 152); capital repair by industrial sector (p. 153);
return on capital by industrial sector in different
periods (p. 155); use of productive capacity of industri-
al enterprises by output of various types of production
(p. 156); and the use and assimilation of projects
commissioned in 1981-85 (p. 157).
Labor Statistics. A variety of new information ap-
pears in the labor statistics section:
? Three new tables on worker participation in govern-
ment have been added (Narkhoz 1986, pp. 382,
385).
? Time and motion study data have been added for
cities and farms on running a household (pp. 427-
429).
? Two tables are included comparing sources and uses
of family income for workers in the oil and textile
industries and for peasants of four oblasts (pp. 446-
447). Aline item has been added to the tables on
family budgets which gives family expenditures on
alcohol (pp. 443-445). Information on average sala-
ries of workers of leading professions by sector of
the economy is also new (p. 432).
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
? Tables on the number of places in homes for the
elderly and the average size of pensions have been
added (p. 439).
? Tables on the educational level and age structure of
managers and specialists are new (p. 421).
? New information is provided on deaths of the
working age population by reason of death, fertility
by age group, deaths by age group, average life
expectancy by age group, and reasons for loss of
work time because of illness (pp. 405, 408, 595).
Deletion of Data
Meanwhile, some important data series have been
deleted from the 1986 handbook. Data are no longer
given for NIU and its components--consumption and
accumulation-in comparable 1973 prices. Without
these comparable price data, it is impossible to com-
pare real and current price growth of the components
of NIU in 1986. Also missing are production data on
trolleybuses, one of the few remaining statistics on
production of transport equipment. Nor does Narkhoz
1986 give an overall value for production of the
collection of consumer goods referred to as tovary-
durable consumer goods and housewares (Narkhoz
1985, p. 171). Some investment data are missing as
well. Statistics on investment by branch of industry
have been deleted.
Revision of Data
Some statistical series have been revised. Coal pro-
duction in standard fuel units was reduced by roughly
10 percent for 1985 and 1986, but production num-
bers for coal in tons for those years are unchanged,
the implication being that the quality of coal is
deteriorating. Data on cotton production have been
revised downward, confirming press reports that these
statistics were overstated during 1976-84. Soviet in-
vestment as a share of US investment in 1985 has
been revised downward from 100 percent in Narkhoz
1985 to 90 percent in Narkhoz 1986. Ratios of the
production of 25 products in CEMA countries relative
to the EEC have been reduced an average of 10
percent.
It is also worth noting that the price base of the table
on the "balance of fixed capital" has been changed
from "comparable 1973 prices" to unspecified "com-
parable prices." The capital stock data given for the
beginning of 1986 match well with data in Narkhoz
1985 (p. 51) for end of year 1985-specified as being
in 1973 prices. However, the value of commissionings
is 33 percent above its 1985 level. According to this
same table, retirements also are more than double
their reported level in 1985. Elsewhere in Narkhoz
1986, on page 104, data-show that the coefficient of
retirement of productive fixed capital only increased
from 1.9 in 1985 to 2.1 percent in 1986. The inconsis-
tencies of these trends suggest that the data on
commissionings and retirements may be in a different
price base than last year-presumably a price base
different from that used to value total fixed capital.
A footnote has been appended to the table of NIP in
current prices (Narkhoz 1986, p. 122). According to
this footnote, in 1986 net output of industry is calcu-
lated excluding subsidies. This exclusion makes the
figure for 1986 lower than the figure for 1985. This
implies that corresponding values before 1986 are
overstated and raises questions about how subsidies
are handled in the other economic sectors.
Finally, growth of total retail trade has been revised
downward from 6.4 percent in TsiJrakh 1986 to 6.3
percent in Narkhoz 1986. There is still a disconnect
between movements in retail prices reported in official
price indexes and the implicit price indexes calculated
by comparing growth of retail trade and its compo-
nents in current and comparable prices. Growth in
current and comparable prices move in different
directions for all these aggregates, and-for sales of
food and beverages and for alcohol-explicit and
implicit price indexes move in opposite directions. The
deflation implied by comparing current and compara-
ble price growth for retail trade is also implied by
comparison of data for NIU and NIP in current and
comparable prices. (See tables 1 to 4.)
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 1
Growth in Indicators of
Retail Sales of Alcohol
Average annual growth in percent Table 3
Changes in Official Price Indexes
for Retail Trade
Sales
Comparable
Prices
Sales
Quantities
(dekaliters)
Price
Index
1981-84
-0.3
0.7
4.7
1985
-14.5
-8.9
5.6
1986
-37.8
-34.7
24.4
Sources: Narkhoz 1985, pp. 459, 471, 478; Narkhoz 1986, pp. 451,
468, 480.
Table 2
Implicit Price Changes
in Retail Trade
Total
Trade
Food and
Beverages
Nonfood
Goods
1981
1.5
2.2
1.0
1982
3.4
4.2
3.0
1983
0.6
0.9
0.5
1984
-1.0
-0.3
-1.9
1985
-1.1
-2.1
-0.9
1986
-3.5
-5.7
-0.9
1987
1.6
NA
NA
Sources: Narkhoz 1983, p. 462; Narkhoz 1985, pp. 458, 469;
Narkhoz 1986, pp. 462, 464-465; Pravda, 18 October 1987
and 24 January 1988.
Other Sources for New Data
In addition to the new types of information contained
in the latest statistical handbook, some new data have
begun to appear in several other sources. The first of
these are press releases.. As of 11 September 1987,
280 such releases had been made.
The releases, distributed daily, range in size from one
to 15 pages and are intended for use by newspapers,
journals, radio, and TV. Some of the information is
unavailable elsewhere, but much of it is advance
release of information usually made available in
statistical handbooks or in reports on fulfillment of
Total
Trade
Food and
Beverages
Nonfood
Goods
1981
1.0
1.9
1.0
1982
3.8
3.8
2.9
1983
0.0
0.9
0.0
1984
-0.9
0.0
-1.9
1985
0.9
1.8
-1.0
1986
].9
5.4
0.0
1987
NA
NA
NA
Sources: Narkhoz 1984, p. 494; Narkhoz 1985,
p. 480; Narkhoz 1986, p. 482.
the state plans. Some of the data also eventually
appear in such journals as Vestnik statistiki and
Ekonomicheskaya gazeta.5
Examples of the kinds of new statistics that have been
reported in these releases include:
? The value of machinery meeting what the Soviets
define as "world" standards, which reportedly fell
from 18 percent in 1976-80 to 15 percent in 1981-85
and to 14 percent in 1986.
? Stocks of uninstalled equipment at construction
sites of civilian machine-building ministries. This
table provides data on both the total and above-
norm stocks of uninstalled equipment at the begin-
ning of 1986 and 1987.
? Expenditures for the introduction of new equipment
in *nachine building.
? Plan fulfillment data on production costs and profits
for each of the 11 civil machine-building ministries
and the machine-building complex.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 4
Comparison of Growth in Major
Aggregate Economic Indicators
Current Comparable Implicit Current
Prices Prices Price Prices
Deflators
Comparable
Prices
Implicit
Price
Deflators
Current
Prices
Comparable
Prices
Implicit
Price
Deflators
1981 5.3 3.1 2.1 5.2
3.2
1.9
5.7
4.1
1.5
1982 7.6 4.2 3.3 7.3
3.8
3.4
3.4
0.0
3.4
1983 4.7 4.0 0.7 4.6
3.6
1.0
3.4
2.8
0.6
1984 4.0 3.3 1.2 4.2
2.3
1.9
3.4
4.4
- 1.0
1985 1.4 3.2 -1.7 1.7
2.3
-0.7
2.6
3.7
-1.1
1986 1.5 4.1 -2.5 1.3
3.3
-1.9
2.4
6.1
-3.5
1987 2.1 2.3 -0.2 1.7
NA
NA
2.8
1.2
1.6
Sources: Narkhoz 1980, p. 429; Narkhoz 1983, pp. 37-38, 40,
Narkhoz 1985, pp. 39, 458; Narkhoz /986, pp. 122, 123, 430,
462;
462,
464-465; Tsz?/rakh 1985, p. 192; Tsilrakh 1987, p. 195; Pravda,
18 October 1987 and 24 January 1988.
The data in these press releases demonstrate the
ambiguities and inconsistencies that are pervasive in
Soviet statistical data. For example, one press release
(No. 269, p. 14) indicates that investment in the
machine-building complex increased by 14 percent in
1986; a second release (No. 4) claims an increase of 17
percent.
Other sources of new information this year are the
reports of monthly plan fulfillment published by
Goskomstat. In addition to growth in output for
individual industrial products, these reports indicate
how output relates to planned performance for pro-
duction of that product. Using this information in
conjunction with data on annual plans for the produc-
tion of individual products given in Tsifrakh 1986, it
is possible to aggregate these production plans and
look at their stability over time.
Prospects for the Release of More Data
Plans recently released by Goskomstat call for the
publication of additional statistics on the economy.
According to Nikolay Belov, first deputy chairman of
Goskomstat, the purpose of releasing more data is to
increase the effectiveness of economic research in the
USSR. To facilitate the release of more information,
a Goskomstat press center-the Information Publica-
tion Center-has been created. This center is to
become the organizational focal point of all "informa-
tion propaganda activity" in the area of statistics.
Center director Leonid Umanskiy outlined publica-
tion plans for 1988 in an article in the October 1987
issue of Vestnik statistiki: 6
? Goskomstat plans to give wide circulation to statisti-
cal collections. In 1988 Goskomstat will publish-in
addition to the statistical yearbook Narkhoz 1987-
10 sectoral and thematic statistical handbooks
(collections).'
? In addition to daily press releases, a press bulletin
will be published three times a month containing
economic results and data worked up by the statisti-
cal agency-i.e., summary measures and
aggregations.
6 See "Improving Work on Statistics" in Vestnik statistiki, No. ] 0,
1987, p. 38.
' According to Belov, the following are among the statistical
collections planned: Agriculture USSR, Capital Construction
USSR, Labor in the USSR, Consumer Goods, and Population
USSR.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
? Along with statistics on the fulfillment of state plans
at the quarter, half year, nine months, and yearend,
data on fulfillment of the 12th Five-Year Plan will
be published.
? Short statistical collections with commentary on
important questions that interest the public and
information gained from special social/economic
surveys on the most important problems of the
development of the USSR will also be released.
? Finally, separate materials on important one-time
work-selective investigations carried out by statis-
tical organs-will be made available.
There also are indications that Moscow may be
intending to release more defense-related data. In a
speech to the UN Conference on Disarmament and
Development in August 1987, Soviet Deputy Foreign
Minister Vladimir Petrovskiy disclosed the composi-
tion of the official Soviet defense budget.8 Petrovskiy
hinted that the USSR might be more forthcoming
with information on defense spending than in the past,
stating that it would be possible to "compare overall
military spending realistically" once the Soviets have
implemented a price reform. If released, such data
probably would be disclosed to the UN and be tied to
revived Soviet proposals for international agreements
to freeze or reduce military expenditures. A UN
experts group established to study the feasibility of
the limitation of military expenditures has designed a
standard format for the reporting of military spend-
ing.
In an article in the No. 36 issue of Moscow News, the
author reported that the USSR had tabled a plan at a
UN Conference to transfer resources spent on prepa-
rations for war to Third World development. Accord-
ing to the author, the USSR is "prepared to publish,
e Petrovskiy announced that the defense budget includes Ministry
of Defense expenses for maintaining military personnel, military
pensions, logistics, military construction, and "a number of other
articles." According to Petrovskiy, military research and develop-
ment and weapons procurement are included in other parts of the
USSR state budget.
for the sake of strengthening mutual confidence, not
only the defense budget figures directly connected
with expenditures by the USSR Defense Ministry but
also those that are connected with the financing of
research and development work and with the purchase
of arms and military hardware."
In an article in Pravda on 17 September 1987,
Gorbachev himself stated that "within the next two to
three years we will be able to compare the figures that
are of interest to us and our partners and which would
symmetrically reflect the expenditures of the sides."
Recent statements on this subject also were made by
Georgiy Arbatov, director of the USA and Canada
Institute of the USSR Academy of Sciences, and
Marshal Sergey F. Akhromeyev, chief of the Soviet
General Staff. In an interview in the Japanese publi-
cation Yomiuri Shimbun, Arbatov said that "the
Soviet TJnion will make the components of its defense
outlays clear and in a couple of years bring them to a
level where they can be compared with the United
States. The cost for assembling tanks and the prices of
metals or energy are cheaper in the Soviet Union and,
therefore, this entails complicated work to make them
comparable. But this is necessary-not only to the
work, but also to the Soviet Union."
Marshal Akhromeyev also touched on the subject in
an interview in the 30 October 1987 issue of the New
York Times. In response to a written question, he
stated that the Kremlin will make its first public
accounting of its military budget in two to three years.
He also reaffirmed that the budget figure for military
spending-20.2 billion current rubles in 1987-re-
flects only personnel, pensions, training, and logistics
and does not include money for development and
acquisition of weapons.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Appendix A
Data Publication:
How Decisions Are Made
Although the workings of Goskomstat are generally
surrounded by secrecy, some information is available
on how decisions are made concerning the publication
of economic statistics. In some cases, information
disappears from multiple sources simultaneously, as it
did in 1977-78. This is probably the result of an
addendum to the perechen', a register of material
restricted from publication. This list is updated regu-
larly by Goskomstat based on decisions made at very
high levels. In less sensitive cases, data not listed on
the perechen' are dropped from the Narkhoz. The
difference is that these data remain available in other
sources.
Although changes in the perechen' explain some of the
broader decisions to remove data from Narkhoz and
all other statistical handbooks at once, many changes
in the Narkhoz have been small and limited to
specific sections. Many of the sections are monitored
by different departments within Goskomstat and each
sector and subsector seems to have had its own unique
history. This suggests that a lot of lower level deci-
sionmaking within Goskomstat has also affected the
composition of the handbook. These departments and
regional Goskomstats apparently have considerable
leeway in interpreting what information is "detrimen-
tal to the state" if it is not specifically proscribed in
the perechen'.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Appendix B
Goskomstat Press Releases,
January-September 1987
No. 4, 21 January, eight pages, Capital Construction
1. Fulfillment of the Plan for Commissioning of
Fixed Capital and Use of the Limit of State
Capital Investment by Various Complexes in
1986-percent.
No. 27, 19 February, one page, Changes in the
Administrative-Territorial Structure of the Units-
1923, 1947, 1957, 1977, 1987.
No. 30, 24 February, one page, Production of Goods
Jor Children and Youth-units, 1985, 1986.
2. 1986 Commissionings of Projects To Be No. 32, 24 February, one page, Deliveries to Trade of
Turned Over According to the Nomenclature of - Wood and Construction Materials in January 1987-
the State Plan by Ordering Ministries-units. units.
3. Fulfillment of an Established Volume of
Contract Work on the Most Important Con-
struction Projects of the Economic Complexes in
1986-percent.
4. Fulfillment of the Established Volume of
Contract Work and Tasks for Growth of Labor
Productivity in 1986-percent.
5. Fulfillment of the Plan for Capital Construc-
tion for the Agroindustrial Complex Using State
Funds in 1986-rubles, percent.
6. Commissionings of the Most Important Pro-
ductive Capacities for the Agroindustrial Com-
plex in 1986-units, percent.
7. Measures for Environmental Protection and
Rational Use of Natural Resources-rubles,
percent, 1986.
No. 11, 4 February, four pages, For 70 Years of
Soviet Power
1. Various Indicators About Growth of the
Material Well-Being and Cultural Level of the
People, 1913-1986 (includes 2 pages of text).
No. 34, 24 February, one page, Stocks o,J' Uninstalled
Equipment in Warehouseslor Capital Construction
ollndividual Ministries-rubles, 1985, 1986.
No. 35, 4 March, 13 pages, Statistical Materials,for
Discussion of the Drc~t Law of the USSR on State
Enterprises
1. Productive and Scientific-Productive Associa-
tions, Combines, and Enterprises on Indepen-
dent Balances by Individual Sectors of Industry
in 1985-units.
2. Productive and Scientific-Productive Associa-
tions in Industry-units, 1970, 1975, 1980,
1985, 1986.
3. Scientific-Productive Associations in Indus-
try-units, 1973, 1975, 1980, 1985, 1986.
4. Industry GVO (enterprise wholesale prices of
the corresponding year)--rubles, 1970, 1975,
1980, 1985, 1986.
5. Annual Average Number of Industrial-Pro-
duction Personnel-units.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
6. Productive Fixed Capital in Industry (compa-
rable 1973 prices, at yearend}-rubles, 1970,
1975, 1980, 1985, 1986 estimate.
7. Number of Enterprises and Organization of
the APK at Yearend 1986-units.
8. Basic Indicators of Kolkhozes (less fishing
cooperatives}-unit and ruble measures, 1970,
1975, 1980, 1985, 1986.
9. Average Size of Kolkhozes (less fishing coop-
eratives, per farms-unit and ruble measures,
1970, 1975, 1980, 1985, 1986.
10. Basic Indicators of Sovkhozes-unit and
ruble measures, 1970, 1975, 1980, 1986.
11. Average Size of Sovkhozes (per farm unit
and ruble measures, 1970, 1975, 1980, 1985,
1986.
12. Industrial Enterprises of the APK in 1985-
units, rubles.
13. Enterprises of the Food Industry-units,
rubles, 1970, 1975, 1980, 1985.
14. Enterprises of Retail Trade and Public
Catering by State and Cooperative Organiza-
tions and of Everyday Services for the People-
units, rubles, 1970, 1975, 1980, 1985.
15. Number of Scientific Establishments (at
yearend units, 1970, 1975, 1980, 1985.
No. 36, 4 February, two pages, Women in the Nation-
al Economy
1. Women in the Economy by Sector-units,
1922, 1940, 1960, 1980, 1986.
2. Number of Women Specialists With Higher
and Secondary Specialized Education Working
in the National Economy-units, 1941, 1960,
1980, 1986.
3. Number of Women Workers and Employees
by Union Republic-units, Y 922, 1940, 1960,
1980, 1986.
No. 38, 4 March, one page, About Collective Fruit
and Vegetable Gardens (according to data oja survey
of 5,000 families of workers and employees, living in
urban areas in 1986)
1. Distribution of Families of Urban Workers
and Employees Having Fruit or Vegetable Plots
According to the Size of Their Plots-percent.
2. Output and Disposition of Farm Products
From Fruit and Vegetable Plots-units, percent.
No. 42, 10 March, three pages, Basic Indicators of
the Work of Urban Electrical Transport in 1986
Tables
1. Tramways-units.
2. Trolleys-units.
3. Subways-units.
No. 48, 18 March, two pages, About Production of
Sporting and Tourism Goods-units, percent, 1986.
No. 53, 24 March, two pages, Statistical Materials
for Geology Day
1. Volume of Geological Exploratory Work (at
the expense of state budget funds and capital
investment in comparable 1985 prices rubles,
1980, 1985, 1986.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
2. Deep Exploratory Drilling for Oil and Gas-
units, 1980, 1985, 1986.
3. Basic Technical-Economic Indicators for the
Ministry of Geology SSSR for January-Febru-
ary 1987-rubles, percent.
No. 54, 24 March, one page, Statistical Materials
For Discussion of the Drajt Law on State Enterprises
(Association)-text.
1. Number of Workers and Employees by En-
terprises and Organizations of Ministries and
Departments, Working in Conditions of Full
Khozraschet and Self-Financing in 1987-
units, percent.
No. 55, 24 March, one page, About the Rhythm of
Construction-text.
No. 58, 24 March, one page, Results of Verification
ojthe Quality ojGoods at Enterprises ojthe Ministry
ojLight Industry ojthe USSR by Wholesale Organi-
zations or the Ministries of Trade of the Union
Republics-units, percent, 1985, 1986.
No. 62, 27 March, two pages, Forestry in the USSR
in 1986
Tables
1. Forests, 1 January-1978, 1983.
2. Forestry Work-units, 1980, 1985, 1986.
No. 66, 27 March, three pages, Results ojthe
Development of the Economy ollndividual Socialist
Countries in 1986-percent.
No. 75, 10 April, one page, Production ojCotton in
the USSR and Several Foreign Countries-units, per
capita.
No. 79, 15 April, one page, Number ojForeigners
Studying in Higher and Secondary Specialized Edu-
cational Institutions in the USSR
1. Number of Foreigners Studying in Higher
and Secondary Specialized Educational Insitu-
tions in the USSR-units, 1975, 1980, 1985,
1986.
2. Study of Foreign Language in Educational
Institutions of the Country in 1986-units.
No. 88, 27 April, two pages, Number ojDays of
Temporary Disability jor Various Reasons (per 100
workers)-1980, 1985, 1986.
No. 89, 27 April, 12 pages, For 70 Years ojGreat
October: Science and Technical Progress in the
USSR-text.
1. Number of Scientific Institutions-units,
1913, 1940, 1960, 1970, 1980, 1985, 1986.
2. Number of Scientific Workers-units, 1960,
1970, 1980, 1985, 1986.
3. Expenditures on Science From the State
Budget and Other Sources-rubles, 1940, 1960,
1970, 1980, 1985, 1986 (plan).
4. Number of Discoveries, Registered in the
State Register of Discoveries of the USSR-
units, 1970, 1975, 1980-86.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
5. Renewal of the Output of Machine Build-
ing-percent, 1970, 1975, 1980, 1985, 1986.
6. Creation of the First Models in the USSR of
Machines, Equipment, Instruments, and Means
of Automation-units, 1971-75, 1976-80, 1981-
85, 1986.
7. Assimilation of New Types of Industrial
Output-units, 1971-75, 1976-80, 1981-85,
1986.
8. Removal From Production of Obsolete De-
signs of Machinery, Equipment, Apparati, In-
struments, Means of Automation, and Articles
of Machinery-units, 1971-75, 1976-80, 1981-
85, 1986.
9. Mechanization and Automation of Productive
Processes in Industry-units, 1971, 1975, 1981,
1985.
10. Installation of Automated Systems of Man-
agement and Information-average annual
units, 1971-75, 1976-80, 1981-85.
11. Expenditures for Introduction of Measures
for New Technology in Industry and Their
Economic Effect-rubles, 1971-75, 1976-80,
1981-8'S.
12. Inventions and Rationalization Proposals in
the Economy-units, rubles, 1971-75, 1976-80,
1981-85, 1986.
13. Production of the Most Important Types of
Progressive, Highly Effective Output of the
Fuel-Energy and Metallurgy Complexes-units,
1980, 1985, 1986, 1987 (plan).
14. Production of the Most Important Types of
Progressive Highly Effective Output of Machine
Building-units, 1980, 1985, 1986, 1987 (plan).
15. The Most Important Types of Progressive,
Highly Effective Output in the Chemical-Forest
Complex and in the Construction Materials
Industry-units, 1980, 1985, 1986, 1987 (plan).
No. 99, 7 May, eight pages, Resource Savings
1. Average Annual Growth of Labor or Produc-
tivity-percent, 1961-65, 1966-70, 1971-75,
1976-80, 1981-85, 1986.
2. Growth of Output and Work Received on
Account of an Increase in Labor Productivity-
percent, 1971-75, 1976-80, 1981, 1985, 1986.
3. Correlation of Growth of Wages and Labor
Productivity for Individual Sectors of the Econ-
omy-percent, 1981-85, 1986.
4. Number of Brigades and Number of Workers
in Them in Individual Sectors of the Economy
in 1985-units, percent.
5. Mechanization of Labor (correlation of num-
ber of workers occupied in mechanized and
manual labor; according to data of a one-time
inventory, in percent)--Industry, Agriculture,
and Construction-1975, 1982, 1985.
6. Average Annual Growth of Produced Nation-
al Income, Productive Fixed Capital, and Re-
turn on Capital-percent, 1971-75, 1976-80,
1981-85, 1986.
7. Lowering of the Prime Cost of Output in
Industry and Construction-percent, 1971-75,
1976-80, 1981-85, 1986.
8. Average Annual Rate of Reduction of Mate-
rial Expenditures-percent, 1971-75, 1976-80,
1981-85, 1986.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
No. 100, 7 May, seven pages, Introduction of Highly
Productive Equipment, Expenditures,for the Intro-
duction of Measures for New Technology, and the
Economic F~/fectiveness of increasing the Technologi-
cal Level o1'Production
1. Introduction of Highly Productive Equipment
in Industry-units, 1986, 1987-first quarter.
2. Expenditures for the Introduction of Mea-
sures for New Technology in Industry and Their
Economic Effectiveness-units, rubles, 1981-85,
1985, 1986.
3. Expenditures for the Introduction of Mea-
sures for New Technology in the Fuel-Energy
Complex and Their Economic Effectiveness-
units, rubles, 1981-85, 1985, 1986.
4. Expenditures for the Introduction of Mea-
sures for New Technology in the Metallurgical
Complex and Their Economic Effectiveness-
units, rubles, 1981-85, 1985, 1986.
5. Expenditures for the Introduction of Mea-
sures for New Technology in the Machine-
Building Complex and Their Economic Effec-
tiveness-units, rubles, 1981-85, 1985, 1986.
6. Expenditures for the Introduction of Mea-
sures for New Technology in the Chemical-
Forest Complex-units, rubles, 1981-85, 1985,
1986.
7. Expenditures for the Introduction of Mea-
sures for New Technology in the APK and
Their Economic Effectiveness-units, rubles,
1981-85, 1985, 1986.
No. 106, 13 May, eight pages, Housing and Social-
Cultural Construction
1. Commissionings of Housing (million square
meters of total [useful] space)-1918-40, 1956-
60, 1961-65, 1966-70, 1971-75, 1976-80, 1981-
85, 1986.
2. Commissionings of Housing in Rural Areas
(million square meters of total [useful] area}-
1961-65, 1966-70, 1971-75, 1976-80, 1981-85,
1986.
3. Housing Construction on the Account of
State Capital Investment-units, 1961-65,
1966-70, 1971-75, 1976-80, 1981-85, 1986.
4. Commissionings of Housing Built by the
Population and Housing-Construction Coopera-
tives (million square meters of total [useful)
area}--1961-65, 1966-70, 1971-75, 1976-80,
1981-85, 1986.
5. Number of Apartments Built and the Num-
ber of People Improving Their Living Condi-
tions-units, 1961-65, 1966-70, 1971-75, 1976-
80, 1981-85, 1986.
6. Commissioning of Projects of Social-Cultural
Significance-places, 1971-75, 1976-80, 1981-
85, 1986.
7. Commissioning of Projects of Social-Cultural
Significance in Rural Areas-places, 1971-75,
1976-80, 1981-85, 1986.
8. Commissionings of Hospital and Ambulatory-
Polyclinic Establishments-beds, 1971-75,
1976-80,1981-85,1986.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
No. 113, 14 May, two pages, About Fulfillment of the
Planfor Prime Cost of Output and the Financial
Results of the Work of Industry in the First Quarter
of 1987
1. Fulfillment of the Plan for Prime Cost and
Profit in First Quarter 1987-percent, gives
component ministries for each complex.
No. 118, 19 May, one page, Deliveries to Trade of
Wood and Construction Materials-units, Jan-Feb
1987.
No. 135, 28 May, one page, Use of Established Work
Time by Individual Sectors of the Economy (in
average for one worker; days)-1985, 1986.
No. 137, 28 May, one page, Changes in the Popula-
tion of the Economic Regions of the RSFSR-units,
1976, 1981.
No. 138, 28 May, one page, Materials for Interna-
tional Protect the Children Day
1. Preschool Establishments-units, places, stu-
dents, 1980, 1985, 1986.
No. 120, 19 May, one page, Sales and Prices of Farm
Products on City Markets (according to 264 citiesJ-
units, rubles, percent, January-April 1987.
No. 130, 25 May, five pages, About Preservation of
the Natural Environment of Lake Baikal
1. Disposal of Waste Water in the Basin of Lake
Baikal-units, 1980, 1985, 1986.
2. Disposal of Insufficiently Cleaned Waste
Water in the Basin of Baikal by Enterprises-
units, 1985, 1986.
3. The Pace of Construction of Basic Water
Preservation Projects in the Basin of Lake Bai-
kal (on 1 January 1987) by Project-rubles,
percent.
4. Characteristics of Air Preservation Activities
in the Region of Lake Baikal-units, 1984,
1985, 1986.
No. 134, 28 May, one page, Structure of Workers
and Employees Diverted (Attracted) From Their Ba-
sic Jobs-percent, 1985, 1986.
2. General Education Schools-units, students,
1980-81, 1985-86, 1986-87.
3. Schools With Extended Days-units, stu-
dents, 1980-81, 1985-86, 1986-87.
4. School Libraries for Children-units, books,
users, 1980-81, 1985-86, 1986-87.
5. Children's Library System, Ministry of Cul-
ture of the USSR-units, holdings, use, 1980,
1985, 1986.
6. Children's Out-of-School Establishments-
units, users, 1980, 1985, 1986.
7. Study Groups in Children's Out-of-School
Establishments-units, users, 1980, 1985, 1986.
8. Study Groups for Children in the Club
Establishments of the Ministry of Culture of the
USSR and Trade Union Organizations-units,
users, 1980, 1985, 1986.
9. Study Groups in General Education Day
Schools-units, users, 1980-81, 1984-85, 1985-
86.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
10. Movie Showings and Movie Establishments
of Goskino and VTsSPS-units, users, 1980,
1985, 1986.
11. Children's Theaters-units, users, 1980,
1985, 1986.
12. Children and Youth Sports Schools-units,
students, 1980, 1985, 1986.
13. Number of Children and Youth Attending
Children's Health Resorts-users, 1980, 1985,
1986.
No. 139, 1 June, two pages, Economic Cadres of the
Country
No. 140, 1 June, 15 pages, For 70 Years of Great
October-The Development ollndustry
1. Growth of Industry GVO for 1917-87-
(1917=1), 1940, 1950, 1960, 1970, 1980, 1985,
1986, 1987 (plan).
2. Growth of Industry GVO for 1940-87-
(1940=1), 1960, 1970; 1980, 1985, 1986, 1987
(plan).
3. Growth of Industry GVO by Sector for the
Period 1940-861940=1), 1960, 1970, 1980,
1985, 1986.
4. Growth of Labor Productivity by Sector of
Industry for the Period 1940-851940=1),
1960, 1970, 1980, 1985, 1986.
5. Increase of Industrial Output Due to the
Growth of Labor Productivity by FYP-per-
cent, 1971-75, 1976-80, 1981-85, 1986.
6. Basic Directives for Economic and Social
Development 1986-90 and Until the Year
2000-text.
7. Production of the Most Important Types of
Industrial Output in Physical Units-1917,
1940, 1945, 1950, 1965, 1970, 1975, 1980,
1985, 1986, 1987 (plan).
8. Growth of Industry GVO by Union Repub-
lic-{1940=1), 1960, 1970, 1980, 1985, 1986,
1987 (plan).
9. Brief Discussion of the Economies of Each
Republic-text.
No. 146, 5 June, six pages, Materials for Land
Improvers' Day
1. Presence and Use of Irrigation and Drainage
of Agricultural Areas on Kolkhozes, Sovkhozes,
Intersectoral, and Other Productive Agricultur-
al Enterprises-units, 1970, 1980, 1985, 1986.
2. Area of Irrigated Land-units, 1970, 1980,
1985, 1986.
3. Use of Irrigated Areas-units, 1970, 1980,
1985, 1986.
4. Sowing Area of Agricultural Crops on Irri-
gated Land-units, 1970, 1980, 1985, 1986.
5. Gross Harvest and Yield of Agricultural
Crops on Irrigated Land for Four Crops-units,
1971-75, 1976-80, 1981-85, 1986.
6. Area of Drained Land-units, 1970, 1980,
1985, 1986.
7. Use of Drained Agricultural Land on Kol-
khozes, Sovkhozes, Interfarm and Other Pro-
ductive Agricultural Enterprises-units, 1970,
1980, 1985, 1986.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
8. Sowing Area of Agricultural Crops on
Drained Land for Four Crops-units, 1970,
1980, 1985, 1986.
9. Gross Harvest and Yield of Agricultural
Crops on Drained Land for Four Crops-units,
1971-75, 1976-80, 1981-85, 1986.
10. Gross Output of Crops on Irrigated and
Drained Lands on Kolkhozes, Sovkhozes, Inter-
farm and Other Productive Agricultural Enter-
prises (comparable 1983 prices}-rubles, per-
cent, 1970, 1971-75, 1976-80, 1981-85, 1986.
11. Amelioration Construction on Account of
State Capital Investment and Kolkhoz Means-
units, rubles, 1971-75, 1976-80, 1981-85, 1986.
No. 158, 17 June, one page, Sales and Prices ojFarm
Product in City Markets (264 cities)-units, percent,
January-May 1987
No. 161, 18 June, one page, About the Certification
of Work Places-units, 1986
No. 166, 22 June, four pages, For Youth Day
1. Number of Young People Under 30 Years of
Age Working in the National Economy (Includ-
ing Kolkhozes) on 15 November 1985-by edu-
cation by union republic, percent.
2. Number of Managers and Specialists Under
30 Years of Age on 1 November 1985-by
occupation, units, percent.
3. Job Placement and Further Study of Young
People According to a One-Time Survey on 1
December-percent, 1974, 1981, 1986.
No. 167, 22 June, five pages, Inventions and Ratio-
nalizing Proposals in the Economy
1. Total Indicators-units, rubles, 1971-75,
1976-80, 1981-85, 1986.
2. Basic Indicators of Invention in the Econo-
my-units, percent, rubles, 1971-75, 1976-80,
1981-85; 1986.
3. Basic Indicators of Rationalizing Work in the
Economy-units, 1971=75, 1976-80, 1981-85,
1986.
4. Inventions and Rationalization in Industry-
units, rubles, 1971-75, 1976-80, 1981-85, 1986.
5. Invention and Rationalization in Agricul-
ture-units, rubles, 1971-75, 1976-80, 1981-85,
1986.
No. 168, 23 June, one page, Average Wage by Sector
and Average Bonus Size
No. 177, 29 June, nine pages, Scientific Potential
Tables
1. Number of Scientific Institutions (at Year-
end)---units, 1970, 1975, 1980, 1985, 1986.
2. Number of Scientific Institutions by Union
Republic (at Yearend}--units, 1970, 1975,
1980, 1985, 1986.
3. Distribution of Scientific Institutions by Sec-
tor of the Economy (ati Yearend)---units, 1975,
1980, 1985, 1986.
4. Number of Scientific Workers (at Yearend
units, 1970, 1975, 1980, 1985, 1986.
5. Distribution of Scientific Workers (at End
1986) by Sector of Science-units.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
6. Number of Graduates With Advanced De-
grees-persons, 1970, 1975, 1980, 1985, 1986.
7. Graduates With Advanced Degrees by Sector
of Science-persons, 1970, 1975, 1980, 1985,
1986.
8. Average Annual Number of Workers and
Employees Working in Science and Scientific
Service-persons, 1970, 1975, 1980, 1985,
1986.
9. Creation in the USSR of First Models of
New Types of Machines, Equipment, Appara-
tus, and Means of Automation-units, percent,
1976-80, 1981-85, 1986.
No. 178, 29 June, three pages, Supply of Consumer
Durables to the Rural and Urban Population
1. Supply of Urban and Rural Populations With
Consumer Durables (Per 100 Families at Year-
end}-units, 1980, 1985, 1986.
2. Supply of the Urban Population With Con-
sumer Durables (Per 100 Families at Year-
end~units, 1980, 1985, 1986.
3. Supply of the Rural Population With Con-
sumer Durables (Per 100 Families at Year-
end)-units, 1980, 1985, 1986.
No. 179, 1 July, one page, Average Supply ojHousing
to the Population by Union Republic
No. 188, 9 July, nine pages, Materials jor Trade
Workers
1. State and Cooperative Retail Trade-rubles,
percent, 1970, 1980, 1985, 1986.
2. State and Cooperative Retail Trade (Per
Capita}-rubles, 1970, 1980, 1985, 1986.
3. Retail Trade and Public Catering Enterprises
of State and Cooperative Organizations (at
Yearend units, 1970, 1980, 1985, 1986.
4. Trade Areas in Stores and the Number of
Places in Public Catering Enterprises (Per
10,000 persons at Yearend-units, 1970, 1980,
1985, 1986.
5. Number of Self-Service Stores in State and
Cooperative Trade-units, 1970, 1980, 1985,
1986.
6. Grouping of Enterprises of Public Catering of
State and Cooperative Organizations by Loca-
tion-units, 1970, 1980, 1985.
7. The Provision to Various Contingents of the
Population With Places in Enterprises of Public
Catering-units, percent, 1980, 1985, 1986.
8. Material-Technical Base of State and Coop-
erative Trade for Preserving Potatoes, Vegeta-
bles, and Fruits (at Yearend units, capacity,
1970, 1980, 1985, 1986.
9. General Warehouses and Refrigerators in
Retail and Wholesale Trade (at Yearend)-
1970, 1980, 1985, 1986.
No. 195, 16 July, 13 pages, The Agroindustrial
Complex of the USSR
1. Role of the Agroindustrial Complex in Accel-
erating Progress and Raising the People's Well-
Being-text.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
2. Output, Productive Fixed Capital, and Num-
ber of Persons Working in the APK Complex-
rubles, persons, 1970, 1975, 1980, 1985, 1986.
3. Number of APK Enterprises and Organiza-
tions-at yearend 1986, units.
4. Industrial Enterprises of the APK in 1986-
units, rubles, fixed productive capital.
5. Enterprises of the Food Industry-units, ru-
bles, fixed productive capital, 1970, 1975, 1980,
1985, 1986.
6. Basic Indicators of the Development of Agri-
culture in the USSR-1913, 1940, 1950, 1960,
1970, 1980, 1985, 1986.
7. Gross Output of Agriculture per 100 Hect-
ares of Agricultural Land by Union Republic-
rubles, 1960, 1970, 1980, 1985, 1986.
8. Purchases of Livestock by Type and Nutri-
tional State at All Categories of Farms (With-
out Additional Weight) for the USSR-units,
1960, 1970, 1975, 1980, 1985, 1986.
9. Quality of Milk Sold to the State-percent,
1975, 1980, 1985, 1986.
10. Economic-Technical Indicators of Work by
Enterprises of Sugar and the Oil-Fats Industry
of the USSR-percent, 1940, 1960, 1970, 1975,
1980, 1985, 1986.
11. Growth of Agricultural Gross Output by
Union Republic-(1913 =1), 1940, 1960, 1970,
1980, 1985, 1986.
12. Indicators of the. Use of Irrigated and
Drained Agricultural Land on Kolkhozes, Sov-
khozes, Interfarm and Other Productive Agri-
cultural Enterprises-percent, rubles, 1971-75,
1976-80, 1981-85, 1986.
No. 199, 17 July, two pages, State Capital Construc-
tion in the Social Sphere, First Half 1987-text,
pinpoints laggards.
No. 207, 28 July, seven pages, For Railworkers Day
1. Basic Indicators of Work of Rail Transport
for General Use-units, 1940, 1960, 1970,
1980, 1985, 1986.
2. Number of Brigades and Number of Workers
in Them in General Purpose Rail Transporta-
tion-units, 1940, 1960, 1970, 1980, 1985,
1986.
3. Density of the Network of Railroads and
Density of Freight and Passenger Transport on
Railroads of the MPS SSSR-units, 1940,
1960, 1970, 1980, 1985, 1986.
4. Use of the Rail Lines of the-MPS SSSR-
units, 1940, 1960, 1970, 1980, 1985, 1986.
5. Meeting Train Schedules-percent, 1980,
1981, 1982, 1983, 1984, 1985, 1986.
6. VUZ Graduates Specializing in Rail Trans-
portation-persons, 1985, 1986.
7. Secondary School Graduates Specializing in
Rail Transportation-persons, 1985, 1986.
8. Sending of Freight by MPS Rail Transport in
- the First Half of 1987-percent of plan and of
1986 level.
No. 208, 28 July, one page, Public Catering Coopera-
tive, 1 July 1987
No. 212, 30 July, 12 pages, Transport and Communi-
cation in the USSR
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
1. Basic Indicators of the Development of All
Types of Transportation-units, 1913, 1917,
1940, 1950, 1960; 1980, 1985, 1986.
2. Basic Indicators of the Development of Rail
Transportation of the MPS SSSR-units, 1913,
1917, 1940, 1945, 1960, 1980, 1985, 1986.
3. Basic Indicators of the Development of River
Transportation-units, 1913, 1917, 1940, 1960,
. 1980, 1985, 1986.
4. Basic Indicators of the Development of High-
way Transport-units (kilometers, ton-kilome-
ters, tons, persons) 1913, 1917, 1940, 1980,
1985, 1986.
5. Basic Indicators of the Development of Air
Transport of the Ministry of Civil Aviations of
the USSR-units, 1928 1940, 1945, 1960, 1980,
1985, 1986, 1987 (plan).
6. Basic Indicators of the Development of Mari-
time Transport of the Ministry of the Maritime
Fleet of the USSR-units (ton-miles, tons, pas-
senger-miles, persons) 1913, 1917, 1940, 1960,
1980, 1985, 1986.
7. Basic Indicators of the Development of Oil
and Oil Product Transport-units, 1913, 1917,
1940, 1960, 1980, 1985, 1986.
8. Basic Indicators of Development of Gas
Transport-units, 1950, 1960, 1970, 1980,
1985, 1986.
9. Basic Indicators of the Development of Com-
munication by the Ministry-units, 1913, 1917,
1940, 1960, 1980, 1985, 1986.
10. Development of Radio and Television-
units, 1940, 1960, 1970, 1980, 1985, 1986.
No. 213, 30 July, five pages, For Physical Education
Day
1. Physical Culture and Sport-units, 1980,
1985, 1986.
2. Number of Sports Facilities-units, 1980,
1985, 1986.
3. Supply to the Population of Sports Facilities
by Union Republic-persons, units per 10,000
persons, 1985, 1986:
4. VUZ Graduates Specializing in Physical
Culture and Sport by Union Republic-persons,
1985, 1986.
5. Specialized Secondary School Graduates Spe-
cializing in Physical Culture-persons, percent,
1985, 1986.
6. VUZ Graduates Specializing in Physical
Culture by Specialization (includes initial mili-
tary instruction and physical education)---per-
sons, percent, 1985-86.
No. 214, 3 August, 10 pages, For Builders Day
1. Commissionings of Fixed Capital-rubles,
1961-65, 1966-70, 1971-75, 1976-80, 1981-85,
1986.
2. Number of Contract Construction and Instal-
lation Organizations-units, 1960, 1970, 1980,
1985, 1986.
3. Growth of Labor Productivity in Construc-
tion-index, 1960, 1970, 1980, 1985, 1986.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
4. Number and Average Size of Newly Con-
structed Apartments-units, 1961-65, 1966-70,
1971-75, 1976-80, 1981-85, 1986.
5. Commissionings of Projects of Soci-Cultural
Significance-units, 1961-65, 1966-70, 1971-
75, 1976-80, 1981-85, 1986.
6. Capital Construction in First Half 1987-
text.
7. Specialists in Construction Graduating From
VUZ-persons, percent, 1985, 1986.
8. Specialists in Construction Graduating From
VUZ by Union Republic-persons, percent,
1985, 1986.
9. Secondary Specialized School Graduates
With Specialities in Construction-persons, per-
cent, 1985, 1986.
10. Secondary Specialized School Graduates
With Specialities in Construction by Union
Republic-persons, percent, 1985, 1986.
5. Age-Specific Birth Rates-children per 1,000
women, 1969-70, 1974-75, 1979-80, 1981-82,
1983-84, 1985-86.
6. Age-Specific Death Rates-deaths per 1,000
persons, 1896-97, 1958-59, 1969-70, 1978-79,
1982-83, 1984-85, 1985-86.
7. Deaths of the Working Age Population by
Reason of Death-deaths per 100,000 persons,
1970, 1980, 1985, 1986.
8. Infant Mortality by Union Republic-deaths
per 1,000 births, under age 1, 1970, 1980, 1985,
1986.
9. Average Life Expectancy for Men and Wom-
en-years, 1896-97, 1926-27, 1938-39, 1955-56,
1958-59, 1971-72, 1978-79, 1983-84, 1984-85,
1985-86, 1984, 1985, 1986.
10. Average Life Expectancy for Men and
Women by Union Republic-year, 1969-70,
1979-80, 1985-86. Accents improvement since
1984.
No. 222, 10 August, 10 pages, Vital Statistics (Basic No. 233, 13 August, three pages, National Income
Indicators of the Reproduction of'the Population) and Labor Productivity by Union Republic
1. General Coefficients of Fertility, Death, and
Natural Growth of the Population by Union
Republic-per thousand persons, 1970, 1980,
1985, 1986.
2. Summary Coefficients of Fertility in the
USSR-total, city, country, children, 1938-86.
3. Summary Coefficients of Fertility by Union
Republic-total, city, country, children, 1969-
70, 1985-86.
1. Growth of National Income Produced,
1970=100-total and per capita, by union re-
public, index, 1980, 1985, 1986.
2. Growth of National Income on Account of an
Increase in Labor Productivity by Union Re-
public-percent, 1971-75, 1976-80, 1981-1985,
1986.
3. Growth of Social Labor Productivity,
1970=100-index, 1980, 1985, 1986.
4. Distribution of Births by Birth Order-chil-
dren, birth order 1-5, 1970, 1975, 1980, 1981,
1982, 1983, 1984, 1985, 1986.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
No. 247, 24 August, seven pages, For Educators Day
1. Number of Educational Institutions by
Type-units, 1970-71, 1975-76, 1980-81, 1985-
86, 1986-87.
2. Wage Data for Teachers at Different Levels
of Education-rubles, 1984, 1987.
3. Number of Students by Type of School-
students, 1970-71, 1975-76, 1980-81, 1985-86,
1986-87.
4. Number of the Population Having Higher
and Secondary Education-persons, 1970,
1979, 1986, 1987.
5. Number of Students Including All Types of
Schools by Union Republic-students, 1970-71,
1980-81, 1985-86, 1986-87.
3. Emergency and First-Aid Facilities by Union
Republic in 1986 (System of the Ministry of
Health of the USSR-independent stations,
departments, and number of persons giving out-
patient aid per 1,000 persons.
4. Mobile Types of Medical Aid by Union
Republic in 1986-number of mobile surgeries,
stomatological installations, flu-orographic in-
stallations, X-ray units, and institutions having
mobile clinic-diagnostic labs.
5. Number of Persons Under Dispensary Super-
vision by Union Republic-persons, percent of
total population, 1980, 1985, 1986.
6. Sanitoriums and Rest Institutions-beds,
population per bed, 1980, 1986.
7. Incidence of Disease With Loss of Work
Time by Union Republic per 100 Workers-
cases, number of days, 1980, 1985, 1986.
6. Publications-units, 1970, 1975, 1980, 1985,
1986. No. 264, 4 September, one page, Brigades in 1986
No. 263, 4 September, eight pages, Drctf't Decree on
Development o~'Public Health in the 12th Five-Year
Plan and to the Year 2000
1. Medical Institutions by Union Republic in
1986-number of hospitals, health-ambulatory
polyclinics, women's consultation centers and
children's polyclinics, number of independent
fel'dsher obstetrical centers, number of sanitary-
epidemiological stations and departments of
united regional hospitals.
2. Number of Dispensaries by Union Republic
in 1986-units.
No. 269, 8 September, 15 pages, For 70 Years oI
Great October: Capital Construction-text, general
discussion of construction by five-year plan. Includes
two numbers for 1987: Share of total productive
capital investment going to renovation 38.7 percent in
1985, 43.0 percent in 1986, planned for 45.0 percent
in 1987. Capital investment in MBMW grew 17
percent in 1986, planned for 30 percent in 1987.
No. 280, 11 September, one page, Distribution of
Housing by Union Republic
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Perestroyka and Soviet
Statistics
STAT
Introduction
An assessment of the impact of perestroyka on the
quality and reliability of Soviet economic statistics at
this time would be premature since there has been
little fundamental change in the Soviet statistical
system since March of 1985. The winds of change
finally reached the Central Statistical Administration
(now the State Committee on Statistics) in the spring
of 1987, and there is evidence that the tempo of
change is accelerating. Still, given the serious difficul-
ties facing the state statistical system, we should not
expect to see radical improvements for at least two to
three years.
A survey of Soviet statistics and of changes which
have been introduced into the system since Gorbachev
became General Secretary is, nevertheless, quite use-
ful. It will better prepare us for the evaluation of
future, more fundamental changes which we hope lie
ahead. Careful monitoring and analysis of recent
developments also promise to yield valuable insights
into the quality and reliability of the Soviet economic
statistics Western specialists have been using for the
last 25 to 30 years.
"Sovietologists" have achieved a good deal of success
in interpreting the published Soviet statistics, filling in
gaps in methodologies and classifications, correcting
shortcomings, and estimating and reconstructing un-
available data. But we have been less successful in
fully understanding the internal workings of the Sovi-
et statistical system and the role played by official
statistics in decisionmaking in the USSR.
For a long time we have been assuming that published
Soviet statistics represent but the tip of the iceberg
and that the TsSU and other agencies in the USSR
have at their fingertips large and well-integrated sets
of economic statistics which were removed from pub-
lic view either because they presented economic per-
formance or the welfare of Soviet people in an
unfavorable light or because of military or commer-
cial security considerations. The evidence which be-
came available over the past 10 to 15 years suggests
that the submerged part of the iceberg was neither as
large as we assumed nor of particularly high quality.
There seems to be more and more evidence indicating
that in the past such adverse social phenomena as
prostitution, drug abuse, and certain types of crime
had not been studied in secrecy by the police, medical
authorities, or the TsSU but simply dismissed a priori
as nonexistent.
Indeed, the TsSU probably was responsible for con-
cealing or ignoring evidence in the late 1970s of
deterioration of the quality of life, the rapid growth of
the "second" or underground economy, and factors
contributing to the slowdown of economic growth.
And, it appears now that these adverse developments
were concealed not only from the general public but
from central authorities as well. In the 1970s and the
early 1980s the TsSU operated, as did many other
Soviet institutions and organizations, without chal-
lenges from central authorities or from the statistical
and economic professions. Except for the gradual
introduction of computers and the organization of
separate accounting and computational facilities, one
does not see any evidence of innovation, revision of
methodologies, or even expressions of need for im-
provement of statistical series on the part of TsSU
functionaries.
Gorbachev's accession to power and his early de-
mands for modernization and increased efficiency of .
the Soviet economy had little or no effect on TsSU
leadership.2 A careful study of conferences, articles on
a A detailed, fully documented chronology of perestroyka of the
state statistical system is appended at the end of this paper.
STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
new topics, discussion of plans, and reform desiderata
published in Vestnik statistiki between mid-1985 and
late 1986 shows little evidence of recognition of the
need of fundamental changes of the state statistical
system. One of the main reasons for the slow progress
probably was the administrative staff of TsSU, which
remained essentially intact after Gorbachev came to
power. Thus, while hundreds of key officials were
being replaced in the state and the party apparatus,
the "old guard" continued to run the TsSU.3 Virtually
nothing of importance happened in 1986 as evidenced
by the lack of change in Narkhoz 1985 and the
statistical appendices in Vestnik statistiki.
A full year after Gorbachev became Secretary Gener-
al the TsSU leadership was not even considering
large-scale release of previously suppressed statistics.
This can be seen in discussions during several confer-
ences of TsSU officials held in 1986. For example,
Korolev's only reference to a possible release of more
economic statistics was a promise "to discuss the issue
of accessibility of a wider circle of specialists to
statistical information" (Markovich 1986, p. 64).
More statistical data, including data which have not
been published for many years, began to appear only
in early 1987 in the form of periodically issued
statistical bulletins by a newly created Center for
Publication and Information.
To put the TsSU's behavior in its proper context, it
should be stressed that Gorbachev's policy of glasnost
and democratization was not formulated and imple-
mented immediately after his accession to power but
evolved gradually over time. Modernization and in-
creased economic efficiency were high on the General
Secretary's agenda from the very beginning. But the
policy of glasnost, with its stress on the need to openly
expose and discuss the shortcomings of the Soviet
economic system, emerged later, when Gorbachev
came to realize the immense difficulties he was
facing. It is clear that the urgent need for wide-
reaching economic reforms would be more readily
acceptable if the majority of people were to realize
how bad the economic situation had become by the
' Lev Volodarsky, the head of the TsSU since 1975, retired in
December of 1985. His replacement, Mikhail Korolev, served as
Volodarskiy's first deputy for 10 years and can hardly be consid-
ered a "young Turk."
mid-1980s. Thus in early 1987 we hear Gorbachev
referring to the "near-crisis" and "stagnation" of the
economy. At the same time, his demands for openness
were becoming more vocal and the definition of
glasnost was expanding.
The pressure to reform the statistical system thus
came from above and in stages. The Central Commit-
tee reviewed the work of the state statistical system in
January of 1986, but the directives issued stressed
only the need for the TsSU to reduce the level of
unauthorized statistical work and to eradicate falsifi-
cation in reported data. Early in 1987 the Central
Committee returned to the question of improving
state statistics, but this time it voiced strong criticism
of shortcomings in the work of the TsSU. Almost at
the same time, the TsSU and the whole statistical
system were harshly criticized in the journal Novyy
mir in an article written by an economist named
Khanin and a journalist specializing in economic
matters named Selyunin. The authors showed that, by
using improper statistical techniques and by intention-
ally distorting data, the TsSU systematically overstat-
ed the true rates of economic growth of the Soviet
economy. Alternative measures developed by Khanin
show that national income of the USSR increased in
the 1928-1985 period 6.6 times compared with official
figures of 89, thus suggesting that official figures
overstated the true growth by a factor of 13. Official
capital and labor productivity indices were shown to
be overstated by a similar order of magnitude.?
The appearance of Selyunin and Khanin's article in
Novyy mir could conceivably be viewed as accidental
but it should be noted that a much shorter article by
the two authors with essentially the same message was
published a couple of weeks earlier in Pravda (Se-
lyunin and Khanin 1986, p. 2). Thus we have grounds
to believe that the appearance of their criticism was a
? Selyunin and Khanin 1987, pp. 182-201 and Khanin 1987, pp. 21-
28. Khanin published several studies criticizing official statistics in
the past (Khanin 1981, pp. 62-73; 1984, pp. 58-67; 1986, p. 2) but
none were as direct and as strongly critical as the Novyy mir
article. For an excellent summary of Khanin's work, see Ericson
1988.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
part of a strategy designed to put pressure on the
TsSU and, at the same time, to offer evidence to
support Gorbachev's demands for urgent reforms of
the economy.
Neither the criticism voiced by the Central Commit-
tee nor the attack by Selyunin and Khanin had much
effect on the TsSU. At a special conference of the
TsSU party organization and management held in the
spring of 1987 the director, Korolev, made a speech
on the tasks of perestroyka of statistics. The speech
and other statements made at the conference indicate
that the TsSU leadership did not take the mounting
criticism seriously. Apparently they felt that the
release of a few formerly suppressed statistical series
and routine exhortations to the cadres to improve
their efficiency were sufficient to satisfy the
reformers.
The pressure from above, however, continued to in-
crease. The Polituro stressed the need for radical
improvement in state statistics in April and empha-
sized that glasnost policy applied also to the TsSU. In
May, the Central Committee returned to the same
issue. Finally, in July the Central Committee and the
Council of Ministers of the USSR issued a decree
demanding radical changes in the state statistical
system and reorganized the TsSU into a more presti-
gious State Committee on Statistics, or Goskomstat.
A conference of the TsSU party organization and
management similar to the one held in March was
convened in August. Korolev again discussed the tasks
of perestroyka but this time added "radical" to the
title of his address and more soul searching to the
text. The new 1986 Narkhoz, which was apparently
ready for printing in the early summer, was held up
and several sets of long-suppressed data were added to
it. Thus, things were finally changing, particularly the
availability of statistical data in the open literature.
According to Belov, a total of "90,000 units of
statistical information" were removed from the
"closed" list in 1987 (Belov 1988, p. 4).
The pace of release of previously unavailable data was
thus accelerating, but the whole process shows
evidence of haste and confusions Unfortunately, there
was no noticeable progress in the area of statistical
methodology and in the quality of statistics released
by the Goskomstat in 1986 and 1987, but the de-
mands for more credible information generated by the
glasnost campaign continued to increase. Two devel-
opments late in 1987 were particulary disturbing for
Goskomstat. The prestigious Institute of World Econ-
omy and International Relations (IMEMO) published
a detailed study which showed that, in the 1917-1986
period, Soviet national income was growing at rates
3.5 times lower than the rates reported in official
sources, and the industrial product rates were lower
by a factor of almost 5 ("Sovetskiy Soyuz ... " 1987).
The Soviet data were converted to 1980 US dollars,
making the results not quite comparable to official
statistics, but the order of the difference was never-
theless striking and has without doubt adversely
affected the credibility of official data in the eyes of
the public. At the same time, a leading Soviet econo-
mist, Abel Aganbegyan, published an article on the
progress of perestroyka in which he cited official
national income growth rates but added that they
were overstated because they did not correctly reflect
inflationary trends, particularly in consumer goods,
machinery, and construction (Aganbegyan 1987, p. 7).
Neither the IMEMO nor Aganbegyan's study at-
tacked the Goskomstat directly, and the differences
were not as large as suggested by Selyunin and
5 Not all data, by far, which disappeared from the public domain in
the mid-1970s were restored in Goskomstat publications. In fact,
some new deletions in published statistics were made. The data on
the ruble value of unfinished construction, which had been routine-
ly published in Narkhoz for years, were deleted in the 1985
Narkhoz and not restored in the 1986 Narkhoz. The format of the
tabulation of ruble value of working capital was completely
changed in the 1986 Narkhoz. Furthermore, in constrast to the
traditional format, the working capital data were given for one year
only, making comparisons with earlier years impossible. The confu-
sion with the declassification of statistics can also be illustrated by
the following case. The short statistical handbook, SSSR v tsiJrakh
v 1986 godu, published a table with rates of death per 100,000 of
working age population and the absolute number of deaths. These
data, of course, made it possible to estimate the absolute number of
the working age population of the USSR, which has not been
published since the early 1970s. Apparently, the publication of both
rates and absolute numbers was an error, and the latter were
deleted in the 1986 Narkhoz published in August ] 987. The
December issue of Vestnik statistiki, however, carried a table with
sex-age specific population data for 1987, including a separate row
with the working age population statistics.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Khanin earlier. The implied criticism, however, came
from within the official Soviet establishment and
therefore could not be dismissed by Goskomstat as
easily as was Selyunin and Khanin's study. Conse-
quently the need to review and alter national income
methodology will be felt even stronger.
Weaknesses in Soviet Statistics
Shortcomings of Soviet economic and social statistics
and of the statistical system itself have been analyzed
by Western scholars many times. There is evidence
that the paucity of explanatory and methodological
material extends to internal and closed data as well.
Discontinuities in time series, residuals in tabulated
statistics, and obvious inconsistencies among different
sets of data are seldom if ever explained in official
publications.
Methodology. The TsSU relies almost exclusively on
total coverage in the collection of data. It seldom
relies on samples and surveys.b Biased formulas, in-
consistent use of formulas, the use of excessively long
periods of time between benchmark years in calculat-
ing constant price output and price indexes, and
chain-linking of indexes make most time series errat-
ic. Aggregate measures such as national income and
gross social product are particularly suspect because
separate components (that is, industry, construction,
etc.) are derived on the basis of different methods and
definitions.
Accuracy and Reliability. The methodological short-
comings described above, of course, reduce the reli-
ability of processed summary statistics made available
by TsSU. In the past, Western specialists believed
that at least the primary data collected by the TsSU
from various reporting units were fairly reliable,
particularly the data in physical units or in current
prices. The natural tendency to distort the data by
reporting units was believed to produce unbiased
results because distortions were likely to be in the
upward or downward direction, thus canceling each
Among the few known sets of Narkhoz statistics based on
sampling are the household budget data reported by a group of
62,000 families (being increased to 90,000 in 1988) and average
urban kolkhoz market prices and quantities collected from markets
in 251 cities. The latter have not been published for some time. In
addition to these, the TsSU would periodically conduct special
sample surveys.
other. The continuous concern expressed by TsSU
with inflated statistical reporting (pripiski) and nu-
merous anecdotal references found in the literature,
however, suggest that primary data collected by the
TsSU may not be too reliable and the production
statistics and aggregate data very often contain up-
ward biases.
Functions oJ'the TsSU. Unlike most other state
statistical agencies in the world which collect and
process national statistics, the TsSU performs two
distinctly different functions for the Soviet Govern-
ment. The first is the basic task of collecting, process-
ing, and publishing statistical data for government
users. The second function is to serve as a central
recordkeeping organization responsibile for the accu-
racy of the data obtained from enterprises and organi-
zations, including the authority to audit suspect re-
porting units. This second function determines certain
aspects of and even interferes with the first function
of statistical collection of the TsSU. For example, the
predilection for total coverage in collection of data
over sampling techniques referred to above is directly
related to the auditing function.' The perennial con-
cern of the TsSU over the possibility of willful
distortions and falsifications (pripiski) on the part of
the reporting units is legitimate. On the other hand,
were the TsSU simply a state agency collecting,
processing, and analyzing national statistics, the
whole issue of distortions would become irrelevant.
Such an agency would probably rely mainly on sam-
pling and properly designed sampling techniques,
which would have significantly reduced the possibility
of obtaining biased data. Other aspects of TsSU work,
' In the 1920s the TsSU was performing purely statistical functions
and, as far as one can see from published statistical compendia,
extensively used sampling techniques. In 1931 the TsSU ceased to
be an independent statistical agency and was subordinated to
Gosplan under the new name of the Central Administration of
National Economic Accounting (TsUNKhU) and the TsSU house
organ, Vestnik statistiki, was merged with the Gosplan's Planovoye
khozyaystvo. Assumption of the auditing and control functions on
behalf of Gosplan was the logical result of this reorganization. In
1941 the name TsSU was restored but the agency continued to.be
subordinated to Gosplan until 1948, when it regained its indepen-
dence. Either by inertia or by design, the TsSU kept the responsibil-
ity for the auditing of all governmental statistics.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
such as the concentration on production at the ex-
pense of consumption, neglect of such aggregate
statistics as national income and product, and lack of
interest in preparation of data by economic regions, in
contrast to republic or oblast' divisions,$ can also be
explained by the requirements of the auditing role of
TsSU. Nonetheless, Goskomstat is not likely to sepa-
rate the two functions, since this would reduce its
prestige and status.9
Political Factors
The issue of the authority of the central statistical
agency over other state agencies and organizations is
interesting. Conventional wisdom has it that the
TsSU has had a monopoly on the collection, process-
ing, and publication of all economic, social, and
demographic statistics in the USSR. In addition, the
statistical authorities have had almost unlimited au-
thority to request data from other agencies and to
prohibit collection of statistical information which
TsSU deems to be unnecessary. It is difficult to
document this issue fully, but I have always believed
that the authority of the TsSU over other agencies
varies according to the importance and prestige of the
latter. For example, the Ministry of Foreign Trade
and Ministry of Finance (and Gosbank) were suffi-
ciently powerful to deliver statistics they felt were
appropriate and in the form they wanted to TsSU.
Both ministries published their own detailed statisti-
cal compendia, and the TsSU publications essentially
reprinted certain series from these publications with-
out change.10
e As in many other countries, the administrative divisions of the
USSR, that is, republics, kray, and oblast, do not concide with true
economic regions. Spokesmen for economic geography (a weI]-
developed and prestigious discipline in the USSR) have long been
demanding statistical groupings by economic regions but without
success. The main reason is, of course, that the TsSU has to audit
performance of existing governmental geographic divisions and is
not interested in regions which have no administrative functions.
' Nikolay Belov specifically singled out complete coverage as the
strong point of the Soviet statistical system, in contrast to statistical
systems of capitalist countries, which rely on sampling (Belov
1987c, p. 3).
1? Vneshnyaya torgovlya SSSR is published annually by the Minis-
try of Foreign Trade and budgetary compendia by the Ministry of
Finance every five years. Vneshnyaya torgovlya 1986, which
appeared in the spring of 1987, showed absolutely no changes in
coverage or in methodology. The handbook of budgetary statistics
covering the 1981-1985 period was published in the fall of 1987. As
with foreign trade data, there were no changes in the format or
coverage, and the same is true of the CEMA statistical handbook
which was released in December of 1987.
This sometimes resulted in inconsistencies among the
data in TsSU publications. For example, the Ministry
of Foreign Trade continues to use its own foreign
trade commodity classification, which is different in
several respects from the unified classification used by
TsSU and Gosplan. TsSU apparently could not revise
or change the data submitted by the Ministry of
Foreign Trade and had to publish them without
correcting any inconsistencies."
The case of the so-called "moral" statistics is even
more complex. The responsibility for the collection
and processing of "moral" statistics-covering such
areas as crimes of various types, prostitution, arrests,
alcohol and probably drug abuse, certain types of
mental illnesses, accidents (industrial and street), sui-
cides, and the like-was transferred from the TsSU
(then TsUNKhU) to the NKVD in the early 1930s
and is still controlled by the internal security agency.
We know little about their scope, coverage, or accura-
cy. Officials of the Ministry of Internal Affairs
apparently are not inclined to share statistics collected
by them with other state agencies, including the
TsSU.'Z Thus, TsSU has not been processing or
publishing "moral" statistics for years.
One probable reason for the unexpected elevation of
the TsSU to the position of a state committee is to
give the statistical agency more power to deal with
influential and independent organizations such as the
Ministry of Foreign Trade and the Ministry of Inter-
nal Affairs. In fact, the Central Committee decree
creating Goskomstat specifically provided for "cen-
tralization of all foreign trade statistics in statistical
" For example, according to the foreign trade classification, con-
sumer durables and appliances and ship repair are not classified as
machinery products as they are under the general classification. As
a result, the structure of foreign trade obtained by the TsSU from
the Ministry of Foreign Trade data and published in Narkhoz
shows machinery which is not comparable to machinery data
reported elsewhere in the compendium.
12 The case of the statistics on the stock of motorcycles will illustrate
this. The traffic division of the MVD, GAI, has the responsibility of
registering all motor vehicles in the country and thus has all
information on the stock. The TsSU, however, apparently never
succeeded in obtaining these data from the GAI, and the stock of
privately owned motorcycles reported in the Narkhoz is based on
cumulative sales of motorcyles in retail trade and some rough
assumptions of the useful life of an average motorcycle (Gorchak
and Lobko 1979, p. 44).
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
agencies" (O korennoy ...1987, p. 183). It will be
interesting to see whether foreign trade statistical
compendia will be transferred to the Goskomstat and
modified to agree with Goskomstat and classifications
and statistical conventions."
As for social statistics, it is too early to say whether
the Goskomstat will succeed in wresting them from
the MVD. The absence of reliable and comprehensive
social and socioeconomic statistics in the USSR was
singled out for harsh criticism by several Soviet
writers, particularly Tatyana Zaslavskaya. The depu-
ty head of the Goskomstat, Nikolay Belov, timidly
suggested in an otherwise strongly critical and frank
article that the time has come to "re-create" moral
statistics in order to ensure systematic analysis of all
social processes, including deviant behavior (Vestnik
statistiki, No. 17, 1987, p. 12). Still, it is too early to
say whether and how this might be accomplished.14
The Central Committee decree on the improvements
in the statistical system did not direct the transfer of
social statistics to Goskomstat, but only called for
increased coordination of all studies of sociodemogra-
phic processes among statistical agencies (O koren-
noy ...1987, p. 182).
The More Recent Data: A Look at Their Quality
I do not propose to review and assess in a comprehen-
sive manner the new statistical data which became
available recently. The information contained in the
new Narkhoz and the "press information" released by
the TsSU in 1987 are evidence that statistical meth-
odologies have not been improved or defective time
series data revised. Most of the price and output index
series, and the indicators of aggregate performance-
such as national income, gross social product, or value
" In a rather dramatic and completely unexpected move, the Soviet
Government disbanded the Ministry of Foreign Trade and the
State Committee on Foreign Economic Relations, creating a new
Ministry of Foreign Economic Relations in January 1988 (TASS
dispatch, January 16, 1988). Several foreign trade organizations
which operated under the old ministry were transferred to manu-
facturing ministries. By all accounts the new Ministry of Foreign
Economic Relations will have a much more restricted scope of
operations than the old and the Goskomstat will be in a better
position to assume the responsibility for collection and publication
of foreign trade statistics.
'? It is interesting to note that Belov used in this context the Russian
verb vossozdat', which could be interpreted as "to reintroduce" or
"to re-create," implying that moral statistics are not being collected
by the TsSU as they had been in the past.
of gross output of industry-are continuations of data
series presented in the 1985 and earlier issues of the
Narkhoz. Price data represent one of the areas that
needs to be improved the most. One high-level Gos-
komstat official recently referred to financial and
price statistics (along with the statistics of the agroin-
dustrial complex) as serious bottlenecks (Ryabushkin
and Remizov 1987, p. 28). Still, a review of all articles
and reports dealing with TsSU plans published in
Vestnik statistiki yielded only a few references to
plans to study prices, relations between prices and
work incentives, and relations between prices and
productivity of new machinery.15 It means that Gos-
komstat may be planning marginal improvements, but
no major changes in the methodology of constructing
price indexes and the employment of such indexes for
the computation of output indexes in constant prices
appear in the works.
Several Western analysts have recently concluded
that Goskomstat-with or without Gorbachev's
knowledge-has "doctored" retail trade sales data
and their incorporation into the national income
accounts in 1985 and 1986. In both years, sales of
alcoholic beverages were reduced substantially as the
result of Gorbachev's antidrinking campaign. With
liquor sales constituting a large share of retail trade
turnover, such cutbacks should have resulted in a drop
in consumption and national income. Instead, these
show significant increases in those }ears. In my
opinion, Soviet authorities have not deliberately falsi-
fied these data. Rather, the inconsistencies in the
national income data in constant prices probably are
caused by the use of biased formulas and the mixing
of different types of price indexes.
The evidence of improvements in methodology or
revisions in the data reflecting important changes
found in the 1986 Narkhoz is not impressive. Al-
though these are minor, they are worth mentioning
because of their novelty. As was widely reported in
15 About the only innovation in the field of prices which emerged
from the Goskomstat in the last three years is an index reflecting
changes in prices of material inputs purchased by agriculture. The
Council of Ministers directed the then TsSU to prepare such a price
index for the purposes of indexing procurement prices. The price
index was prepared and published without documentation or any
technical detail, so we cannot evaluate its effectiveness.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
the media, Soviet authorities uncovered a major scan-
dal involving large-scale overstatement of raw cotton
production and procurement in Central Asia, particu-
larly in Uzbekistan. The overstatement of procure-
ment, which generated tens of millions of rubles in
illegal income shared by hundreds of officials, was
particularly flagrant in the 1975-83 period. Despite
the widely publicized prosecutions and numerous esti-
mates of fictitious cotton tonnage given in the media,
the 1985 USSR and Uzbekistan statistical yearbooks
listed the cotton procurement figures for earlier years
without revisions (Narkhoz 1985, p. 210). The 1986
Narkhoz, however, reported substantially lower fig-
ures for the years in question (p. 228). Comparison of
the two sets of figures indicates that, in the 1976-80
period, procurement of raw cotton by all producers
was overstated by 1,925,000 tons (about 5 percent
over the correct figures), with Uzbekistan accounting
for 1,725,000 tons. In the 1980-83 period the over-
statement increased to 3,935,000 tons (about 14 per-
cent), 2,645,000 of which was in Uzbekistan.
Data revisions contained in the national income sec-
tion in Narkhoz 1986 could mean that the methodolo-
gies used to calculate these data have been changed,
something which has not been done in Soviet national
income accounting for many years. National income
produced in industry in 1986 was reported as 258
billion rubles (in prices of the current year) compared
with 263.1 billion rubles in 1985. A footnote in the
table explained that the drop in 1986 was due to
increased subsidies and added that, without this ad-
justment, the 1986 figure would have been 265.6
billion rubles (Narkhoz 1986, p. 122). It should be
noted that, while references to agricultural subsidies
are found frequently in the Soviet literature, official
statistical compendia such as Narkhoz and budgetary
handbooks have never identified these subsidies or
even recognized their existence. The implied subsidy
of 7.6 billion rubles must represent the subsidy of
manufactured goods sold to agriculture (machinery,
fertilizer, electrical power, etc.), but we need more
information to understand the accounting principles
involved. The subsidy on manufactured goods sold to
agriculture has been in effect since 1967 and has
ranged between 4 billion and 5 billion rubles in the
early 1980s. The puzzling aspect of this adjustment is
that it affected only 1986 and that national income
values for earlier years were left unadjusted in the
new Narkhoz. Unfortunately, a much higher subsidy
of some 70 billion rubles on purchases of agricultural
raw materials by food-processing and light industries,
which is concealed in national income originating in
industry, was neither identified nor corrected.16 Na-
tional income flows between agriculture and industry
basically remain as distorted as before." Thus, as is
the case with other recently introduced improvements,
the correction in national income accounts in 1986 is
not fully explained and does not go far enough.
There was another development in Goskomstat's pub-
lications of national income data. The July 1987
decree, among other specific instructions, directed the
Goskomstat to expand international economic ana-
lyses and to construct more meaningful measures of
international comparisons (O korennoy ...1987, p.
183). As with other instructions, Goskomstat respond-
ed with a reorganization and expansion of a special
section on international comparisons in the 1986
Narkhoz (pp. 653-698). The section consisted of sim-
ple tabulations of production data by countries and
did not offer any aggregate measures. There is one
interesting detail-for many years Narkhoz compen-
dia have been publishing a table of main aggregate
measures for the USSR as a percent of the United
States. For purposes of comparison, Soviet net materi-
al product (NMP) was converted to US dollars and
divided by a similar measure estimated for the United
States. Over the years these ratios became almost
meaningless-the ratio of Soviet and US NMP re-
mained almost constant at about 66 percent between
1970 and 1985, while the same section reported that
the average annual growth of Soviet NMP was 4.5
16 State procurement agencies sell agricultural raw materials to
processing industries at much lower prices than they pay to
agricultural producers. This subsidy averaged over 25 billion rubles
in the late 1970s and increased to some 70 billion rubles by the mid-
1980s. In what I consider to be an incorrect accounting convention,
this subsidy is not recorded in national income originating in
agriculture but in industry (Treml 1982 and 1988).
"Although a large share of turnover taxes is levied on industrial
goods of agricultural origin (alcoholic beverages, sugar, vegetable
oil, wool, and other textiles), all turnover taxes are recorded in
national income originating in industry. Recognizing this distortion,
the national income tables in the Narkhoz carried a footnote
reimputing national income between industry and agriculture pro-
portionally to labor costs (see for example Narkhoz 1984, p. 424).
Unfortunately, this useful correction was deleted in 1985 and not
restored in 1986.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
percent compared with 2.8 percent for the United
States (Narkhoz 1986, p. 582). It is, of course,
impossible for one country to grow much faster than
another country while the ratio of their national
income remains unchanged. Realizing the obvious
incongruity between the ratios and the rates of
growth, the editors of the 1986 Narkhoz deleted the
tabulation of the ratios from the international com-
parison section and placed it in a different part of the
compendium (p. 13), that is, as far from the table on
growth rates as possible. In doing this they have, of
course, removed one of the few aggregate measures of
international comparisons from the appropriate sec-
tion. There was one sign of progress in this respect.
Responding to some criticism, the Goskomstat con-
structed ameasure of Soviet Gross National Product
as used in the West, which was never done before.
The 1987 plan fulfillment summary prepared by
Goskomstat included a statement that the 1987
USSR GNP increased by 3.3 percent compared with
1986 ("Ob itogakh ... " 1988, p.5). No explanation of
the methodology of conversion or documentation was,
however, given. Since the traditionally measured So-
viet NMP grew at a disappointingly low 2.5 percent,
the introduction of the GNP rates of growth was
probably dictated by a desire to show the development
of the Soviet economy in a more favorable light.
Overall, the national income accounts remain the
weakest part of published Soviet statistics. The meth-
odology of recomputing national income (net material
product) in constant prices has serious shortcomings
which make the official series almost useless. The
method is as follows: as a first step, gross value of
output in five major branches of the economy-
industry, construction, agriculture, transportation and
communications, and trade and distribution-in
prices of the current year are deflated by appropriate
price indexes. Separate price indexes for material
inputs into production and for depreciation are used to
convert the material component of each GVO to
constant prices. Net material product for each branch
is estimated by subtracting the material component
and depreciation computed in constant prices from
GVO in constant prices. NMP distributed (that is,
calculated by the expenditures on final material
goods) is broken into major components-industrial
goods, agricultural goods, other goods, and construc-
tion-and each is deflated by specially prepared price
indexes.
The problem with this dual approach is that in the last
27 years for which we have data, the rates of growth
of constant-price NMP produced consistently exceed
the rates of growth of constant-price NMP distribut-
ed. The annual differences are in the range of 0.5 to
1.5 percentage points and for the whole 1960-1987
period amount to 56 percentage points. Since the
difference always has the same sign, it suggests that
the method used to deflate has abuilt-in bias.18
The problem of estimating NMP growth is com-
pounded by a bizarre choice of base years for the
different components of NMP produced and NMP
distributed. Aggregate NMP values were measured in
1956 constant prices in the 1956-58 period, 1958
prices in the 1959-65 period, 1965 prices in the 1966-
75 period, 1973 prices in the 1976-85 period, and
1983 prices starting with 1986. Price indexes for gross
output of industry, agriculture, capital investment,
consumer services, commodity exports and imports,
and goods sold in retail trade have been derived with
different base-year weights, and in virtually all in-
stances the introduction of new weights occurred in
different years, none of which coincided with NMP
periods or base years. One can only hope that the
criticism of official national income rates of growth
implicit in the IMEMO study and by Aganbegyan
discussed above will force the Goskomstat to finally
address the thorny issue of revising the methodology
of national income accounting and of price indices.
1e NMP growth rates by republics present a mixed picture. For most
years rates of growth of NMP produced exceed rates of growth of
NMP used but not always. The range of differences between the
two rates are also wider. For example, the 1985-86 rate of growth
of Armenian NMP produced was reported as 3.9 percent and the
rate for NMP used was an unusually high 9.9 percent. Major
components of the NMP used, such as retail trade turnover, grew
only by 3.5 percent in constant prices, which makes the inconsisten-
cy between reported rates even more marked (Armenia 1986, p.
215).
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Appendix
Perestroyka of State
Statistics in the USSR
Under Gorbachev: A Chronology
Spring 1985
The SSSR v tsifrakh v 1984 godu (short handbook) is
published. No evidence of improvement in quantity or
quality of statistics. In fact, some statistical series
which have been published for years, such as produc-
tion of wine and beer, are deleted. The same is true of
the Narodnoye khozyaystvo SSSR v 1984 godu
(Narkhoz 1984) published in the late summer.
December 1985
Lev Volodarskiy, the head of the TsSU since 1975, is
replaced by the former first deputy, Mikhail Korolev.
No major administrative changes. Because of Volo-
darskiy's age (74) it is impossible to say whether he
was dismissed or simply retired.
January 1986
Vestnik statistiki publishes the traditional January
article summarizing plans for state statistical agencies
in the current year. As in the past, the article is signed
by Ivan Matyukha, the head of the Summary Statis-
tics and Statistical Methodology Division-the key
division of TsSU. The plan does not contain anything
new (Matyukha 1986, pp. 3-13).
The Central Committee reviews the problems of
improving the organization of state statistics and
stresses the need to eliminate unneccessary paperwork
and unauthorized statistical reporting. The TsSU is
criticized for not exercising sufficient control over
state statistics("V Tsentral'nom Komitete :::-"-1986;
p. 3).
volume of government reporting by one-half, but more
effort on reduction of the level of unauthorized statis-
tics is needed (reported in Pravda, June 5, 1986).
May-June 1986
Three conferences on the improvement of statistical
work are convened in Moscow under TsSU's auspices.
Two are for "in-house" specialists: a conference of
heads of summary statistics and statistical informa-
tion divisions of republic TsSUs, and a conference on
statistics of the newly created Gosagroprom (State
Agroindustrial Committee). At the third conference,
TsSU statisticians meet with their counterparts from
other state organizations. The issues of excessive
paperwork and of unauthorized statistics are dis-
cussed again. Speakers suggest improvements in some
statistical methods, stress the need to avoid discrepan-
cies between TsSU and ministerial data, but, in
general, the tone of criticism of the state statistical
system is mild and the issues raised are trivial. The
issue of wider availability of statistical information is
barely mentioned (Dubnov 1986, pp. 67-71; Marko-
vich 1986, pp. 63-67; Somova 1986, pp. 60-62).
October 1986
The new 1985 Narkhoz offers some evidence of the
announced policy of glasnost with publication of some
statistical series which were not available for some
time, such as the grain harvest statistics (discontinued
in 1981) and data on production, sale, and per capita
consumption of alcohol (not published since the early -
1930s); but, generally speaking, the new Narkhoz is a
disappointment.
May 1986
The Central Committee returns to the issues of
improvement of organization of state statistics. The
TsSU is reported to have succeeded in reducing the
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
The Central Committee publishes a resolution with an
unusually strong condemnation of falsification in sta-
tistical reporting in Moldavia, Kirovograd Oblast of
the Ukraine, and the Ministry of Automobile
Production.
December 1986
Korolev discusses the newly released Narkhoz 1985 in
an Izvestiya interview, stressing the need for open
discussion of information reflecting both successes
and weaknesses of the Soviet economy, improvements
in quality and timeliness of data, etc. (Korolev 1986,
p. 3).
January 1987
TsSU begins distribution of "press-information,"
three- to five-page bulletins appearing once or twice
weekly. The bulletins offer statistics which have not
been published before but also some routine data.
Major Soviet dailies and periodicals, such as Kom-
munist and Ekonomicheskaya gazeta, selectively
publish some of the "press-information."
February 1987
A long article with unprecedentedly harsh criticism of
official statistics, written by Vasiliy Selyunin and
Grigoriy Khanin, appears in Novyy mir (1987a,
pp. 182-201). The same two authors published a
shorter version of the same article earlier in Pravda
(1986, p. 2), while Khanin published an article of a
more historical nature in January (1987, pp. 21-28).
The authors reject official price and output indices as
biased, criticize the TsSU for data distortions and the
use of improper methods, and offer their own esti-
mates of rates of growth of national income, which
are much lower than the official rates. The open
criticism of official statistics creates a sensation.
Khanin is invited to speak to different audiences such
as the Moscow State University and the Central
Economic-Mathematical Institute (Ericson 1988).
After an unusual delay of almost a month, Vestnik
statistiki publishes the traditional article by Ma-
tyukha summarizing the TsSU work plan for 1987
(Matyukha 1987, pp. 10-20). The plan, however, does
not contain anything radically new.
March 1987
Vestnik statistiki publishes an unsigned lead article
which probably originated not in the TsSU but with
the Central Committee or with somebody on Gorba-
chev's staff. The article discusses the tasks ahead for
the Soviet economy and criticizes the TsSU for not
providing timely and accurate information ("Statis-
tika ... " 1987, pp. 3-9). A conference of the party
organization and management of the TsSU meets to
discuss the criticism of the statistical system voiced at
the Central Committee plenary session ("Rasshiren-
noye ... " 1987a, pp. 32-48). In his main address
Korolev also briefly refers to Selyunin and Khanin's
article, labeling it as "profoundly misleading" while
noting that the poor quality of statistical information
in the country gives support to such criticism. Impor-
tance of timely publication of statistics and wider
availability is stressed (Korolev 1987a, p. 6).
Pravda reports that the Politburo reviewed the need
for "radical improvement" of statistics in the country.
The issued instructions stress the need to improve the
reliability of statistics and to increase their
availability.
May 1987
The Central Committee reviews the status of statisti-
cal work in the country and criticizes the excessive
amount of data collected outside of regular TsSU
channels.
Vestnik statistiki publishes an article by two authors
(Skobtsova and Adler 1987, pp. 52-56) with relatively
strong criticism of methodological and classificational
shortcomings of published state statistics. Inconsisten-
cies among data reported by different republic TsSUs
are noted.
June 1987
Responding to Selyunin and Khanin's criticism, three
professors of statistics defend the state statistical
system in Vestnik statistiki, but the arguments are
rather weak and miss the most important points
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
(Knyazevskiy et al. 1987, pp. 53-60). Selyunin and
Khanin's conclusions are also attacked in Ekonomi-
cheskaya gazeta (Adamov 1987, p. 14).
The deputy director of the TsSU, Nikolay Belov, is
interviewed in Argumenty i.1akty. Belov notes with
regret that so far there has not been real progress in
the struggle with falsification in statistical reports. He
also states that new methodological and specialized
statistical compendia will be published by the TsSU in
the next five years (Belov 1987c, pp. 2-3).
July 1987
The July issue of Vestnik statistiki publishes an
article by the deputy head of TsSU, Belov, about the
tasks facing the statistical system in connection with
the reforms (pp. 8-19). The article is more critical
than similar articles published earlier and demands
major improvements. Belov focuses on the struggle
with "illegal" statistics (that is, statistical data col-
lected outside of the TsSU and without its approval)
and on falsification in statistical reporting.
The Central Committee and the Council of Ministers
of the USSR publish a decree concerning "measures
of radical improvement of statistics work" in the
country. TsSU is reorganized into aunion-republic
State Committee for Statistics (Goskomstat) with
Korolev retained as the director. Among other mea-
sures, the Central Committee demands a reduction in
the level of statistical reporting taking place outside
the regular channels, an end to falsified reporting, and
greater reliability of statistics. Goskomstat is also
instructed to develop new statistical techniques for
measuring the progress of perestroyka and quantita-
tive measures of "success indicators." The decree
addresses a number of specific issues such as the need
to increase the size and the representativeness of the
household budget survey, wider use of surveys of
public opinion, and an increase in the number of
statistical publications, including both specialized
compendia and methodological sources. The decree
also asks for the concentration of foreign trade statis-
tics under the auspices of Goskomstat and for Gos-
komstat coordination of all studies of sociodemogra-
phic processes in the country (O korennoy ...1987,
pp. 178-190).
Goskomstat organizes a large conference to discuss a
set of experimental formulas designed to measure the
effectiveness of productive processes under the new
conditions. Reaction of outside specialists invited to
the conference is basically negative (Ippolitov 1987).
August 1987
The new status and the tasks of a radical change of
the statistical system are discussed at a special confer-
ence of the party organization and management of the
Goskomstat. The tone of the criticism is much sharper
than earlier, and a series of far-reaching changes are
discussed. Korolev's speech at the conference is pub-
lished in Vestnik statistiki No. 9, 1987, and summar-
ies of statements made by others are published in the
October issue. The head of the newly created Center
for Publications and Information of the Goskomstat;
Leonid Umanskiy, reports that 10 statistical compen-
dia devoted to different subjects will be published in
1988 ("Rasshirennoye ... " 1987b, p. 38).
Korolev discusses the tasks of the reorganized Gos-
komstat in an interview in Pravda (Korolev 1987c,
p. 2).
The Deputy Minister of Foreign Affairs of the USSR,
speaking in New York, states that what was always
claimed to be the Soviet defense budget (and pub-
lished as such in official statistical compendia) covers
only military pay, operational costs, military pensions,
and military construction. Procurement of weapons
and military research and development are covered
out of other (unspecified) budgetary categories (Pe-
trovskiy 1987, p. 4).
October 1987
The 1986 Narkhoz is published.19 The new Narkhoz
offers several sets of new statistics (some of which had
" In the second part of the 1970s the annual issue of Narkhoz
would be approved for publication (podpisano k pechati) in July; in
the early 1980s the approval date moved to August. A puzzling
feature of the 1986 Narkhoz is that it carries two dates of
approval-June 5 and August 25. The probable explanation is that,
having promised to expedite their publications, the Goskomstat was
ready at an unusually early date in June but then had the
manuscript returned. We do not know what happened, but the final
approval was delayed by two months.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
already appeared in Goskomstat press releases), but
the whole compendium shows evidence of haste and
confusion, and some data are contradictory. A meth-
odological appendix to the compendium which has not
appeared for several years is restored but does not
contain anything new. The format and the table of
contents are completely reorganized with new sections
and divisions reflecting economic reforms: The new
format is, however, more confusing than helpful and
special sections do not contain new aggregate
measures.
As a followup on the decision of the Central Commit-
tee and the Council of Ministers of July of 1987, the
latter adopts a decree with further details on the
organization and restructuring of the Goskomstat
("Uskorit' ... " 1988, pp. 3-6). The Ministry of
Finance of the USSR publishes its statistical compen-
dium covering state finances for 1981-85 (Minis-
terstvohnansov SSSR 1987). The compendium fol-
lows the format of similar compendia published in the
past; no new data are given and all summary budget-
ary tables contain unidentified residuals.
Nikolay Belov (1987b, p. 16) reports on progress in
restructuring of the state statistical system and says
that five new statistical compendia will be published
in 1988 (population, labor, consumer goods, capital
investment, and agriculture).
November 1987
Policy of glasnost is not evident in the newly pub-
lished 1987 Statistical Handbook of CEMA Coun-
tries (Sovet Ekonomicheskoy Vzaimopomoshchi
1987). The format and contents are identical to
CEMA handbooks published in the past with no new
statistics added.
November 1987-January 1988
Direct and indirect criticism of official Goskomstat
statistics, particularly national income data, appear
with increasing frequency. Aganbegyan, one of the
key engineers of perestroyka, notes in a lead article in
the journal Ekonomika i organizatsiya promyshlen-
nogo proizvodstva (Aganbegyan 1987, p.7) that offi-
cial Soviet national income statistics understate infla-
tionary factors and therefore overstate the true rates
of growth. The house organ of the prestigious Insti-
tute of World Economy and International Relations
(IMEMO) publishes atwo-part study of aggregate
indices of growth (national income, industrial and
agricultural product measured in constant 1980 dol-
lars) for the 1913-1987 period. National income
growth rates reported in the study are three times
lower than the official rates and industrial product
growth rates are five times lower ("Sovetskiy
Soyuz ... " 1987). Novyy mir publishes an article by
Selyunin and Khanin in which the authors defend
their analysis and repeat the main points of criticism
of official state statistics published earlier (1987b,
pp. 255-257). Selyunin attacks some policies of peres-
troyka in an article in Sotsialisticheskaya industriya
and criticizes official statistics for distorting the real
growth of standards of living in the country (1988, p.
3). Sharp criticism of official statistics is also voiced
by an average man in the street. Thus, the publication
of Goskomstat data on household budgets in Izvestiya
produced a large number of angry letters to the editor
dismissing the budgetary data as total fabrication
(Izvestiya, December 19, 1987). Selyunin's article in
Sotsialisticheskaya industriya also produced many
letters to the editor in which the readers fully support-
ed Selyunin's criticism of official statistics. The statis-
tical section of Vestnik statistiki and Goskomstat
press releases continue to publish data which have not
been published for years, such as age-sex specific
demographic data for 1987, alcohol mortality by
major causes, ruble values of different categories of
pensions, and more detailed breakdowns of national
income accounts in current prices. Belov reports that a
total of 90,000 "units of statistical information" were
removed from the "closed" list, that is, declassified, in
1987 (1988, p. 4).
January 1988
Goskomstat report on the plan fulfillment shows that
the economy did not do well in 1987; national income
grew 2.5 percent compared with the planned rate of
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
4.1 percent. A novel feature of the plan fulfillment
report is the use of the Western concept of GNP
measure applied to the USSR ("Ob itogakh ... "
1988, pp. 15-26.)
Goskomstat publishes new regulations governing col-
lection of statistics in the country aimed at reduction
of "unauthorized statistics." Two sets of statistical
data are recognized-the interbranch (mezhotrasle-
vaya) statistics, which are uniform for the whole
economy, and branch statistics established in separate
branches. The authority of Goskomstat to approve all
forms of statistical reporting is reiterated ("Porya-
dok ... " 1988, pp. 19-21).
February 1988
The Goskomstat announces changes in financial ar-
rangements with the state treasury-starting in 1988
the agency will receive (negotiate) a government order
(zakaz) for certain types of statistical services and
data analyses. A long article in Vestnik statistiki
(Matyukha 1988, pp. 3-14) spells out tasks and assign-
ments given to the Goskomstat. The order specifies
that for purposes of expansion of glasnost, Goskom-
stat will publish several topical compendia covering
industry (last published in 1964), agriculture (last
published in 1971), and population (1975).
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Bibliography
Media interviews are listed under the name of the
interviewee; the annual statistical compendia pub-
Iished by the TsSU, Narodnoye khozyaystvo SSSR,
are cited as Narkhoz with the last year of coverage;
and unsigned articles are cited by the title.
V. Adamov 1987. "Chto stoit za indeksami," Ekono-
micheskaya gazeta, No. 29, June, p. 14.
A. G. Aganbegyan 1987. "Programme korennoy per-
estroyki," Ekonomika i organizatsiya promyshlen-
nogo proizvodstva, No. 11, pp. 3-19.
N. Belov 1987a. "Zadachi statistiki v usloviyakh
reformy khozyaystvennogo upravleniya," Vestnik sta-
tistiki, No. 7, pp. 8-19.
N. Belov 1987b. "Statistika: vremya peremen,"
Ekonomicheskaya gazeta, No. 42, p. 14.
N. Belov 1987c. "Perestraivayetsya li nasha Statisti-
ka?" Argumenty i fakty, No. 23, 13-19 June, pp. 2-3.
N. Belov 1988. "Schet v nashu pol'zu," Moskovskiye
novosti, No. 2, January 10, p. 4.
S. Bushev 1987. "Statisticheskaya informatsionnaya
sistema na putyakh perestroyki," Vestnik statistiki,
No. 10, pp. 3-10.
A. Dergachev 1987. "Voprosy perestroyki statistiki
finansov i tsen," Vestnik statistiki, No. 6, pp. 3-8.
B. Dubnov 1986. "Soveshchanye po statistike agro-
promyshlennogo kompleksa," Vestnik statistiki, No.
9, pp. 67-71.
R. Ericson 1988. "The Soviet Statistical Debate:
Khanin vs. TsSU," paper presented at the March
1988 Hoover Institution-Rand Corporation confer-
ence on the defense sector and the Soviet economy,
Stanford University, p. 39.
V. Gorchak and A. Lobko 1979. Rynok tekhniches-
kikh-slozhnykh tovarov, Moscow.
S. Ippolitov 1987. "Kruglyy stol-ekonomicheskaya
effektivnost' proizvodstva: yeye opredelenye i analiz,"
Vestnik statistiki, No. 8, pp. 30-44.
G. Khanin 1981. "Al'ternativnyye otsenki resul'tatov
khozyaystvennoy deyatel'nosti proizvodstevnnykh ya-
cheyek promyshlennosti," Izvestiya akademii nauk
SSSR, seriya ekonomicheskaya, No. 6, pp. 62-73.
G. Khanin 1984. "Puti sovershenstvovaniya informat-
sionnogo obespecheniya svodnykh planovykh narod-
nozkhozyaystvennykh raschevtov," Izvestiya akade-
mii nauk SSSR, seriya ekonomicheskaya, No. 3, pp.
57-67.
G. Khanin 1986. "Sochtem fondy ... ," Sotsialisti-
cheskaya industriya, August 27, p. 2.
G. Khanin 1987. "V poiskakh boleye tochnykh ekono-
micheskikh otensok," Izvestiya sibirskogo otdeleniya
akademii nauk SSSR, No. 1, January, pp. 21-28.
V. Knyazevskiy, T. Kozlov, and N. Sheremet 1987.
"Po povodu lukavoy stat'i o tsifrakh," Vestnik statis-
tiki, No. 6, pp. 53-60.
M. Korolev 1986. "Chto znayet Statistika," Izvestiya,
December 10, p. 3.
M. Korolev 1987a. "Zadachi perestroyki statistiki,"
Vestnik statistiki, No. 4, pp. 3-12.
M. Korolev 1987b. "Statistika znayet vse-a chto
govorit?" Izvestiya, June 23, p. 2.
M. Korolev 1987c. "Statistika-instrument upravlen-
iya i glasnosti," Pravda, August 11, p. 2.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
M. Korolev 1987d. "O korennoy perestroyke raboty
organov gosudarstevennoy statistiki," Yestnik statis-
tiki, No. 9, pp. 3-16.
M. Markovich 1986. "Soveshchaniye nauchnykh ra-
botnikov po voprosam uluchsheniya dela statistiki v
svete reshenii XXVII syezda KPSS," Vestnik statis-
tiki, No. 10, pp. 63-67, and No. 11, pp. 62-64.
"Materialy ... " 1968. "Materialy vsesoyuznogo so-
veshchaniya statistikov 1968 goda," Vestnik statis-
tiki, No. 7, pp. 15-41.
I. Matyukha 1986. "Plan statisticheskikh i metodolo-
gicheskikh rabot TsSU SSSR na 1986 god," Vestnik
statistiki, No. 1, pp. 3-13.
I. Matyukha 1987. "Flan raboty organov gosudarst-
vennoy statistiki na 1987 god," Vestnik statistiki, No.
2, pp. 10-20.
I. Matyukha 1988. "Gosudarstvennyy zakaz na razra-
botku statisticheskoy informatsii v 1988 g.," Yestnik
statistiki, No. 2, pp. 3-14.
Ministerstvo finansov SSSR 1987. Gosudarstvennyy
byudzhet SSSR, 1981-85, Moscow.
"Ob itogakh ... " 1988. "Ob itogakh vypolneniya
gosudarstevnnogo Plana ekonomicheskogo i sotsial'-
nogo razvitiya SSSR v 1987 godu," Vestnik statistiki,
No. 2, pp. 15-26.
O korennoy ... 1987. O korennoy perestroyke uprav-
leniya ekonomikoy: Sbornik dokumentov, Moscow.
V. F. Petrovskiy 1987. "Vybor v pol'zu razoryzheniya
i razvitiya," Izvestiya, August 27, p. 4.
"Poryadok ... " 1988. "Poryadok wedeniya novoy i
peresmostra deystvuyushchey statisticheskoy otchet-
nosti v narodnom khozyaystve SSSR," Vestnik statis-
tiki, No. 1, pp. 19-21.
"Rasshirennoye ... " 1987a. "Rasshirennoye zase-
daniye kollegii ipartiyno-khozyaystvennogo aktiva
TsSU SSSR," Vestnik statistiki, No. 5, pp. 32-48.
"Rasshirennoye ... " 1987b. "Rasshirennoye zase-
daniye kollegii ipartiyno-khozyaystvennogo aktiva
Goskomstata SSSR," Vestnik statistiki, No. 10,
pp. 31-45.
R. Ryabushkin and A. Remizov 1987. "Napravleniya
razvitiya otchetnogo balansa narodnogo khozyayst-
va," Vestnik statistiki, No. 10, pp. 22-30.
V. Selyunin 1988. "Tempy rosta na vesakh potreblen-
iya," Sotsialisticheskaya industriya, January 5, p. 3.
V. Selyunin and G. Khanin 1986. "Pyl' v glaza,"
Pravda, December 30, p. 2.
V. Selyunin and G. Khanin 1987a. "Lukavaya tsifra,"
Novyy mir, No. 2, pp. 181-201.
V. Selyunin and G. Khanin 1987b. "Statistika znayet
vse?" Novyy mir, No. 12, pp. 255-257.
V. M. Simehera, ed. 1986. Ucheniye zapiski po
statistike. Problemy sotsial'noi statistiki, Vol. 50,
Moscow.
F. Skobtsova and G. Adler 1987. "Povysit' kachestvo
statisticheskikh publikatsiy," Vestnik statistiki, No.
5, pp. 52-56.
Z. Somova 1986. "Soveshchaniye rabotnikov otdelov
svodnoy statistiki i statisticheskoy informatsii," Yest-
nik statistiki, No. 8, pp. 60-62.
"Sovetskiy Soyuz ... " 1987. "Sovetskiy Soyuz v
mirovoy ekonomike (1917-1987 gg)," Mirovaya ekon-
omika i mezhdunarodnyye otnosheniya, No. 11, pp.
145-157, and No. 12, pp. 141-148.
Sovet Ekonomicheskoy Vzaimopomoshchi 1987. Sta-
tisticheskiy yezhegodnik strap-chlenov Soveta Ekon-
omicheskoi Vzaimopomoshchi, Moscow.
V. Starovskiy 1967. "Polveka sovetskoy statistiches-
koy nauki i praktiki," Vestnik statistiki, No. 10, pp.
17-35.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
"Statistika ... " 1987. "Statistika i sovershenstvovan-
iye upravleniya v narodnom khozyaystve," Vestnik
statistiki, No. 3, pp. 3-9.
V. Treml 1982. "Subsidies in Soviet Agriculture:
Record and Prospects," in US Congress Joint Eco-
nomic Committee, Soviet Economy in the 1980s, Part
II, Washington, DC., pp. 171-185.
V. Treml 1987. "Income from Private Services Rec-
ognized by Official Soviet Statistics," in M. Alexeev
et al., Studies on the Soviet Second Economy. Berk-
ley-Duke Occasional Papers on the Second Economy
in the USSR, Paper No. 11, December,
pp. 4.0-4.27.
V. Treml 1988. Notes on Soviet Agricultural Subsi-
dies, staff paper, Center for International Research,
Census Bureau, US Dept of Commerce, Washington,
DC (forthcoming).
Trezvost'... 1987. Trezvost' i kul'tura, No. 2,
p. 21 (untitled and unsigned box of selected undocu-
mented references).
V. Tsvelikh 1987. "Povysit' kachestvo statisticheskikh
publikatsiy," Vestnik statistiki, No. 12,
pp. 40-43.
"Uskorit' ... " 1988. "Uskorit' perestroyku deyatel'-
nosti i organizatsionnoy struktury statisticheskikh or-
ganov," Vestnik statistiki, No. 12, pp. 3-6.
"V Tsentral'nom ... " 1986. "V Tsentral'nom Komi-
tete KPSS," Vestnik statistiki, No. 6, pp. 3-4.
"V Tsentral'nom Komitete ... " 1986. "V Tsentral'-
nom Komitete KPSS," Ekonomicheskaya gazeta,
No. 24, June, p. 3.
Yu. Vasil'ev 1968. "Simpozium 'Sovetskaya Statistika
cherez 20 let,' "Vestnik statistiki, No. 11,
pp. 76-83.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
The New Look at Soviet
Statistics: Implications
for CIA Measures of the
USSR's Economic Growth
STAT
Introduction
The criticisms of Soviet statistics that have been
reviewed at this conference today obviously raise
questions about CIA measures of economic growth.
Unfortunately, glasnost, with some exceptions, has
not added a great deal to what we already know about
the deficiencies of these statistics. For production
statistics-the principal focus of my presentation this
afternoon-the major deficiencies are:
? First, the availability of economic data is still
extremely limited compared with the situation in
most developed countries. Moreover, in many in-
stances, we still know little about how the statistics
are collected and processed, or when changes in
definitions and methods occur.
? Second, statistics on value of output still seem much
more suspect than physical production statistics.
What Soviet critics have been saying reinforces
substantially what Westerners have argued for
years-that value statistics are flawed by the hidden
inflation resulting from new product pricing.
quality and his reorganization of statistics will intro-
duce adisconnect between pre-Gorbachev and post-
Gorbachev statistics-and analyze what this could
mean for our measures of Soviet growth.
Physical Versus Value Statistics
Turning first to the kinds of output statistics used in
CIA's GNP accounts, I would note that these data
can be differentiated according to the degree of
disaggregation they reflect. Table 1 shows some of the
possibilities. Consider the example of metal-cutting
machine tools. The growth of the total number of
units in a category is compared with that based on a
breakdown of these tools into types; each type is
valued at a representative price. Alternatively, there
are the official figures on total value of production of
metal-cutting machine tools in so-called constant
prices. When the measure supposedly represents the
summation of all or almost all ps times gs, it can be
called amodel-based measure. In the case of coal and
gas, I am stretching the definition a bit. Here, the
prices are the conversion coefficients for translating
tons and billion cubic meters of different kinds of coal
and gas into standard fuel units.
In this connection, although more is being written
about overreporting (and sometimes underreporting)
of production in physical terms, I do not have the
sense that the degree of distortion increased or de-
creased in the pre-Gorbachev decades. At any rate, I
will be talking mainly about the effects of inflation in
Soviet value statistics on CIA GNP accounts.
I want to review our estimates of Soviet GNP growth
and try to gain some appreciation of how much they
may be biased. First, I will illustrate how the selection
of production statistics can lead to bias. I will then
make a preliminary appraisal of the net bias in our
sector-of-origin and end-use measures. Finally, I will
discuss the possibility that Gorbachev's stress on
Some generalizations can be drawn from the compari-
sons. As might be expected, the disaggregated mea-
sures of production tend to grow faster than the
aggregated measures for entire categories-like the
number of all machine tools, production of passenger
cars, or total output of cement in tons. However, for
most fuels and basic materials, measures based on
entire categories do not behave very differently from
measures that break categories into types. This sug-
gests that if changes in the production profile are
important, they result from improvements within
types of products. Thus, the big differences are be-
tween measures based on a disaggregation by product
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 1
USSR: Comparison of Different Measures
of Industrial Production
Average annu
al percentage change
Metal-cutting machine tools
1951-65
1951-55
1956-60
1961-65
Category-units
6.7
10.7
5.9
3.6
Type-SPIDER avalues (1 July 1955 prices)
7.2
10.2
6.7
5.0
Model-Narkhoz values (1 July 1955 prices)
13.6
19.4
12.7
8.8
Tractors
1971-85
1971-75
1976-80
1981-85
Type-units
1.6
3.7
0.2
1.1
Type-horsepower (basis for SPIDER)
4.0
7.1
2.6
2.4
Model-value (based on sample)
2.7
5.6
t.7
0.9
Agricultural machinery
1966-82
1966-75
1976-82
Type-value (Treml)
2.8
3.0
2.5
Model-value (Narkhoz, 1 July 1967 prices)
7.4
9.0
5.1
Light automobiles
1951-85
1951-70
1971-85
Category-units
9.3
9.2
9.4
Model-value (SPIDER-1970 rubles)
10.0
14.3
4.6
Finished rolled ferrous metals
Category-units (million tons)
5.3
7.8
2.0
Type-value (SPIDER)
5.4
8.2
2.0
Coal
1956-86
1956-75
1976-86
Category-units
2.1
3.0
0.6
Model-standard fuel units
1.2
2.1
-0.4
Natural gas
1961-85
1961-80
1981-85
Category-units
1 1.2
12.0
8.1
Model-standard fuel units
11.0
11.9
7.6
Cement
Type-units (million tons)
7.6
11.8
2.1
Model-value (SPIDER)
8.1
12.6
2.3
Canned goods
1951-85
1951-70
1971-85
Category-units (million standard cans)
7.3
10.2
3.5
Type-value (SPIDER)
7.4
9.9
4.2
Processed meat
Category-units (thousand tons)
5.7
7.9
2.8
Type-value (SPIDER)
5.8
7.9
3.2
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 2
Bias in CIA Index of Soviet GNP by Sector of Origin:
A First Approximation
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 2
Bias in CIA Index of Soviet GNP by Sector of Origin:
A First Approximation (continued)
Branch and Sector
Measure
Assessment of Bias
Percentage Share
Under-
statement
About
Right
Over-
statement
of GNP at Factor
Cost in 1982
Metal structures
X
(0.2)
Sanitary engineering products
X
(0.1)
Machinery repair
X
(1.9)
Chemicals and petrochemicals branch
2.6
Mineral chemicals sector
X
(0.1)
Basic chemicals sector
X
(1.3)
Aniline dye products sector
X
(0.02)
Synthetic resins and plastics sector
X
(0.2)
Synthetic fibers sector
X
(0.2)
Organic synthetic products sector
X
(0.2)
Paints and lacquers sector
X
(0.1)
Pulp and paper sector
X
(0.3)
Wood chemicals sector
X
(0.02)
Construction materials branch
2.0
Other construction materials
X
(0.6)
Glass and porcelain sector X
(0.1)
Light industry branch
2.3
Cotton fabric sector
X
(0.6)
Silk fabric sector
X
(0.2)
Wool fabric sector
X
(0.2)
Linen fabric sector
X
(0.06)
Hosiery and knitwear sector
X
(0.2)
Sewn goods sector
X
(0.8)
Other light industry sector X
(0.2)
Processed food branch
X
2.6
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 2
Bias in CIA Index of Soviet GNP by Sector of Origin:
A First Approximation (continued)
Branch and Sector Measure
Assessment of Bias
Percentage Share
Under-
statement
About Over-
Right statement
of GNP at Factor
Cost in 1982
Dairy products sector
X
(0.4)
Sugar sector
X
(0.2)
Flour and cereal sector
X
(0.1)
Bread products sector
X
(0.2)
Confectionary products sector
X
(0,2)
Vegetable oils sector
X
(0.06)
Fruit and vegetable products sector
X
(0.09)
Other foods sector
X
(0.8)
Construction
X
8.0
Agriculture
X
20.2
Transportation
X
9.9
Communications
X
0.9
Trade
X
6.5
Services
I8.5
Housing
X
(4.7)
Utilities
X
(1.2)
Repair and personal care
X
(1.6)
Recreation
X
(0.9)
Education
X
(3.4)
Health
}(
(1.8)
Science
X
(1.9)
Credit and insurance
X
(0.3)
Administration and miscellaneous
X
(2,6)
Military personnel
X
2.0
Other branches
X
0.6
type and those based on the disaggregation by model Bias in the Sector-of-Origin Accounts
implied in the Soviet value series-e.g., machine tools I want to demonstrate my sense of the direction of
and agricultural machinery. But I would note that bias in the individual branch and sector indexes in the
where we have tried to disaggregate machinery pro- GNP by sector-of-origin accounts (see table 2). In
duction by model (that is, for light automobiles, general I think the bias is negligible in most of the
trucks, and buses), the model-based measure does not
behave much differently from the measures based on
total unit production.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
fuels and basic materials branches of industry. How-
ever, some exceptions were revealed in the new Soviet
statistical handbook (Narkhoz) for 1986: coal and gas.
In my opinion, there are only two machinery sectors
in which bias is not, prima facie, a problem. These are
the automotive sector and military machinery, where
we build up indexes of value, model by model. The
other machinery sectors are represented by indexes
that probably either understate growth because they
use physical measures to move values for a given type
of equipment or overstate growth because they use
Soviet value indexes.
In most of the other major sectors of origin-con-
struction, agriculture, transportation, communica-
tions, and trade-I assume, as a first approximation,
that there is little or no net bias in the indexes of
growth. But many of the services are represented by
indexes based on employment or, in the case of
housing, on the growth of square meters of housing.
The employment-based indexes very likely do not take
account of productivity gains, while growth in housing
stock does not capture quality change.
To obtain an initial sense of the net effect of the
biases, I first used a -1, 0, and + 1 scoring scheme-
that is, a -1 for negative bias, 0 for no bias, and ~-1
for positive bias. If one interprets these values as
percentage points of average annual growth and
weights them by their share of GNP at factor cost, the
net result would be a downward bias in CIA's GNP
measure of about one-tenth of a percentage point.
This, of course, is not a very satisfying answer. Going
a step further, one can try to attach some more
plausible values to the degree of understatement and
overstatement. Consider first the metals sector of
industry. I accept the current estimates for nonferrous
metals, which are based on the index of GVO for the
Ministry of Nonferrous Metals. However, I allow for
the possibility that the ferrous metals index may
understate growth by as much as 1 percentage point
per year because of a failure to capture changes in
assortment and quality within product types like cold
rolled sheets and seamless pipes and tubes. Next are
the fuels branches. We can substitute the recently
released information on the production of coal and gas
in standard fuel units for the physical production
measures we have been using.
Machine building and metalworking (MBMW), of
course, is the toughest call. The first approach that I
tried involves abranch-level correction based on a
comparison of factor productivity growth. in MBMW
with productivity growth in other sectors where infla-
tion in product prices is not likely to be much of a
factor. The underlying premise is that labor and new
machinery should be, on average, about as productive
in nonmachinery sectors as in MBMW. It turns out
that factor productivity in MBMW increased about
1.3 percentage points faster than a simple average of
factor productivity in electric power, ferrous metals,
construction materials, and light industry between
1965 and 1975. In 1976-80, this gap narrowed to six-
tenths of a percentage point. In 1981-85, however,
factor productivity in these other four sectors grew
six-tenths of a percent faster than factor productivity
in MBMW. Since the measure of MBMW output is
CIA's measure, one might conclude that in the late
1960s and early 1970s it may exaggerate growth of
MBMW by more than a percentage point and since
1975 by half a percentage point or even less.
I also tried to take a more disaggregated approach to
evaluating the bias in the machinery index, relying on
some work done a few years ago by Professor Vlad
Treml. He compiled an index of civil machinery
production based entirely on samples reported in
physical units for various sectors of MBMW. For a
number of sectors where we also rely on physical
production, our estimates of growth are not very
different. But our indexes are very different for those
sectors of MBMW for which we use value of output
as reported by the Soviets-almost 40 percent of all
MBMW. On the argument that the physical series
understate growth because they do not reflect quality
change while the value series are flawed by hidden
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
inflation, I simply split the differences between
Treml's and the CIA's rates of growth and substituted
the adjusted indexes for those that are now in
SPIOER. For this part of MBMW, the adjusted
series grew by 6.4 percent per year between 1965 and
1982, compared with the 9.2-percent-per-year rate for
this portion of the present SPIOER sample. Carried
over to the entire machinery sample, the adjustment
cuts the rate of growth of value added in MBMW
from 5.3 to 4.7 percent for the same period.
This excursion suggests an overstatement of CIA's
index of MBMW growth of about half a percentage
point. This is, of course, before any account is taken of
the likely understatement of growth in other machine-
building sectors where we rely on physical measures
of output. Applying this discount factor to the whole
period, 1950 to 1986, reduces MBMW growth from
5.8 percent per year to 5.0 percent per year. This may
not seem like much, but the effect is to lower CIA's
estimated value for MBMW output in 1986 by 24
percent.
Because the degree of inflation in MBMW output is a
controversial issue, it is worth trying still another
approach to try to appraise its significance. Our
analysis of the Soviet literature and some of our own
research 'suggests strongly that inflation in the price
of new products is the preeminent cause of inflation in
overall machinery prices. I asked some of my col-
leagues to model the relationship between the extent
of inflation in new product prices, the share of new
products in total output, and average inflation in a
sector. To repeat, the assumption is that a new
product comes in at a price greater than is warranted
by its productivity, durability, labor-saving potential,
and the like, and that the new price becomes the
comparable price in the Soviet index of gross or net
output. Thereafter, the comparable price is retained
until a general price revision comes along.
The results of the modeling exercise are shown in
table 3. The values in the table indicate, for example,
that if the share of new products in output is 3
Table 3
The Dependence of Overall Inflation
on the Extent of New Product
Price Inflation and the Share of New Products
in Total Output
Overall Price Share of New Products in Total Output for a Given
Inflation Year
1 percent
per year
2 percent
per year
3 percent
per year
4 percent
per year
5 percent 2,000 385 213
per year
percent, then the level of new product prices would
have to be 1.49 times the level of existing prices to
generate an overall inflation rate of 1 percent. This
year's Narkhoz reports the share of products assimi-
lated for the first time in the machine-building com-
plex as a share of total tovarnaya produktsiya. The
shares were 4.5 percent in 1970, 3 percent in 1980, 3
percent in 1985, and 4.5 percent in 1986. How these
shares are calculated is somewhat a mystery. The
share, for example, might not count one-time orders
or some kinds of batch production. Although the
exercise can only be suggestive, it does suggest to me
that it would be hard to defend an estimate of the
hidden inflation component of Soviet value statistics
for MBMW of more than 2 to 3 percent per year,
which is roughly what my 1-percentage point discount
factor for MBMW implies.
STAT
..,~,.
Robert E. Leggett, "Measur- 5 ~ ~..~ ~
ing Inflation in the Soviet Machine Sector, 1960-1973," Journal of
Comparative Economics, vol. 5 (June 1981), pp. 169-84.
3 percent 4 percent 5 percent 6 percent 7 percent
49 33 25 20 16
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 4
USSR: Alternative
Estimates of Growth
of Value Added
in Services, 1951-84
Table 5
USSR: Alternative
Prell
CIA
Difference
Housing
6.0
3.2
2.8
Education
4.2
3.1
1.1
Health
4.7
3.2
1.5
Science
5.9
6.6
-0.7
Government
2.0
1.3
0.7
Municipal services
5.2
3.2
2.0
For chemicals, I simply assume a notional average
understatement of 1 percentage point per year for the
aniline dye, rubber products, and the three synthetics
sectors and a 1-percent overstatement for the one
Soviet value series-mineral chemicals. The net result
is an implied understatement of growth in chemicals
output of three-tenths of a percentage point. Not
much happens, largely because the sector adjustments
cover only about 25 percent of value added in
chemicals.
In other industrial branches, several sector indexes
suspected of being biased have so tiny a weight
(construction ceramics in construction materials and
confectionary products in processed foods) that I
ignored them. For other sectors, where I have no way
of judging the degree of bias, an arbitrary 2-percent-
age-point adjustment was made to the growth rates
presently used.
This brings me to the service sectors, where-as I
noted earlier-we have been accused of ignoring
productivity advances. Here I can take advantage of
some work carried out by Mark Prell while writing a
Ph.D. dissertation for Professor Martin Weitzman at
Massachusetts Institute of Technology. Essentially,
he estimated new indexes of output for housing and
other services. For housing, he uses the value of
housing stock as reported by the Soviets rather than
square meters of housing. For education, health,
Estimates of Growth of GNP
by Sector of Origin
at Factor Cost, 1951-86
Current CIA Adjusted
Estimate Estimate
a Construction, transportation, communications, trade, military
personnel, and other branches.
science, municipal services, and government he con-
structs indexes based on a weighted average of labor
and capital inputs. His estimates and CIA's are
compared in table 4. I believe that the implied
annnual growth in the value of a square meter of
housing construction is too high, so I take the average
of the Prell and CIA growth factors. But for the rest
of the services in question I simply replace the CIA
estimates with Prell's.
The end result of this primitive round of sensitivity
testing is not very exciting. The comparisons of the
sector-of-origin estimates that the CIA now carries
with my adjusted estimates are set out in table 5. The
figures in the column representing the current CIA
estimate reflect the values for' GNP in 1950 and 1986
at factor cost in 1982 prices. To obtain the adjusted
estimate, I first took the values for 1950 in our
current estimate as sector weights and applied adjust-
ed indexes in the instances I have indicated to obtain
adjusted values for 1986. As you can see, the net
effect of the adjustments is to raise the rate of growth
of GNP slightly as industrial growth is reduced and
that of services raised. If one accepts the adjustment
for industry and leaves services alone, GNP growth
declines-but only to 3.8 percent per year.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Bias in the End-Use Accounts
I have carried out the same kind of exercise for the
end-use accounts. Here again, I believe we are most
likely to understate growth when we track service
activity by some kind of physical index-like housing
stock--0r employment plus material inputs, which are
the basis for measuring trends in education, health,
and administration. As I noted earlier, we do not in
these instances allow for improvements in productivity
related in part to the rising stock of fixed capital per
worker in these sectors.
For the food and soft goods components of consump-
tion, and for the other services, the CIA's subindexes
are based on Soviet reported per capita consumption,
physical production, and deflated retail sales. The
data on per capita consumption and physical produc-
tion may not capture some quality change, but I do
not believe that this is important in food and soft
goods. Deflated retail sales, which rely on published
Soviet price indexes, are likely to reflect some hidden
price inflation, but for the moment I ignore it.
The indexes for the various components of defense are
our own and rely on what we believe is an almost
complete enumeration of Soviet activities, each with
its own price weight. I would not exclude the possibili-
ty of hidden quality change here, but I am not sure of
its direction. For example, the average educational
attainment of military personnel has increased, but
the complaints about the dedication and toughness of
the armed forces parallel what has been said about
industrial workers.
The most serious danger of overstating growth arises,
I believe, in the machinery sector-consumer dura-
bles, the machinery and equipment component of new
fixed investment, and capital repair. The index for
consumer durables is based on deflated retail sales,
that for investment in machinery and equipment relies
on the Soviet constant price series, and the index for
capital repair is derived by deflating reported Soviet
data in current prices. Indeed, the problem is poten-
tially worse on the end-use side than it is on the
sector-of-origin side because the end-use indexes do
not include a deflationary component like the physical
production series in the machinery production index.
My sense of where the biases may be on the end-use
account is set out in table 6; the trick is to weight
these various biases in some reasonable fashion. My
approach was much like that taken on the sector-of-
origin side. Instead of a consumer durables index
based on the official tovary series, I assumed that the
degree of inflation in this series was equal to half the
difference between the average growth in Treml's
indexes for producer durables, which are based on
physical production, and our indexes for equivalent
sectors, which are based on reported ruble values of
output. For the period 1950 to 1986, this worked out
to a 5.6-percent instead of a 9.6-percent average
annual rate of growth. For the machinery and equip-
ment component of investment I used the same
discount factor that was applied for the production of
machinery and equipment. This reduced growth of
this end-use component from 9.2 percent to 6.4
percent. Similarly, the discount for growth of capital
repair is carried over from the capital repair sector of
MBMW output. Finally, the adjusted indexes for end-
use services were obtained by applying the same
adjustment factors that were used on the sector-of-
origin service accounts.2
The results of this exercise can be seen in table 7.
Remember that the rates of growth for GNP are the
same as those in the table for GNP by sector of origin
because the sector-of-origin estimates provide the
control totals for GNP. The main differences between
CIA's current estimate and the adjusted estimate lie
in (a) a somewhat slower growth for consumer goods
in the adjusted estimate because of slower growth of
consumer durables, (b) substantially faster growth of
consumer and government services, and (c) an appre-
ciably slower growth of new fixed investment and
capital repair. The effect of these adjustments to
CIA's GNP measures is not negligible, however. If
these differential rates of inflation had indeed pre-
vailed over the whole period from 1950 to 1986, the
' More work needs to be done here, because the understatement of
growth of the materials components of the various services is not
likely to be as large as that for the value-added part.
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 6
Bias in CIA Index of Soviet GNP by End Use:
A First Approximation
Under- About Over-
statement Right statement
Outlays, not elsewhere counted (including inventory change 0.5
and net exports)
Percentage Share
of GNP at Factor
Cost in 1982
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 7
USSR: Alternative
Estimates of Growth
of GNP by End Use at
Factor Cost, 1951-86
Table 8
USSR: Alternative Estimates and
Distribution of GNP by End Use at
Factor Cost, 1986
Current CIA
Estimate
Adjusted
Estimate
Consumption
3.8
4.0
Consumer goods
3.7
3.4
Consumer services
4.1
5.0
Investment
7.1
6.2
New fixed investment
6.6
5.6
Capital repair
10.3
9.2
Defense
2.8
2.8
Other government
expenditures a
2.4
3.2
Residual n
~
4.3
GNP
3.9
4.0
a Administration, other services, and civilian research and
development.
n Includes net exports, inventory change, and statistical
discrepancy.
Declines from 8.5 billion rubles to -1.2 billion rubles.
distribution of GNP by 1986 would have been quite
different. The composition of consumption shifts
markedly, for instance, and investment's share of
GNP drops by almost one-third (see table 8).
Even more striking would be the implications of the
adjustments for how we view factor productivity in
the Soviet economy or-even before we can do this-
how we estimate factor cost weights for our measures
of GNP by sector of origin and end use. To demon-
strate, asimple exercise looking at the stock of
machinery and equipment (M&E) over time by means
of the perpetual inventory method was undertaken.
Investment was assumed to enter the stock, retain its
initial value through time, and then drop out after 20
years. We further assumed an average rate of growth
of investment of 7.6 percent, the value the Soviets
report as the average annual growth of the machinery
and equipment component of new fixed investment
from 1950 to 1986. The behavior of the capital stock
of M&E under alternative assumptions regarding the
inflation component in new fixed investment in M&E
is shown in table 9 and 10. We look at two end points:
Current CIA
Estimate
Adjusted
Estimate
Consumptio^
52.1
53.7
Consumer goods
34.0
29.8
Consumer services
18.1
23.9
Investment
29.2
21.2
New fixed investment
23.4
17.2
Capital repair
5.8
3.9
Defense
15.8
] 5.2
Other government
expenditures a
3.1
3.8
Residual b
-0.2
6.1
GNP
100.0
100.0
a Administration, other services, and civilian research and
development:
b Includes net exports, inventory change, and statistical
discrepancy.
letting the model run for 20 years and, alternatively,
for 40 years. The results simply illustrate the pro-
found effect on the value of the capital stock if we
conclude that the official investment numbers have an
inflation component of as little as 2 percent.
Comparability of Soviet Statistics:
Pre- and Post-1985
Finally, let me turn to glasnost and its effects on our
probably fair to say that it has not been a bonanza so
far. We live in hope, however. In this connection, the
promised publication of some of the specialized statis-
tical handbooks will be welcome after a lapse of so
many years. And, of course, we will be greatly
interested to see what the USSR decides to reveal
about its defense budget. Will the revelations put us
STAT
SCAT
STAT
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 9
Deflated and Undeflated Growth in Capital Stock
With Initial Investment of 5 Billion Rubles a
Rate of Inflation Billion rubles
in Percent K K*
a. Capital stock after 20 years
(no retirements)
Where:
1. Nominal average annual growth of investment = 7.6 percent.
2. Initial stock of capital = 100 billion rubles.
3. K = undeflated capital stock and K* =deflated capital stock.
out of work or become another area of dispute
between the United States and the Soviet Union?
have a problem this year and perhaps in the next few
years in evaluating Soviet economic performance. I
would think, though, that after a relatively brief
interval a new equilibrium would be established.
Overreporting should stabilize, probably at a lower
level, although Gorbachevian pressure to fulfill the
plan while undergoing reform encourages enterprises
to overstate their performance. Similarly, state in-
spection will be extended throughout industry, and
enterprises are likely to find a fairly stable combina-
tion of compliance and evasion.
The other effects of perestroyka, however, are still
very hard to figure out. Perhaps the most important
potential effects are (a) the wringing of some of the
water out of the statistics reported up through the
statistical hierarchy, (b) the impact of the campaign
for higher quality on both physical production and
value statistics, (c) the possibility that some produc-
tion of goods and services that used to go unreported
will now be reported under the new regulations on
private and cooperative activity, and (d) the effect of
self-financing and a partial decentralization of pricing
decisions on Soviet value statistics.
Because of stricter controls over reporting and state
inspection, the level of output reported may be less
than would have been reported in the pre-Gorbachev
era. We have seen a good deal of evidence, for
example, that Gospriyemka has held production
down, especially in machine building. Therefore, we
Another new development-the new encouragement
to private activity-works the other way. To the
extent that activities that had been carried out in the
second economy are brought into the open, statistical
reporting will give an inflated picture of the growth,
particularly in the service sector. In a year or so,
however, the Soviets are likely. to have published
enough about the results of the new decrees to give us
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Table 10
Deflated and Undeflated Growth in Capital Stock
With Initial Investment of 10 Billion Rubles a
a. Capital stock after 20 years
(no retirements)
a Whers;:
1. Nominal average annual growth of investment = 7.6 percent.
2. Initial stock of capital = 100 billion rubles.
3. K = undeflated capital stock and K* =deflated capital stock.
an idea of how much of a shift has occurred. And, as
with tighter supervision of reporting, a new balance
between the second and the first or open economy is
likely to be struck.
measures, as the Soviets did in the case of cotton. Our
best hope is that Goskomstat will provide more statis-
tics on production. We need to be able to disaggregate
our samples so that we can have more based on
individual models rather than on broader categories.
In other words, there is no quick fix to dealing with
distortion in Soviet statistics. It requires a substantial
investment in research time and some luck in getting
access to more data.
What the combination of conversion to self-financing
and some decentralization of pricing decisions will
mean for Soviet data in so-called comparable prices is
a much tougher problem. Certainly, the Soviets worry
about enterprises pushing up prices. My hypothesis,
nonetheless, is that in a long-term perspective,
changes in the degree of exaggeration in Soviet
reporting will not make much difference for our
measures of GNP.
I would like to conclude with some observations about
improving. CIA's GNP estimates. First, we should
assume that the Soviets are not going to go back and
correct their statistics for distortions that arise from
the pricing mechanism. We may, however, see some
isolated instances of changing series given in physical
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
 Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9
Declassified in Part -Sanitized Copy Approved for Release 2012/12/19 :CIA-RDP89T014518000400480001-9