1. This statement of the “standard” hypothesis is somewhat oversimplified. There are in fact several versions, which differ from one another mainly because of differences in the way in which the “agricultural sector” has been defined. Alec Nove, for example, considers primarily the relation between the peasantry and the state in An Economic History of the US.S.R. (London, 1969), pp. 148-86. Alexander Erlich adopts essentially the same approach in The Soviet Industrialisation Debate, 1924-1928 (Cambridge, Mass., 1960), pp. 119-21. However, the main concern of most “economists has been the relation between the agricultural sector taken as a whole and the nonagricultural sector, and this is the form in which the hypothesis has passed into the general literature on economic growth. See, for example, Baran, Paul, The Political Economy of Growth (New York, 1962), pp. 266–71 Google Scholar; Nicholls, William H.,” The Place of Agriculture in Economic Development,” in Eicher, Carl K. and Witt, Lawrence W., eds., Agriculture in Economic Development (New York, 1964), pp. 22–24 Google Scholar; F. Johnston, Bruce and W. Mellor, John, “The Role of Agriculture in Economic Development,” American Economic Review, 51, no. 4 (September 1961): 579 Google Scholar; Wilber, Charles K., “The Role of Agriculture in Soviet Economic Development,” Land Economics, 45, no. 1 (February 1969): 87–96 CrossRefGoogle Scholar. For a Soviet view see Politicheskaia ekonomiia: Kommunisticheskii sposob proisvodstva (Moscow, 1963), p. 286. For an excellent statement and analysis of the standard hypotheses see Campbell, Robert W., The Soviet-Type Economies Performance and Evolution, 3rd ed. (Boston, 1974), pp. 62–76 CrossRefGoogle Scholar, which appeared after this manuscript had already been drafted.

2. Karcz, Jerzy F., “From Stalin to Brezhnev: Soviet Agricultural Policy in Historical Perspective,” in Millar, James R., ed., The Soviet Rural Community (Urbana, 1971), esp. pp. 37–60 Google Scholar; Fallenbuchl, Z. M., “Collectivization and Economic Development,” Canadian Journal of Economics and Political Science, 33, no. 1 (February 1967): 1—15 CrossRefGoogle Scholar; Millar, James R., “Soviet Rapid Development and the Agricultural Surplus Hypothesis,” Soviet Studies, 22, no. 1 (July 1970): 77–93 CrossRefGoogle Scholar; R. Millar, James and A. Guntzel, Corinne, “The Economics and Politics of Mass Collectivization Reconsidered: A Review Article,” Explorations in Economic History, 8, no. 1 (Fall 1970): 103–16.CrossRefGoogle Scholar

3. Barsov, A. A., Batons stoimostnykh obmenov mezhdu gorodom i derevnei (Moscow, 1969)Google Scholar, and “Sel'skoe khoziaistvo i istochniki sotsialisticheskogo nakopleniia v gody pervoi piatiletki (1928-1933),” Istoriia SSSR, 1968, no. 3, pp. 64-82 (hereafter Barsov 1969 and Barsov 1968 respectively). Barsov has evidently exploited fully the statistical data available in published sources pertaining to the period, and the bibliography that may be compiled from his footnote citations is essentially definitive. However, and more important, Barsov extensively uses new data series not previously available in published form, derived from the Tsentral'nyi gosudarstvennyi arkhiv narodnogo khoziaistva SSSR (esp. fonds 1562, 4372, and 7733), which was founded in 1961. For a brief description of this new archive see Grimsted, Patricia K., Archives and Manuscript Repositories in the USSR: Moscow and Leningrad (Princeton, 1972), pp. 133–34.Google Scholar

4. There are two related but separable issues involved in the standard hypothesis on the contribution of Soviet agriculture. One pertains to the actual net flow of real resources between agriculture and nonagriculture (however defined) and is pertinent to the questions addressed by the typical two-sector growth model. The other has to do with the welfare implications of the net flow—for example, for the agricultural population. Subject to the qualifications set forth below in the text and notes, the first issue lends itself to straightforward empirical measurement. The second, however, involves a choice of value standard and must therefore be answered using data (such as hypothetical price weights) not contained in the actual historical record. Since Barsov’s choice of value standard is not likely to gain acceptance among Western students of the period, his principal contribution consists in the collection and reconstruction of the actual data series, and therefore is pertinent to the first issue. Thus this review is focused on the question of the measurement of the net resource contribution of agriculture, as opposed to the issue of peasant welfare.

5. This is sometimes referred to as the index number problem, or, more technically, the formula error. The comparison of any two years, for example, involves two sets of quantities and two sets of prices. The quantity change may be measured using as weights prices in either the base or the given year. Ordinarily, the two alternative formulas will not yield answers that agree precisely, and the difference can be quite substantial. However, since the two formulas stand on an equal footing, there is no basis for choosing between them. Obviously, some other set of price weights, drawn from a third time or place, will be likely to yield a third answer.

6. For a detailed explanation see Millar, “Soviet Rapid Development,” esp. pp. 87-

7. Golev, la. I., Sel'skokhoziaistvennyi kredit v SSSR (Moscow, 1958), pp. 19–21 Google Scholar; D'iachenko, V. P. et al., 50 let sovetskikh finansov (Moscow, 1967), pp. 49–50 Google Scholar; Barsov (1969), p. 82.

8. Malafeev, A. N., Istoriia tsenoobrazovaniia v SSSR (1917-1963 gg.) (Moscow, 1964), pp. 131–32, 172-73Google Scholar; Atlas, M., Razvitie gosudarslvennogo banka SSSR (Moscow, 1958), pp. 129–30 Google Scholar; Barsov (1969), pp. 115-24.

9. It would be extremely useful to be able to calculate the trade surplus of agriculture using price weights drawn, say, from 1932, because the size of the discrepancy between it and the 1928 price-weight index could serve as a measure of reliability. Unfortunately, Barsov does not provide price data for the years following 1928. It is doubtful, however, that the discrepancy would be larger than exists between the 1913 and the 1928 price-weighted indices.

10. As was indicated above, different sectoring criteria and different price weights may be expected to yield different concepts and measures of the “contribution” of agriculture. It is possible, therefore, that what I have called the standard interpretation of the role of agriculture may be valid for some subset of sectoring criteria and price weights. However, in my opinion the sectors designated by Barsov are appropriate to an attempt to measure the contribution of the agricultural sector to economic development, conceived within the framework of a two-sector growth model. In addition, 1928 price-weights are all that are available from the actual historical period.

11. Much depends, of course, on the reliability of the wholesale price index for “means of production,” which indicates constant prices on equipment and materials directly supplied by the state to sovkhozes, the MTS, and for state-financed purchases only to kolkhozes (see note b, table 3). One of the (anonymous) referees of this paper has called attention to the relatively high prices of tractors in 1928. He has also suggested that this index may have been applied (inappropriately) to construction costs and other elements where price controls would have been much less effective than for equipment and materials. The latter criticism, if correct, implies that the favorable change in the terms of trade for agriculture may be overstated, and possibly substantially so. There does not appear to be any way to resolve this question directly. However, as I pointed out in the text above, the available data on transfers, financial flows, and the change in the terms of trade are consistent with a net import surplus for the agricultural sector. Moreover, the generally high quality of Barsov’s works provides another reason for placing confidence in the sign of the change in the terms of trade, if not the precise magnitude of the change.

12. Karcz, “From Stalin to Brezhnev,” p. 42.

13. Millar, “Soviet Rapid Development,” pp. 78-82.

14. See, for example, Erlich, Alexander, “Stalin’s Views on Economic Development,” in Simmon, Ernests, ed., Continuity and Change in Russian and Soviet Thought (Cambridge, Mass., 1955).Google Scholar

15. Gross agricultural output exceeded the precollectivization 1928 level only in 1937 and 1940, according to Soviet official figures; see, for example, Narodnoe khoziaistvo SSSR v 1958 godu (Moscow, 1959), p. 350. Western estimates have presented an even bleaker picture; see Kahan, Arcadius, “Soviet Statistics of Agricultural Output,” in Laird, Roy D., ed., Soviet Agricultural and Peasant Affairs (Lawrence, Kans., 1963), pp. 134–60.Google Scholar

16. Barsov (1969), pp. 84, 87, 90; Iu. V. Arutiunian, “Osobennosti i znachenie novogo etapa razvitiia sel'skogo khoziaistva SSSR,” collected in Istoriia sovetskogo krest'ianstva i kolkhoznogo stroiteV stva v SSSR (Moscow, 1963), p. 409. According to Arutiunian, average annual agricultural output per capita changed as follows (1926-29 = 100): 90.3 in 1913; 100.0 in 1926-29; 86.8 in 1930-32; 90.0 in 1938-40; and 94.0 in 1950-53.

17. Iu. A. Moshkov, “Zernovaia problema v gody kollektivizatsii sel'skogo khoziaistva,” in Istoriia sovctskogo krest'ianstva i kolkhoznogo stroitel'stva v SSSR, p. 272; Barsov (1969), p. 85, n. 53.

18. Istoriia Velikoi Otechestvennoi voiny Sovctskogo Soiuza, 1941-1945, vol. 6, p. 43.

19. It has frequently been argued that mass collectivization did at least serve to ensure an adequate flow of labor out of the countryside to fill the growing number of industrial occupations. It would be inappropriate to include the net flow of labor between agriculture and the nonrural sector as they are defined by Barsov, for it would involve double-counting where sectors are identified by type-of-product criteria. It would, of course, be suitable where geographical criteria are used. However, as an economic rationale for collectivization, the mobilization-of-labor argument is without force. In the first place, there is no evidence to suggest that the supply of labor was deficient prior to the initiation of collectivization. In the second place, it is clear that collectivization encouraged an excessive off-farm flow of labor and population, and its continuation at the present time remains one of the most intractable obstacles to the modernization of Soviet agriculture today. On this subject see Norton T. Dodge, “Recruitment and the Quality of the Soviet Agricultural Labor Force,” in Millar, Soviet Rural Community, pp. 180-213.

20. It must be remembered, of course, that we remain essentially in the dark with respect to the balance of trade during the Second Five-Year Plan, and, according to Barsov, who has seen the archival material available for that period, we are likely to remain ignorant. Very little information apparently exists, and what is to hand is evidently insufficient to permit the reconstruction of the necessary price and physical-volume time series (Barsov 1969, pp. 186-90). Consequently, the force of this conclusion obtains (with the reservations noted above) primarily for the period 1928-32. However, Barsov's findings do appear to be consistent with Bergson's findings for the 1937 data, which suggest that agriculture contributed considerably less to Soviet capital formation than had previously been supposed. See Bergson, Abram, The Real National Income of Soviet Russia Since 1928 (Cambridge, Mass., 1961), p. 257 Google Scholar. On this same point see also Karcz, “From Stalin to Brezhnev,” pp. 48-51.