In 1953, Wassily Leontief made the then startling discovery that the empirical structure of American factor proportions generated in bilateral exchange with the rest of the world contradicted the theoretically anticipated result based on the Heckscher-Ohlin hypothesis. Using the logic of general equilibrium analysis in a two-factor neoclassical model, the Heckscher-Ohlin theorem predicted that America, being relatively abundantly endowed with capital, would produce export goods that intensively embodied capital and import replacement goods intensively embodying the relatively scarce factor labor. By applying the apparatus of input-output analysis, Leontief showed that paradoxically American factor proportions exhibted precisely the opposite relationship to the one anticipated on the basis of the Heckscher-Ohlin theorem. Leontief's finding initiated an avalanche of theoretical and empirical research directed towards unravelling the causal factors responsible for this unexpected and perverse result. Robert Baldwin in a recent article surveyed the evolution of the factor proportions controversy, and since all students of international trade can be expected to be reasonably familiar with the issues at stake, we shall not review them here.

Abram Bergson (1987a) has recently attempted to show that socialist economic systems are underproductive and underefficient.' Close evaluation, however, reveals that this conclusion is too sweeping. Using the USSR as a case in point, this comment demonstrates that Soviet productivity growth using Bergson's data (1987b) and methods (Bergson, 1978) matched Western Europe's and exceeded the U.S. achievement in 1960-75. Moreover the pronounced duality between Soviet industrial and nonindustrial productivity discovered by Bergson for 1960 appears to have increased thereafter through 1975, according to the statistics of the United States Central Intelligence Agency (CIA) similar to those employed in Bergson (1987b). This duality is confirmed further by a separate calculation of comparative machine-building productivity. Using CIA sector of origin data that are consistent with Bergson's counterpart end use data (Imogene Edwards et al., 1979) and his factor share weights (Bergson, 1978), Soviet machine-building productivity is calculated to be 79 percent of the American level in 1975 (Table 1). Revised data for the same year, published by the CIA in 1982, raise this figure to 88 percent. These findings clearly suggest that socialist economies need not be comprehensively underproductive (underefficient). Although they do not dispose of many issues of legitimate controversy2 and pose serious questions about the potential shortcomings of the adjusted data on which Bergson relies, they do demonstrate nonetheless that the positivist debate over the comparative productivity (efficiency) of socialism even in the Soviet case still cannot be completely laid to rest.3