An earlier version of this article was presented at the 1968 meetings of the Western Economic Association in Corvallis, Oregon. The author acknowledges the helpful comments of Professor James Tattersall in the preparation of the early drafts.

¹ The original statement of this thesis appeared in Gerschenkron's, “Economic Backwardness in Historical Perspective,” in Hoselitz, Bert F. ed., The Progress of Underdeveloped Areas (Chicago: University of Chicago Press, 1952), pp. 3–29.Google Scholar At that time not all these aspects of development were included. The hypothesis appeared in the above form in Alexander Gerschenkron's, “The Early Phases of Industrialization in Russia: Afterthoughts and Counterthoughts,” in Rostow, W. W. ed., The Economics of Take-off into Sustained Growth (New York: St. Martin's Press, 1963), pp. 151–69.Google Scholar Gerschenkron's original essay, along with others relevant to his hypothesis, was reprinted in his Economic Backwardness in Historical Perspective: A Book of Essays (New York: Frederick A. Praeger, 1962). The hypothesis was again restated by Gerschenkron, in “The Discipline and I,” Journal of Economic History, XXVII (Dec. 1967), 443–59.CrossRefGoogle Scholar

² To my knowledge this is the first attempt to make a quantitative test of Gerschenkron's hypothesis. The specific measures of backwardness selected are my choices, and the method of calculating relative backwardness is based upon my interpretations of the hypothesis. Were the study conducted by Gerschenkron, the results might well be quite different. For two other discussions of the hypothesis, see Rosovsky, Henry, Capital Formation in Japan 1868–1940 (New York: The Free Press of Glencoe, 1961), pp. 55–104Google Scholar; and Landes, David S., “Technological Change and Industrial Development in Western Europe, 1750–1914,” in The Industrial Revolutions and After (I), Vol. VI of The Cambridge Economic History of Europe, Habakkuk, H. J. and Postan, M., eds., 6 vols. (Cambridge, ]Eng.[: The University Press, 1965), pp. 587–91.Google Scholar

³ While the propositions concerning the level of consumption and size of plant and enterprise are quantifiable, the lack of adequate data prevents their inclusion in the tests.

⁴ Gerschenkron, “Early Phases,” p. 163.

⁵ It is not clear that France experienced a spurt as defined by Gerschenkron. Gerschenkron's use of France in the development of his hypothesis, however, justifies its inclusion in the tests. England is not included because Gerschenkron's hypothesis concerns follower countries that could make extensive use of borrowed technology from more advanced countries. England, as the pioneer of industrial growth, could not be a follower.

⁶ Gerschenkron, Alexander, “Notes on the Rate of Industrial Growth in Italy, 1881–1913,” Journal of Economic History, XV (Dec. 1955), 365.Google Scholar Reprinted in Gerschenkron, Economic Backwardness, pp. 72–89. Alexander Gerschenkron, “Patterns and Problems of Russian Economic Development,” in Gerschenkron, Economic Backwardness, pp. 119–51. Reprinted from Black, C. ed., The Transformation of Russian Society (Cambridge: Harvard University Press, 1955).Google Scholar

⁷ Rostow, W. W., The Stages of Economic Growth (Cambridge, ]Eng.[: The University Press, 1960), Table I, p. 38.Google Scholar

⁸ Gerschenkron, comparing the concepts of the take-off and the great spurt, notes that they both stress discontinuity in the growth process, but that ‘… great spurts are confined to the area of manufacturing and mining, whereas take-offs refer to national output.” Gerschenkron, “The Approach to European Industrialization: A Postscript,” in Economic Backwardness, in. 1, pp. 353–54.

⁹ Gerschenkron's second criterion, that a country maintain its rapid rate of manufacturing during a period of international recession, was not difficult to satisfy. Information presented by Mitchell indicates brief international recessions occurred every few years. Mitchell, Wesley C., Business Cycles, The Problem and Its Setting, NBER (New York: J. J. Little and Company, 1927), Table 32, pp. 425–37.Google Scholar

¹⁰ Youngson, A. J., Possibilities of Economic Progress (Cambridge, ]Eng.[: The University Press, 1959), pp. 201–2.Google Scholar Unfortunately, manufacturing statistics for Denmark are not available for periods prior to 1870, so Youngson's statement cannot be verified. All that can be confirmed is that the growth of manufacturing output in Denmark following 1870 was more rapid than that which occurred during the spurts of France and Germany (see Table 3, p. 9).

¹¹ “Reflections on the ‘Prerequisites’ of Modem Industrialization,” in Economic Backwardness, p. 44. Reprinted from L'industria (no. 2, 1957).

¹² Unreliable data and the ambiguity of index numbers are two of the greatest problems. By the use of only one source for per capita income statistics, however, some of the problems are reduced. Colin Clark has converted all national currencies into what he calls International Units and estimated nonmarketed farm income, adding it to his estimates of national income in his computations of per capita income. The Conditions of Economic Progress (3d ed.; London: Macmillan, 1957).Google Scholar For a detailed description of Clark's methods, see ibid., pp. 18–87. It might be argued that more recent sources for national income data than Clark should be used. The difficulty in using a variety of sources is in converting data from the various countries into a common denominator.

¹³ Even Gerschenkron expresses his doubts as to the usefulness of per capita income as a measure of backwardness. “Reflections,” pp. 42–43.

¹⁴ See, for example, ibid., p. 44, and Gerschenkron, “Notes,” p, 361.

¹⁵ Simon Kuznets has referred to it as “intriguing” in his “Notes on the Take-off,” in Rostow, W. W., ed., The Economics of Take-off into Sustained Growth (New York: St. Martin's Press, 1963), p. 38.Google Scholar

¹⁶ This article is concerned primarily with the macroeconomic characteristics of development which result from countries beginning industrialization with different amounts and kinds of technology available for use in that development. While the measures and the methods chosen for assigning the rankings of relative backwardness may have value for this article, it does not follow that they will be adequate for some other purpose. They may not be suitable for examining, for example, the changing social institutions during development.

¹⁷ Use of a single country is a second choice, but one made necessary by the lack of criteria for using a combination of countries.

¹⁸ The use of England as the standard is not without pitfalls. One possible criticism might be that England is an island country without much room for agricultural expansion, but with rich mineral resources. This combination of factor endowments could make the characteristics of the English economy significantly different from those of the continental European countries. This may be true. When the rankings of relative backwardness using England as the standard are compared to those resulting from the use of lateness, however (a measure not requiring this kind of comparison), they are nearly identical (see Table 2).

¹⁹ The computations are carried out as follows:

²⁰ Gerschenkron, “Notes,” and “Patterns and Problems.”

²¹ It is not clear that the specific length of time selected for the tests is a critical choice. The one- and two-decade tests normally give very similar results (in terms of the rankings of the countries), implying that the characteristics of the spurt in each country were fairly consistent over at least this period of time. It should be emphasized that I am choosing the lengths of time for the tests of the hypothesis, not of the spurts. There is no implication that a spurt ends after twenty years.

²² Spearman's coefficient, r_s, is used.

²³ Gerschenkron, “Economic Backwardness,” p. 6. Gerschenkron refers to the situation in which there is a difference between the actual and potential conditions of a country as one of “tension.” The greater the1 unused opportunities, the greater the “tension.”

²⁴ Gerschenkron attributes this to “… complementarity and indivisibilities in economic process.” Ibid., pp. 8, 9.

²⁵ Ibid., p. 7.

²⁶ Gerschenkron, “Early Phases,” p. 157.

²⁷ Gerschenkron, “Economic Backwardness,” p. 7.

²⁸ Gerschenkron, “Patterns and Problems,” p. 122.

²⁹ The rB coefficient between lateness and the initial share of producers’ goods is 0.94, with a level of significance of 0.01.

³⁰ This suggestion is supported by Fogel, who states that if the rate of structural change during a country's ‘take-off” is not unique, neither is “… the quality or degree of technological innovation.” Fogel, Robert William, Railroads and American Economic Growth: Essays in Econometric History (Baltimore: The Johns Hopkins Press, 1964), p. 128.Google Scholar Fogel is concerned with the take-off, but because of the similarities between Gerschenkron's “great spurt” and Rostow's “take-off,” the conclusion Fogel reaches should not be any different were he discussing the “great spurt.”

³¹ The rs between lateness and per capita manufacturing output at the beginning of the spurt is —0.89 with a significance level of 0.02.

³² There was also some indication that per capita output of producers’ goods was higher in the later countries than the earlier ones. This possible relationship, while not statistically significant, merits further investigation.

³³ In Rostow's terminology the higher producers’ goods share of manufacturing might be referred to as a higher degree of satisfaction of one of the prerequisites for growth (the existence of capital equipment).

³⁴ This is similar to the Rodenstein-Rodan argument in which social overhead capital creates further opportunities for investment and must precede, to a certain extent, the development of a country. See Rodenstein-Rodan, P. N., “Problems of Industrialization of Eastern and Southeastern Europe,” Economic Journal, LIII (June-Sept. 1943).Google Scholar