¹ Clark, Gregory, “Why Isn't the Whole World Developed? Lessons from the Cotton Mills,” this Journal, 47 (03 1987), pp. 141–74.Google Scholar

² Hanson, John R. II, “Why Isn't the Whole World Developed? A Traditional View,” this Journal, 48 (09 1988), pp. 668–75.Google Scholar

³ Brassey, Thomas, Foreign Work and English Wages (London, 1879), pp. 157–96.Google ScholarJeans, J. S., “On the Comparative Efficiency and Earnings of Labour at Home and Abroad,” Journal of the Royal Statistical Society (Dec 1884), pp. 623–24. Schulze-Gaevernitz even expressed a view somewhat similar to the one I put forward to explain England's supremacy in cotton textiles in the late nineteenth century.Google Scholarvon Schulze-Gaevernitz, G., The Cotton Trade in England and on the Continent (London, 1895), see esp. pp. 85–130.Google Scholar

⁴ See, for example, British Parliamentary Papers, Indian Factory Labour Commission, Vol. 2: Evidence. Cd. 4519 (1909). Buchanan himself is unclear about how much of excess machine manning in India to attribute to lower labor efficiency and how much to capital-labor substitution in response to India's low wages.Google ScholarBuchanan, Daniel, The Development of Capitalist Enterprise in India (New York, 1934).Google Scholar

⁵ Pack, Howard, Productivity, Technology and Industrial Development: A Case Study in Textiles (New York, 1987).Google Scholar

⁶ Ibid., pp. 139–46. It should be noted that Pack interprets these differences in a completely different way than I do, putting much more emphasis on capital-labor substitution in determining manning levels.

⁷ Buchanan argues that once we correct for shift working the ratio would actually be only 1.33 to 1 in jute. Buchanan, Capitalist Enterprise in India, p. 368.Google Scholar

⁸ Clark, W. A. Graham, Linen, Jute and Hemp Industries in the United Kingdom with Notes on the Growing and Manufacture of Jute in India. U.S. Department of Commerce, Bureau of Foreign and Domestic Affairs, Special Agents Series, No. 74 (Washington, D.C., 1913), pp. 101–3, 168–72.Google Scholar As late as 1925–1929 Britain was still manufacturing 186,000 tons of jute per year, compared to 217,000 tons a year in 1896–1900. On exports, see Great Britain, Board of Trade, Working Party Reports, Jute (London, 1948), pp. 116, 121.Google Scholar

⁹ Great Britain, Board of Trade, Working Party Reports, Jute, p. 13.Google Scholar

¹⁰ The profit rate calculated for India represents a lower bound, since if the mill operated there it might have been able to use more cheap labor to economize on more expensive factors such as machinery.Google Scholar

¹¹ Moms, Morris D., “The Growth of Large Scale Industry to 1947,” in Kumar, Dharma, ed., The Cambridge Economic History of India (Cambridge, 1983), p. 572. At the same time the net profits of Bombay textile mills averaged only 10.3 percent.Google Scholar

¹² Davis, Lance and Huttenback, Robert A., Mammon and the Pursuit of Empire: The Political Economy of British Imperialism, 1860–1912 (Cambridge, 1986), p. 107. If the profit rate in Indian mills was only 10 percent per year, then at the same price for outputs British profits on circulating capital would be −2 percent, even if the fixed capital in buildings and machinery was a free good for British manufacturers in 1912 because it had no value in other uses. Consequently all the British jute manufacturers should have liquidated their circulating capital and exited the business long before 1912.Google Scholar

¹³ Clark, “Why Isn't the Whole World Developed?” table 3.Google Scholar

¹⁴ Buchanan, Capitalist Enterprise in India, p. 368.Google Scholar

¹⁵ Another good example of such a task would be sweeping the floors, where if Indian workers were slower it would be easy to use more of them without this having any influence on machine productivity.Google Scholar

¹⁶ In Britain workers of different abilities could get very different outputs out of the same looms; for example, in one loom shed with 13 experienced weavers each operating 3 cotton looms, the average efficiency of weavers varied from 70.2 to 92.6 percent. Looms operated by the same weaver had very similar outputs, almost all the differences in loom efficiency stemming from differences in weavers. In silk weaving on one class of cloth the efficiency of weavers varied from 55.0 to 82.2 percent, and on another from 39.3 to 70.6 percent. S. Wyatt, Variations in Efficiency in Cotton Weaving, Great Britain, Medical Research Council, Industrial Fatigue Research Board, Report No. 23 (1923).Google Scholar

¹⁷ That is, no substitution of labor for capital is possible if wages are lower.Google Scholar

¹⁸ British Parliamentary Papers, Indian Factory Labor Commission, pp. 20, 23, 36, 65, 77, 79, 85, 90–92, 125, 139, 181, 315, 354, 377.Google Scholar

¹⁹ Clark, W. A. Graham, Cotton Goods in Japan. U.S. Department of Commerce, Bureau of Foreign and Domestic Affairs, Special Agents Series, No. 86 (Washington, D.C., 1914), p. 191.Google ScholarWinterbottom, James, Cotton Spinning Calculations and Yarn Costs (2nd ed.London, 1921), p. 213.Google Scholar

²⁰ British Parliamentary Papers, Indian Factory Labor Commission, pp. 91, 149.Google Scholar

²¹ Ibid., pp. 5 79,91, 94–95, 149–50, 204, 206, 214, 347.

²² I should note that Buchanan believes output per machine was close to that in Britain in India, but I think the evidence to the contrary is persuasive. See Buchanan, Capitalist Enterprise in India, pp. 369–71.Google Scholar

²³ British mills operated 55 hours per week in 1906 and 54 hours in 1912. Indian mills operated 90 hours per week in 1904 and 83.25 hours per week in 1912. The spindlage, looms, and jute consumption figures are from Clark, Linen, Jute and Hemp Industries, pp. 103, 148, 157.Google Scholar

²⁴ Not all the raw jute went to the looms since some was sold as yarn, but there is no reason to believe this proportion was any greater in India than in Britain, since the ratio of spindles to looms was the same.Google Scholar

²⁵ This is calculated using the fact that the maximum weight of output per loom-hour in pounds on different types of cloth is given by Z = (ppmlppi). width. weight. 1.67, wher ppm equals picks per minute, ppi equals picks per inch, width equals the width of the cloth in yards, and weight equals pounds per square yard of cloth. Sacking looms produced cloth weighing about twice as much per square yard. The cloth was narrower than hessian, but this was offst by higher ppm and lower ppi. See Clark, Linen, Jute and Hemp Industries, p. 165, Board of Trade, Jute, p. 16.Google Scholar

²⁶ Clark, Linen, Jute and Hemp Industries, pp. 96, 101–3, 137.Google Scholar

²⁷ Ibid., pp. 148, 168–72. Buchanan. Capitalist Enterprise in India, p. 245. We can calculate the value of total output in the Indian mills from the value of exports, since from 1911 to 1920 exports represented 80 percent of Indian mill output.

²⁸ Value added per loom in India was not lower because of lower prices for the same output in India than in Britain. As noted above, the Indians had access to markets where Britain sold, except that the effective price to British producers would often be lower because of higher transport costs.Google Scholar

²⁹ They could have paid higher wages per hour because longer hours economized greatly on capital.Google Scholar