¹ The trade statistics are tabulated in Allen, Robert C., “International Competition and the Growth of the British Iron and Steel Industry: 1830–1913” (Ph.D. diss., Harvard Univ., 1975), pp. 420–23Google Scholar. The British product classification was used for all countries.

² In this paper, iron and steel production for all countries in 1881 and iater years equals the total tonnage of final products produced by the iron and steel sector. Final products includes final rolling mill products, iron castings (or pig iron consumption by foundries), exports of pig iron, and exports of semifinished products. The underlying data were derived from the mineral and trade statistics of the countries concerned. Details of the sources and calculations are available in ibid., pp. 132–39. It should be noted that the divergence between final iron and steel production and pig iron production is not large. For 1880 and earlier years, pig iron production is used in this paper as a proxy for final iron and steel production. Pig iron production figures are tabulated in ibid., pp. 404–07.

³ See Table 8.

⁴ Payne, Peter L., “Iron and Steel Manufactures,” in Aldcroft, Derek H., ed., The Development of British Industry and Foreign Competition: 1875–1914 (Toronto, 1968), p. 85Google Scholar.

⁵ This phenomenon has been noted in other trades as well. See Aldcroft, Development, passim.

⁶ Mitchell, Brian R. and Deane, Phyllis, Abstract of British Historical Statistics (Cambridge, 1971), p. 148Google Scholar.

⁷ Carr, James C. and Taplin, Walter, History of the British Steel Industry (Cambridge, 1962), pp. 251–53Google Scholar.

⁸ Harley, C. Knick, “Skilled Labour and the Choice of Technique in Edwardian England,” Explorations in Economic History, 11 (Summer 1974), 391–414CrossRefGoogle Scholar.

⁹ Notable advocates include Landes, David S., The Unbound Prometheus (Cambridge, 1969)Google Scholar, and Aldcroft, Derek H., “The Entrepreneur and the British Economy, 1870–1914,” Economic History Review, 2nd ser., 17 (April 1964), 113–34CrossRefGoogle Scholar. McCloskey, Donald N. and Sandberg, Lars G., “From Damnation to Redemption: Judgments on the Late Victorian Entrepreneur,” Explorations in Economic History, 9 (Fall 1971), 89–108CrossRefGoogle Scholar, summarize the literature and vigorously defend the British entrepreneur. Burn, Duncan L., The Economic History of Steelmaking: 1867–1939 (Cambridge, 1961)Google Scholar, develops the productivity hypothesis in the context of iron and steel.

¹⁰ McCloskey, Donald N., Economic Maturity and Entrepreneurial Decline (Cambridge, 1974)Google Scholar.

¹¹ Other, miscellaneous inputs had an average share in costs of 13 Excluding these inputs from equation (1) is equivalent to assuming that their productivity was the same in the two situations compared.

¹² Same sources as Table 2.

¹³ According to the United Kingdom, Final Report on the First Census of Production of the United Kingdom (1907) (1912), p. 175, sales by the irohand steel industry were £105,322,000 and material costs were £75,274,000. Employment was 261,666. In conjunction with the annual earnings per worker estimate of £84.5 in McCloskey, Economic Maturity, p. 122, the implied cost shares are 8 percent for capital, 21 percent for labor, and 71 percent for other inputs. For German shares, see the iron and steel production statistics in Statistisches Jahrbuch für das Deutsche Reich, annually.

¹⁴ Landes, Unbound Prometheus, p. 255.

¹⁵ For German data, see the production statistics annually recorded in Statistisches Jahrbuch. For United States, British, and Continental data in 1888–89, see U.S. Commissioner of Labor, Sixth Annual Report. Cost of Production: Iron, Coal, Steel, Etc. (Washington,. 1891), pp. 152–54, 183–86Google Scholar. These sources are summarized in Allen, “International Competition,” pp. 260–64. For mid-nineteenth-century American and British figures, see Table 5 in this paper.

¹⁶ For Britain, production is taken to equal pig iron production as given by Mitchell and Deane, British Statistics, pp. 131–32. The number of workers' equals the number of people in England, Wales, and Scotland returned in the 1861 census with the occupations “iron manufacture (including moulders and founders),” “steel manufacturing workers,” and “tin plate workers.” In fact, the 1861 occupational statistics record the industry in which an individual was employed rather than his occupation per se. See Armstrong, W. Alan, “The Use of Information about Occupation,” in Wrigley, Edward A., ed., Nineteenth Century Society: Essays in the Use of Quantitative Methods for the Study of Social Data (Cambridge, 1972), pp. 191–310CrossRefGoogle Scholar. It should be noted that the iron and steel employment figures used. here are larger than the employment estimates for that industry made by Armstrong (p. 260). For the United States, production and employment were derived from the 1860 and 1870 censuses. Production equals the output of rolling mills (exclusive of nail plate), nail factories, steelworks, and foundries. Casting production was estimated to equal pig iron production less pig iron consumption by rolling mills and steelworks. Number of workers equals the number employed in these industries plus the number of blast furnace workers. Foundry employment was estimated by dividing casting production by output per worker in foundries in 1870. For Prussia, all data were derived from Prussia, Zeitschrift fur das Berg., Hütten-, und Salinenwesen (1861) pp. 28–32, table entitled “Produktion der Hiitten in dem Preussischen Staate M. J. 1860,” pt. II: Hütten, sect. 1: Eisen. Employment equals total employment in that category. Production equals total production in that category less pig iron production.

¹⁷ The British figures were computed from the regional production figures and input-output coefficients given in Louis E. Gruner and Charles Lan, Etat présent de la métallurgie dufer en Angleterre (Paris, 1862). The national fuel productivity figure is a weighted average of the average product of fuel for the following products, given the weights shown in parentheses: Staffordshire bars (252), South Wales bars (1125), Scottish bars (0855), South Wales rails (2739), Cleveland rails (0561), Staffordshire plates (2200). The energy content of coal was taken to be 14,135 BTU per pound. The regional product weights are from ibid., pp. 516, 551, 553, 617–18, 632, 661. Input-output coefficients for each product were derived from ibid., pp. 530, 548, 554, 554–55, 621–22, 628, 662.

¹⁸ Allen, Robert C., “The Peculiar Productivity History of American Blast Furnaces, 1840–1913,” this Journal, 37 (Sept. 1977), 605–33Google Scholar.

¹⁹ Great Britain, Final Report on the Third Census of Production of the United Kingdom (1924): The Iron and Steel Trades, the Engineering Trades, and the Non-Ferrous Metals Trades (1931), reports the returns for the aborted 1912 census of production. The returns for the iron and steel industry were unusually complete. Steelworks and rolling mill production equals tonnage reported on p. 36 for categories d through p, inflated in proportion to value to account for unenumerated production (category r). Employment equals the total on p. 54 less estimated blast furnace employment derived from Ministry of Labor, Gazette.

²⁰ Germany, Produktion der Kohten-, Eisen-, und Hüttenindustrie, im Jahre 1912, Vierteljahreshefte zur Statistiks des Deutschen Reichs (1914), Vol. I, pp. 360–65Google Scholar. Production is tonnage of finished rolling mill products plus exports of semi-rolled products (from Statistisches Jahrbuch für das Deutschen Reich [1913], p. 214). The number of workers equals employment in puddling works, steelworks, and rolling mills.

²¹ U.S. Bureau of the Census, Census of Manufactures, 1914 (1919), Vol. II, pp. 199–267Google Scholar. Production is the output of finished rolled products plus the net output of semi-products. The number of workers equals employment in steelworks and rolling mills. In 1914 American productivity was unusually low for cyclical reasons. As Table 2 indicates, output per worker had been 73 tons in 1909.

²² The productivity of fuel in British steelworks was computed by dividing the production of final products by energy consumption computed from Jones, J. H., “The Present Position of the British Coal Trade,” Journal of the Royal Statistical Society, 93 (1930), 33CrossRefGoogle Scholar. On the productivity of fuel in blast furnaces, see Allen, “Peculiar Productivity,” and Berck, Peter, “Hard Driving and Efficiency: Iron Production in 1890,” this Journal, 38 (Dec. 1978), 879–900Google Scholar.

²³ In particular, the price of iron ore for Westphalian furnaces near the ore is very similar to the value per ton of ore at the mine in the early 1860s given annually in Prussia, Zeitschrift.

²⁴ For the wage rates of puddlers and rollers in the three countries, see U.S. Commissioner of Labor, Fifteenth Annual Report: Wages in Commercial Countries, Vol. 2 (1900), pp. 1174, 1218Google Scholar.

²⁵ Temin, Peter, Iron and Steel in Nineteenth-Century America: An Economic Inquiry (Cambridge, 1964), pp. 266–67Google Scholar.

²⁶ Mitchell and Deane, British Statistics, p. 131.

²⁷ Robert C. Allen, “Entrepreneurship and Technical Progress in the Northeast Coast Pig Iron Industry, 1850–1913,” Research in Economic History, vol. 7, forthcoming.

²⁸ Gruner and Lan, Métallurpe du fer, pp. 593–94.

²⁹ Habakkuk, H. J., American and British Technology in the Nineteenth Century (Cambridge, 1962)Google Scholar, and the vast literature it spawned.

³⁰ Bell, I Lowthian, Principles of the Manufacture of Iron and Steel (London, 1884), p. 562Google Scholar. See also Temin, Iron and Steel, p. 116.

³¹ Ibid.

³² Gruner and Lan, Métallurgie du fer, pp. 84, 249, 287.

³³ Until just before World War I less than one fifth of the Cleveland ironstone mined each year was converted into steel. This result is obtained by dividing basic steel ingot production on the Northeast coast (as given in British Iron Trade Association, Annual Statistical Report) by one third (the ore yield) times the output of Cleveland ore (as given by Mitchell and Deane, British Statistics, pp. 129–30).

³⁴ Campbell, Harry H., The Manufacture and Properties of Iron and Steel (New York, 1903), p. 706Google Scholar.

³⁵ Ibid., p. 701.

³⁶ The cost data for 1888–89 reported in U.S. Commissioner of Labor, Sixth Annual Report, pp. 33–193, also confirm this contention. Average costs computed from these data are reported and discussed in Allen, “International Competition,” pp. 314–15.

³⁷ Holley, A. L., Holley's Reports to the Bessemer Steel Co., Limited (New York, 1881)Google Scholar, Ehrenberg, Hans, Die Eisenhuttentechnik und der Deutsche Hüttenarbeiter, Müchener Volkswirtschaftliche Studien, Vol. 80 (Stuttgart, 1906), pp. 177–86Google Scholar; and the prices averaged to form Table 1. Berck, “Hard Driving,” however, computed high ratios of price to average total cost for American blast furnaces in the mid-1880s.

³⁸ Prussia, Zeitschrift, 1882.

³⁹ Lambi, Ivo N., “The Protectionist Interests of the German Iron and Steel industry,” this Journal, 22 (March 1962), 59–70Google Scholar.

⁴⁰ Martin Fritz, “Järnmalmsproduktion och järnmalmsmarknad, 1883–1913,” Meddelanden från Ekonomisk-historiska institutionen vid Göteborgs Universitet, 11 (Göteborg, 1967), 40–41; and Martin Fritz, “Svensk järnmalmsexport, 1883–1913,” ibid., 12 (1967), 154–59.

⁴¹ Fritz, “Järnmalmsproduktion,” p. 25; Campbell, Iron and Steel, p. 745; Pounds, Norman J. G., The Ruhr (Bloomington, Ind., 1952), pp. 111–12Google Scholar; Germany, Produktion der Kohlen-, Eisen-, and Hüttenindustrie, 1913.

⁴² Mussey, Henry R., Combination in the Mining Industry (New York, 1905)Google Scholar; U.S. Bureau of Corporations, Report of the Commissioner of Corporations on the Steel Industry, Vol. I (1911), pp. 377–82Google Scholar, and Vol. III (1913), pp. 351–52, 377–90, 480–507.

⁴³ The accounting scheme used here is developed formally in Robert C. Allen, “Accounting for Price Changes,” Univ. of British Columbia, Dept. of Economics Discussion Paper, No. 78–14 (1978).

⁴⁴ Tables 7 and 8.

⁴⁵ Report of the Tariff Commission, Vol. I, The Iron and Steel Trades (London, 1904), paragraphs 992, 1015Google Scholar.

⁴⁶ U.S. Bureau of Corporations, Report, Vol. 3, pp. 86, 144, 211.

⁴⁷ Tariff Commission, paragraph 993.

⁴⁸ The ratio of average realized price to unit cost in 1902–06 was 1.16 for structural shapes, 1.11 for steel merchant bars, and 1.15 for plates. Cost in these calculations is book cost (which includes depreciation but not interest on investment) plus interest at 5% on industrial fixed capital and working capital. Price and book cost from U.S. Bureau of Corporations, Report, Vol. 3, pp. 223, 230, and 236; investment costs from ibid., pp. 529–33. In the absence of investment costs for bars, the costs for Bessemer steel rails were used. It should be noted that the markups for structural steel and plates reported in this note are similar to those in Table 8. The markup for bars in Table 8 is less than that reported here. That anomaly is a consequence of the abnormally high relative price of British steel bars (see Table 1). Burn, Steelmaking, p. 110, n. 3, noted the same phenomenon in a different context, but could not account for it.

⁴⁹ McCloskey, Economic Maturity, pp. 120–24.

⁵⁰ This reallocation of exports presumes the price elasticity of demand for iron and steel is zero.

⁵¹ Temin, Peter, “The Relative Decline of the British Steel Industry, 1880–1913,” in Rosovsky, Henry, ed., Industrialization in Two Systems (New York, 1966), pp. 140–55Google Scholar, explored this question. Berck, “Hard Driving,” also considered it, but discussed only the consequences of Britain's using hard driving extensively.

⁵² Burn, Steelmaking, pp. 167–82.

⁵³ McCloskey, Economic Maturity, pp. 56–72.

⁵⁴ On the technology of the basic open hearth process see ibid., pp. 68–72; Pratt, Arthur E., “The Future Development of the Metal-Mixer and the Open-Hearth Process,” Journal of the Iron and Steel Institute, 3 (1908), 156–92Google Scholar, and “Correspondence,” 193–205; Arthur W. Richards, “Manufacture of Steel from High-Silicon Phosphoric Pig Iron by the Basic Bessemer Process,” ibid., 1 (1907), 104–08, and “Discussion,” 109–13; Ward, Robert G., An Introduction to the Physical Chemistry of Iron and Steelmaking (London, 1962)Google Scholar; Committee on Physical Chemistry of Steelmaking, Basic Open Hearth Steelmaking (New York, 1944), pp. 504–49Google Scholar.

⁵⁵ Reserve estimates from Max Roesler, The Iron-Ore Resources of Europe, U.S. Geological Survey, Bulletin 706 (1921), pp. 12, 45–46. Less conservative estimates are found in Henry Louis, “The Iron Ore Resources of the United Kingdom of Great Britain and Ireland,” in The Iron Ore Resources of the World, XI International Geological Congress (Stockholm, 1910), Vol. II, pp. 623–41. Production of ore from Mitchell and Deane, British Statistics, p. 130.

⁵⁶ Temin, “Relative Decline,” p. 151, produced some calculations in support of an argument of this kind. McCloskey, Economic Maturity, pp. 105–13, has recently remedied several difficulties with Temin's calculations and, on the basis of the revised calculations, rejected the possibility that differences in the average age of capital in Britain, America, and Germany could have been important. It is important to realize that neither Temin nor McCloskey ever measured the average age of capital in these countries. They only computed estimates on the basis of certain equilibrium assumptions. Nelson, Richard R., “Aggregate Production Functions and Medium-Range Growth Projections,” American Economic Review, 54 (Sept. 1964), 575–605Google Scholar, derives the requisite formulae and clarifies the relevant economic assumptions. The assumption that the rate of fall of unit costs is independent of the scale of new investment is questionable for the steel industry. Moreover, in their calculations, both Temin and McCloskey assume the depreciation rate exclusive of obsolescence to be 5% per year. No justification for this value is offered. Since other plausible values lead to widely different average ages, estimates of the equilibrium ages of capital must be regarded as inconclusive evaluations of the embodiment hypothesis.

⁵⁷ Allen, “Entrepreneurship and Technical Progress.”