¹ Solow, Robert, “Technical Change and the Aggregate Production Function,” Review of Economics and Statistics, XXXIX, No. 3 (Aug. 1957), 312–20CrossRefGoogle Scholar; Abramovitz, Moses, “Resource and Output Trends in the U.S. Since 1870,” American Economic Review Papers and Proceedings, XL, No. 2 (May 1956), 1–23Google Scholar; Massell, Benton, “Capital Formation and Technological Change in U. S. Manufacturing,” Review of Economics and Statistics, XLII, No. 2 (May 1960), 182–88.CrossRefGoogle Scholar

² The two most important works which should be consulted by anyone interested in this area of research are Habakkuk, H. J., American and British Technology in the 19th Century (Cambridge [England]: The University Press, 1962)Google Scholar, and Strassmann, W. Paul, Risk and Technological Innovation: American Manufacturing Methods During the Nineteenth Century (Ithaca: Cornell University Press, 1959).Google Scholar

³ Rostow, W. W., “The Take-Off into Self-Sustained Growth,” Economic Journal, LXVI, No. 261 (Mar. 1956), 25–48CrossRefGoogle Scholar, and Lewis, W. A., “Economic Development with Unlimited Supplies of Labour,” The Manchester School, XXII, No. 2 (May 1954), 139–91.CrossRefGoogle Scholar

⁴ Kuznets, Simon, “Quantitative Aspects of the Economic Growth of Nations: VI. Long-Term Trends in Capital Formation Proportions,” Economic Development and Cultural Change, Vol. IX, No. 4, Part 2 (July 1961).Google Scholar

⁵ Gallman, Robert, “Commodity Output, 1839–1899,” Trends in the American Economy in the Nineteenth Century (Studies in Income and Wealth, 23 [Princeton, N. J.: Princeton University Press for NBER, 1960]) 13–67Google Scholar; Kuznets, Simon, Capital in the American Economy (Princeton for NBER, 1961), ch. iv.Google Scholar

⁶ “There were mules and steam-engines before there were any labourers whose exclusive occupation it was to make mules and steam-engines; just as men wore clothes before there were such people as tailors.” Marx, Karl, Capital, I (Modern Library Edition [New York: Random House, 1936]), 417.Google Scholar

⁷ The experiences of Eli Whitney with his government musket contract are by now legendary. See Mirsky, Jeanette and Nevins, Allan, The World of Eli Whitney (New York: Macmillan Company, 1952).Google ScholarHabakkuk, Compare H. J., “The Historical Experience on the Basic Conditions of Economic Progress,” International Social Science Bulletin, VI, No. 2 (1954), 189–96.Google Scholar

A skeptical note on Whitney's role in developing the “American system” has recently been struck in Woodbury, Robert S., “The Legend of Eli Whitney and Interchangeable Parts,” Technology and Culture, II, No. 2 (Summer 1960), 235–53.CrossRefGoogle Scholar

⁸ It is not an historical coincidence that specialized machinery producers began to emerge on the national scene during precisely the same period as the development of our national railway network. Until roughly 1840, machinery production was not only relatively unspecialized—each producer typically undertaking a wide range of output—but it was also, because of the high cost of transporting machinery, a highly localized operation—each producer typically producing for a very limited geographic radius. The growing specialization in machine production after 1840, characterized by the emergence of large numbers of producers each of whom typically concentrated on a very narrow range of machines, was closely linked with the transportation improvements and consequent reduction in freight costs during the period.

⁹ A Chronicle of Textile Machinery 1824–1924 (Boston: Saco-Lowell Shops 1924) pp. 16–20.Google Scholar

The number of locomotives built by the Locks and Canals Company and by the Lowell Machine Shop between 1835 and 1861 is given, together with the purchasing railroad companies, in Gibb, George S., The Saco-Lowell Shops (Cambridge: Harvard University Press, 1950), appendix 6, p. 641.CrossRefGoogle Scholar

¹⁰ Hayes, John L., American Textile Machinery (Cambridge, 1879), pp. 57–58Google Scholar; Lincoln, Jonathan T., “Machine Tool Beginnings,” American Machinist, LXXVI (Aug. 3, 1932), 902–3.Google Scholar Although Mattias Baldwin in 1852 described himself as a “Manufacturer of Locomotive, Marine, and Stationary STEAM ENGINES,” his advertisements also added: “All kinds of Machinery furnished to order.” Journal of the Franklin Institute, L (June 1852)Google Scholar, opposite p. 444. A report of British observers for the same period states, “The practice which prevails of combining various branches of manufacture in the same establishment, would … render separate descriptions of each somewhat complicated. In some cases the manufacture of locomotives is combined with that of mill-gearing, engine-tools, spinning, and other machinery. In others, marine engines, hydraulic presses, forge-hammers, and large cannon are all made in the same establishments. The policy of thus mixing together the various branches arises, in addition to other causes, from the fact that the demand is not always sufficient to occupy large works in a single manufacture.” The Industry of the United States in Machinery, Manufactures and Useful and Ornamental Arts, (compiled from the official reports of Messrs. Whitworth, Joseph and Wallis, George [London, 1854]), p. 3.Google Scholar

¹¹ Roe, Joseph W., “Machine Tools in America,” journal of the Franklin Institute (May 1938), pp. 499–511.CrossRefGoogle Scholar

¹² As early as 1855, a team of British engineers which had traveled to the United States for the purpose of inspecting American methods of arms manufacture felt compelled to make the following observations on the extent to which specialized machinery had been developed in American industry: “As regards the class of machinery usually employed by engineers and machine makers, they are upon the whole behind those of England, but in the adaptation of special apparatus to a single operation in almost all branches of industry, the Americans display an amount of ingenuity, combined with undaunted energy, which as a nation we would do well to imitate, if we mean to hold our present position in the great market of the world.” Report of the Committee on Machinery of U. S. (H. C., 1855), p. 32.Google Scholar Similar observations and admonitions appear through a succession of reports by British observers to international exhibitions through the subsequent decades of the nineteenth century. See, for instance, the report on machine tools in Reports on the Paris Universal Exhibition, 1867, Presented to both Houses of Parliament (1868), IV, especially 370–73Google Scholar, and “Machines and Tools for working Metals, Wood and Stone,” in Reports on the Philadelphia International Exhibition of 1876, Presented to both Houses of Parliament (1877), I, especially 228–35. In the latter report, Mr. John Anderson, the author, states: “To realize the nature of the competition that awaits us, their [American] factories and workshops have to be inspected, in order to see the variety of special tools that are being introduced, both to insure precision and to economize labour; this system of special tools is extending into almost every branch of industry where articles have to be repeated. This applies to furniture, hardware, clocks, watches, small arms ammunition, and to an endless variety of other things. The articles so made are not only good in quality, but the cost of production is extremely low, notwithstanding that those employed earn high pay.” (p. 235). Mr. Anderson closes with a rhapsody on the tool-using and tool-making abilities of the Americans, and on the urgency of Britain's girding her loins for the coming industrial and engineering competition.Google Scholar

¹³ Twelfth Census of the United States (1900), X, Part 4, “Manufactures,” 385. This report also refers to the existence of a total of ninety metal-working machinery firms in the five leading centers of Cincinnati, Philadelphia, Providence, Hartford, and Worcester. For a breakdown by value of output of major categories of metal-working machinery in 1900 and 1905, see Special Reports of the Census Office (1905), “Metal-working Machinery,” p. 227.

¹⁴ Census of Manufactures (1914), II, “Reports for Selected Industries,” 269. For the same year there were 277 metal-working machinery plants, other than those producing machine tools, with an output valued at $17,419,526.

¹⁵ American Machinist, XL (Jan. 29, 1914), 210.

¹⁶ We suggest, only in passing, that such a focus may also provide a more fruitful approach to a theory of the multiproduct firm.

¹⁷ “…industrial differentiation … has been and remains the type of change characteristically associated with the growth of production. Notable as has been the increase in the complexity of the apparatus of living, as shown by the increase in the variety of goods offered in consumers’ markets, the increase in the diversification of intermediate products and of industries manufacturing special products or groups of products has gone even further.” Young, Allyn, “Increasing Returns and Economic Progress,” Economic Journal XXXVIII, No. 152 (Dec. 1928), 537.Google Scholar

¹⁸ Usher, A. P., History of Mechanical Inventions (Cambridge: Harvard University Press, 1954), especially chs. xiii-xv.Google Scholar

¹⁹ On this point it is difficult to avoid the conclusion that we are still suffering, in our understanding of technological innovation, from a Schumpeterian blight. For all his profound understanding of the capitalist process, Schumpeter never quite overcame his preoccupation with the charismatic aspects of leadership and its role in instituting changes in the operation of the economic system. As a result, his own towering intellectual leadership in this area has led to an excessive concern with the more dramatic and discontinuous aspects of innovation, with the circumstances surrounding the initial “breakthrough,” and to a neglect of the less spectacular dimensions of innovation. We refer, in particular, to two other aspects: (1) the cumulative impact of relatively small innovations (which were of great importance in the design, development, and adaptation of machines), and (2) the determinants of the rate and the area over which an innovation, once made, is eventually diffused. These points will be treated subsequently in this paper.

²⁰ Stigler, George, “The Division of Labor is Limited by the Extent of the Market,” The Journal of Political Economy, LIX, No. 3 (June 1951), 190.Google Scholar

²¹ Woodbury, “Legend of Eli Whitney.” (See n. 7.)

²² Fitch, Charles, “Report on the Manufactures of Interchangeable Mechanism,” Tenth Census of the United States (1880), II, 13–19. (Hereafter cited as Fitch, Report.)Google Scholar

²³ The application to the last item was apparently too quick, for Blanchard was upheld, as late as 1849, in an infringement suit involving the application of his lathe to turning shoemaker's lasts. See Journal of the Franklin Institute XLVII (1849), 259–62.Google Scholar

The British commission on arms manufacture which visited the United States in 1853 was particularly impressed with the gunstock lathe, at the time apparently still unknown in England. “It is most remarkable that this valuable labour-saving machine should have been so much neglected in England, seeing that it is capable of being applied to so many branches of manufacture, its introduction into the armory will prove a national benefit.” Report of the Committee on Machinery of U.S. (H. C., 1855), p. 39. The British Government subsequently purchased these machines from the Ames Manufacturing Company of Chicopee, Massachusetts.

²⁴ Fitch reports that die-forging machines were employed at Harper's Ferry as early as 1827. (Fitch, Report, p. 20.) The most significant subsequent improvements were achieved at the Colt armory under the guidance of its ingenious superintendent, Elisha K. Root. Although the sewing-machine industry relied more heavily upon casting than on forging, drop forging with dies was introduced into sewing-machine manufacture as early as 1856. (Fitch, Report, p. 37.) Compare also Durfee, W. F., “The History and Modern Development of the Art of Interchangeable Construction in Mechanism,” Transactions, American Society of Mechanical Engineering, XIV (1893), especially 1,250–51.Google Scholar

²⁵ Fitch, Report, pp. 22–26; Roe, Joseph W., “History of the First Muling Machine,” American Machinist, XXXVI (June 27, 1912), 1,037–38Google Scholar; Woodbury, Robert S., History of the Milling Machine (Cambridge: M. I. T. Press, 1960), chs. i and ii. Woodbury's book, which is part of a series devoted to the history of machine tools, is an invaluable guide to the detailed technical development of the milling machine.Google Scholar

²⁶ Fitch, Report, pp. 3 and 25; Deyrup, Felicia J., Arms Makers of the Connecticut Valley (Northampton: Smith College Press, 1948), pp. 153–54.Google Scholar

²⁷ Fitch, Report, p. 26. Fitch cites examples for 1880 of armsmaking plants where milling machines constituted between 25 and 30 per cent of the total number of machines in use (p. 22).

²⁸ ibid., p. 28.

²⁹ Parkhurst, E. G., “Origin of the Turret, or Revolving Head,” American Machinist, XXIII (May 24, 1900), 489–91.Google Scholar Compare also Hubbard, Guy, “Development of Machine Tools in New England,” American Machinist, LX (Feb. 21, 1924), 272–74.Google Scholar

³⁰ Fitch, Report, pp. 27–28; Roe, Joseph W., English and American Tool Builders, (New Haven: Yale University Press, 1916), p. 143.Google Scholar The successors of the Robbins and Lawrence Company, the Jones and Lamson Machine Company, remained one of the leaders in turret lathe production for many years. See Hartness, James, Machine Building for Profit (Springfield, Vt.: Jones & Lamson Machinery Co., 1909). Hartness introduced the flat-turret lathe and was a pioneer in the application of hydraulic feeds to machine tools.Google Scholar

³¹ Hubbard, Guy, “Development of Machine Tools in New England,” American Machinist, LXI (Aug. 21, 1924), 314; Roe, Tool Builders, p. 176.Google Scholar

³² Manufactures of the United States in 1860 (compiled from the original returns of the Eighth Census), p. clxxxix. The Twelfth Census conveniently collates the basic earlier census data on the sewing machines; Twelfth Census (1900), X, 404. An illustration and description of the Singer Sewing Machine, as it appeared when patented in 1851, appears on the front page of Scientific American, Vol. VII, Number 7 (November 1, 1851).

³³ “In the Exhibition at London in 1851, only two very imperfect sewing machines were exhibited. In 1855, at Paris, there were 14 varieties of sewing machines, some of which were so perfect that little or no material advance has since been made; and in 1862, in the London Exhibition, about 50 different arrangements of machines were shown.… In the ‘Exposition Universelle,’ now open in Paris, there are no less than 87 exhibitors of sewing machines. Their manufacture is now very general in European countries. France has 27 exhibitors, America 21, and England 12. Even so small a principality as Hesse has two exhibitors of sewing machines; the colony of Canada, five.’ Hichens, Captain on “Apparatus for Sewing and Making up Clothing,” in Reports on the Paris Universal Exhibition, 1867. Presented to both Houses of Parliament (1868)Google Scholar; V, 131–32.

³⁴ See the summary of census statistics for 1880 and 1890 for industries in which the sewing machine was employed extensively, in Depew, Chauncey M. (ed.), One Hundred Years of American Commerce (New York, 1895), II, 538.Google Scholar See also the description of sewing machines for specialized purposes in Paget, Frederick A., “Report on the Machines and Apparatus used in Sewing and Clothing,” Reports on the Philadelphia International Exhibition of 1876. Presented to both Houses of Parliament (1877)Google Scholar; I, 242–43.

³⁵ The sewing-machine industry relied much more heavily than did the firearms industry upon cast iron and was instrumental in bringing about important improvements in foundry operations. The molding press, for example, was introduced by Albert Eames, who was at the time (around 1873), foreman of the foundry at the Wheeler and Wilson Sewing Machine Company and who was earlier employed in firearms manufacture. The molding press played an important role in the production of sewing-machine parts, in hardware generally, and in other industries dependent on casting. Fitch, Report, pp. 36–37; also Fitch, , “Report on the Manufacture of Hardware, Cutlery, and Edge-tools,” Tenth Census of the United States (1880); II, 10.Google Scholar

³⁶ Roe, Tool Builders, pp. 202–6; Woodbury, Robert S., History of the Gear-Cutting Machine (Cambridge: M. I. T. Press, 1958), pp. 80–81. Brown and Sharpe undertook the production of automatic gear cutters in 1877.Google Scholar

³⁷ Woodbury, , History of the Grinding Machine (Cambridge: M. I. T. Press, 1959), pp. 60–61.Google Scholar

³⁸ ibid. Compare Burlingame, Luther D., “The Universal Milling Machine,” American Machinist, XXXIV (Jan. 5, 1911), 9.Google Scholar

³⁹ I am grateful to Professor Duncan McDougall for kindly placing at my disposal his data on the sale of machinery output of the Brown and Sharpe Company.

⁴⁰ Woodbury, Grinding Machine, pp. 60–61.

⁴¹ ibid., pp. 61–62; Hubbard, Guy, “100 Years of Progress in American Metalworking Equipment,” Automotive Industries, CXIII, No. 5 (Sept. 1, 1955), p. 315.Google Scholar

⁴² Woodbury, Grinding Machine, pp. 64–71. Brown and Sharpe applied the hydraulic feed technique to the grinding machine in 1902, a technique which makes it easier to operate any machine tool automatically. ibid., p. 139.

⁴³ Burlingame, “Milling Machine” (cited in n.38).

⁴⁴ An admirable descriptive survey, profusely illustrated, of the “state of the arts” in machine tools in 1880 may be found in Tenth Census (1880), Vol. XXII, “Report on Power and Machinery Employed in Manufactures.”

⁴⁵ It is of more than passing interest as evidence of technological convergence between sewing machines and bicycles to note that, in England as well as in the United States, the earliest bicycles were produced in sewing-machine plants. Colonel Pope in 1878 produced his first “Columbia” in a corner of the Weed Sewing Machine Company plant at Hartford. In England in the late 1860's, the Coventry Sewing Machine Company played a similar role.

See Twelfth Census (1900), X, 331; Albert A. Pope, “The Bicycle Industry,” in Depew, One Hundred Years, II, 550; Allen, G. C., The Industrial Development of Birmingham and the Black Country, 1860–1927 (London: G. Allen, 1929), p. 243Google Scholar; Clapham, J., An Economic History of Great Britain (Cambridge: The University Press, 1952), II, 96–97.Google Scholar

⁴⁶ Special Reports of the Census Office (1905), Part 4, “Selected Industries,” p. 289. Compare Thirteenth Census of the United States (1910), X, “Manufactures,” 825–28.

⁴⁷ Some of the earliest bicycles are reported to have weighed over one hundred pounds.

⁴⁸ Twelfth Census (1900), “Manufactures,” pp. 385–88. See also Colvin, Fred, Sixty Years with Men and Machines (New York: Whittlesey House, 1947), 88–89.Google Scholar

⁴⁹ “The successful ball bearing depends upon having the balls themselves perfectly spherical and all of identical diameter. The ball must run in races perfectly circular, perfectly concentric, and of exact dimensions. Not only must the balls and their races be machined to a fine surface finish, but all these dimensions must be held to close tolerances, and all these parts must be hardened. Only grinding could deal with this problem.” Woodbury, Grinding Machine, p. 110.

⁵⁰ ibid., p. 111.

⁵¹ A former Brown and Sharpe employee, who had been initiated earlier into the problems of grinding at the Seth Thomas Clock Company.

⁵² Woodbury, Gear-Cutting Machine, pp. 78–126.

⁵³ “The bicycle impinged upon the locomotive through the medium of machine tools when the new turret lathes, which had been developed to form wheel hubs, were applied to the manufacture of locomotive crankpins …” Colvin, Sixty Years, p. 89.

⁵⁴ In 1894, Pope's bicycle plant in Hartford built “… what was then the most up-to-date mill in the country for making cold drawn steel tubing. There was also a research department for testing metals and improving bicycle design.” Rae, John B., American Automobile Manufacturers (Philadelphia: Temple Press, 1959), p. 9.Google Scholar

⁵⁵ Special Reports of the Census Office (1905), Part 4, “Manufactures, p. 269. In the 1900 Census Reports, the brief treatment of motor vehicles appears in the chapter on locomotives. Most of the 4,192 motor vehicles mentioned in that report were in fact powered by steam or electricity. Twelfth Census (1900), “Manufactures,” pp. 255–59.

⁵⁶ Census of Manufactures (1914), II, “Reports for Selected Industries,” 731–32.

⁵⁷ Woodbury states that after 1900 the automobile industry became “…the largest single customer of the machine tool industry, taking 25 to 30 per cent of the output.….” Woodbury, Grinding Machine, p. 120.

⁵⁸ ibid., p. 99.

⁵⁹ “When the demand for bicycles decreased some manufacturers turned to the automobile, and many establishments that made only bicycles in 1900 are now [1905] devoted primarily to the manufacture of automobiles, while others make them to a greater or less degree in connection with the manufacture of bicycles.…” Special Reports of the Census Office (1905), Part 4, “Manufactures,” p. 289. See also Rae, American Automobile Manufacturers, pp. 8–10. The emergence of early automobile firms out of bicycle firms is a sequence which was reproduced in Great Britain. See Saul, S. B., “The Motor Industry in Britain to 1914,” Business History V, No. 1 (Dec. 1962), 22–44.CrossRefGoogle Scholar

⁶⁰ Hinde, H. J., “Relation of Power Presses and Dies to the Automobile Industry,” Mechanical Engineering, XLIII (Aug. 1921), 531.Google Scholar

⁶¹ Just as, at an earlier date, the working action of the Colt revolver was incorporated into machines used for producing the revolver. “…the same arrangement of parts characteristic of the Colt revolver seems to have been carried through the principal machines for its manufacture, the horizontal chucking lathes, cone-seating and screw machines, barrel-boring, profiling, and mortising machines, and even the compound crank-drops, exhibiting the same general arrangement of working parts about a center.” Fitch, Report, pp. 27–28.

⁶² Einstein, S., “Machine-Tool Milestones, Past and Future,” Mechanical Engineering, LII (Nov. 1930), 961Google Scholar; Hendrickson, F. K., “The Influence of the Automobile on the Machine-Tool Industry in General,” Mechanical Engineering, XLIII (Aug. 1921), 530.Google Scholar In discussing the relationship between the milling machine and the automobile, Einstein concluded: “The automotive industry asked of the machinetool designer automatic machines, either through modification of standard designs or machines of entirely special design—quick-acting fixtures, either hand-operated or automatically operated—more powerful and stronger machines occupying a minimum amount of floor space. On the other hand, the automotive industry supplied the machine-tool designer with a vast variety of highly successful machanisms and constructive details, from which he could draw freely such elements and such ideas as could be adapted to the design of machine tools.” “Discussion at the Machine Shop Practice Session,” Mechanical Engineering, XLIII (Aug. 1921), 534.Google Scholar

⁶³ For example, the redesigning of milling cutters at the Cincinnati Milling Machine Company, when it was discovered that “…the cutters of the time were not as strong as the machines that were driving them and therefore gave out long before the maximum power of the machine was reached.” Woodbury, Milling Machine, p. 80.

⁶⁴ Hubbard, Guy, “Metal-Working Plants,” Mechanical Engineering, LII (Apr. 1930), 411.Google Scholar See also Frederick Taylor, W., The Art of Cutting Metals (New York: American Association of Mechanical Engineers, 1907)Google Scholar; Special Reports of the Census Office (1905), Part 4, “Metal-working Machinery,” pp. 232–33; Flanders, Ralph, “The Influence of the Automobile on Lathe Practice,” Mechanical Engineering, XLIII (Aug. 1921), 532Google Scholar; Oxford, Carl J., “One Hundred Years of Metal Cutting Tools,” Centennial of Engineering 1852–1952 (Chicago: American Association of Engineers, 1953), pp. 346–50Google Scholar; Einstein, S., “Machine-Tool Milestones, Past and Future,” Mechanical Engineering, LII (Nov. 1930), pp. 959–62.Google Scholar

⁶⁵ Just as it simplified, for example, the development of the typewriter, whose problems remained unsolved until it was placed in the hands of the skilled machinists and technical experts of E. Remington and Sons, gun manufacturers at Ilion, New York. Twelfth Census (1900), “Manufactures,” p. 442.

⁶⁶ Accuracy for Seventy Years, 1860–1930 (Hartford, Conn.: Pratt & Whitney Co., 1930).Google Scholar

⁶⁷ Hubbard, Guy, “Development of Machine Tools in New England,” American Machinist, LX (Jan. 31, 1924), 171–73.Google Scholar

⁶⁸ Owen, C. B., “Organization and Equipment of an Automobile Factory,” Machinery, XV (Mar. 1909), 493.Google Scholar

⁶⁹ The government's current role in subsidizing research and development activity has an earlier and interesting parallel in the innovations emerging out of firearms production—an industry where, as we have seen, the government played a role as a major producer as well as “consumer.”