¹ Rosenberg, N., “The Direction of Technological Change: Inducement of Mechanisms and Focusing Devices,” in Perspectives on Technology (New York, 1976), pp. 110–11.CrossRefGoogle Scholar

² Schmookler, J., Invention and Economic Growth (Cambridge, Massachusetts, 1966);CrossRefGoogle ScholarJewkes, J., Sawers, D., and Stillerman, R., The Sources of Invention (New York, 1958).Google Scholar For surveys of the induced innovation issues, see Kennedy, C. and Thirlwall, A. P., “Surveys in Applied Economics: Technical Progress,” Economic Journal, 82 (March 1972);CrossRefGoogle ScholarDavid, P., Technical Choice, Innovation and Economic Growth (Cambridge, 1975), ch. 1;Google ScholarBinswanger, H. P., Ruttan, V. W. et al. , Induced Innovation (Baltimore, 1978);Google ScholarRosenberg, N., “The Historiography of Technical Progress,” in Inside the Black Box: Technology and Economics (New York, 1982), pp. 3–33.Google Scholar

³ Binswanger and Ruttan, Induced Innovation, p. 159.Google Scholar

⁴ Schumpeter, J., “The Creative Response in Economic History,” this Journal, 7 (11 1947).Google Scholar For a recent application of the managerial/entrepreneurial distinction see Lazonick, W., “Competition, Specialization, and Industrial Decline,” this Journal, 41 (03. 1981), 31–8.Google Scholar

⁵ For similar views see Salter, W. E. G., “Productivity, Growth and Accumulation as Historical Processes,” in Problems in Economic Development, ed. Robinson, E. A. G. (London, 1965), especially pp. 266–68CrossRefGoogle Scholarand David, Technical Choice, p. 16.Google Scholar

⁶ Penrose, E., The Theory of the Growth of the Firm (Oxford, 1959), p. 29.Google Scholar

⁷ Witham, G. S., Modern Pulp and Paper Making (New York, 1920, 2nd ed., New York, 1942).Google Scholar

⁸ Three other aspects of the pattern of technological change in the industry—non-fourdrinier process innovations, the adaptation of the kraft pulping process for the utilization of southern U.S. pine, and product innovations—are not included here but are discussed in Cohen, A., “The Economic Determination of Technological Change,” chs. 4, 5. Other notable exclusions here are the Canadian pulp and paper industry and a discussion of the role of capital goods suppliers. By the twentieth century, pulp and paper was truly a North American industry and should ideally be discussed in that context. The chronicle of technological change covers all North American innovations. Most information on innovation originated from the Technical Association of the Pulp and Paper Industry (TAPPI). which did not distinguish between U.S. and Canadian advances. The TAPPI reports that appeared in the Paper Trade Journal usually were reproduced in the major Canadian trade journal, The Pulp and Paper Magazine of Canada. Only U.S. economic data are used to explain the pattern of innovation because of the frequent lack of comparable Canadian data and problems in combining data from two countries. The inclusion of Canadian economic data would not substantially alter the explanation, since the United States was by far the major market for industry output. Much of the innovative activity in the industry was performed by capital goods suppliers responding to the needs of pulp and paper firms. Implications of the role of capital goods suppliers are not analyzed since the discussion is limited to the focusing of innovative effort.Google Scholar For analysis of the role of capital goods suppliers see Rosenberg, N., “Technological Change in the Machine Tool Industry, 1840–1910,” in Perspectives;Google ScholarPeck, M. J., “Inventions in the Postwar American Aluminum Industry,” in The Rate and Direction of Inventive Activity, National Bureau of Economic Research (Princeton, 1962);Google ScholarBrown, W. H., “Innovations in the Machine Tool Industry,” Quarterly Journal of Economics, 71 (August 1957);CrossRefGoogle ScholarSchmookler, J., Invention and Economic Growth.Google Scholar

⁹ Computed from Series W30–54, “Indexes of Output Per Man-Hour for Production Workers, Selected Industries: 1909–1970”, U.S. Bureau of Census, The Statistical History of the United States (New York, 1976), p. 950.Google Scholar

¹⁰ Paper Trade Journal (hereafter PTJ) (11 January 1923), p. 46; PTJ (15 April 1920), P. 207; PTJ (24 March 1932), p. 46. Because of space limitations, the titles of Paper Trade Journal articles have been omitted. A complete bibliography is available from the author on request.Google Scholar

¹¹ PTJ (15 April 1920), p. 211.Google Scholar

¹² The first installation of any directly connected electric motors was reported to be “about 1906,” PTJ (13 April 1922), p. 43. The first installation of a sectional electric drive occurred in 1908, PTJ (5 November 1931), p. 48.Google Scholar

¹³ PTJ (9 October 1931), p. 49; PTJ (5 December 1931), p. 49. The first General Electric sectional paper drive was installed one year later, in 1920, PTJ (11 August 1921), p. 38.Google Scholar

¹⁴ PTJ (13 December 1940), p. 29;Google ScholarPTJ (13 June 1929), p. 83.Google Scholar

¹⁵ PTJ (11 January 1923), p. 46;Google ScholarPTJ (13 August 1936), p. 94; PTJ (25 February 1937), p. 149;Google ScholarPTJ (26December 1940), p. 30.Google Scholar

¹⁶ PTJ (26 December 1940), p. 31;Google ScholarPTJ (5 November 1931), p. 49.Google Scholar

¹⁷ PTJ (13 August 1936), p. 92.Google Scholar

¹⁸ PTJ (5 November 1931), p. 49.Google Scholar

¹⁹ PTJ (26 October 1933), p. 23;Google ScholarPTJ (13 June 1929), p. 83;Google ScholarPTJ (22 March 1923), p. 53.Google Scholar

²⁰ PTJ (13 June 1929), p. 83.Google Scholar

²¹ Longer wires also helped solve drainage problems, see PTJ (13 June 1929), p. 82;Google ScholarPTJ (13 August 1936), pp. 94–95;Google ScholarPTJ (11 January 1923), p. 46.Google Scholar

²² PTJ (9 June 1921), pp. 22–25;Google ScholarPTJ (11 January 1923), p. 47;Google ScholarWitham, Modern Pulp and Paper Making, pp. 312–13;Google ScholarPTJ (1 July 1937), p. 84.Google Scholar

²³ PTJ (5 November 1931), p. 48;Google ScholarPTJ (13 August 1936), p. 92;Google ScholarPTJ (1 July 1937), p. 84; “Progress in the Design of Paper Machines: 1902–1942,” Pulp and Paper Magazine of Canada (July 1942), p. 597.Google Scholar

²⁴ Lorant, J. H., The Role of Capital-Improving Innovations in American Manufacturing During the 1920's (New York, 1975), p. 135.Google Scholar

²⁵ For a detailed description of the Pope innovation by the inventor, see PTJ (18 April 1918), pp. 44–48 and PTJ (6 June 1918), pp. 44,46. For additional references to the Pope and Sheehan systems and paper web transferring in general,Google Scholar see PTJ (13 April 1922), p. 43;Google ScholarPTJ (11 January 1923), p. 47;Google ScholarPTJ (22 March 1923), p. 53;Google ScholarPTJ (24 March 1932), p. 46;Google ScholarPTJ (13 August 1936), p. 92.Google Scholar

²⁶ PTJ (13 June 1929), p. 84.Google Scholar

²⁷ “Progress in the Design of Paper Machines: 1902–1942,” pp. 597–98;Google ScholarPTJ (5 November 1931), p. 49.Google Scholar

²⁸ PTJ (5 November 1931), p. 49. As the following quotation from a 1929 papermaker indicates, the changes in bearings and lubrication were dramatic: “A few years ago we preferred a chunk of suet for lubrication as we felt sure that the journal [bearing] would always have some grease… Then we passed through several stages of oiling, but on the paper machine of today the dryer is supported entirely in … anti-friction bearings tightly concealed, with oil circulating through these bearings out to a filter and to a cooler with a pump return”. PTJ (13 June 1929), p. 84.Google Scholar

²⁹ For example, sectional electric drive could be installed on existing paper machines, as the following advertisement from Westinghouse boasts:Google Scholar“Sectional individual motor drive with automatic control … [makes] possible increased speed and production of 25–50 per cent on many existing machines.”Google ScholarPTJ (19 05 1921), p. 31,Google Scholarcited in Lorant, The Role of Capital-Improving Innovations, p. 136.Google Scholar

³⁰ PTJ (6 June 1918), p. 44.Google Scholar

³¹ “Progress in the Design of Paper Machines: 1902–1942,” p. 596.Google Scholar

³² PTJ (13 May 1920), p. 40.Google Scholar

³³ Smith, D. C., History of Papermaking in the U.S. 1691–1969 (New York, 1970), p. 620.Google Scholar

³⁴ Jerome, H., Mechanization in Industry (New York, 1934), pp. 88–89.Google Scholar

³⁵ Lorant, The Role of Capital-Improving Innovations, p. 146.Google Scholar

³⁶ According to one 1942 paper mill engineer, “With each successive increase in speed, there has been some setback to sheet formation, with the result that compensating changes have been required. Thus, in spite of much higher speeds, formation is today better than for lower speeds several years ago.” “Progress in the Design of Paper Machines: 1902–1942,” p. 597.Google Scholar

³⁷ PTJ (24 March 1932), p. 46.Google Scholar

³⁸ On the use of patent data as a proxy for innovative activity, see Schmookler, Invention and Economic Growth, pp. 18–56,Google Scholarand Comanor, W. S. and Scherer, F. M., “Patent Statistics as a Measure of Technological Change,” Journal of Political Economy, 77 (05/June 1969).CrossRefGoogle Scholar

³⁹ The concept of internal resources is similar to Chandler's, A. “firm structure” in Strategy and Structure (Cambridge, Massachusetts, 1962), p. 14.Google Scholar

⁴⁰ Penrose, Theory, pp. 4–5.Google Scholar

⁴¹ For discussion of the evolutionary bias created by interdependence, see Frankel, M., “Obsolescence and Technological Change in a Maturing Economy,” American Economic Review, 45 (06 1955), 296–319;Google ScholarAmes, E. and Rosenberg, N., “Changing Technological Leadership and Industrial Growth,” Economic Journal, 73 (03 1963);CrossRefGoogle ScholarKindleberger, C., “Obsolescence and Technical Change,” Oxford University Institute of Statistics Bulletin (August 1961);Google ScholarHicks, J. R., The Theory of Wages (London, 1932);Google ScholarSalter, W. E. G., Productivity and Technical Change (Cambridge, 1966);Google ScholarHamberg, D., “Invention in the Industrial Research Laboratory,” Journal of Political Economy, 71 (04 1963), 95–115.CrossRefGoogle Scholar

⁴² See Mansfield, E., The Economics of Technological Change (New York, 1968), pp. 92–93.Google Scholar

⁴³ As one industry scientist put it, a new innovation “needs to be adaptable to current hardware if it is to be used in existing plants because of the tremendous investments that have already been made in these facilities. If the new process does not fit present configurations, it must then be installed as a new mill—and this requires faith and audacity.” Strange, J. G., The Paper Industry: A Clinical Study (Appleton, Wisconsin, 1977), p. 77.Google Scholar

⁴⁴ For independent estimates of increasing returns to scale cost functions see PTJ (20 February 1941), pp. 129–32,Google Scholarand Armstrong, G. R., “An Economic Study of New York's Pulp and Paper Industry,” monograph (Syracuse, 1968), p. 73. Scale economies also were produced by labor requirements, which were fixed by the size of the machinery and were relatively invariant to operating speed. One paper manufacturer wrote: “a machine running at three or four hundred feet per minute will require approximately the same amount of help in the machine room as a machine running eight hundred or a thousand feet per minute or over.”Google ScholarJerome, Mechanization, pp. 249–50. Unit labor costs decreased from about $34/ton for small mills (daily capacity of 10 tons or less) to $8/ton for the largest mills (over 200 tons daily capacity), American Paper and Pulp Association, Monthly Review (March/April/May/ 1935), p. 13.Google Scholar The importance of labor factors, however, should not be overestimated. Federal Trade Commission reports on 1915 newsprint and bookpaper mills found that direct labor costs constituted only 10.0 percent of total costs, cited in Clarke, P. J., “Monopoly and Competition in the American Paper Industry” (Ph.D. diss., Graduate School of Business Administration, New York University, 1941), pp. 65–66.Google Scholar

⁴⁵ “Economics of Paper,” Fortune (October 1937), p. 116.Google Scholar

⁴⁶ U.S. Tariff Commission, Report on Wood Pulp and Pulpwood, Report No. 126, Second Series (Washington, D. C., 1938), p. 85.Google Scholar

⁴⁷ Stevenson, L. T., The Background and Economics of American Papermaking (New York, 1940), p. 83.Google Scholar Another author notes that “Mills and machines are massive as well as highly specialized, and it is impractical to move them once they are built.” Whitney, S. N., “Paper” ch. 6 in Antitrust Policies: American Experience in Twenty Industries (New York, 1958), p. 368.Google Scholar

⁴⁸ PTJ (21 July 1921), p. 30.Google Scholar

⁴⁹ Witham, Modern Pulp and Papermaking, p. 281.Google Scholar

⁵⁰ Stevenson, Background and Economics, p. 83. An added consequence of the complexity and delicacy of the machine was a long teardown time, which resulted in a significant loss of production. Given the importance of volume of production in the industry, this imposed an additional expense on the selling or complete replacement of a fourdrinier machine.Google Scholar

⁵¹ U.S. Tariff Commission, Report on Wood Pulp, pp. 92, 94. Mills included in the data produced 93 percent of all U.S. wood pulp in 1934.Google Scholar

⁵² Whitney, “Paper,” p. 363.Google Scholar

⁵³ Lockwood Trade Journal Company, Directory of the Paper and Applied Trades (New York, 1910, 1920, 1930, 1940).Google Scholar

⁵⁴ “A few experiments, to be sure, have been tried with the sulphite process, but these have led to somewhat negative results, on account of the excessive costs of using a commercial plant for experimental purposes. This is one reason why the paper trade up to the present has not done more along this line.” 60th Congress, 2nd Session, House of Representatives, Select Committee, Pulp and Paper Investigation Hearings, Doe. No. 1502 (Washington D. C., 1909), p. 1461. Also see Carter, C. F. and Williams, B. R., Industry and Technical Progress (New York, 1957), p. 232Google Scholar and Rauch, J., ed., The Kline Guide to the Paper and Pulp Industry (Fairfield, New Jersey, 1976), p. 52.Google Scholar

⁵⁵ PTJ (13 April 1922), p. 41. Fortune magazine noted that in 1937 there were six fourdrinier machines in use that had been operating since 1860.Google Scholar“Economics of Paper,” p. 184.Google Scholar

⁵⁶ PTJ (3 October 1935), p. 13.Google Scholar

⁵⁷ American Paper and Pulp Association, Monthly Review, 6 (February 1939), p. 19,Google Scholar cited in Clarke, , “Monopoly and Competition,” p. 172.Google Scholar

⁵⁸ Rosenberg, “The Direction of Technological Change,” p. 111.Google Scholar

⁵⁹ Smith, History of Papermaking, ch. 10. By December 1917, the Paper Trade Journal reported that: “The paper mills in this city [Holyoke, Mass.” and locality continue to do a flourishing business. Manufacturers … state they have never seen such a strong demand, with the volume becoming greater and greater as the weeks slip by. Conditions show that the market is not a speculative one. The first indications of the returns to good business were discerned last February and since that time the improvement has been gradual and has become more and more intensified. Taken on the whole the local paper mills are now booked up with orders that bring the business up to more than 115 per cent normal.”Google ScholarIbid, pp. 324–25.

⁶⁰ Neubrech, W. L. and Schumacher, A. C., “The Pulp and Paper Industry in War and Peace,” Survey of Current Business, 22 (12 1942), p. 15.Google Scholar

⁶¹ Guthrie, J., The Newsprint Paper Industry, An Economic Analysis (Cambridge, Massachusetts, 1941), p. 121.Google Scholar

⁶² “What Were and Are the Markets for All Grades of Paper,” ch. 11 in Lockwood Trade Journal Company, The Progress of Paper (New York, 1947), p. 147.Google Scholar

⁶³ PTJ (2 June 1921), p. 44.Google Scholar

⁶⁴ “15 Paper Companies,” Fortune (November 1937), p. 196.Google Scholar

⁶⁵ See Devine, W. D. Jr, “From Shafts to Wires: Historical Perspective on Electrification,” this Journal, 43 (June 1983), 347–72.Google Scholar

⁶⁶ “The years between 1931 and 1935 were characterized by decreasing markets for paper and demands on the part of publishers for improved quality. With considerable excess productive capacity in the industry, no new machines were installed in this country [U.S.], and with special emphasis placed on quality, there was little incentive to the individual mill to increase speed of existing equipment. PTJ (13 August 1936), p. 94.Google Scholar

⁶⁷ PTJ (11 April 1935), p. 20 (emphasis added).Google Scholar

⁶⁸ PTJ (13 August 1936), p. 95.Google Scholar

⁶⁹ Usher, A. P., A History of Mechanical Inventions, rev. ed. (Cambridge, Massachusetts, 1954), p. 41.Google Scholar

⁷⁰ Although Binswanger will not relinquish the technique of constrained optimization, which is not helpful for explaining innovative attempts to transcend constraints, he arrives at the period-by- period historical approach as the most fruitful for the analysis of technological change: “Equilibriurn growth models are unable to capture many important features of the development process precisely because of their orientation toward equilibrium states and because they take as given so many things that will change in the development process itself, such as market structures, infrastuctural equipment, technology, tastes, the structure of property rights, and many other institutions. Development should rather be viewed as an iterative adjustment process to disequilibria caused by the development and production process itself and by exogenous shocks and influences. A fruitful approach might be to view development as a sequence of periods. At the beginning of each period, technology, tastes, capital stocks, and institutions are given and can be taken as parameters that enter the optimization problems of the participants in the development process. Individuals and organizations then allocate resources to production, investment, technical change, and institutional change. In addition, exogenous shocks such as weather, developments in international markets, advances in basic and supporting sciences, and changes in ideas and tastes occur. Individual allocation decisions and exogenous shocks jointly determine production and consumption flows and changes in stocks, technology, institutions, and possibly, tastes. New disequilibria emerge. The next period's problem then starts with those newly created parameters, and a new round is initiated. In a way, this is a sort of dynamic programming model, except that individual actors have different objective functions. Such an approach lacks all the neatness of equilibrium growth models and will be extremely difficult to put into strict modeling frameworks. But thinking and looking at subproblems in these terms might allow us to learn more about development than we can learn from growth models.” Binswanger and Ruttan, Induced Innovation, pp. 157–58.Google Scholar

⁷¹ According to Schumpeter in “The Creative Response in Economic History,”, innovation, as an entrepreneurial or creative response, “can always be understood ex post; but it can practically never be understood ex ante; that is to say, it cannot be predicted by applying the ordinary rules of inference from the pre-existing facts. That is why the ‘how’ [of the mechanism through which it acts] must be investigated in each case (p. 150).” Schumpeter's “mechanism” that must be individually investigated is the focusing process of the theoretical framework—the interaction of firms attempting to overcome constraints on accumulation within the historically specific structure of production and demand. For a general discussion of the methodological significance of causal mechanisms in economic explanations see Cohen, A., “The Methodological Resolution of the Cambridge Controversies,” Journal of Post Keynesian Economics, 6 (Summer 1984).CrossRefGoogle Scholar