¹ Chandler, Alfred D. Jr., The Visible Hand: The Managerial Revolution in American Business (Cambridge, Mass., 1977)Google Scholar; Chandler, , Scale and Scope: The Dynamics of Industrial Capitalism (Cambridge, Mass., 1990)Google Scholar.

² The use of “spectrum” here should not be taken to imply a simple linear matching of firm and format. As the notion of “bridge” firms suggests, large enterprises like General Electric made custom-designed turbines, batch-produced parts, and mass-produced light bulbs. Smaller metalworking firms could mix making long runs of bulk items with batch work to fill idle machine time or to respond to seasonal slackening of orders. “Spectrum” thus indicates a range of choices for decision, though institutional routines and technical capacities did lead many firms to confine their operations to one format.

³ Lamoreaux, Naomi, The Great Merger Movement in American Business, 1895–1904 (New York, 1985)CrossRefGoogle Scholar; Hounshell, David, From the American System to Mass Production (Baltimore, Md., 1984)Google Scholar; Strasser, Susan, Satisfaction Guaranteed: The Making of the American Mass Market (New York, 1989)Google Scholar.

⁴ Engineering Magazine 19 (1900): 100–104Google Scholar.

⁵ Industrial Management 52 (1916–1917): 633–37Google Scholar.

⁶ The distinction drawn between bulk and mass production is based on observing that, alongside the structured sequential systems for achieving throughput and uniformity (Hounshell, American System), less elaborate formats also operated in which staple products were created without throughput flows. Whereas mass production aptly covers the making of Model Ts or the canning of foodstuffs, bulk production may better suit the weaving of print cloths, cutting of dimensional lumber, or manufacture of nails. In these latter cases, no integrated machine system was involved, yet enormous volumes of goods were made for stock.

⁷ Industrial Management 52 (1916–1917): 633Google Scholar.

⁸ Quantifying the role of batch production in overall industrial output is a difficult challenge, given the mix in many “industry” census categories of outputs from differing formats. A crude estimation, based on 1909 Manufacturing Census tables, suggests that custom and batch sectors contributed a bit less than one half the value added in manufacturing that year; Scranton, Philip, “Industrial Formats and Prospects for Research in American Business History,” unpub. paper, 1991, tables 1–3Google Scholar.

⁹ Iron Age 137 (25 June 1936): 22–63Google Scholar; Reutter, Mark, Sparrows Point (New York, 1988)Google Scholar; Meyer, Stephen, “Technology and the Workplace: Skilled and Production Workers at Allis-Chalmers, 1900–1941,” Technology and Culture 29 (1988): 839–64CrossRefGoogle Scholar; Chandler, Alfred D. Jr., and Tedlow, Richard S., The Coming of Managerial Capitalism (Homewood, Ill., 1985), 680–91Google Scholar; Hughes, Thomas, American Genesis: A Century of Invention and Technological Enthusiasm, 1870–1970 (New York, 1990), 175–77Google Scholar.

¹⁰ Chandler, Visible Hand, 455–63; Chandler and Tedlow, Coming of Managerial Capitalism, 696–706.

¹¹ Metal Industry 10 (1912): 71Google Scholar.

¹² Machinery 14 (1907–1908): 291–95Google Scholar. A photograph of the completed planer, adorned by the workers who built it, has been reproduced from the Machinery article in American Heritage of Invention and Technology 4 (Fall 1988): 60–61Google Scholar.

¹³ Machinery 15 (1908–1909): 686–87Google Scholar. This feat was repeated in sixteen days for a comparable engine in 1911; Machinery 18 (1911–1912): 380Google Scholar.

¹⁴ On shipbuilding, see Palmer, David, “Organizing the Shipyards” (Ph.D. diss., Brandeis University, 1990Google Scholar), introduction, chaps. 1–5; Mattox, W. C., Building the Emergency Fleet (Cleveland, Ohio, 1920Google Scholar), chap. 15. For Baldwin, see Westing, Fred, The Locomotives Baldwin Built (Seattle, Wash., 1966)Google Scholar; Vauclain, Samuel, Steaming Up! (New York, 1930)Google Scholar; and Custer, Edgar, No Royal Road (New York, 1937)Google Scholar. At the University of Virginia, John Brown is presently completing a dissertation on Baldwin through the First World War, which promises to add greatly to our understanding of this firm; Business History Bulletin 4: 2 (1990): 508Google Scholar.

¹⁵ Here “scope” carries a meaning distinct from that employed in Chandler's recent Scale and Scope. There, scope seems to have been used in two ways, to refer to the addition of product lines by diversifying corporations and the extension of existing production capacities to manufacturing new and related items, both oriented to discovering goods with large market potential, a process that might be termed “scale-seeking” scope. Some batch firms pursued such strategies, and the multiple-format “bridge” giants all did so, but for most batch operations, achieving scope meant having a sufficient range of technology and skill on hand to take on novel products within a firm's chosen area of activity, learning from the experience in making them, and, at times, testing out new input materials, stylistic forms, or channels for contracting. The search criteria for them, if such a term be permitted, were not framed to yield products with large runs and large, durable market share, but to stretch the capacity of the enterprise to cover novel profit possibilities emerging from the uncertainties of technical or fashion shifts.

¹⁶ Engineering Magazine 19 (1900): 104Google Scholar.

¹⁷ Scranton, Philip, Figured Tapestry: Production, Markets, and Power in Philadelphia Textiles, 1885–1941 (New York, 1989)Google Scholar; Sabel, Charles, Work and Politics (New York, 1982)CrossRefGoogle Scholar. By no means were all batch producers located in urban districts; locational isolation was long critical to the success of generalist metalworking firms, for example.

¹⁸ Michael Storper, “The Transition to Flexible Specialization,” Cambridge Journal of Economics (forthcoming).

¹⁹ Jean Saglio, “Rationalité, Economique, et Comportement des Acteurs dans un Système Industriel Localisé,” and Rudolf Boch, “The Rise and Decline of Flexible Production: The Cutlery Industry of Solingen since the Eighteenth Century,” papers presented at the International Working Group on Historical Alternatives to Mass Production, Paris and Lyon, 1989–90. For early work on the American sites, see de Geer, Sten, “The American Manufacturing Belt,” Geografisker Annaler 9 (1927): 233–357CrossRefGoogle Scholar.

²⁰ On governance, see Michael Storper and Bennett Harrison, “Flexibility, Hierarchy, and Regional Development,” paper presented at the Industrial Policy Conference, Bologna, 1989. For examples of local institutional development, see Ransom, Frank, City Built on Wood (Grand Rapids, Mich., 1955)Google Scholar; Harris, Howell, “Getting It Together: The Metal Manufacturers' Association of Philadelphia, c. 1900–1930,” in Masters to Managers, ed. Jacoby, Sanford (New York, 1991), 111–31Google Scholar; and Scranton, Philip, “A Productive Diversity: New Jersey Industrialization, 1870–1930,” New Jersey History Seminar, Princeton, 1990Google Scholar.

²¹ Stocking, George, Basing-Point Pricing and Regional Development (Chapel Hill, N.C., 1954)Google Scholar; Federal Trade Commission (FTC), “In the Matter of U.S. Steel Corporation, et al.,” Docket 760, 8 FTC Decisions 1 (1924)Google Scholar; Strasser, Satisfaction Guaranteed, 269–85. Resale price maintenance of branded goods was hotly opposed by retailers into the 1930s. For a particularly vivid critique, see Q. F. Walker's brief for Macy's and the Retail Code Authority's protest of proposals before the NRA (National Archives, Records of the National Recovery Administration, Record Group 9, box 3497, file 103–20 [hereafter, NA, RG9]).

²² In 1910 machine tool builders estimated that 80 percent of their trade was handled through dealers; Machinery 16 (1909–1910): 872Google Scholar.

²³ For detailed treatment of these relations in fashion textiles, see Scranton, Figured Tapestry, which also discusses the decline that emerged once a “permanent” buyers' market was institutionalized in the 1920s.

²⁴ Mass, William, “Technological Change and Industrial Relations” (Ph.D. diss., Boston College, 1985)Google Scholar; Pratt, and Whitney Company, Accuracy for Seventy Years (Hartford, Conn., 1930)Google Scholar; Hoe, R. and Company, Catalogue of Printing Presses and Printers' Materials (New York, 1881)Google Scholar; Tucker, Stephen, “History of R. Hoe and Company,” Proceedings of the American Antiquarian Society 82 (1972): 351–453Google Scholar; Rice, Barton and Fales, , A Line of Men One Hundred Years Long (Worcester, Mass., 1937)Google Scholar; Hounshell, American System, chaps. 2 and 6.

²⁵ Reutter, Sparrows Point; Meyer, , “Technology and the Workplace”: Machinery 7 (1900–1901): 1–4Google Scholar; 8 (1901–2): 337–43; 9 (1902–3): 285–88; 24 (1917–18): 1073–81 (for floor plans, plates, and portable tools). For GE, see Ripley, Charles, Romance of a Great Factory (Schenectady, N.Y., 1919), 72–79, 108–17Google Scholar; and for Westinghouse at Pittsburgh, East, Machinery 17 (1910–1911): 841–42Google Scholar.

²⁶ Williams, Karel et al. , “The End of Mass Production?” Economy and Society 16 (1987): 423–24Google Scholar.

²⁷ Lazonick, William, Competitive Advantage on the Shop Floor (Cambridge, Mass., 1990)Google Scholar, chaps. 7 and 8.

²⁸ Chandler, Visible Hand, 249–50, 290–91; Cooper, Patricia, Once a Cigar Maker (Urbana, Ill., 1987)Google Scholar.

²⁹ Manufacturing Jeweler 2 (1885–1886): 170, 215Google Scholar; 24 (1899): 171, 184. Real dollar conversions made using Wholesale Price Indexes E–1 and E–13, Department of Commerce, Historical Statistics of the United States (Washington, D.C., 1976)Google Scholar.

³⁰ Nelson, Daniel, Managers and Workers: Origins of the Factory System in the United States (Madison, Wisc., 1975), 96–99Google Scholar; Noble, David, America By Design (New York, 1977), 300–305Google Scholar. For apprenticeship at Baldwin, see Engineering Magazine 27 (1904): 321–33Google Scholar; Pratt, and Whitney, , Machinery 23 (1916–1917): 455–61Google Scholar; Brown, and Sharpe, , Engineering Magazine 26 (1903–1904): 511–19Google Scholar; and GE, , Machinery 18 (1911–1912): 619–23Google Scholar.

³¹ These reconstructions were both contested by craft unions and intensely debated in the industrial press. As A. E. Outerbridge remarked crisply: “As new machines can be produced more rapidly than skilled operatives can be created, the balance between human skill to direct and mechanical power to execute… has been overturned, and instead of a surplus of skilled labor crowded out by labor-saving machines, we find an opposite condition.” Engineering Magazine 10 (1895–1896): 228Google Scholar. For other views, see Machinery 3 (1896–1897): 58–59Google Scholar; 4 (1897–98): 312; Manufacturing Jeweler 3 (1886–1887): 478Google Scholar; 8 (1890–91): 803; 24 (1899): 453–54; 28 (1901): 502. In batch operations, specialists mastered a class of work or tools, becoming moderately versatile, rather than being assigned to repetition production on dedicated machines, as in line assembly or Northrop weaving.

³² Lazonick, Competitive Advantage, chap. 8. Jacoby, Sanford differs from Lazonick's assessment; see his Employing Bureaucracy (New York, 1985)Google Scholar.

³³ Vauclain, Steaming Up!, 158–71. Vauclain described the locomotive trade thus: “Sharp fluctuation of demand; impossibility of manufacturing in advance of orders; impracticability of mass production; necessity for employing specialists; high cost of each completed item and varying conditions under which it is sold and used call for unusual adjustability, flexibility, ingenuity, and resourcefulness by executives” (p. 146). Infrastructure firms were especially hard hit in general depressions. After 1893, GE sales dropped 75 percent and the company laid off nearly two-thirds of its workers, retaining a core of 3,000 even though “there was not enough work to keep [them] occupied full time.” Broderick, John, Forty Years at General Electric (Albany, N.Y., 1929), 31Google Scholar.

³⁴ Vauclain, Steaming Up!, 149, 158, 167–68; Vauclain, , Optimism (Philadelphia, Pa., 1924), 293Google Scholar; Machinery 2 (1895–1896): 15, 389Google Scholar; Engineering Magazine 8 (1894–1895): 57Google Scholar. For veteran workers at GE, see Ripley, Romance of a Great Factory, 63–66.

³⁵ Scranton, , Figured Tapestry; Manufacturing Jeweler 9 (1981): 803Google Scholar; Machinery 3 (1896–1897): 58, 126–27Google Scholar; 6 (1899–1900): 280–81; 8 (1901–2): 322–23; 11 (1904–5): 551; Kleiman, Jeffrey, “The Great Strike: Religion, Labor and Reform in Grand Rapids, 1890–1916” (Ph.D. diss., Michigan State University, 1985Google Scholar). the “floater” or “tramp” printer, machinist, or woodworker was a folkloristic character, frequently discussed in trade press columns. Writers claimed that, although often these travelers carried wide experience into the shops, at times they turned out to be worthless “fakers,” trying to “steal the trade” by pretending competences. On these issues and much else concerning skilled labor, see Montgomery, David, Workers' Control in America (New York, 1979)Google Scholar, and Montgomery, , The Fall of the House of Labor (New York, 1987), chaps. 1, 4, and 5CrossRefGoogle Scholar.

³⁶ Scranton, Philip, “The Politics of Production: Technology, Markets, and the Two Cultures of American Industry,” in Technology, Aesthetics, Culture, Politics, ed. Block, Fred, Rieber, Alfred, and Siefert, Marsha (New York, forthcoming)Google Scholar.

³⁷ Baron, Ava, “Questions of Gender: Deskilling and Demasculinization in the U.S. Printing Industry, 1830–1915,” Gender and History 1 (1989): 178–99CrossRefGoogle Scholar; Milkman, Ruth, ed., Women, Work and Protest (Boston, 1986)Google Scholar; Jensen, Joan and Davidson, Sue, eds., A Needle, A Bobbin, A Strike (Philadelphia, 1984)Google Scholar; Manufacturing Jeweler 29 (1901): 380Google Scholar.

³⁸ Vauclain, Optimism, 4–5. Vauclain thought firm loyalty was secured through fair dealing with individuals, but others relied on familistic gestures, including sports teams, banquets, holiday bonuses and excursions, plus hiring preferences for workers' kin. Even Vauclain laid off single men first, another common practice. Unlike corporate welfare schemes, which trended toward insurance, pensions, and literacy classes, paternalist provision embodied a direct gift relation, the manly proprietor caring for his community, rather than boards of directors' cost-benefit calculations. See also Montgomery, Workers' Control, 13–15; Montgomery, Fall of the House of Labor, 17–22, 454–57; and Scranton, Philip and Licht, Walter, Work Sights (Philadelphia, Pa., 1986), 188–91Google Scholar.

³⁹ Nelson, Managers and Workers, chap, 4.

⁴⁰ On payment controversies, see Engineering Magazine 17 (1899): 444–54, 925–36Google Scholar; 18 (1899–1900): 572–54, 690–96; 19 (1900): 373–80; 20 (1900–1901): 631–40; 25 (1903): 533–38, 699–713; and Montgomery, Fall of the House of Labor, chap. 5. For simple means to speed work, see Machinery 7 (1900–1901): 201–6Google Scholar; 12 (1905–6): 33–34. On woodworking and piece rates, see The Wood-Worker 26 (Sept. 1907): 31Google Scholar; 31 (May 1912): 23–24, (Sept. 1912): 45–46, and (Dec. 1912): 27. For brief accounts of bonus plans and basic bibliography, see George, Claude, A History of Management Thought (Englewood Cliffs, N.J., 1967)Google Scholar. Guaranteeing piece rates against cuts minimized conflict and was long standard practice at Baldwin, Brown and Sharpe, and Sellers machine tool; Nelson, Managers and Workers, 45. During the First World War, the Hog Island Shipyards installed such terms, securing extraordinary results; Industrial Management 60 (1920): 409–12Google Scholar. Absent guarantees, workers widely restricted output to avoid killing the rate, as documented over the last century. See Wright, Carroll, Eleventh Special Report of the Commissioner of Labor (Washington, D.C., 1904)Google Scholar; Mathewson, Stanley, Restriction of Output among Unorganized Workers (New York, 1931)Google Scholar; Roy, Donald, “Banana Time: Job Satisfaction and Informal Interaction,” Human Organization 18 (1958): 158–68CrossRefGoogle Scholar; and Burawoy, Michael, Manufacturing Consent (Berkeley, Calif., 1985)Google Scholar.

⁴¹ Engineering Magazine 21 (1901): 586–90Google Scholar; Kleiman, “Great Strike”; Scranton, Figured Tapestry, 210–26.

⁴² For examples of defensive unionism paired with hard-times mushroom firm starts in textiles, see McLewin, Philip, “Labor Conflict and Technical Change,” in Silk City, ed. Scranton, Philip (Newark, N.J., 1985), 135–58Google Scholar; David Goldberg, “The Battle for Labor Supremacy in Paterson,” in ibid., 107–34; Goldberg, , A Tale of Three Cities (New Brunswick, N.J., 1989Google Scholar), and Scranton, Figured Tapestry, 427–96.

⁴³ Astute union leaders dealing with batch trades crafted local market agreements that recognized such differences, notably Sidney Hillman's Amalgamated Clothing Workers. See Zaretz, Charles, The Amalgamated Clothing Workers (New York, 1934)Google Scholar, and Carpenter, Jesse, Competition and Collective Bargaining in the Needle Trades (Ithaca, N.Y., 1967)Google Scholar.

⁴⁴ My sense of contemporary labor union discourse is that leaders sought national presence and uniformity, but often without full understanding of the power of local specificities. Struggling for national wage and hours standards really made sense only in sectors that were product-standardized, but this was poorly perceived by union members and officers.

⁴⁵ Reeves, William, “Machinery Salesmanship,” Journal of the Association of Engineering Societies 53 (1914): 222Google Scholar.

⁴⁶ Vauclain, Steaming Up!, 200–212; Engineering Magazine 17 (1899): 545–61, 968–76Google Scholar; Chandler, Scale and Scope, 276–77, 351–53; R. C. Brill Collection, Historical Society of Pennsylvania, Photograph Albums.

⁴⁷ On the difficulty of fitting this species of demand neatly to business cycles, see Kuznets, Simon, Economic Change (New York, 1953), chaps. 3–5Google Scholar, especially appendix A, chap. 3, on railway demand, 1891–1930. For seasonality issues, see Kuznets, , Seasonal Variations in Business and Trade (New York, 1933), 242–75, 312–21Google Scholar.

⁴⁸ Vauclain, Steaming Up!, 129–30, 186–90, 206; Broderick, Forty Years, 31, 42–46. In 1891 Baldwin ran one of its early compound locomotives 50,000 miles, coast to coast, at a reported expense of $200,000, a stunt Vauclain credited with helping sell some 3,000 compounds (114, 186). See also Nye, David, Electrifying America: Social Meanings of a New Technology (Cambridge, Mass., 1990)Google Scholar, chap. 3. Trenton's Roebling wire cable works ensured demand by buying large blocks of Otis Elevator shares; Zink, Clifford and Hartman, Darcy, “Spanning the Industrial Century: Development of the John A. Roeblings' Sons Company,” unpub. paper, Trenton-Roebling Project, 1989, 80Google Scholar.

⁴⁹ Cramp's Shipyards (Philadelphia, Pa., 1910), 68–71, 78, 81, 88–89Google Scholar; Vauclain, Steaming Up!, 121–29, 139–40. For continuity in such relations after the 1930s, see Scherer, F. M., Industrial Market Structure and Economic Performance (Chicago, Ill., 1970), 160Google Scholar.

⁵⁰ Lester, Bernard, Marketing Industrial Equipment (New York, 1935)Google Scholar. The survey had been conducted by American Machinist, following up a 1925 study.

⁵¹ Seville, Lloyd, “Price Determination in the Gray-Iron Foundry Industry” (Ph.D. diss., Columbia University, 1950)Google Scholar; Rice, Barton and Fales, Line of Men, 11–14. For a brief theorization of pricing and markets in multi-product batch firms, stressing their sales of “not specific products, but the capacity to produce,” see Scherer, F. M., Industrial Pricing (Chicago, Ill., 1970), 131–33Google Scholar. Scherer concludes that assessing the allocative efficiency of such firms “is so complex that no confident generalizations seem feasible” (p. 133).

⁵² Jerome, Harry, Mechanization in Industry (New York, 1934), 179, 300–304Google Scholar; Wagoner, Harless, The U.S. Machine Tool Industry from 1900–1950 (Cambridge, Mass., 1968), 133–35Google Scholar. Wagoner noted that no evidence was found that the machine tool industry's sales problems were “essentially different from those of builders of other types of capital equipment” (p. 95).

⁵³ Machinery 15 (1909–1910): 920–21Google Scholar; Wagoner, U.S. Machine Tool Industry, 95–97.

⁵⁴ Intense product competition was the other side of this absence of brand loyalty. See Broehl, Wayne, Precision Valley (Englewood Cliffs, N.J., 1959), 64–65, 122–23Google Scholar. On exporting, see ibid., 78, 89, 133, 152, and Wagoner, U.S. Machine Tool Industry, 98–101, 214–16. In 1913, a tough year, tool exports rose above one-quarter of sales, but in 1929 they were only one-tenth, rising again above a quarter in 1932.

⁵⁵ See Bethlehem Steel Corporation, Catalog No. 132 (Bethlehem, Pa., 1935)Google Scholar; W. S. Tyler Co., Catalog No. 70 (Cleveland, Ohio, n.d.).

⁵⁶ Casanave Supply Company, Philadelphia, listed about 30,000 items in its 1929 catalogue, whereas the same city's Theo C. Ulmer offered over 20,000 in 1942, both handling factory supplies and equipment. Specialty printers, R. R. Donnelly and Sons of Chicago, better known for telephone books, were major printers of trade catalogues. As early as 1879, P. & F. Corbin's New Britain hardware products catalogue spanned 834 pages, up from 320 in 1871. Corbin sold direct from the factory and a New York office. Corbin, P.&F., History of the House of P. & F. Corbin (New Britain, Conn., 1904), 53–58, 63–65Google Scholar.

⁵⁷ Scranton, Figured Tapestry, 69–80, 142–49, 329–47.

⁵⁸ Zappalla, Frederick, “A Financial History of the American Woolen Company” (MBA Thesis, University of Pennsylvania, 1947)Google Scholar; Marchand, Roland, Advertising the American Dream (Berkeley, Calif., 1985), 123–27Google Scholar.

⁵⁹ See U.S. House of Representatives, Committee on Ways and Means, Tariff Hearings: 1913 (Washington, D.C., 1913)Google Scholar, for sharp exchanges over expected duty reductions by the new Democratic administration.

⁶⁰ For textiles, see Scranton, Figured Tapestry; on jewelry, Manufacturing Jeweler 1–32 (1884–1903): passim, and Shapiro-Perl, Nina, “Labor Process and Class Relations in the Costume Jewelry Industry” (Ph.D. diss., University of Connecticut, 1983)Google Scholar. For furniture, Kleiman, “The Great Strike,” and Ransom, City Built on Wood, cover Grand Rapids well. Trade discounts on invoices were often scaled—for example, 6 percent if paid in ten days, 4 percent until thirty or sixty days, net thereafter. Late payers often took expired discounts, a nagging trade abuse. For department stores, see Benson, Susan, Counter Cultures: Saleswomen, Managers, and Customers in American Department Stores, 1890- 1940 (Urbana, Ill., 1986)Google Scholar, and National Retail Dry Goods Association, Twenty-Five Years of Retailing (New York, 1936)Google Scholar.

⁶¹ Manufacturing Jeweler 28 (1901): 570Google Scholar. The catalogue firm was S. O. Bigney of Attleboro, Mass. In 1929, textiles and apparel represented 69 percent of department store sales versus 4 percent for furniture and 2.6 percent for jewelry, silverware, and books combined. Beckman, Theodore, The Changing Competitive Position of Department Stores (Ann Arbor, Mich., 1957Google Scholar; rev. ed., 1961), tables 14–16. On buyers, see also Henry, Robert, “The Changing Roles and Status of Department Store Buyers, 1870–1960” (Ph.D. diss., Syracuse University, 1965Google Scholar). For furniture, see Selby, John, “Industrial Growth and Worker Protest… High Point, N.C.” (Ph.D. diss., Duke University, 1984)Google Scholar; Ransom, City Built on Wood; and Nicklason, Clarence, Furniture Distribution in the Midwest (Washington, D.C., 1932)Google Scholar.

⁶² Scranton, Figured Tapestry, 397. Furniture contracts for business were reported in Furniture Manufacturer and Artisan's trade notes columns, with Jamestown companies strongly oriented toward office purchasers. Analogous reports on contracts for furnishing textiles to hotels, moving picture chains, and the like were printed in American Carpet and Upholstery Trade.

⁶³ Jewelry's problems commenced before the First World War, as will be noted below, but fashion textiles, furniture, machine tools, and heavy transportation equipment had serious difficulties in the 1920s. On the other hand, batch specialists attuned to the parts needs of the auto and electrical sectors, firms often located in small cities and towns, expanded operations, making long runs of varied components and using lower skilled (and lower paid) labor than did earlier metropolitan batch companies. My thanks to the anonymous Business History Review referee for pointing out this variant, which has conceptual implications beyond the scope of this essay.

⁶⁴ Scranton, Figured Tapestry, 256–58. A 1909 report of a metal trade association discussion noted: “upon sounding the sentiments of the different members as to what should constitute a fair percentage of profit to add to manufacturing costs to obtain selling prices[,] there was found to be but little difference between the highest and lowest estimates.” Metal Industry 8 (1910): 24Google Scholar. the markups averaged about 20 percent, from which selling expenses were to be deducted.

⁶⁵ Broderick, Forty Years, 75–82; Seville, “Gray-Iron Foundry,” 44–52, 68–71; Machinery 18 (1911–1912): 97, 301Google Scholar.

⁶⁶ For examples, see Manufacturing Jeweler 15 (1894): 3–4Google Scholar; Furniture Manufacturer and Artisan 62 (1911): 45–47, 240–42, 313Google Scholar; 78 (Jan. 1919): 20–222; Potter, H. D., Cost-Finding Principles for Furniture Factories (Grand Rapids, Mich., 1924Google Scholar), esp. chap. 8. The key person in devising accounting methods for batch production, whereby costs could be tracked for multiple items in process, was Church, Alexander. See his Proper Distribution of Expense Burden (New York, 1908)Google Scholar; Production Factors in Cost Accounting (New York, 1910)Google Scholar; and Overhead Expense (New York, 1930)Google Scholar, among other works. On Church's place in the history of accounting, see Vangermeersch, Richard, The Contributions of Alexander Hamilton Church (New York, 1986)Google Scholar, and Johnson, H. Thomas and Kaplan, Robert, Relevance Lost: The Rise and Fall of Management Accounting (Boston, Mass., 1987), 52–56Google Scholar.

⁶⁷ Lamoreaux, Great Merger Movement, chaps. 2 and 3, esp. 25–26, 55, 59; Scranton, , Figured Tapestry; Manufacturing Jeweler 12 (1893): 710Google Scholar; 13 (1893): 5, 39, 193, 377–78; 14 (1894): 199.

⁶⁸ For response to the 1893 collapse by staple textile mills, see Scranton, Figured Tapestry, 173–78, and for bulk watch movement makers, Manufacturing Jeweler 13 (1893): 146, 445Google Scholar.

⁶⁹ Smith, Edwin, Reducing Seasonal Unemployment (New York, 1931), 212Google Scholar. See also Kuznets, Seasonal Variations, 313–14.

⁷⁰ Smith, Reducing Seasonal Unemployment, 218–19; Scranton, Figured Tapestry, chap. 6.

⁷¹ Scranton, Figured Tapestry, 30–31; and see note 118 below. There is no mention, however, of similar rental practices at Grand Rapids, as the spatial requirements for case goods apparently were too great.

⁷² Marx, Thomas, “The Diesel-Electric Locomotive Industry” (Ph.D. diss., University of Pennsylvania, 1973)Google Scholar; Reutter, Sparrows Point; Ripley, Romance of a Great Factory, 12–17; Iron Age 137 (25 June 1936): 23–68Google Scholar (on Westinghouse).

⁷³ For an extensive discussion of contract bidding by a locomotive works, see Fry, George Jr., “Bidding in a Job-Lot Iron Works: A Case Study” (MBA thesis, University of Pennsylvania, 1949)Google Scholar.

⁷⁴ Comparato, Frank, Books for the Millions (Harrisburg, Pa., 1971), 123–30, 268–70Google Scholar. Comparato notes that Appleton destroyed 300,000 volumes of warehoused “dead” titles after 1893 rather than remainder them.

⁷⁵ Roe, Joseph, English and American Tool Builders (New York, 1926)Google Scholar; Broehl, Precision Valley; Wing, George, “The History of the Cincinnati Machine Tool Industry” (DBA diss., Indiana University, 1964)Google Scholar; Stewardson, Gary, “The Relationship between Changing Technology and the Training of Machinists and Tool and Die Makers in the U.S. Machine Tool Industry, 1880 to 1980” (Ph.D. diss., University of Maryland, 1987), 47Google Scholar.

⁷⁶ American Machinist 2 (28 June 1879); 8–9Google Scholar, cited in Wing, “Cincinnati Machine Tool Industry,” 93.

⁷⁷ Wing, “Cincinnati Machine Tool Industry,” 94–96. The New England and Mid-Atlantic states' share of sectoral employment fell to 55 percent by 1919 and 47 percent by 1939; Wagoner, U.S. Machine Tool Industry, 359.

⁷⁸ Pratt and Whitney, Accuracy, 39, 41; Wagoner, U.S. Machine Tool Industry, 359, 370; Wing, “Cincinnati Machine Tool Industry,” 140–41; Stewardson, “Changing Technology,” 91–93; Machinery 4 (1897): 101–9Google Scholar; 9 (1902–3): 296–98.

⁷⁹ Wagoner, U.S. Machine Tool Industry, 362; Wing, “Cincinnati Machine Tool Industry,” 217. These production figures are for metal-cutting machines only. In 1914 dollars, the 1918 output reduces to c. $100 million. If all metalworking machinery (including forges and presses) is considered, 1918 output reached $321 million ($145 million in 1914 dollars).

⁸⁰ Wagoner, U.S. Machine Tool Industry, 63, 362; American Machinist 60 (3 Jan. 1924): 1–2Google Scholar. In terms of reduced value added, machine tools stood at the head of thirtythree industries surveyed, 1919 versus 1921, at minus 72 percent, compared with an all-industries average of minus 26 percent; NA, RG 9, “Consequences of Depression: Experiences of Selected Industries in 1921 as Compared with 1919,” box 3498, file 103–22.

⁸¹ Wagoner, U.S. Machine Tool Industry, 119–21; Wing, “Cincinnati Machine Tool Industry,” 230–36. Wing's real output series covers only the local area, but deflation of figures in Wagoner's series (362) suggests a similar national pattern.

⁸² Machinery 28 (1921–1922): 425, 457, 873–74Google Scholar; 29 (1922–23): 357–58; Wagoner, U.S. Machine Tool Industry, 121–29, 133–35, 362. In unit terms, 28,000 tools were shipped in 1927, 50,000 in 1929, and 5,500 in 1932 (136–37). See also Wing, “Cincinnati Machine Tool Industry,” 240–43.

⁸³ Machinery and Allied Products Institute, Capital Goods and the American Enterprise System (Chicago, Ill., 1939), 39Google Scholar.

⁸⁴ Wing, “Cincinnati Machine Tool Industry,” 220–21; Wagoner, U.S. Machine Tool Industry, 363.

⁸⁵ NA, RG 9, Alexander Sachs, “Material Bearing on the Machine Tool Industry: Description of the Industry,” box 3499, file 103–26, 8–10, 16; Wagoner, U.S. Machine Tool Industry, 25; Stewardson, “Changing Technology,” 58–60.

⁸⁶ Sachs, “Material Bearing on the Machine Tool Industry,” 3–4.

⁸⁷ Ibid., 2.

⁸⁸ Wing, “Cincinnati Machine Tool Industry,” 83–87, 185–90; Harris, Howell, “Getting It Together”; Engineering Magazine 17 (1899): 21–22Google Scholar (Orcutt). Lodge had been a journeyman machinist and president of the Mechanics' Union in the 1870s. He started in business in 1880 and his “rapid success encouraged many of his old friends” to follow his lead. Machinery 22 (1915–1916): 375Google Scholar. See also Stewardson, “Changing Technology,” 91–93.

⁸⁹ Wing, “Cincinnati Machine Tool Industry,” 149–50.

⁹⁰ Ibid., 128.

⁹¹ For a full description of the system installed at Oberlin Smith's Ferracute Machine Company, near Philadelphia, see Parkhurst, Frederick, Applied Methods of Scientific Management (New York, 1912), esp. 126–29, 143–46Google Scholar.

⁹² Machinery 17 (1911–1912): 394, 765Google Scholar. The quotes are from remarks by Ralph Flanders of Vermont, later NMTBA president. See also Stewardson, “Changing Technology,” 62, and for an overview of particular tools, Woodbury, Robert, Studies in the History of Machine Tools (Cambridge, Mass., 1972)Google Scholar. Cincinnati Milling Machine was the most dedicated “unit system” advocate.

⁹⁸ Engineering Magazine 21 (1901); 586–90Google Scholar; Harris, “Getting It Together”; Wing, “Cincinnati Machine Tool Industry,” 179–80. 230–37; NA, Records of the Federal Mediation and Conciliation Service, RG 280, files 33/106 (Brown and Sharpe, 1915), 33/119, 33/529, 33/985, 170/804 (all Cincinnati, 1915–20). In Philadelphia, Cincinnati, and Providence, metal trade groups sponsored labor bureaus that served to organize labor markets and to screen out unionists and strike veterans. Wing shows that war decade Cincinnati strikes, all failures, were followed by wage increases that brought local rates to eastern levels, eliminating a profitable, forty-year differential.

⁹⁴ Machinery 2 (1895–1896): 15Google Scholar; 3 (1896–97): 189; Engineering Magazine 12 (1896–1897): 833Google Scholar; 18 (1899–1900): 572–84, 689–96 (Norris). For later accounts of pay system tangles, see Engineering Magazine 45 (1913–1914): 566–70Google Scholar and sources in note 40.

⁹⁵ Machine tool employment in 1931 was 55 percent of the 1929 level, versus an all-industries average of 74 percent; “A Comparison of Selected Industries for the Years 1929 and 1931,” NA, RG 9, box 3498, file 103–22. See also T. D. Doan to Herman Lind, NMTBA, 4 Jan. 1934, in NA, RG 9, box 3497, file 103–9, and F. Selfert to W. H. Dillingham, National Research Project, 25 May 1936, in RG 9, box 3499.

⁹⁶ Machinery 4 (1897–1898): 101–9Google Scholar; 15 (1908–9): 299–301; 16 (1909–10): 223–24, 782–83.

⁹⁷ Machinery 7 (1900–1901): 127–28Google Scholar; 17 (1910–11): 782, 872; 18 (1911–12): 265–67, 345–46, 755; 23 (1916–17): 677–78 (Heald).

⁹⁸ Wagoner, U.S. Machine Tool Industry, 93–97, 149; Sachs, “Material Bearing on the Machine Tool Industry: Related Industries,” 2–3; Machinery 28 (1921–1922): 452–53, 873–74Google Scholar; 29 (1922–23): 357–58; 30 (1923–24): 853–54; 32 (1925–26): 543–14.

⁹⁹ Parkhurst, Applied Methods, 22.

¹⁰⁰ “Reports on 59 Machine Tool Builders, showing the influence of depressions,” tables I and II, NA, RG 9, box 3498, file 103–22.

¹⁰¹ Sachs, “Materials Bearing on the Machine Tool Industry,” 13, table II. That leading firms like Brown and Sharpe were “close corporations” with nominal stock issues (for B&S, $100,000, supporting employment of 4,500 workers in the 1930s) surely added to this flexibility. See also A. L. Faulkner to William Wilson, Secretary of Labor, 26 Oct. 1915, NA, RG 280, file 33/106.

¹⁰² Machinery 31 (1924–1925): 359, 367Google Scholar; 32 (1925–26): 24, 238; 33 (1926–27): 138, 331–32; Wagoner, U.S. Machine Tool Industry, 120, 151.

¹⁰³ Of the 139 active machine tool firms whose dates of founding were recorded in a 1933 report, the pattern of New starts was as follows:

¹⁰⁴ Landes, David, Revolution in Time (Cambridge, Mass., 1985)Google Scholar.

¹⁰⁵ Carsten, Oliver, “Work and the Lodge: Working Class Sociability in Meriden and New Britain, Connecticut, 1850–1940” (Ph.D. diss., University of Michigan, 1981), 78–81, 87–92Google Scholar. See also U.S. Tariff Commission, Silverware (Washington, D.C., 1927)Google Scholar.

¹⁰⁶ Manufacturing Jeweler 1 (1884–1885): 335, 397Google Scholar; 2 (1885–86): 169–70; 9 (1891): 268; 14 (1894): 12, 14, 16, 324; Frankovich, George, The Jewelry Industry (Cambridge, Mass., 1955), 9–10Google Scholar; New Jersey Commerce and Finance 2 (1906): 145Google Scholar; Parsons, Floyd, ed., New Jersey: Life, Industries, and Resources (Newark, N.J., 1928), 289–91Google Scholar.

¹⁰⁷ Bovin, Peter, “A Study of the Contemporary Precious Jewelry Industry in New York City” (E. D. diss., New York University, 1975), 61Google Scholar; Manufacturing Jeweler 8 (1890–1891): 318Google Scholar; 10 (1892): 284; 14 (1894): 414.

¹⁰⁸ Manufacturing Jeweler 1 (1184–85): 173, 218; 2 (1885–86): 215; 3 (1886–87): 476; 29 (1901): 522.

¹⁰⁹ Ibid. 4 (1887–88): 192–96; 13 (1893): 103, 391; 26 (1900): 56; Shapiro-Perl, “Labor Process,” 52–54. Such household contractors could not afford shop rents, as in apparel outwork in New York or Philadelphia, and used family living spaces to set up their manufacturing efforts. For homework in clothing manufacturing at Philadelphia, and the union battle to eradicate it, see LaMar, Elden, The Clothing Workers in Philadelphia (Philadelphia, Pa., 1940), 120–32Google Scholar.

¹¹⁰ Manufacturing Jeweler 24 (1899): 171Google Scholar; 26 (1900): 56; 28 (1901): 40; 29 (1901): 522.

¹¹¹ Ibid. 24 (1899): 241, 261–62, 266, 284, 317, 360. An earlier effort to form a jobbers' trust proved equally futile; ibid. 7 (1890): 52.

¹¹² By 1909, Rhode Island and Massachusetts held 471 firms and 17,000 workers, to which New York and New Jersey added 649 shops with 12,000 employees; Manufacturing Jeweler 49 (1911): 214Google Scholar; 50 (1912): 1382. For discussion of demoralization, see ibid. 50 (1912): 1064, 1154; 80 (9 June 1927): 16, 18; and for the sellers' market, ibid. 65 (1919): 374, 402, 546. Wartime attrition seems to have reduced the number of firms by half; ibid. 66 (1920): 1228, but new starts, both the fruit and the source of trade miseries, surged during the 1920s to reach a 1929 total of 1,536 firms. Sachs, Alexander, “Material Bearing on the Precious Jewelry Producing Industry,” 1, NA, RG 9, box 4409. Peak 1919Google Scholar employment was 31,000, but fell below 24,000 by 1925 (ibid., 7). For earlier complaints about new firms' practices, see Manufacturing Jeweler 15 (1894): 3Google Scholar; 29 (1901): 155.

¹¹³ Massachusetts Institute of Technology, Industrial Survey of Metropolitan Providence for 1926 (Providence, R.I., 1928), 4Google Scholar. For Newark troubles, see Journal of Finance and Industry 3 (1928–1929): 58Google Scholar.

¹¹⁴ A gold surface mechanically plated around a core of base metal.

¹¹⁵ Frankovich, , Jewelry; Machinery 17 (1910–1911): 181–84Google Scholar; 18 (1911–12): 255–57; Manufacturing Jeweler 24 (1899): 15Google Scholar (on Ferracute's presses for jewelers' use); American Machinist 49 (1918): 47–52Google Scholar.

¹¹⁶ Manufacturing Jeweler 3 (1886–1887): 210Google Scholar; 12 (1893): 395–96; 13 (1893): 711; 24 (1899): 362; 25 (1899): 486; 19 Oct., anniversary no., separately paginated [hereafter, AN]: 28–34; 29 (7 Nov. 1901, AN): 32; 56 (1915); 647; Metal Industry 13 (1915): 125–26, 336, 425–26Google Scholar; Modern Plastics 13 (Aug. 1936): 9–11, 54–56Google Scholar; 17 (July 1940): 32–35, 82. By 1934 three-quarters of costume jewelry was being made from plastics; Modern Plastics 17 (July 1940): 32Google Scholar. See also “Hearing on… Medium and Low Priced Jewelry,” 13 Oct. 1933, 447, 456–62, NA, RG 9, box 7247.

¹¹⁷ Sachs, “Materials Bearing on Precious Jewelry Industry,” 6a.

¹¹⁸ Manufacturing Jeweler 4 (1887–1888): 234, 578–84Google Scholar; 5 (1889): 140; 9 (1891): 306–22; 10 (1892): 284; 11 (1892): 871–72; 12 (1893): 592–600; 25 (1899): 28; 29 (1901): 91; (7 Nov., AN): 42–46; 49 (1911): 304; Commonwealth of Massachusetts, 45th Annual Report on the Statistics of Labor (Boston, Mass., 1914), 90–109Google Scholar; U.S. Senate, 61st Cong., 2d sess., Report on Condition of Women and Child Wage-Earners in the United States (Washington, D.C., 1913), 18: 213–19Google Scholar, doc. no. 645; Children's Bureau, Department of Labor, Industrial Home Work of Children (Washington, D.C., 1922), 28–40, 58–72Google Scholar; B. Leinwand to General Hugh Johnson, 24 Feb. 1934, NA, RG 9, box 4414, file 130–24 (on continuity of bench rental).

¹¹⁹ Manufacturing Jeweler 57 (1915): 1126Google Scholar. By “card,” Makinson referred to the pasteboard backings on which cheaper jewelry was mounted for display in retail shops. This work was then done by apprentices and later by women.

¹²⁰ Ibid. 2: (1885–86): 170, 215; 25 (1899): 277; 29 (1901): 380.

¹²¹ Ibid. 5 (1888–89): 307; 9 (1891): 803–4; 11 (1892): 253; 24 (1899): 59, 453–54; 25 (1899): 183–84, 260; 28 (1901): 590; 55 (1914): 492–94; 58 (1916): 940; 60 (1917): 201–2; 64 (1919): 144–46; 70 (1922): 269.

¹²² Ibid. 25 (1899): 210, 230, 243, 260, 264, 277, 284, 388, 420, 429; 31 (1902): 334, 380, 447, 512, 768; 32 (1903): 378, 386; 34 (1903): 128, 163, 228, 239–40; 49 (1911): 214; 65 (1919): 84, 128, 354, 590, 638, 686, 1026, 1078; 66 (1920): 38–39, 548; 67 (1920): 1024. After intensive 1902 organizing efforts, the 1903 strike plans aborted. The Inter national Jewelers continued to have little success, especially in the eastern district, into the 1940s. See IJU, Report of the Proceedings of the Sixth Convention (Chicago, Ill., 1926)Google Scholar, and Report of the Ninth Convention (New York, 1942), 96, 258Google Scholar.

¹²³ IJU, Monthly Bulletin, quoted in Shapiro-Perl, “Labor Process,” 69. Skilled jewelers were paid by the hour or week, with piecework reserved for assembly, outwork, and related tasks usually assigned to women. As late as 1928, a New England survey found half the jewelry firms having “no employees on a piece work basis” and only 21 percent of all workers compensated by output; Manufacturing Jeweler 82 (2 Feb. 1928): 29Google Scholar. This pattern may help to account for the union's weak showing, as craft organizations used the advance of piecework and rate manipulations as a key argument for solidarity (for example, among machinists). The absence of this issue, the ease with which workers could become petty proprietors, and the large presence of women workers rarely considered worth organizing combined to frustrate unionizers' efforts.

¹²⁴ Manufacturing Jeweler 1 (1884–1885): 7, 9, 19, 404Google Scholar.

¹²⁵ Manufacturers advocated selling direct by 1884 and actively practiced it by 1888; ibid.: 9; 10 (1892): 283. On scheme and prize goods, see ibid. 3 (1886–87): 210, and for jobbers' boycotts, ibid. 5 (1888–89): 581, 587; 10 (1892): 283. For improved trade terms, see ibid. 6 (1890): 116–17, 197, 216.

¹²⁶ Ibid. 3 (1886–87): 84, 174; 5 (1888–89): 193, 531; 9 (1891): 1207, 1327, 1363; 11 (1892): 417.

¹²⁷ Ibid. 49 (1911): 329; 50 (1912): 334–36, 1064; 53 (1913): 1048–49; 55 (1914): 1006, 1142; 55 (1914): 838, 924; 57 (28 Oct. 1914, AN): 46–51; 58 (1916): 1180a; 70 (1922): 922, 934; 71 (1922): 1073; 76 (18 June 1925): 20–22; (17 Sept. 1925): 20; (5 Nov. 1925): 18–20; 80 (7 April 1927): 22; (9 June 1927): 11, 18.

¹²⁸ Schurz, William, “Affadavit” (Washington, D.C., 1935), 17Google Scholar, NA, RG 9, box 4409. Given the sample and duplicate form of jobbers' ordering, middlemen had retailers' confirmed orders in hand when they returned to arrange terms and prices. The ease with which styles could be copied by “mushroom” firms only added to the pressure for manufacturers' concessions. In real terms (1929 dollars), the earnings drop was 60 percent, with real product value down 64 percent; Schurz, “Affadavit,” 13, adjusted using Series E–14, Historical Statistics of the United States. In Philadelphia textiles, real earnings, 1929–33, fell 37 percent and real product 43 percent; Scranton, Figured Tapestry, tables 6.9 and 6.10. See also Manufacturing Jeweler 82 (29 March 1928): 5Google Scholar; (1 Nov. 1928): 23. In the jobbers' defense, the retailers they served were among the worst merchandisers in the nation, averaging 1.5 stock turns annually (vs. six for department stores), and jobbers' bad debt losses were double those of wholesalers in other fields; ibid. 54 (1914): 939–40; 82 (7 June 1928): 13.

¹²⁹ Manufacturing Jeweler 6 (1889–1889): 222Google Scholar; 32 (1903): 269–70; 51 (1912): 217–18, 552, 1270, 1450; (24 Oct., AN): 68–70; 52 (1913): 378–82; 54 (1914): 170, 1193; Schurz, “Affadavit,” 17.

¹³⁰ Manufacturing Jeweler 2 (1885–1886): 170Google Scholar; 15 (1894): 494–96; 25 (1899): 429; (19 Oct., AN): 3, 28–34; 82 (2 Feb. 1928): 29; 84 (2 May 1929): 14; Bureau of Foreign and Domestic Commerce, Jewelry Distribution (Washington, D.C., 1931), 26–33Google Scholar. A 1917 profile of 108 Attleboro and Providence firms showed the following pattern of founding dates:

¹³¹ Batch specialists operating in isolated locations were constrained in layoff policies by the absence of alternative work opportunities for dismissed employees, the fear of their relocation, and the difficulties and costs of new recruitment when demand rose. In urban districts, laid-off workers could take their skills to other companies in the same or related sectors, in some cases shifting among firms with complementary seasonalities (as in textiles and radio at Philadelphia, c. 1920–40). The larger labor pool made rehiring less a problem, whereas aspiring entrepreneurs in the work force could, through job shifting, accumulate experience in varied trade branches. However, in cities where industry was dominated by a single sector or a few large corporations, these possibilities were less marked.

¹³² “Lining” is the practice of gapped pricing by retailers (the $15, $22.50, and $30 men's suit), which forced suppliers to make goods for ranges of retail selling prices set in advance and downstream from factories. In the 1920s, Woolworth's (1,100 stores) spent $4 million yearly for five- and ten-cent goods, commanding nearly 10 percent of the cheap jewelry market. Department store and jobbing buyers sought $1, $3, and $5 lines, resisting makers' efforts to raise prices; Manufacturing Jeweler 68 (1921): 64Google Scholar; 86 (27 March 1930): 24.

¹³³ The 1930s surge of plastics in cheap jewelry devalued metalworking machinery and skills, while bringing back casting, in its injection molding variant; Modern Plastics 13 (Aug. 1936): 55–56Google Scholar; 14 (May 1937): 70–72.

¹³⁴ Makers of school, college, and fraternity jewelry were a partial exception. Josten's, Balfour, and others contracted directly with institutions and sent salesmen on rounds to test-market designs and gather orders. This subsector alone remained a solid performer into the Depression era. “Hearing on Medium and Low Priced Jewelry Industry,” 82–124.

¹³⁵ Chandler, Scale and Scope, 605–28.

¹³⁶ For one effort at reassessment, see Best, Michael, The New Competition (Cambridge, Mass., 1990)Google Scholar, esp. chaps. 2 and 5.

¹³⁷ What I have termed “bridge” firms in electrical manufacturing are an exception, and all the more fascinating for that. Similarly, a number of European “mixed format” giants (Siemens, I.G. Farben) drew creatively on a variety of production forms to generate enormous market power and technological momentum.

¹³⁸ For seminal discussions, see Hodgson, Geoffrey, Economics and Institutions (Philadelphia, Pa., 1988Google Scholar) and Douglas, Mary and Isherwood, Baron, The World of Goods (New York, 1979)Google Scholar.

¹³⁹ Williamson, Oliver, The Economic Institutions of Capitalism (New York, 1985)Google Scholar.

¹⁴⁰ Granovetter, Mark, “Economic Action and Social Structure: The Problem of Embeddedness,” American Journal of Sociology 91 (1985): 481–510CrossRefGoogle Scholar.