¹ For more detail on these and some statistics on uptake, see Emmet, Boris, Profit-Sharing in the United States, U.S. Bureau of Labor Statistics Bulletin No. 208 (Washington, D.C., 1916), pp. 97–98, 107.Google Scholar For how the terms should be understood in the context of the argument I lay out below, see my Wage Determination and the Five-Dollar Day a: Ford: A Detailed Examination (Harvard University, unpublished monograph, 1987), pp. 163–94;Google Scholar or (for the same analytical points in comparative perspective) my “Ford Welfare Capitalism in Its Context” (Harvard University, unpublished manuscript, 1987) in preparation for S. Jacoby, ed., Managing the Workplace: A New Industrial History (forthcoming).

² In American system production, single-purpose machine tools were used to make parts to very high tolerances. (Thus the phrase “interchangeable parts.”) Fitting these together to assemble the product did not require a mechanic's training and skill. For the implications of that, read on.Google Scholar

³ The Detroit automotive industry generally had high turnover rates, but the numbers at Ford were half again and twice the common ones. See Slichter, Sumner, The Turnover of Factory Labor (New York, 1921), pp. 33–34.Google Scholar

⁴ The subject of the study (J. C. Mead, “Salvage of Men,” mimeograph in Ford Archives) was the ability of the Ford Motor Company to make use of partly or severely handicapped individuals in production jobs. The sorts of handicaps under consideration ranged from one or more amputated or hopelessly crippled limbs, partial blindness and the like, to deafness and hernia. The author first discovered that some of these people could in fact do heavy work the company needed done. Others could not do light work the company required. So jobs had to be surveyed one by one. Though the period was slightly later, the production process studied was for practical purposes the same as the one in place in 1914. I should add that no information in the study is inconsistent with the notion that the jobs surveyed constitute a rough random sample of the jobs there were.Google Scholar

⁵ Emmet, Profit-Sharing in the United States, p. 95, gives a table indicating that as of the preceding October about 2 percent of the jobs were clearly skilled, 72 percent were clearly unskilled, and 26 percent were of ambiguous content. Hounshell, David, From the American System to Mass Production: The Development of Manufacturing Technology in the United States (Baltimore, 1984), pp. 247, 374, observes that the percentages of skilled to unskilled in the auto industry elsewhere at the time were roughly 83 and 17 respectively.Google Scholar

⁶ Arnold, Horace and Faurote, Fay L., Ford Methods and Ford Shops (New York, 1915), p. 41.Google Scholar

⁷ Nevins, Allan, Ford: The Times, The Man, The Company (New York, 1954), pp. 377, 379.Google Scholar

⁸ Information in Emmet, Profit-Sharing in the United Stares, p. 95, makes plain that most mechanics and sub-foremen earned in the range of $4.32 to $5.85 per day.Google Scholar

⁹ Their raise was proportionately smaller than that received by the less skilled. This is consistent with the explanation I lay out below.Google Scholar

¹⁰ These numbers are slight underestimates, since they do not allow time for getting the new employee through the employment office and a medical examination to the job site. The company estimated the time involved in this as slightly over two hours. Plausible guesses at the salaries of all the people involved lead to trivial changes in the savings estimates in the text.Google Scholar

¹¹ Slichter, The Turnover of Factory Labor, p. 32, summarizes research he conducted for the Commission on Industrial Relations. The memo he cites does not seem to have survived among his personal papers or either of the two Commission hoards.Google Scholar

¹² It must be said that this calculation may be a little facile: it casually assumes that each worker was trained only once. With as many as 10.5 percent absent each day—in late 1913, this would have been upwards of 1,300 men—and a convention that the absent had to miss five days running before they were deemed to have quit, it is clear that at least some workers learned more than one task. It is not clear from the surviving records what the most accurate way of modeling this excess training would be, and Mead's later survey, “Salvage of Men,” did report 7000-odd different jobs. I strongly suspect that a great many of these jobs differed chiefly in location in the plant, the type of single-purpose machine tool to be tended, or the diameter of the bolt in question and would not pose any serious challenge to many who had been trained for some other job in the plant. My reading suggests that the training costs will have been mainly first-time costs, as much about the discipline of the line as anything else. All things considered, this seems unlikely to be an important qualification. Reducing turnover costs remains an unlikely central motive. Note that I have been treating turnover as problem rather than symptom. For more on the latter, see below.Google Scholar

¹³ See Yellin, Janet, “Efficiency Wage Models of Unemployment,” American Economic Review, 74 (05 1984), pp. 200–5.Google Scholar

¹⁴ Meyer, Stephen, The Five-Dollar Day: Labor Management and Social Control in the Ford Motor Company 1908–1921 (Albany, 1981), p. 41.Google Scholar

¹⁵ Heliker, George, “Labor in Detroit 1910–1916” (unpublished typescript, Ford Archives, undated), p. 85.Google Scholar

¹⁶ Martin, I. T., “The Melting Pot at Ford's: Conditions of Employees under the New Profit-Sharing System,” Official Journal of the Carriage, Wagon, and Automobile Workers Union, 4 (08 1915), p. 8.Google Scholar

¹⁷ Perhaps the division of labor was radical enough that all such exercises of discretion were obvious to the supervisor.Google Scholar

¹⁸ The figures are derived from Arnold and Faurote, Ford Methods and Ford Shops, p. 46; “Model T Production Statistics,” Ford Archives Accession 125; and “List of Trades and Occupations,” Ford Archives Accession 940, box 16. A finer time series would obviously be desirable. But there are no data in the Ford Archives from which to construct one, and the State of Michigan's Department of Labor and Industry, which probably did collect them, lost all its manuscript records from this period in a fire in 1951. No copies survive.Google Scholar

¹⁹ Matthewson, Stanley, Restriction of Output Among Unorganized Workers (New York, 1929).Google Scholar

²⁰ See Dickens, william T. and Katz, Lawrence F., “Inter-Industry wage Differences and Theories of Wage Determination” (unpublished manuscript, University of California, Berkeley, 1986);Google Scholar and Krueger, Alan and Summers, Lawrence H., “Efficiency Wages and the Inter-Industry Wage Structure” (unpublished manuscript, Harvard University, 1986).Google Scholar See also Dickens, , “Wages, Employment, and the Threat of Collective Action by Workers” (unpublished manuscript, University of California, Berkeley, 1986).Google Scholar

²¹ Dickens, “Wages, Employment, and the Threat of Collective Action by Workers,” p. 3.Google Scholar

²² For interesting research on the history of Detroit employers' efforts in this line, see Thomas Klug on the Employers Association of Detroit (Wayne State University, dissertation in progress).Google Scholar

²³ The most famous example is a machine tool that drilled 45 holes in 4 sides of the engine block. The locations and sizes of the holes were exactly what the Model T design required. The machine design was such that they could not be modified or adjusted. See Hounshell, From the American System to Mass Production. p. 233.Google Scholar

²⁴ Galamb, Joseph, “Reminiscences” (unpublished typescript, Ford Archives, 1952), p. 148.Google Scholar It is noteworthy that in a city with correspondents for Reuters, the national wire services, and major out-of-town papers, only reporters from the three local papers were invited to the press conference. The publicity was directed at workers inside, not outside, the factory. (See also Nimmo, H. M., “Talking It Over with Henry Ford,” Detroit Saturday Night, 01 10, 1914, pp. 9–10.)Google Scholar

²⁵ Sorensen, Charles, My Forty Years with Ford (New York, 1956), pp. 137–38.Google Scholar

²⁶ Interview with Perry, Percival, Ford Archives Accession 834, pp. 3–5; and “The English Works of the Ford Motor Company: Notes on the More Interesting Processes,” The Automobile Engineer (July 1915), pp. 188ff.Google Scholar

²⁷ Letter from Percival Perry to Henry Ford dated February 26, 1913 in Ford Archives Accession 62, box 59.Google Scholar

²⁸ Vera Howard interview in “FMC: 1908–1914” file, Ford Archives Accession 581, box 5.Google Scholar

²⁹ On the Sorensen presence, see the Perry interview, pp. 4–5.Google Scholar

³⁰ Carbon copy of letterfrom Sorensen to H. S. Jenkins dated 9 September 1926 in Ford Archives Accession 38, box 52.Google Scholar

³¹ So much was this true that when, slightly later in the 1920s, the Manchester managing director sought permission from Dearborn to cut wages, he seems to have felt he ought to begin his arguments with a ritual obeisance. The key passage begins “High wages are a sound investment on our part, but …” Letter from Perry to Edsel Ford of 7 August 1928 in Ford Archives Select File.Google Scholar

³² The first references to widespread rising wages in this sector in Detroit that I have found are contemporaneous with the arrival of orders connected with World War I. This seems unrelated.Google Scholar

³³ For more on which see my “Efficiency Wage Theory, the Inter- and Intra-Industry Structure of Earnings, and Evolution in Production Processes” (unpublished manuscript, Harvard University, 1987).Google Scholar

³⁴ For Ford, contrasts, and some thought-provoking comparisons, see my “Ford Welfare Capitalism.”Google Scholar