Hostname: page-component-78c5997874-4rdpn Total loading time: 0 Render date: 2024-11-18T12:27:12.081Z Has data issue: false hasContentIssue false
  • English
  • Français

Labor Mobility and Lengthy Jobs in Nineteenth-Century America

Published online by Cambridge University Press:  03 March 2009

Susan B. Carter  and
Elizabeth Savoca
Elizabeth Savoca
Affiliation:
Associate Professor of Economics, Smith College, Northampton, MA 01063. Assistant Professor of Economics, Smith College, Northampton, MA 01063 and Fellow, Institution for Social and Policy Studies, Yale University, New Haven, CT 06520.
Get access Rights & Permissions [Opens in a new window]

Abstract

Extensive amounts of geographic mobility and high rates of labor turnover before World War I gave rise to the notion that the industrial labor force was “casual” and “impermanent.” But data from firms' payrolls and from nineteenth-century surveys conducted by state labor statistics bureaus show that male workers averaged about four years of experience in their current jobs. Data from an 1892 survey of San Francisco workers show that the average non-union male could expect to remain with his current employer almost 13 years.

Type
Articles
Information
The Journal of Economic History , Volume 50 , Issue 1 , March 1990 , pp. 1 - 16
Copyright
Copyright © The Economic History Association 1990

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 The quotation is from Hall, Robert E., “The Importance of Lifetime Jobs in the U.S. Economy,” American Economic Review, 72 (09 1982), p. 716.Google Scholar On dating the transition to stable, lengthy employment spells, see Ross, Arthur M., “Do We Have a New Industrial Feudalism?” American Economic Review, 48 (12 1958), pp. 903–20Google Scholar; Sachs, Jeffrey, “The Changing Cyclical Behavior of Wages and Prices: 1890–1976,” American Economic Review, 70 (03 1980), pp. 78–90Google Scholar; Jacoby, Sanford M., Employing Bureaucracy: Managers, Unions, and the Transformation of Work in American Industry, 1900–1945 (New York, 1985)Google Scholar; and Long, J. Bradford De and Summers, Lawrence H., “The Changing Cyclical Variability of Economic Activity in the United States,” in Gordon, Robert J., ed., The American Business Cycle: Continuity and Change (Chicago, 1986), pp. 679–734.Google Scholar

2 See Nelson, Daniel, Managers and Workers: Origins of the New Factory System in the United States, 1880–1920 (Madison, 1975)Google Scholar; Jacoby, Employing BureaucracyGoogle Scholar; Keyssar, Alexander, Out of Work: The First Century of Unemployment in Massachusetts (Cambridge, MA, 1986)Google Scholar; Gutman, Herbert G., Work, Culture and Society in Industrializing America (New York, 1977)Google Scholar; and Monkkonen, Eric H., ed., Walking to Work (Lincoln, NE, 1984).Google Scholar

3 Thernstrom, Stephan, Poverty and Progress: Social Mobility in a Nineteenth Century City (Cambridge, MA, 1975), p. 85.Google Scholar

4 Easterlin, Richard, “Regional Growth of Income: Long Term Tendencies,” in Population Redistribution and Economic Growth: United States, 1870–1950, vol. 2 (Philadelphia, 1960), pp. 141–204.Google Scholar For other evidence of labor market integration, see Lebergott, Stanley, Manpower in Economic Growth (New York, 1964)Google Scholar; Margo, Robert A. and Villaflor, Georgia C., “The Growth of Wages in Antebellum America: New Evidence,” this Journal, 47 (12 1987), pp. 873–95Google Scholar; Rothenberg, Winifred B., “The Emergence of Farm Labor Markets and the Transformation of the Rural Economy: Massachusetts, 1750–1855,” this Journal, 48 (09 1988), pp. 537–66Google Scholar; and Rosenbloom, Joshua L., “Labor Market Institutions and the Geographic Integration of Labor Markets in the Late Nineteenth Century United States” (Ph.D. diss., Stanford University, 1988).Google Scholar

5 See Williamson, Jeffrey G., “Did British Labor Markets Fail During the Industrial Revolution?” (Harvard Institute for Economic Research, Discussion Paper No. 1209, 02 1986), for a discussion of the literature on labor market failure.Google Scholar

6 See Wright, Gavin, Old South, New South: Revolutions in the Southern Economy Since the Civil War (New York, 1986), on the failure of the southern labor market.Google Scholar See Field, Alexander James, “Sectoral Shift in Antebellum Massachusetts: A Reconsideration,” Explorations in Economic History, 15 (04 1978), pp. 146–71, on the female labor market in the antebellum Northeast.CrossRefGoogle Scholar

7 Jacoby, Employing Bureaucracy, pp. 36–37.Google Scholar

8 Ross, “New Industrial Feudalism,” p. 911.Google Scholar

9 Jacoby, Employing Bureaucracy, provides a good summary of this evidence.Google ScholarCarter, Susan B., “The Changing Importance of Lifetime Jobs, 1892–1978,” Industrial Relations, 27 (Fall 1988), pp. 287–300, shows that lengthy employment spells were more common in the 1890s than case study evidence seems to indicate.Google ScholarCarter, Michael J. and Carter, Susan B., “Internal Labor Markets in Retailing: The Early Years,” Industrial and Labor Relations Review, 38 (07 1985), pp. 586–98CrossRefGoogle Scholar; and Sundstrom, William A., “Internal Labor Markets before World War 1: On-the-Job Training and Employee Promotion,” Explorations in Economic History, 25 (10 1988), pp. 424–45, present evidence of internal labor markets prior to the 2920s.CrossRefGoogle Scholar

10 Ware, Caroline, The Early New England Cotton Manufacture: A Study of Industrial Beginnings (Boston, 1931), p. 224.Google Scholar

11 Sachs, “The Changing Cyclical Behavior of Wages and Prices,” pp. 78–90Google Scholar; Gordon, Robert J., “A Century of Evidence on Wage and Price Stickiness in the United States, the United Kingdom, and Japan,” in Tobin, James, ed., Macroeconomics, Prices, and Quantities (Washington, DC, 1983), pp. 85–120Google Scholar; and De Long and Summers, “Changing Cyclical Variability,” pp. 679–734. For a model of the effect of community standards of “fairness” on wage flexibilityGoogle Scholar, see Akerlof, George A., “A Theory of Social Custom, of Which Unemployment May Be One Consequence,” The Quarterly Journal of Economics, 94 (06 1980), pp. 749–75.CrossRefGoogle Scholar

12 Clark, Kim and Summers, Lawrence, “Labor Market Dynamics and Unemployment: A Reconsideration,” Brookings Papers on Economic Activity, 1 (1979), pp. 13–60CrossRefGoogle Scholar; and Hall, “The Importance of Lifetime Jobs,” p. 716.Google Scholar See also Akerlof, George A. and Main, Brian G. M., “An Experience-Weighted Measure of Employment and Unemployment Durations,” American Economic Review, 71 (12 1981), p. 1003.Google Scholar

13 Slichter, Sumner, The Turnover of Factory Labor (New York, 1919), p. 43.Google Scholar

14 Wright, Old South, New South, p. 140.Google Scholar

15 The estimate of average length of residence of 10 years was computed by assuming that the average residence of the 65 percent of unskilled laborers who had left Newburyport by the end of the decade was five years; assuming that the remaining 35 percent of laborers would leave in the following decade after an average residence of 15 years and so forth. This exercise implies that the average resident stayed 10.38 years.Google Scholar

16 A number of recent investigations suggest that Thernstrom's estimates overstate the extent of geographic mobility. Donald Parkerson concludes that “the record-linkage process itself has been responsible for a significant underestimation of nineteenth-century persistence.”Google Scholar See Parkerson, Donald H., “How Mobile Were Nineteenth-Century Americans?” Historical Methods, 15 (Summer 1982), p. 105.CrossRefGoogle ScholarKousser, Morgan, Cox, Gary, and Galenson, David find, not a “floating proletariat,” but rather a “youthful mobility, comparatively settled middle age, and the accumulation of human and physical capital over the lifecycle.”Google Scholar See Kousser, J. Morgan, Cox, Gary W., and Galenson, David W., “Log-Linear Analysis of Contingency Tables,” Historical Methods, 15 (Fall 1982), p. 164.CrossRefGoogle Scholar See also Galenson, David W. and Levy, Daniel S., “A Note on Biases in the Measurement of Geographic Persistence Rates,” Historical Methods, 19 (Fall 1986), pp. 171–79CrossRefGoogle Scholar; and Guest, Avery M., “Notes from the National Panel Study: Linkage and Migration in the Late Nineteenth Century,” Historical Methods, 20 (Spring 1987), p. 76.CrossRefGoogle Scholar

17 For discussions of turnover and careers of nineteenth-century female school teachers, see Elsbree, Willard S., The American Teacher: Evolution of a Profession in a Democracy (New York, 1939), pp. 293–95Google Scholar; Bernard, Richard M. and Vinovskis, Maris A., “The Female School Teacher in Antebellum America,” Journal of Social History, 10 (03 1977), pp. 332–45CrossRefGoogle Scholar; and Carter, Susan B. and Savoca, Elizabeth, “Teacher Turnover, Teacher Tenure: Tales Out of Iowa School Data, 1884” (paper presented at the Social Science History Association Meetings, 11 1989).Google Scholar

18 The 18-year estimate for the modern era is from Akerlof and Main, “An Experience-Weighted Measure.” It is an average across all sectors.Google Scholar

19 California Bureau of Labor Statistics, Fifth Biennial Report (Sacramento, 1893). The report contains no explicit discussion of how the data were collected.Google Scholar

20 Issel, William and Cherney, Robert W., San Francisco, 1865–1932: Politics, Power and Urban Development (Berkeley, 1986), p. 54.Google Scholar

21 See Monkkonen, , ed., Walking to WorkGoogle Scholar; and Goldin, Claudia, “Monitoring Costs and Occupational Segregation by Gender,” Journal of Labor Economics, 1 (01 1986), pp. 1–27. The exclusion of union males only marginally affects the representativeness of our sample since unionized males comprised less than 5 percent of the labor force.CrossRefGoogle Scholar See Friedman, Gerald, “Politics and Union Growth: Unions and the Labor Movement in France and the United States” (Ph.D. diss., Harvard University, 1985). Analysis of women's job tenure awaits further research.Google Scholar

22 Workers in this era experienced substantial periods of seasonal unemployment. We do not know if they deducted these from the length of job tenure to date which they report. In any case the point is irrelevant, as our interest is in the elapsed time since the initiation of the employment relation. Some of the other variables in Table 2 also require discussion. What we label SCHOOL is computed in the standard way, subtracting from “Age began Work” the number six to reflect the typical school entry age. We excluded workers for which this procedure yielded implausibly high levels of schooling, in the range of 15 to 20 years.Google Scholar

23 Carter, Susan B. et al. , “Codebook and User's Manual: A Survey of 3,493 Workers in California, 1892” (Historical Labor Statistics Project, University of California, 1989).Google Scholar

24 While these data allow us to estimate job duration, they do not give us any insight into why jobs come to an end when they do. We suspect that most jobs in our sample were ended by a voluntary quit since, with the exception of trough years of business cycles, voluntary quits were the proximate cause of most job separations in the turnover data collected after 1909. See Slichter, The Turnover of Factory Labor, p. 85Google Scholar; Brissenden, Paul and Frankel, Emil, Labor Turnover in Industry: A Statistical Analysis (New York, 1922), p. 79Google Scholar; and Woytinsky, W. S., Three Aspects of Labor Dynamics (Washington, DC, 1942), p. 2. The California survey was taken at the peak of a business cycle. Wesley Mitchell estimated that Jan. 1892 marked the eighth month of an expansion that peaked in Jan. 1893.Google Scholar See Mitchell, Wesley, What Happens During Business Cycles? A Progress Report (Cambridge, MA, 1951), p. 12. While it is likely that quits were the cause of most of the separations of workers in our sample, variables were included in our specification to represent factors expected to influence all types of job terminations—layoffs, firings, retirements, and deaths, as well as voluntary quits.Google Scholar

25 Preliminary analysis which stratified the sample by nativity found no statistically significant difference in coefficients between native- and foreign-born workers. Hence, the full sample estimates are reported in Table 3.Google Scholar

26 Gutman, Work, Culture and Society, p. 24Google Scholar; Nelson, Managers and Workers; and Jacoby, Employing Bureaucracy.Google Scholar

27 In a proportional hazard model individuals with different background characteristics have hazards which are proportional to each other at any given level of tenure, t. In a proportional hazard of the form: the impact of a 0−1 dummy variable, say Ζ1, can be measured as the ratio of the hazard for an individual for whom Ζ1 = 1 to the hazard for a person for whom Ζ1 = 0: Similarly, the impact of a change in a continuous variable, say Ζ2, can be expressed as exp(β2ΔΖ2).Google Scholar

28 Flinn, Christopher J. and Heckman, James J., “Models for the Analysis of Labor Force Dynamics,” in Basmann, Robert L. and Rhodes, George F., eds., Advances in Econometrics (Greenwich, 1982).Google Scholar

29 Abraham, Katherine B. and Farber, Henry S., “Job Duration, Seniority, and Earnings,” American Economic Review, 77 (06 1987), pp. 278–97.Google Scholar

30 Weiner, Stuart E., “Labor Market Flows and Nested Survival Models” (Research Working Paper No. 85–05, Federal Reserve Bank of Kansas City, July 1985).Google Scholar

31 Weiner, “Labor Market Flows”Google Scholar; Meitzen, Mark E., “Differences in Male and Female Job-Quitting Behavior,” Journal of Labor Economics, 4 (04 1986), pp. 151–67CrossRefGoogle Scholar; Burdett, Kenneth, Kiefer, Nicholas, and Sharma, Sunil, “Layoffs and Duration Dependence: A Model of Turnover,” Journal of Econometrics, 28 (04 1985), pp. 51–69CrossRefGoogle Scholar; and Abraham and Farber, “Job Duration.” These ambiguous effects of schooling and age may be due to the fact that in our specification these variables capture offsetting effects. While the more educated may have greater alternative opportunities and hence a higher probability of quitting, they may also be of greater value to the firm and therefore less likely to be laid off. The coefficient on AGE may also reflect several counteracting forces. On the one hand, youths may enter the job with less knowledge of their labor market opportunities and engage in more job switching. As workers age, the number of years over which returns to job mobility can be captured decline. These factors make the probability of job leaving Fall with age. On the other hand, older workers who have acquired a greater stock of general skills may have greater job opportunities and thus be more likely to switch jobs if this tendency is not offset by firm efforts to retain more experienced workers. Older workers are also more likely to leave jobs through retirement and death. These factors cause job leaving to rise with age.Google Scholar

32 See Parsons, Donald O., “Specific Human Capital: An Application to Quit Rates and Lay Off Rates,” Journal of Political Economy, 80 (11/12 1972), pp. 1120–43CrossRefGoogle Scholar; and Pencavel, John H., “wages, Specific Training, and Labor Turnover in U.S. Manufacturing Industries,” International Economic Review, 13 (02 1972), pp. 53–64.CrossRefGoogle Scholar

33 Alternatively, the above-market wages offered by firms in capital-intensive industries, which in turn lead to low quit rates, may result from efforts to discourage shirking and/or to share risks or rents. For a survey of these models, see Akerlof, George A. and Yellen, Janet L., eds., Efficiency Wage Models of the Labor Market (Cambridge, 1986), pp. 1–21CrossRefGoogle Scholar; and Raff, Daniel M. G., “Wage Determination Theory and the Five-Dollar Day at Ford,” this JOURNAL, 48 (06 1988), pp. 387–99. These explanations are consistent with our findings as well.Google Scholar

34 See Cochran, Thomas C., The Pabst Brewing Company: The History of an American Business (New York, 1948).Google Scholar

35 See testimony of brewers reported in California Bureau of Labor Statistics, Third Biennial Report (Sacramento, 1888).Google Scholar

36 Jovanovic, Boyan, “Firm-Specific Capital and Turnover,” Journal of Political Economy, 87 (12 1979), p. 1246.CrossRefGoogle Scholar

37 See Burdett, Kiefer, and Sharma, “Layoffs and Duration Dependence”; Abraham and Farber, “Job Duration”; Weiner, “Labor Market Flows”; and Meitzen, “Differences in Male and Female Job-Quitting Behavior.”Google Scholar

38 Lancaster, Tony, “Econometric Methods for the Duration of Unemployment,” Econometrica, 47 (07 1979), pp. 939–56.CrossRefGoogle Scholar

39 See Flinn and Heckman, “Models for the Analysis of Labor Force Dynamics.” There are further econometric reasons for questioning the reliability of the estimates of duration dependence. Heckman and Burton Singer find that estimates of duration dependence are highly sensitive to the manner in which covariates that vary across the course of the job, such as age, are introduced into the hazard. They also point out that in single spell duration models it may be impossible to separate the effects of time-varying regressors from true duration dependence, particularly in the case of a regressor such as age which has no independent variation from time on the job.Google Scholar See Heckman, James and Singer, Burton, “Econometric Analysis of Longitudinal Data,” in Griliches, Zvi and Intriligator, Michael, eds., Handbook of Econometrics, vol. 3 (New York, 1986).Google Scholar

40 In our computation we assume that workers do not work beyond age 65. For evidence on the timing and extent of retirement in the late nineteenth century, see Ransom, Roger L. and Sutch, Richard, “The Labor of Older Americans: Retirement of Men On and Off the Job, 1870–1937,” this JOURNAL, 46 (03 1986), pp. 1–30.Google Scholar

41 The absence of spells of less than one year arises from the fact that short spells have a low probability of being sampled at any point in time. Abraham and Farber also find that short spells are virtually nonexistent among the sample of workers whose spells were in progress at the time the survey used in their analysis was taken. Only 0.6 percent of these workers were predicted to have completed employment spells of less than a year. See Abraham and Farber, “Job Duration.”Google Scholar

42 In 1890, 39.2 percent of males engaged in gainful, nonagricultural occupations were single. See U.S. Department of the Interior, Bureau of the Census, Report on the Populalion of the United States at the Eleventh Census: 1890 (Washington, DC, 1897), vol. 1, part 2, p. 744, table 119. Average “family” size was 4.9, where “family” is defined as “not only the normal family, as generally understood, but also all persons living alone, and all larger aggregations of people having only the tie of a common roof and table, as the inmates of hotels, hospitals, prisons, and asylums.” The census goes on to express the opinion that “considered as regards great bodies of population, the presence of these large ‘census’ families does not probably have any very appreciable effect on the average size of the family.“Google Scholar See Eleventh Census: 1890, vol. 1, part 1, p. clxxxviii.Google Scholar We took 3.9 as the average number of dependents per male wage earner. Thirty-seven percent of nonfarm families owned their own homes. See Eleventh Census: 1890, vol. 13, p. 27.

43 De Long and Summers, “Changing Cyclical Variability,” p. 41.Google Scholar

44 Akerlof and Main's estimated mean completed tenure of 18 years refers to all industries. For manufacturing, the relevant comparison to our calculations, their estimates are 18.3 and 19.5 for durable and nondurable manufacturing, respectively. See Akerlof and Main, “ExperienceWeighted Measures,” p. 1007.Google Scholar

45 Ibid., p. 1008.

46 Salant, Stephen W., “Search Theory and Duration Data: A Theory of Sorts,” Quarterly Journal of Economics, 91 (02 1977), pp. 39–57.CrossRefGoogle Scholar

47 Amemiya, Takeshi, Advanced Econometrics (Cambridge, MA, 1985). Note that this result relies on the assumption of a constant entry rate. That is, the probability of beginning a job is unrelated to calendar time.Google Scholar

48 Berndt, Ernst K. et al. , “Estimation and Inference in Nonlinear Structural Models,” Annals of Economic and Social Measurement, 3/4 (1974), pp. 653–65.Google Scholar