Hatton, Timothy J. 2010. THE CLIOMETRICS OF INTERNATIONAL MIGRATION: A SURVEY. Journal of Economic Surveys, Vol. 24, Issue. 5, p. 941.
Eckard, E. Woodrow 2010. Anti-Irish Job Discrimination circa 1880. Social Science History, Vol. 34, Issue. 04, p. 407.
Minns, Chris 2000. Income, Cohort Effects, and Occupational Mobility: A New Look at Immigration to the United States at the Turn of the 20th Century. Explorations in Economic History, Vol. 37, Issue. 4, p. 326.
BUFFUM, DAVID and WHAPLES, ROBERT 1995. FEAR AND LATHING IN THE MICHIGAN FURNITURE INDUSTRY: EMPLOYEE-BASED DISCRIMINATION A CENTURY AGO. Economic Inquiry, Vol. 33, Issue. 2, p. 234.
Rosenbloom, Joshua L. 1994. Looking for Work, Searching for Workers: U.S. Labor Markets after the Civil War. Social Science History, Vol. 18, Issue. 03, p. 377.
Carter, Susan B. Ransom, Roger L. and Sutch, Richard 1991. The Historical Labor Statistics Project at the University of California. Historical Methods: A Journal of Quantitative and Interdisciplinary History, Vol. 24, Issue. 2, p. 52.
CARTER, SUSAN B. 1988. The Changing Importance of Lifetime Jobs, 1892–1978. Industrial Relations, Vol. 27, Issue. 3, p. 287.
Late nineteenth-century immigrants tended to concentrate in large cities despite the fact that they experienced less occupational mobility there than they did in small cities. This paper suggests that variation across cities in labor management systems and in the associated froms of discrimination may help to explain this apparent paradox. Analysis of data from Michigan's agricultural implements and iron-working industries in 1890 indicates that discrimination in hiring made it more difficult for immigrants to break into the small-city labor force. But in large cities, immigrant mobility was restricted by discriminatory barriers to entry into higher level jobs.
1 Hutchinson E. P., Immigrants and Their Children, 1850–1950, Census Monograph (New York, 1956), p. 22.
2 Gallaway Lowell E., Vedder Richard K., and Shukla Vishwa, “The Distribution of the Immigrant Population in the United States: An Economic Analysis,” Explorations in Economic History, 11 (Spring 1974), 225.
3 It should be noted that the Gallaway, Vedder, and Shukla study leaves open the possibility that immigrant settlement choices could have been influenced by economic factors other than labor market conditions. For example, immigrants may have preferred large cities because economies of scale in the provision of special goods and services demanded by ethnic groups resulted in the concentration of such things as ethnic food markets and restaurants or foreign-language newspapers in large cities.
4 The larger scale and greater complexity of the division of labor in large cities have been posited as explanations for this relationship. See Mueller Charles M., “City Effects on Socioeconomic Achievements: The Case of Large Cities,” American Sociological Review, 39 (10. 1974), 658–67;Duncan Otis D. and Reiss Albert, Social Characteristics of Urban and Rural Communities, 1950 (New York, 1956); and Schnore Leo, “The Socioeconomic Status of Cities and Suburbs” American Sociological Review, 28 (02 1963), 76–85.
5 It should be noted that Gallaway, Vedder, and Shukla attempted to include “job opportunities” as an explanatory variable in their settlement model. But, their proxy for job opportunities— per capita income growth in the preceding two decades—is not a very good proxy for mobility. It seems quite possible that they have omitted a crucial decision variable by falling to include a more direct measure of mobility.
6 Thernstrom Stephen, Poverty and Progress (Cambridge, Massachusetts, 1964), pp. 201–03.
7 Kirk Gordon W., The Promise of American Life: Social Mobility in a Nineteenth Century Immigrant Community, Holland, Michigan, 1847–1894 (Philadelphia, 1978);Kirk Gordon W. and Kirk Carolyn T., “The Immigrant, Economic Opportunity, and Type of Settlement in Nineteenth Century America,” this JOURNAL, 38 (03 1978), 226–34.
8 U. S. Congress, Senate, Abstracts of Reports of the Immigration Commission, S. Doc. 747, 61st Cong., 3rd Sess. (Washington, D. C., 1911), p. 780.
9 Bureau Michigan of Labor and Industrial Statistics, Eighth Annual Report, 1891 (Lansing, 1891).
10 The Eighth Annual Report provides no information about the firms included in the survey except their location and product. Judging by the numbers surveyed, however, the sample seems to include a wide range of firm sizes with a bias towards large firms. The 2,078 Michigan firms producing similar products listed in the 1890 Census employed a total of 18,507 workers or 8.91 employees per firm. The size of the Bureau of Labor survey implies an average firm size of at least 44 employees.
11 The smaller cities and towns range in size from 46,322 to 505. One would like to divide this further into medium and small cities but the resulting sample sizes would be too small.
12 The introduction to the Eighth Annual Report indicates that the canvassers simply attempted to interview as many employees as possible in each firm included in the survey.
13 Eighth Annual Report, p. xi.
14 It should be noted that the foreign-born sample is dominated by immigrants from Northern and Western Europe. Germans alone make up almost half, and Northwestern Europeans together with British and Canadians over 90 percent of the foreign born. Thus a very small proportion represent the so-called second-wave nationalities who the literature suggests were the most likely targets of discrimination.
15 Variations in weeks worked of course may reflect voluntary individual decisions about labor supply or employer decisions about labor demand. The Bureau of Labor also collected information on the reasons for time lost from work, however. For the sample as a whole, 76 percent of the time lost was attributed to layoffs, 15 percent to sickness and accidents, 7 percent to holidays and vacations, and 2 percent was unexplained. Moreover, there was little variation across nativity groups in the reasons for lost time: 78 percent of the foreign born and 74 percent of the natives attributed their lost time to layoffs. Thus it seems unlikely that the fewer weeks worked by the foreign born resulted from voluntary decisions. See Hannon Joan Underhill, “The Immigrant Worker in the Promised Land: Human Capital and Ethnic Discrimination in the Michigan Labor Market, 1888–1890,” Unpub. Ph.D. dissertation (Madison, 1977).
16 The occupation index shown in Table 1 is Otis D. Duncan's socioeconomic index. The index weights are based on the 1950 Census and attempt to capture both the income and education levels associated with each occupation in 1950. See Duncan Otis D., “A Socioeconomic Index for all Occupations,” in Reiss Albert J., et al. , Occupations and Social Status (New York, 1961).The use of 1950 index weights is questionable, of course. Yet the construction of alternative indices with a wide variety of weights suggests that the results are very insensitive to the choice of weights. See Hannon, “The Immigrant Worker”, pp. 375–82.
17 The least squares estimators of the regression coefficients for equation (1) are exactly the same as those that would be obtained from three separate regressions (for natives, second generation, and foreign born). The difference between the two approaches concerns σ2. If the variance of the error terms is unchanged throughout the entire sample, then the estimate of σ2 obtained from equation (1) will be efficient, whereas the three estimates obtained from the three subsamples would not be. One problem with estimating equation (1), however, is that the binary nativity variables are likely to be highly correlated with the interaction terms; i.e., N is likely to be highly correlated with NS and P with PS. Since multicollinearity tends to increase the variance of the estimated regression coefficients, it will tend to reduce the significance of the coefficients on the nativity variables, and thus reduce the possibility that one will find significant evidence of discrimination. See Kmenta Jan, Elements of Econometrics (New York, 1971), pp. 387–89, 419–21.
18 Recent debate over discrimination against the foreign born has focused on wage discrimination. See Higgs Robert, “Race, Skills, and Earnings: American Immigrants in 1909”, this JOURNAL, 31 (06 1971), 420–28;Blau Francine, “Immigration and Labor Earnings in Early Twentieth Century America”, Research in Population Economics, 2 (01 1980), 21–41; and McGouldrick Paul F. and Tannen Michael B., “Did American Manufacturers Discriminate Against Immigrants Before 1914?” this JOURNAL, 37 (03 1977), 723–46. Employment instability and occupational segregation have been the focus of much of the historical literature. See Shergold Peter R., “Wage Rates in Pittsburgh During the Depression of 1908,” Journal of American Studies, 9 (1975), 163–88; United States Commissioner of Labor, Eighteenth Annual Report of the Commissioner of Labor, 1903 (Washington, D. C., 1904), pp. 288–309; and Handlin Oscar, The Uprooted, 2nd Edition (Boston, 1973), p. 81.
19 Standard human capital theory has tended to focus on earnings as the major job reward. The sociological research on the distribution of job rewards has been dominated by the causal modelling of Blau and Duncan, where the emphasis is on occupational status as the job reward of interest. A direct extension of the model views income as determined jointly by occupation and human capital. See Becker Gary, Economic Theory (New York, 1971);Blaug Marc, “The Empirical Status of Human Capital Theory: A Slightly Jaundiced Survey,” Journal of Economic Literature, 14 (1976), 829–55;Mincer Jacob, Schooling Experience and Earnings (New York, 1974);Blau Peter M. and Duncan Otis D., The American Occupational Structure (New York, 1967);Sewell W. H. and Hauser R. M., Education, Occupation, and Earnings: Achievement in the Early Career (New York, 1975);Duncan G. J., “Earnings Functions and Nonpecuniary Benefits,” Journal of Human Resources, 11 (1976), 462–83; and Kalleberg Arne L. and Sorensen Page B., “The Sociology of Labor Markets” Annual Review of Sociology, 5 (1979), 351–79.
20 Equation (2) provides a test for wage discrimination. In the absence of data on turnover or unemployment rates by nativity, equation (3) provides a test of the hypothesis that the foreign born faced discrimination in employment security. Of course, the test is a very rough one since the variation in weeks worked may reflect individual choices, but see note 15 above. Equation (4) provides a test for occupational segregation.
21 See Mincer Jacob, Schooling, Experience, and Earnings (New York, 1974).
22 Higgs also argues that the ability to speak English was a crucial determinant of foreign born earnings in the United States. Higgs Robert, “Race, Skills, and Earnings”, pp. 420–8. For contemporary evidence on differential returns to experience in the old country versus experience in the United States, see Chiswick Barry R., “The Effect of Americanization on the Earnings of Foreign-born Men”, Journal of Political Economy, 86 (10 1978), 897–921.
23 Comparing the return to YRSUS for the foreign born with the return to AGE for natives may be a very strict test for discrimination. In the absence of discrimination, one might expect the foreign born's return to exceed (rather than simply equal) that of natives for two reasons. First, recent immigrants are likely to have less knowledge of the customs and language and less information about U.S. job opportunities. This would produce lower earnings initially, but as immigrants informally gain information and skills relevant to the U.S. labor market, their earnings might rise at a faster rate than the earnings of natives. Second, economic theory suggests that migration is likely to be a selective process bringing those more able and highly motivated. This self-selection in migration suggests that for the same schooling, age, and other characteristics, the foreign born might have more innate ability or be more highly motivated than natives. This also might produce a steeper experience-earnings profile for the foreign born than for natives. See Chiswick, “The Effect of Americanization”, pp. 899–901.
24 Contemporary evidence suggests that married men tend to have higher labor force participation rates, invest more in human capital, and have better health than men who are not married. For the same age, schooling, and place of residence, married men have higher earnings. See Chiswick, “The Effect of Americanization”, p. 900.
25 Several alternative forms of the equation, including a simple linear earnings function and one in which the experience variables are entered as AGE, NAGE, PAGE, and YRSUS (providing an estimate of the difference between the return to AGE and YRSUS for the foreign born), have been estimated. The results do not differ from those presented in Table 2.
26 The rates of return to U. S. experience are dLn(Y)/dYRSUS = b2 + 2d1 AGE for the foreign born, dLn(Y)/dAGE = b3 + 2(d1 + d2)AGE for the second generation, and dLn(Y)/dAGE = b3 + b4 + 2(d1 + d2 + d3)AGE for natives, where b2, b3, and b4 are the estimated coefficients on YRSUS, AGE and PAGE, and d1, d2, and d3 are the estimated coefficients on AGE2, NAGE2, and PAGE2. Table 2 implies the following rates of return by age: Detroit and Grand Rapids
27 The quality effect is given by (b2 - b1)(18), where b2 is the coefficient on YRSUS and b1 is the coefficient on AGEMIG. Discrimination, defined as differences in the estimated intercepts and returns to U.S. experience, is measured by (a2 + a3) + (b3 + b4 - b2)(18) + (d2 + d3)(182), where a2 and a3 are the coefficients on N and P; b2, b3 and b4 are the coefficients on YRSUS, AGE, and PAGE; and d2 and d3 are the coefficients on NAGE2 and PAGE2.
28 See Flanagan Robert J., “Segmented Market Theories and Racial Discrimination”, Industrial Relations, 12 (10. 1973), 253–73.
29 These two occupational categories actually comprise 81 percent of the sample.
30 The monetary implications of these occupational gaps can be assessed by using the structural weeks worked and earnings equations to predict the resulting differences in annual earnings. According to the weeks worked equation for Grand Rapids and Detroit, the 4.47 index point gap between 18 year old native and foreign-born workers would produce a 0.79 percent difference in weeks worked. Together with the gap in occupational status, this would produce a 7.5 percent gap in annual earnings. By age 40, the earnings gap attributable to occupational discrimination would have widened to 10.2 percent. These predictions are generated by using the same coefficients (those for the foreign born) on OCC and Ln(WW) for both groups so that the only differences between them are in occupational status and the associated difference in weeks worked.
31 This interpretation is weakened somewhat by the fact that AGEMIG also enters the small city weeks worked equation with a significant positive coefficient, suggesting that experience in general—rather than firm specific training in particular—was an important determinant of weeks worked for the foreign born. It should also be noted that experience indirectly influences weeks worked for both natives and the foreign born since experience affects occupational status and occupation is a significant determinant of weeks worked.
32 It should be noted that recent arrivals who had been in the United States for less than one year would obviously work for fewer than 52 weeks. This alone might make YRSUS a significant determinant of weeks worked. Yet only 15 of the foreign-born workers (1 percent of the large city sample and 2 percent of the small city sample) had been in the U.S. less than one year. On average, the foreign born in large cities had been in the U.S. for 12.6 years and those in small cities for 17.7 years.
33 A movement from semiskilled to skilled (a 13 point increase in OCC), for example, would produce a 2.3 percent increase in weeks worked for the foreign born, compared with a 6.2 percent increase for natives of native parentage. According to the structural earnings equation, these improvements in job stability would yield a 2.9 percent increase in annual earnings for the foreign born and a 7.7 percent increase in annual earnings for natives. These earnings predictions are generated by using the same coefficient (that for the foreign born) for both groups so that the only difference between them is in the number of weeks worked.
34 This pattern also appears in other Michigan industries. See Hannon, “The Immigrant Worker”.
35 The average age of the second generation in Detroit and Grand Rapids is only 24.13 compared to 30.84 for the foreign born and 30.58 for natives of native parentage. In the small city sample, the average age of the second generation is 28.24, compared to 33.96 for the foreign born and 32.97 for natives of native parentage.
36 Becker Gary, Human Capital (New York, 1964), pp. 7–29.
37 See Chiswick, “The Effect of Americanization”, p. 900. This explanation is consistent with the data on weeks worked for Grand Rapids and Detroit. The weeks worked equation predicts that for a given age, the second generation would work slightly more weeks than either the foreign born or natives. But the average second generation worker in the large city sample actually worked slightly fewer weeks than either natives or the foreign born, suggesting that the second generation's youth may have been associated with a higher quit rate. In small cities, where the second generation is slightly older, however, they worked more weeks on average than did the foreign born or natives. Yet the second generation's estimated earnings profile is somewhat steeper than that of natives here as well. Chiswick suggests that the foreign born, like youth, might be more likely to invest in general rather than firm specific training. If so, we are left with no explanation for the difference between the estimated returns to experience for the second generation versus the foreign born in large cities.
38 The occupation profiles for Grand Rapids and Detroit are consistent with Thernstrom's finding that the second generation were more mobile than their parents in both directions. See Thernstrom Stephan, The Other Bostonians (Cambridge, Massachusetts, 1973).
39 The regressions, of course, are based on cross-sectional data and are only suggestive of the lifetime earnings or occupational experience of individuals. I have shown elsewhere, however, that changes over time in the composition of the immigrant stock or in the degree of discrimination are unlikely explanations of the cross section profiles. See Hannon, “The Immigrant Worker”, pp. 340–46.
40 See Hannon, “The Immigrant Worker,” pp. 283–89.
41 Although the evidence is somewhat weaker, the same pattern is found in Michigan's furniture industry. See Hannon, “The Immigrant Worker” pp. 161–223.
42 Contemporary evidence suggests that the degree of discrimination against blacks varies positively with their numerical importance. See Reich Michael, Racial Inequality: A Political Economic Analysis (Princeton, 1981); and Bergman Barbara R. and Lyle Jerolyn R., “The Occupational Standing of Negroes by Areas and Industries,” Journal of Human Resources, 6 (Fall 1971), 411–33. Sociologists have explained this by suggesting that an increase in the numerical importance of a minority group tends to increase prejudice against them since the majority begins to fear their growing power. The historical literature certainly suggests that native employees feared that their wages would be lower and the likelihood of unemployment higher if large numbers of immigrants were allowed to compete with them. See Allport Gordon, The Nature of Prejudice (Cambridge, Massachusetts, 1955), p. 227;Saenger Gerhardt, The Social Psychology of Prejudice (New York, 1953), p. 99; and Williams R. W., The Reduction of Intergroup Prejudice (New York, 1947), p.47. Even given the level of prejudice, models of labor market discrimination as different as Becker's model of employer discrimination and the crowding model developed by Bergman and others imply that the degree of discrimination will increase as the relative size of the minority group increases. See Becker Gary, The Economics of Discrimination, 2nd Ed. (Chicago, 1971);Bergman Barbara R., “The Effect on White Incomes of Discrimination in Employment,” Journal of Political Economy, 79 (03/04 1971), 294–313;Bergman and Lyle, “The Occupational Standing of Negroes,” pp. 411–33;Chiswick Barry R., “Racial Discrimination in the Labor Market: A Test of Alternative Hypotheses,” Journal of Political Economy, 81 (11/12. 1973), 330–52; and Thurow Lester, Poverty and Discrimination (Washington, D. C. 1969).
43 The following discussion draws heavily on Edwards Richard, Contested Terrain (New York, 1979);Nelson Daniel, Managers and Workers: Origins of the New Factory System in the United States, 1880–1920 (Madison, Wisconsin, 1975);Gordon David M., Edwards Richard, and Reich Michael, Segmented Work, Divided Workers (Cambridge, Massachusetts, 1982); and Korman Gerd, Industrialization, Immigrants, and Americanizers: The View From Milwaukee (Madison, 1967).
44 This is a straightforward extension of the argument above about the role of information networks.
45 Oscar Handlin, for example, argues that in mass production industries, “The labor force was decisively divided … Completely segregated, they (immigrants) could hardly hope to advance their positions.” See Handlin, The Uprooted, p. 81. For further evidence on discrimination by foremen and other hired bosses, see Hannon, “Ethnic Discrimination”, and Korman, Industrialization, Immigrants, and Americanizers.
46 In Detroit and Grand Rapids, the average number surveyed per firm was 52, compared with 38 for firms in small cities and towns. Only 25 percent of the large city sample, compared with over 40 percent of the small city sample, came from firms where fewer than 50 employees were canvassed. Almost 30 percent of the large city sample, compared with 17 percent of the small city sample, came from firms where over 200 employees were canvassed.
Email your librarian or administrator to recommend adding this journal to your organisation's collection.
Full text views reflects the number of PDF downloads, PDFs sent to Google Drive, Dropbox and Kindle and HTML full text views.
* Views captured on Cambridge Core between September 2016 - 23rd November 2017. This data will be updated every 24 hours.