Did the black ghetto have a golden age? Class structure and class segregation in New York City, 1949-1970, with initial evidence for 1990

Journal of Urban History

Volumn 23, Issue 1, 1996, Pages 3-28

  Fainstein, Norman ^a   Nesbitt, Susan ^b  

^a VASSAR COLLEGE   (United States)

^b St Joseph's College

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0345290603     PISSN: 00961442     EISSN: None     Source Type: Journal    
DOI: 10.1177/009614429602300101     Document Type: Article

References (50)

1. See Norman Fainstein, "Race, Class, and Segregation: Discourses about African Americans," Journal of Urban and Regional Research 17, 3 (1993), 384-403; quotation is from William J. Wilson, The Truly Disadvantaged (Chicago, 1987), 7.
2. See Norman Fainstein, "Race, Class, and Segregation: Discourses about African Americans," Journal of Urban and Regional Research 17, 3 (1993), 384-403; quotation is from William J. Wilson, The Truly Disadvantaged (Chicago, 1987), 7.
3. Wilson, The Truly Disadvantaged, 7.
4. Loic Wacquant, "Urban Outcasts: Stigma and Division in the Black American Ghetto and the French Urban Periphery," International Journal of Urban and Regional Research 17, 3 (1993), 367. His reference is to Kenneth Clark, Dark Ghetto (New York, 1965). Ibid.
5. Loic Wacquant, "Urban Outcasts: Stigma and Division in the Black American Ghetto and the French Urban Periphery," International Journal of Urban and Regional Research 17, 3 (1993), 367. His reference is to Kenneth Clark, Dark Ghetto (New York, 1965). Ibid.
6. Loic Wacquant, "Urban Outcasts: Stigma and Division in the Black American Ghetto and the French Urban Periphery," International Journal of Urban and Regional Research 17, 3 (1993), 367. His reference is to Kenneth Clark, Dark Ghetto (New York, 1965). Ibid.
7. Douglas Massey and Nancy Denton, American Apartheid (Cambridge, MA, 1993).
8. See Fainstein, "Race, Class, and Segregation."
9. As to whether the relatively greater interclass contact between the poor and nonpoor among African Americans compared with whites has a positive effect on uplifting the poor, Massey remains uncertain. Absent racial integration, it is unclear the extent to which nonpoor blacks can redefine the pathological social dynamics of neighborhoods with high levels of poverty.
10. Joe W. Trotter, Jr., Black Milwaukee (Urbana, 1985).
11. St. Claire Drake and Robert Cayton, Black Metropolis (New York, 1945). Elijah Anderson, Streetwise (Chicago, 1990), 58.
12. St. Claire Drake and Robert Cayton, Black Metropolis (New York, 1945). Elijah Anderson, Streetwise (Chicago, 1990), 58.
13. Drake and Cayton, Black Metropolis, 521, 604.
14. Thus the most important national study of residential segregation at mid-century outlines the ways in which segregation inhibits the social mobility of the better-off by reducing class spatial segregation among blacks; see Davis McEntire, Race and Residence (Berkeley, 1960), ch. 5. More recently, Roger Lane, "Black Philadelphia Then and Now," in Adele Harrell and George Peterson, eds., Drugs, Crime, and Social Isolation (Washington, DC, 1992), 43, finds the roots of black economic disadvantage in contemporary Philadelphia (described by him in terms of an isolated underclass) in its ghettoized past: "Although the negative effects of social isolation, which rob whole neighborhoods of working role models, are real, so were the costs of the tight residential segregation that preceded it. The costs to the middle class of living in high crime districts, not the least of which was the problem of transmitting values to their children, were widely recognized in earlier years."
15. Thus the most important national study of residential segregation at mid-century outlines the ways in which segregation inhibits the social mobility of the better-off by reducing class spatial segregation among blacks; see Davis McEntire, Race and Residence (Berkeley, 1960), ch. 5. More recently, Roger Lane, "Black Philadelphia Then and Now," in Adele Harrell and George Peterson, eds., Drugs, Crime, and Social Isolation (Washington, DC, 1992), 43, finds the roots of black economic disadvantage in contemporary Philadelphia (described by him in terms of an isolated underclass) in its ghettoized past: "Although the negative effects of social isolation, which rob whole neighborhoods of working role models, are real, so were the costs of the tight residential segregation that preceded it. The costs to the middle class of living in high crime districts, not the least of which was the problem of transmitting values to their children, were widely recognized in earlier years."
16. On this point, see the review of the literature in Norman Fainstein, "Black Ghettoization and Social Mobility," in Michael P. Smith and Joe Feagin, eds., The Bubbling Cauldron (Minneapolis, 1996).
17. Because census tracts are treated here as unitary wholes, we are unable to analyze contact or isolation within tracts, much less at the levels of individual apartment houses and tenements. Whereas the census provides block data after 1950, they do not permit a class analysis at the sub-tract level. It should also be added that the census count certainly underestimates the presence of the most impoverished and unstable elements of all communities at each period. Nonetheless, census data permit an accurate description of the racial forest - if not of every tree - and of how its configuration changed over time.
18. We include the suburbs in our analyses of class structure and we report on research that has examined class inequality at the metropolitan area. At this point, however, our own findings on class segregation are limited to the central city. Because we end the main part of this article at 1970, the effect of concentrating on the central city are minimal for blacks (as black suburbanization does not become significant until the 1970s). The effect on white class segregation is to somewhat understate its magnitude. After 1970, middle-class whites were much more successful than comparable blacks in isolating themselves in the suburbs, so we expect that further study of white class segregation at the metropolitan level will make the black situation even more divergent from the white over the entire postwar period, and especially in the most recent decades; on this point, see the strong evidence provided in John Logan and Richard Alba, "Locational Returns to Human Capital: Minority Access to Suburban Community Resources," Demography 30 (May 1993), 243-68.
19. By geography we mean the physical positioning of census tracts over the urban terrain. Our data tell us whether households occupy the same tracts. But when they do not, we have no way of knowing whether their different tracts are adjacent or miles apart.
20. The Index of Dissimilarity represents one-half the mean of the absolute value of the differences between the percentage of two groups in each census tract or other unit of analysis. It is commonly interpreted as the percentage of the population of either group that would have to change census tracts so as to achieve a distribution in which the percentage of the two groups is identical in each tract to their percentage in the total population of the city. D ranges between 0.00 for a completely unsegregated situation to 1.00 for complete segregation. Values of D for racial segregation run between .60 and .85 for major U. S. urban areas. Values above .60 are considered high. See Douglas Massey and Nancy Denton, "The Dimensions of Racial Segregation," Social Forces 67, 2 (1988), 281-316.
21. Otis Duncan and Beverly Duncan report in The Negro Population of Chicago (Chicago, 1957, Table 69) on D for employed males in major occupational groups in Chicago in 1950. They find that D between professional/managerial occupations at the upper end and laborers at the lower end of the class hierarchy runs at about .42 for nonwhites and about .33 for whites. (The white figure is probably reduced because much larger "community areas" were utilized instead of census tracts.) E. S. Uyeki in "Residential Distribution and Stratification, 1950-1960," American Journal of Sociology 69, 5 (1964), Table 4, looks at all employed males in Cleveland (with no racial breakdown) and finds a D of .52 at the extremes of the distribution. A. Wilkins examines occupational segregation for all males in eight middle-sized cities in 1950 and finds similar patterns across the occupational grid in such diverse locations as Syracuse, Memphis, and Fort Worth; see "The Residential Distribution of Occupational Groups in Eight Middle-sized American Cities," (doctoral dissertation, University of Chicago, 1956). Karl Taeuber and Alma Taeuber in Negroes in Cities (Chicago, 1965, Table C-1) provide data on occupational segregation for nonwhite males only in 10 large and middle-sized cities (excluding New York) in 1950 and 1960. Using block data, they find a range in D for the occupational extremes from .27 in Philadelphia to .47 in St. Louis in 1950. Albert Simkus offers similar data broken down for race in 10 urbanized areas (excluding New York) in 1960 and 1970; see "Residential Segregation by Occupation and Race in Ten Urbanized Area, 1950-1970," American Sociological Review 43 (February 1978), Table 3. Nathan Kantrowitz, in the only study devoted entirely to New York City (in fact, the only one we have found that even includes New York prior to 1970), examines D for income and education in 1960 for the SMSA. His results (Table 3.1) on income are virtually identical to ours for nonwhites, and only slightly different for whites (we have not examined tracts outside the NYC boundaries), with D at the extremes of the distribution running .43 for whites and .50 for nonwhites, but with even closer figures for the parts of the income range where significant numbers of nonwhites were actually present; see Ethnic and Racial Segregation in the New York Metropolis (New York, 1973). Reynolds Farley examines average levels of D by race for occupation, income, and education in the 29 largest U. S. cities and their suburbs in 1970. He finds average levels of D for income of .21 for whites in central cities and .25 for blacks; using occupation, the comparable scores are .22 for whites and .21 for blacks. In his detailed breakdown by city, he shows that New York is exactly on the mean for D by occupation in the city sample; see "Residential Segregation in Urbanized Areas of the United Sates in 1970: An Analysis of Social Class and Racial Differences," Demography 14, 4 (1977), 497-518. Finally, Farley presents the most thorough analysis of trends in the 1970s. He finds rather similar levels of class segregation using the index of dissimilarity for major categories of metropolitan areas. City by city comparisons show virtually identical patterns of spatial segregation by income between the poor and nonpoor among blacks and whites, with an average index of .29 - compared, of course, with indexes in the .70s and .80s for racial segregation. He also shows no consistent pattern of increase in class segregation among blacks during the 1970s (as had been posited by many theorists of the underclass); see "Residential Segregation of Social and Economic Groups Among Blacks, 1970-1980," in Christopher Jencks and Paul Peterson, eds., The Urban Underclass (Washington, 1991), 274-98.
22. Otis Duncan and Beverly Duncan report in The Negro Population of Chicago (Chicago, 1957, Table 69) on D for employed males in major occupational groups in Chicago in 1950. They find that D between professional/managerial occupations at the upper end and laborers at the lower end of the class hierarchy runs at about .42 for nonwhites and about .33 for whites. (The white figure is probably reduced because much larger "community areas" were utilized instead of census tracts.) E. S. Uyeki in "Residential Distribution and Stratification, 1950-1960," American Journal of Sociology 69, 5 (1964), Table 4, looks at all employed males in Cleveland (with no racial breakdown) and finds a D of .52 at the extremes of the distribution. A. Wilkins examines occupational segregation for all males in eight middle-sized cities in 1950 and finds similar patterns across the occupational grid in such diverse locations as Syracuse, Memphis, and Fort Worth; see "The Residential Distribution of Occupational Groups in Eight Middle-sized American Cities," (doctoral dissertation, University of Chicago, 1956). Karl Taeuber and Alma Taeuber in Negroes in Cities (Chicago, 1965, Table C-1) provide data on occupational segregation for nonwhite males only in 10 large and middle-sized cities (excluding New York) in 1950 and 1960. Using block data, they find a range in D for the occupational extremes from .27 in Philadelphia to .47 in St. Louis in 1950. Albert Simkus offers similar data broken down for race in 10 urbanized areas (excluding New York) in 1960 and 1970; see "Residential Segregation by Occupation and Race in Ten Urbanized Area, 1950-1970," American Sociological Review 43 (February 1978), Table 3. Nathan Kantrowitz, in the only study devoted entirely to New York City (in fact, the only one we have found that even includes New York prior to 1970), examines D for income and education in 1960 for the SMSA. His results (Table 3.1) on income are virtually identical to ours for nonwhites, and only slightly different for whites (we have not examined tracts outside the NYC boundaries), with D at the extremes of the distribution running .43 for whites and .50 for nonwhites, but with even closer figures for the parts of the income range where significant numbers of nonwhites were actually present; see Ethnic and Racial Segregation in the New York Metropolis (New York, 1973). Reynolds Farley examines average levels of D by race for occupation, income, and education in the 29 largest U. S. cities and their suburbs in 1970. He finds average levels of D for income of .21 for whites in central cities and .25 for blacks; using occupation, the comparable scores are .22 for whites and .21 for blacks. In his detailed breakdown by city, he shows that New York is exactly on the mean for D by occupation in the city sample; see "Residential Segregation in Urbanized Areas of the United Sates in 1970: An Analysis of Social Class and Racial Differences," Demography 14, 4 (1977), 497-518. Finally, Farley presents the most thorough analysis of trends in the 1970s. He finds rather similar levels of class segregation using the index of dissimilarity for major categories of metropolitan areas. City by city comparisons show virtually identical patterns of spatial segregation by income between the poor and nonpoor among blacks and whites, with an average index of .29 - compared, of course, with indexes in the .70s and .80s for racial segregation. He also shows no consistent pattern of increase in class segregation among blacks during the 1970s (as had been posited by many theorists of the underclass); see "Residential Segregation of Social and Economic Groups Among Blacks, 1970-1980," in Christopher Jencks and Paul Peterson, eds., The Urban Underclass (Washington, 1991), 274-98.
23. Otis Duncan and Beverly Duncan report in The Negro Population of Chicago (Chicago, 1957, Table 69) on D for employed males in major occupational groups in Chicago in 1950. They find that D between professional/managerial occupations at the upper end and laborers at the lower end of the class hierarchy runs at about .42 for nonwhites and about .33 for whites. (The white figure is probably reduced because much larger "community areas" were utilized instead of census tracts.) E. S. Uyeki in "Residential Distribution and Stratification, 1950-1960," American Journal of Sociology 69, 5 (1964), Table 4, looks at all employed males in Cleveland (with no racial breakdown) and finds a D of .52 at the extremes of the distribution. A. Wilkins examines occupational segregation for all males in eight middle-sized cities in 1950 and finds similar patterns across the occupational grid in such diverse locations as Syracuse, Memphis, and Fort Worth; see "The Residential Distribution of Occupational Groups in Eight Middle-sized American Cities," (doctoral dissertation, University of Chicago, 1956). Karl Taeuber and Alma Taeuber in Negroes in Cities (Chicago, 1965, Table C-1) provide data on occupational segregation for nonwhite males only in 10 large and middle-sized cities (excluding New York) in 1950 and 1960. Using block data, they find a range in D for the occupational extremes from .27 in Philadelphia to .47 in St. Louis in 1950. Albert Simkus offers similar data broken down for race in 10 urbanized areas (excluding New York) in 1960 and 1970; see "Residential Segregation by Occupation and Race in Ten Urbanized Area, 1950-1970," American Sociological Review 43 (February 1978), Table 3. Nathan Kantrowitz, in the only study devoted entirely to New York City (in fact, the only one we have found that even includes New York prior to 1970), examines D for income and education in 1960 for the SMSA. His results (Table 3.1) on income are virtually identical to ours for nonwhites, and only slightly different for whites (we have not examined tracts outside the NYC boundaries), with D at the extremes of the distribution running .43 for whites and .50 for nonwhites, but with even closer figures for the parts of the income range where significant numbers of nonwhites were actually present; see Ethnic and Racial Segregation in the New York Metropolis (New York, 1973). Reynolds Farley examines average levels of D by race for occupation, income, and education in the 29 largest U. S. cities and their suburbs in 1970. He finds average levels of D for income of .21 for whites in central cities and .25 for blacks; using occupation, the comparable scores are .22 for whites and .21 for blacks. In his detailed breakdown by city, he shows that New York is exactly on the mean for D by occupation in the city sample; see "Residential Segregation in Urbanized Areas of the United Sates in 1970: An Analysis of Social Class and Racial Differences," Demography 14, 4 (1977), 497-518. Finally, Farley presents the most thorough analysis of trends in the 1970s. He finds rather similar levels of class segregation using the index of dissimilarity for major categories of metropolitan areas. City by city comparisons show virtually identical patterns of spatial segregation by income between the poor and nonpoor among blacks and whites, with an average index of .29 - compared, of course, with indexes in the .70s and .80s for racial segregation. He also shows no consistent pattern of increase in class segregation among blacks during the 1970s (as had been posited by many theorists of the underclass); see "Residential Segregation of Social and Economic Groups Among Blacks, 1970-1980," in Christopher Jencks and Paul Peterson, eds., The Urban Underclass (Washington, 1991), 274-98.
24. Otis Duncan and Beverly Duncan report in The Negro Population of Chicago (Chicago, 1957, Table 69) on D for employed males in major occupational groups in Chicago in 1950. They find that D between professional/managerial occupations at the upper end and laborers at the lower end of the class hierarchy runs at about .42 for nonwhites and about .33 for whites. (The white figure is probably reduced because much larger "community areas" were utilized instead of census tracts.) E. S. Uyeki in "Residential Distribution and Stratification, 1950-1960," American Journal of Sociology 69, 5 (1964), Table 4, looks at all employed males in Cleveland (with no racial breakdown) and finds a D of .52 at the extremes of the distribution. A. Wilkins examines occupational segregation for all males in eight middle-sized cities in 1950 and finds similar patterns across the occupational grid in such diverse locations as Syracuse, Memphis, and Fort Worth; see "The Residential Distribution of Occupational Groups in Eight Middle-sized American Cities," (doctoral dissertation, University of Chicago, 1956). Karl Taeuber and Alma Taeuber in Negroes in Cities (Chicago, 1965, Table C-1) provide data on occupational segregation for nonwhite males only in 10 large and middle-sized cities (excluding New York) in 1950 and 1960. Using block data, they find a range in D for the occupational extremes from .27 in Philadelphia to .47 in St. Louis in 1950. Albert Simkus offers similar data broken down for race in 10 urbanized areas (excluding New York) in 1960 and 1970; see "Residential Segregation by Occupation and Race in Ten Urbanized Area, 1950-1970," American Sociological Review 43 (February 1978), Table 3. Nathan Kantrowitz, in the only study devoted entirely to New York City (in fact, the only one we have found that even includes New York prior to 1970), examines D for income and education in 1960 for the SMSA. His results (Table 3.1) on income are virtually identical to ours for nonwhites, and only slightly different for whites (we have not examined tracts outside the NYC boundaries), with D at the extremes of the distribution running .43 for whites and .50 for nonwhites, but with even closer figures for the parts of the income range where significant numbers of nonwhites were actually present; see Ethnic and Racial Segregation in the New York Metropolis (New York, 1973). Reynolds Farley examines average levels of D by race for occupation, income, and education in the 29 largest U. S. cities and their suburbs in 1970. He finds average levels of D for income of .21 for whites in central cities and .25 for blacks; using occupation, the comparable scores are .22 for whites and .21 for blacks. In his detailed breakdown by city, he shows that New York is exactly on the mean for D by occupation in the city sample; see "Residential Segregation in Urbanized Areas of the United Sates in 1970: An Analysis of Social Class and Racial Differences," Demography 14, 4 (1977), 497-518. Finally, Farley presents the most thorough analysis of trends in the 1970s. He finds rather similar levels of class segregation using the index of dissimilarity for major categories of metropolitan areas. City by city comparisons show virtually identical patterns of spatial segregation by income between the poor and nonpoor among blacks and whites, with an average index of .29 - compared, of course, with indexes in the .70s and .80s for racial segregation. He also shows no consistent pattern of increase in class segregation among blacks during the 1970s (as had been posited by many theorists of the underclass); see "Residential Segregation of Social and Economic Groups Among Blacks, 1970-1980," in Christopher Jencks and Paul Peterson, eds., The Urban Underclass (Washington, 1991), 274-98.
25. Otis Duncan and Beverly Duncan report in The Negro Population of Chicago (Chicago, 1957, Table 69) on D for employed males in major occupational groups in Chicago in 1950. They find that D between professional/managerial occupations at the upper end and laborers at the lower end of the class hierarchy runs at about .42 for nonwhites and about .33 for whites. (The white figure is probably reduced because much larger "community areas" were utilized instead of census tracts.) E. S. Uyeki in "Residential Distribution and Stratification, 1950-1960," American Journal of Sociology 69, 5 (1964), Table 4, looks at all employed males in Cleveland (with no racial breakdown) and finds a D of .52 at the extremes of the distribution. A. Wilkins examines occupational segregation for all males in eight middle-sized cities in 1950 and finds similar patterns across the occupational grid in such diverse locations as Syracuse, Memphis, and Fort Worth; see "The Residential Distribution of Occupational Groups in Eight Middle-sized American Cities," (doctoral dissertation, University of Chicago, 1956). Karl Taeuber and Alma Taeuber in Negroes in Cities (Chicago, 1965, Table C-1) provide data on occupational segregation for nonwhite males only in 10 large and middle-sized cities (excluding New York) in 1950 and 1960. Using block data, they find a range in D for the occupational extremes from .27 in Philadelphia to .47 in St. Louis in 1950. Albert Simkus offers similar data broken down for race in 10 urbanized areas (excluding New York) in 1960 and 1970; see "Residential Segregation by Occupation and Race in Ten Urbanized Area, 1950-1970," American Sociological Review 43 (February 1978), Table 3. Nathan Kantrowitz, in the only study devoted entirely to New York City (in fact, the only one we have found that even includes New York prior to 1970), examines D for income and education in 1960 for the SMSA. His results (Table 3.1) on income are virtually identical to ours for nonwhites, and only slightly different for whites (we have not examined tracts outside the NYC boundaries), with D at the extremes of the distribution running .43 for whites and .50 for nonwhites, but with even closer figures for the parts of the income range where significant numbers of nonwhites were actually present; see Ethnic and Racial Segregation in the New York Metropolis (New York, 1973). Reynolds Farley examines average levels of D by race for occupation, income, and education in the 29 largest U. S. cities and their suburbs in 1970. He finds average levels of D for income of .21 for whites in central cities and .25 for blacks; using occupation, the comparable scores are .22 for whites and .21 for blacks. In his detailed breakdown by city, he shows that New York is exactly on the mean for D by occupation in the city sample; see "Residential Segregation in Urbanized Areas of the United Sates in 1970: An Analysis of Social Class and Racial Differences," Demography 14, 4 (1977), 497-518. Finally, Farley presents the most thorough analysis of trends in the 1970s. He finds rather similar levels of class segregation using the index of dissimilarity for major categories of metropolitan areas. City by city comparisons show virtually identical patterns of spatial segregation by income between the poor and nonpoor among blacks and whites, with an average index of .29 - compared, of course, with indexes in the .70s and .80s for racial segregation. He also shows no consistent pattern of increase in class segregation among blacks during the 1970s (as had been posited by many theorists of the underclass); see "Residential Segregation of Social and Economic Groups Among Blacks, 1970-1980," in Christopher Jencks and Paul Peterson, eds., The Urban Underclass (Washington, 1991), 274-98.
26. Otis Duncan and Beverly Duncan report in The Negro Population of Chicago (Chicago, 1957, Table 69) on D for employed males in major occupational groups in Chicago in 1950. They find that D between professional/managerial occupations at the upper end and laborers at the lower end of the class hierarchy runs at about .42 for nonwhites and about .33 for whites. (The white figure is probably reduced because much larger "community areas" were utilized instead of census tracts.) E. S. Uyeki in "Residential Distribution and Stratification, 1950-1960," American Journal of Sociology 69, 5 (1964), Table 4, looks at all employed males in Cleveland (with no racial breakdown) and finds a D of .52 at the extremes of the distribution. A. Wilkins examines occupational segregation for all males in eight middle-sized cities in 1950 and finds similar patterns across the occupational grid in such diverse locations as Syracuse, Memphis, and Fort Worth; see "The Residential Distribution of Occupational Groups in Eight Middle-sized American Cities," (doctoral dissertation, University of Chicago, 1956). Karl Taeuber and Alma Taeuber in Negroes in Cities (Chicago, 1965, Table C-1) provide data on occupational segregation for nonwhite males only in 10 large and middle-sized cities (excluding New York) in 1950 and 1960. Using block data, they find a range in D for the occupational extremes from .27 in Philadelphia to .47 in St. Louis in 1950. Albert Simkus offers similar data broken down for race in 10 urbanized areas (excluding New York) in 1960 and 1970; see "Residential Segregation by Occupation and Race in Ten Urbanized Area, 1950-1970," American Sociological Review 43 (February 1978), Table 3. Nathan Kantrowitz, in the only study devoted entirely to New York City (in fact, the only one we have found that even includes New York prior to 1970), examines D for income and education in 1960 for the SMSA. His results (Table 3.1) on income are virtually identical to ours for nonwhites, and only slightly different for whites (we have not examined tracts outside the NYC boundaries), with D at the extremes of the distribution running .43 for whites and .50 for nonwhites, but with even closer figures for the parts of the income range where significant numbers of nonwhites were actually present; see Ethnic and Racial Segregation in the New York Metropolis (New York, 1973). Reynolds Farley examines average levels of D by race for occupation, income, and education in the 29 largest U. S. cities and their suburbs in 1970. He finds average levels of D for income of .21 for whites in central cities and .25 for blacks; using occupation, the comparable scores are .22 for whites and .21 for blacks. In his detailed breakdown by city, he shows that New York is exactly on the mean for D by occupation in the city sample; see "Residential Segregation in Urbanized Areas of the United Sates in 1970: An Analysis of Social Class and Racial Differences," Demography 14, 4 (1977), 497-518. Finally, Farley presents the most thorough analysis of trends in the 1970s. He finds rather similar levels of class segregation using the index of dissimilarity for major categories of metropolitan areas. City by city comparisons show virtually identical patterns of spatial segregation by income between the poor and nonpoor among blacks and whites, with an average index of .29 - compared, of course, with indexes in the .70s and .80s for racial segregation. He also shows no consistent pattern of increase in class segregation among blacks during the 1970s (as had been posited by many theorists of the underclass); see "Residential Segregation of Social and Economic Groups Among Blacks, 1970-1980," in Christopher Jencks and Paul Peterson, eds., The Urban Underclass (Washington, 1991), 274-98.
27. Otis Duncan and Beverly Duncan report in The Negro Population of Chicago (Chicago, 1957, Table 69) on D for employed males in major occupational groups in Chicago in 1950. They find that D between professional/managerial occupations at the upper end and laborers at the lower end of the class hierarchy runs at about .42 for nonwhites and about .33 for whites. (The white figure is probably reduced because much larger "community areas" were utilized instead of census tracts.) E. S. Uyeki in "Residential Distribution and Stratification, 1950-1960," American Journal of Sociology 69, 5 (1964), Table 4, looks at all employed males in Cleveland (with no racial breakdown) and finds a D of .52 at the extremes of the distribution. A. Wilkins examines occupational segregation for all males in eight middle-sized cities in 1950 and finds similar patterns across the occupational grid in such diverse locations as Syracuse, Memphis, and Fort Worth; see "The Residential Distribution of Occupational Groups in Eight Middle-sized American Cities," (doctoral dissertation, University of Chicago, 1956). Karl Taeuber and Alma Taeuber in Negroes in Cities (Chicago, 1965, Table C-1) provide data on occupational segregation for nonwhite males only in 10 large and middle-sized cities (excluding New York) in 1950 and 1960. Using block data, they find a range in D for the occupational extremes from .27 in Philadelphia to .47 in St. Louis in 1950. Albert Simkus offers similar data broken down for race in 10 urbanized areas (excluding New York) in 1960 and 1970; see "Residential Segregation by Occupation and Race in Ten Urbanized Area, 1950-1970," American Sociological Review 43 (February 1978), Table 3. Nathan Kantrowitz, in the only study devoted entirely to New York City (in fact, the only one we have found that even includes New York prior to 1970), examines D for income and education in 1960 for the SMSA. His results (Table 3.1) on income are virtually identical to ours for nonwhites, and only slightly different for whites (we have not examined tracts outside the NYC boundaries), with D at the extremes of the distribution running .43 for whites and .50 for nonwhites, but with even closer figures for the parts of the income range where significant numbers of nonwhites were actually present; see Ethnic and Racial Segregation in the New York Metropolis (New York, 1973). Reynolds Farley examines average levels of D by race for occupation, income, and education in the 29 largest U. S. cities and their suburbs in 1970. He finds average levels of D for income of .21 for whites in central cities and .25 for blacks; using occupation, the comparable scores are .22 for whites and .21 for blacks. In his detailed breakdown by city, he shows that New York is exactly on the mean for D by occupation in the city sample; see "Residential Segregation in Urbanized Areas of the United Sates in 1970: An Analysis of Social Class and Racial Differences," Demography 14, 4 (1977), 497-518. Finally, Farley presents the most thorough analysis of trends in the 1970s. He finds rather similar levels of class segregation using the index of dissimilarity for major categories of metropolitan areas. City by city comparisons show virtually identical patterns of spatial segregation by income between the poor and nonpoor among blacks and whites, with an average index of .29 - compared, of course, with indexes in the .70s and .80s for racial segregation. He also shows no consistent pattern of increase in class segregation among blacks during the 1970s (as had been posited by many theorists of the underclass); see "Residential Segregation of Social and Economic Groups Among Blacks, 1970-1980," in Christopher Jencks and Paul Peterson, eds., The Urban Underclass (Washington, 1991), 274-98.
28. Otis Duncan and Beverly Duncan report in The Negro Population of Chicago (Chicago, 1957, Table 69) on D for employed males in major occupational groups in Chicago in 1950. They find that D between professional/managerial occupations at the upper end and laborers at the lower end of the class hierarchy runs at about .42 for nonwhites and about .33 for whites. (The white figure is probably reduced because much larger "community areas" were utilized instead of census tracts.) E. S. Uyeki in "Residential Distribution and Stratification, 1950-1960," American Journal of Sociology 69, 5 (1964), Table 4, looks at all employed males in Cleveland (with no racial breakdown) and finds a D of .52 at the extremes of the distribution. A. Wilkins examines occupational segregation for all males in eight middle-sized cities in 1950 and finds similar patterns across the occupational grid in such diverse locations as Syracuse, Memphis, and Fort Worth; see "The Residential Distribution of Occupational Groups in Eight Middle-sized American Cities," (doctoral dissertation, University of Chicago, 1956). Karl Taeuber and Alma Taeuber in Negroes in Cities (Chicago, 1965, Table C-1) provide data on occupational segregation for nonwhite males only in 10 large and middle-sized cities (excluding New York) in 1950 and 1960. Using block data, they find a range in D for the occupational extremes from .27 in Philadelphia to .47 in St. Louis in 1950. Albert Simkus offers similar data broken down for race in 10 urbanized areas (excluding New York) in 1960 and 1970; see "Residential Segregation by Occupation and Race in Ten Urbanized Area, 1950-1970," American Sociological Review 43 (February 1978), Table 3. Nathan Kantrowitz, in the only study devoted entirely to New York City (in fact, the only one we have found that even includes New York prior to 1970), examines D for income and education in 1960 for the SMSA. His results (Table 3.1) on income are virtually identical to ours for nonwhites, and only slightly different for whites (we have not examined tracts outside the NYC boundaries), with D at the extremes of the distribution running .43 for whites and .50 for nonwhites, but with even closer figures for the parts of the income range where significant numbers of nonwhites were actually present; see Ethnic and Racial Segregation in the New York Metropolis (New York, 1973). Reynolds Farley examines average levels of D by race for occupation, income, and education in the 29 largest U. S. cities and their suburbs in 1970. He finds average levels of D for income of .21 for whites in central cities and .25 for blacks; using occupation, the comparable scores are .22 for whites and .21 for blacks. In his detailed breakdown by city, he shows that New York is exactly on the mean for D by occupation in the city sample; see "Residential Segregation in Urbanized Areas of the United Sates in 1970: An Analysis of Social Class and Racial Differences," Demography 14, 4 (1977), 497-518. Finally, Farley presents the most thorough analysis of trends in the 1970s. He finds rather similar levels of class segregation using the index of dissimilarity for major categories of metropolitan areas. City by city comparisons show virtually identical patterns of spatial segregation by income between the poor and nonpoor among blacks and whites, with an average index of .29 - compared, of course, with indexes in the .70s and .80s for racial segregation. He also shows no consistent pattern of increase in class segregation among blacks during the 1970s (as had been posited by many theorists of the underclass); see "Residential Segregation of Social and Economic Groups Among Blacks, 1970-1980," in Christopher Jencks and Paul Peterson, eds., The Urban Underclass (Washington, 1991), 274-98.
29. Operationally, two households are in spatial contact if they live in the same census tract; they are isolated from each other if they do not. Keep in mind that whereas contact specifies a rather high level of geographical contiguity (most tracts in a very dense city like New York comprise only a few dozen acres), isolation tells us much less about actual distance. Even in 1950, the black population was concentrated in the Harlem (Manhattan) and Bedford-Stuyvesant quarters, separated by about 10 miles (though, significantly, lying astride the same express subway line - the A Train). At that time, two "black" tracts could well have been quite far apart. By 1970, blacks occupied a dozen or more major districts of the city, in every borough but Staten Island, and no inferences could be drawn from the noncontact of two households regarding their spatial separation. Of course, if contact between class strata is relatively high, then the whole question of isolation (and its geographic magnitude) becomes much less vexing.
30. An individual (or household or family) chosen at random in the contacting group is, in effect, the average individual in the contacting group. Thus the statistic P* reflects the distribution of individuals in the contacted group in all census tracts weighted by the distribution of individuals in the contacting group. It answers the question, what are the characteristics of those who live in the census tract of the mean member of the contacting group? Unlike D, P* is highly sensitive to the relative sizes of contacting and contacted groups, and of course, is asymmetric as a result. See Bridgette Erbe, "Race and Socioeconomic Segregation," American Sociological Review, 40 (1975), 801-12.
31. Ibid. More recently, Massey, Condron, and Denton (1987) used measures of probable contact in their important study of the city of Philadelphia in 1980; a regression model was employed by Richard Alba and John Logan in "Minority Proximity to Whites in Suburbia: An Individual-level Analysis of Segregation," American Journal of Sociology 98 (May 1993), 1388-427, and in "Locational Returns," loc. cit.
32. Ibid. More recently, Massey, Condron, and Denton (1987) used measures of probable contact in their important study of the city of Philadelphia in 1980; a regression model was employed by Richard Alba and John Logan in "Minority Proximity to Whites in Suburbia: An Individual-level Analysis of Segregation," American Journal of Sociology 98 (May 1993), 1388-427, and in "Locational Returns," loc. cit.
33. Ibid. More recently, Massey, Condron, and Denton (1987) used measures of probable contact in their important study of the city of Philadelphia in 1980; a regression model was employed by Richard Alba and John Logan in "Minority Proximity to Whites in Suburbia: An Individual-level Analysis of Segregation," American Journal of Sociology 98 (May 1993), 1388-427, and in "Locational Returns," loc. cit.
34. Data on incomes and earnings in the decennial censuses always refer to the previous year, whereas occupational and other questions refer to the current year.
35. In 1950, the city's nonwhite population stood at 756,000; in 1970, it stood at 1.8 million, and changed very little thereafter; see Norman Fainstein and Susan Fainstein, "Governing Regimes and the Political-Economy of Development in New York City, 1946-84," in John Mollenkopf, ed., Power, Culture, and Place (New York, 1988, Table 7.1). Of the nonwhite population in 1970, Mollenkopf estimates that about 1.53 million were non-Hispanic black; by 1990, this figure had reached about 1.85 million; see A Phoenix in the Ashes (Princeton, 1992), Table 3.3.
36. In 1950, the city's nonwhite population stood at 756,000; in 1970, it stood at 1.8 million, and changed very little thereafter; see Norman Fainstein and Susan Fainstein, "Governing Regimes and the Political-Economy of Development in New York City, 1946-84," in John Mollenkopf, ed., Power, Culture, and Place (New York, 1988, Table 7.1). Of the nonwhite population in 1970, Mollenkopf estimates that about 1.53 million were non-Hispanic black; by 1990, this figure had reached about 1.85 million; see A Phoenix in the Ashes (Princeton, 1992), Table 3.3.
37. See Mollenkopf, Power, Culture, and Place; John Mollenkopf and Manual Castells, Dual City (New York, 1991); and Susan Fainstein et al., Divided Cities (New York, 1992).
38. See Mollenkopf, Power, Culture, and Place; John Mollenkopf and Manual Castells, Dual City (New York, 1991); and Susan Fainstein et al., Divided Cities (New York, 1992).
39. See Mollenkopf, Power, Culture, and Place; John Mollenkopf and Manual Castells, Dual City (New York, 1991); and Susan Fainstein et al., Divided Cities (New York, 1992).
40. Our evidence is drawn from the printed report of census tract data for the 1950 U.S. Census of Population and Housing, from Summary Tape Files (STF) 3 A for the 1960 and 1970 U.S. Censuses, and from Public Use Microdata Samples (PUMS) for 1950 and 1970. (PUMS data were unavailable for metropolitan or smaller areas in 1960.) To establish the critical baseline of 1950 in measuring class segregation before the enormous expansion of the black population, we created computerized files of census tract data available only on microfilm. Doing so required encoding by hand more than 100,000 figures. We use the term "computerized tract data" in the sources for tables to indicate this constructed set. For measures of segregation the unit of analysis in each case is the census tract, a small and compact geographical unit. New York City is divided into about 2,800 census tracts, of which between 250 and 350 encompass nearly the entire black population. Tract populations average 2,700. 1950 tract data include only those nonwhites living in tracts with a nonwhite population of at least 250. PUMS data for 1950 and 1970 report on the New York Standard Metropolitan Area, rather than on New York City. Although we use the categories of "black" and "white" in all of our tables, the actual categories employed by the Census Bureau changed somewhat over the period. In 1950 and 1960, the census category was nonwhite, so these figures included other races (mainly Asians). In 1970, the category was "black." In all three censuses, some Hispanics were included in the nonwhite or black categories.
41. "The Ethnic Niche and the Structure of Opportunity," in Michael Katz, ed., The Underclass Debate: Views from History (Princeton, 1993), 161-93; the discussion of employment by industry is based on PUMS data for 1950 and 1970.
42. These trends can be seen most clearly in an industry such as professional services, encompassing health care, education, hotels, and other fast-growing sectors of the transforming city economy. Both black men and women increased their occupational status in the service sector and converged with the white pattern.
43. Model, "The Ethnic Niche," Table 5.4.
44. See Annemette Sorensen, Karl Taeuber, and Leslie Hollingsworth, Jr., "Indexes of Racial Residential Segregation for 109 Cities in the United States, 1940 to 1970," Sociological Focus 8 (1975), 125-42.
45. The tables that follow report first on D, then on P* for an average of about 300 units of analysis (census tracts) for the nonwhite or black population, and 2,400 units for the white population. (We have computed the white population for each tract in 1950 because only total population was reported in Census Bureau reports.) These units encompass roughly 200,000-450,000 nonwhite or black households over the period, and 1-2 million white households.
46. "Residential Segregation in Urbanized Areas of the United States in 1970: An Analysis of Social Class and Racial Differences," Demography 14 (4), 497-518.
47. It is important to remember, however, that P* is the mean value of a statistic that has a distribution around P*. It is possible that different phenomena may appear at the extremes of the distribution. Thus some tracts could very well have much more homogeneous income composition than the average tract. Further analysis will examine the distributional properties of P*, and in particular, the percentage of the poor living in homogeneously poor census tracts and how this changes in the period 1950-1970, as well as in the succeeding decades.
48. Remember that "isolation" is defined operationally as the probability of not being in contact with a particular group. Because contact is measured by P*, isolation is measured by 1-P*. The isolation index is a standard indicator of segregation. "Exposure" is the same as contact, and is defined here as P*. See Massey, "The Dimensions of Racial Segregation." Reference to William J. Wilson, The Declining Significance of Race (Chicago, 1978), and to Robert Weaver, The Negro Ghetto (New York, 1948).
49. Remember that "isolation" is defined operationally as the probability of not being in contact with a particular group. Because contact is measured by P*, isolation is measured by 1-P*. The isolation index is a standard indicator of segregation. "Exposure" is the same as contact, and is defined here as P*. See Massey, "The Dimensions of Racial Segregation." Reference to William J. Wilson, The Declining Significance of Race (Chicago, 1978), and to Robert Weaver, The Negro Ghetto (New York, 1948).
50. Remember that "isolation" is defined operationally as the probability of not being in contact with a particular group. Because contact is measured by P*, isolation is measured by 1-P*. The isolation index is a standard indicator of segregation. "Exposure" is the same as contact, and is defined here as P*. See Massey, "The Dimensions of Racial Segregation." Reference to William J. Wilson, The Declining Significance of Race (Chicago, 1978), and to Robert Weaver, The Negro Ghetto (New York, 1948).