-
Notes
- Russell Sage Foundation
- Chapter
- Additional Information
Notes Chapter 1 1. For both surveys, we used census data to assist in the selection of a random sample of households and then, within each household, the interviewer randomly selected an adult at the doorstep and interviewed him or her for about one hour. Chapter 3 l. Determining how many jobs are located in a metropolis is challenging , even though numerous figures are published by various statistical agencies. For the most part, we rely on censuses and household surveys, since they include questions asking large samples whether they are working. We assume that those enumerated in the metropolis also work in the metropolis. This is not entirely accurate, since some of the jobs reported by those living in metropolitan Detroit are located in Windsor, Flint, or Toledo. The census of 1990 allows an analysis of these flows: Residents of the three-county metropolis whose reported place of work was in the three counties (Macomb, Oakland and Wayne Counties) 1,655,106 Residents of the three-county metropolis whose reported place of work was outside the three counties 60,444 Persons who lived elsewhere in the United States, but whose reported place of work was in the three counties 111,931 Net difference +51,487 Sources: U.S. Department of Commerce 1993. 2. To look at recent changes, we use a different data source. Most employers report how many workers are on their payroll to several DETROIT DIVIDED federal agencies for a variety of purposes, such as forwarding their employees' Social Security payments and income taxes to the right agency. Since 1972, the Bureau of Labor Statistics has tabulated the number of workers reported by employers for a six-county Detroit metropolitan area. In addition to the Macomb, Oakland, and Wayne county metropolis used in this book, it includes Lapeer, Monroe, and St. Clair counties. In 1990, 93.5 percent of employment in the six-county area was within Macomb, Oakland, and Wayne counties. 3. General Motors' Poletown plant also occupies land in Hamtramckan incorporated suburb completely surrounded by the city of Detroit . 4. The occupational coding system used by the Bureau of the Census for 1980 and later differs substantially from that used previously. For long-term trends in this chapter, the occupational data were obtained from the Integrated Public Use Microdata Sample created by Steven Ruggles at the University of Minnesota. Occupations reported in all censuses were coded, insofar as possible, into categories used in the census of 1950. (For additional information, see Ruggles and Sobek 1997.) 5. To describe racial and gender differences in employment, occupational achievement and earnings data were analyzed from Public Use Microdata Samples of decennial censuses from 1940 through 1990. Unfortunately, the Public Use Sample from the enumeration of 1960 does not identify either metropolitan Detroit or the city of Detroit. 6. The estimates of net racial differences in hours of employment come from ordinary least square regression models that took annual hours of work as their dependent variable. The independent variables included dichotomous variables identifying eight age groups, dichotomous variables identifying four educational attainment categories , one variable distinguishing married-spouse-present people from others, and one variable distinguishing blacks from all other races. Separate models were fit for each year and each gender. 7. Indexes of dissimilarity assess how differently blacks and whites were arrayed across the broad occupational categories used in this analysis. If blacks and whites had identical occupational distributions , this index would take on its minimum value of O. Were there total racial segregation of occupations, the index would take on its maximum value of 100. Declines in the values of the index of dissimilarity , shown here, reveal that blacks and whites increasingly work at similar jobs. The gap has narrowed much more among women than among men. 268 [44.222.161.54] Project MUSE (2024-03-29 05:56 GMT) 1940 1950 1960 1970 1980 1990 Men 38 38 37 34 25 24 Source: Ruggles and Sobek 1997. Index of Occupational Dissimilarity Women 47 45 39 20 13 9 NOTES 8. These estimates of the net effects of race were derived from ordinary least squares regression models that used occupational prestige scores as dependent variables and the array of independent variables listed in note 6. Separate models were run for each year and for each gender. 9. To determine the net effects of race on earnings, the log of earnings was used as the dependent variable. Independent variables are those listed in note 6. A dichotomous variable also...