Detroit Divided

5. The Detroit Labor Market: The Workers' Perspective

5 The Detroit Labor Market: The Workers' Perspective THIS CHAPTER focuses on the extent to which the continuing concentration of the African American population in Detroit, while jobs relocate to the suburbs, contributes to the deteriorating employment outcomes of black residents in the city. How does the location of one's residence influence the location of one's job search? Do blacks who seek and find suburban jobs have relatively better employment outcomes than those working in the city? If so, why don't African Americans search more frequently or more intensely for suburban work? Answers to these questions help explain why a spatial mismatch between black central-city residents and suburban jobs may persist over time, despite strong economic incentives that should induce African Americans to seek and find suburban jobs, even if they do not live there. We also investigate several other issues related to searching for employment . For instance, Lawrence Mead (1992) contends that low black employment rates are largely due to a lack of commitment to work. In his view, unemployment and nonparticipation in the labor force may be high among African Americans because • their expectations regarding acceptable wages are higher than the wage offers they obtain (Holzer 1986); • relatively high incomes and status can be gained from their participation in criminal activity; and • African Americans may be "discouraged" or "defeated" because they do not expect a job search to yield job offers (Mead 1992). The Detroit Area Survey contains extensive data on the locations of both jobs and residences for whites and blacks, and also data on the search patterns of prospective workers, which provide important and innovative information about these issues. 126 THE DETROIT LABOR MARKET: THE WORKERS' PERSPECTIVE TABLE 5.1 Locations of Residence and Workplace of Detroit-Area Workers, 1990 Percentage of Employed White Residents Black Residents Persons Who Work in City Suburbs Total City City of Detroit 47 12 14 59 Three-county suburban ring 51 88 86 39 Outside three-county metropolis 2 < 1 < 1 2 Total 100 100 100 100 Number of employed persons (thousands) 82 1,303 1,384 234 Source: U.S. Department of Commerce 1993. Residential and Workplace Locations of Detroit Residents Suburbs Total 22 52 73 46 5 2 100 100 61 295 We begin by comparing the tendencies of whites and blacks to live in the city of Detroit with their tendencies to work there, and how these two factors interact. Then we compare the characteristics of jobs and workers in each location, and how they differ by race and gender. Table 5.1 shows place of residence by place of work for whites and blacks in 1990.1 Residential and work patterns are highly correlated for each racial group-in other words, city residents are much more likely to work in the city than are suburban residents and vice versa. Among whites living in the city, about half held a job in the city and one-half in the ring, but among whites who lived in the ring, only one in eight worked in Detroit. The overwhelming majority of whites and nearly half of African Americans worked in the ring, reflecting the long-term trend toward the suburbanization of employment. Differences in residential location account for a substantial part of the greater tendency of blacks to work in the city. However, within each residential area, blacks are more likely than whites to work in the city. For example, among city residents, 47 percent of whites but 59 percent of blacks held city jobs. These results are consistent with evidence presented earlier that fewer African Americans work in the suburbs, partly because employers there receive fewer applications from them, and partly because these firms hire a smaller percentage of black applicants. 127 DETROIT DIVIDED About half of the central-city residents engage in reverse commuting to suburban jobs. But relatively small fractions of suburbanites commute to the city: 12 percent of employed whites and 22 percent of blacks. These data confirm not only the declining fraction of metropolitan jobs located in city but also the much greater dependence of blacks than whites on this dwindling pool of jobs. Table 5.2 presents information for employed adults and for each racial group separately, about the characteristics of workers and their jobs, classified by job location. We consider weekly and hourly earnings, the educational attainment of workers, their occupational and industrial distributions, and two characteristics of commuting: the percent who use their cars and vans to get to work and the mean commuting time. These data confirm several findings regarding characteristics of city and suburban jobs noted in the previous chapter. For instance, weekly and hourly wages were higher in the city. Educational attainment among employees is generally higher for those who work in the city, especially when we consider those who have graduate degrees; and the fraction of workers in professional or managerial occupations is considerably higher as well. Clerical and sales jobs made up a larger share of the suburban than the city workforce (32 versus 16 percent), whereas the concentration of blue-collar jobs there is a bit higher as well. Finally, the share of total jobs in manufacturing was also higher in the suburbs than in the city (29 versus 16 percent), while those in the services are concentrated in the city (51 versus 33 percent). Among service occupations and service industries , the story is very different. The service occupations tend to have few educational requirements and generally pay little, since they include jobs such as waiter, counter help, cook, bartender, and building cleaner. Service industries, on the other hand, often demand a highly trained workforce, since this sector includes hospitals, clinics, law offices , educational institutions, and consulting firms. City-suburban differences in worker characteristics vary dramatically by race. Earnings are typically higher for African Americans employed in the suburbs than for those working in the city. For instance, blacks working in suburban jobs earned $855 per week, whereas those working in the city earned only $723 per week (in 1997 dollars). Similar results can be found with respect to educational attainment of black workers and the occupations in which they are employed; that is, those working in the suburbs have higher levels of education and are more likely to work in professional or managerial occupations than are central -city workers. The opposite, however, is true for whites: central-city workers typ128 THE DETROIT LABOR MARKET: THE WORKERS' PERSPECTIVE TABLE 5.2 Characteristics of Employed Persons Aged Twenty-Five to Sixty-Four in 1990, by Race and Workplace Location Employed in City of Detroit Employed in Suburban Ring Characteristics of Jobs Total Whites Blacks Total Whites Blacks Mean wages in 1997 dollars Hourly $19.31 $20.97 $16.85 $19.44 $19.65 $17.22 Weekly $826.00 $919.00 $682.00 $844.00 $855.00 $723.00 Educational attainment High school dropout 12% 11 % 14% 11% 9% 19% High school diploma 51 47 57 55 56 58 Associate's degree 8 8 9 9 9 9 Bachelor's degree 16 20 11 16 17 10 Graduate degree 13 14 9 9 9 4 Total 100% 100% 100% 100% 100% 100% Occupational distribution Professional, managerial and technical 37% 42% 29% 34% 35% 22% Clerical 17 16 23 16 16 19 Sales 8 8 4 12 12 7 Crafts 11 13 7 13 14 9 Operatives 12 11 13 12 11 24 Laborers 3 2 3 3 2 4 Service occupations 13 8 20 10 9 15 Total 100% 100% 100% 100% 100% 100% Industrial distribution Construction 4% 5% 3% 5% 5% 2 Manufacturing, durable goods 15 18 11 25 24 32 Manufacturing, nondurable goods 5 6 3 4 4 3 Transportation, communication 9 9 10 6 6 9 Wholesale trade 4 5 2 5 5 4 Retail trade 8 9 7 14 15 12 Finance, insurance, and real estate 7 7 7 6 6 7 Service industries 39 34 44 30 31 26 Public-sector employment 9 6 13 3 3 4 Total 100% 100% 100% 100% 100% 100% Journey to work Uses own car or van 89% 94% 83% 96% 96% 93% Mean one-way commute time (minutesJ 28 31 24 25 24 26 Source: U.S. Department of Commerce 1993. 129 DETROIT DIVIDED ically have higher earnings, educational levels, and occupational status than whites who work in the suburbs. Thus, gaps in earnings between whites and blacks employed in the city were quite high, while the earnings of blacks working in the ring are more nearly comparable to those of whites employed there. For example, black/white weekly earnings ratio is .74 in the city and .85 in the suburbs. The pattern of higher earnings for whites working in the city, but higher earnings for African Americans working in the suburbs, has been documented previously.2 Data on relative educational attainment, occupations , and industries of employment in table 5.2 suggest four possible causal forces at work: (1) Because most African Americans live in the city and most whites live in the suburbs, each group must be compensated for time spent commuting to the more distant location. Thus, white suburban residents earn a premium for working in the city and black centralcity residents earn one for working in the suburbs; (2) Because African Americans are "crowded" into the city by discriminatory suburban housing and labor market barriers, there is an excess supply of central-city black labor relative to available jobs, which depresses wages. In contrast, an excess demand for whites relative to available jobs for them in the city raises their wages (Hirsch and MacPherson 1994); (3) The quality of black workers who either live in or commute to work in the suburbs is higher than that of those who stay and work in the central city, whereas the opposite is true for whites; (4) The quality of jobs to which blacks have access in the suburbs exceeds the quality of those available to them in the city. There is evidence in our data supporting each hypothesis. For instance , earnings rise with commute times for whites but not for blacks.3 For each additional ten minutes of commuting time (in each direction), wages for whites increased by 5 percent per hour, whereas for blacks they remained essentially unchanged.4 Indeed, the average commute time of central-city black workers was only seven minutes shorter than that of whites who worked there, even though the earnings of the former are much lower than those of the latter. The concentration of African Americans in central-city workplaces certainly suggests "crowding" there, thus contributing to lower wages. The fact that automobile transportation is highly correlated with the likelihood of blacks being employed in the suburbs suggests that this crowding is linked to differential access to suburban jobs. The simul130 THE DETROIT LABOR MARKET: THE WORKERS' PERSPECTIVE taneous effects of housing segregation and labor market discrimination are thus reinforced by transportation factors.5 The relatively high educational levels of black suburban workers reflect two factors: • the higher education levels of black suburban residents relative to central-city residentsj and • the fact that, among central-city residents, higher-skilled blacks are more likely to obtain jobs in the suburbs either because they "selfselect " into such jobs or because suburban employers are more likely to hire them. Our data support both interpretations. Census data report that central -city whites had less educational attainment than white suburbanites did.6 The higher education attainment of whites who work in the city is driven entirely by the higher attainments of those commuting from the suburbs relative to those working there. Finally, differences in the characteristics of city and suburban jobs appear to contribute to the observed outcomes in table 5.2 as well. As noted, manufacturing and/or blue-collar jobs are more concentrated in the suburbs than in the city, as are clerical and sales jobsj whereas jobs requiring at least a college diploma are more concentrated in the city, as are low-wage jobs in the service occupations. Whites capture the vast majority of jobs at the top end of the educational and occupational distributions in the city. And among whites who work in the city, more than four in ten have managerial, professional , or technical jobs. Similar percentages of black and white suburban workers are employed in mid-level white-collar and blue-collar jobs and in manufacturing jobs. But a greater percentage of metro-area whites than blacks work in the suburbs, thus allowing them to capture more of these types of jobs as well. In contrast, African Americans are overrepresented among the lower-wage blue-collar and service occupations-jobs that remain in the city. All four of these factors contribute to the geographic pattern of racial earnings differentials observed for blacks and whites in table 5.2, and this finding is supported by regression analyses. The characteristics of workers and jobs account for some, but not all, of these geographic earnings differentials by race. In regression equations that have weekly earnings as the dependent variable, the inclusion of controls for worker education, occupation and industry reduces the positive coefficient on central-city workplace location for whites, but it does not totally disappear . Net of other characteristics determining wages, whites working in the city earn more than whites working in the ring. The same is true for 131 DETROIT DIVIDED the negative coefficient on city workplace location for African Americans / The higher education and skill levels of suburban whites enable them to obtain most of the high-paying jobs in the city, even though few live there. Highly educated people are generally well informed about employment opportunities throughout the entire metropolis and have little difficulty commuting to good jobs. In 1990, 94 percent of white suburban households reported ownership of at least one car or van, while two-thirds owned two or more. Whites can choose both where to live and where to work in ways that maximize their overall incomes and their personal preferences with regard to housing amenities. In contrast, less-educated black city residents have difficulty gaining access to the better-paying jobs requiring less education, which are more heavily concentrated in the suburbs, than do less-educated whites who live in those suburbs. This differential access appears to be due to "space"-that is, the transportation and informational problems associated with their residential concentration in the central city. In 1990,35 percent of the city's black households reported no car or van. It is also due to "race"-that is, apparent discrimination in both the housing and labor markets.8 Even the skills and employability of central-city African Americans are limited by residential segregation, which reduces the quality of their schooling and increases their"social isolation."9 Thus, the housing patterns and racial attitudes of Detroit area residents limit the labor market opportunities of less-educated blacks in many ways.10 Given the relative paucity of good central-city jobs for less-educated workers, and the costs and barriers they face in gaining suburban employment , it is important to analyze how frequently African Americans look for work in the suburban labor market. Table 5.3 presents information about whether whites and blacks have ever searched for work in each of five local areas within the metropolis. Job search questions were asked of all respondents who had searched at any time in the ten years prior to the 1992 Detroit Area Study. Results are presented for all white and black job seekers, as well as for subgroups of each-younger (below age thirty-five) versus older workers, those living in Detroit versus those living elsewhere, those living in each of the particular cities being considered , and those with and without access to cars during their search. The key finding is that African Americans are much more likely than whites to search in the city of Detroit, and are more likely to search in the city than anywhere else. About three-quarters of blacks but only 40 percent of whites had searched in Detroit. The differences hold within all subgroups of whites and blacks as well, and are even more pronounced among the young. Only among the small fraction of 132 TABLE 5.3 Percentage of Workers Who Have Searched in Each Area During Previous Ten Years Not Living Living in in Living in Living in Specific Specific Has Access No Access Access Total Young Old Detroit Suburbs Location Location to Car to Car Whites Detroit 42 39 45 72 39 72 39 44 53 Southfield 40 38 42 48 39 100 39 43 30 Troy 43 48 38 35 44 84 41 47 22 Warren 34 32 35 40 33 94 31 37 30 Downriver suburbs 24 26 23 30 24 84 20 27 17 Sample size 398 186 212 46 3S2 299 8 Blacks Detroit 73 83 62 77 S9 77 59 74 7S Southfield 56 66 46 54 67 75 55 65 43 Troy 31 34 27 31 33 31 39 13 Warren 30 35 24 28 36 30 38 15 Downriver suburbs 29 25 34 26 40 29 33 22 Sample Size 377 190 185 339 36 223 92 Source: University of Michigan, Detroit Area Study 1992. Note: These data refer to the residents of the three-county metropolis and to persons who searched for a job within the last decade. DETROIT DIVIDED whites who actually live in Detroit was the tendency to search there about the same as for blacks. City of residence clearly influences where people look for work. African Americans were somewhat more likely than whites to have searched in Southfield and the Downriver suburbs, but somewhat less likely to have searched in Warren and especially Troy. Over half of blacks searched in Southfield, whereas less than a third ever searched in the other three suburban locations. Among younger blacks, the tendency to search in the Downriver area, where there are manufacturing plants, was somewhat lower than it was for older blacks, whereas their tendency to search in Southfield, Troy, and Warren was greater. For those living in the suburban ring, residence also affected job search. Thus, for both blacks and whites, residents in each specific location were far more likely to search where they live than are those living elsewhere. All of the entries in column 6 are above 80 percent for whites, except for Detroit, at 72 percent. About three-fourths of blacks living in Detroit or Southfield have searched where they live. White residents of the city of Detroit differed from African Americans who lived there in an important way: they were more likely to have looked for work in the outer suburbs. Finally, access to cars during the search period raised the tendency of African Americans to seek employment in the suburbs. However, there is a puzzling finding: even blacks with access to cars had search probabilities under 40 percent in Troy, Warren, and the Downriver area.l1 Why job seekers search in some areas but not others may involve factors beyond access to cars and place of residence. Respondents who reported not having searched in a particular location were asked why not, and their responses were coded into several categories. The predominant reason (46 percent) given for not searching in any area was travel distance or transportation problems. This is consistent with our demonstration that city of residence strongly predicts where people search. About 14 percent of respondents stated that they just did not bother to search in an area, which suggests to us some combination of a lack of information and distance. Ten percent of the answers stated that the jobs in an area or the employers there were not appropriate for the skills of the respondent. About 9 percent of the reasons indicated that the respondent believed no jobs were available there. Four percent of the answers stated that no search was carried out because of "crime in the area" or because the person would not feel safe there. These were primarily suburban whites who refused to look for a job in the city. That is, 16 percent of the whites who did not look for work in Detroit gave a fear of crime as their reason. The tendency to mention travel and distance as the predominant 134 THE DETROIT LABOR MARKET: THE WORKERS' PERSPECTIVE reasons for not searching does not explain why much larger fractions of blacks sought work in Southfield than in Warren or the Downriver suburbs . The latter are geographically close to black neighborhoods of Detroit . More African Americans live in Southfield than in the other areas, but it is a small fraction of all metropolitan blacks-only 2.4 percentwho live there. Residential differences thus cannot fully explain the geographic search differential among blacks. Personal networks and contacts between African Americans living in Southfield and those living in Detroit probably account for some of the tendency of blacks to seek employment in Southfield. It is also probable that African Americans feel more welcome in some areas than in others, perhaps due to the great reputations of these areas, as well as their own experiences there. People are likely to be more comfortable searching for work in areas that are perceived as being friendlier, and may seek jobs in areas viewed as hostile only if they have an extremely promising lead. Table 5.4 shows how white and black respondents in the household survey thought white residents in five specific suburbs would react if blacks moved there. Because the question refers to residence rather than employment, it is not a perfect measure of attitudes toward the latter. Presumably, whites will be much less hostile to blacks who work in their area and live elsewhere than to blacks who move into their neighborhoods . In any case, it is African American perceptions that affect their job search, and they may be less likely to distinguish between attitudes toward residence or employment. The areas about which the questions were asked overlap somewhat with those from the search questions , though Dearborn and Taylor are included in these attitude questions rather than Detroit and Downriver. African Americans-and whites-perceive Southfield as being much more welcoming toward blacks than are the other areas. For example, about 70 percent of blacks thought of Southfield as a place where they would be welcome, whereas only about one-third perceived Troy or Taylor as such a place. African Americans saw Warren-renowned as the home of "Reagan Democrats" (Greenberg 1995; chap. 2)-and Dearborn -with its legendary Mayor Orville Hubbard (Good 1989)-as hostile areas. Whites think that blacks would generally receive a warmer welcome in these places but, when looking at all five suburbs, we find that both races pretty much share the same cognitive map about where black movers would be welcomed or would face hostility. Data in the lower panel of Table 5.4 classify the search efforts of blacks by how they perceive three suburbs. They indicate that African Americans are more likely to search in particular areas if they personally perceive the areas as being friendly to blacks.12 Thinking that the 135 DETROIT DIVIDED TABLE 5.4 Perceptions of Friendliness to Blacks and Tendencies for Blacks to Search in Area Perception of Whether Current Residents Would Welcome or Not Welcome Blacks Moving into Their Area White Respondents' Perception Black Respondents' Perception Not Not Welcome Welcome Indifferent Welcome Welcome Indifferent Southfield 59% 26% 8% 69% 16% Troy 27 60 7 32 51 Warren 34 48 9 16 73 Dearborn 29 58 8 9 86 Taylor 42 36 10 38 40 Percentage of Blacks Who Have Ever Searched in the Area by Their Perceptions of Friendliness of Whites There Southfield Troy Warren How Blacks Perceive Their Reception Not Welcome Welcome Indifferent 61% 39 32 48% 33 31 47% 28 25 Source: University of Michigan, Detroit Area Study 1992. 12% 7 3 2 13 Note: Sample sizes for respondents in the upper panel are 539 and 496 for whites and blacks, respectively. Sample size for the lower panel is 360. The percentages in the upper panel do not sum to 100% because some respondents said they did not know what kind of reception blacks would receive. whites who live in a suburb are hostile to African Americans discourages them from looking for work there. But there are other factors. Blacks are more likely to search in Southfield than in Troy or Warren, even when they perceive similar friendliness among whites in these suburbs. The greater concentrations of black residences and social contacts in Southfield probably account for this. Nevertheless, the differences in perceived friendliness contribute to these differing search patterns , perhaps in ways that are understated by these data.13 Thus, the subjective impressions that African Americans have about friendliness or hostility toward them in Detroit suburbs may influence where they seek employment, and therefore whether they obtain jobs. The lower tendency of black applicants to be hired in the suburbs, despite the apparently higher relative skills of those who search there, also suggests that these perceptions might be based on actual experiences of the respondents or their friends and family when seeking employment. 136 THE DETROIT LABOR MARKET: THE WORKERS' PERSPECTIVE The racial attitudes of whites and African Americans, the high degree of racial residential segregation, and the absence of good public transit clearly affect where white and black people search for work. These results are strongly consistent with the spatial mismatch hypothesis , and give us more insight into why these effects tend to persist over time.14 Other Characteristics of the Job Search The geographic dimension of job search is one of many along which whites and blacks might differ. It is possible, for instance, that blacks are choosier than whites with regard to the jobs that are available to them; or that blacks do not search as thoroughly as do whites; or that they use search methods that are not as productive; or that they have less accurate perceptions about employers' hiring requirements and preferences than whites. These hypotheses have been stated forcefully by Lawrence Mead (1992), and his arguments have been frequently repeated as a major explanation for why blacks fare poorly in urban labor markets. He contends that blacks with modest educations often do not seek or accept available jobs, albeit they might be jobs with low wages. He attributes this to "defeatism" or discouragement.IS Data from the household Detroit Area Study allow us to test his ideas, since they provide information about "reservation wages," defined as the minimum wage a job seeker requires for accepting employment . This household survey also inquired about search effort, search methods, tendencies to contact employers and receive offers, and general perceptions of labor market characteristics and employer behavior. We analyze these data by race and sometimes by age as well, as these questions frequently center around the behavior and motivation, or lack of motivation, of younger blacks. We also emphasize search behavior among the nonemployed, as their current or most recent behaviors and attitudes are most relevant to their current employment status. Information about the job search efforts and search outcomes of whites and blacks is presented in table 5.5. Only nonemployed persons who had searched for work in the last thirty days are included. This table shows the fraction of nonemployed persons who looked for work, the techniques they used, the average number of methods used, hours per week spent searching, access to cars during the period of search, number of employers contacted, and whether an offer had been received.16 Several findings emerge. First, nonemployed African Americans were much more likely than nonemployed whites to be searching for 137 DETROIT DIVIDED TABLE 5.5 Characteristics of Tob Search for Nonemployed Persons in Detroit Area Study Who Searched for Employment in the Thirty Days Prior to Their Interview Whites Blacks Percentage of non-currently employed persons who searched in last thirty days 13 22 Sample size 592 532 Among those who have searched, percentage who used specific means of search: Talked to relatives and friends 74 79 Read newspaper ads 96 84 Checked with labor unions 11 5 State employment agency 17 42 School placement officers 13 20 Looked for "help wanted" signs 29 53 Mean number of search methods used 2.86 3.04 Hours spent searching per week 8.17 9.81 Had access to car during search 98% 69% Employers contacted per week Mean 2.01 l.53 Median 0.83 0.50 Job offers received per week 0.09 0.06 Sample size 77 115 Source: University of Michigan, Detroit Area Study 1992. work (22 versus 13 percent). This is true even when we limit the samples to the young. Both races reported similar methods, although blacks were more likely than whites to have looked for Help Wanted signs and to have had recent contact with the Michigan State Employment Service Commission. African American searchers spent more time and used more methods of search than did white searchers. These results strongly imply that the longer durations of search and nonemployment that have been observed among blacks cannot be attributed to lack of effort. They also cast considerable doubt on the notion that African Americans lack motivation or display a "defeatist" attitude toward work, as Mead suggests.I? These data also show that almost all white searchers have access to automobiles, where almost one-third of African Americans do not. This may be why white searchers contact many more employers (means of 2.01 versus 1.53 per week, respectively) and receive more offers (means of .09 versus .06 per week, respectively) than do black searchers. Job 138 THE DETROIT LABOR MARKET: THE WORKERS' PERSPECTIVE offers received per hour of search are almost twice as high for white job seekers as for blacks (.Oll versus .006). On average, it took a white job seeker 91 hours of search time to generate an offer; it took an African American 167 hours. Thus, the rates at which employers make job offers to white and black job seekers have large effects on the search and nonemployment durations of each group. To what extent are the difficulties of black job seekers caused by their greater unwillingness to accept available low-wage jobs? The ratio of reservation wages among any group of workers to the wage offers they receive would be the most informative measure of that group's willingness to accept jobs that are available to them. But since we cannot observe potential wage offers from employers, we go by the previously received wages of workers in the group.IS Data on reservation wages and received wages among all nonemployed white and black searchers and among the young (under age thirty-five) appear in table 5.6. Several summary measures of the reservation and received wage distributions-the means and medians of each distribution-are shown, as well as percentage of the relevant sample willing to accept a job with a wage below certain low-income benchmarks : $4, $5, or $6 per hour. There is modest support for the notion of relatively high reservation wages among young blacks. In particular, we find that mean and median reservation wages for nonemployed African Americans are somewhat higher, relative to their received wages, than are the comparable numbers for whites, especially among the young. The ratio of mean reservation to received wages is comparable for all nonemployed blacks and whites (.81), but is higher among young blacks than whites (.95 versus .87). Ratios of median reservation to median received wages are higher among blacks than whites in both groups (.79 versus .67 among all the nonemployed, and 1.06 versus .75 among the young). Reservation wages are higher for young whites than blacks in absolute terms, whereas the received wages of whites exceed those of blacks by considerable amounts (with respective medians of $8.00 versus $5.64 among the young). These findings are consistent with Harry J. Holzer (1986) and Stephen Mark Petterson (1997), who also find relatively high reservation wages among young black males using a nationwide survey of data from the 1980s.19 On the other hand, the magnitudes of these differences in relative reservation wages-that is, reservation wages relative to received wages-are generally on the order of 10 to 30 percent. These differences are not sufficient to account for large parts of the black-white differences in labor force participation and unemployment described in chapter 3.20 Furthermore, considerably higher fractions of young black males than whites are willing to accept very low wage employment. Forty-two 139 DETROIT DIVIDED TABLE 5.6 Received Wages and Reservation Wages of Young Nonemployed Workers, by Race and Age (in 1992 Dollars) Whites All nonemployed workers Previously received wages Mean $10.32 Median $8.98 % ~ 4.00 4 % ~ 5.00 18 % ~ 6.00 26 Reservation wages Mean $8.44 Median $6.00 % ~ 4.00 17 % ~ 5.00 34 % ~ 6.00 52 Sample Size 60 Young nonemployed workers Previously received wages Mean $9.23 Median $8.00 % ~ 4.00 8 % ~ 5.00 26 % ~ 6.00 33 Reservation wages Mean $8.01 Median $6.00 % ~ 4.00 19 % ~ 5.00 27 % ~ 6.00 52 Sample Size 31 Blacks $8.47 $6.33 17 43 50 $6.87 $5.00 29 53 68 134 $8.04 $5.64 18 46 55 $7.65 $6.00 20 42 63 65 Source: Author's tabulations of the Multi·City Study of Urban Inequality Employer Sur· vey. Note: A "young" worker is defined in this chapter as being under the age of thirty·five. percent of Detroit's young nonemployed African Americans were willing to work for less than $5.00 per hour (in 1992 dollars), in contrast to only 27 percent of similar whites. An important caveat is in order. It seems probable that the DAS household survey underrepresents those men who have very high reservation wages relative to received wages, including those who have or anticipate high earnings from illegal activity. Presumably it is difficult to include such respondents in a survey.21 On the other hand, the racial 140 THE DETROIT LABOR MARKET: THE WORKERS' PERSPECTIVE differences in employment outcomes among those who were interviewed are very large, and differences in their reported reservation wages cannot account for much, if any, of the large observed differences in employment. Thus, we find, at best, limited support for the hypothesis that young African Americans are relatively less likely than whites to accept low-wage employment available to them.22 In table 5.7 we consider additional measures of job search perceptions among those under age thirty-five in the upper panel, and among those young persons who were not employed at the time of the survey in the lower panel. These data present the fractions of each group who perceive that particular characteristics are very or somewhat important to employers in deciding whom to hire. Young African Americans consider "hard" skills to be important in the job search process, at least as important as do young whites. On the other hand, young nonemployed blacks differed from whites in judging the emphasis that employers place on "soft" skills. With regard to how someone looks and dresses, 44 percent of young nonemployed whites, but only 28 percent of African Americans, said this was important to employers. With regard to being a team player, it was 56 percent for whites, 37 percent for blacks. There was also a racial difference, although small, with regard to how employers rate the demographic characteristics of applicants: African Americans, more so than whites, thought employers took these characteristics into consideration when deciding whom to put on the payroll. While young blacks realize the importance of education, experience, and a good command of English, they may underestimate the importance to employers of soft skills, such as attitudes and communication. This has been stressed by Philip Moss and Chris Tilly (1995), who find that employers claiming to put the most emphasis on soft skills in the hiring process are less likely to employ African Americans than employers who chiefly emphasize hard skills. However, it is unclear from their data to what extent this reflects real racial differences in soft skills, as opposed to discriminatory employer perceptions.23 Conclusion Why do African Americans, especially men, fare so poorly in Detroit's labor market? In these chapters we tested the most frequent explanations , and we can now draw unambiguous conclusions: • We reject the idea that African Americans underestimate the importance of skills such as educational credentials and work experience. They emphasize them more than whites. 141 DETROIT DIVIDED TABLE 5.7 Perceptions of What Employers Emphasize When Hiring, Persons in Metropolitan Detroit Under Age Thirty-Five White Black Racial Difference Respondents Respondents (Black Minus White) Total young sample IIHard" skills Specific experience in your line of work 54% 53% -1 Formal education 39 50 11 References 26 31 5 Ability to speak English well 51 53 2 IISoft" skills How someone looks and dresses 33 34 1 Being a team player 62 51 -11 Demographic characteristics of job candidate Where someone lives 2 13 11 Age 2 8 6 Race 3 4 1 Gender 5 6 1 Sample size 184 190 Currently nonemployed young sample IIHard" skills Specific experience in your line of work 49% 61 % 12 Formal education 24 53 29 References 28 26 -2 Ability to speak English well 63 63 0 IISoft" skills How someone looks and dresses 44 28 -16 Being a team player 56 37 -19 Demographic characteristics of job candidate Where someone lives 4 10 6 142 THE DETROIT LABOR MARKET: THE WORKERS' PERSPECTIVE TABLE 5.7 Age Race Gender Sample size Continued White Respondents 3 3 3 33 Black Respondents 11 8 4 65 Source: University of Michigan, Detroit Area Study 1992. Racial Difference (Black Minus White) 8 5 1 • We reject the idea that African American do not search much for jobs or that they consistently use the least effective search methods. Black job seekers spend more time looking than whites. • The evidence casts some doubt on the idea that high unemployment among African Americans is due to unrealistically high estimates of what they are worth in the labor market. Indeed, more nonemployed blacks than whites say they will work for the minimum wage or less. • There is some evidence that blacks may underestimate the importance of soft skills in the job search process. • There is some evidence that blacks may not search frequently in the outlying suburbs where jobs may be available. • There is strong evidence of a skills mismatch. Inner-city employersand it is in the inner city where blacks concentrate their job searchreport many openings, and the long duration of these vacancies suggests that applicants do not meet their standards. • There is evidence of a reluctance of suburban employers to hire African American applicants. Although black applicants for suburban jobs seem somewhat more qualified than black applicants for city jobs, they are turned down more frequently, strongly suggesting an employer preference for white workers. • The simultaneous effects of housing market and labor market discrimination seem considerable. Blacks remain in the city while the jobs for which they are most qualified-manufacturing-sector jobssuburbanized . The city retained employment in the industrial sectors where the skills deficits of African Americans make it difficult for most of them to be employed-the professional services sector. Given that both black and white job seekers concentrate their search fairly close to home, blacks are at a disadvantage. And whites are advantaged, since they retain a great share of the best jobs in the city and dominate jobs in all occupational ranks in the suburbs. 143 ...