3
Disparities in Outcomes

This chapter examines the state of existing knowledge on the disparities in outcomes experienced by individuals and jurisdictions in metropolitan areas. For individuals and groups, these disparities may be grouped into three categories: income disparities (median family income levels, per capita income levels, and poverty rates); labor market disparities (employment and unemployment rates for working-age men); and educational disparities (high school and college completion rates). These simple measures do not, of course, fully capture the differences among groups in metropolitan areas. Moreover, they reflect outcomes that result from many factors, not simply from inequalities in the metropolitan opportunity structure.

We then go on to consider the extent to which these disparities are caused by differences in the metropolitan opportunity structure; we also consider the spatial components of causes less directly related to metropolitan phenomena. To give an indication of the costs to individuals and groups of these disparities, the committee reports the work of researcher Harry Holzer who calculated at our request the costs of unequal opportunity borne by those most directly affected, as well as by the entire metropolitan area.

Finally, we shift to consider disparities in fiscal capacities among local governments, in terms of taxes and services.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 40
3 Disparities in Outcomes This chapter examines the state of existing knowledge on the disparities in outcomes experienced by individuals and jurisdictions in metropolitan areas. For individuals and groups, these disparities may be grouped into three categories: income disparities (median family income levels, per capita income levels, and poverty rates); labor market disparities (employment and unemployment rates for working-age men); and educational disparities (high school and college completion rates). These simple measures do not, of course, fully capture the differences among groups in metropolitan areas. Moreover, they reflect outcomes that result from many factors, not simply from inequalities in the metropolitan opportunity structure. We then go on to consider the extent to which these disparities are caused by differences in the metropolitan opportunity structure; we also consider the spatial components of causes less directly related to metropolitan phenomena. To give an indication of the costs to individuals and groups of these disparities, the committee reports the work of researcher Harry Holzer who calculated at our request the costs of unequal opportunity borne by those most directly affected, as well as by the entire metropolitan area. Finally, we shift to consider disparities in fiscal capacities among local governments, in terms of taxes and services.

OCR for page 40
Disparities Among Individuals Central-City and Suburban Residents Analyses of intrametropolitan differences in economic and social conditions are nearly always presented in terms of central-city and suburban differences, even though there is clearly substantial variation across individual suburbs (Danielson and Wolpert, 1992). In many metropolitan areas there are inner-ring suburbs and industrial suburbs whose residents share the same environment and consequent problems as residents of the central city. However, because of the way data are presented, virtually all of the research examines only central-city and suburban differences; for that reason, most of the data and the discussion presented in this chapter focus on the central-city/suburban distinction. In 1990, central-city residents had incomes that were considerably lower than those of suburban residents (Table 3-1). The median household income of urban residents was only 74 percent of that of their suburban counterparts, their per capita income was 84 percent of suburban per capita income, and their poverty rate was over twice as high. Disparities in employment and education were somewhat smaller, but central-city residents appear to be more deprived in these realms as well. In particular, compared with suburban communities, the unemployment rate of men of prime working age was 1.7 times higher in central cities, and the rate of young men (ages 25-34) who completed high school in central cities was only 94 percent that of the suburbs. Perhaps more significant than the magnitude of these disparities are their trends over time. Table 3-2 shows, for selected measures, the ratios of suburban to central-city outcomes for the decades from 1960 through 1990. With the TABLE 3-1 Comparison of Selected Outcomes for Central-City and Suburban Residents, 1990 Outcome Central City Non-Central City Ratio of Central-City to Suburban Outcome Median household income $26,727 $36,314 .74 Per capita income $13,839 $16,527 .84 Poverty rate, families 14.1% 6.0% 2.35 Employed men ages 25-54a 83.3% 90.0% .93 Unemployment rate, men ages 25-54 6.8% 4.0% 1.7 High school graduates, men ages 25-34 80.7% 85.8% .94 a Members of the armed forces are counted as employed. Source: U.S. Bureau of the Census (1990a).

OCR for page 40
TABLE 3-2 Ratios of Selected Outcomes for Central-City and Suburban Residents, 1960-1990   Ratio of Central-City to Suburban Outcome Outcome 1960 1970 1980 1990 Median family income .89 .85 .81 .77 Per capita income NA .92 .88 .84 Family poverty rate 1.52 1.75 2.08 2.35 Percent employed, men age 16 and oldera 0.96 0.93 0.92 .91 Unemployment rate, men age 16 and oldera 1.37 1.33 1.34 1.59 Percent high school graduates, age 25 and older 0.85 0.86 0.9 .94 Percent college graduates, age 25 and older 0.82 0.84 0.94 .96 a For 1960, the employment and unemployment rates correspond to men at least 14 years of age. Source: U.S. Bureau of the Census (1960, 1970, 1980, 1990a). exception of education, every measure in the table suggests that the relative status of central-city residents has consistently declined over the last three decades. Relative to their suburban counterparts, central-city residents now have lower incomes, higher poverty rates, and lower employment rates than they did 10, 20, and 30 years ago.1 The income differences are particularly stark. In 1960, the income of central-city families was 89 percent of that of suburban families. By 1990, this ratio had fallen to 77 percent. Per capita income follows a similar pattern, with ratios falling from .92 in 1970 to .84 in 1990. Poverty is becoming increasingly concentrated in central-city areas. In 1960, poverty was 1.5 times as frequent for families living in central cities; by 1990, the ratio had risen to 2.4. Finally, in 1960 the unemployment rate in central cities was 1.5 percentage points higher than the unemployment rate in the suburbs; by 1990 it was 3 percentage points higher. These central-city/suburban disparities have grown much more rapidly for blacks than they have for the population as a whole (Ihlanfeldt, this volume). The deepening of central-city disparities over time reflects both the selective migration of better-off people to the suburbs and the declining fortunes of lower-skilled and low-income groups remaining in the city. Their declines in fortune result from changes in the national economy, referred to earlier, that reduce the relative earning power of low-skilled workers. As noted, education is a marked exception to the general trend of continued central-city decline. Because it is easier to increase the proportion of high school

OCR for page 40
and college graduates in a population that has fewer of them to start with, the relative fortunes of central-city residents improved as a result of overall societal progress. Between 1960 and 1990, for example, although the high school graduation rate rose by 66 percent in non-central-city areas, it rose by 79 percent in central cities. These comparisons lump together all suburbs as a single entity, but suburbia is hardly monolithic (Orfield, 1997). Data on neighborhoods within suburban areas provide some sense of their diversity. Although, as noted above, the median household income in suburban America is significantly larger than that in central cities ($36,314 compared with $26,727), in approximately one-fifth of suburban neighborhoods, the median household income is actually less than the central-city median. (And without the commercial tax base of a central business district, these suburbs may be even worse off than their central cities.) The story is similar with the poverty rate. Nationwide, the poverty rate is 8 percent in the suburbs and 18 percent in central cities. But approximately 10 percent of suburban census tracts had poverty rates of at least 18 percent. How do these city-suburban disparities vary by region or by size of the metropolitan area? Table 3-3 indicates that the disparities appear to be greater in the Northeast and the Midwest and lower in the South and the West. On average, the per capita income of central-city residents in the South and the West nearly matched that of their suburban counterparts. In the average metropolitan area in the Midwest and the Northeast, by contrast, the per capita income of central-city residents was just 76 and 65 percent, respectively, of that of suburban residents. A similar geographic pattern applies to the relative poverty rates. In the average metropolitan area in the Midwest, the central-city poverty rate was more than 4 times that of the suburban poverty rate. In the West, central-city poverty was only 1.8 times that of suburban poverty. In larger metropolitan areas, central-city residents fare worse relative to their suburban counterparts than they do in smaller ones. Table 3-4 shows that in metropolitan areas with over 1 million residents, per capita income in central cities is an average of 80 percent of per capita income in suburban locations. In contrast, in metropolitan areas with under 250,000 residents, the per capita income in central-city and suburban areas is almost identical. Similarly, in metro- TABLE 3-3 Mean Central-City/Suburban Disparities in 1990 by Region Region Per Capita Income Poverty Rate Northeast .65 3.75 South .96 2.31 Midwest .76 4.24 West .94 1.77   Source: Tabulations from U.S. Bureau of Census (1996a).

OCR for page 40
TABLE 3-4 Mean Central-City/Suburban Disparities in 1990 by Population Size of Metropolitan Statistical Area   Mean Ratio of Central City to Suburban Outcome in Set of Metropolitan Statistical Areas MSA-Size Class Per Capita Income Poverty Rate Employment Rate, Men 16-64 1,000,000 + .8 3.3 .9 250,000-999,999 .91 2.6 .92 100,000-249,999 .97 2.2 .94 0-99,999 .97 2.2 .94   Source: Tabulations from U.S. Bureau of Census (1996a). politan areas with over 1 million residents, the central-city poverty rate is more than 3 times larger on average than the suburban rate. In metropolitan areas with under 250,000 residents, the ratio is just over 2. In contrast to income and poverty measures, the difference in employment rates between central-city and non-central-city residents varies only modestly with population size. Ellen (this volume) has constructed an index of disparity and ranked all metropolitan areas by their degree of city-suburban disparity.2 Significantly, she found that not all central cities are worse off than their surrounding suburbs. In 1990, central cities were actually as prosperous or more so than their surrounding suburban communities in 101 (31 percent) of metropolitan areas (see also Hill and Wolman, 1997a). Table 3-5 presents the mean characteristics of selected variables for each of four quartiles of disadvantage. Once again, the regional pattern is pronounced. TABLE 3-5 Regional Distribution of Central Cities, Ranked by Degree of Central-City Disadvantage (percentage)   Quartiles of Central-City Disadvantage Region Q1: Greatest Disadvantage 2nd Quartile 3rd Quartile Q4: Least Disadvantage Northeast 43.9 25.6 8.5 3.6 Midwest 31.7 36.6 22.0 10.8 South 20.7 22.0 45.1 60.2 West 3.7 15.9 24.4 25.3 Total 100 100 100 100 Mean Population 934,560 537,981 542,056 277,366

OCR for page 40
In the quartile with the most disadvantaged central cities, 44 percent are located in the Northeast, 32 percent in the Midwest, 21 percent in the South, and 4 percent in the West. In the quartile with the least disadvantaged central cities, by contrast, just 4 percent are in the Northeast, 11 percent are in the Midwest, 60 percent are in the South, and 25 percent are in the West. With respect to size, the metropolitan areas with greater disparities tend to have larger populations. Variations across metropolitan areas in income disparities between cities and suburbs are substantial. Table 3-6 displays the ratios between central-city and suburban outcomes for the 15 largest metropolitan areas. The ratio of central-city to suburban per capita income ranges from a low of .56 in Detroit to a high of 1.0 in San Diego and Los Angeles. The ratios of employment rates range from a low of .74 in Detroit to a high of .98 in San Diego, and high school completion ratios vary from .68 in Anaheim to a high of 1.03 in San Diego. Consistent with the results above, these disparities tend to be larger in the metropolitan areas located in the Northeast and the Midwest and smaller in those in the South and the West. Anaheim is the one TABLE 3-6 Disparities Between Central-City and Suburban Residents in the 15 Largest Metropolitan Areas, 1990     Ratio of Central-City to Suburban Outcome Metropolitan Statistical Area Elasticitya Per Capita Income Employment Rate, Men 16-64 Percent High School Graduates, Men 25-34 Atlanta, GA .139 .89 .86 .94 Riverside-San Bernadino, CA .151 .96 .96 .97 Washington, DC .155 .93 .90 .92 St. Louis, MO .162 .69 .87 .91 Boston, MA .2 .79 .91 .94 Anaheim-Santa Aria, CA .232 .58 .95 .68 Detroit, MI .235 .56 .74 .81 Minneapolis-St. Paul, MN .26 .85 .92 .97 Philadelphia, PA .326 .63 .83 .86 Dallas, TX .394 .96 .92 .89 Los Angeles-Long Beach, CA .442 .998 .96 .93 San Diego, CA .445 1.00 .98 1.03 Chicago, IL .459 .68 .85 .85 Houston, TX .494 .89 .93 .87 New York, NY .857 .68 .89 .90 a Elasticity is the proportion of the metropolitan area population that lives in the central-city named in the title of the Primary Metropolitan Statistical Area. (When two central cities are named in the title of the Primary Metropolitan Statistical Area, the population of both are counted as living in the primary central city.) Source: U.S. Bureau of the Census (1990a).

OCR for page 40
notable exception, probably because of the high proportion of low-skilled, Hispanic immigrants living in its central city—nearly 50 percent of its central-city population is Hispanic. There is also a substantial gap in educational performance between children who attend public schools in central cities and children who attend public schools elsewhere. Only 43 percent of fourth graders attending central-city schools scored at a basic or higher level on the 1994 National Assessment of Educational Progress (NAEP) reading test, compared with 63 percent for children attending public schools outside the central city. The difference was essentially the same (42 versus 66 percent) for eighth graders on the NAEP math test, and it was even greater (38 versus 65 percent) for eighth graders on the NAEP science test (Education Week, January 8, 1998). Minorities and Whites In the last 30 years, racial minorities have gained access to some opportunities previously closed to them. The black middle class has grown (O'Hare and Frey, 1992), racial segregation has declined (albeit modestly), and attitudes about racial prejudice appear to have softened (Farley et al., 1993; Firebaugh and Davis, 1988). Still, racial disparities remain pronounced, especially between African Americans and non-Hispanic whites (Table 3-7). Income differences are particularly striking: The median income of black households living in a metropolitan area was just 61 percent that of the median TABLE 3-7 Comparison of Selected Outcomes for Different Racial Groups, 1990   Population in Metropolitan Areas       Outcomes for Different Racial Groups Ratios of Outcomes Outcome NHW Black Hispanic Black to NHW Hispanic to NHW Median household Income $34,676 $21,247 $25,009 .61 .72 Per capita income $17,559 $9,414 $8,603 .54 .49 Poverty rate 7.7% 27.5% 24.5% 3.6 3.18 Percent employed, men 25-54 90.3% 72.5% 82.6% .8 .91 Unemployment rate, men 25-54 3.9% 11.5% 14.1% 2.95 3.62 Percent high school graduates, men 25-34 89.4% 76.0% 55.7% .85 .62 Note: NHW, non-Hispanic white. Source: U.S. Bureau of the Census (1990a).

OCR for page 40
non-Hispanic white household in 1990; the median income of Hispanic households was 72 percent that of non-Hispanic white households. Racial differences in per capita income are even greater, since minority households are larger on average than white households. Finally, 3.6 times as many metropolitan-area blacks and 3.2 times as many Hispanics lived in poverty in 1990 as non-Hispanic whites. In pan, these racial and ethnic gaps in income reflect differences in employment rates. As the table also shows, over 90 percent of prime working-age, non-Hispanic white men in metropolitan locations held jobs in 1990, compared with 83 percent of Hispanic men of prime working-age and just 73 percent of black men of similar ages. There are sharp differences in the numbers of men explicitly looking for work as well. The unemployment rate was 3 times as great for prime working-age black men as for working-age white men, and the disparity between Hispanics and non-Hispanic whites was even greater. Some fraction of the income gap, as well as the employment gap itself, may be explained by significant racial and ethnic disparities in educational attainment. Among men between the ages of 25 and 34, for example, nearly 90 percent of non-Hispanic whites have at least high school diplomas. The corresponding proportions for African Americans and Hispanics are just 76 and 56 percent, respectively. As Table 3-8 illustrates, these relationships have generally not changed much over the decade from 1980 to 1990.3 In general, blacks improved their educational attainment relative to whites, and the poverty gap closed modestly as well. Still, the gains were small, and blacks actually lost ground relative to whites with respect to their labor market success and per capita income. Racial differences vary both by region and by metropolitan-area size. Table 3-9 illustrates that the gap between black and white per capita income is smaller TABLE 3-8 Percentage Change in Selected Outcomes by Race, 1980-1990   Population in Metropolitan Areas   Percent Change in Outcome, 1980-1990 Outcome Whites Blacks Median household income (constant dollars) 5.6 6.7 Per capita income (constant dollars) 19.4 13.4 Poverty rate 2.4 -.7 Percent employed, men ages 25-54 -.6 -4.9 Unemployment rate, men ages 25-54 .0 15.0 High school graduates, men age 25 and older 12.0 19.9   Source: U.S. Bureau of the Census (1980, 1990a).

OCR for page 40
TABLE 3-9 Mean Racial Disparities in Per Capita Income in Metropolitan Areas in 1990 by Region   Mean Ratio of Per Capita Incomes in Set of Metropolitan Statistical Areasa Region Black/Non-Hispanic White Hispanic /Non-Hispanic White Northeast .56 .51 South .53 .631 Midwest .544 .602 West .607 .501 a Ratios of black and non-Hispanic white and Hispanic and non-Hispanic white per capita income were first calculated for each individual metropolitan area. These ratios were then averaged within the region, weighted by the population of the metropolitan area. Source: Tabulations from U.S. Bureau of Census (1996a). in the West than in the other three regions. An examination of the poverty rate reveals a similar pattern—the poverty rate for African Americans is 1.8 times higher than the white rate in the Western region of the country, and a full 4 times greater in the Midwest (Current Population Reports, 1993). For Hispanics, the income gaps tend to be lower in the South and the Midwest and higher in the Northeast and the West. Significantly, the regional differences are considerably larger for the Hispanic-white differentials than for the black-white ones, probably because of the large differences in background between the Hispanic population groups living in different areas of the country. The magnitude of racial disparities appears to vary with the size of a metropolitan area and to be somewhat larger in larger areas (Table 3-10). In metropolitan areas with over 1 million people, per capita income for blacks was just 55 percent of that for white households on average. In metropolitan areas with less than 100,000 residents, by contrast, black per capita income was, on average, 60 percent of white per capita income. The Hispanic-white disparities in per capita income followed a similar pattern. The per capita income of Hispanics in metropolitan areas of over 1 million people was an average of 53 percent of that for non-Hispanic whites; in metropolitan areas with less than 100,000 residents, the corresponding ratio was just 62 percent. For Hispanics, the difference may be due to the fact that recent Hispanic immigrants tend to settle in larger metropolitan areas (Bartel, 1989). These regional and population size averages, of course, conceal considerable variation across individual metropolitan areas. In certain metropolitan areas, minority groups appear to do far better than in others, and the extent to which central-city residents fall behind their suburban counterparts varies a great deal as well. To give a sense of the range of disparities across metropolitan areas, Tables 3-11 and 3-12 display racial, ethnic, and urban-suburban disparities for the 15

OCR for page 40
TABLE 3-10 Mean Racial Disparities in Per Capita Income in Metropolitan Areas by Metropolitan Statistical Area Population Size, 1990   Mean Ratio of Per Capita Incomes in Set of Metropolitan Statistical Areasa Metropolitan Statistical Area Size Class Black to Non-Hispanic White Hispanic to Non-Hispanic White 1,000,000 + .545 .531 250,000-999,999 .566 .608 100,000-249,999 .577 .616 0-99,999 .603 .624 a Ratios of black and non-Hispanic white and Hispanic and non-Hispanic white per capita income were first calculated for each individual metropolitan area. These ratios were then averaged within the size-class, weighted by the population of the metropolitan area. Source: Tabulations from U.S. Bureau of Census (1996a). largest metropolitan areas (Nassau County, NY, is omitted since it has no central city), for three outcomes: per capita income, employment rates for men of prime working age (between 25 and 54), and high school completion for men ages 25 to 34. In Table 3-11, comparing blacks and non-Hispanic whites, the ratio of per capita income ranges from a low of .43 in New York City to a high of .64 in Riverside. Black-white employment ratios meanwhile vary from a low of .70 in Detroit to a high of .95 in Anaheim. Black-white high school graduation rates range from a low of .78 in New York City to a high of .99 in Anaheim. As expected, these disparities generally tend to be greater in the metropolitan areas in the Northeast and the Midwest. The disparities are also larger in the metropolitan areas with larger and more segregated black populations. Indeed, the simple correlation between the segregation index for blacks (measured by the index of dissimilarity) and Ellen's index of disparity (in this volume) is -.91. More work would be useful to try to probe further into the roots of these metropolitan-area differences. The range in disparities for these three outcomes is even greater for Hispanics and non-Hispanic whites (Table 3-12), consistent with the large cultural variation in the Hispanic population living in the United States. The Hispanic-white ratio of per capita income ranges from a low of .32 in Los Angeles to a high of .77 in St. Louis, and employment ratios range from .79 in Philadelphia to .97 in Washington, DC. High school completion ratios vary from .45 in Dallas to .89 in St. Louis. In general, the disparities are smaller in metropolitan areas, such as St.

OCR for page 40
TABLE 3-11 Disparities Between Blacks and Non-Hispanic Whites in the 15 Largest Metropolitan Areas, 1990   Ratio of Black to Non-Hispanic White Outcome Metropolitan Statistical Area Per Capita Income Employment Rate, Men 16-64 Percent High School Graduates, Men 25-34 Anaheim, CA .63 .95 .99 Atlanta, GA .51 .86 .92 Boston, MA .55 .83 .86 Chicago, IL .45 .73 .8 Dallas, TX .44 .83 .89 Detroit, MI .54 .7 .81 Houston, TX .44 .82 .86 Los Angeles, CA .48 .8 .86 Minneapolis, MN .5 .76 .88 New York, NY .43 .81 .78 Philadelphia, PA .53 .76 .81 Riverside, CA .64 .84 .99 St. Louis, MO .52 .77 .85 San Diego, CA .53 .88 .95 Washington, DC .55 .87 .88   Source: U.S. Bureau of the Census (1980, 1990a). TABLE 3-12 Disparities Between Hispanics and Non-Hispanic Whites in the 15 Largest Metropolitan Areas, 1990   Ratio of Hispanic to Non-Hispanic White Outcome Metropolitan Statistical Area Per Capita Income Employment Rate, Men 16-64. Percent High School Graduates, Men 25-34 Anaheim, CA .38 .95 .48 Atlanta, GA .66 .95 .71 Boston, MA .46 .85 .71 Chicago, IL .4 .94 .55 Dallas, TX .39 .94 .45 Detroit, MI .64 .89 .82 Houston, TX .38 .95 .47 Los Angeles, CA .32 .95 .47 Minneapolis, MN .52 .92 .86 New York, NY .34 .86 .67 Philadelphia, PA .43 .79 .66 Riverside, CA .51 .95 .6 St. Louis, MO .77 .94 .89 San Diego, CA .43 .93 .63 Washington, DC .51 .97 .65   Source: U.S. Bureau of the Census (1980, 1990a).

OCR for page 40
residents with more than a high school education was associated with a 1.41 percentage point difference in per capita income. Decline in Demand for Less-Skilled Labor The national economy has undergone substantial structural change over the past 25 years that has greatly reduced the proportion of employees in manufacturing jobs and increased the proportion in service-sector jobs. In 1970, 26 percent of all U.S. workers were employed in the manufacturing sector, compared with only 16 percent in 1994. In addition, technological change has increased skill requirements in many sectors of the economy, including manufacturing. These changes have shifted labor demand from less-educated to more-educated portions of the workforce. Well-paid, less-skilled jobs in the manufacturing and other sectors have largely disappeared. When combined with differences in human capital on the supply side, the result is a skills mismatch that disadvantages less-skilled workers in general and blacks, a higher proportion of whom are less skilled, in particular (see Wilson, 1987, 1996). Research results suggest that a skill mismatch exists and that declining manufacturing employment has contributed to black-white disparities by reducing the relative earnings and employment of blacks, with particular impact on the youngest and least educated black males (see Holzer, 1994, for a review of this literature). Holzer emphasizes, however, that substantial earnings and employment gaps between blacks and whites remain even after the decline of manufacturing employment and skill mismatch are taken into account. Spatial factors also interact with the decline in demand for less-skilled jobs. Kasarda (1995) and Holzer (1996) both found that job growth in the central city has been in sectors that require higher skills, such as finance and business services, than do the sectors of suburban job growth, such as retail trade, personal services, and traditional blue-collar manufacturing. Kasarda (1995) notes that this exchange of goods-processing for information-processing jobs (requiring higher skills) in central cities has meant that central-city jobs are no longer functionally accessible to less-educated city residents, even if they are physically accessible. Meanwhile, opportunities for the relatively less-skilled jobs that might be appropriate for less-educated city residents are constrained by the spatial mismatch problems discussed earlier. Racial Discrimination in Employment Part of the black-white gap in labor market outcomes may simply be due to employer discrimination in labor markets. Indeed, there is very good reason to believe this is the case. Holzer (1994) summarizes recent evidence, which appears in Turner et al. (1991) and is based on an Urban Institute audit study, in which matched pairs of black and white job applicants, with employment histories and

OCR for page 40
skills designed to be exactly comparable, applied to several hundred firms in two major metropolitan areas. Blacks received significantly fewer job offers than did whites (19 versus 29 percent). Furthermore, there appears to be a spatial element in employer discrimination. Holzer (1996) found that the ratio of black new hires to black applicants is lower in the suburbs than in the central city. He concludes that there is a "relatively lower employer preference for black applicants in suburban areas than elsewhere" (1996:95). In a multivariate analysis, he found that the probability of a firm's hiring blacks is lower for those with mostly white customers, which are more likely to be located in the suburbs. Finally he indicates that the evidence suggests "that discriminatory employers may deliberately choose locations for their firms that make them inaccessible to blacks" (1996:95). Costs of Unequal Opportunity The costs of unequal opportunity are clearly borne by those who are most directly affected in terms of lower income, higher unemployment, and constrained choices. But in addition, there are very real costs imposed on the entire metropolitan area as well as on the nation as a whole. In work undertaken at the request of the committee, Harry Holzer arrived at a rough estimate, for the purposes of this report, of the costs that segregation imposes on young blacks (ages 20-30) in metropolitan areas, as well as on the metropolitan areas overall. He first used regression results from Cutler and Glaeser (1997) to calculate the effects of differences in the levels of segregation across metropolitan areas (as measured by the dissimilarity index) on several major outcomes for young blacks in these areas: high school graduation, college graduation, the probability of not being employed, the log of annual earnings, and the probability of being a single mother. He then linked the costs borne by the young people themselves to their outcomes for the metropolitan areas overall. Holzer's calculations estimate the impact of segregation by contrasting the most highly segregated metropolitan areas with the least segregated metropolitan areas. He derived two estimates of the effects of segregation on young blacks. The first is based on the difference between high and low segregation levels (i.e., dissimilarity indexes of between .70 and .75 at the high end and between .45 and .50 at the low end). Examples of metropolitan areas with high segregation levels include New York and Philadelphia; those with low levels include Phoenix and Raleigh; the difference constitutes a change of two standard deviations in the distribution of segregation levels in metropolitan areas. The second estimate is based on differences in segregation between the most and the least segregated areas (dissimilarity indexes of between .80 and .90 for the most segregated and between .30 and .40 for the least segregated). Examples of metropolitan areas with the most segregation include Detroit, Chicago, and Gary; those with the least include Albuquerque and San Jose. This difference

OCR for page 40
constitutes a change of four standard deviations in the distribution of segregation levels (almost the entire range). Using this technique, Holzer found very large negative effects of segregation on blacks when comparing the least segregated to the most segregated metropolitan areas. An increase in segregation from low to high levels caused a drop in high school graduation rates of 6 percentage points. An increase in segregation from the level of the least segregated areas to that of the most segregated reduced high school graduation rates by 12 percentage points. Similar increases in segregation caused college graduation rates of young blacks to drop by 2.5 to 5 percentage points, the percentage of those not employed or in school to increase by 7.5 to 15 points, annual earnings to decline by 20 to 40 percent, and the probability of being a single mother to increase by 10 to 20 percentage points. Holzer then linked these adverse consequences on individuals to costs for the metropolitan area and beyond. Assuming that worker productivity is proportional to earnings and that blacks are about 15 percent of the population in large metropolitan areas, then, in a steady state over time, a 20 to 40 percent reduction in earnings due to high segregation translates into a 3 to 6 percent decline in productivity for the area as a whole. Segregation also affects labor turnover through increasing high school dropout rates. Turnover imposes major costs on employers, in the form of fixed costs for recruiting, hiring, and training, as well as in lower productivity from lost accumulation of work experience. Turnover rates are at least 40 percent higher per year among high school dropouts than among high school graduates (Bernhardt et al., 1997). An increase in dropout rates of 6 to 12 percentage points implies increases in dropout rates of 30 to 60 percent (since the dropout rate of blacks in low-segregation cities is about 20 percent). This in turn implies an increase in dropout rates of 4.5 to 9 percent for the metropolitan area as a whole (if blacks constitute 15 percent of the population) and an increase in turnover rates of 2 to 4 percent for area employers. Finally, Holzer notes that the result in effects on area crime also are substantial. Freeman (1992) has shown that well over 50 percent of black male dropouts between the ages of 16 and 34 are in the criminal justice system, compared with only 10 percent for black male high school graduates. Thus, an increase in black high school dropout rates due to an increase in segregation will lead to huge increases in crime. Assuming young black males account for approximately 30 to 40 percent of serious crime in less-segregated metropolitan areas, Holzer calculated that the impact of moving from low to high segregation levels will increase serious crime in the metropolitan area by 45 to 60 percent. The impact of moving from segregation levels of the least segregated areas to those of the most segregated areas will increase serious crime by 90 to 120 percent. Holzer's estimates should be taken as rough approximations of the impact of high compared with low degrees of segregation. As such, they are an overstate-

OCR for page 40
ment of the impact for most metropolitan areas. And, as he cautions, they may also overestimate somewhat the effect on employers of labor turnover (because employers are likely to pay lower wages to employees who have a higher probability of turnover) and the effect on crime (since the causality may run both ways: high school dropouts may be more likely to engage in crime, but those who have a propensity to engage in crime may be more likely to drop out of school). Nonetheless, they provide an illustrative estimate of the potential costs, both to blacks and to the metropolitan area as an entity, that are imposed by highly segregated areas compared with what would be the case if these areas had very low levels of segregation. To what extent are these overall costs in metropolitan areas borne by white suburban residents as well as by black inner-city residents? Although it is difficult to make Precise estimates, it is very likely that some significant costs are borne by whites as well as blacks. For instance, the lower earnings of less-educated black workers brought about by unequal opportunity will reduce white income through the multiplier effect, as blacks purchase fewer goods and services throughout the area. In addition, the lower earnings and productivity of these black workers will generate lower tax revenues from them and higher transfer payments to them, both of which are ultimately borne by middle-class taxpayers of both races. Higher poverty rates among blacks will also generate other costs to the white middle class, such as greater expenditures on law enforcement (discussed below) and health care for these groups. Lower productivity of black employees in a given set of jobs will almost certainly result in lower profits to their employers. Even if their lower productivity could be fully offset through lower wages, the "surplus" that accrues to employers as a result of their efforts will still be reduced. And, to the extent that the lower productivity of these employees generates higher turnover, this results in higher direct costs to their employers (in the form of recruitment, screening, and training costs) as well as foregone output while their jobs are vacant. When labor markets are tight, the costs to employers of attracting and keeping good employees can be especially high. Finally, the costs of central-city crime to suburban residents are likely to be substantial. The direct costs of administering the criminal justice system (including the building and staffing of prisons) are currently about 2 percent of gross domestic product nationally, or $150 billion, borne entirely by taxpayers. Although the majority of crime victims are themselves minorities who live in high-crime neighborhoods, some urban crime clearly spills over into neighboring suburban areas; a fear of crime often pervades those areas, above and beyond what actually occurs. Even for those who have effectively escaped the risk of crime by residing in outlying suburbs, the loss of enjoyment of urban amenities may entail some real costs.

OCR for page 40
Disparities in Fiscal Capacity In turning from disparities in outcomes among individuals or groups of individuals to disparities among jurisdictions, we focus on disparities in fiscal capacity among local governments. Tax/service disparities—using the term synonymously with fiscal capacity disparities—are an important part of the metropolitan spatial opportunity structure; they themselves contribute to the disparities in outcomes between central-city and suburban residents and between minority and white households. Tax/Service Disparities Local governments in metropolitan areas differ, sometimes widely, in the tax burden they impose on their citizens and in the level of services they provide. These observable variations in tax burden and services may be termed tax/service differences.12 They reflect a variety of causes, including variations in public preferences, in local government efficiency, and in tax capacity and expenditure need. The committee's concern, however, is not with these differences per se, but with tax/service disparities (sometimes called fiscal disparities), which we define as differences based on factors outside the control of local officials in the ability of their governments to provide a reasonable level of services with a reasonable tax burden on their residents (Yinger, 1996:1). In the absence of equalizing aid from higher levels of government, residents in local governments with low tax capacity or higher expenditure needs are faced with a difficult choice: they must either pay a higher percentage of their income in taxes than do wealthier communities in order to obtain equal levels of public services, or, if they tax themselves at a comparable rate, they must accept lower levels of public services. To the extent that tax/service disparities reflect differences in tax capacity and service need rather than preference, the committee considers them undesirable both in terms of their impact on the spatial opportunity structure of metropolitan areas and in terms of equity.13 The problems stem from the fact that local services are financed largely out of locally raised revenues. In 1991-1992, 62 percent of all local government general revenue (and 72 percent of all municipal government general revenue) was raised from local government's own sources, with the remainder coming from intergovernmental aid (U.S. Bureau of the Census, 1996c). This means that, to the extent that quality of service is related to amount spent, residents of low-income communities face a difficult dilemma. Unless they live in a community with a substantial business tax base, or unless their community is compensated for its low tax base with disproportionate amounts of intergovernmental aid, they will be faced with paying more out of their income in order to receive similar levels of service or, if they do not wish to take on this extra burden, with receiving inferior levels of service. This has

OCR for page 40
particularly severe effects with respect to services, such as education, that are related to life chances. As a consequence, the opportunity structure for residents of low-income communities is spatially constrained. The equity problem is also serious. Let us consider two similar households, each with an income of $30,000, one residing in community A with a local tax base of $10,000 per capita and the other in community B with a local tax base of $40,000 per capita. For the sake of illustration, assume both communities impose only a tax on income to raise local revenues. In order to spend $2,000 per capita in public services, community A would have to impose a tax rate of 20 percent on income, and community B would have to impose a tax of only 5 percent. The $30,000 household in the lower-income community would thus face a tax of $6,000, and the $30,000 household in the wealthier community would face a tax of only $1,500. This clearly fails the test of horizontal equity. Now consider a poor household with an income of $15,000 in community A and a middle-class household with an income of $40,000 in community B. In order to receive the same $2,000 of services per capita, the poor household in community A would have to pay $3,000 in taxes, and the wealthier household in community B would have to pay only $2,000. In short, poor people living in lower-income communities must pay a higher proportion of their income, in this example, a larger absolute amount of income, than wealthier people in more prosperous communities in order to receive the same level of spending on services. Thus, the test of vertical equity is also failed. The fiscal system in metropolitan areas operates either to reduce the income of residents of low-income communities relative to those of wealthier communities, thus contributing to disparities, or to reduce the level of spending on public services financed out of local revenues, with negative effects on life chances. It is true that the tax/service disparities described above are moderated to some extent by the capitalization of the relative taxes into house values (Oates, 1969; Hamilton, 1975). Thus, the purchase price of a house in a high-tax/low-service area would be lower than that of a similar house in a low-tax/high-service area. However, capitalization of these fiscal factors into house values are unlikely to fully offset tax/service disparities. In addition, they serve to reduce the wealth of existing owners of homes in central cities, most of whom, as the result of selective out-migration, have low and moderate incomes. As the tax burden of central-city residents increases and public services deteriorate, the value of the homes of existing homeowners declines to reflect these changes, thus bringing about a real decline in the value of their capital assets. There has been relatively little research on intrametropolitan fiscal disparities, although Ladd and Yinger (1991) have shown that fiscal disparities among large cities are substantial. Empirical studies are fraught with difficulty. Obviously gross differences in own-source revenues and spending reflect differences among communities in preferences and in the efficiency with which services are provided, as well as differences in tax capacity and expenditure need. Further-

OCR for page 40
more, communities may export taxes to those outside the area, so that own-source revenues may not always come from local residents. In addition, disparities due to tax base and expenditure need may be reduced to some extent through aid from higher levels of government. (For a more extensive discussion of this, see Ladd, 1994a, 1994b). Finally, as Pagano points out (this volume), most disparity studies have not taken into account the increasing reliance of local governments on fees and services and the effect of proliferating residential community associations and their ''voluntary'' contributions on service receipt and taxpayer burden. One study calculated estimates of fiscal capacity for a sample of general-purpose local governments in the Chicago metropolitan area for 1987 (Rafuse, 1991). Own-source revenue-raising ability was estimated by calculating how much revenue each local government would raise from each of 10 separate categories of revenues (e.g., property tax, sales tax, fees, and charges) if they were to levy rates at the average level for all governments in the area. Representative expenditure need was estimated by calculating what it would cost each local government to provide a standard (average) set of local public services, given the underlying socioeconomic and demographic composition of the community (and assuming the average level of efficiency for local governments in the area. Both the revenue and the expenditure capacity estimates were indexed to the average for the entire metropolitan area, so that the average overall fiscal capacity value was 100. An overall fiscal capacity index was created by dividing the own-source general revenue index for a local government by its index of representative expenditures. Using this mechanism, the city of Chicago had an own-source general revenue index of 80 and an index of expenditure need of 111; this yielded an overall index of fiscal capacity of 72 (28 percent below the average for the metropolitan area). However, when Rafuse factored in federal and state grants and added them to own-source revenue, Chicago's index of fiscal capacity rose to 87, still 13 percent below the average for local governments in the metropolitan area. By contrast, Evanston's overall index of fiscal capacity (including federal and state grants) was 117, Lake Forest's was 266, and Winnetka's was 207. Several lower-income suburbs had fiscal capacity indexes considerably lower than Chicago's: Maywood's was 54, North Chicago's was 60, and Burbank's was 68. Unfortunately, similar data are not available for a broad range of metropolitan areas. However, a similar analysis was performed for all municipal governments with populations over 2,500 in the state of Wisconsin for 1987-1991 (Green and Reschovsky, 1994). Using a slightly different methodology, Green and Reschovsky construct a need-expenditure gap similar to Rafuse's fiscal capacity index. The need-capacity gap for the "average" city was set at 0. Milwaukee had a gap between needed expenditures and revenue-raising capacity of $111 per capita after taking into account intergovernmental aid. Some other older suburbs also had very high gaps. These included Cudahy ($148), South Milwaukee ($143), and West Allis ($67). By contrast, many of Milwaukee's suburbs had

OCR for page 40
negative need-capacity gaps: Brookfield (-$249), Glendale (-$198), Menomonee Falls (-$114), Whitefish Bay (-$102), and Wauwatosa (-$47). Causes of Tax/Service Disparities Why is there variation among citizens in a metropolitan area with respect to the local tax burden they face and the level and quality of local public services they receive? The first and most obvious reason is because, in virtually all metropolitan areas in the United States, there are many separate local governments, each with its own authority (albeit constrained authority) to make taxing and spending decisions and each highly dependent on its own local tax base for its revenues. As a consequence, residents in different local jurisdictions will naturally face variations in the tax rates they pay and in the level and type of services they receive. As has been noted, these variations occur for a variety of reasons, including differences in preferences and in local government efficiency as well as differences in fiscal capacity. However, it is the differences in fiscal capacity with which we are here concerned. Fiscal capacity differences can reflect differences in both revenue-raising capacity and in expenditure needs. Differences in the revenue-raising capacity of local governments (defined as the amount of revenue each local government could raise from various tax bases—the major ones are property, sales, and earnings—if it imposed a standard tax burden on its residents) arise for two reasons, according to Ladd (1994a:238); there are "differences in the average per capita income of residents and differences in the ability of cities to export tax burdens to nonresidents. For example, at the standard tax burden a city with richer residents can raise more revenue per resident from its residents than can a city with poorer residents. And a city with a large proportion of its property tax base in the form of business property, with a large proportion of its retail sales to nonresident commuters or tourists, or with a large proportion of earnings generated in the city accruing to nonresident commuters, can substantially increase its revenues by exporting tax burdens to nonresidents." Differences in the expenditure needs of local governments (defined as the amount of money local governments must spend in order to achieve a standardized package of public services) reflect differences in the costs they face of providing public services. Costs may vary because of differences in input costs, such as the market-determined costs of attracting workers, and because of environmental factors (Ladd and Yinger, 1991). Environmental factors include both the composition of the population (a local government with a higher proportion of school-age children in the population will have greater expenditure need for education than will one with a lower proportion) and because of the difficulty in providing services (e.g., a local government with social and economic conditions

OCR for page 40
that breed crime will have to spend more on police protection than will a typical upper-middle-class suburban community). Using regression analysis to determine the impact of various environmental conditions on the cost of providing average levels of public service, Ladd reports that (1994a:239) "poverty, for example, has a big impact on the costs of providing public safety. A city with a poverty rate 1 percentage point higher than that of another city will have police costs that are 5.5 percent higher on average. Furthermore, a city with a poverty rate one standard deviation above the 1982 mean must pay 36.4 percent more for police services than a city with average poverty. The size of a city, as measured by its population, and the amount of its economic activity, as measured by private employment per capita, also increase the costs of public services." It is clear that, given their population characteristics, central cities are likely to have both low standardized revenue-raising capacity and high expenditure needs. Suburbs may vary, with poorer, inner-core suburbs increasingly resembling central cities with respect to relative fiscal disparities or even being worse off, and with higher-income suburbs at the top end. The actual fiscal condition of a city reflects not only its standardized revenue-raising capacity and expenditure needs, but the actual set of fiscal institutions it operates under (Ladd and Yinger, 1991). Local governments are constrained by their state with respect to the kinds of taxes they are actually permitted to levy and the rate at which they can levy them (see Pagano, this volume:Table 5, for variations among states in the extent to which they permit local governments to levy income and sales taxes); they may also be constrained by state limitations on the total amount of taxes they can raise or expenditures they can make. As Pagano notes, a city that is prohibited access to an earnings or income tax that can be imposed on nonresident employees who work in the city will find itself in a poorer revenue-generating position than one that is permitted to do so. State governments also set forth the service responsibilities for local governments. Local governments with a broader range of service responsibilities will have greater expenditure needs than those with a narrower range of responsibility. Although most large city governments do not have responsibility for financing welfare, there are some, such as New York City, that do. The activity of overlying units of local government also affects the actual fiscal condition of a specific local government: the greater the taxes levied by overlying units of government, the less revenue-raising capability exists for the local government. Finally, actual fiscal condition is also affected by the amount of state and federal aid that a local government receives, since such aid makes up, to some extent, for the gap between local revenue-raising capacity and expenditure need. Ladd and Yinger found substantial variations in actual fiscal health among large cities across metropolitan areas. Some of these variations reflect differences that are likely to be greater among metropolitan areas than within metropolitan areas, which is our concern. Thus, for example, differences facing local

OCR for page 40
governments in the cost of labor are likely to be smaller within a metropolitan area, and state fiscal rules conditioning local government behavior (taxes allowed, tax and expenditure limits) and defining service responsibilities are likely to vary much less among jurisdictions within a metropolitan area. Thus, differences in state-imposed fiscal rules may explain the extent of fiscal disparities among local governments across metropolitan areas, particularly the relative fiscal health of cities. Pagano notes (this volume) that only 11 states permit some of their local governments to levy an income tax on nonresidents or a payroll tax on employees, regardless of residence. In most cases, the ability to levy such taxes is severely limited to the very largest municipal governments, although Ohio and Pennsylvania permit large numbers of municipalities access to these taxes.14 Notes 1   Data from the Current Population Survey presented in Ihlanfeldt (this volume) suggest that the city-suburban disparity in family income continued to rise between 1989 and 1995. 2   The index is calculated from three ratios of central-city to suburban outcomes: per capita income, employment rate for men between 16 and 64, and the proportion of persons 25 or older who have at least a high school diploma. In order that each of the three measures is weighted equally, they are each standardized to a scale of 0-100, and the standardized ratios are then averaged to create an overall index of disparity. The higher the number, the better off the central city relative to its suburban counterparts. In order to give equal weight to each of the three ratios, they were each standardized to a scale of 0-100 before being averaged. The minimum value was assigned as zero and the maximum value 100. The ratios in between are assigned values between 0 and 100 according to the following formula: y = (x-xmin)/(xmax - xmin). 3   The figures here compare all whites and all blacks, since separate non-Hispanic white estimates were not available for 1980. 4   A "bad neighborhood" is defined as a neighborhood in which more than 40 percent of teenagers are high school dropouts, more than 40 percent of families are headed by single females, and less than 10 percent of employed persons hold professional or managerial jobs. 5   Hanushek (1996:59) does not argue that schools or school inputs do not matter. He argues, instead, that, although commonly measured input characteristics, such as class size and teacher education and experience, do not make a difference, there are other unmeasured characteristics of schools and teachers that do. 6   In particular, they were able to utilize a "value added" approach to track changes in educational performance over time for individual students, to isolate instructional spending from total educational spending, and to utilize measures of actual class size rather than school or district averages of teacher-pupil ratios. 7   In metropolitan areas in which there is more than one central city, they included the largest central city, but included smaller central cities only if their population was over 100,000. 8   Blair et al. (1996) examine the effect of elasticity on both city and metropolitan-area growth in population, employment, per capita income, and poverty in 117 metropolitan areas between 1980 and 1990. They found that elasticity is related significantly to growth in all of the variables for central cities (positively for the first three and negatively for poverty). It is significantly related to growth in population and employment, but not for per capita income or poverty for metropolitan areas. However, the only variable controlled for in the analysis was the state change (net of the

OCR for page 40
    metropolitan area) in each of the dependent variables examined. These findings do not address the question of the impact of elasticity on central-city/suburban disparities. 9   The finding that more fragmented systems of government lead to lower per capita spending does not necessarily mean they are more efficient, since it is virtually impossible in these studies to separate out expenditure differences that reflect efficiency from those that reflect differences in the quantity or quality of services provided. Indeed, we would expect the demand for services to vary with different sizes of government over which demand is aggregated and with different spillover effects associated with different sizes of government. 10   The most commonly used operational measures is the number of local governments in a metropolitan area or number of local governments per capita. But it is not clear that this is a valid measure of the concept. Should all local governments be counted in fragmentation research? To the extent the research is concerned with coordination difficulties, then perhaps the most commonly used measures are reasonable: the greater the number of local governments (or local governments per capita), the greater the difficulties in arranging coordinated activity across the area. But, if the concern is the "sorting" consequences of fragmentation, as discussed above, perhaps only those local governments that have an important effect on the sorting process, i.e., those with local land use powers (primarily general-purpose units of local government rather than special districts) should be counted. However, it might well be argued that local school districts, although a special district and without land use powers, play such a significant role in the sorting process that they should be included as well. 11   Lewis' measure of fragmentation differs dramatically from previously utilized measures. He constructs a fragmentation index that is equal to TE(1-SSP), in which TE is total expenditures per capita in the metropolitan area and SSP is the sum of the squared percentages of total expenditures accounted for by each local government. The greater the number of governments, each with a lower share of total government expenditure, the greater will be the fragmentation index. 12   Bald (1994:297) reports on tax/service differences in 35 metropolitan areas between central cities and suburbs in the aggregate. On average in these 35 areas, central cities spent $1.51 per capita for every $1.00 per capita spent by the area's suburbs. (The difference was due to the much higher level of spending by cities on noneducational expenditures; suburban governments spent more per capita on education than did cities.) However, taxes as a percentage of family income (tax burden) were an average of 44 percent higher in the central cities, and Bahl notes (1994:297) that the tax burden disparity is increasing over time. 13   Oakland (1994:7-8) argues, however, that fiscal disparities are not necessarily undesirable and that efforts to reduce or eliminate them could have perverse efficiency consequences. 14   Pagano observes (this volume) that, on the basis of the Ladd and Yinger measures, the cities with the greatest actual revenue-raising capacity are in Ohio, which permits the most progressive earnings taxes on commuters; Cleveland's revenue-raising capacity was 41 percent higher than the average U.S. city and Dayton's 59 percent higher, both substantially above their standardized revenue-raising capacity.