SEVERE DISTRESS AND CONCENTRATED POVERTY: TRENDS FOR NEIGHBORHOODS IN CASEY CITIES AND THE NATION

Transcription

1 ANNIE E. CASEY FOUNDATION MAKING CONNECTIONS INITIATIVE SEVERE DISTRESS AND CONCENTRATED POVERTY: TRENDS FOR NEIGHBORHOODS IN CASEY CITIES AND THE NATION G. Thomas Kingsley and Kathryn L.S. Pettit October 2003 THE URBAN INSTITUTE WASHINGTON, DC

2 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 1 SEVERE DISTRESS AND CONCENTRATED POVERTY: TRENDS FOR NEIGHBORHOODS IN CASEY CITIES AND THE NATION G. Thomas Kingsley and Kathryn L.S. Pettit The Urban Institute October 2003 INTRODUCTION O Hare and Mather (2003) have examined national trends in the scope of severe neighborhood distress over the 1990s. Kingsley and Pettit (2003) have done the same for a different measure, concentrated poverty. This paper was prepared for the Annie E. Casey Foundation. It takes a preliminary look at the overlap between these two concepts, for all metropolitan areas and then for Casey neighborhoods: neighborhoods selected for intensive work by the Foundation in its 10 Making Connections cities and three Civic Sites. DEFINITIONS AND DIFFERENCES This research measures the two concepts for 1980, 1990, and 2000 for all U.S. metropolitan areas (the official 1990 definition incorporating 50,502 census tracts is held constant for all time periods) using the Neighborhood Change Database (NCDB) which contains data for tracts with consistently defined boundaries over time. 1 As to concentrated poverty, we present some data for what we have previously termed extreme-poverty neighborhoods (tracts with poverty rates of 40 percent or more), but generally present results for high-poverty tracts (poverty rates of 30 percent or more). 1 These data are consistent with those used and further defined in Kingsley and Pettit, For further descriptions of the NCDB see In contrast, O Hare and Mather, 2003: (1) analyze the period only (not ); (2) focus on summary data for all U.S. tracts (not just metropolitan); (3) look at data for tracts with contemporaneous boundary definitions as of 1990 and 2000 as presented in the census; and (4) although they seldom refer to them, use the new 2003 definitions of metropolitan areas.

3 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 2 The definition of severely distressed neighborhoods is almost the same as that used by O Hare and Mather, but there are a few differences (noted in parentheses). Severely distressed tracts are those where three or four of the following four conditions apply in any year: 1. Poverty rate: 27.4 percent or more (same as O Hare and Mather). 2. Share of families with own children headed by females: 37.1 percent or more (O Hare and Mather used families with related children, a slightly larger universe -- e.g., includes families headed by grandmothers) Share of population years who are not in school and have no high school degree: 23.0 percent or more (i.e., high school drop outs, same as O Hare and Mather). 4. Share of civilian males 16 years or older who are not employed: 44.2 percent or more (O Hare and Mather used males 16-64) 3 The definitional differences seem to have little effect on outcomes. For example, O Hare and Mather now calculate the share of all children in severely distressed tracts nationally as 7.5 percent in 1990 and 7.7 percent in Our calculations (metropolitan areas only) yield 7.5 percent in both years. The severe distress concept is a modification of the underclass concept developed by Ricketts and Sawhill (1988). It sets the threshold for distress for each measure as one standard deviation above the mean. Along with O Hare and Mather, we hold the threshold values as calculated for 1990 constant for all years, whereas Ricketts and Sawhill used the values as calculated for each year separately. We agree that O Hare and Mather s constant threshold approach is more meaningful 4 and provides a useful interim measure. But we expect to assess alternative formulations in the next stages of our own research. We are not confident we know enough about how these different measures relate to distress. The overall determinations depend on a complex interaction among the four constituent indicators which is not well understood, and there are other indicators that may be equally important. The percent of all metropolitan tracts exceeding the threshold for each indicator and year under the above definition is as follows: 2 We did not match precisely in this case because the O Hare and Mather variable was more difficult to construct from our NCDB source data. 3 We did not match precisely in this case because we wanted to take the comparisons back to 1980 and data on employment for the cohort is not available for (The cut-off here is on standard deviation above the mean for the 16 years or over group as of 1990 i.e., the same concept applied to a somewhat different universe). 4 There have been a number of criticisms of the Ricketts and Sawhill approach on this and other grounds. See Jargowsky, 1997, for a summary.

4 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 3 1. Poverty rate, 27.4 percent or more: 9.2 percent of tracts in 1980, 12.7 percent in 1990, 12.7 percent in Share of families with own children headed by females 37.1 percent or more: 11.0 percent of tracts in 1980, 14.8 percent in 1990, 19.4 percent in Share of population years who are not in school and have no high school degree 23.0 percent or more: 17.4 percent of tracts in 1980, 13.5 percent in 1990, 11.3 percent in Share of civilian males 16 years or older who are not employed, 44.2 percent or more: 10.6 percent of tracts n 1980, 11.9 percent in 1990, and 17.1 percent in Clearly, any overall increase in the number of severely distressed tracts in the 1990s was driven by increases in only two of these indicators: the share of families with children headed by females and the share of civilian males 16 years or older who are not employed. TRENDS FOR ALL METROPOLITAN AREAS Table 1 shows the data for all U.S. metropolitan areas. Are things getting better or worse? It depends on which measure you select. We think the share of poor children in troubled neighborhoods is a particularly important indicator and, on this score, circumstances clearly improved in the 1990s after serious deterioration in the previous decade. This share of all poor kids in severely distressed tracts went up from 24.2 percent to 25.1 percent in the 1980s but then down again to 23.5 percent in The share in high-poverty tracts went up from 30 percent in 1980 to 35 percent in 1990 but back down to 30 percent in The share in lesstroubled neighborhoods (neither severely distressed nor high poverty) went from 68 percent in 1980 down to 63 percent in 1990, but back up to 67 percent in We did not match precisely in this case because the O Hare and Mather variable was more difficult to construct from our NCDB source data.

5 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 4 Table 1 CONCENTRATION ANALYSIS -- All Metropolitan Areas Number Number of Tracts Total Population (thous.) Total Children (thous.) Total Poor Children (thous.) Poverty rate 30% or more 3,856 5,432 5,224 12,764 17,987 17,957 4,089 5,423 5,421 2,061 2,889 2,615 Severely distressed 2,581 3,393 3,548 8,556 10,446 11,224 2,865 3,288 3,557 1,530 1,882 1,786 Not severely distressed 1,275 2,039 1,676 4,208 7,541 6,733 1,223 2,135 1, , All other tracts 46,646 45,070 45, , , ,139 44,204 44,127 52,012 4,919 5,273 6,108 Severely distressed ,548 1,723 2, Not severely distressed 46,145 44,547 44, , , ,447 43,760 43,681 51,252 4,756 5,105 5,844 Total 50,502 50,502 50, , , ,096 48,293 49,550 57,432 6,979 8,162 8,723 Severely distressed 3,082 3,916 4,335 10,104 12,169 13,916 3,309 3,734 4,317 1,692 2,049 2,050 Not severely distressed 47,420 46,586 46, , , ,180 44,983 45,816 53,115 5,287 6,113 6,673 Percent of Total Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed

6 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 5 If you look at the share of all children in these neighborhoods, conditons worsened in the 1980s but did not change much in the 1990s. Over the three censuses, the share in severely distressed tracts went from 6.9 percent to 7.5 percent to 7.5 percent. The share in high-poverty tracts went from 8.5 percent to 10.9 percent to 9.4 percent. The share in less-troubled neighborhoods went from 91 percent to 88 percent to 89 percent. If you count the number of tracts in each category (places) the trends are more complicated. The number of severely distressed tracts increased in both decades: from 3,082 in 1980, to 3,916 in 1990 and then to 4,335 in But the number of high poverty tracts decreased modestly in the 1990s: from 5,432 to 5,224 after having expanded sizeably (by 41 percent) in the 1980s. In 1980, less-troubled tracts made up 91 percent of the total, but their share dropped to 88 percent in 1990 and, as a result of the above interactions, wound up at that same level in The overlap of the severe distress designation and concentrated poverty has not been calculated in any research published before. Table 1 shows that, as might be expected, it is substantial. In 2000, 82 percent of severely distressed tracts were also high-poverty tracts; 68 percent of high-poverty tracts were also severely distressed (proportions that have not changed greatly since 1980). Altogether in 2000, there were: 3,548 tracts that were both high-poverty and severely distressed (such tracts were the location of 22 percent of all poor children in 1980, 23 percent in 1990, and 21 percent in 2000) 1,676 tracts that were high-poverty, but not severely distressed (location of 8 percent of poor children in 1980, 12 percent in 1990 and 10 percent in 2000) 787 tracts that were severely distressed but not high-poverty (location of 2 percent of poor children in 1980 and 1990, and 3 percent in 2000). Table 2 shows the changes for a more detailed set of poverty rate categories. The most important finding here probably relates to the substantial declines in all of the measures for the extreme poverty tracts (those with poverty rates of 40 percent or more) in the 1990s (as previously identified and analyzed by Jargowsky, 2003, and Kingsley and Pettit, 2003). For example, the share of all poor children in such tracts dropped from 20 percent to 14 percent. Those declines were offset by a slight increase in the percent poverty rate category and a more sizeable one in the percent category.

7 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 6 Table 2 CONCENTRATION ANALYSIS -- All Metropolitan Areas, Percents by Detailed Poverty/Distress Categories Number of Tracts Total Population (thous.) Total Children (thous.) Total Poor Children (thous.) Poverty rate 40% or more Severely distressed Not severely distressed Poverty rate 30-40% Severely distressed Not severely distressed Poverty rate 20-30% Severely distressed Not severely distressed Poverty rate less than 20% Severely distressed Not severely distressed Total Severely distressed Not severely distressed

8 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 7 Table 3 CONCENTRATION ANALYSIS -- Central Cities of Largest 100 Metro Areas Number Number of Tracts Total Population (thous.) Total Children (thous.) Total Poor Children (thous.) Poverty rate 30% or more 2,595 3,366 3,231 8,569 10,883 10,617 2,844 3,403 3,317 1,502 1,859 1,631 Severely distressed 1,966 2,328 2,381 6,597 7,166 7,553 2,241 2,304 2,443 1,224 1,335 1,237 Not severely distressed 629 1, ,973 3,716 3, , All other tracts 11,465 10,694 10,829 39,886 39,768 43,666 9,895 9,083 10,403 1,595 1,557 1,857 Severely distressed ,006 1, Not severely distressed 11,167 10,391 10,356 38,945 38,763 42,088 9,620 8,818 9,947 1,493 1,457 1,698 Total 14,060 14,060 14,060 48,455 50,651 54,283 12,740 12,486 13,720 3,097 3,415 3,488 Severely distressed 2,264 2,631 2,854 7,537 8,172 9,130 2,516 2,568 2,899 1,326 1,435 1,395 Not severely distressed 11,796 11,429 11,206 40,918 42,479 45,152 10,223 9,917 10,821 1,772 1,980 2,093 Percent of Total Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed

9 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 8 Table 4 CONCENTRATION ANALYSIS -- Suburbs of Largest 100 Metro Areas Number Number of Tracts Total Population (thous.) Total Children (thous.) Total Poor Children (thous.) Poverty rate 30% or more ,349 2,532 3, Severely distressed ,296 1, Not severely distressed ,237 1, All other tracts 23,566 23,276 23,202 80,155 94, ,559 22,636 23,795 28,722 2,001 2,284 2,694 Severely distressed Not severely distressed 23,469 23,177 23,043 79,838 93, ,956 22,544 23,705 28,560 1,968 2,250 2,638 Total 23,974 23,974 23,974 81,505 96, ,590 23,080 24,570 29,689 2,200 2,668 3,136 Severely distressed ,068 1,640 2, Not severely distressed 23,654 23,491 23,358 80,437 94, ,403 22,726 24,060 28,998 2,040 2,406 2,825 Percent of Total Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed

10 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 9 PRIMARY CITIES AND SUBURBS OF LARGER METROPOLITAN AREAS In 2000, the population of America s 330 metropolitan areas totaled million (80 percent of the nation s total). The population of the largest 100 among these metropolitan areas was million (75 percent of the metropolitan total). The population of the primary cities these large metropolitan areas was 54.3 million (24 percent of the all metropolitan total). Yet those cities accounted for 61 percent of the children in high-poverty tracts in all metropolitan areas and 67 percent of their children in severely-distressed tracts. The dominance of these 100 large cities in this regard, however, has been declining. Their share of all metropolitan children in high-poverty tracts had dropped from 70 percent in 1980, and to 63 percent in 1990, before reaching 61 percent in Their share of all metropolitan children in severely distressed tracts dropped from 76 percent in 1980, and to 69 percent in 1990, before reaching 67 percent in Most of the difference was accounted for by growing shares in the suburbs of the largest 100 metropolitan areas. Their share of all metropolitan children in high-poverty tracts grew from 11 percent in 1980, to 14 percent in 1990, and to 18 percent in Their share of all metropolitan children in severely-distressed tracts increased similarly from 11 percent in 1980, to 14 percent in 1990, and to 16 percent in Tables 3 and 4 show all results for the primary cities and suburbs of the 100 largest metropolitan areas. As would be expected, the incidence of troubled neighborhoods is much higher in the large cities than it was on average in all metropolitan areas. In 2000, 24 percent of all children lived in high-poverty tracts and 21 percent lived in severely-distressed tracts. In the suburbs, in contrast, only 2.7 percent of all children lived in high-poverty tracts and 1.9 percent lived in severely-distressed tracts. CONCENTRATED POVERTY AND SEVERE DISTRESS IN CASEY NEIGHBORHOODS OVERALL Table 5 presents results in the same categories for the neighborhoods selected for intensive treatment in the 10 Making Connections cities (Denver, Des Moines, Hartford, Indianapolis, Louisville, Milwaukee, Oakland, Providence, San Antonio and Seattle) and three Civic Sites (Atlanta, Baltimore and Washington). Altogether the selected neighborhoods incorporate a total of 138 tracts with a 2000 population of 471,000 (down from 499,000 in 1990 and 525,000 in 1980).

11 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 10 Table 5 CONCENTRATION ANALYSIS -- All Casey Neighborhoods Number Number of Tracts Total Population (thous.) Total Children (thous.) Total Poor Children (thous.) Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed Percent of Total Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed

12 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 11 These tracts were selected by the Foundation because they were thought to be tough neighborhoods so it is not surprising that in 2000 they were predominantly highpoverty (90 tracts, 65 percent) and high-poverty (84 tracts, 61 percent). Only 30 percent of them (41) were in neither of those categories. Overall, the absolute numbers of children (total and poor) in high-poverty and severely distressed tracts in these areas went down over the 1990s, but this was a function of their overall population loss. The share of total children in high-poverty tracts went down while the share in severely distressed tracts increased slightly. More specifically: The number of poor children in high-poverty tracts went down from 72,000 in 1990 to 52,000 in 2000 while the share went down from 85 percent to 78 percent. In severely-distressed tracts, the number of poor children went down from 54,000 to 44,000 but the share went up from 64 percent to 67 percent. The share of poor children in Casey neighborhoods in less-troubled tracts (neither of these categories) went up from 12 percent to 18 percent. The total number of children (poor and not) in high poverty tracts went down from 124,000 to 103,000 and the share went down from 75 percent to 68 percent. In severely distressed tracts, the number went down from 90,000 to 88,000 but the share went up from 55 percent to 58 percent. The share of all children living in less-troubled tracts went up from 21 percent to 28 percent. As with the all-metropolitan experience, the number of tracts in the highpoverty category declined (from 95 to 90) while the number in the severely-distressed category increased (from 79 to 84). Again, the overlap of categories is substantial. In 2000, 92 percent of severely distressed tracts were also high-poverty tracts; 86 percent of high-poverty tracts were also severely distressed. RANGE OF PROBLEMS AND TRENDS ACROSS INDIVIDUAL CASEY NEIGHBORHOODS Table 6 shows the share of all children in the various categories of tracts in 1980, 1990 and 2000 for the 13 Casey neighborhoods individually. There are strong contrasts. With respect to concentrated poverty in 2000, 100 percent of the children in the Louisville and Providence neighborhoods lived in high-poverty tracts. Next in order was Milwaukee at 92 percent, followed by a cluster in the percent range (Washington, Hartford, Baltimore, Atlanta, and San Antonio), and then Oakland at 61 percent. The smallest shares of children in high-poverty tracts were in Indianapolis (36 percent), Denver (33 percent), Des Moines (27 percent), and Seattle (12 percent).

13 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 12 Table 6 CONCENTRATION ANALYSIS -- Individual Casey Neighborhoods Atlanta Baltimore Denver Des Moines Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed Hartford Indianapolis Louisville Milwaukee Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed Oakland Providence San Antonio Seattle Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed Washington Poverty rate 30% or more Severely distressed Not severely distressed All other tracts Severely distressed Not severely distressed Total Severely distressed Not severely distressed The pattern was generally similar with respect to the shares in severelydistressed tracts in 2000, but there were some notable individual differences in position. Baltimore tops this distribution at 100 percent, followed by a similar cluster in the percent range (Providence, Washington, Milwaukee, Hartford, Louisville, Atlanta and San Antonio). Denver came next (66 percent), followed by Des Moines (45 percent), Indianapolis (40 percent) and Seattle (16 percent). Surprisingly, by 2000 Oakland no longer had any tracts that met the severely-distressed criteria.

14 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 13 Perhaps the best way to summarize directions of change is to look at trends in the shares of all children living in less-troubled tracts in these neighborhoods (those that were neither high-poverty nor severely-distressed). First we note four that experienced major improvements. The largest was Denver where the share living in less-troubled tracts in the Casey neighborhoods went up from 0 in 1990 to 35 percent in Gains were also experienced in Oakland (from 11 to 39 percent), and Atlanta and San Antonio (both improving from 0 to 17 percent). Closely aligned with this group are two others which maintained high shares in less-troubled tracts, even though they did not evidence much or any further improvement: Indianapolis (slight decline from 53 to 50 percent) and Seattle (modest improvement from 84 to 88 percent). Des Moines is in something of an intermediate category, still having a fairly high score on this indicator, but suffering a more notable worsening of conditions in the 1990s (drop from 62 percent to 55 percent). At the other extreme, this measure evidenced a significant worsening of conditions in three sites: Washington (decline from 35 percent in less-troubled tracts in 1990 to 12 percent in 2000); Providence (from 16 percent to 0); and Louisville (from 14 percent to 0). These neighborhoods are not much more of a worry, however, than those in the remaining three sites where none or very small shares of children lived in less-troubled conditions and no (or negligible) improvements occurred. Outstanding in this regard is the Baltimore, where none of the tracts has ever been in the less-troubled group since 1980 (i.e., all children live in tracts that are both high-poverty and severely distressed). The others in this group are Hartford (share in less-troubled tracts changed from 6 percent in 1990 to 7 percent in 2000) and Milwaukee (change from 7 percent to 9 percent). IMPLICATIONS The evidence cited above confirms our earlier analysis with respect to the declining incidence of concentrated poverty in the 1990s. It also shows a slight decline in the share of poor children in severely-distressed neighborhoods. We interpret this as good news in part, as in our earlier paper, because it showed at least that positive change for a time is possible. In our view, the earlier widely held expectation that

15 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 14 continuing further concentration of poverty was inevitable itself created a context that inhibited sound policy. However, we did not suggest that this evidence provided confidence that trends would remain positive in the future. It appears that neighborhood poverty trends can fluctuate and we said we thought it likely that with the economic downturn since 2000, poverty might well be moving toward more concentration again. The data examined in this paper show a mix of positive and negative changes depending on the indicators chosen. None of these trends is so dramatic as to establish an unshakable trajectory in either direction. High-poverty and severely-distressed neighborhoods are not likely to vanish, nor are their numbers likely to double, any time soon. Recognizing that interventions can make a difference, the purpose of research on these topics ought to be to learn more about what interventions work best in what circumstances; not just to determine whether recent trends should be judged to be positive or negative overall. In that light, we think more research is warranted on how varying neighborhood characteristics interact with poverty rates to suggest how the mix of appropriate interventions might best be constructed in different circumstances. The Making Connections theory of change suggests that a mix of place-based and people-based strategies is appropriate in tough neighborhoods. It seems the placebased component is probably less important in low-poverty neighborhoods. But our guess is that place-based strategies are likely to have a role to play in all neighborhoods with poverty rates at least in excess of 20 percent. However, the best mix is probably different in a neighborhood just above that threshold than one with a poverty rate above 40 percent and high scores on all severe distress indicators. Further research is needed to help us understand such differences in conditions and potentials. As a part of this, we think it may be worth exploring a wider range of indicators of distress. The original Ricketts and Sawhill concept was built around underclass theory and that limited the possibilities. 6 There are other measures, like the other risk factors identified in the Foundation s 1999 Kid s Count essay, that may be worth examining in this context. 6 Again see Jargowsky, 1997, pps

16 Severe Distress and Concentrated Poverty: Trends for Neighborhoods in Casey Cities and the Nation 15 REFERENCES Ellen, Ingrid G., and Margery Austin Turner Does Neighborhood Matter? Assessing Recent Evidence. Housing Policy Debate 8: Frey, William H Melting Pot Suburbs: A Census 2000 Study of Suburban Diversity. Washington, D.C.: Center on Urban and Metropolitan Policy, The Brookings Institution. Jargowsky, Paul A Poverty and Place; Ghettos, Barrios, and the American City. New York: Russell Sage Foundation Stunning Progress, Hidden Problems: The Dramatic Decline of Concentrated Poverty in the 1990s. Washington D.C.: Center on Urban and Metropolitan Policy, The Brookings Institution. Katz, Bruce, and Alan Berube Cities Rebound Somewhat. The American Enterprise 13(4): 47. Kingsley, G. Thomas, and Kathryn L. S. Pettit Concentrated Poverty: A Change in Course. Washington, D.C.: The Urban Institute. May. O Hare, William and Mark Mather The Growing Number of Kids and Severely Distressed Neighborhoods: Evidence from the 2000 Census. Baltimore: The Annie E. Casey Foundation Tatian, Peter A Census CD Neighborhood Change Database (NCDB) Data Users Guide: Short Form Release. Washington, D.C.: The Urban Institute. Wilson, William Julius The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy. Chicago: University of Chicago Press.