BLACK EDUCATIONAL ADVANTAGE IN T H E I N N E R C I T Y 1, 2

Katrina Bell McDonald and Thomas A. LaVeist

Analyses of contemporary urban poverty have overwhelmingly focused on the plight of African Americans residing in inner-city neighborhoods (Gregory, 1998). Preoccupation with the racial dimension of poverty is driven primarily by the increasingly disproportionate representation of blacks in segregated inner cities where they are said to suffer the deleterious effects of social dislocation (Auletta, 1982; Farley and Frey, 1994; Glasgow, 1981; Litcher, 1988; Massey, Gross, and Shibuya, 1994; Venkatesh, 1994). The gravity and persistence of these trends has led many Americans to equate poverty with African-American existence (Devine and Wright, 1993; Fainstein, 1995; Hartigan, 1997; Katz, 1995). While Wilson's thesis (1987), drawn upon heavily in these analyses, concerns the black urban context, it broadly demonstrates how poverty and social disadvantage can concentrate in segregated inner-city communities to severely threaten the life chances of persons living in those communities. Wilson argues that the compounding of multiple disadvantage in these urban communities, not race per se, produces dislocation of poor inner-city blacks from mainstream social institutions. That these structural constraints are more numerous and pronounced in communities predominated by blacks helps to explain the observations of large racial differences in social status overall. His thesis also implies "that whites exposed to similar socio-environmental conditions as poor urban blacks would suffer a similar fate. However, the vast body of research examining the relationship between urban poverty and individual life chances fails to adequately test this idea, further feeding the perception that whiteness is protective even under the bleakest of social circumstances. The high degree of residential segregation that persists despite many other advances in race relations undermines the ability of researchers to study disadvantaged blacks and whites residing in communities ecologically and economically on par (Jencks and Mayer, 1990; Tiggs, Browne,

26

The Review of Black Political Economy/Summer 2001

and Green, 1998). Perhaps this is why few researchers have conducted race comparisons in inner-city settings. One exception is the work of Krivo and Peterson who overcome this problem by studying Columbus, Ohio, "a city with a relatively high prevalence of black and white disadvantage" (1996, p. 623). Their study of crime rates across extremely disadvantaged Columbus neighborhoods lends strong support to Wilson's thesis, though it does not directly evaluate differences in criminal outcomes at the individual level. While inner-city neighborhoods, frequent sites of poverty research, are grossly segregated by race, predominantly white areas and predominantly black areas may cluster together in close proximity and by virtue of their poverty suffer a similar fate. Clusters of neighborhoods, what Wilson refers to generally as "heterogenous groupings of families and individuals" (1987, p. 8), constitute communities where residents of different racial backgrounds regularly interact, willingly or unwillingly, over resources and institutions common to them all (e.g., schools, social service facilities, and workplaces). An additional feature of poor communities is the grim prospect of residents breaking free to more mainstream educational, occupational, and economic activity (MacLeod, 1995). Thus, clusters of neighborhoods provide an alternative context in which to study poor urban dwellers and racial difference in life outcomes. The purpose of this article is to employ this broader conceptualization of urban community in examining differences in life outcome among black and white children in the inner city. Specifically, we seek to determine the extent to which urban disadvantage differentiates young adult educational outcomes by race under such conditions. Our focus on inner-city educational outcomes rests on the substantial research linking educational success to occupational and economic outcomes in later life and on concerns over persistent racial disparity in college enrollment and completion (Farley, 1997). Interestingly, past studies of racial differences in educational attainment suggest that controlling for urban-rural residence and family background factors, African American youth tend to attain higher levels of education than their white counterparts (Alexander et al., 1987; Hauser, 1990; Hauser and Phang, 1993; Jencks et al., 1972; Margo, 1990; Mare, 1981; Pallas, 1984). This research, however, is not designed to highlight the experience of children from inner-city families where the effects of urban poverty are said to be most severe. Geographic residence, when controlled for, is intended to capture a relative advantage of urban-dwellers over rural-dwellers (Bauman, 1998); inner-city residence is not made distinct from other

McDonald and LaVeist

27

urban residence. Proceeding from a Wilsonian framework, we predict that given their socioeconomically poor geographic context, black and white children of this study will experience low levels of educational attainment relative to the nation at large. More importantly, given their shared inner-city community context, no significant racial differences in educational attainment are anticipated. We use longitudinal data from the Pathways to Adulthood study (Hardy et al., 1997) to construct and estimate models of educational attainment for a sample of black and white children born between 1959 and 1965 to inner-city women in Baltimore, Maryland. The study subjects represent a cohort of children who graduated from high school and began their college careers during the late 1970s and throughout much of the 1980s, a period marked by a setback in educational progress for African Americans as a whole (Baker and V~lez, 1996; Washington and Newman, 1991), but particularly for young adults of the inner city. They began their lives in neighborhoods within an area of East Baltimore, extending approximately 10 blocks in each direction around the Johns Hopkins University Hospital. This is an area of the city that experienced increased urban deterioration and social dislocation over the past 40 years. For the most part, Pathways respondents remained in these neighborhoods through young adulthood. It is the fact of their close neighborhood proximity, their births to inner-city women, and their similar levels of low family income that leads us to characterize these children as "similarly situated." Education and Race It is well known that the educational attainment of African Americans improved substantially during the 1960s and 1970s due in large part to the policies emerging from the civil rights movement. The 1960s witnessed the enactment of major civil rights laws and Supreme Court decisions promoting the equalization of opportunity across the races (Farley, 1977, 1997; Katz, 1995; Lewin-Epstein, 1986; NCES, 1995). During this period, racial convergence in educational attainment occurred on many levels. For those age 25-35 years, the gap between the high school graduation rates of white and non-whites was 25.1 percentage points in 1960; by 1980 that gap was reduced to only 7.6 percentage points (Statistical Abstract of the United States, 1990). Day and Curry (1996) report that by the mid-1990s black high school graduation rates virtually matched that of whites. Impressive gains for blacks were also made in college comple-

28

The Review of Black Political Economy/Summer 2001

tion rates, rising from 11.8 to 25.3 over the same period. The improvement in college participation among blacks is particularly significant given that "the baccalaureate may be the most significant education credential" in the educational process and "a prerequisite to most of the better jobs" (Rothman, 1999, p. 195). While Baker and V~lez (1994) report a widening of the racial gap in college participation around the mid-1980s, black improvements in secondary and postsecondary educational attainments have continued through the 1990s. In a recent study, Bauman (1998) cites several studies confirming that net family background factors, African-American rates of educational attainment outpace that of whites. These studies emphasize familial context and inter-generational processes believed to have a large impact on a child's educational and economic status in later life. At least two studies show that this may have been true for many decades (see Margo, 1990; Perlmann, 1987). Bauman's own analysis, covering time periods from 1973 to 1991, produces similar results, though he finds differences between the sexes and across educational transitions (i.e., attending high school, competing high school, and attending college).

Urban Poverty As the 1950s and 1960s mark some important changes in the educational landscape, so do they mark intense reinvestment by policy makers in the elimination of urban poverty. (American sociologists had been commenting and researching the issue for more than a half century [DuBois, 1899; Drake and Cayton, 1962].) Concerns over the growing number and concentration of impoverished inner-city dwellers--a new form of urban poverty--taking shape in American central cities were hotly debated (Devine and Wright, 1993; Katz, 1995). The launching of Lyndon Johnson's "War on Poverty" in 1964 ultimately led to focused attention on those trapped in inner-city areas plagued by persistent racial segregation, out-of-wedlock childbearing, crime, and socioeconomic dislocation of various kinds. Poverty programs were designed, among other things, to bridge the gap between the educational achievements of the poor and other groups and between racial groups (Dworkin and Dworkin, 1999; Kantor and Brenzel, 1993). Interest in the consequences of being impoverished heightened dramatically in the 1980s, and the impact of poverty on the socioeconomic outcomes of inner-city dwellers has been a major focal point ever since (Kasarda, 1993; Massey, Gross, and Shibuya, 1994; Wilson, 1987).

McDonald and LaVeist

29

The harshness of poor inner-city life during the formative years has been shown, among other things, to significantly depress the number of years of schooling children will be able to complete (Crane, 1991; BrooksGunn et al., 1993; Corcoran et al., 1989; Datcher, 1982). Yet given that the poverty rate among African-American children is far greater than that of white children, and counter-intuitively black children tend to fair better educationally than white children, there is a surprising lack of research that disaggregates educational data by race and economic status allowing for even simple comparisons between poor inner-city blacks and whites. Kantor and Brenzel, (1993) for example, do not present data showing racial differences among the poor students they observe, though they state that despite the educational gains made by African Americans by the 1960s, "increasingly large numbers of poor and minority students" living in central cities failed to complete school in the decades that followed (1993, p. 380). Devine and Wright argue that disaggregation of educational data'"'would inevitably yield a different and more troubling portrait" for blacks as compared to whites, given that "city schools in poverty neighborhoods cannot escape the pathology that surrounds them" (1993, p. 160). In another study, Hochschild (1996) concludes that by 1988 a larger percentage of poor black children lacked adequate proficiency in basic reading and math skills than poor white children, a fact she attributes to the greater likelihood that black children live in poor neighborhoods with low-quality schools. Still, neither of these two reports speak to the potential commonality of educational experience between poor urban blacks and poor urban whites living in close proximity to one another. Finally, Bauman (1998) found that across the board, increased educational attainment was found to be associated with urban versus rural residence; but he does not explore racial differences among urban residents or among those who are poor. Continued systemic racial discrimination against blacks contributes to the general assumption that disparate educational outcomes are ubiquitous and are at their extreme in poor inner-city neighborhoods (Rothman, 1999). It is generally assumed that whites raised in racially segregated communities with concentrated poverty will realize greater socioeconomic success than blacks raised under similar circumstances. The factors conditioning the experiences of poor inner-city populations are said to "depress" the educational potential of blacks more severely than that of whites (Alexander, 1997). Yet there is little direct evidence that race exerts significant educational drawbacks for blacks relative to whites in this context.

30

The Review of Black Political Economy/Summer 2001

The research reported in this article seeks to advance the study of urban poverty and individual life outcomes in several ways. It seeks to free scholars from the data limitations that typically confine studies of urban poverty and race to'"'apple-and-orange" comparisons, i.e., poor blacks in racially segregated poor black neighborhoods and poor whites in socioeconomically heterogenous white neighborhoods. Bound within the catchment area for the Pathways to Adulthood study, the subjects of this investigation have resided in inner-city neighborhoods very close in proximity to one another, offering us an opportunity to observe the educational progress of black and white children with a common poor innercity residential base. Still, we examine variation in neighborhood context, investigating the role of racial segregation in predicting educational outcomes. Moreover, by employing the Pathways data we provide a unique longitudinal view of these children as they mature over a 30-year period. Most importantly, this research explores the educational attainment of poor inner-city white populations, a group whose life chances have attracted little attention. DATA AND METHODS

Data

Pathways to Adulthood is a longitudinal multi-generational family study conducted at the Johns Hopkins University School of Hygiene and Public Health and School of Medicine. A detailed description of the study design can be found in Hardy et al. (1997). The first generation (G1) consists of pregnant women who came to the Johns Hopkins obstetric clinic for prenatal care during the years 1959 to 1965 (Time 1). These women were randomly selected to participate in the study. Most G1 women lived within the 10-square block immediately surrounding the Johns Hopkins University Hospital in East Baltimore (extending about 10 city blocks in each direction), a low-income inner-city community that continues to have one of the highest poverty rates of any community in the city. Twenty-three percent of the 2,307 G1 respondents were white and the remaining 77 percent were African American. G1 respondents were followed during their pregnancy and their babies (G2s) were enrolled in the study at birth. For this early phase of the project, G2s were followed until their eighth birthday (Time 2, 1967 to 1973). Because some Gls gave birth to more than one child (either twins or more than one pregnancy during the Time 1 field period), there are more G2 re-

McDonald and LaVeist

31

TABLE 1 Interview Status of 1992 Pathways G2 Respondents at Follow-Up % of Those Located

Sample Size

% of Original Sample (n = 2,694)

for Follow-Up (n = 2,220)

Located:

Completed interview

1,758

65.3

79.2

Proxy interview

71

2.6

3.2

Deceased

88

3.3

4.0

Refused

135

5.0

6. I

In field at end of study

157

5.8

7.1

Unavailable

11

0.4

0.5

Sub-totals

2,220

82.4

100.0 *

474

17.6

N/A

2,694

100.0

100.0

Other:

Not located

Totals

1 Percentages o f 100 may actually total slightly more due to rounding.

spondents (2,694) than GI respondents. Data collected on G2s included psychological, developmental, and neurological examinations along with assessments of the socioeconomic circumstances of the child' s family. At Time 3 (1992-1994) a questionnaire was administered to G1 and G2 respondents. Seventy percent of the Time 3 interviews were conducted in person and the remaining interviews (persons who moved outside of the Baltimore area) were conducted by telephone. At Time 3, G2 respondents ranged in age from 27 to 33. In Table 1 we present the results of efforts to locate and interview the adults who were study participants as children (G2s). The table shows that although we had not been in contact with the study participants for more than 25 years, we were able to locate 82.4 percent (2,220) of the G2 respondents (column 2). The third column of Table 1 shows that of those G2s that could be

32

The Review of Black Political Economy/Summer 2001

located, we completed interviews with 79.2 percent, conducted proxy interviews with 3.2 percent, and obtained information indicating that 88 (4%) had died. Thus, we obtained data on 1,829 G2 respondents (and their mothers), amounting to 82.4 percent of the located respondents. Of these, only the first G2s born to the Gls during the study period (including twins) were retained for the study reported here, a total of 1,555. Bias due to sample attrition, i.e., whether G2s who could not be interviewed at Time 3 may have differed significantly from those who were interviewed, was considered by comparing the distributions of several variables from Time 1 and Time 2. To examine this concern we conducted analysis to test the differences among three groups of respondents: (1) respondents who were re-interviewed, (2) respondents who were located, but not re-interviewed, and (3) respondents who were not located. The groups were compared with regard to sex, race, reading level at age 7, mother's income at birth of child, mother's education at birth of child, and whether on public assistance at birth of child. This analysis found no differences for the selected variables comparing respondents who were located but not interviewed with respondents who were interviewed. There are, however, a few differences between respondents who were not located compared with respondents who were interviewed. Respondents who were not located were more likely to be female. Their mothers were more likely to subsist below 85 percent of official poverty at time 1. And, they were more likely to have been on public assistance at birth. However, although we acknowledge this source of potential bias we believe it is of lesser importance within the context of the present study because we did not find significant differences in sample attrition by race. Additionally, because of the high proportion of respondents that were interviewed at Time 3 (86.2% of G2's were located), the characteristics of the total sample (interviewed and not interviewed combined) were similar to those interviewed. Measures

In the present analysis we use data from each of the three time periods of the study. The independent variables are chosen from among those identified in previous analyses of the Pathways data, including the analysis of educational attainment (Hardy et al., 1997), and found elsewhere to influence educational trajectory (Baumann, 1998; Kalmijn and Kraaykamp, 1996; Pascarella, 1985). These capture the G2's family, neighborhood, and individual characteristics. Family background factors are mother's

McDonald and LaVeist

33

marital status, educational attainment, age, and poverty status at the G2's birth, and whether the father was present in the household at G2's age 8. All but the latter are measured at Time 1. Marital status is specified as a binary variable scored 1 for married and 0 for all others. Educational attainment is also dichotomized, where 1 is for high school graduate and 0 for less than high school graduate. Age is measured in years. Poverty status is measured continuously and is assessed as the ratio of the annualized income to the federally-determined poverty level for the G1 's family in 1960. The presence of the father in the household is scored 1 and absence is scored 0. Tract-based measures are employed as proxies for neighborhood measures (White, 1987). These are the G2's neighborhood poverty status, the G2's residential mobility, and segregation status. Poverty status is measured both during childhood and in early adulthood (using the 1970 and 1990 census, respectively). Measuring poverty status at both of these points, rather than just during childhood, allows us to construct the residential mobility measure to help determine whether movement in and/or out of area neighborhoods over time may account for any potential significant effect on educational outcome. Following Wilson (1987), a census tract was determined to be poor if 20 percent or more of the tract's residents lived below the federal poverty level. Thus, residence in a poor neighborhood is scored 1 and residence in a non-poor neighborhood is scored 0. Ideally, we would have liked to have known the degree to which residential mobility (or non-mobility) occurred outside of the original East Baltimore area neighborhoods; however, we were unable to make such a determination from the data at hand. Nor are we comfortable with the quality of the 1960 census tract information in the dataset, which could at least have allowed us to measure neighborhood characteristics earlier in the G2's lives. Alternatively, residential mobility is represented by four dummy variables, measuring the change in neighborhood poverty status from young childhood (1970) to adulthood (1990): remained in or returned to a poor neighborhood (stayed poor), moved from a non-poor neighborhood to a poor one (downward mobility), moved from a poor to a non-poor neighborhood (upward mobility), or remained in or returned to a non-poor neighborhood (stayed non-poor.) In the regression analyses, stayed poor is the reference category. The criteria for defining census tracts based on racial composition (segregation) is wildly inconsistent across previous analyses. We first considered a standard set by Lee and Wood (1991), who defined "predominantly black" as greater than 89 percent. Their justification is as

34

The Review of Black Political Economy/Summer 2001

follows: "...because of the high level of racial residential segregation in most cities--and a corresponding scarcity of tracts that are neither all white nor all black--sample size might be jeopardized by the imposition of stricter guidelines" (p. 24). However, our area-bounded inner-city tract distribution was such that the majority-black tracts (51% or more) were numerous and clustered more at the far end of the scale; so we adjusted the guidelines to better reflect our tract distribution. We define a tract as predominantly black if its 1970 population was 95 percent African American or greater (scored 1). The relevant and available individual-level factors employed for this analysis are the G2's sex and reading level at age 7. Sex was measured at Time 3 (1992-93) and is specified as 1 for female and 0 for male. Measured at Time 2, reading level for the Pathways study was assessed using the Wide Range Academic Tests (WRAT) (Jastak and Jastak, 1965), determined as reading above, at, or below grade level. Each of these is represented by a dummy variable; reading below grade level is the reference category. The race of the G2s was recorded at Time 3 (1992-93). For race, a score of 1 refers to African American and 0 white. The dependent variables were also recorded at Time 3 for the G2s. High school completion is measured dichotomously where high school graduate is scored 1 and less than high school graduate is scored 0. Likewise, college completion is scored 1 for college graduate and 0 for less than college graduate. Only those at risk for college entry (i.e., high school graduates) are analyzed for college completion.

Analytic Strategy In this research it is hypothesized that controlling for family background, neighborhood, and individual factors, inner-city black and white children's educational attainment will not differ significantly. The first phase of the analysis was designed to determine whether and to what extent the subjects of this research differed by their race status in important and relevant ways, despite their being situated geographically in the same inner-city community as young children and subject to similar social dislocations. Factors determined not to significantly differentiate the racial groups are excluded from the multivariate logistic regression procedures, which are organized around those factors that potentially advantage one racial group over the other in regards to educational attainment. In the second phase, two logistic regressions are performed. The first of

McDonald and LaVeist

35 TABLE 2 Study Sample Characteristics

Variable

Total

White

Black

n

I, 555

285

1266

% High school graduate***

79.8

62.1

84.0

% Co[lege graduate*

12.7

6.2

13,7

GI mean age at G2 birth*

25.0

24.1

25.2

% GI high school graduate at G2 birth***

29.0

12.0

33.0

% GI married at G2 birth***

66.9

87.0

62.4

1.8

1.6

1.9

61.8

73.0

59.4

37.9

1.4

44.9

% G2 in poor neighborhoods, 1970"**

35.7

15.8

40,2

% G2 in poor neighborhoods, 1990"**

32.6

7.4

38.3

% Stay poor***

20.0

4.2

23.5

% Downward* * *

12.6

3.2

14.8

% Upward

15.7

11.6

16.7

% Stay non-poor***

51.7

81.1

45.0

53.1

49.5

53.9

% Reading above grade level

4.5

5.3

4.3

% Reading at grade level

82.9

80.4

83.5

% Reading below grade level at age 7

11.3

12.6

l 1.0

G2 educational uttainment at young adulthood."

(7,2 Family Characteristics:

GI mean family poverty index (% of poverty) at G2 birth* % Father living in household at G2 age 8*** G2 Neighborhood characteristics: % G2 in predominantly black neighborhoods, 1970'**

Residential mobility of G2, 1970-1980:

Individual Characteristics for {72 in childhood: % Female

Reading skillsat age 7

*** p = .001 **p = .01 *p = .05, significance o f b l a c k / w h i t e difference. *** p = .001 * * p = .01 *p = .05, s i g n i f i c a n c e o f b l a c k / w h i t e difference.

36

The Review of Black Political Economy/Summer 2001

these is a model to estimate the effect of race on high school and college completion and the next to control for family background characteristics. This is followed by an expansion of the analysis in a third model to include neighborhood characteristics. RESULTS

Table 2 presents some general characteristics of the sample. Chi-square or t-test analyses were conducted to test for race differences for all of the independent variables. The table shows that although the respondents originally lived in a limited section of East Baltimore, there nonetheless are differences in educational outcomes and social status, African-American G2s graduated from both high school and college at significantly higher rates than did white G2s. Where about 84 percent of blacks graduated high school and 13.7 percent of those graduates completed college, the figures for whites are only 62.1 percent and 6.2 percent, respectively. Mothers for both groups were about the same age when the G2s were born (25 years of age on average), though the mothers of whites were significantly younger than those of blacks (24.1 versus 25.2 years, respectively). Like their children, the mothers of the African-American G2s were more likely to be high school graduates (33%) than their white counterparts (12%), and they were less likely to have been married when they gave birth (62.4% versus 87%). Family income status was 180 percent of poverty on average when the G2s were born. The families of white G2s, however, subsisted on significantIy less than that of black G2s (160% versus 190%, respectively). Fathers of white G2s were significantly more likely (73%) than those of black G2s (59.4%) to have resided with these children when they were 8 years old_ On average, the proportion of G2s residing in poor neighborhoods during young childhood is about 36 percent, but the proportion for African-Americans is significantly higher than their white counterparts (about 40%). A similar and significant disproportionality exists at Time 3, about 25 years later. Over time, white G2s remained in non-poor neighborhoods at a rate of about 81 percent, a rate significantly higher than that for their black counterparts. (Recall that "remained" in a non-poor or poor neighborhood also includes having returned to a neighborhood of that same status.) African-Americans were significantly more likely than whites to have remained in poor neighborhoods and to have experienced downward and upward residential mobility. And as expected, nearly 45 percent of black G2s resided in predominantly black neighborhoods dur-

McDonald and LaVeist

37

TABLE 3 Logistic Regression of Race Differences in Achievement in Young Adulthood Educational Attainment (Models 1-2) Independent Variable n

MODEL 1 H.S. Grad College Grad 1,555 1,241

MODEL 2 H_S, Grad College Grad 1,555 1,209

Race (1 = black)

1.18"**

1.08"**

(3.26)

.88** (2.42)

G1 married at G2 birth (1 = yes)

GI H.S. grad at G2 birth (1 = yes)

G1 age at G2 birth

n

1,555

1,241

G1 family poverty mdex at G2 birth

(2.95)

(2.18)

-.007

.25

(.99)

(1.28)

1.62"**

.68***

(5.06)

(1.98)

.004

.006

(1.00)

(I.Ol)

1,555

1,209

-.007

.06

(.99)

(1.06)

.51"**

Father living in HH at G2 age 8 (l=yes)

.78"

.29

(1.67)

(1.34)

Constant

1.06"**

-2.28***

1_46"**

-2.54***

~2

64.87***

9.21"*

159.59"**

34.666***

*** p = .001 **p = .01 *p = .05 for significance of black/white difference. Numbers in parenthesis are odds ratios. ing young childhood, where only about 1 percent of whites did so. No significant racial differences exist for either sex or reading skills level. Table 3 presents the results of logistic regression analysis for Model 1, which tests for race differences in high school and college graduation. This serves as the baseline equation. The model shows a significant effect of race on educational attainment. The relative rate (in parentheses) for Model 1 shows that African American G2s were 3.26 times more likely to graduate from high school and 2.42 times more likely to complete college compared to whites. As explained earlier, Models 2 4 control for those family, neighbor-

38

The Review of Black Political Economy/Summer 2001 TABLE 4 Logistic Regression of Race Differences in Achievement in Young Adulthood Educational Attainment (Model 3) MODEL Independent Variable

H.S. Grad

3

College Grad

n

1,303

1,042

Race (1 = black)

1.43'**

.78*

(4.20)

(2.17)

GI married at G2 birth (1 = yes)

GI H.S. grad at G2 birth (1 ~yes)

GI age at G2 birth

GI family poverty index at G2 birth

-.01

.21

(.99)

(1.23)

1.62"**

.57**

(5.07)

(1.77)

.006

.01

(1.Ol)

(1.Ol)

-. 03

.04

(.97)

(1.04)

Father living in HH at G2 age 8 (I = yes)

.40*

.05

(l.~)

(i.o5)

.08

.29

G2 neighborhood characteristics: Predominantly black (1 = yes)

-

(.97)

(I .34)

Residential mobility:' Downward

Upward

Stay non-poor

.12

.07

(I.13)

(1,07)

.51"

.93"

(1.66)

(2.52)

,93" **

1.67" **

(2.53)

(5.31)

Constant

1.64'**

2.36"* *

X2

161.98'**

70.13"**

*** p = .001 * * p = .01 * p = .05 f o r s i g n i f i c a n c e o f b l a c k / w h i t e d i f f e r e n c e . N u m b e r s i n p a r e n t h e s i s are o d d s ratios.

McDonald and LaVeist

39

hood, and individual-level factors found to significantly differentiate the race groups and that may account for or even reverse this race effect. Adding controls for G2 family background characteristics (Table 3) diminishes the effect of race somewhat, but African Americans are still significantly more likely to graduate from high school and college than their white counterparts. Several other independent variables have significant effects; neither the mother's marital status, her age, nor the G2 family's poverty statuses are among them. As anticipated, G1 educational attainment proved to significantly predict G2 educational attainment for both groups. The effect is larger for high school completion; if the G1 had herself graduated from high school at G2's birth, the G2 was 506 percent more likely to graduate from high school. Fathers' presence in the home at age 8 proved to be significant only for graduation from high school. Table 4 reports the results of further expanding our model of educational attainment to include indicators of the socio-environmental context in which white and black G2s lived at Time 2 (Model 3). The high collinearity anticipated and observed between neighborhood poverty status and residential mobility required choosing one or the other for inclusion in the regression models; we chose the latter of these. Inconsistent with past studies, being brought up in a predominantly black neighborhood--generally thought to be synonymous with a poor neighborhood-does not appear to have a significant impact on educational outcomes for the inner-city residents of this study. What does appear to matter for some is having moved from a poor to a non-poor environment or having remained in neighborhoods with poverty levels below 20 percent over time. Upward mobility has a significant positive effect on high school and college completion relative to remaining in a poor neighborhood. Those G2s who remain in non-poor inner-city neighborhoods also were significantly more likely to graduate from high school and complete college. Most importantly, these data show that the racial differences in educational attainment are consistent with previous models. Furthermore, models adjusting for living in racially segregated and impoverished areas does not eliminate the effect of race: net other factors, African-Americans were 420 percent and 217 percent more likely to graduate from high school and college, respectively, compared to white G2s.

DISCUSSION Children born to inner-city families during the early 1960s were clearly not a socioeconomically homogeneous group, even with regards to a limited, sociologically relevant set of individual, family, and neighborhood characteristics. Yet, the blacks and whites among them were bound by the confines, literally and figuratively, of poor urban existence. We have argued that both groups were born and raised in closely adjacent inner-city neighborhoods plagued by a host of familiar inner-city problems, such as poor schools, stagnant economic growth, and drug infestation. Whether the interaction of the two groups in that space was willing or unwilling, that interaction formed the basis for similarly situated urban existence and potentially similar long-term life outcomes. Thus, it is fair to say that our analysis first proceeds by giving greater weight to the unobserved homogeneity within this population than to the observed differences. Still, our statistical analysis gives proper attention to addressing those differences we did observe. The sample profile shows that the African American G2s were more likely to be disadvantaged by their mothers' unmarried status at their birth and the absence of their fathers in childhood. In addition, the families of the black G2s were significantly worse off than those of whites in regards to income, and their neighborhoods were significantly more likely to be impoverished and sharply segregated. White G2s suffered fewer of the disadvantages we studied. They were significantly more likely, however, to be disadvantaged by their mothers' failure to complete high school and young maternal age. Our bivariate logistic regression analysis, which showed African Americans substantially more likely than whites to graduate from high school and college, could easily be explained by the relatively better socioeconomic status of the G1, but only if "better" is qualified to mean more educated and slightly older at the G2's birth. The African Americans in this sample were clearly disadvantaged by a greater number of factors than were whites, and these other considerations would favor whites' attaining higher educational statuses. Moreover, statistical controls for G2's age and education did not eliminate the race difference in educational attainment. In controlling factors potentially advantaging one racial group over another, we expected the significance of race on each outcome would be eliminated. Instead, race continued to exert a significant influence on educational outcome'independent of potentially advantaging factors. The inner-city African Americans of this cohort were consistently more likely

McDonald and LaVeist

41

than their white counterparts to reach two important educational milestones. It is important to acknowledge that this analysis is based on a sample from East Baltimore, Maryland, and as such our findings may not generalize to other settings. While East Baltimore is not unlike other inner-city communities in the United States, there may be subtle differences between the study setting and other communities. As such, it would be important to replicate our findings in other settings.

Black Advantage The apparent advantage of inner-city black youth in achieving educational mobility is revealed in at least two other studies. Hardy et al. (1997), who also employ the Pathways to Adulthood data, conduct research to "identify conditions in the lives of children, born in a poor, urban environment that lead to...economic independence in young adult life" (p. 81). In the report of their multivariate analysis they state that blacks are "in a more favorable position" with regard to graduating from high school (p. 84). Some years prior, Alexander et al. (1982) found in a national study of the Class of 1972 that "among low-status youth, blacks are more likely to complete college" (1982, p. 317). In neither case were possible explanations for black advantage posited, something we find curious given that these findings defy what most social theorizing on race and urban poverty purports but not surprising in light of what we contend is a long-held preoccupation with the arrested social advancement of black "urban underclass" populations. Baumann (1988) argues that because black educational advantage was evident before the enactment of the 1960s affirmative action programs (and therefore before the birth of our cohort), this change in social policy could not have been "the major force behind" (p. 524) findings such as those reported here. It is certainly plausible, however, that the anti-discrimination legislation of 1964 and the executive order mandating the development of affirmative action strategies to address past and present discrimination of minorities played an important role. The African-American study participants were among the first to reap the benefits of these acts, aimed at dismantling state-enforced racial segregation across all domains of government control, including public schools and universities. A near frenzy erupted around efforts to rectify past discrimination against blacks (and other groups) and to help fashion a new U.S. workforce inclusive of educated African Americans. In the words of Bowen and

42

The Review of Black Political Economy/Summer 2001

Bok, "these efforts soon bore fruit" (1998, p. 7). Black educational attainment improved steadily and dramatically for the following two decades. Further, our research suggests that affirmative action programs may have had dramatically positive consequences for poor inner-city African Americans. Vigorously seizing new-found equal opportunity by and on behalf of minority populations may have ultimately placed black rates of educational attainment ahead of that of whites in the inner city. Additionally, we believe that Wilson (1987) has painted a fairly accurate picture of what African-American urban life was like prior to the 1980s. For decades blacks survived largely by maintaining strong ties with one another within and across socioeconomic strata, providing the less advantaged access to mainstream American society and its socioeconomic rewards (McDonald, 1997). Such activity would qualify as part of the African-American cultural tradition favoring education and of blacks' desire to do as well if not better than whites, as evidenced by Margo (1990) and Perlmann (1987, 1988), among others. It is likely that affirmative action bolstered the infrastructure of social capital that already existed among blacks, helping to advance inner-city poor black children educationally. Social capital embodies the coping abilities of parents and extended kin under socioeconomically stressed conditions, their ability to make connections with social mechanisms that help advance their children's success, and their investment in future generations as a means to escape poverty (Mulsow and Murray, 1996). The ability and desire of African Africans to advance each other, particularly once it was clear that they had a legal right to educational opportunity, may have been sorely underestimated or even unrecognized by social scientists, the architects of affirmative action, and the nation at large (Gregory, 1998).

White Disadvantage The preoccupation of social researchers and policy makers with the urban black underclass has resulted in a paucity of work on the plight of poor urban white populations, thereby handicapping our ability to explain the relative educational disadvantage observed here for white innercity children. That which we have learned is based primarily on ethnographic or journalistic studies of rural and/or post-1980 urban communities. Among those who have addressed this problem, it is agreed that at its root is a poor understanding of race. In Hacker's words, "...neither sociologists nor journalists have shown much interest in depicting poor whites as a 'class.' For whites, poverty tends to be viewed as atypical or

McDonald and LaVeist

43

accidental. Among blacks, it is seen as a natural outgrowth of their history and culture" (1992, p. 100). This binary logic of either/or social designations precludes us from making whites part of the underclass and from imparting race as a factor in the condition of poor urban whites. Past research on poor whites offers a few clues, however. Relative to minority families of color, poor inner-city white families apparently attracted few community and political advocates. Despite their poverty, their whiteness (racial privilege) effectively excluded them from the affirmative action debates of the 1960s, rendering the plight of the white urban poor virtually invisible in scholarly discourses. According to Gibson, "Advocates for other disadvantaged groups do not represent the generalized interests of the poor because all exist in a racial hierarchy which makes identification of a common white enemy an important organizing principle" (1997, p. 387). Gibson also suggests that more privileged whites did not come to their aid in an act of racial solidarity. Instead, they actively distanced themselves from poor whites using tactics similar to those used against blacks--by constructing "physically differentiated stereotypes" (p. 379) to distinguish the poor's behaviors and values from those of their own. This privileged white practice of distancing and devaluating poor whites would have also made it difficult for inner-city families to erect and maintain an infrastructure of social capital that would have supported their children's educational advancement. Contrary to popular wisdom, it appears that some inner-city white children are more likely to spurn the achievement ideology than are their African-American counterparts, "the reigning social perspective that sees American society as open and fair and full of opportunity" (MacLeod, 1995, p. 3). In a rare glimpse into a low-income urban neighborhood, MacLeod finds that the central factor contributing to the "fundamental incongruity" (p. 128) between the predominantly white ("Hallway Hangers") and the predominantly black ("Brothers") male teen cliques he observed is race: Whereas poor blacks have racial discrimination to which they can point as a cause of their family' s poverty, for the Hallway hangers to accept the achievement ideology is to admit that their parents are lazy or stupid or both. Thus, the achievement ideology not only runs counter to the experiences of the Hallway Hangers, but is also a more serious assault on their self-esteem...The severe emotional toll that belief in the achievement ideology exacts on poor whites relative to poor blacks explains why the Hallway hangers dismiss the ideology while the Brothers validate it. (p. 129)

44

The Review of Black Political Economy/Summer 2001

In essence, what M a c L e o d p r o c l a i m s is that to be white and p o o r is far worse psychologically than to be b l a c k and poor: whites are not e x p e c t e d to "fail." W h e n rates o f "failure" a m o n g certain groups o f whites b e g i n to rival blacks, those whites are in a sense " c o l o r e d " black in the public imagination, a notion implicit in Charles M u r r a y ' s c o m m e n t a r i e s (1986, 1993). T o be white and o f a p u r p o r t e d elite racial g r o u p offers little c o m f o r t and confers no elite status to those whose families h a v e experienced little or no social mobility. On the other hand, discrimination battles w o n by earlier generations o f p o o r blacks is a source o f pride for innercity black children, and instead o f indicting their parents for their p o o r s o c i o e c o n o m i c status they m o r e willingly e m b r a c e the disciplined w a y their parents support their educational a d v a n c e m e n t . NOTES 1. Address correspondence to Thomas A. LaVeist, The Johns Hopkins University, 624 North Broadway, Baltimore, Maryland 21205. E-mail: [email protected] We thank Karl Alexander and Mark King for their very helpful comments. 2. "Black" and "African American" are used interchangeably in this article. REFERENCES Alexander, Karl L. 1997. "Public Schools and the Public Good.'" Social Forces, 76 (1): 1-20. Alexander, Karl L., Doris R. Entwisle, and Nader Kabbani. 1999. "The Dropout Process in Life Course Perspective: Risk Factors at Home and School." Paper presented at the annual meetings of the Southern Sociological Association, Nashville, Tennessee. Alexander, Karl L., Aaron M. Pallas, and Scott Holupka. 1987. "'Consistency and Change in Educational Stratification: Recent Trends regarding Social Background and College Access." Pp. 161-185 in Research in Social Stratification and Mobility, vol. 6, edited by Robert V. Robinson. Greenwich, Conn: JAI. Alexander, Karl L., Cornelius Riordan, James Fennessey, and Aaron M. Pallas. 1982. "Social Background, Academic Resources, and College Graduation: Recent Evidence from the National Longitudinal Survey." American Journal of Education 90:315-333. Auletta, Ken. 1982. The Underclass. New York: Random house. Baker, Therese L. and William Vblez. 1996. "Access to and Opportunity in Postsecondary Education in the United States." Sociology of Education 69: 82101. Bauman, Kurt J. 1998. "Schools, Markets, and Family in the History of AfricanAmerican Education." American Journal of Education 106:500-531. Bowen, William G. and Derek Bok. 1998. The Shape of the River: Long-term Consequences of Considering Race in College and University Admissions. Princeton, NJ: Princeton University. Brooks-Gunn, Jeanne, Greg J. Duncan, Pamela Kato Klebanov, and Naomi Sealand.

McDonald and LaVeist

45

1993. "Do Neighborhoods Influence Child and Adolescent Development?" American Journal of Sociology 99 (2): 353-95. Corcoran, Mary, Roger Gordon, Deborah Laren, and Gary Solon. 19899 "Effects of Family and Community Background on Men's Economic Status" (Working Paper 2896). National Bureau of Economic Research, Cambridge, MA. Crane, Jonathan, 1991. "The Epidemic Theory of Ghettos and Neighborhood Effects on Dropping Out and Teenage Childbearing." American Journal of Sociology 96: 1226-59. Datcher, Linda. 1982. "Effects of Community and Family Background on Achievement." The Review of Economics and Statistics 64: 32~t1. Day, Jennifer and Andrea Curry. 1996. Educational Attainment in the U.S.: March 1995. Current Population Reports P-20, no. 489. Washington, DC: Government Printing Office. Devine, Joel A. and James D. Wright9 1993. The Greatest of Evils: Urban Poverty and the American Underclass. New York: Aldine De Gruyter. Drake, St. Clair and Horace R. Cayton. 1962. Black Metropolis: A Study of Negro Life in a Northern City. New York: Harper and Row. DuBois, W. E. B. 1899. The Philadelphia Negro. Philadelphia: Publication for the University9 Dworkin, Anthony Gary and Rosalind J. Dworkin. 1999. The Minority Report: An Introduction to Racial, Ethnic, and Gender Relations. Fort Worth, TX: Harcourt Brace College Publishers9 Fainstein, Norman. 1995. "Black Ghettoization and Social Mobility." Pp. 123-14 I in the Bubbling Cauldron: Race, Ethnici~, and the Urban Crisis, edited by Michael Peter Smith and Joe R. Feagin. Minneapolis, MN: University of Minnesota Press. Farley, Reynolds. 1977. "Trends in Racial Inequalities: Have Gains of the 1960s Disappeared in the 1970s?" American Sociological Review 42: 189-208, 9 1984. Blacks and Whites: Narrowing the Gap? Cambridge, MA: Harvard University Press. 9 1997. "Racial Trends and Differences in the United States 30 Years after the Civil Rights Debate." Social Science Research 26: 235-262. Farley, Reynolds and William H. Frey. 1994. "Changes in the Segregation of Whites from Blacks During the 1980s: Small Steps Toward a More Integrated Society." American Sociological Review 59: 23-45, Gibson, Jane W. 1996. "The Social Construction of Whiteness in Shellcracker Haven, Florida." Human Organization 55 (4): 379-89 Glasgow, Douglas G. 1980. The Black Underclass: Poverty, Unemployment, and Entrapment of Ghetto Youth9 San Francisco: Jossey-Bass Publishers. Gregory, Steven. 1998. Black Corona: Race and the Politics of Race in an Urban Community. Princeton, NJ: Princeton University Press. Hardy, Janet B., Sam Shapiro, E. David Mellits, Elizabeth A. Skinner, Nan Marie Astone, Margaret Ensminger, Thomas LaVeist, Rosemary Baumgardner, and Barbara H. Starfield. 1997. "Self-suffiiciency at Ages 27 to 33 Years: Factors Present Between Birth and 18 Years That Predict Educational Attainment among Children Born to Inner-city Families." Pediatrics 99 (1): 80-87. Hartigan, John, Jr. 1997. "Green Ghettos and the White Underclass." Social Research 64 (2): 339-659 Hauser, Robert M. 1990. "The Decline in College Entry among African Americnas: Findings in Search of Explanations." Working Paper No. 90-20. University of Wisconsin, Madison, Center for Demography and Ecology.

46

The Review of Black Political Economy/Summer 2001

Hochschild, Jennifer L. 1996. Facing Up to the American Dream: Race, Class, and the Soul of the Nation. Princeton, NJ: Princeton University Press. Jastak, Joseph Florian and Jastak, S. R. 1965. "The wide range achievement test (WRAT)." Wilmington, DE: Guidance Associates. Jencks, Christopher and Susan E. Mayer. 1990. "The Social Consequences of Growing up in a Poor Neighborhood." Pp. 111-86 in Inner-city Poverty in the United States, edited by Laurence E. Lynn, Jr., and Michael G. H. McGeary. Washington, DC: National Academy Press. Jencks, Christopher S., Marshall Smith, Henry Akland, Mary Jo Bane, David Cohen, Herbert Gintis, Barbara Heyns, and Stephen Michelson. 1972. Inequality: A Reassessment of the Effect of Family and Schooling in America. New York: Harper & Row. Kalmijn, Matthijs and Gerbert Kraaykamp-Gerbert. 1996. "Race, Cultural Capital, and Schooling: An Analysis of Trends in the United States." Sociology of Education 69 (1): 22-34. Kantor, Harvey and Barbara Brenzel. 1993. "Urban Education and the 'Truly Disadvantaged': The Historical Roots of the Contemporary Crisis, 1945-1990." Pp. 366-402 in The "Underclass" Debate: Views from History, edited by Michael Katz. Princeton, NJ: Princeton University Press. Kasarda, John D. 1992. "Inner-City Concentrated Poverty and Neighborhood Distress: 1970 to 1990." Housing Policy Debate 4:253-302 Katz, Michael B. 1995. Improving Poor People: The Welfare State, the "Underclass," and Urban Schools as History. Princeton, NJ: Princeton University Press. Krivo Luren J. and Ruth D. Peterson. 1996. "Extremely Disadvantaged Neighborhoods and Urban Crime." Social Forces 75 (2): 619-50. Lee, Barrett A. and Peter B. Wood. 1991. "Is Neighborhood Racial Succession Place-specific?" Demography 28:21-40. Lewin-Epstein, Noah. 1986. "Effects of Residential Segregation and Neighborhood Opportunity Structure on the Employment of Black and White Youth." The Sociological Quarterly 27 (4): 559-70. Lichter, Daniel T. 1988. "Racial Differences in Underemployment in American Cities." American Journal of Sociology 93 (4): 771-92. MacLeod, Jay. 1995. Ain't No Makin' It: Aspirations and Attainment in a LowIncome Neighborhood. Boulder, CO: Westview Press. Mare, Robert D. 1981. "Market and Institutional Source of Educational Growth." Research in Social Stratification and Mobility 1: 20545. Margo, Robert A. 1990. Race and Schooling in the South, 1880-1950: An Economic Histo~. Chicago: University of Chicago Press. Marks, Carole C. 1995. "Separate Societies: Negotiating Race and Class in the '90s." Sociological Focus 28 (1): 49-61. Massey, Douglas S., Andrew B. Gross, and Kumiko Shibuya. 1994. "Migration, Segregation, and the Geographic Concentration of Poverty." American Sociological Review 59: 425-45. McDonald, Katrina Bell. 1997. "Black Activist Mothering: A Historical Intersection of Race, Gender, and Class." Gender & Society 11: 773-95. Mulsow, Miriam H. and Velma McBride Murry. 1996. "Parenting on the Edge: Economically Stressed, Single, African American Adolescent Mothers." Journal of Family Issues 17 (5): 704-21. Murray, Charles. 1986. "White Welfare, White Families, 'White Trash.'" National Review 38 (March 28).

McDonald and LaVeist

47

Murray, Charles. 1993. "The Coming White Underclass." Wall Street Journal (October 29). National Center for Education Statistics (NCES). 1995. "Findings from the Condition of Education, 1994: The Educational Progress of Black Students." NCES 95-765. U.S. Department of Education. Pallas, Aaron. 1984. "The Determinants of High School Dropout." Ph.D. dissertation, The Johns Hopkins University. Pascarella, Ernest T. 1985. "Racial differences in factors associated with bachelor's degree completion: A nine-year follow-up." Research in Higher Education 23 (4): 351-73. Perimann, Joel. 1987. "A Piece of the Educational Pie: Reflections and New Evidence on Black and Immigrant Schooling since 1880." Sociology of Education 60: 54-61. Perlmann, Joel. 1988. Ethnic Differences: Schooling and Social Structure among the Irish, Italian, Jews, and Blacks in the American Ci~, 1880-1935. Cambridge: Cambridge University Press. Rothman, Robert A. 1999. Inequality and Stratification: Race, Class, and Gender. Upper Saddle River, NJ: Prentice Hall. Statistical Abstract of the United States. 1990. U.S. Government Printing Office. Tiggs, Leann M., Irene Brown, and Gary P. Green. 1998. "Social Isolation of the Urban Poor: Race, Class, and Neighborhood Effects on Social Resources." The Sociological Quarterly 39 (1): 53-77. Venkatesh, Sudhir Alladi. 1994. "Getting Ahead: Social Mobility among the Urban Poor." Sociological Perspectives 37 (2): 157-82. Washington, Valora and Joanna Newman. 1991. "Setting Our Own Agenda: Exploring the Meaning of Gender Disparities among Blacks in Higher Education." Journal of Negro Education 60 (1): 19-35. White, Michael J. 1987. American Neighborhoods and Residential Differentiation. New York: Russell Sage Foundation. Wilson, William Julius. 1987. The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy. Chicago: University of Chicago Press.