1031, 1032; Tr. 4/25/00, at 8.) Dr. Armor convincingly demonstrated that
the differences reflected in standardized test scores for
African-American and white students in the District are not based on
racial discrimination in the schools but rather are the result of
socioeconomic factors, including poverty. Dr. Armor's conclusions are
supported by decades of social science research, and his education and
experience as a sociologist give him the knowledge necessary to interpret
achievement data and give meaning to the results of his regression
Dr. Armor evaluated the academic achievement of students in the
District to determine whether any vestiges of past segregation remain.
(Commw. Ex. 1031; Tr. 4/25/00, at 9-13.) To do so, he compared average
Iowa Test for Basic Skills ("ITBS") scores for black and white students
in reading and math. Dr. Armor also compared ITBS scores for black and
white students at various grade levels with the same students' first
grade and kindergarten ITBS scores. Finally, he compared the ITBS scores
for blacks and whites taking into account each students' participation in
the free or reduced lunch program. (Commw. Ex. 1031; Tr. 4/25/00, at
Dr. Armor included early test scores and free lunch participation as
variables in his regression, because those are valid measures of
socioeconomic factors for children in the District. Kindergarten and
first grade scores reflect the students' skill level as developed under
conditions in the students' homes and communities and without influence
from the schools. Free lunch status is a commonly used indicator of,
poverty. (Tr. 4/25/00, at 78.)
Socioeconomic factors and family background have been widely recognized
as influential in students' academic performance.*fn18 (Tr. 4/25/00, at
13-16.) Differences in the socioeconomic backgrounds of black and white
students are reflected nationally in an achievement gap. This gap appears
at all ages in virtually every school system throughout the United States
in reading, mathematics and science. A comparison of national scores on
achievement tests highlights this disparity. (Commw. Ex. 1031; Tr.
4/25/00, at 17.)
Taking socioeconomic factors into account, Dr. Armor found that the
achievement gap between black and white students in the WHSD is somewhat
smaller than the national achievement gap.*fn19 (Commw. Ex. 1033; Tr.
4/25/00, at 11, 20-21.)
When the effects of poverty and first-grade or kindergarten tests
scores are removed, the remaining or unexplained portion of the
achievement gap is very small. (Commw. Exs. 1031, 1034, 1035; Tr.
4/25/00, at 26-27.)
In 10 out of 12 analyses (reading and math for six grade cohorts), the
residual achievement gap between black and white students is not
(Commw. Ex. 1031, 1036; Tr. 4/25/00, at 26-27.
In the two cases where the unexplained achievement gap is statistically
significant, Dr. Armor demonstrated that the achievement gap is caused by
unmeasured family background differences that continue to operate as a
child moves through school. (Commw-. Ex. 1031, 1036; Tr. 4/25/00, at
In preparing his report, Dr. Armor relied in part upon socioeconomic
data prepared by Dr. Vijai B. Singh, Associate Chancellor and Professor
of Sociology at the University of Pittsburgh, who testified as an expert
for the defendants in socioeconomic and demographic data.*fn20
Dr. Singh reported stark differences between the socioeconomic
circumstances of the African-American and white residents of the Woodland
Hills School District. His data shows that:
• 46.5% of the students enrolled in WHSD in the
1997-98 school year were from low-income families.
(Commw. Exs. 1089, 1099 & 1116; Tr. 4/17/00, at
• The rate of unemployment for blacks in the WHSD
area was almost three times the rate of unemployment
among whites (16% compared to 5.8%). (Commw. Exs.
1089, 1093 & 1110; Tr. 4/17/00, at 163.)
• The number of black households in WHSD with
incomes below $25,000 was one and a half times
greater than the numbem of white households.
(Commw. Exs. 1089, 1094 & 1111; Tr. 4/17/00, at
• In 1989, 41% of blacks and 14.6% of whites in
WHSD were living in poverty, defined as an income
below $12,674 for a family of four. In other words,
theme were almost three times as many black families
with children living in poverty as white families.
(Commw. Exs. 1089, 1095 & 1112; Tr. 4/17/00. at
• The rate of blacks without high school degrees
was 1.38 times that of whites. (Commw. Exs. 1089,
1098 & 1115; Tr. 4/17/00, at 167-68.) The rate
of bachelor's degrees among whites was three times
greater than the rate among blacks. (Commw. Exs.
1089, 1107 & 1125; Tr. 4/17/00. at 175-76.)
• The rate of single-parent families among blacks
was 3.35 times that of white families. (Commw. Exs.
1089, 1100 & 1117; Tr. 4/17/00, at 169.)
• Between 1993 and 1995. the average rate of
teen-age pregnancies among blacks in WHSD was almost
6 times greater than the rate among whites (57.45
per thousand compared to 9.90 per thousand).
(Commw. Exs. 1089, 1101 & 1118; Tr. 4/17/00, at
• On average between 1995 and 1997, the rate of
live births to single mothers for blacks was almost
3 times that of white mothers. (Commw. Exs. 1089,
1103 & 1120; Tr. 4/17/00, at 172.) In the same
period, black mothers in WHSD were almost twice as
likely as white mothers to give birth to a
low-weight baby (13.3% to 7.3%). (Commw. Exs. 1089,
1104, 1121; Tr. 4/17/00, at 173), and the infant
mortality among blacks was 1.25 times greater than
the rate among whites (13.54 per thousand compared
to 10.79 per thousand),
(Commw. Exs. 1089, 1102 & 1119; Tr. 4/17/00, at
African-American and white students, then, may be subject to
substantially different influences outside the schools, and the
environment outside the schools for many black students is likely to be
less conducive to academic success.
Research has determined that students from socioeconomic backgrounds
which are less conducive to academic achievement will score somewhat
lower, on average, on standardized tests than students whose environments
are more advantageous for academic success. (Commw. Ex. 1031; Tr.
4/25/00, at 20-27.)
Plaintiffs countered Dr. Armor's testimony with a statistical analysis
of test score data by Dr. de Leeuw, who criticized Dr. Armor's
methodology and conclusions.
Dr. de Leeuw was highly critical of the use of early test scores as a
proxy for socioeconomic status ("SES"). He noted that SES is only one of
many factors which affects early test scores. (Tr. 5/2/00, at 46.) We
agree with Dr. de Leeuw that many different factors likely affect early
test scores; however, none of those other, unnamed factors can be
school-related, since the students being tested in kindergarten or in
first grade are just beginning their school careers. We are in accord
with Dr. Armor's conclusion that the District is not responsible for the
achievement gap that already exists when children first enter school.
(Tr. 4/25/00, at 17.)
Dr. de Leeuw also disagreed with Dr. Armor's analysis of the
achievement gap over time, which essentially demonstrated that the
achievement gap present when children enter the Woodland Hills School
District is still present when they are tested in the sixth grade.
(Commw. Ex. 1034; Tr. 4/25/00, 23, 24.) Dr. de Leeuw's conclusion from
this data is that the schools have clearly failed to provide a uniform
education to all or their students. (Tr. 5/2/00, at 54). One possible
explanation, he suggested, is that children with low test scores are
placed in lower level courses. (Tr. 5/2/00, at 50.)
The record in this case, however, clearly establishes that children in
the Woodland Hills elementary schools have been taught in homogeneous,
detracked classrooms at least since 1990. (1990 R & R at 32.)
Moreover, it is likely that the children who have tested the lowest have
been provided with remedial tutoring and other compensatory programs. We
agree with Dr. Armor's testimony that the achievement gap is caused by
family and socioeconomic factors outside the control of the school
district. (Tr. 4/25/00, at 17, 24.)
Conclusions of Law:
Although our remedies have not addressed test scores directly, they
provide significant evidence that a racial disparity in achievement
unfortunately remains. Courts considering scores on standardized tests as
evidence of an achievement gap, however, generally accord them little
weight in determining the question of unitary status, and the Supreme
Court has suggested that measuring student achievement against national
norms may not be the proper test to use. Missouri v. Jenkins, 515 U.S. at
100-01, 115 S.Ct. 2038. This case, though, stands in a slightly different
posture from at least one of the cases cited by the defendants.*fn21
inasmuch as we have previously determined that the disparity in
achievement is causally related to the constitutional violation in this
case. Thus we conclude that the defendants have the burden of showing
that the remaining gap is caused by external factors, and that the
District has in fact done all that is practicable to narrow the divide.
Other federal courts considering the achievement gap in the context of
school desegregation have determined that
socioeconomic factors and family background explain most if not all of
the difference in students test scores. See Jenkins, 515 U.S. at 120, 115
S.Ct. 2038, 132 L.Ed.2d at 89 ("numerous external factors beyond the
control of [schools] affect minority student achievement"); People Who
Care, 111 F.3d at 537 ("The social scientific literature on educational
achievement identifies a number of other variables besides poverty that
explain differences in scholastic achievement, such as the educational
attainments of the student's parents and the extent of their involvement
in their children's schooling"); Coalition, 90 F.3d at 766 n. 17
(sub-standard academic achievement is "the product of many complex
socio-economic factors"); Manning v. School Bd. of Hillsborough County,.
Supp.2d 1277, 1334 (M.D.Fla. 1998) ("differences in academic performance
. . . are the result of socioeconomic factors unrelated to the
schools"); Capacchione, 57 F. Supp.2d at 272 ("numerous external factors
beyond the control of a school district affect educational outcomes");
Flax, 725 F. Supp. at 330 ("Poor achievement scores are often an
incidence of poverty and family environment, matters not remediable by a
school desegregation plan").
The Constitution does not require that a school district eliminate
every trace of racial disparity between standardized test scores.
Although the Constitution requires that all of its
citizens have equal access to the pursuit of
education, and that they be given equal breaks while
attending school, it does not insist that they all
finish even. The proper test under the Constitution is
equality of opportunity, not of results.
Coalition, 90 R.3d at 766.
We must also recognize that there are limits to what court-ordered
remedies can accomplish. As the Supreme Court stated in Swann:
We are concerned in these cases with the elimination
of discrimination inherent in the dual school
systems, not the myriad factors of human existence
which can cause discrimination in a multitude of ways
on racial, religious, or ethnic grounds. The target of
the [desegregation cases] was the dual school system.
The elimination of racial discrimination in public
schools is a large task and one that should not be
retarded by efforts to achieve broader purposes lying
beyond the jurisdiction of school authorities. One
vehicle can carry only a limited amount of baggage.
Swann, 402 U.S. at 22, 23, 91 S.Ct. 1267.