Two Cities' Tracking and Within-School Segregation

by Jeannie Oakes - 1995

Evidence from two school systems whose ability grouping and tracking systems were scrutinized in 1993 in conjunction with school desegregation cases demonstrates how grouping practices can create within-school segregation that discriminates against black and Latino students. In both cases, grouping practices created a cycle of restricted opportunities and diminished outcomes. (Source: ERIC)

Evidence from two school systems whose ability grouping and tracking systems were subject to scrutiny in 1993 in conjunction with school desegregation cases demonstrates how grouping practices can create within-school segregation and discrimination against African-American and Latino students. In both school systems, tracking created racially imbalanced classes at all three levels—elementary, middle, and senior high, with African-American or Latino students consistently over represented and white and Asian students consistently underrepresented in low-ability tracks in all subjects. Neither district’s placement practices created classrooms with a range of measured student ability and achievement in classrooms sufficiently narrow to be considered homogeneous “ability groups,” and African-American and Latino students were much less likely than whites or Asians with comparable scores to be placed in high-track courses. These disproportionate lower-track placements worked to disadvantage minority students’ achievement outcomes. Whether students began with relatively high or relatively low achievement, those who were placed in lower-level courses showed lesser gains over time than similarly situated students placed in higher-level courses. In both systems, grouping practices created a cycle of restricted opportunities and diminished outcomes, and exacerbated differences between African-American and Latino and white students.

Since the 1920s, most elementary and secondary schools have tracked their students into separate “ability” groups designed for bright, average, and slow learners and into separate programs for students who are expected to follow different career routes after high school graduation. Tracking has seemed appropriate and fair, given the way psychologists have defined differences in students’ intellectual abilities, motivation, and aspirations. Tracking has seemed logical because it supports a nearly century-old belief that a crucial job of schools is to ready students for an economy that requires workers with quite different knowledge and skills. According to this logic, demanding academic classes would prepare bright, motivated students heading for jobs that require college degrees, while more rudimentary academic classes and vocational programs would ready less able and less motivated students for less-skilled jobs or for post–high school technical training. With the development early in the century of standardized tests for placement, most people viewed a tracked curriculum with its “ability-grouped” academic classes as functional, scientific, and democratic—an educationally sound way to accomplish two important tasks: (1) providing students with the education that best suits their abilities, and (2) providing the nation with the array of workers it needs.

Despite its widespread legitimacy, there is no question that tracking, the assessment practices that support it, and the differences in educational opportunity that result from it limit many students’ schooling opportunities and life chances. These limits affect schoolchildren from all racial, ethnic, and socioeconomic groups. However, schools far more often judge African-American and Latino students to have learning deficits and limited potential. Not surprisingly, then, schools place these students disproportionately in low-track, remedial programs.

Educators justify these placements by pointing out that African-American and Latino children typically perform less well on commonly accepted assessments of ability and achievement. Moreover, conventional school wisdom holds that low-track, remedial, and special education classes help these students, since they permit teachers to target instruction to the particular learning deficiencies of low-ability students. However, considerable research demonstrates that students do not profit from enrollment in low-track classes; they do not learn as much as comparably skilled students in heterogeneous classes; they have less access than other students to knowledge, engaging learning experiences, and resources.1 Thus, school tracking practices create racially separate programs that provide minority children with restricted educational opportunities and outcomes.

In what follows, I will illustrate these points with evidence from two school systems whose ability grouping and tracking systems have been subject to scrutiny in the past year in conjunction with school desegregation cases. The first system, Rockford Public Schools, in Rockford, Illinois (previously under an interim court order), was the target of a liability suit brought by a community group, The People Who Care. Among other complaints, the group charged the school system with within-school segregation through ability grouping and discrimination against the district’s nearly 30 percent African-American and Latino students. The second system, San Jose Unified School District, in San Jose, California, approached the court hoping to be released from its desegregation order of 1985. The plaintiffs in the San Jose case argued, among other things, that the district had used its ability-grouping system to create within-school segregation and, thereby, circumvented the intent of the court order with regard to its approximately 30 percent Latino student population. I analyzed data about the grouping practices in both these cities, prepared reports for the court, and testified. The San Jose system reached a settlement prior to the formal hearing date. The Rockford system was found liable by the court.

To shed light on the grouping practices in these two systems, I conducted analyses and reported my conclusions about tracking and ability grouping practices around several questions:

1. Does the school system employ tracking and/or ability grouping? If so, what is the specific nature of these practices?

2. Does the system’s use of tracking and/or ability grouping create racially imbalanced classrooms?

3. Does the system’s use of these grouping practices reflect sound, consistent, and educationally valid considerations?

4. Are the racial disproportionalities created by the system’s ability grouping practices explained by valid educational considerations?

5. What are the consequences of the system’s grouping and tracking practices for the classroom instructional opportunities of Latino children?

6. What are the consequences of the system’s grouping and tracking practices for the educational outcomes of Latino children?

7. Does the system have the necessary support and capacity to dismantle racially identifiable tracking and create heterogeneously grouped classrooms?

I addressed these questions with analyses using data specific to the two school systems. These data were gathered from a variety of sources: district and individual school curriculum documents (e.g., curriculum guides, course catalogs, course descriptions, etc.); school plans; computerized student enrollment and achievement data; prior reports prepared by court monitors; and depositions taken from school district employees in the course of the discovery process.2

Several analytic methods were applied to these data, all of which had been used in prior published research on tracking and ability grouping. In both systems, I used statistical methods to calculate the achievement range within each track; the distribution of students from various ethnic groups into various tracks; and the probability of placement of students from each ethnic group whose prior achievement “qualified” them for various tracks. In San Jose, but not in Rockford, I was also able to calculate rather precisely the impact of track placement on achievement gains of students with comparable prior achievement. I applied “content analysis” techniques to district and school curriculum documents in order to classify courses into various track levels, determine placement criteria and processes, and identify curricular goals, course content, and learning opportunities. These documents constitute official district policy statements about the levels and content of the districts’ programs and courses, as well as the criteria and procedures by which students enroll in various programs and courses.

The scope of possible analyses was limited, more in Rockford than in San Jose, by a lack of some essential data. Even so, the available data permitted quite comprehensive analyses of many aspects of the district’s grouping practices. They provided a clear picture of tracking and ability grouping in the two systems, and enabled me to place the district’s practices in light of national research.


Grouping practices and their effects on minority children were remarkably similar in both systems. Both systems used tracking extensively. At most grade levels and in most academic subject areas at nearly all schools, educators assigned students to classes based on judgments about students’ academic abilities. The schools then tailored the curriculum and instruction within classes to the students’ perceived ability “levels.” The districts’ tracking “systems” were not only very comprehensive (in terms of the subject areas and grade levels that are tracked); they were also quite rigid and stable. That is, the districts tended to place students at the same “ability level” for classes in a variety of subject areas, and to lock students into the same or a lower ability-level placement from year to year.


In both school systems, tracking had created racially imbalanced classes at all three levels—elementary, middle, and senior high. This imbalance took two forms: (1) White (and Asian, in San Jose) students are consistently over represented, and African-American and Latino students are consistently underrepresented, in high-ability classes in all subjects; (2) in contrast, African-American or Latino students were consistently over represented, while white and Asian students were consistently underrepresented, in low-ability tracks in all subjects.


The criteria used to assign students to particular tracks were neither clearly specified nor consistently applied. Accordingly, neither district’s tracking policies and practices could be construed as the enactment of valid educational purposes; neither did either district present an educational justification for the racial imbalance that results from tracking. Moreover, my analyses demonstrate quite clearly that neither district’s placement practices—practices that result in racially imbalanced tracked classrooms— could be justified by a racially neutral policy of creating classrooms that are distinctly different from one another in terms of students’ academic ability or achievement. To the contrary, neither district had enacted ability grouping and tracking in ways that narrow the range of measured student ability and achievement in classrooms sufficiently so that these classrooms can be considered bona fide ability groups.

Both school systems honored parent requests for students’ initial track placements and for subsequent changes. This policy undermined the basis of student assignments in either “objective” measures of students’ abilities or more subjective professional judgments. Making matters worse, not all parents were informed about tracking practices or about parents’ right to influence their children’s placements. Specifically, African-American and Latino parents had less access than others to this knowledge.

Additionally, teacher and counselor recommendations at the critical transitions between elementary and middle school and between middle and high school included a formal mechanism to take into account highly subjective judgments about students’ personalities, behavior, and motivation. For example, the screening process for gifted programs usually began with a subjective teacher identification of potentially gifted children, who were then referred for formal testing. Such referrals were often based on subjective judgments about behavior, personality, and attitudes.


The “theory” of tracking argues that, to facilitate learning, children should be separated into groups so that they may be taught together with peers of similar ability and apart from those with higher or lower abilities. But in both school systems, classes that were supposed to be designated for students at a particular ability level actually enrolled students who spanned a very wide range of measured ability. These ranges demonstrate quite dramatically that in both Rockford and San Jose racially imbalanced tracked classes have borne little resemblance to homogeneous ability groups—even though they have been labeled and treated as such by schools. While the mean scores in each of the tracks followed expected patterns—with average achievement score for students in the low track less than average score for students in the standard or accelerated tracks—the extraordinarily broad range of achievement in each of the three tracks makes clear how far these classes are from being homogenous ability groups. In sum, the district’s practices do not represent what tracking advocates would claim is a trustworthy enactment of a “theory” of tracking and ability grouping.

For example, at one Rockford middle school, the range of eighth-grade reading scores in Honors English (31–99 National Percentile [NP])overlapped considerably with the range in Regular English (1–95 NP), which overlapped considerably with the range in Basic English (1–50 NP). At one of the senior highs, the math scores of tenth-graders in the normal progress college prep math track (26–99 NP)overlapped considerably with those in the slow progress college prep courses (1–99), and both overlapped considerably with the scores of those in non–college preparatory classes (1–99). I found similar patterns of large, overlapping ranges of qualifying scores throughout the system.

The same was true in San Jose. For example, sixth-graders placed in a low-track mathematics course demonstrated abilities that ranged all the way from rock-bottom Normal Curve Equivalent (NCE) achievement scores of 1 to extraordinarily high scores of 86. Even more striking, sixth-graders in standard-track math classes had achievement scores that spanned the entire range, from NCE scores of 1 to 99. And, while sixth-graders in accelerated courses had a somewhat more restricted ability range, they too scored all the way from 52 to 99 NCE scores. I found similar patterns in a number of other subjects in most middle and senior high school grades.


As a group, African-American and Latino students scored lower on achievement tests than whites and Asians in Rockford and San Jose. However, African-American and Latino students were much less likely than white or Asian students with the same test scores to be placed in accelerated courses. For example, in San Jose, Latino eighth-graders with “average” scores in mathematics were three times less likely than whites with the same scores to be placed in an accelerated math course. Among ninth-graders, the results were similar. Latinos scoring between 40 and 49, 50 and 59, and 60 and 69 NCEs were less than half as likely as their white and Asian counterparts to be placed in accelerated tracks. The discrimination is even more striking among the highest scoring students. While only 56 percent of Latinos scoring between 90 and 99 NCEs were placed in accelerated classes, 93 percent of whites and 97 percent of Asians gained admission to these classes.

In Rockford’s tracks and class ability levels, the groups of higher track students whose scores fell within a range that would qualify them for participation in either a higher or lower track (i.e., their scores were the same as students in the lower track) were consistently “whiter” than groups of students whose scores fell within that same range but were placed in the lower track. In a number of cases, Rockford’s high-track classes included students with exceptionally low scores, but rarely were these students African Americans. Conversely, quite high scoring African Americans were enrolled in low-track classes; again, this was seldom the case for high-scoring whites. For example, in 1987, none of the African-American students who scored in the top quartile (75–99 NP) on the California Assessment Program (CAP) reading comprehension test at two of Rockford’s large high schools were placed in high-track English, compared with about 40 percent of top-quartile whites who were enrolled in the high track at those schools. In contrast, at three of the system’s senior high schools a small fraction of white students who scored in the bottom quartile (1–25 NP) were in high-track classes, while no similarly low-scoring African Americans were so placed. At two other senior highs, while some top-quartile African Americans were placed in Honors English, many more top-scoring African Americans were in the basic classes. No low-scoring whites were so placed. I found similar patterns in other subjects at the district’s high schools.

I found other striking examples of racially skewed placements in Rockford’s junior highs. For example, at one, the range of reading comprehension scores among eighth-graders enrolled in Basic English classes was from the first to the seventy-second national percentile. Of these, ten students scored above the national average of 50 NP. Six of the highest scoring, above-average students were African-American, including the highest achieving student in the class. One other of the above-average students was Latino.

In both San Jose and Rockford, placement practices skewed enrollments in favor of whites over and above that which can be explained by measured achievement.


In both school systems, African-American and Latino students in lower-track classes had fewer learning opportunities. Teachers expected less of them and gave them less exposure to curriculum and instruction in essential knowledge and skills. Lower-track classes also provided African-American and Latino students with less access a whole range of resources and opportunities: To highly qualified teachers; to classroom environments conducive to learning; to opportunities to earn extra “grade points” that can bolster their grade point averages; and to courses that would qualify them for college entrance and a wide variety of careers as adults.


Not only did African-American and Latino students receive a lower quality education as a result of tracking in San Jose and Rockford; their academic achievement suffered as well. In Rockford the initial average “achievement gap” (i.e., the difference in group mean achievement scores) between white and African-American and/or white and Latino students (i.e., that found on district-administered achievement tests in first grade) did not diminish in higher grades. To the contrary, eleventh-graders exhibited gaps somewhat larger than first graders. For example, on the 1992 Stanford Achievement Test in reading comprehension, the gap between African-American and white first-graders was 25 percent; that between African-American and white eleventh-graders was 30 percent. Undoubtedly more telling, at the time of the seventh-grade test—probably the last point before considerable numbers of lower-achieving minority students drop out of school—the achievement gap between African Americans and whites had grown considerably wider, to 36 percent. A similar pattern was found in students’ raw scores in reading comprehension and mathematics for grades 1–6 on the 1992 Stanford Achievement Test. Here, the reading achievement gap between African-American and white students at first grade was .88 of a standard deviation and grew to .99 by grade 6. The Latino-white gap grew from .67 to .70 over the same grades. In math, the African American–white gap grew from .87 top 1.01; in contrast, the Latino-white gap dropped from .98 to .79. Clearly, the district’s tracked programs failed to close the minority-white gap between average group scores. Neither did these practices correct the overrepresentation of black and Latino students in the group of lowest scoring students in the district. For example, in 1992, 37 percent of the first-grade children scoring between the first and the twenty-fifth national percentiles in reading comprehension on the Stanford Achievement Test were African-American; at seventh-grade, the percentage of African Americans in this low-scoring group had risen to 46 percent, and by grade 11 (following a disproportionately high incidence of dropping out by low-achieving African-American students), African-American students still made up 35 percent of this group. Neither did student placements in various instructional programs enable minority students to rise into the group of district’s highest achievers. In fact, the proportion of minority students in the highest achieving group of students dropped quite precipitously. For example, in 1992, 10 percent of the first-grade children scoring between the seventy-fifth and the ninety-ninth national percentiles in reading comprehension on the Stanford Achievement Test were African-American; at seventh grade, the percentage of African Americans in this high-scoring group had dropped by half, to only 5 percent (28 in number); this low proportion was also found at grade 11 (even though the actual number of students, twenty, was smaller).

Rockford’s grouping practices that created racially identifiable classrooms and provided unequal opportunities to learn (with fewer such opportunities provided to minority students) did not serve a remedial function for minority students. To the contrary, these practices did not even enable minority students to sustain their position, relative to white students, in the district’s achievement hierarchy.

In San Jose, better data permitted me to analyze the impact of track placement on individual students over time. Students who were placed in lower-level courses—disproportionately Latino students—consistently demonstrate lesser gains in achievement over time than their peers placed in high-level courses. For example, among the students with pre-placement math achievement between 50 and 59 NCEs, those who were placed in a low-track course began with a mean of 54.4 NCEs, but lost an average of 2.2 NCEs after one year, and had lost a total of 1.9 NCEs after three years. Students who scored between 50 and 59 NCEs and were placed in a standard-track course, by contrast, began with a mean of 54.6 NCEs, gained 0.1 NCEs after one year, and had gained 3.5 NCEs after three years. The largest gains were experienced by students who were placed in an accelerated course, who began with a mean of 554 NCEs, gained 6.5 NCEs after one year, and had gained a total of 9.6 NCEs after three years.

These results are consistent across achievement levels: Whether students began with relatively high or relatively low achievement, those who were placed in lower-level courses showed lesser gains over time than similarly situated students who were placed in higher-level courses.


The findings from my analyses of San Jose and Rockford support disturbing conclusions about tracking and within-school segregation and discrimination. The districts’ tracking system pervade their schools. The harm that accrues to African Americans and Latinos takes at least three demonstrable forms : (1) unjustifiable disproportionate and segregative assignment to low-track classes and exclusion from accelerated classes; (2) inferior opportunities to learn; and (3) lower achievement. In both systems, grouping practices have created a cycle of restricted opportunities and diminished outcomes, and exacerbated differences between African-American and Latino and white students. That these districts have not chosen to eliminate grouping practices that so clearly discriminate against its African- American and Latino children warrants serious concern and strong remedial action.


Is it technically possible or politically feasible to abandon these discriminatory practices in San Jose, Rockford, or other school systems that are like them? The two systems are currently charged with making significant progress toward that end.

Both Rockford and San Jose school systems have considerable technical capacity to reform their placement practices so that they teach all children in heterogeneous settings, including the gifted, for part or all of the school day in most or all core academic courses. Conspicuous examples of successful heterogeneous grouping exit currently in San Jose schools. Much of the professional expertise and some of the support structures needed to implement such practices district wide are already in place. Moreover, in both systems administrative and teaching staff demonstrate considerable knowledge of the harms of tracking, and ample ability to implement educationally sound alternatives.

Further, both districts are situated in a national and state policy environment that encourages the development and use of such alternatives. For example, such national policy groups such as the National Governors’ Association and federally supported efforts to create “national standards” in each of the curriculum areas all recommend against tracking. In California, the State Department of Education’s major policy documents on the reform of K–12 schooling (It’s Elementary, Caught in the Middle, and Second to None) and the state’s subject matter frameworks caution schools about problems with tracking and strongly recommend that they not use it.3 Similar state-led initiatives promote heterogeneity in Illinois—for example, the state’s involvement in middle-school reform and its adoption of the Accelerated Schools model.

However, racially mixed school systems that have tackled this issue around the country have experienced considerable difficulty creating alternatives. Amy Stuart Wells and I are currently studying ten such schools.4 While each has made considerable progress toward integrated classrooms and a more even distribution of educational opportunities, most have been the target of considerable fear and anger. As with the nation’s experiences with between-school segregation, the pursuit of court sanctions against tracking and ability grouping may be critical to ensuring educational equality. However, like that earlier effort, remedies are neither easily specified nor readily accepted.

Cite This Article as: Teachers College Record Volume 96 Number 4, 1995, p. 681-690 ID Number: 49, Date Accessed: 10/26/2021 3:59:25 AM

Purchase Reprint Rights for this article or review