Abstract
The Coleman Report concluded that if students of color attended schools with more white students, they were likely to garner significantly better achievement results. Several contextual factors should be considered, however, before we make strong conclusions about the strength of the effect of racial and ethnic minority students attending school with white students. This paper makes a case for considering the necessity of both the “material” and “sociocultural” domains of schooling if we are to move toward a more holistic understanding of “school effects” on students’ academic well-being. Based on a case-study analysis of four high-performing, diverse schools in different regions, using ethnographic observations and random student surveys, this paper offers evidence that students who hail from similar socioeconomic and family backgrounds but who attend schools with different sociocultural contexts have disparate academic experiences. The findings draw our attention to new directions to consider as we seek to understand how educational outcomes vary depending on the school's organizational culture.
What does equality of educational opportunity mean? Does it mean the same opportunity to get an education? Or does it mean an opportunity to get the same education? Or the opportunity to be educated up to the level of one’s capabilities and future occupational prospects? Or the opportunity to learn whatever one needs to know to develop one’s own peculiar potentialities? Is only racially integrated education equal, irrespective of whether lack of integration is intentional or accidental? Is equality of educational opportunity a moral as well as a mathematical concept? (Thompson 1968).
More than six decades after the landmark Supreme Court decision of Brown v. Board of Education, and fifty years since the 1966 publication of the seminal report Equality of Educational Opportunity (or the Coleman Report), the United States continues to face the daunting task of equalizing educational opportunity to enhance the life chances of its low-income students and those of its racial and ethnic minority students. Inequality, stubbornly rooted in the foundation of the nation, continues to entrench historically disadvantaged groups at the bottom of the educational and mobility ladders in the United States. Significant disparities abound across social class, race, and ethnicity in the United States, affecting low-income and African American, Latino/a, and Native American student populations disproportionately more than other groups (Duncan and Murnane 2011).
The Coleman Report, written by the sociologist James Coleman and his colleagues, found that family background, not school effects, significantly predicted academic achievement. It laid the groundwork for an enduring narrative about the limited effects of between-school differences on student achievement in the presence of the robust effects of family background, concluding thatschools bring little influence to bear on a child’s achievement that is independent of his background and general social context; and that this very lack of an independent effect means that the inequalities imposed on children by their home, neighborhood, and peer environment are carried along to become the inequalities with which they confront adult life at the end of school. For equality of educational opportunity through the schools must imply a strong effect of schools that is independent of the child’s immediate social environment, and that strong independent effect is not present in American schools. (Coleman et al. 1966, 325)
Coleman and his team also concluded that the social composition of a school’s student body is more highly related than any other school factor to achievement, independently of the student’s social background. The Coleman Report therefore yielded two significant takeaways: first, that the advantages and disadvantages conferred by the family context follow students throughout their elementary and secondary schooling; and second, that if disadvantaged students (namely, African Americans, Native Americans, and Mexican Americans) attend schools with more white students, then they are likely to garner significantly better achievement results.
As I elaborate in the following pages, several contextual factors should be considered before we draw our own conclusions about what Coleman and his colleagues reported as a lack of school effects and the strength of the effect of racial and ethnic minority students attending school with white students. First, any analyses of school effects should not be ahistorical; we cannot ignore the near-homogeneous nature of schools for disadvantaged racial and ethnic minority students in the United States in the mid-twentieth century. At the time of publication of the Coleman Report, institutionally sanctioned discrimination and lack of access to a quality education for students of color were as old as the United States itself. Southern states had made it illegal to teach an enslaved person to read, and this policy persisted through the Emancipation and Jim Crow eras; well into the twentieth century, African Americans faced de facto and de jure exclusion from public schools, as did Native Americans and, frequently, Mexican Americans (Kluger 1976; Tyack 1974). Even in the North, problems of exclusion, segregation, and lack of resources were severe for African Americans who migrated from the South to urban areas like Chicago (Neckerman 2007). Therefore, the variation in the proportions of low-poverty and high-poverty schools to which African Americans, Asian Americans, Mexican Americans, and Native Americans were exposed was extremely limited (Anderson 1988; Kluger 1976; Tyack 1974).1 Finally, as Samuel Lucas (this issue) argues, “failure to gather truly school-specific [as opposed to district-specific] expenditure data in the mid-1960s could underlie the finding of small or even nonexistent differences in resources by race and, by restricting the range of variables and introducing measurement error, may explain why school-level factors of the period appeared to be of little consequence for racial inequality.”
The current fiscal and social landscape of schools is similarly dismal for a significant proportion of these students in the second decade of the twenty-first century. The National Center for Education Statistics (Kena et al. 2015) reports that in the 2012–2013 school year, nearly half of African American and Latina/o students (45 percent for both groups) attended high-poverty schools (schools where more than 75 percent of the students were eligible for free or reduced-price lunch), compared to only 8 and 16 percent of white and Asian students, respectively. Taken altogether, these economic facts should inform the social science community that certain social groups were and continue to be significantly more likely to attend low-quality schools with extremely limited variation in fiscal resources and a concentration of poverty. Arguably, such an enduring material landscape could create an illusion of either limited or absent independent effects of schools.
Second, the Coleman Report points to the positive influences of student composition in a school—namely, the presence of white students. This finding indicates the comparative advantages of diverse schools attended by both students of color and white students. Congruent with Coleman’s findings, some newer studies have shown positive effects of desegregated schools, including stronger retention rates, or less dropping out, and better performance of African Americans in these schools than in segregated ones ( Guryan 2004; Hanushek, Kain, and Rivkin 2009; Mickelson 2001). In addition, low-income, high-achieving black students in desegregated schools have been found to have a higher chance of getting out of poverty with their high school diplomas (Braddock et al. 1986). We might ask whether these effects stem from (1) students of color sitting next to white students, (2) students interacting in ways that change attitudes and aspirations, or (3) the additional resources that are concomitant with white students’ enrollment.
Other studies, in contrast, appear to suggest that school desegregation has no positive influence on the reduction of the racial test-score gap (Vigdor 2011), although data on enrollment in honors courses suggest that within-school segregation increases when schools are more highly diverse, thus potentially offsetting the benefits of school desegregation (Card and Rothstein 2007).2
In an assessment of educational outcomes other than test scores, some studies, including that of Geoffrey Borman and Jaymes Pyne (this issue), have shown that racial and ethnic minority students fare better sociopsychologically and culturally in mono-racial and mono-ethnic schools than in desegregated schools (see also Bates 1990). Students of color, according to these studies, feel more attached to school when they attend it with greater proportions of students of their own race or ethnicity (Johnson, Crosnoe, and Elder 2001). In predominantly white higher education settings, black students report problems of cultural alienation, social adjustment, racial discrimination, and strained interpersonal relations with other students, faculty, and staff (Allen 1988, 1992; Bennett 1984; Chavous et al. 2002; Cureton 2003; Hurtado 1998; Kraft 1991; Willie 2003). Fewer studies, however, have explicitly examined the cultures of secondary schools and how they influence the incorporation of specific groups of students.
Persistent achievement differences, in addition to social and cultural challenges, beg the question of whether simply ameliorating a school’s “mobility context”—that is, its material or resource conditions—is sufficient to eradicate academic achievement disparities. Researchers—in the intellectual tradition of such scholars as Samuel Bowles and Herbert Gintis (1976) and Pierre Bourdieu (1977)— have suggested that a school’s cultural environment affects students’ incorporation and their educational behaviors ( Delpit 1997; Lareau and Horvat 1999; Soudien 2001). Lucas (this issue) describes the omission of the analysis of in-school processes as a major methodological drawback of the Coleman Report. Instead, the Coleman Report adopted an input-output model, or a material framework (albeit limited in measurement), because consensus around a more complex educational process within schools did not exist (Coleman et al. 1966, 239–40; Lucas, this issue).
Perhaps we social scientists and educational researchers in the United States have not heeded sufficiently the caution of the sociologist W. E. B. Du Bois (1935), who foreshadowed problems that desegregated (and even resource-rich) schools would confront. Placing diverse bodies next to each other would not heighten student achievement, Du Bois argued, since multiracial schools characterized by thick social (that is, racial) boundaries were just as “bad” as segregated schools with limited material resources. In the case of the former, Du Bois suggested, teachers, principals, staff, and students would have to tackle enduring social and symbolic boundaries that compelled individuals and groups to act in ways that reproduced inequality (see also Tilly 1999). And failure to meet the “intangible” goals of integration (see Wells 2000)—that is, bridging the racial, ethnic, and socioeconomic chasms and facilitating cross-cultural communication and relationships among social groups previously isolated from one another—would undermine the goals of equal opportunity within schools.
I argue that to fully understand the school’s role in students’ incorporation, it is important to differentiate between the school’s mobility and sociocultural contexts. The mobility context comprises the material or resource conditions of the school. On the one hand, high rates of poverty and other disadvantaged socioeconomic conditions characterize many urban, minority-dominant schools, while greater wealth and improved conditions such as smaller teacher-student ratios, more experienced teachers, more course offerings, more funds and diversity in extracurricular programs, and wider networks of information about college attainment are found in desegregated schools (Crain 1970; Crain and Mahard 1983; McIntire, Hughes, and Say 1982; Orfield 2001). The sociocultural context, in comparison, comprises the school’s norms of academic achievement, its logic for student conduct and presentations of self, its pedagogical content and practices, and its climate of teacher-student, student-student, and other inter- or intragroup dynamics. The school’s cultural environment can also include the meanings that students and educators attach to certain curricular patterns, such as which students belong in specific courses and who plays which sports or participates in particular school activities.
An underlying premise of this paper is that student success, engagement, or well-being is not simply indicated by tests. Several other outcomes that pertain to students’ relationships to schooling, including their sense of “engagement” and “incorporation” within their school, are probably related, either directly or indirectly, to students’ attainment and achievement. Specifically, I examine the following questions: Are there associations between schools’ sociocultural environments—specifically in minority-dominant versus majority-white, multiracial schools—and the incorporation of black and Latino students? If so, what are the features of these environments that enable certain racial and ethnic groups to be incorporated more easily at some schools than at others?
Conceptually, a social group’s incorporation within a school entails more than the nature of its members’ participation in classes and activities or an individual student’s propensity to either stay in school or drop out. Here I use “incorporation” to refer to the reception by an institutional context (a school) of different social groups and that context’s ability to move them toward the center of the school organization and away from its margins in their social relations and academic participation. That is, I argue that the relational aspects of the school context matter greatly too, and that the nature of educators’ and students’ social relations—their position, status, and location within the school—can either facilitate or reproduce educational inequality in ways that are masked by a focus only on either the students’ family backgrounds or the school’s material resources or inputs. To be clear, the objective of this paper is not to present a predictive model of school achievement among social groups across different school contexts (although some test score data comparisons are shown here), but rather to illuminate how the social status of African American and Latino students with similar socioeconomic backgrounds can differ based on the school’s sociocultural context and composition. Significant divergences in access and participation in myriad educational activities—which, conceivably, are conducive to greater academic mobility—are found between high-performing majority-minority and majority-white schools.
METHODS
Using a comparative case-study analysis of four schools located in two different regions of the United States, I compare the experiences of students who hailed from similar socioeconomic and family backgrounds but who attended different schools with varying social compositions. Relatively speaking, each school was a high-performing high school (grades 9 through 12) in a metropolitan area, two in a Southern capital city and two in a Northeastern capital city. One school in each city was minority-dominant (one was predominantly black and one was predominantly black and Latino), and one was multiracial and predominantly white. A team of seven researchers (six assistants and I) visited the four schools weekly for six months from January through June 2007. The research team comprised three African Americans, one Egyptian American, and three white Americans. Except at South City Honors, which was visited mainly by two African American researchers, a mixed-race pair of researchers attended an array of classes and extracurricular and lunchroom activities four to five days a week, spending anywhere from three to seven hours at the school.
The four schools in the study shared two main criteria. First, they were all multiracial in terms of student composition, though the racial majority varied in each school. Second, to hold constant the average academic orientation of the schools, the four schools chosen were similar in their state’s accountability ratings as it pertained to the No Child Left Behind law: all four schools were relatively high-performing in their district. Also, we selected two cities that were in areas where desegregation struggles had been fraught with racial and ethnic strife (Dalbey and Harris 2001; Eaton 2001) but were also in different areas of the country; thus, regional history might account for variations in the extent of interracial and interethnic contact and the permeability of group boundaries (Farley and Frey 1994). While the four schools were not representative of all schools in their district, they were typical of schools that could be classified as multiracial in their district and metropolitan area.
The data collection began in Southern Capital City (SCC) in January 2007 at South County Prep High School (a pseudonym), a school of grades 9 through 12, with 1,400 students, located at the fringe of a medium-sized, urban Southern capital city (2010 population 175,000).3 According to the 2010 census, there were 540,000 people residing within the SCC’s metropolitan statistical area (MSA). The racial makeup of the MSA was 53 percent white, 45 percent African American, 1 percent Asian, and 1 percent Hispanic or Latino of any race. At South County Prep High School, about 77 percent of the students were racially classified as white, 21 percent as black, 1 percent as Asian, and 1 percent as Hispanic. Sixteen percent of students enrolled qualified for either free or reduced-price lunch.4 South County Prep’s student-teacher ratio was seventeen to one, and all of the teachers and staff, except for three of them, were white. Ten miles west of South County Prep was South City Honors High School, an urban, comprehensive high school of grades 9 through 12 with a notable Advanced Placement (AP) program. South City Honors had 1,300 students, 93 percent of whom were racially classified as black and 6 percent as white. Sixty-four percent of its students were on either free or reduced-price lunch. With a student-teacher ratio of twenty to one, the school was led by a multiracial staff of sixty-eight teachers (see table 1).
Similar school types were selected for observational, interview, and survey data collection in Northern Capital City (NCC). NCC, a large urban center with a population (2010 population 620,000) that was 54 percent white, 24 percent black, 18 percent Hispanic, and 9 percent Asian, had a public school student population of 60,000. According to the census of 2010, there were about 4 million people who resided within NCC’s MSA. The student racial demographics in NCC’s public schools did not, however, reflect the residential demographics. With the exception of three exam schools (an organizational structure that attracted a critical percentage of white students), NCC’s public schools were highly segregated. Forty-two percent of students were black, 14 percent white, 9 percent Asian, and 35 percent Hispanic.
North City Tech School was one of the three exam schools; in spite of the influx of white students, more than 80 percent of its 1,200 students could be categorized as “minority” or students of color. Although many of these students performed better than the majority of NCC students on their elementary state tests, most North City Tech students had not scored high enough to be accepted to the two other exam schools, which were considered more elite and ranked higher. Out of the four schools in this study, North City Tech was the most racially and ethnically diverse, with about 45 percent black students, 23 percent Asian, 20 percent Hispanic, and 11 percent white. Nearly two-thirds of the students (63 percent) were on free or reduced-price lunch.5 North City Tech’s student-teacher ratio was eighteen to one, and like South City Honors in SCC, its teaching staff and administration of about sixty-five was multiracial.
To find a comparative majority-white school with a critical mass of black and Latino students, I had to look in a suburban, upper-middle-class school district that participated in NCC’s metropolitan Voluntary Desegregation Program (VDP).6 North Village Prep High School, located twenty-three miles northwest of North City Tech High School, had participated in the VDP since 1967 to attract African American or black and Hispanic students, most of whom entered a lottery to attend affluent suburban schools like North Village Prep High School. Of North Village Prep’s student population of 1,200, 84 percent were white, 6 percent were black, 5 percent were Asian, and 3 percent were Latino. Only 2 percent of North Village Prep’s students were on free or reduced-price lunch. With the exception of one African American and one Asian American teacher, all of North Village Prep High School’s eighty-eight teachers were white, and its student-teacher ratio was fourteen to one.
REPORT OF FINDINGS
Variations in Educational Experiences by School Context and Race-Ethnicity
While the Coleman Report documented the positive influences of racial and ethnic minority students attending school with white students, he and his colleagues were unlikely to have anticipated that the potential effects of interracial contact might be mitigated by organizational mechanisms within such schools.7 There is no shortage of research showing the stratifying and consequential effects of school tracking by race, ethnicity, and social class (Card and Rothstein 2007; Gamoran 1987; Oakes 2005; Tyson 2011). Tracking has begun to be referred to as a form of resegregation because it has evolved into an educational practice that frequently excludes and stratifies on the basis of perceived ability by these social identity categories. Further, from a researcher’s perspective, I have seen that the social organization of students within classes and racially demarcated extracurricular activities reinforces the establishment of de facto ethnically and racially segregated spaces.
At North Village Prep, 88 and 71 percent of the Asian and white students, respectively, were enrolled in at least one honors or AP course, compared to 30 percent of their black and Latino peers (see table 2). English and history were not tracked at North Village Prep, but other courses were; we learned that none of the black and Latino students in the school were enrolled in the advanced math and science classes. At South County Prep, 62 percent of the white students surveyed were enrolled in at least one honors or AP course, compared to only 44 percent of the African American students. In contrast, at both South City Honors and North City Tech, surveyed students from all racial and ethnic groups enrolled in such courses at nearly the same rate, ranging from an average of 75 percent at the former and 45 percent at the latter.
Comparing himself to the more privileged students at his school, one fifteen-year-old sophomore, Judah, commented on the academic divide between students participating in the Voluntary Desegregation Program and non-VDP students:When I got here, looking at kids my age who are taking trigonometry, and I’m here in geometry . . . I’m here in Algebra 2; I’m looking at kids who are in calculus; and I’m still here in Algebra 2, being where I’m supposed to be. And I feel like I have to catch up to them because if I don’t, then thirty years down the line, who is going to be the clerk and who is going to be the one who is leading the company?
Judah anticipated a long-term economic impact of tracking in his school, one that would reinforce the socioeconomic class privilege already present at North Village Prep. In his view, different levels of course access and preparation would lead to differential levels of higher education and economic opportunities.
A similar pattern could be seen at South County Prep. Although our research team observed five African American students (out of 302 overall) in several AP and honors courses there, these were usually the same students enrolled in the higher-level classes. Mrs. Spann, one of the three black teachers at South County Prep, gave me her take on the issue of low black representation in the most advanced and demanding classes. I met Mrs. Spann, who taught communications skills, in her first year at South County Prep. I attended one of her classes during the last period of the day and remained after school to talk to her. When I mentioned the low number of high achievers among the African American students, Mrs. Spann remarked, “There are many more smart students, but you should notice the color of the skin of those who get classified as ‘smart.’ You’ll notice that some of the students, because of the way they look, will never get deemed as smart, even though they are quite competent and intelligent,” she continued.
While tracking was also practiced at majority-minority South City Honors and North City Tech, there were significant differences between these two schools and the other two in terms of which students were represented as the “smartest” or the “brightest.” On a typical school day in the spring at South City Honors, I sat in on the AP English class of Cate Gilman, a 2007 ACT (American College Test) Star Teacher who was a tall, slender, thirty-something white female with medium-length, fine brown hair and fair skin. Her classroom walls were lined with sketches of the faces of famous authors, including Jane Austen, D. H. Lawrence, and George Orwell. Ms. Gilman had also posted a picture of Toni Morrison on the bulletin board to the right of the door in a collage of other writers’ pictures; Morrison was the only writer of color in the mix. Ms. Gilman’s second-period class had twenty-two students: fourteen black females, three black males, two white females, and three white males. I was familiar with some of the faces, including Xavier, the African American male who was one of the two National Achievement semifinalists in that year’s senior class. Also, there was Benson, a white male who was the high school valedictorian and had obtained a perfect score on the ACT. Ms. Gilman had asked her students to complete an AP practice test exercise, and they worked for about twenty minutes on a difficult prose passage excerpted from a literary work on how a nineteenth-century aristocracy dealt with the virtues of charity and humility. As they worked, Ms. Gilman walked around and passed out cards, each marked with the number of a particular question; as students answered the questions correctly she handed them a card with remarks such as: “You got #1 correct; you’re a genius!”
Nearly 1,500 miles away at North City Tech, our research team witnessed similarly diverse classrooms and student and teacher dynamics. The AP physics class of Mr. Jimenez—a thirty-something bald and bespectacled Latino who sported a well-worn University of Northern Capital City sweatshirt embossed with a logo promoting a local youth program at the university—was large (around thirty students), interactive, and lively. It was also racially mixed and fairly balanced in terms of gender. Mr. Jimenez’s classroom had the feel of a physics lab: the entire back wall was lined with multiple kinds and sizes of gears, pulleys, and levers and student-made, physics-themed mobiles were hung from the ceiling. Students sat together at tables, interacting across race and gender. On this particular day, for example, a black male, upon entering the room, gave pounds and handshakes to a group of very diverse boys. As they worked the students shared ideas and answers, supporting one another in solving the assigned problems. They moved freely from table to table to help one another, and all conversations between them appeared to be physics-related.
The two classroom observations made at South City Honors and North City Tech High Schools were fairly representative of many of the courses at these schools, especially the ones for high-achieving students. While not all of the classes at these high schools seemed to be as engaging as these, these particular classes for high-achieving students presented a contrast between these two schools—whose predominant groups were characterized as racial and ethnic minorities in the U.S. context—and the two other schools in the study. At South City Honors and North City Tech, it was not a rarity to see the higher-level courses—those considered the most rigorous or even for the “smartest”—filled with significant percentages of black and brown students. In comparison, beyond their class time with white peers in their general comprehensive and college preparatory classes, nearly all of the students of color at North Village Prep—namely, the VDP participants—had limited social contact with white students. Whether they were completing paper assignments, getting tutorial assistance, playing chess, talking politics, or merely hanging out with one another, most VDP students spent at least some time in the “VDP room,” away and separate from their white peers.
Other issues threatened to undermine the initial attention that student leaders and faculty at North Village Prep had given to the social distance and separation between the students. During our first week of research at North Village Prep, I sensed a buzz of eager anticipation in the air: some students and a few fortunate teachers were preparing to head to Japan that spring on a band and orchestra trip for a two-week exchange. While one of only two black teachers at North Village Prep, Mr. Moman, was going on the trip as a chaperone, none of the African American and Latino students who participated in the school’s VDP were headed to Japan. Yearly, students of this affluent school took trips abroad to Europe, Asia, Latin and Central America—not only after numerous car washes and bake sales but with parental financial support. By contrast, North Village Prep’s black and Latino students, the majority of whom were lower-income and voluntarily bused to the school via the VDP, could not afford these trips and were not financially subsidized by the school to participate in them.
Additionally, the school’s efforts were challenged by the impact of residential segregation on VDP student participation in non-academic aspects of school life, as well as their social cohesion with schoolmates. Almost all of North Village Prep’s VDP students were bused into the district from the urban center and surrounding areas of Northern Capital City. Several voiced their dislike for the lack of neighborhood proximity; they could not fully participate in every aspect of school life, and their principal and teachers had found no easy way around the matter. Although the school did provide a “late” bus, which left at 5:30 PM daily, taking it would have prevented VDP students from participating in any evening programs or early morning meetings that occurred before the bus arrived. One of those students was Neela, who had been on the basketball team and had been able to sleep over at a teammate’s house a few evenings when she did not feel like making the long trek back home to the city from the suburbs after the evening practices held from 6:00 to 8:00 PM. Neela did not enjoy having to stay overnight away from her home so often, however, and with no prospect of a late bus, she stopped playing basketball.
“It’s a Matter of ‘Tastes’”: Social Distance and Racial-Ethnic and Class Relations
The school communities of North Village Prep and South County Prep High Schools, the two multiracial, majority-white schools, experienced thick social boundaries (Lamont and Molnár 2002) that endured behind school walls and in classrooms.8 A stroll through North Village Prep’s cafeteria on any given day would reveal this. Filled with vibrant chatter and myriad social dynamics, the school cafeteria is a sociological laboratory for the study of adolescent group relations. The most noticeable and immediate observation at North Village Prep was the sight of the VDP students—nearly all of the African American and Latino students—sitting on the right-hand side near the back door entrance, while the white and Asian students sat on the left-hand side in duos and groups according to clique preferences.
The school’s personnel had begun to acknowledge some of its social and cultural difficulties and even attempted to do something about it with the “Challenge Day” program, which would reach at least one hundred students. On an assembly day, as students walked into the auditorium, they had been given randomly assigned, color-coded pieces of paper and told to sit in the area designated by its color. The students had chosen to ignore these directions, not only because they were rebellious adolescents, but also because randomly assigned seating would have taken them out of their comfort zone of sitting with their friends. After they watched video excerpts featuring the Challenge Day program from The Oprah Winfrey Show, students were encouraged by two twelfth-grade co-facilitators, an African American female and a white male, to participate in North Village’s own upcoming Challenge Day.9 While Will, a graduating white male senior, anticipated a positive long-term impact of such a program, he worried about the impact of more immediate practices already in place at his school:
Interviewer: If you were able to change one thing about your school, what would that be?
Will: I think one thing I would change is I would try to make this school more welcoming for . . . all people. I kind of, I mean, as with any high school, probably, you tend to form . . . tend to break down into groups. I mean, I hang out with people who do a lot of the same activities as me and, you know, get the same grades and are in the same classes, and I think that’s one disadvantage of having kind of tracked classes. I’m glad that we don’t have them in English or history. I think that helps a little bit, but, yeah, I’m certainly glad that we have the . . . via the [Be the] Change program this year, I think . . . I really think that will make a difference in the school.
Interviewer: You do?
Will: Yeah. I really do. I mean, I participated in one of the days, and I thought it was an amazing experience, but I think people . . . I think it’s . . . if not already, then in the future it will help people to kind of recognize others and not just judge them. It will probably take a while for it to make a really big difference in the school.
How do we make sense of this social distance between students? Like Will, some other students expressed their racial separation as a matter of “interests” and shared understandings. Vonda, an African American sophomore at North Village Prep, explicitly attributed “all the black kids [sitting] together” to their common interests:10
I think that people just notice that we sit together because we’re a different race or we’re different . . . we’re like . . . most obviously different from outside than everybody else. [The VDP kids] may very well be a group of basketball players, a group of nerds, a group of hockey players, a group of, you know, ballets or a group of, like, musical. You know what I mean? In the cafeteria they all sit with each other and have cliques, but you can’t tell because they all look similar and then we are in the corner over there [laughs]. Like obviously different, so people are just like, “Oh, they don’t like us.” But it’s not that. . . . In general people find similarities with other people, and they relate to them and be friends with them and we all understand each other, so we end up doing that. It’s the same thing with all the hockey players who understand each other and talk about hockey all day or all the like [pause] you know, singers will all sit with each other and sing with each other all day.
Vonda tried to convince me that racial, ethnic, and ostensibly class divisions lead to a sense of shared interests and a social cohesion that presumably would lead any humans with something in common to form groups and distinguish themselves from others spatially in the cafeteria. Unlike their white schoolmates, however, Vonda and her VDP peers did not have the luxury of forming groups according to specific social, extracurricular, academic, or hobby interests. Their low numbers and high visibility precluded that sort of social organization. Still, two obvious and critical questions remain: Why did the other interest groups not include some of these VDP students, who presumably shared their taste for music, sports, language, academics, or other interests? And why did Vonda and her VDP peers choose not to integrate those groups?
Similarly, at South County Prep, basketball and step-dancing are to “black” and “black female” as baseball and cheerleading are to “white” and “white female,” respectively. There students formed close relationships with those participating in the same activities. Cheerleading, baseball, the Fellowship of Christian Athletes, and the Young Republicans Club had few African American students involved, but these students were well represented in basketball and football. The real and symbolic distinctions students saw mirrored in their social organization within schools were teaching them their places, their boundaries, their “lines” (for more discussion, see Carter 2012). As Ashley, a white female eleventh-grader at South County Prep, noted:We have, like, kind of blurry lines a lot of the times . . . but like, you know, you have that group and you can’t really, like, relate to that group. You can individually but, like, not as a whole group, but a lot of the other groups, just, like, they’ve learned their lines a lot. Like we have a lot of people that are in the AP classes, and they hang out together a lot, and there’s, like, theater groups and stuff like that, and they hang out and just, like, random small groups from, like, different . . . just from being in high school together for so long and stuff like that.
Not only did Ashley call out the social “lines” or boundaries between groups of students, but she also distinguished between interpersonal and intergroup contact and prejudices (“You have that group and you can’t really, like, relate to that group. You can individually . . .”). That is, she could cross racial boundaries and relate to an African American peer as a friend, for example, but her friendship with one person did not necessarily allow her to view African Americans as a whole differently. Further, Ashley noted that the social organization of her school influenced her thinking and her relationships with students who differed from her socially. Social psychologists have noted (for example, Pettigrew 1998) that although cross-race friendships may benefit in-group members’ perceptions of out-group members, the extent to which the befriended out-group member is perceived to be representative of the out-group actually determines in-group members’ ability to change their attitudes about the out-group. Otherwise, the in-group members tend to maintain a “but you’re not like the other members of your group” attitude about out-group members.
Thoughts and beliefs like those of Vonda, Ashley, and Will about the consequences of the boundary-making that occurs within schools—the site, we would assume, of deep interracial and ethnic contact—emerged repeatedly in the study. Similarly, at South City Honors and North City Tech, students admitted to having cliques and peer groupings, but the boundaries seemed more permeable at both of these schools. White kids were in the minority at South City Honors. They spoke openly with me about it, and students like Adam and Meredith shared some poignant stories about the consequences of interracial dating in their communities (see Carter 2012). According to Fred, a white male senior, the diminution of racial boundaries at school was undermined by the social climate outside in the communities of his Southern city:At South City Honors for . . . [pause] you know, people tend to disperse to who they’re more related to, and I guess being white and the minority [laughs] . . . is, I don’t know . . . I guess you can relate maybe a little . . . easier to white people and it’s kind of easier to get along with them, but . . . uh . . . and it’s kind of how it is, but . . . uh . . . I mean, everyone is cool. Everyone is generous, you know, they’re good people, and it’s . . . it’s kind of like a school friendship but away from school you don’t really keep the same friendships and I don’t know.
At North City Tech, where more salient ethnic immigrant distinctions arose, students formed ethnic social clubs: at lunch students referred to the Puerto Rican table, the African American table, the Chinese table, the Japanese table, the Haitian table, the Vietnamese table, the “geek” table, and the “loser” table, to name some. Strikingly, the loser table was the most diverse of all the tables in terms of race, ethnicity, and gender. These students would attend each other’s programs and events, even if they were not in-group members. We met many cultural boundary crossers, however, at North City Tech. For instance, Cherise, an African American ninth-grader at North City Tech, spoke frankly about why she had decided to enroll in a Chinese language class at school: most of her friends, since seventh grade, had been Asian American—Chinese, to be specific—and she wanted to converse with them in their language. Cherise’s was not an exceptional case. Two of her schoolmates were white female seniors, one an Italian-Irish American girl and the other an Irish Catholic, who perceived themselves as “minorities” in this predominantly black, Asian, and Latino school. One of them, Natalia, discussed her own cultural flexibility: “I was the only Italian-Irish girl, and everybody thought I was some type of Latina, so they were like, ‘Oh, what are you?’ and they thought I was Cuban and white. Like, ‘No. I’m Italian.’ I hang out with a lot more Latino people just because, I don’t know, so I, like, learn how to speak Spanish and stuff.”
The Coleman Report pointed to the positive effects on achievement of interracial contact between students in schools. But how does that contact facilitate achievement in diverse schools where homophily (friendship preferences for one’s own group) prevails? Further, how might the social organization of schools and classrooms undermine the goals of integration? As the sociologist James Moody (2001) found with a nationally representative data set, there is still a strong and generally positive relationship between heterogeneity (diversity) and friendship segregation in schools. Thus, simple exposure to one another does not promote integration. In addition, Moody found that friendship integration only occurs in spaces where students are encouraged to mix through extracurricular activities and classes in which status equality is permitted.
In some instances, the morality and merits of school diversity are clear (Allen 2004; powell 2005). In other instances, we still seek to understand why other research has not shown this increased contact to yield the strongest academic effects. As the legal scholar john powell has argued, “true integration” has not occurred in most U.S. schools. That is, desegregation (or spatial proximity) is not the same as integration—the transformation of a school’s cultural, social, and political environment to incorporate all. (For a more theoretical discussion on this, see Fine, Weis, and Powell 1997; powell 2005.) The findings of my study suggest that researchers must deal with the racial, ethnic, and class “stuff”—the residual baggage and emotions fomented by historical and fraught relations—before they disavow the argument that school integration does not yield strong academic effects or fulfill the promise of the Brown v. Board of Education decision. At the same time, we have to question any assumptions created by the Coleman Report that sending students of diverse racial and ethnic backgrounds to school together will actually pay off academically for all groups. The school’s sociocultural context matters. In actuality, South County Prep and North Village Prep had some of the finest educational resources in their communities, and students’ families knew this. Paradoxically, these schools’ resource contexts did not appear to be strongly associated with either students’ propensity to cross social boundaries or educators’ ability to facilitate that navigation so that all groups of students would have equal educational opportunity within these wealthier learning environments.
The Paradoxes of “Good Schools”
When it comes to the academic incorporation of historically disadvantaged groups of students, often traditional, high-performing majority-minority schools face significantly greater material (resource) challenges to lifting the attainment of all their students (Darling-Hammond 2010). High-performing majority-white schools, in comparison, tend to face significant social and cultural challenges (Carter 2012; Lewis and Diamond 2015; Tyson 2011). Meanwhile, some majority-minority schools, such as South City Honors and North City Tech, produce excellence with fewer resources, while some affluent majority-white schools produce mediocre outcomes, on average, among their African American and Latino students. Like most urban schools in the United States, South City Honors and North City Tech faced the pressure to produce high test-score results. South City Honors prided itself on its multiyear history as a highly regarded high school in the state—and even in the nation, according to national ranking polls. Some students lamented the school’s being “very obsessed with being a level five school.” Darlene, an African American sophomore, declared to me: “They ain’t worried about our future; they ain’t worried about us. All they want is to look good [on the tests].” One of Darlene’s peers, Danielle, sitting next to her in the in-school suspension room (the school auditorium), where I spoke to them, agreed: “They [the educators] worried about the school’s future, not the students.” Despite the school’s reputation, some of its most academically vulnerable students still recognized the relative material disadvantages of South City Honors compared to a school like South County Prep down the road. Darlene continued: “I mean, for us to be such a good school and they always saying that South City Honors is so good, why can’t we get some of the stuff that, like, South County Prep and all them get?” Although average test scores were excellent at South City Honors, Darlene and some of her peers did not perceive that they had access to the best learning environments. Certainly, high test-score performance masked the differential resource levels between South City Honors and South County Prep, but whether Darlene and Danielle were correct demands deeper levels of inquiry.
Much farther north, the families of students in the Voluntary Desegregation Program in Northern Capital City schools actively chose to send their children across municipal boundaries so that they would have access to better schools. “I don’t like the schools in Northern Capital City! They’re bad, [and] the kids don’t like to learn anything!” exclaimed tenth-grader Briana when asked why she chose to go to North Village Prep High School. Briana’s explanation was echoed by other VDP students when we interviewed them. Most of the youth in the VDP were relatively high-achieving students in their local communities; they were the kids who might have attended North City Tech, one of a very few competitive high schools in the central city. Once they entered North Village Prep, however, the standards and characterizations of “good” students changed drastically for them. VDP students encountered a more rigorous academic climate than they had yet encountered. They were reasonably aware of these academic differences, and students like tenth-grader Judah (introduced earlier) grappled with the greater difficulty of the classes. One afternoon Judah debated with some other classmates and fellow participants in the VDP about whether the achievement gap was the “fault” of the individual student or the school administration:I think I have kind of a [long pause] theory. I don’t know if you guys feel this way, but I think it’s more of kind of feeling intimidated by the North Village kids just because there’s just some of them that are so much smarter than you. Like, it’s hard to compete with them in how they do and so you get discouraged. Like, am I going to do better? Like basically, like say you’re in a class and you want to know . . . like the worst kid is really smart . . . and you want to know, am I going to do better than that worst kid? So you’re trying to strive; you get nervous and you want to do better than him and you end up doing bad. It’s kind of like, really . . . it’s like a big weight falling on your chest and therefore you have to carry that weight around the whole year. Whole year! Thinking you want to do better than that worse kid and that’s only the worse kid. Like saying, and I don’t know if you guys feel that way, but sometimes it feels like . . . like I’m basically the dumb one in the class. Like I have to compete with all the smart kids, and it’s really hard to do that when, like, all the kids are really smart, so therefore I think it hurts me on my test because I’m going into a test thinking that I’m going to do bad when I could probably do better.
Not only did Judah recognize how his North Village Prep peers differed from him in their academic preparation, but he also hinted at what the psychologist Claude Steele and his colleagues (Steele and Aronson 1995) have discovered about the concept of “stereotype threat.” The pressure of competing with peers from better-resourced backgrounds produces an anxiety that may stifle performance. An analysis of the mean academic self-esteem of North Village Prep students by race and ethnicity confirmed this: black students possessed significantly lower academic self-esteem, on average, than their white (p < 0.01) and marginally lower than their Asian peers (p < 0.10), while the Latino students possessed marginally lower academic self-esteem than their white and Asian peers. (No significant differences like these were found at the other three schools.) Academic self-esteem, or what some social psychologists refer to it as “academic self-concept,” is known to have a reciprocal relationship with academic performance. That is, academic self-concept is found to influence academic performance, and conversely, academic performance influences academic self-concept (see, for example, Marsh and O’Mara 2008).
When we examined the state test results for the students surveyed, we found various between- and within-school differences (see table 3).11 In keeping with the main comparative focus of the study, first we analyzed student test scores within race between majority-minority and majority-white schools. Our analyses show that African American and Latino students performed better in math at the majority-minority high school than their African American peers at the majority-white school in Northern Capital City; no between-school differences were found on English test scores among African Americans and Latinos in either city. The converse is true for Asian and white students at North Village Prep only: these students scored significantly higher on the state English exam than their racial or ethnic counterparts at North City Tech, but no differences were found in students’ math scores between the schools in Northern Capital City. As for regional comparisons, we found no significant within-race differences among students on test scores between Northern Capital City and Southern Capital City.
In contrast, within each school some significant racial and ethnic differences were found. At the most affluent school, North Village Prep, African American and Latino students scored significantly lower than their Asian and white peers on both the English and math state exams. Farther away at South County Prep, attended by only African American and white students, the same significant pattern held in both English and math. At the majority-minority North City Tech, the African American students surveyed scored significantly higher on their English exam than the other ethnic and racial groups, but African American students scored significantly lower on their math exams than their white schoolmates at South City Honors. In sum, across the board we found significantly higher outcomes for African American students at the schools where they were incorporated more deeply into various facets of school life.
These test-score findings and others converge with what other studies have found about the strong positive relationship between students’ sense of belonging and their school performance. In experimental intervention studies, results show that when racial and ethnic minority students feel that they belong in a school, they perform better (Walton and Cohen 2007). As I explored other dimensions of the schools’ sociocultural contexts, I examined students’ sense of attachment or belonging to their schools. A survey question—using Likert-type response categories, with 1 being “strongly agree” and 4 being “strongly disagree”—asked a random-stratified sample of 647 students attending the four high schools to respond to the following statement: “I feel like I am a part of this school.” As table 2 shows, African American students in the majority-minority schools (North City Tech and South City Honors) reported significantly higher levels of attachment to their schools than their African American peers in the majority-white schools (South County Prep and North Village Prep). At South City Honors, white and African American students felt a part of their school to nearly equal degrees—76 and 72 percent, respectively. At the other majority-minority high school, North City Tech, 83, 62, 69, and 66 percent of the African American, Asian American, Latino, and white students, respectively, felt a part of their school. Noticeably, white students felt less a part of North City Tech (the most racially and ethnically diverse of the four schools), where they were a numerical minority, than the African American students did, although white students’ level of attachment was similar to that felt by Asian American and Latino students.
In comparison, at majority-white and high-income North Village Prep, the overwhelming majority of the three groups surveyed— 90, 84, and 74 percent of white, Asian, and Latino students, respectively—said that they felt they were a part of their school, while only 38 percent of the African American students agreed with this statement. Latino students felt more a part of North Village Prep than African Americans did, but to a still significantly lesser extent than their white peers. At the other majority-white school, South County Prep, the difference was not as wide between blacks and whites as at North Village Prep, but it was noticeable and significant nonetheless: 67 percent of South County Prep’s white students agreed that they felt a part of their school, compared to 59 percent of African American students.
Students’ sense of belonging can also be marked by a sense of group inclusion on a larger level. At both North Village Prep and South County Prep, faculty and students were unaccustomed to viewing people of color, even teachers of color, as models of high achievement or intelligence. One morning while engaging with students in a class, I offered some clarification on sociological terms like “inequality,” “race,” “socioeconomic status,” and “ethnicity,” ideas that they conflated. One of the few African American males in the class subsequently expressed his incredulity at how “smart” I was and noted that I was “his color.” The comment was meant to be a compliment, but for me it was disheartening because it signaled that the students—Asian, black, Latino, white, and multiracial alike—clearly had not had the opportunity to interact with many highly educated persons of color. They were not getting the message in school that “black” or “brown” could be synonymous with intelligence. I asked Diane Newsome, the director of the VDP, why North Village Prep had so few African American (and no Latino) faculty or staff members—only two, including Diane—and she quipped, “Why is it that Harvard [my place of employment at the time] has so few tenured black faculty?” Good question. Diane and I began a long conversation after her pointed question about racism, discrimination, exclusion, and the idea that the impact of representation and incorporation on student performance and success at both South County and North Village High Prep (or even elite universities) cannot be underestimated. Teacher representation, like student belonging, is likely to interact with some social-psychological processes known to influence student achievement (Dee 2005; Gershensona, Holta, and Papageorgec 2016; Irvine 1990).
DISCUSSION
The Coleman Report is quite explicit about the benefits of schooling black and white youth together. But is spatial proximity alone sufficient for improving the educational experiences of students of color? Even within resource-rich schools such as North Village Prep and South County Prep, educational inequality occurs daily through unequal access to classrooms and programs. School practices such as tracking shape not only students’ everyday engagement in school and the rigor of their academic experience (Carter 2012; Lewis and Diamond 2005; Tyson 2011) but also their preparation for higher education (Lucas, this issue). In turn, the separation of students in classrooms and extracurricular activities inhibits the development of interracial peer and friendship networks (Granovetter 1985; Moody 2001), which are potential forms of social capital.
Moreover, the resegregation of students within schools influences their attitudes toward one another, their classes, and their teachers, and it also influences how their teachers perceive them. A string of studies over the last five decades has found that teachers’ attitudes and expectations are associated with student performance (Bates 1990, 11; see also Alexander, Entwisle, and Thompson 1987; Dee 2005; Gershensona, Holta, and Papageorgec 2016; Rosenthal and Jacobson 1968). At North Village High, we found no evidence on the surface that teachers did not care about their VDP students. In fact, several volunteered for extra tutorial work with these students and even spent weekends on “cultural tours” of their students’ neighborhoods (see Carter 2013). They did not expect, however, that these students would ultimately fare as well as the majority of their students in terms of grades and further educational attainment. In contrast, my field notes from South Country Prep are filled with observations of more lackadaisical treatment of students by educators. Stunningly, of the teachers and students I casually queried, few perceived any of the African American students as either “bright” or “smart” or worthy of being pushed into the more academically rigorous classes. Those beliefs were mirrored in the actual course-taking patterns of the African American students at South County Prep (see table 2).
The observational and survey data from these four school case studies provide conceptual insights into how the schools’ sociocultural contexts both directly and indirectly shape the educational experiences and well-being of historically marginalized groups of students. Other papers in this issue make strong, plausible arguments that the Coleman Report suffered from some significant measurement errors, which precluded the finding that schools’ resource contexts are not significantly related to differences in student achievement by race.
My research team and I witnessed radically different resources among the four schools in this study. The two majority-minority schools, with nearly two-thirds of its students on free or reduced-price lunch, did not have nearly as many resources as the two more affluent, white-dominant schools. Still, the students of color attending the poorer, majority-minority schools covered the entire spectrum of school involvement and engagement: from those at one end who were taking college preparatory AP and honors courses and participating in orchestra, chorus, and Model UN to those at the other end who had disciplinary problems and tended to drop out of school. In contrast, their coethnic peers at the more resource-rich schools found themselves more heavily represented at the lower end of the educational and extracurricular status hierarchies. These findings offer a cautionary tale about what happens within schools even when students of diverse socioeconomic and racial-ethnic backgrounds share a school building.
These findings also help us distinguish between the “material” and “sociocultural” contexts of schooling. The following debate among a group of African American boys attending affluent North Village Prep High School sharply captured that distinction:
Judah: The system doesn’t encourage us to interact. Think about it.
Jarvis: Why should we interact?
Jonathan: Yeah, why?
Marcus: Why do you need to be encouraged?
Judah: Because the VDP is not just about coming out to suburban schools, doing the homework and going back to our own homes. The VDP program is about teaching suburban schools, suburban students, what it’s like to live in the inner city, what it feels like to be a person of color. It’s about . . .
Jarvis: I thought that the [purpose of] VDP was to come out and get a good education.
Marcus: Yeah.
Jonathan: Yeah.
Judah: No, it’s not. . . . That’s its main purpose, but the other purpose is to teach each other what we have to offer.
In this debate about the academic and social purposes of the Voluntary Desegregation Program and going to school with more affluent white youth, most of the boys took the conventional, instrumental perspective: students of color should attend a “high-quality” school with good teachers, high scores, and strong financial resources if they want to do well. However, Judah, who was in the minority, challenged them to think beyond the material context because of their experiences and his observations at North Village Prep. To assume that social contact with one another will lead to better outcomes in educational engagement and attainment is to assume that status equality exists in schools, but that is often not the case (Carter 2012; Fine, Weis, and Powell 1997; Lewis and Diamond 2015; Moody 2001; Tyson 2011).
In conclusion, both observed and reported educational experiences from my study highlight a central tension in the American educational sphere. Policymakers and educators—informed, at times, by social science research—focus on the provision of “high-quality” schools for all groups without considering what that actually looks like in daily practice. Here is where various qualitative research studies are useful, although the findings of qualitative researchers are rarely featured in the debates about the roles and causes of “school effects.” Such studies, however, frequently illuminate conceptual areas that large-scale research studies either miss or ignore in their analyses.
To fully understand the impact of school context on student outcomes, we must consider how schools maintain environments that compel students and educators to behave in ways that send strong signals about in- and out-groups. That is, the research must be vigilant about the social and symbolic boundaries embedded in school contexts that can privilege certain groups and marginalize others. Therefore, if researchers want to truly understand the persistence of educational disparities, then they must examine more deeply how inequality penetrates social relationships in school environments—from the lunchroom to the orchestra room to the classroom. They must also commit to understanding the relationship between the school's organizational and cultural context and educational opportunity gaps that stubbornly persist both within and across communities.
Acknowledgments
This research was supported by grants from the William T. Grant Foundation (7895), the Spencer Foundation/NAE Postdoctoral Fellowship, and the William F. Milton Fund at Harvard University. I am also grateful for the research assistance of several former graduate and undergraduate students: Grace Atukpawu, Jakeya Caruthers, Jessica Foster, Megan Holland, Rejoice Nsibande, Lazeena Rahman, Lauren Rivera, Carla Shalaby, Graziella da Silva, Kathleen Sullivan, Rebecca Sullivan, Audrey Alforque Thomas, Dana Van Deman, Charnise Virgil, Josh Wakeham, Benita Washington, and Laura Wentworth.
FOOTNOTES
↵1. Today there are highly visible material disparities and significant variation among the states, with per-pupil expenditures in 2008 ranging from nearly $18,000 in Vermont to just over $6,000 in Utah (Baker, Sciarra, and Farrie 2010).
↵2. The most recent data from the Civil Rights Project at UCLA (Orfield et al. 2014) on segregation shows that although desegregation progress was very substantial for African Americans in the South from the mid-1960s to the late l980s, the South has now lost all of the additional progress made after 1967. Still, the South is the least segregated region for black students. Meanwhile, the growth of segregation has been most dramatic for Latino students, particularly in the West, where there was substantial integration in the l960s. Latinos are also significantly more segregated than blacks in suburban America.
↵3. Pseudonyms are used throughout this paper to protect the identities of the schools, students, and staff and to mask their locations.
↵4. National Center for Education Statistics (NCES), Common Core of Data (CCD): public schools, 2005–2006 academic year.
↵5. NCES, CCD: public schools, 2005–2006 academic year.
↵6. VDP is a state-funded program designed to eliminate racial imbalance through the busing of children of color from NCC to public school systems in surrounding suburban communities.
↵7. Some of this section draws heavily on work previously described in Carter (2012).
↵8. Michèle Lamont and Virág Molnar (2002) define “social boundaries” as the lines of demarcation between, or categorizations of, people or practices that are often used to justify the allocation of unequal resources; they define “symbolic boundaries” as the conceptual and cultural tools used to categorize groups to separate the “us” from the “them.”
↵9. Founded in 1987, Challenge Day, a program that is part of the Be the Change Movement—which promotes kind and compassionate acts for others and the infusion of humanist tendencies throughout society—was developed by a husband-wife team of motivational speakers. The founders travel to high schools and have students and teachers spend time in one-day retreats discussing and breaking across social boundaries, whether black versus white, poor versus middle-class, popular versus unpopular, overweight versus slim, or bully versus nerd.
↵10. See Tatum (1997) for social and psychological explanations of why youth in school tend to separate by race and ethnicity.
↵11. Because the test results come from two different states, the range of scores varies among the four schools but are the same for the two schools in each state.
- Copyright © 2016 by Russell Sage Foundation. All rights reserved. Printed in the United States of America. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Reproduction by the United States Government in whole or in part is permitted for any purpose. WThis research was supported by grants from the William T. Grant Foundation (7895), the Spencer Foundation/NAE Postdoctoral Fellowship, and the William F. Milton Fund at Harvard University. I am also grateful for the research assistance of several former graduate and undergraduate students: Grace Atukpawu, Jakeya Caruthers, Jessica Foster, Megan Holland, Rejoice Nsibande, Lazeena Rahman, Lauren Rivera, Carla Shalaby, Graziella da Silva, Kathleen Sullivan, Rebecca Sullivan, Audrey Alforque Thomas, Dana Van Deman, Charnise Virgil, Josh Wakeham, Benita Washington, and Laura Wentworth. Direct correspondence to: Prudence L. Carter at plcarter{at}berkeley.edu, Graduate School of Education, 1501 Tolman Hall, Berkeley, CA 94720.
Open Access Policy: RSF: The Russell Sage Foundation Journal of the Social Sciences is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.