Counting on Context: Cross-Sector Collaborations for Education and the Legacy of James Coleman’s Sociological Vision

A National Study of Cross-Sector Collaboration for Education

To explore these questions, we draw from two components of a comprehensive study of cross-sector collaborations for education in which we have been engaged. ¹ The first is a national scan of collaborations around the United States. Our research team sought collaborations that were anchored in goals related to education, that involved the local K–12 school system and at least one additional sector (e.g., city government, social agencies, business, charitable foundations, higher education), and that operated at the level of a neighborhood, city, school district, or region (rather than being affiliated with just one school). We looked for programs that were currently in operation and had a presence on the Internet as of January 2015, using a broad keyword search strategy with additional efforts to locate collaborations in the one hundred largest cities and one hundred largest school districts in the country. We identified 182 collaborations and analyzed information presented on their websites, including links to published reports and other sources (see Henig et al. [2016] for further information about the scan methodology and findings).

The second component is a set of three intensive case studies of cross-sector collaboration for education. These include the Milwaukee Succeeds initiative in Wisconsin, affiliated with the StriveTogether network; Say Yes Buffalo in New York, one of several citywide initiatives across the country sponsored by the Say Yes to Education foundation; and the All Hands Raised partnership in Portland and Multnomah County, Oregon, another StriveTogether affiliate. For these case studies, we and other members of our research team visited the sites at least twice over the course of a year and conducted more than two hundred interviews, along with observations, document collection, and check-in conversations before and after the site visits. The analysis of these data is under way; in this article, we report preliminary findings.

The Coleman Report and What It Set in Motion

The release in 1966 of EEO, colloquially known as the Coleman Report, was a pivotal event that eventually became thought of as marking the beginning of a new federal role in research on schooling. Although the government had been collecting data about students, teachers, and schools and colleges since 1870, these data were mostly descriptive documentations of educational inputs and expenditures. A study a few years earlier, the Project Talent study shared many characteristics with EEO but did not have the same impact (Hutt 2016). The Coleman Report was notable for being comprehensive: it used a large, nationally representative survey database—more than 600,000 students in a 5 percent random sample of public schools (see Hill, this volume, for additional details). Methodologically, it was notable for using standardized tests as measures of student performance across elementary and secondary grades, and for employing statistical methods (primarily stepwise decomposition of variance) that, while inadequate by today’s standards, linked contextual conditions and inputs in a production function analysis of outcomes. This approach took hold and a few years later the government began to monitor educational outcomes regularly through the National Assessment of Educational Progress and also began numerous longitudinal and cross-sectional survey research programs for purposes of correlational and causal analysis (Snyder 1993). The study did not draw from, articulate, or test a specific theory about how and why student academic achievement might vary across schools and students. Theory-based analyses of instruction and learning were not as well developed at the time as they are now, and the government’s intentions behind EEO constrained Coleman to focus on statistical associations rather than theoretical elaboration. But the selection of measures included in the survey suggested a conceptual framework that linked student performance to individual and contextual background variables and to rudimentary features of the instructional program, and this led to new ways of thinking about educational productivity and equality.

The EEO acquired some notoriety before it was even released. Controversies erupted about the wording and intent of survey items, superintendents of prominent districts declined to participate, data were lost and surveys had to be readministered, and the project’s advisory board refused to sign off on the report because of disagreements over the statistical analyses (Egalite 2016). Because its findings were in many ways unexpected and controversial, it spawned an extensive amount of examination, discussion, and reanalysis over the years (Hill, this volume; Mosteller and Moynihan 1972).

As consequential as it was for education research, the Coleman Report was a “fountainhead study” for those who more generally wanted to promote the use of evidence to inform policy (Hanushek 2016). It was commissioned in 1964 as Section 402 of the Civil Rights Act, tacked onto a bill that was known at the time mostly for its provisions regarding voting rights and public accommodations (Grant 1973). Initially, the bill called for the Commissioner of Education to “investigate … the extent to which equal educational opportunities are denied to individuals” (Grant 1973, 18), but the wording was changed to a less incendiary call for a survey to document “the lack of availability of equal educational opportunities for individuals by reason of race, color, religion, or national origin in public educational institutions at all levels” (Coleman et al. 1966, iii). The report itself contained no recommendations, but as the U.S. Commissioner of Education, Harold Howe II, wrote in his letter of transmittal, it was conducted in such a way as to be “useful to those concerned with public education in the United States” (p. iv).

As the study moved toward completion, Coleman and others in the Office of Education were well aware that it could impact policy in unintended and potentially undesirable ways. Hence, much attention was paid to crafting the executive summary. It was seen as so politically sensitive that three versions were prepared, none written by Coleman himself (Grant 1973). While the controversy swirled around the summary, Coleman “had barricaded himself in a motel in southwest Washington,” with one change of clothing, working nonstop on the full report while meals were delivered along with daily analysis reports sent down from the computer center in Princeton via Greyhound bus (Grant 1973, 28). In the end, in the third and final version of the summary, the main finding was tucked beneath a deluge of tables and graphs: “It appears that differences between schools account for only a small fraction of differences in pupil achievement” (Coleman et al. 1966, 22). A few prominent newspapers reported the study’s release. It garnered a short announcement, which included what might be the first public mention of the “achievement gap,” on page 24 of the Saturday edition of the New York Times on July 2, 1966, under the headline “Negro Education is Found Inferior: U.S. Public Survey Confirms Racial Disparity.” The article buried the lead that “the survey clearly showed that ‘family background is more important than schools’” (New York Times 1966a).

For the next two years, nothing happened programmatically in the Office of Education because of the study. But at the same time, largely through the efforts of Daniel Patrick Moynihan, the report was pored over by Congress and eventually by the media and the public (Grant 1973; Hill, this volume). Senator Abraham Ribicoff of Connecticut accused the Office of Education of suppressing the report and asserted that “it should be subjected to exposure, argument and debate in every school board and PTA in the country” (New York Times 1966b). The report was invoked on different sides of education policy debates during the Johnson and Nixon administrations, as justification both for a doubling down on desegregation and for a proposed reduction in education spending for schools, which Nixon needed to offset costs of the Vietnam War. By 1970, the report was widely known, and Coleman was brought back to Washington as a consultant to the Cabinet Committee on Desegregation. This led eventually to a second Coleman Report (Coleman, Kelly, and Moore 1976) that documented extensive white flight from big cities following desegregation. This laid Coleman open to intense criticism and even vilification from his professional colleagues and eventually helped to fuel a subtle but pervasive retreat from desegregation as a core strategy for educational equity (Peterson 2010).

“Schools don’t matter” was an immediate and frequently invoked summary of EEO’s findings, but one that does not capture the full story. Whether this conclusion was just plain wrong or merely simplistic, the Coleman Report energized both those who wanted to refute this implication as well as those who accepted it. It ushered in a long era of attention to two sets of factors influencing student performance: policies, resources, and practices that could strengthen the ability of schools to make a difference for their students; and, conversely, policies and practices that bypassed the schools and were meant to help ameliorate the dampening effects of students’ own backgrounds and the compositional effects generated by the social backgrounds of their peers.

Coleman’s Legacy for Strengthening the Impact of the School

The Coleman Report found that school-related conditions such as teacher quality seemed to influence students from minority backgrounds more than white students, even if the effects were relatively small compared to other factors. Coleman and colleagues wrote: “It is for the most disadvantaged children that improvements in school quality will make the most difference in achievement” (1966, 22). This was heartening news to those looking to increase resources for schools. Since then, analysts and policy-makers have sought ways to increase the effect schools can have on equality of educational opportunity, in part by examining the many in-school factors the study did not measure, and by attending to the factors that the study did explore (like teacher quality), to make them more influential and equitable. Since Coleman was not a pedagogical expert and, as he admitted, he had only crude measures of school quality, many stones were left unturned in the study itself. The ensuing years have seen much research on the effects of curriculum content, pedagogical techniques, instructional pacing, academic tasks, class size and composition, and many other in-school factors on academic excellence and equity. Thus, in seeking to link cross-sector collaborations to the Coleman Report, we first ask: Do cross-sector collaborations extend the legacy of EEO in trying to bolster the ability of schools to make a difference for students?

Coleman’s Legacy for Addressing Family and Social Context Effects

The final paragraph of text in the Coleman Report offered a stark conclusion:

Taking all these results together, one implication stands out above all: That schools 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. (Coleman et al. 1966, 325)

Like the results about school effects, this finding has also been subjected to much scrutiny. EEO was not able to go very far in presenting a conceptual framework for explaining the impact of a student’s background or the composite social background of a school’s student population on achievement. For the most part, however, the impact of a student’s own family and social background continues to hold up as “far more important than school social composition and school resources for understanding student outcomes” (Borman and Dowling 2010, 1202). Some have interpreted this to endorse the importance of families doing particular things to support their children’s education and have sought strategies for enhancing family involvement (see Liu and White, this volume; Kraft and Monti-Nussbaum, this volume). Many interpreted it to mean that poverty was the culprit causing a negative family effect on academic performance. In response, direct poverty reduction strategies, urban community development, and school desegregation were employed as ways to improve the social backgrounds of students.

Others looked for ways in which social policy might help to provide some of the tangible benefits associated with family social class that seem to enhance student performance in school. For example, more advantaged students have access to health and social services that help to prepare and enable them to learn, early childhood education and out-of-school services that enhance their development, cultural experiences that enrich their studies, and assistance with the finances and logistics of moving from high school to higher education. This has led to policy initiatives for comprehensive or “wraparound” services to reduce disparities in what is available to advantaged and disadvantaged students (Egalite 2016; Rebell and Wolff 2011; Rothstein 2004). This also leads to our second research question: Do cross-sector collaborations help to ameliorate the impact of students’ social backgrounds on school performance?

Coleman’s Legacy for the Social Milieu and Collective Socialization

It is possible to interpret contemporary cross-sector collaborations as a legacy of the Coleman Report in a third way. To draw this through-line, we place EEO in the context of Coleman’s broader sociological vision and body of scholarship. In much of his writing, Coleman was motivated by an interest in social institutions and systems. His approach was often structural-functionalist and Durkheiman in nature (Marsden 2005; Pallas 1999), exploring how social institutions could influence the actions of individuals and create a cohesive society. He saw this happening primarily through the social transmission of values, expectations, and motivations. In a seemingly sentimental analysis, which he repeated many times, Coleman extolled the family as the primordial social institution responsible for providing structure and socialization for the young. When economic activity was centered in family farms or small enterprises, children were watched over, educated, and prepared for their future social and economic lives within the household unit. Communities also served as socializing contexts for transmitting values and motivations, which were reinforced because community status hierarchies were congruent with them. To succeed, members of a community followed its norms and avoided its sanctions.

But these social forms have weakened over time, Coleman argued. Men’s work migrated from the home to industrial organizations and businesses, and young people were sent to school to be prepared for those settings. Once women (especially mothers of young children) also sought work outside the home, the family as an institution was further “detonated” and “evacuated,” “shrinking in size, function, and capability” and schools were relied upon for custodial care as well as education (Coleman and Hoffer 1987, 19). In some contexts, families and communities had become so buffeted by social and economic upheavals that the schools had to take over basic socialization functions for them. Some of Coleman’s portrayals of these patterns sound downright alarming: Communities found themselves separating children from adults in their own households (now “emptied out during the day”) and from the places where adults were to be found (Coleman 1961). As schools became remote sites for intellectual development, families had “an extremely variable ability to socialize children for achievement” (Coleman 1971, 648). The family had turned into “a kind of backwater in society, cut off from the mainstream” (Coleman 1987, 33).

When family and community fail to perform their socialization functions, what happens to the schools? Coleman explored a version of this question in The Adolescent Society (1961). How, he wondered, were youth going to learn important social lessons that would motivate them to do well in school when their main relations were with each other and with large, impersonal social institutions like the school? In his analysis of the closed social system of peer group culture in high schools, he found that to some degree, students’ attitudes about academic achievement and aspirations for college were responsive to adult value systems in the school—a “schools matter” argument. Community social class and values and students’ family background also had some impact on student academic performance. But these socializing forces did not necessarily predict what was most important to the peer status culture or who would be in the “leading crowds” in schools, and adolescent peer culture became a “small society” on its own that was often at cross-purposes with educational values. Coleman implicated the wider social system in this analysis:

It is the adult community that fixes the activities of the adolescent community. It does so by fixing the activities of the school—for example, by using high-school sports as community entertainment, and as contests between communities—and by restricting adolescent activities outside of school. The adolescent has little or no possibility of responsibility today; adults have shut him out of the job market—have told him, in effect, to go off to school and not bother them. (Coleman 1961, 305)

Coleman’s The Adolescent Society (1961) preceded his massive study for the U.S. government by half a decade. Five years after EEO, he again took up the matter of socialization, and once again he saw a role for schools but also for the larger society. In “How do the young become adults?” (Coleman 1972), he once more noted that the home had become depleted as a site for socialization:

In the family, the young remain, while the activities from which they could learn have moved out; in workplaces, the activities from which the young could learn remain, but the young themselves have been excluded. This exclusion places youth more on the fringes of society, outside its important institutions. …The young have no institutional base, they are a lumpen proletariat. (Coleman 1972, 433)

Coleman’s proposed solution at this point became even more dramatic: attention to students’ academic and intellectual work in schools needed to be supplemented by other efforts to enlist the “economic institutions of society” in giving all students opportunities to learn the knowledge and skills they would need to be well-rounded and productive citizens. Academic and vocational education had to go hand in hand; students should not be tracked into curricula emphasizing just one or the other.

In Coleman’s early writings, these relationships among families, communities, and schools were described in sentimental and romantic terms, as essential elements of a tightly woven social fabric. Later on, he used a more hard-edged vocabulary to describe the same phenomena. Social relationships in value-rich “functional communities,” especially those whose number, strength, and quality of ties created “intergenerational closure,” created “social capital” that enhanced the ability to socialize the young and develop their human capital (Coleman 1988; Coleman and Hoffer 1987). Once again, he seemed provincial in blaming the loss of social capital and subsequent socialization problems on things like the structural deficiency of a one-parent household, the weakness of communities no longer held together by shared values and geographic proximity, and even “increasing psychic involvement of youth with mass media” (Coleman and Hoffer 1987, 230). But his point was that this long decline for families and communities will continue, and society must now look to the large institutions like schools to replace families and communities and to provide the norms and obligations that serve as socializing guides and motivations for the “engine of action” that young people would employ via rational self-interest (Coleman 1988, 1990). These large institutions cannot simply be more intensive versions of schools as we know them. Instead, children need to encounter not just the traditional demands, opportunities, and rewards of schooling, which are not motivating enough on their own, but also socializing features that could “induce the kinds of attitudes, effort, and conception of self that children and youth need to succeed in school and as adults” (Coleman 1987, 38). Whether these socializing features are engendered by adults manipulating the adolescent society in schools (Coleman 1961), or by using social capital in functional communities to transmit values (Coleman 1988), or by developing bonds between teachers and students around the shared expectation of being accountable for performance (Coleman et al. 1997), they turn schools into active agents of socialization, not just beneficiaries of the socialization accomplished through families and communities.

The strongest clue that socialization is what Coleman cared most about was in “A Vision for Sociology” (Coleman 1994), a reflection he offered late in his career, in which he reminisced about the watershed period for sociology in the 1950s. During this period, he explained, a shift was under way between, on one hand, sociologists’ traditional interest in community studies—where the community was the unit of analysis and the goal was to examine the norms and values, status systems, and cleavages of social life—and, on the other hand, their burgeoning interest in survey research, which used systematic data and the statistical analysis of variables to develop causal accounts of individuals’ actions and attitudes. In a more analytic piece on this same theme (Coleman 1986), he suggested that sociologists were unable to unify their “social theory” agenda with their “social research” innovations because they were interrupted by the government’s need for applied social research that could provide guidance for programmatic interventions with specific social problems and populations. Their research pivoted to empirical studies with individuals as the unit of analysis.

In this so-called watershed moment, Coleman tried to bridge community studies and survey research by exploring the closed social system of high school and its impact on students. Although his data did not enable him to fully execute the intended analysis, this work became The Adolescent Society (1961).

Then, he had another chance with EEO. This time, he leaned more deliberately in the direction of the statistical analysis of individual outcomes. Given the intended aims of the government-sponsored study, Coleman focused on monetized resources in the school—the books, labs, debate teams, music programs, and well-trained and experienced teachers—that could be manipulated by social policy, and he examined their impact on academic outcomes for members of social groups defined by race. This was the analysis that galvanized the study’s many audiences. It was empirical and policy-relevant, destined to become a landmark. But Coleman himself described it as a detour from his enduring interests (Marsden 2005).

As it turned out, however, the study did enable him to document at least some effects of a social system on individuals, because it included numerous measures of the social milieu of students’ families and communities, including parent education, family structure, educational resources, and expectations for their children. These were factors that Coleman already believed produced variability in the socialization of youth toward academic performance. He also studied teachers’ expectations for their students, and students’ own self-efficacy or sense of control, two more indicators of the value climates in schools that could affect performance.

Presented as a production function analysis, EEO ended up bearing a close resemblance to classic sociological studies of social context, with the added bonus of measuring effects on individuals. Although it lacked an explicit conceptual framework and explanation of mechanisms (Sørensen and Morgan 2000), it led to the finding that mattered most to Coleman: that the “inequalities imposed on children by their home, neighborhood, and peer environment” are inequalities of socialization and social context, and they have a stronger effect on children and youth than the educational features of the schools themselves.

It also was the finding that Alexander Mood, the U.S. Office of Education official who brought Coleman in to direct the study, pointed to in his second version of the EEO executive summary, which never saw the light of day: “Thus these data suggest that the major source of variability among schools is not a result of the school effects, but of differences in the community that affect the children’s receptiveness to the school” (Grant 1973, 26). And it probably was the finding that spurred Daniel Patrick Moynihan to pay so much attention to the report, because it seemed to resonate with his own controversial work, The Negro Family (Moynihan 1965).

In 1987, Coleman summarized the EEO findings about family background as follows:

The outputs of education result from the interaction of qualities the child brings from home—which can be loosely characterized as attitudes, effort, and conception of self—with qualities of the school. As the social capital in home and neighborhood shrinks, school achievement and other growth will not be increased by replacing these resources with more school-like resources—that is, those that produce opportunities, demands, and rewards—but by replacing them with resources which produce attitudes, efforts, and conception of self—that is, those qualities that interact with the ones provided by the school. (p. 38)

Coleman understood full well the importance of the educational resources that schools provide. Indeed, in his biographical musings, he explained that he was motivated in this work by his own frustrations with dismal, stultifying, uninspiring schooling (Pallas 1999). But he cared even more that students have ample opportunities to develop the social values that would lead to their well-being and prosperity. Although EEO gave birth to a school effects industry that pursued more research evidence about, and practical development of, the inner workings of schools, and although EEO also ushered in equally ambitious efforts to understand and intervene around family/social background effects, what really interested Coleman was the school as a functional social system and how the social milieu affected youth. In this regard, EEO was not such a detour after all. This brings us to our final question: Do cross-sector collaborations lead to better social contexts and enhance the collective socialization capacities of cities?

Conclusion

In a pragmatic sense, cross-sector collaborations for education aim to improve academic outcomes for children and youth in their locales. To the degree that they seek to do so by supporting schools and school systems in their efforts to improve, and by expanding and aligning comprehensive, wraparound social services and other resources for children and youth, they are pursuing strategies consistent with the EEO findings and are carrying forward Coleman’s legacy.

We argue that cross-sector collaborations also have the potential to help realize James Coleman’s deeper vision for social institutions that provide a consistent, nurturing, and normative context in which all children and youth can learn the values and habits that will enable them to thrive academically, economically, and socially. These collaborations set their sights on whole cities or, in some cases, regions defined by other boundaries. It is daunting indeed to imagine how a social context on that scale, in our era, can be a functional community, with cohesive values and strong social ties, that binds actors of many different backgrounds and with varying resources and disparate interests in a common pursuit on behalf of a diverse population of young people. It requires a level of political, economic, and educational acuity and will that was not present, for example, in many of the attempts to desegregate schools following the Coleman Report. The EEO was an empirical study that led to a solution embroiled in politics and culture. Cross-sector collaboration may be a newer solution, but one embedded in politics and culture as well. Coleman sought a social theory of the school because he thought the school was an institution of a scale that one could examine and understand in its entirety, and it was a target of policy efforts, always a compelling draw. Cross-sector collaborations invite thinking about a social theory of the city. If public policy looks for place-based solutions to the many problems connected to inequality, a social theory of the city might be just what James Coleman would prescribe.