SOCIOLOGY OF EDUCATION

In the broadest perspective, education refers to all efforts to impart knowledge and shape values; hence, it has essentially the same meaning as socialization. However, when sociologists speak of education, they generally use a more specific meaning: the deliberate process, outside the family, by which societies transmit knowledge, values, and norms to prepare young people for adult roles (and, to a lesser extent, prepare adults for new roles). This process acquires institutional status when these activities make instruction the central defining purpose, are differentiated from other social realms, and involve defined roles of teacher and learner (Clark 1968). Schools exemplify this type of institutionalization.

The central insight of the sociology of education is that schools are socially embedded institutions that are crucially shaped by their social environment and crucially shape it. The field encompasses both micro- and macro-sociological concerns in diverse subfields such as stratification, economic development, socialization and the family, organizations, culture, and the sociology of knowledge. To understand modern society, it is essential to understand the role of education. Not only is education a primary agent of socialization and allocation, modern societies have developed formidable ideologies that suggest that education should have this defining impact (Meyer 1977).

Durkheim (1977) was the intellectual pioneer in this field, tracing the historical connections between the form and content of schools and larger social forces such as the rise of the bourgeoisie and the trend toward individualism. Largely because the field focuses so intensively on stratification-related issues (e.g., the impact of family background on educational attainment), the larger issues raised by Marx and Weber are readily evident in current scholarship. However, as Dreeben's (1994) historical account indicates, the direct contribution of the discipline's founders to the development of the sociology of education in the United States was minimal; indeed, even the foremost early American sociologists in the field did not decisively shape its development.

In The Sociology of Teaching, Waller (1932) examined teaching as an occupational role and school organization as a mechanism of social control. He emphasized the role of the school in the conflict-ridden socialization of the young as well as the interpersonal and organizational mechanisms that furthered students' acceptance of the normative order. Although now recognized as a classic, Waller's analysis stimulated little work for several decades.

Although less focused on education per se, Sorokin (1927) portrayed schools as a key channel of mobility with their own distinctive form of social testing. He argued that increasing opportunities for schooling would stratify the society, not level it. However, Blau and Duncan's (1967) paradigm-setting study of status attainment (see below) did not refer to Sorokin's analysis of education despite their appreciation of his larger concern for the significance of social mobility. Warner's and Hollingshead's community studies considered education integral to community social organization, especially through its connection to the stratification system, but their influence, like that of Waller and Sorokin, was more a matter of suggesting general ideas than of establishing a cumulative research tradition.

As a subfield within the sociological discipline, the sociology of education has been propelled largely by a host of practical, policy-related issues that emerged with the development of the mass educational system. Essentially, research has focused on whether education has delivered on its promise of creating more rational, culturally adapted, and productive individuals and, by extension, a "better" society. The field was particularly energized by the egalitarian concerns of the 1960s: How "fair" is the distribution of opportunity in schools and in the larger society, and how can disparities be reduced? These questions continue to animate the field.

THEORETICAL DEBATES

Much research, even the most policy-oriented, has been grounded, often implicitly, in more general analytic perspectives on the role of education in modern society. The two main orientations are functionalism and conflict theory, though other, less encompassing perspectives also have shaped the field significantly.

Functionalism. In the functionalist view, schools serve the presumed needs of a social order committed to rationality, meritocracy, and democracy. They provide individuals with the necessary cognitive skills and cultural outlook to be successful workers and citizens (Parsons 1959; Dreeben 1968) and provide society with an efficient, fair way of sorting and selecting "talents" so that the most capable can assume the most responsible positions (Clark 1962). Complementing this sociological work is human capital theory in economics, which contends that investment in education enhances individual productivity and aggregate economic growth (Schultz 1961). The criticism in the 1980s that poor schooling had contributed significantly to America's decline in the international economy reflects a popular version of this theoretical orientation.

However, in the 1970s, both the increasing prominence of critical political forces and the accumulated weight of research spurred a theoretical challenge. Important parts of the empirical base of functionalism were questioned: that schools taught productive skills, that mass education had ushered in a meritocratic social order, and that education had furthered social equality. A number of conflict-oriented approaches emerged.

Neo-Marxist Theory. Neo-Marxist scholars have provided the most thorough challenge to the functionalist position. For all the diversity within this conflict theory, the main point is that the organization of schools largely reflects the dictates of the corporate-capitalist economy. In the most noted formulation, Bowles and Gintis (1976) argue that education must fulfill the needs of capitalism: efficiently allocating differently socialized individuals to appropriate slots in the corporate hierarchy, transferring privilege from generation to generation, and accomplishing both while maintaining a semblance of legitimacy. Thus, the changing demands of capitalist production and the power of capitalist elites determine the nature of the educational system.

More recent neo-Marxist scholarship (Willis 1981) emphasized that schools are not only agents of social reproduction but also important sites of resistance to the capitalist order. Many neo-Marxists also have emphasized the "relative autonomy" of the state from economic forces and, correspondingly, the partial responsiveness of schools to demands from subordinate groups (Carnoy and Levin 1985). Other scholars in this general critical tradition have turned in "post-Marxist" directions, emphasizing inequities related to gender and race along with class, but the common, defining point remains that educational inequities reflect and perpetuate the inequities of capitalist society and that oppressed groups have an objective interest in fundamental social transformation (Aronowitz and Giroux 1985). This newer critical approach has developed with relatively little connection to mainstream approaches (i.e, positivistic, often reformoriented research) despite some similarities in concerns (e.g., student disruptions and challenges to authority in schools) (Davies 1995).

Obviously, neo-Marxists do not share the essentially benign vision of the social order in functionalist thought, but both perspectives view the organization of schooling as "intimately connected with the changing character of work and the larger process of industrialization in modern society" (Hurn 1993, p. 86). These competing perspectives are rooted in similar logical forms of causal argument: To explain educational organization and change, functionalists invoke the "needs" of the society, while neo-Marxists invoke the "needs" of the capitalist order for the same purpose. Critics contend that both perspectives posit an overly tight, rational link between schools and the economy and concomitantly downplay the institutional autonomy as schools as well as the complexity of political struggles over education (Kingston 1986).

Status Conflict. Arising out of the Weberian tradition, the status conflict approach emphasizes the attempts of various groups—primarily defined by ethnicity, race, and class—to use education as a mechanism to win or maintain privilege (Collins 1979). The evolving structure of the educational system reflects the outcomes of these struggles as groups attempt to control the system for their own benefit. With varying success, status groups use education both to build group cohesion and to restrict entry to desired positions to those certified by "their" schools. However, as lower-status groups seek social mobility by acquiring more educational credentials, enrollments may expand beyond what is technically necessary. In this view, then, the educational system is not necessarily functional to capitalist interests or other imputed system needs.

Consistent with this view, a primary effect of schools, especially at the elite level, is to provide cultural capital, of which educational credentials are the main markers (Bourdieu and Passeron 1977). This form of capital refers to the personal style, social outlooks and values, and aesthetic tastes that make a person suitable for socially valued positions. (The point of comparison is human capital, an individual's productive, technical skills.) In this perspective, education is rewarded because occupational gatekeepers value particular forms of cultural capital, and thus education is a key mechanism of class and status reproduction.

The Interpretative Tradition. Sociologists in the interpretative tradition view schools as places where meaning is socially constructed through everyday interactions. This tradition incorporates the general orientations of phenomenology, symbolic interactionism, and ethnomethodology. Accordingly, micro-level concerns predominate—for example, what do teachers expect their students to learn, and how do those expectations condition their conduct in class?—and research tends to rely on qualitative techniques. This tradition is unified by a general sense of what kinds of questions to ask (and how to ask them) rather than a set of related theoretical propositions or a body of accumulated findings.

EMPIRICAL STUDIES

The highly selective review of empirical studies that follows focuses on the two key questions in contemporary American sociology of education: (1) How is education involved in the distribution of life chances? (2) How are family status and school characteristics connected to educational attainment and/or academic achievement? With few exceptions, analyses of education in other countries are not considered. The field is dominated by American research, and American sociologists have engaged in relatively little comparative research. Baker (1994) speculates that this lack of a comparative research tradition in the United States reflects both a belief in American "exceptionalism" (for instance, an extreme emphasis on mass access) and a strong focus on micro-level issues that do not necessarily call for comparative research designs.

Schooling and Life Chances. Throughout the twentieth century in all industrial countries, there has been a dramatic upgrading in the occupational structure and a dramatic expansion in educational systems. Ever more jobs have come to require academic qualifications, a process that usually is interpreted as being driven by the rationalism and universalism of modernization. In this functionalist perspective, academic skills are presumed to be technically required and meritocratically rewarded, transforming the stratification system so that individual achievements rather than ascriptive characteristics determine life chances.

This interpretation has been subject to empirical test at two levels: (1) the individual level—to what extent, absolutely and relatively, does education affect economic attainment? and (2) the macro level—to what extent have educational expansion and the increasing significance of schools for occupational attainment increased overall equality of opportunity?

At the first level, as part of the general analysis of status attainment, researchers have concentrated on measuring the connection between individuals' schooling and their economic position. Building on Blau and Duncan's (1967) work, researchers have repeatedly documented in multivariate models that education (measured in years of schooling and degree completion) has by far the largest independent impact on adult attainment (Featherman and Hauser 1978; Jencks et al. 1979). By comparison, the net direct effects of family status (usually measured in terms of parental education and occupation) are modest. Indeed, among the college-educated in recent years, higher family status confers no extra advantage at all (Hout 1988).

Earlier in life, however, family status is substantially related to educational attainment. The total effect (direct and indirect) of family status on occupational attainment is therefore substantial, though its impact is mediated very largely through educational attainment. In effect, then, education plays a double-sided role in the stratification process. Education is the great equalizer: It confers largely similar benefits to all regardless of family origins. However, it is also the great reproducer: Higher-status families transmit their position across generations largely through the educational attainment of their children.

The strong connection between schooling and occupational attainment is open to diverse interpretations. Most prominently, human capital theory suggests that education enhances productivity, and because people are paid in accordance with their marginal productivity, the well educated enjoy greater prospects. In favor of this interpretation is the fact that schooling is demonstrably linked to the enhancement of academic competencies (Fischer et al. 1996) and that basic academic skills are substantially correlated with job performance in a wide variety of settings (Hunter 1986).

By contrast, credentials theory portrays the educational institution as a sorting device in which individuals are slotted to particular positions in the occupational hierarchy on the basis of academic credentials, often with little regard for their individual productive capacities. The fact that possessing specific credentials (especially a college degree) has positive career effects, net of both years of schooling and measured academic ability, provides indirect support for this view. That is, there appears to be a "sheepskin effect," so that employers value the degree per se, although people with degrees may have unmeasured productive capacities or dispositions that account for their success ( Jencks et al. 1979). Moreover, the credentialist argument is strengthened by the fact that in some elite segments of the labor market, employers primarily recruit graduates of certain prestigious programs and make little effort to discern differences in the academic-based skills of those included in the restricted applicant pool (Kingston and Clawson 1990).

Both views seem to have some merit; indeed, they may be partially complementary. Employers may generally use educational attainment as a low-cost, rough proxy for productive skill, and for certain positions they may favor holders of particular degrees because of their presumed cultural dispositions and the prestige that their presence lends the organization. The relative explanatory power of the human capital and credentialist perspectives may vary across segments of the labor market.

At the macro level, it might be expected that the great expansion of access to education has reduced the impact of family origins on educational attainment, increasing equality of opportunity, but that has proved to be more the exception than the rule. A rigorous thirteen-country comparative study identified two patterns: greater equalization among socioeconomic strata in the Netherlands and Sweden and virtual stability in the rest, including the United States (Shavit and Blossfeld 1993), where the strata have largely maintained their relative positions as average attainment has increased. Thus, the impact of educational policies designed to promote equality appears minimal; even in Sweden and the Netherlands, the trend toward equalization emerged before reforms were introduced.

Socioeconomic Status and Achievement. Given the centrality of educational attainment in the general attainment process, researchers have focused on the substantial relationship between socioeconomic status and educational attainment. (This relationship appears to be stronger in highly developed societies than in developing societies.) The best predictor of educational attainment is academic achievement (i.e., higher grades and test scores); the school system consistently rewards academic performance and in that sense is meritocratic. Regardless of academic performance, children from socially advantaged families have somewhat disproportionate success in moving through the educational system, but the main reason higher-status students have this success is that they achieve better in schools.

The question here is, Why do higher-status students achieve better in schools? Clearly, there is no simple answer. Research has pointed to the following family-related factors, among others:

1. Material resources. Richer families can purchase the materials (e.g., books) and experiences that foster intellectual development.
2. Parental expectations and/or encouragement. Well-educated parents more actively stress the importance of academic achievement, and their own success through schooling encourages their children to accept that value.
3. Direct parental involvement in home learning activities. Higher-status parents are more willing and able to teach academic lessons at home and help with homework.
4. Verbal and analytic stimulation. In higher-status families, interactions between parents and children are more likely to promote verbal sophistication and reasoning.
5. Family structure and parenting style. The presence of two parents and parenting styles involving warm interactions favor academic achievement, and both factors are related to socioeconomic status (SES).
6. Parental involvement in schools. Higher-status parents are better able to interact effectively with teachers and administrators to secure favorable treatment and understand expectations.
7. Cultural "fit" with schools. The cultural styles of higher-status students are more compatible with the prevailing norms and values in schools.
8. Social capital. Initially Coleman's (1988) idea, this refers to the extent and nature of the connections between parents and children as well as the connections with other family and community members. By providing informational, emotional, and other resources, these connections facilitate adaptions to the demands of schools.
9. Social context. Higher-status families are likely to live in communities where other families promote achievement and their children's peers are committed to academic achievement.
10. Genetic advantage. Early IQ is related to SES, and intelligence is related to academic performance.

Individually, none of these factors seems to account for a large part of the overall relationship between SES and academic achievement, nor is the relative significance of these factors clear, yet the very length of the list suggests the complexity of the issue. Higher-status students are not all similarly advantaged by each of these factors, and lower-status students are not all similarly disadvantaged by each one. The substantial aggregate relationship between SES and achievement undoubtedly reflects complex interactions among the many home-related contributing causes. As is more thoroughly discussed below, the mediating impact of school resources and practices is much less consequential.

The Racial Gap. The black-white disparity in academic performance remains large despite some notable reductions in recent years, and it is economically significant. A number of researchers have shown that for younger cohorts, the racial disparity in earnings is accounted for very largely by differences in basic academic skills as measured by scores on tests such as Armed Forces Qualifications Test (Farkas 1996).

Why this gap persists is unclear, partly because until recently, sociologists and other social scientists were wary of addressing such a politically explosive issue. Most relevant for the discussion here is the fact this gap cannot be explained by blacks' lesser school resources (see "School Effects," below). Largely drawing on the work of scholars in related fields, the sociological consensus appears to be that the racial disparity does not reflect a group-based difference in genetic potential ( Jencks and Phillips 1998). (At the individual level, there is undoubtedly some genetic component to IQ among people of all races.) Moreover, this gap cannot be attributed largely to racial differences in economic advantage: Socioeconomic status explains only about a third of it. However, a broader index of family environment, including parental practices, may account for up to twothirds of the gap (Phillips et al. 1998). A complete explanation probably will involve many of the factors previously noted in the discussion of the relation between SES and achievement but also include the distinctive cultural barriers that "involuntary minorities" face in many societies (Ogbu 1978) as well as subtle interactional processes within schools.

Racial disparities in educational attainment have declined dramatically. High school graduation rates are now virtually the same, and the remaining disparity in college attendance reflects blacks' lower economic resources, not a distinctive racial barrier.

School Effects. The governmental report Equality of Educational Opportunity (Coleman et al. 1966) strongly challenged conventional wisdom about the connections among economic status, schools, and achievement. In doing so, it fundamentally shaped the agenda for further research in this area.

Attempting to identify the characteristics of schools that improve learning, the so-called Coleman Report documented two key points. First, there is a weak relationship between social status and school quality as measured by indicators such as expenditure per pupil, teachers' experience, and class size despite considerable racial segregation. Second, these measures of school quality have very little overall effect on school achievement (scores on standardized tests) independent of students' family background. The Coleman Report also showed, however, that school effects were notably larger for black and Hispanic students than they were for whites and Asians. Among the school effects, the racial composition of schools was the most critical: Blacks did somewhat better in integrated schools.

Later research largely validated the main conclusions of the Coleman Report, but also modified them, often by considering more subtle aspects of school quality. For instance, some school resources, including expenditures, seem to enhance achievement, but the predominance of home factors on achievement remains undisputed. In regard to another between-schools effect, Coleman argued for the educational superiority of Catholic schools, an advantage he attributed to their communal caring spirit and high academic expectations for all students. The Coleman Report did not consider such cultural matters or specific educational practices. Much of the post–Coleman Report research focused on within-school effects because gross between-school effects appeared to be relatively minor.

Ability grouping in elementary schools and tracking in high schools have attracted attention, largely as a source of inequalities of academic performance. The premise of these practices is that students differ substantially in academic ability and will learn more if taught with students of similar ability. Although many different practices are grouped under the term "tracking," students in the "top" groups generally receive a more demanding education, with higher expectations, more sophisticated content, and a quicker pace, and are disproportionately from advantaged families. The obvious but not fully settled issue is whether schools "discriminate" in favor of the socially advantaged in making placements. At the high school level, controlling for measures of prior achievement (themselves affected by family factors), higher SES seems to enhance one's chances modestly, though achievement factors are predominant in placement. Blacks are somewhat favored in the process if one controls for prior achievement. At the elementary school level, research is less consistent, though one study indicates that neither test scores nor family background predicts early reading group placement (Pallas et al. 1994).

Another important but not fully settled issue is whether students in certain ability groups or tracks learn more because of their placement. Gamoran (1992) shows that the effects of tracking are conditioned substantially by the characteristics of the tracking system (for example, how much mobility between tracks is allowed) and subject matter. However, by way of gross summary, higher track placement per se generally seems to have a modestly beneficial impact on achievement and also seems to increase students' educational aspirations and self-esteem. However, to exemplify the important exceptions to this generalization, it appears that within-class grouping for elementary school mathematics may help both low and high groups.

Teacher Expectations. It is commonly supposed that differences in teachers' expectations explain at least some of the racial and socioeconomic disparities in academic achievement. The claim is that a self-fulfilling prophecy is at work: Teachers expect less from socially disadvantaged students and treat them accordingly, and therefore these students perform less well in school. Rosenthal and Jacobson's (1968) small-scale experimental study provided the initial impetus for this argument, but follow-up studies in real classrooms suggest that teachers' expectations have little or no effects on later performance.

If the standard for fairness is race neutrality in light of past academic performance, there is little evidence of racial bias in teachers' expectations, but some limited evidence suggests that teachers' beliefs are more consequential for blacks than for whites (Ferguson 1998). More generally, research has not established that socially discriminatory practices in schools significantly explain the link between family and/or racial status and achievement.

Contextual Effects. Not only do students come to school with different backgrounds that affect learning, schools provide students with different social environments that are importantly shaped by the economic and racial composition of the student body. Because peers are so influential in children's and adolescents' lives, the obvious question is whether the social composition of a school affects individual learning beyond the effects attributable to an individual's status characteristics. This issue has had practical significance in light of ongoing public debates about the impact of racial desegregation initiatives.

Evidence about the impact of social context on learning is mixed, but in any case the impact is not large. To the extent that the SES of a student body is consequential, this appears to result from the connection between SES and a positive academic climate in a school. Greater racial integration generally seems to promote black student achievement slightly, but the benefits are more pronounced for black students when they actually have classroom contact with white students rather than just attending a formerly integrated school.

More recent research suggests an important cautionary note about whether integration "works." Entwistle and Alexander (1992), for example, show that on a yearlong basis, in the early grades black students in integrated schools had better reading comprehension than did black students in segregated schools. However, the apparent advantage of integrated schools totally reflects the fact that black students at integrated schools improved more during the summer than did black students at segregated schools. During the school year black students did slightly better in segregated schools. This analysis exemplifies the increasing recognition that a simple conclusion about integration—works versus does not work—is inadequate.

Learning through the Year. As should be evident, a major issue in the sociology of education is separating the effects of the home from the effects of the school. The perplexing finding is that racial and class disparities in achievement in the early grades become substantially greater as students progress through school. Critics have seized on this finding to indict schools for discriminatory practices that exacerbate social inequality.

However, so-called summer learning research suggests a different interpretation (Alexander and Entwistle 1995; Gamoran 1995). Examining the same students' test scores at the beginning and ending of each of several school years, researchers have shown that (1) despite initial disparities, advantaged and disadvantaged groups have roughly similar gains in achievement during the school year but that (2) advantaged students continue to improve during the summer while disadvantaged students stagnate or decline. As the effects of this process accumulate over the years, initial disparities become ever larger. The important implication is that schools neither reduce nor add to the inequalities that are rooted in homes. Schools in effect passively reproduce existing inequalities.

Enhancing Performance. Although crude measures of school resources (e.g., teacher certification levels) appear at most to be weakly related to school achievement, a burgeoning and increasingly sophisticated line of research finds that effective schools can be identified. These schools are marked by strong leadership committed to academically focused goals and order, high academic demands, and frequent practice of academic skills. This research also directs attention to the benefits of an overall communal culture and classroom interactions that stress cooperative efforts between students and teachers (Lee and Croninger 1994). What appears critical is how resources are organizationally applied.

Macro-Level Effects. This article has focused on the experiences of individuals: how education affects life chances and how personal characteristics and school experiences affect learning. The unit of analysis, in other words, is the individual. Research in the field much less commonly takes the society as the unit of analysis: How do societal features shape the nature of the educational system? How do the features of this system affect other societal arrangements? An important example of macroanalysis is the generally limited impact of increasing educational access on equality of opportunity (see "Schooling and Life Chances," above). Perhaps the most studied macro-level topic is the relationship between educational expansion and economic growth.

If the individual economic benefits of education are clear, the impact of educational expansion on economic growth is less certain. The orthodox view in economics is that educational expansion promotes growth. This view follows from human capital theory: People with more schooling get higher pay because they are more productive, and if more people get more schooling, they will produce more and get paid more, with the aggregate effect being economic growth. Many sociologists are at least partially skeptical of this idea. Undoubtedly, more educated workers get paid more, but the positive (private) rate of return they enjoy reflects greater productivity only if it is assumed that the labor market is perfectly competitive and in equilibrium. This assumption is at least partly problematic given socially discriminatory employment practices, internal labor markets with seniority rules and restricted job mobility, professional and union restrictions of labor supply, and public sector employment with politically determined pay structures.

Allocation theory—which also is called the credentialing perspective—offers an alternative explanation of the link between education and economic rewards. In brief, employers assume that the more educated, as a group, are relatively desirable people to hire (for reasons that may or may not reflect their individual productive capacities); and in turn, how people are ranked in the educational hierarchy becomes linked to how they are ranked in the hierarchy of the existing job structure. Educational expansion, then, does not necessarily promote economic growth; it only affects who gets which of the already existing jobs. To the extent that credentialing processes are operative, it is impossible to infer aggregate effects on growth from individual-level data on income.

Given the ambiguous implications of individual income data, the best way to examine the issue is through aggregate, national-level studies of how education affects economic growth. The accumulated weight of this research undercuts claims about the large universal benefits of more education of all types. Benavot (1992), for example, establishes the following for a large sample of developed and poor countries in the period 1913– 1985: Throughout the period, the expansion of primary education promoted growth; the expansion of secondary education had more modest impact, and only during times of worldwide prosperity; and tertiary education tended to retard growth at all times. In the United States, moreover, tertiary enrollments have never stimulated growth (Walters and Rubinson, 1983).

However, even if more education is not a universal economic "fix," in certain circumstances particular types of education may stimulate growth in specific sectors. Reviewing single-country times series studies that use an aggregate production function model, Rubinson and Fuller (1992) conclude that education had the greatest beneficial impact when it created the kinds of skills that were suited to an economy's sectoral mix and technological demands. However, a good fit between the educational system and the economy is by no means certain because educational expansion and the actual educational content of schools are so often driven by political processes, not technological demands.

Even if the actual economic impact of education is often less than is commonly supposed, the widespread belief in the general modernizing benefits of education is central to an ideology that permeates the entire world. Indeed, in Meyer's institutionalist perspective (Meyer and Hannan 1979; Meyer 1977), the quest to appear modern has induced later-developing societies to mimic the educational practices of the early modernizers so that many school structures, rituals, and formal curricular contents are remarkably similar throughout the world. In turn, this institutionalized similarity means that on a global basis, certain types of knowledge become defined as relatively significant, the elite and mass positions become defined and legitimated by educational certification, and assumptions about a national culture rest on the existence of mass education. Nevertheless, if education is associated at the individual level with certain democratic values, educational expansion per se does not appear to contribute to the emergence of democratic regimes or state power.

THE REFORMIST PROJECT

Policy debates about education have often been contentious, fueled by larger ideological and political struggles. In conservative times, schools have been pressed to emphasize discipline and social and/or intellectual sorting; conversely, in more liberal times, issues of equality and inclusion have come to the fore. The apparent result is cyclical, pendulum-like swings in policy between, say, an emphasis on common core requirements and highly differentiated curricula.

While differences at the rhetorical level have sometimes been sharp, actual changes in practice in much of the twentieth century have been relatively minor. This reflects the institutionalization of the school, meaning that there is a widespread collective sense of what a "real" school is like (Tyack and Cuban 1995). This institutionalization rests on popular legitimization and the recurrent practices of school administrators and teachers. Concrete practices such as the division of knowledge into particular subject areas, the spatial organization of classrooms, and the separation of students into age-based grades are all part of the "real" school. Educational practices that depart from this pattern have had limited acceptance, for example, open classrooms in the 1970s. The lesson for current reformers is that policies that modify institutionalized practices, not fundamentally challenge them, are more likely to be successful and that the political support of in-the-school educators is critical for success.

Indeed, much policy-oriented research has had a mildly reformist bent, primarily concerned with making existing schools "work better." That has largely — and narrowly—meant producing students with higher scores on standardized tests in the basic academic subject areas. Critics have questioned both the validity of these tests and the desirability of evaluating school "success" in these limited terms alone. Proponents contend that scores on these tests have considerable predictive validity for later school and occupational performance and that their standardized results permit rigorous comparisons across groups and school settings.

The welter of policy-related studies is impossible to summarize here (and the distinction between sociological research and educational research is hardly sharp), but two general types of contributions from sociologists stand out. The first is essentially a debunking contribution: Sociologists have shown what does not work despite fervent beliefs to the contrary. The previously discussed Coleman Report is the most prominent example, undercutting the liberal faith of the 1960s that differences in school resources substantially account for racial and socioeconomic differences in academic achievement.

The second contribution is essentially methodological, alerting policymakers to the fact that many apparent school effects may largely or even totally reflect selection biases. That is, if groups of students are subject to different educational practices, are any differences in their performance attributable to the educational practices per se, or are different sorts of students subject to different practices, thus accounting for the association between practice and performance? In recent years, controversies about the efficacy of private and Catholic schools, related to larger debates about school choice plans, have centrally involved the issue of selection bias. In the most sophisticated study, Bryk et al. (1993) demonstrate net positive effects of Catholic schools on academic achievement and show that the gap in achievement between white and minority students is reduced in Catholic schools.

Even with the most sophisticated multilevel, multivariate statistical models, however, sociologists cannot make firm causal claims by analyzing survey data. However, by ruling out many potential sources of spuriousness, these analyses can suggest interventions that are likely to have a positive effect. True experiments, which involve the actual manipulation of the treatment and/or practice, are rare. In a state-sponsored experiment in Tennessee, starting in kindergarten, students were randomly assigned to varyingly sized classes (with and without a teacher's aide). The results showed that students, especially minority students, benefited academically from small classes (thirteen to seventeen students) and that the benefits persisted even when the students later moved to larger classes (Finn and Achilles 1990). Prior nonexperimental analyses had shown, across the range of class size in existing schools, that class size had very little or no effect.

Now that it is accepted that schools can make a difference in learning despite the great significance of family-based factors, the research agenda probably will focus on specifying the conditions in which particular school practices are most effective. This will involve analyzing inside-school practices as well as the links between families and schools and between schools and the workplace.

references

Alexander, Karl, and Doris Entwistle 1995 "Schools and Children at Risk." In Allan Booth and Judith Dunn, eds., Family-School Links: How Do They Affect Educational Outcomes? Hillsdale, N.J.: Erlbaum.

Aronowitz, Stanley, and Henry Giroux 1985 Education under Siege: The Conservative, Liberal and Radical Debate over Schooling. South Hadley, Mass.: Bergin and Garvey.

Baker, David 1994 "In Comparative Isolation: Why Comparative Research Has So Little Influence on American Sociology of Education." Research in Sociology of Education and Socialization 10:53–70.

Benavot, Aaron 1992 "Educational Expansion and Economic Growth in the Modern World, 1913–1985." In Bruce Fuller and Richard Rubinson, eds., The Political Construction of Education. New York: Praeger.

Blau, Peter, and Otis D. Duncan 1967 The American Occupational Structure. New York: Wiley.

Bourdieu, Pierre, and Jean-Claude Passeron 1977 Reproduction in Society, Culture, and Education. Beverly Hills, Calif.: Sage.

Bowles, Samuel, and Herbert Gintis 1976 Schooling in Capitalist America. New York: Basic Books.

Bryk, Anthony, Valerie Lee, and Peter Holland 1993 Catholic Schools and the Common Good. Cambridge, Mass.: Harvard University Press.

Carnoy, Martin, and Henry Levin 1985 Schooling and Work in the Democratic State. Stanford, Calif.: Stanford University Press.

Clark, Burton 1962 Educating the Expert Society. San Francisco: Chandler.

—— 1968. "The Study of Educational Systems." In David L. Sills, ed., International Encyclopedia of the Social Sciences. New York: Macmillan and Free Press.

Coleman, James 1988 "Social Capital and the Creation of Human Capital." American Journal of Sociology 94(Supplement):S95–S120.

—— Ernst Campbell, Carol Hobson, James McPartland, Alexander Mood, Frederick Weinfeld, and Robert York 1966 Equality of Educational Opportunity. Washington, D.C.: U.S. Government Printing Office.

Collins, Randall 1979 The Credential Society. New York: Academic Press.

Davies, Scott 1995 "Leaps of Faith: Shifting Currents in Critical Sociology of Education." American Journal of Sociology 100(6):1448–1478.

Dreeben, Robert 1968 On What Is Learned in School. Reading, Mass.: Addison-Wesley.

—— 1994 "The Sociology of Education: Its Development in the United States." Research in Sociology of Education and Socialization 10:7–52.

Durkheim, Émile 1977 The Evolution of Educational Thought, trans. Peter Collins. London: Routledge and Kegan Paul.

Entwisle, Doris, and Karl Alexander 1992 "Summer Setback: Race, Poverty, School Composition, and Mathematics Achievement in the First Two Years of School." American Sociological Review 59:446–460.

Farkas, George 1996 Human Capital or Cultural Capital? New York: Aldine de Gruyter.

Featherman, David, and Robert Hauser 1978 Opportunity and Change. New York: Academic Press.

Ferguson, Ronald 1998 "Teachers' Perceptions and Expectations and the Black-White Test Score Gap." In Christopher Jencks and Meredith Phillips, eds., The Black-White Test Score Gap. Washington, D.C.: Brookings Institution Press.

Finn, Jeremy and Charles Achilles 1990 "Answers and Questions about Class Size: A Statewide Experiment." American Educational Research Journal 27:557–577.

Fischer, Claude, Michael Hout, Martin Sanchez Jankowski, Samuel Lucas, Ann Swidler, and Kim Voss 1996 Inequality by Design: Cracking the Bell Curve Myth. Princeton, N.J.: Princeton University Press.

Gamoran, Adam 1992 "The Variable Effects of High School Tracking." American Sociological Review 57:812–828.

—— 1995 "Effects of Schooling on Children and Families." In Allan Booth and Judith Dunn, eds., Family-School Links: How Do They Affect Educational Outcomes. Hillsdale, N.J.: Erlbaum.

Hout, Michael 1998 "More Universalism, Less Structural Mobility: The American Occupational Structure in the 1980s." American Journal of Sociology 93:1358–1400.

Hunter, John 1986 "Cognitive Ability, Cognitive Aptitudes, Job Knowledge, and Job Performance." Journal of Vocational Behavior 29(3):340–362.

Hurn, Christopher 1993 The Limits and Possibilities of Schooling, 3rd ed. Boston: Allyn and Bacon.

Jencks, Christopher, and Meredith Phillips, eds. 1998 The Black-White Test Score Gap. Washington, D.C.: Brookings Institution Press.

—— et al. 1979 Who Gets Ahead? The Determinants of Economic Success in America. New York: Basic Books.

Kingston, Paul 1986 "Theory at Risk: Accounting for the Excellence Movement." Sociological Forum 1:632–656.

—— and James Clawson 1990 "Getting on the Fast Track: Recruitment at an Elite Business School." In Paul Kingston and Lionel Lewis, eds., The High Status Track: Studies of Elite Schools and Stratification. Albany, N.Y.: State University of New York Press.

Lee, Valerie, and Robert Croninger 1994 "The Relative Importance of Home and School for Middle-Grade Students." American Journal of Education 102:286–329.

Meyer, John 1977 "The Effects of Education as an Institution." American Journal of Sociology 83:55–77.

——, and Michael Hannan, eds. 1979 National and Political Change, 1950–1970. Chicago: University of Chicago Press.

Ogbu, John 1978 Minority Education and Caste: The American System in Cross-Cultural Perspective. New York: Academic Press.

Pallas, Aaron, Doris Entwisle, Karl Alexander, and M. Francis Stluka 1994 "Ability Group Effects: Instructional, Social or Institutional?" Sociology of Education 67:27–46.

Parsons, Talcott 1959 "The School Class as a Social System." Harvard Educational Review 29:297–308.

Phillips, Meredith, Jeanne Brooks-Gunn, Greg Duncan, Pamela Klebanov, and Jonathan Crane 1998 "Family Background, Parenting Practices, and the Black-White Test Score Gap." In Christopher Jencks and Meredith Phillips, eds., The Black-White Test Score Gap. Washington, D.C.: Brookings Institution Press.

Rosenthal, Robert, and Lenore Jacobson 1968 Pygmalion in the Classroom. New York: Holt, Rinehart, and Winston.

Rubinson, Richard, and Bruce Fuller 1992 "Specifying the Effects of Education on National Economic Growth" In B. Fuller and R. Rubinson, eds., The Political Construction of Education. New York: Praeger.

Schultz, Theodore 1961 "Investment in Human Capital." American Economic Review 51:1–17.

Shavit, Yossi and Hans-Peter Blossfield, eds., 1993 Persistent Inequality: Changing Educational Attainment in Thirteen Countries. Boulder, Colo.: Westview Press.

Sorokin, Pitirim 1927 Social Mobility. Glencoe, Ill.: Free Press.

Tyack, David, and Larry Cuban 1995 Tinkering Toward Utopia: A Century of Public School Reform. Cambridge, Mass.: Harvard University Press.

Waller, Willard 1932 The Sociology of Teaching. New York: Wiley.

Walters, Pamela, and Richard Rubinson 1983 "Educational Expansion and Economic Output in the United States." American Sociological Review 48:480–493.

Willis, Paul 1981 Learning to Labor. New York: Columbia University Press.

Paul W. Kingston