SPECIAL EDUCATION AND THE RISK OF BECOMING LESS EDUCATED

Given the rising importance accorded educational participation, performance, and certification, why do some students still leave school without qualifications? This question reflects that formal schooling shapes the life courses not only of the highly educated, as the educational expectations have risen considerably, but for all young adults. A paradoxical consequence of these higher expectations is a rapidly growing proportion of students who, not performing in school adequately or quickly enough, are referred to compensatory special education programs. As its diverse organizational forms developed over the twentieth century, special education offered assistance not only to children with recognized impairments, but increasingly also to those with a variety of newly defined ‘special educational needs (SEN),’ such as learning disabilities.

Sociological perspectives on special education have emphasized such dimensions as exclusion/inclusion, segregation/integration, in/equality, (de-)institutionalization, learning opportunities, stigma, risk, and certification. In particular, they contribute to our better understanding of the categorical boundaries of SEN and dis/ability among children and youth, the contrasting organizational forms of special and inclusive education, and the often unanticipated or unintended consequences of special education participation for educational opportunities and attainments. Life course research demonstrates the strategy of connecting early inequities with differential life chances. While disability has only recently begun to be analyzed utilizing such an approach (Powell 2003a; Priestley 2003), life course sociology has long emphasized modern nation-states’ increased institutionalization of individuals’ life courses as it exemplifies needs and risks addressed by authorized professionals in legitimate organizations (Mayer and Müller 1986: 234). Here, special education is examined as a paradigmatic case of such institutionalization. Youth in special education have been considered ‘at risk’ for well over a century (Richardson 2000). Although educational expansion facilitated affirmation of the goal of schooling for all children, growing diversity led to differentiation and standardization, attempts to resolve the tension between expanded access and organizational constraints: Rules of ‘access’ and of ‘passage’ governed the exemption of those deemed ‘ineducable’ or ‘disabled’ (Richardson 1999). Informed by disciplinary knowledge, classification systems and groupings or tracks based on ‘ability’ were organizational, pedagogic, and political responses to increasing heterogeneity. The institutionalization – regulative, but often residential as well – of these individuals’ life courses had been steady, until advocates of ‘normalization’ and ‘de-institutionalization’, challenged this status quo in recent decades (Braddock and Parish 2001). Despite massive general educational expansion and the disability movement's successful activism for increased access to (integrated or even inclusive) schooling, more than ever before, being ‘disabled’ is linked to being less educated than one's peers. Conversely, being less educated leads to an increased risk of becoming disabled, of experiencing poverty, and of suffering social exclusion (OECD 2003).

In contrast to abundant good intentions and compensatory investments, special education settings – authorized to offer different educational opportunities – seem to legitimately reduce individual access to opportunities to learn. These reduced opportunities, in combination with regulatory limits on certification, may reduce educational attainment. Individuals’ risk of low (or no) attainment increases in special education, with its students significantly overrepresented in the group of less-educated youth in both Germany and the United States (cf. Powell, 2007). Ironically, educational expansion has increased stigmatization of less-educated youth because they constitute the lowest educational category – that has become smaller and more socially selective over time – while ever more of their peers have earned certificates (Solga 2002: 164, 2005). Indeed, not only does education influence political and economic allocation, but also having credentials has become the ‘primary mechanism by which individuals are defined as full and legitimate societal members’ (Ramirez and Rubinson 1979: 80). Thus, education is seen as a global human right that states must provide and nearly everyone supports the norm of universal educational access and equal opportunity, even given interindividual variations in ability (Meyer 2001). Responding to these principles, states and non-governmental organizations around the world have committed themselves to ‘education for all’ – and to inclusive education (e.g., UNESCO 1994).

Before comparing these nations’ varying accomplishments in achieving their goal of inclusive education, we define ‘special educational needs’ as referring to institutionalized cultural value judgments about behavior, intellectual functioning, and health that result in particular human differences being recognized as deserving of support or professional services. In her pioneering analyses, Tomlinson (1981, 1982) discussed these complex and varied accounts of SEN as behavioral, bodily, functional, intuitive, linguistic, organizational, psychological, social, statistical, statutory, and tautological (as in: ‘a child with special educational needs has special educational needs’). Ambivalent and often contentious, classification as having SEN requires extensive mediation between its many positive and negative consequences: Provision of additional resources and rights, but also prevalent stigmatization, even institutionalized discrimination, frequently lasting throughout the life course (Powell 2003a, b). Analyzing SEN or student dis/abilities requires analytic attention to the relationships between individuals embedded in social situations, but also to cultural contexts, disciplinary perspectives, and translations of concepts into empirical measures that guide classification processes. Educational administrations distinguish student disabilities and regulate access to special educational services and settings according to culturally specific social norms and professional practices: ‘Far from being “scientific facts” based on objective, universally understood definitions of difference, the categories and labels assigned in different societies are contingent, temporary, and subjective’ (Barton and Armstrong 2001: 696). International data presented here provides further evidence to support that claim.

Applied by school gatekeepers – such as teachers, administrators, and school psychologists – at individual level in response to particular behaviors, SEN categories imply deviance from social norms.¹ Official classification furnishes students with specific rights, but simultaneously provides bureaucratic legitimacy and accountability needed to justify compensatory provision of additional expenditures and specialized services. This process is charged with resolving the ‘distributive dilemma of disability’ (Stone 1984) in particular times and places. Student dis/ability and SEN categories are continuously revised, yet the processes of classification in schools, once implemented, resist change – as do the organizations established to serve classified students. Examining current special education classification and school segregation and integration rates emphasizes the contingent development of (special) educational classification systems and school structures.

Given the (challenged) dominance of medical models of disability and the clinical professions that define disability and SEN mainly in terms of individual deficits, we might expect that rates of SEN would be roughly similar across advanced industrialized countries. Yet across Europe, the rates of all children classified and receiving services vary considerably, from less than one percent to nearly a fifth of all students (EADSNE 2003). These programs serve a highly heterogeneous group of children with social, ethnic, linguistic, physical, and intellectual disadvantages. While the group participating in special education includes children with similar disadvantages and disabilities in all countries, we find large differences not only in the size of the group and its demographics, but also in organizational structures (OECD 2004). Across Europe, a remarkable array of organizations provides special education in special schools or classes as well as more mainstream settings (Eurydice 2005: 129). Furthermore, every European country has or is implementing reforms toward more school integration or inclusive education, but at very different paces (EADSNE 1998). Among OECD countries (as within the European Union), not only do the rates of all children classified as having SEN vary considerably by nation and region, but the proportion of those who are integrated in general school settings ranges from almost none to almost all (Eurydice 2000Eurydice 2005; OECD 1999, 2004).

This article explores variance in population sizes, learning opportunities provided, and the resulting inequalities in educational qualification of this diverse group of students. Knowing which groups of students are most likely to participate in expanding special education systems demonstrates which children and youth in these societies are most likely to grow up less educated. To investigate these issues empirically, we analyze (1) students’ classification into special education, (2) their allocation to schooling structures that provide or constrain opportunities (cf. Sørensen 1996), along a continuum from segregation and separation to integration and inclusion, and (3) their resulting educational attainments (Germany, US). While individual characteristics are crucial, the cross-national differences in constructions of and organizational responses to SEN investigated here emphasize the need for institutional explanations in place of those that focus solely or mainly on individual deficits. Embedded in a broader European comparison, the German and American findings presented here accentuate national patterns of association between classification rates that measure selection processes, learning opportunity structures, and the resulting educational attainments.

While some European countries utilize only one or two SEN categories and others more than a dozen, most nations have implemented between six and ten such categories (Eurydice 2002: B-12), depending on extant official disability classifications, assessment procedures, finance regulations, allocated resources, and educational system differentiation. Striking differences among the OECD countries exist. The following 22 categories are currently used to identify students with disabilities, difficulties, and disadvantages: Partially sighted, blind, partially hearing, deaf, emotional and behavioral difficulties, severe/moderate/light learning difficulties, physical disabilities, combinatorial/multiple disabilities, learning disabilities, speech and language disabilities, hospital, autism, gifted and talented, remedial, second language/mother-tongue-teaching, traveling children, disadvantaged, aboriginal, young offenders, and other (OECD 2004: Annex 2). Recent cross-national studies of inclusive and special education and social exclusion utilize just three broad groups of students who receive ‘additional resources to access the curriculum’ (Evans et al. 2002; OECD 2000a, 2004): (A) children with impairments; (B) children with learning difficulties; and (C) children with disadvantages. That typology emphasizes the main groups served by special education programs and policies. However, significant differences between countries are found not only when including disadvantaged students or those with learning disabilities, but also in the seemingly most ‘objective’ categories such as visual or hearing impairments (Powell, forthcoming).

This overview presents the (1) classification rate (proportion of all students classified into special education), (2) segregation rate (proportion of all students that are segregated, defined as attending separate facilities or nearly fulltime separate classes), (3) segregation index (proportion of students with SEN segregated), and (4) type of (special) education system. Despite the quantification and comparison of every aspect of schooling, comparable data on educational outcomes is almost completely lacking for students with SEN, thus ‘future data gathering exercises will focus on collecting outcome data’ (OECD 2004: 131).

The considerable range in proportion of students with SEN of all compulsory school age students in Europe stretches from Greece and Italy with less than 1.5 percent to Finland, where almost 18 percent of all schoolchildren receive some form of special education services (EADSNE 1998, 2003). Germany's 5 percent and the US's 12 percent fall in-between (Table 1: Column 1). What these additional resources consist of and where they are provided requires more detailed analyses than can be presented here, but these rates reflect major differences in national policies and the variable institutionalization of special education, including organizational differentiation, service provision and curricular models, teacher training, and finance.

Across Europe, a multitude of special education arrangements exists, with the proportion of students in separate schools or classes ranging from less than 1 percent to over 6 percent of all students (Table 1: Column 2). By contrast, in the US, while the proportion of students not attending a general school full-time is very low – 0.4 percent of all students – adding the relatively large group of ‘separated’ students who spend more than 60 percent of their school day in separate classrooms places the US in the mid-range, at 2.1 percent (Powell, forthcoming). Such large cross-national differences – often matched or exceeded by regional variance within countries – demonstrate that the proportion of children classified and their learning opportunities depend not solely on individual characteristics, but largely on (special) education systems’ institutional development and classification systems in use.²

Unitary educational systems (e.g., Iceland, Norway), aim for ‘full inclusion’, educating nearly all students in general classrooms. Some OECD countries, including the US, maintain a continuum of settings from inclusive classrooms to segregated special schools, while Germany is gradually moving its highly differentiated, segmented educational system (with mainly special schools) toward such a ‘continuum’ or multi-track model. Although last century many asylums and special schools were closed in favor of students sharing in the mainstream of school life, segregating or separating students with SEN remains part of policies and praxis in most countries, despite international charters and national laws that aim to increase school integration and/or inclusive education. Among the countries presented here, Germany (along with Switzerland and Belgium) has the highest percentage of all students of compulsory school age that are schooled in segregated settings (all above 4 percent), while Greece, Italy, Spain, Norway, Portugal, and Iceland have the lowest rates of segregation (all below 1 percent). However, in many Southern European countries, these rates coincide with fewer support provisions of any kind. Even when most students do attend a general school, the goal of individualized support for accessing the curriculum is not always met. Throughout Europe, while inclusion of students with SEN in compulsory general education is agreed upon as desirable, such models have yet to be universally accepted as appropriate for all children and youth with SEN, especially due to issues of quality (Eurydice 2005: 129ff.) and institutional and organizational barriers that hinder provision of supports flexibly within general school settings (cf. Skrtic 1991).

Based on their educational system structures, these countries can be grouped into a tripartite typology of dual, multiple, and unitary: (1) two-track with parallel development of general schools and legally and organizationally separate special schools; (2) multi-track offering a continuum of settings and services from special schools to schooling partially in separate schools or classrooms to full-time participation in general classrooms; and (3) one-track with a goal of ‘inclusion’ for all children that educate almost all students in general classrooms (EADSNE 1998: 178ff., 2003). In many of these countries, debate centers on legislative advances prioritizing an increase in institutional flexibility (movement toward a continuum of settings and services), growing awareness of funding system consequences (e.g., incentives to classify, segregate or separate), and the importance of parental choice. Even such a cursory cross-national comparison emphasizes the importance of historical and comparative research on the institutionalization of educational systems. Therefore, in the following, we concentrate on the German–American comparison. Whereas in the US, over 95 percent of students receiving special education support attend general schools, in Germany, less than 15 percent do. The next section explores institutional factors that help explain such vast differences.

Having begun 200 years ago with schools for blind and deaf children, the institutions established to provide special education became less and less similar over the twentieth century as they developed isomorphically to national general educational systems (cf. Powell, forthcoming). The presented German–American comparison, while exploratory,³ offers insights, especially since these nations have continuously borrowed each other's educational ideas and concepts (cf. Drewek 2002). Moreover, they unite the unusual mixture of federal democracies with decentralized control over education content and financing with more centralized rules for special, often unequal, groups of students, such as disabled, disadvantaged, and immigrant children (Meyer 1992: 236). While these special education institutions were originally quite similarly exclusionary and reform efforts strive to implement inclusive education, the German and American systems have diverged considerably. Comparing Germany's highly-differentiated special school system and the US's burgeoning lowest comprehensive school track provides a test of educational expansion's impact on the distribution of educational opportunities not as is usually done – from the top – but from the bottom.⁴

Perhaps most fundamentally for the questions raised here, similar ideologies, interests, and institutions relating to dis/ability and ab/normality resulted in the exclusion – in both countries – of a majority of children with impairments from schooling until after WWII, when their citizenship rights were affirmed (cf. Barrett and Kurzman 2004 on the global spread of personhood following eugenic tragedies). The legacies of that legitimated exclusion, amounting to selection based on disability classification at the school gate, are evident in national and Länder/state educational policies and organizational responses to it. These institutions determine which children in Germany and the United States will become disabled and when, who will be integrated and where, and what learning opportunities they will benefit from while in school.

In absolute numbers and proportion, Germany and the US witnessed dramatic growth in their special education populations over the twentieth century (cf. Powell, forthcoming). Official German time-series only include data on ‘Integrationsschüler’, those students with SEN who attend general schools, since schoolyear 1999 (Krappmann et al. 2003). Thus, only recently can integration and inclusion developments across Gemany's Länder be monitored and compared. The overall classification rate has increased to 5.6 percent of the student population in general schools, representing nearly half a million children and youth (KMK 2005: ix). In the US, growth has been continuously upward, at a much faster rate: By 2001, more than 5.8 million students ages 6–21 received special education services – around 12 percent of public school enrollment (US DoED 2005: 21).

Rapid developments in dis/ability concepts, definitions, and labels exhibit the shifting boundaries between special and general education students. In 1994, Germany replaced organizational-administrative categories of ‘the need to attend a certain special school type’ (Sonderschulbedürftigkeit) with educational support categories. By contrast, American individual categories of impairments and disabilities, differentiated over time, have continuously relied on psychometric diagnosis. Yet such changes in categorical labels – despite new procedures for identification, referral, assessment, diagnosis, and classification as well as evolving understandings of disability and the tools to measure its complex interrelation of personal, social, and environmental factors – have not transformed the institutional settings in which students so classified spend their schooldays. Thus, the organizational source of (special) education stigmatization continues. Children may be differentiated through application of categories and labels, but the result is allocation to learning opportunity structures.

The groups participating in special education exhibit considerable heterogeneity. Demographic dimensions of age, gender, poverty, and race or ethnicity vary by disability category and region, as individual and environmental characteristics interact. Early intervention programs have become increasingly important as research shows how vital are early learning experiences and preventive measures for at-risk children. School transitions, horizontally or vertically in stratified school systems, are especially significant. Whereas in Germany, students’ risk of transfer to a special school rises steadily, with the rate peaking at age 14 and falling off sharply thereafter, in the US, nine- and ten-year-old children are most likely to be classified as having SEN (OECD 1999). The remaining non-classified students’ probability of receiving an individualized education program (IEP)⁵ declines gradually thereafter. Reflecting gender patterns in many OECD countries, nearly two-thirds of special school students in Germany are boys, as are US special education students (OECD 1999). Boys in both countries seem to be increasingly disadvantaged given their considerable and growing overrepresentation in special education.

Poverty not only adversely affects health; it ‘is the most consistently associated indicator of poor academic achievement and school failure’ (Land and Legters 2002: 4ff.). Despite much higher poverty rates among children in the US than in Germany (double, by most estimates), a large proportion of special education students in both countries are children from low-income families. In the US, compensatory education was conceived to reduce the adverse effects of childhood poverty, minority status, and other characteristics on learning, but special education provides far more resources targeted to individual students. There, the official definition of ‘learning disability’ as measuring a discrepancy between school grades and measured IQ expressly rejects classification of children who are having difficulty learning due to material disadvantages. In contrast, the German category mainly reflects the effects of low SES, including interpretations of culturally specific behavior and linguistic disadvantages. These crucial qualitative differences belie the quantitative similarity that roughly half of all students with SEN are classified as ‘learning disabled’.

Racial and ethnic disproportionality has attracted considerable criticism to special education and to segregated and separate settings in particular (Powell and Wagner 2002). For over 30 years in the US, black students have had significantly higher probabilities of classification in categories such as ‘mental retardation’ and ‘emotional disturbance,’ but not in ‘specific learning disabilities’. Recent research suggests that while ‘toxic social conditions’ may lead to higher impairment rates among children of some racial and ethnic groups, overrepresentation also results from inappropriate interpretation of cultural differences and biased classification processes and instruments (Oswald et al.2002). And ‘residential, social and school segregation is so profound, especially for blacks, that it often overrides middle-class advantages that some minority children may have’ (Fischer et al.1996: 196). While socio-demographic factors are clearly associated with classification rates and with disproportionality among racial and ethnic groups, these do not always point in the same direction (e.g., Asians/Pacific Islanders in the US and Scandinavians in Germany are clearly underrepresented in special education).

In sum, special education represents children and youth at the nexus of multiple social differences and ascriptive attributes, including disability, gender, and ethnicity. But the effects of social, economic, and cultural disadvantages are evidently hardly separable from impairments, disabilities, and/or SEN that are identified during children's school careers. The wide diffusion of special education programs in nearly every American school, combined with higher poverty rates and greater ethnic diversity, correlate with a proportion of students in American special education programs more than two times higher than the German rate. The organizational proximity, availability of special education, and emphasis on early intervention also leads to American students being identified, referred, and assessed as having a disability earlier than students in Germany.

Despite a growing diversity of organizational forms in some German Länder, there is as yet no significant ‘continuum’ as in the US, but rather the institutionally constituted either-or of special or general school. Despite the existence of comprehensive schools (Gesamtschulen) in some Länder, their institutionalization did not successfully challenge the vertical differentiation of secondary schooling overall (Leschinsky and Mayer 1999). Ten types of segregated special schools were established in the 1960s, effectively blocking decades of integration attempts. American comprehensive schools are outwardly democratic and egalitarian, but many schools continue to stratify within via tracking. A key factor in the systems’ different ‘classification thresholds’ (Powell 2003b) is that American schools allow flexibility in curricular planning and permeability in allocation to courses, while the German structure does not.

Proponents and opponents of integration have struggled since the very beginning of German (special) education. Since the 1970s, criticism of Germany's highly differentiated special school system has led to calls for Integrationspädagogik. Policy elites, nongovernmental organizations, and interest groups at local, regional, and national levels shifted the debate to an inclusive education model that accepts and values heterogeneity, similar to Anglo-American concepts. In response, a multitude of organizational forms have steadily developed, from integrated classes and individual mainstreaming to ambulant services and resource centers, among a host of other (quantitatively insignificant) concepts evolving regionally (Sander 1998). These school reforms attempt to meet individual needs without segregation; however, they differ considerably in the amount of peer contact they provide and in their curricular goals. Diverse forms of ‘joint instruction’ (gemeinsamer Unterricht), most of which do enable classified children to take part in general classes all day, remain marginal, reaching only an estimated one-tenth of all special education students.⁶ For the vast majority, having SEN still means being segregated (see Table 2).

The prevailing structure of Germany's provision of special education is often criticized for failing students because (1) many children in general schools do not receive the support they need and (2) most children attending special schools are sent only after a ‘wait-to-fail’ period of retention of one to two years in which their learning needs have often not been fully addressed. Teachers seem highly reluctant to send their students to special schools, preferring retention (historically, a legal prerequisite), at which point for many children it will already be too late to make up their past lost learning opportunities. Conversely, students that are sent become part of a negatively selected group that suffers all the more stigma and low expectations. They attend schools that serve only students in the same category – often with similar learning, language or behavioral problems. Segregation not only magnifies the challenges faced by special schoolteachers (despite the much smaller classes), but also takes away positive peer role models and isolates students from school and neighborhood friends. Once transferred to a special school, youth are likely to remain in such ‘special’ institutions. With the exception of those in the support category ‘speech,’ less than 5 percent of special school students return to general schools (Preuss-Lausitz 2001: 211). Thus, special school attendance is generally seen as ‘without perspective’.

While several Länder have established cooperative or ‘external’ classes that reflect the American setting of ‘resource rooms’ in which special education students spend most of their day separated from their peers within general schools, the major difference remains the default setting: Inter-school segregation in Germany and intra-school separation in the US. In the former, retention (and delayed school entry) is a far more significant mechanism, affecting a large minority of schoolchildren, than special education; whereas the two are often utilized in conjunction in the US (Entwisle et al.1997).

In terms of permeability and opportunities to access general curricula, the American system is far more flexible than the German. In the US, nearly all students with an IEP attend their local, not necessarily neighborhood, general school. Since the mid-1970s, ‘mainstreaming’ increased access to general schools. More recently, American states’ noncompliance with the laws mandating inclusion have led to repeated calls for ‘restructuring,’ with a goal of reducing separation and maximizing the time students with SEN spend in the general classroom. Although conceived of in a variety of ways, inclusive education envisions teachers replacing separate schools and classrooms with diverse general classrooms in reformed schools. By 2001, 45 percent of all students with an IEP spent fourth-fifth of their schoolday in the general classroom (inclusion), 30 percent spent more than two-fifths there (integration), while 20 percent spent even less time among their peers without SEN (separation) (DoED 2005). Given that continuum of educational settings, the structural conditions for return to general education would seem to be very good (cf. Lucas 1999 on ‘track mobility’). Not only do almost all students with an IEP remain within the general school building, but these IEPs are revised annually. Students may receive a plethora of special education services, for differing durations, and in an array of settings. Findings on special education students’ educational attainments indicate the effects of such disparate organizational forms and of flexible, usually partial separation in the US, but continued full-time segregation in Germany.

The group of less-educated individuals in both countries has contracted considerably, such that by 1998, the proportion of the population aged 25–64 who had completed upper secondary education in both countries exceeded 80 percent (OECD 2000b: 26). Nevertheless, these systems continue to produce graduates with an ‘absolute wealth of competencies’ and school-leavers without any certificates – the educationally impoverished (Allmendinger and Leibfried 2002: 304). As the lowest qualified secondary school certificate (Hauptschulabschluss) in Germany has become ubiquitous, the group of youth without such certification represents a residual category of shrinking proportion, but of increasing societal concern. Indeed, even higher general education certificates are increasingly being taken for granted, regulating access to further education or even most low-paying jobs. Solga (2002, 2005) shows how educational expansion paradoxically led to the increasing exclusion of less-educated youth from vocational training and from many occupations.

German special school-leavers make up two-fifths of youth who do not attain the Hauptschulabschluss. Four-fifths of youth leaving Germany's segregated special schools do not do so (KMK 2002); they are likely to remain in a holding pattern (‘cooling-out’) in state-sponsored, school-based vocational training measures (Wagner 2005: 167ff.). In 2000, only two percent of around 45,000 special school-leavers attained an intermediate school-leaving certificate, and a few dozen the Abitur, necessary for university entrance (KMK 2002).

By contrast, in the US, nearly half of those students with an IEP ages 14 and older graduated with a regular high school diploma (US DoED 2005: 17, 70), the credential necessary, but not sufficient, for most entry-level jobs. As the uniform standard for access to post-secondary schooling, special education students also strive to attain it. However, American special education students were more likely to drop out than their peers in general education, less likely to complete a general equivalency diploma (GED) later on, and less likely to participate in postsecondary education or have paid employment – and if so, these jobs were more likely to be low-status and/or part-time (Marder and D'Amico 1992: 47ff.; NLTS 2003).

Many German Länder and US states simply do not offer special education students the organizational or curricular conditions necessary to earn any type of qualified certification. Despite the traditional importance of Germany's dual system of general and vocational training, this combination of school and practical training is almost never offered to special school-leavers. Ironically, they most often remain in or return to school-like settings, attending special programs sponsored by the federal employment office. Since few school-leavers without certificates (Hauptschule level or higher) are able to secure training opportunities, they face tremendous difficulties in transitioning to vocational training and work (Hillmert and Powell 2005). From their transfer to a segregated school onwards, these youth will populate a parallel world of special programs, without much hope of future employment.

Although vocational training plays a lesser role in the US than it does in Germany, longitudinal American data indicate that the more vocational training disabled youth have received in high school, the more likely they were to succeed in finding paid work (Wagner et al.1993). If vocational training can simultaneously ‘provide a safety net and be a mechanism of social exclusion’ (Shavit and Müller 2000: 449), it remains to be investigated how wide a net vocational training can truly cast. Comparing the German and American (special) educational systems manifests the importance of opportunity structures, not only for learning, but also of attaining credentials necessary for further study or even to access basic vocational training needed to secure employment in low-status occupations. Despite the access they have won to educational systems, school-leavers from special education represent a growing proportion of America's ‘working poor’ and Germany's long-term unemployed and social assistance receivers (Daly 1997: 115).

Remarkable differences across Europe and the OECD countries were found in (1) SEN classification rates and (2) in learning opportunities, provided in settings ranging from special schools to inclusive classrooms. Cross-national studies of disparities in achievement concur that even among the wealthiest nations, some educational systems protect disadvantaged students from experiencing inequality far more than others (Baker et al.2005: 84). While the impact of such national differences is increasingly recognized, results from the OECD's Program of International Student Assessment (PISA) and the US's National Assessment of Educational Progress (NAEP) demonstrate that nations, states, and localities not only have highly variable rates of providing additional resources to access the curriculum, but also of including students with SEN in assessments. Such ‘exclusions’ from data collection and analysis problematize aggregate comparisons of student performance, though these are increasingly required by law (e.g., the US's No Child Left Behind Act) and directly influence students, teachers, and schools. Pressures to perform provide multiple incentives for increased SEN classification and thwart attempts to account for all students’ learning program. Special education's growth – and its students’ disparate participation rates in large-scale assessments – indicate its increasing but not uniform authority.

To evaluate the impact of the recent, at times contradictory, international forces of achieving high standards of educational performance and inclusive education (and ‘education for all’) – as well as of global educational expansion more broadly – we compared the two-track German educational system of general and special schools with the American multi-track system, a continuum ranging from special schools to inclusive classrooms. In broader European comparison, these two cases fall in the mid-range between the most segregating and most inclusive educational systems. The findings accentuate national differences in SEN definitions, inclusion ideologies, and institutionalized learning opportunity structures. In both countries, attempts to realize the goal of education for all may have been achieved, while the aim of inclusive education remains significantly challenged, especially for secondary education. Neither country has fully embraced the unitary model of educating all students together in general classrooms, but the US has moved much more quickly in that direction.

Special education organizations in both societies have served a population of students continuously changing in size and composition, but representing especially poor boys, children belonging to racial, ethnic, migrant or linguistic minority groups, and increasingly integrated children with perceived impairments. It is these diverse student bodies that most challenge rationalized, standardized organizational structures of German and American (special) educational systems. Which disadvantages should be compensated, how much, in which school settings, and what level of school certification should result remains a matter of continued debate. Ambivalence towards special education highlights the tension between equality of opportunity and merit measured in school performance. Resistance to reform and restructuring of the existing special education systems necessary to successfully realize inclusive education relies on the legitimated institutional logic of each national education system – German inter-school segregation versus America intra-school separation. These reflect societal values and educational ideologies as they frame the interests that successfully fought for special education's diffusion and differentiation.

Special education continues to be organized in a tremendous variety of ways, despite movement towards more integration and inclusion. Holding national educational systems accountable for all students’ educational performance and adequate certification requires recognition of persistent segregation or separation and stigmatization of children and youth with SEN, who more often than not are already among the disadvantaged. However, the resulting additional disadvantages in learning opportunities, educational attainments, and life chances are something that few societies, despite egalitarian rhetoric, have eagerly confronted. German debates about quality in (special) education and about the organizational settings in which students with SEN should be educated mirror American discourses (Benkmann 1994). Yet whereas Americans view special education as a tool to realize equality of educational opportunity, Germany's educational policies maintain class stratification and are overshadowed by social policies in state efforts to compensate for disadvantages. Attempts to address and reduce the overrepresentation of male students, racial and ethnic minorities, and poor children and youth in special education programs have largely failed, and these remain key groups at risk of becoming less educated.

There are significant differences between and within these societies in the ways in which groups of disabled students are socially defined, sorted into educational programs, and to which degree they are integrated into general school systems or inclusive classrooms. Considerable inequalities in learning opportunities persist. Inclusive education remains a widely-held goal, as it promises to more fully utilize the diversity of interests and abilities found among all groups of children to develop each individual's intellectual and social competencies. However, inertia throughout educational systems has hindered special education reforms. Throughout the world, the value of schooling has continuously increased. Even youth with SEN, whose training and employment opportunities are seriously limited, partake in ever-longer school careers on the path to certified adult citizenship, if not always salable competencies. Thus, if sociologists wish to understand national disparities in the risk of becoming ‘less educated’, increasingly we will have to ask who ‘has SEN’ as well as why and how special and inclusive education systems provide these children and youth with such differential opportunities to learn.

For helpful comments on earlier versions of this article, I thank Bernhard Ebbinghaus, John Goldthorpe, John Meyer, Lisa Pfahl, Alan Sadovnik, Heike Solga, and the participants of the ECSR conference ‘Causes and Consequences of Low Education in Contemporary Europe’ (2004), the Harvard CES Berlin Dialogue ‘Social Determinism vs. Equal Opportunity’ (2004), and the ASA Centennial Meeting Special Session on the ‘Institutionalization of Human Intelligence’ (2005) as well as the anonymous reviewers.

Justin J. W. Powell teaches Sociology at the University of Göttingen, Germany. From 2000 to 2005, he was a research fellow of the Max Planck Institute for Human Development in Berlin. His book Barriers to Inclusion: Special Education in the United States and Germany is forthcoming from Paradigm.