Sociology Working Paper Paper Number 2005--02

An assessment of the “oppositional culture” explanation for ethnic differences in educational attainment in Britain Catherine Rothon Department of Sociology University of Oxford

Department of Sociology University of Oxford Manor Road Oxford, OX1 3UQ www.sociology.ox.ac.uk/swp.html

1

Abstract This paper examines the utility of one of the most influential American theories applied to differential educational attainment by race: the “oppositional culture” explanation, exemplified by the work of John Ogbu. This thesis argues that certain ethnic minority groups recognise the value of education in an abstract sense. However, as they are made aware of barriers to opportunities in further education and the labour market, they develop a bleak view of their future prospects. This translates into a rejection of competition for scholastic rewards; students adopt antiachievement values in order to reduce the discrepancy between work put in and rewards received on the labour market. These anti-achievement norms are reflected in the behaviour of children at school. This paper examines the educational attainments of blacks, Indians, Pakistanis and Bangladeshis in Britain in the context of oppositional culture theory. Using empirical evidence from a nationally representative survey, limited support is found for Ogbu’s framework. In terms of the school performance of blacks, the direction of attitudes and behaviour supports his thesis, but the extent of them does not. The responses of the south Asian groups cast further doubt on the utility of the thesis. Ogbu predicts that Indians will exhibit significantly different attitudinal and behavioural patterns to Pakistanis and Bangladeshis; this is not supported by the data.

Attainment at age 16 Younger generations of ethnic minorities have generally been improving their attainment rates at GCSE with some groups (notably Chinese and Indians) outperforming Whites. Nevertheless, gaps still remain between the overall outcomes at GCSE for Whites and Bangladeshis, Pakistanis and Black Caribbeans. Whilst showing absolute improvement, some groups appear not to be closing the gap between their attainment levels and those of Whites at all (Performance and Innovation Unit 2001). One of the most enduring findings of social science research is the relationship between family background and academic achievement, yet most research has not revealed clearly how a person’s ethnicity and class influence his or her achievement at school. Such theories are limited in the sense that they have failed to explain why blacks of the same socio-economic background to whites still perform at a comparatively lower level. This paper examines the utility of one of the most influential American theories applied to differential educational attainment by ethnicity, the oppositional culture explanation, in explaining patterns of achievement in Britain.

Theoretical background Traditional cultural theories of student attainment In his first published work, Ogbu (1974) reacted to what he perceived to be weaknesses in traditional theories of black academic achievement. The first explanation he critiqued has been termed the “culturally deprived child” thesis (Riessman 1962). This theory views the school as a victim forced to cope with the “resistance to learning” that poor and ethnic minority children exhibit. This resistance is created within the home; such children are “retarded” in their social and psychological development since the stimulation for normal development is not present (Deutsch 1967, Hunt 1967, 1969). A modified version of the theory argues that the children of subordinate groups fail because the cultures in which they grow up are different from that of the white middle class and their 2

values, attitudes and learning skills are at odds with those in mainstream schools (Inkeles 1966, Riessman 1962). Ogbu (1974, 1978, 1997b, 1999, 2003) argued that such theories do not account for the high performance of some ethnic minority groups; Arabs, Chinese, Filipinos and Japanese, for example, perform well in American schools. He stated that “remedial” programs, set up to mitigate the effects of “inferior” home environments, had been largely unsuccessful (Ogbu 1974, 1986). In addition he pointed out that in Asia and Africa students with very different cultures are able to achieve in Western-style schools. Finally, he stated that the aims of whites and other ethnic groups are not that different: good education, good jobs, high wages and better living conditions. Ogbu (1974) termed the second group of explanations to which he objected “blaming the schools”. These theories question whether the traditional classroom is the most appropriate setting for effective learning, and argue that negative teacher expectations have an adverse effect on the development of poor and ethnic minority students (Rosenthal and Jacobson 1967, 1968). A number of studies on ethnicity and education have shown that teachers often expect black children to fail and label ethnic minority pupils as less intelligent, and it is argued that such perceptions may negatively affect pupils’ school careers (Coard 1971, Aragon 1973, Pumphrey and Verna 1990). Critics of traditional educational methods sought to replace this approach with more informal education. Ogbu stated that this showed a lack of understanding of what minorities wanted or needed, arguing that subordinate minorities believed that formal educational could help them “upgrade” themselves. The final group of theories to which Ogbu (1974, 1986) objected was genetic deficiency theses based on IQ differences. Jensen (1969) had reported that psychologists had reached a general consensus on blacks possessing lower IQs than whites and felt that it was not unreasonable, in view of the fact that intelligence variation has a large genetic component, to hypothesize that genetic factors may play a part in this picture (Jensen 1969: 82, see also Crow 1969).

Jensen calculated that about 80 per cent of IQ differences could be explained by genetic variation, and that the remainder could be attributed to environmental differences. Herrnstein and Murray (1994) argued that blacks do less well because of their lower IQ; this was explained in terms of poor genetic endowment. Herrnstein (1971) had argued that those in the higher echelons of society had a higher level of intelligence than those below them. From this it followed that since blacks were at the lower end of the social spectrum, they were inherently less intelligent. Although Ogbu (1974) accepted that blacks do poorly on IQ tests and on scholastic tests he rejected that idea that there was a causal relationship between the two. Rather, he argued that black children do not take the tests seriously so that the test scores do not reflect any real ability or potential. He pointed out that through miscegenation a number of blacks have inherited white genes, and that there was no evidence to suggest that those with more white genes were significantly more intelligent than those blacks with fewer, or those blacks without. In response to Herrnstein, Ogbu (1974, 1978) asserted that the United States is only one of a number of plural societies with a domination-subordination framework. Others include Great Britain (whites versus Afro-Caribbeans and Pakistanis), British Guyana (blacks versus East Indians) and Fiji (Indians versus native Fijians). If Herrnstein’s theory were correct, one would expect those in the dominant group to consistently exhibit the highest levels of intelligence. The hypothesis could therefore be rewritten to state that in any stratified society, intelligence was determined by social position and that when a subordinate group replaces the dominant group its collective intelligence rises; this seems improbable. Finally, Ogbu (1974) expressed a general suspicion of so-called “scientific 3

proofs”. In the early part of the 20th century the high rate of illiteracy among Eastern and Southern European immigrants was viewed as a function of “inherent racial tendencies”.1 By the time at which Ogbu was writing they were classified alongside the dominant white majority and by “scientific proof” said to have a higher level of intelligence than blacks. “Resistance Models” In response to the inadequacies of these theories, Ogbu (1974) developed a new framework within which to examine the educational attainments of ethnic minority students, most specifically black Americans. In relation to racial stratification in the United States and its impact on academic “success”, Ogbu argued that performance should not be viewed in universal terms but rather placed in the cultural and historical context within which young people develop. He pointed out that black Americans that work as hard as whites receive fewer rewards. In order to reduce the dissonance between effort and reward, he argued that blacks had chosen not to work as hard as whites. One consequence of this lack of effort has been to reinforce white stereotypes about the intellectual and cultural inferiority of blacks. A similar argument was put forward by Wilson (1987). His “resigned adaptation” explanation argued that, in response to high levels of unemployment and discrimination, non-whites have become relatively more tolerant of joblessness. The fact that many ethnic minorities live in communities where unemployment is prevalent restricts their aspirations and decreases their desire for work. This is reinforced by a feeling that obtaining work will not necessarily result in any financial gain compared to benefit or other non-market income. Such norms, it was argued, can become transmitted within and across generations and the incentive to acquire education in order to offset discrimination becomes less and less important. These problems are compounded by the lack of economic opportunities in the inner city areas where less affluent ethnic minority groups are concentrated. This has been termed “spatial mismatch” theory (Holzer 1991). Ogbu used the “resistance model” to further develop his framework. This idea was first introduced by Paul Willis (1977). Willis looked at a group of disaffected working-class boys in a Secondary Modern (“the lads”). This group rejected the school curriculum, subverted the authority of teachers and administrators, and disrupted classes. The lads’ rejection of the school, Willis argued, was partly the result of their deep insights into the economic conditions of their social class under capitalism. However, by equating manual labour with success and mental labour with failure, their cultural outlook limited their options; they did not see that their actions led simply to lower-paying jobs. Seeing no connection between school achievement and social mobility, they chose to join their male relatives on the shop floor; this was a choice made cheerfully and without coercion. Willis’ key innovation came in attributing agency to the lads, who made real choices to continue in working class jobs. This framework has been attractive to those who have sought to explain racial inequalities in educational attainment: The Willis study introduced ‘resistance’ as a force of human agency in the process of the reproduction of class inequality through schooling…It is precisely because of the introduction of students and school personnel as human agents actively involved in the process of cultural reproduction or resistance that this kind of study is relevant to the educational problems of racially stratified groups (Ogbu 1997a: 770). 1

Dillingham Committee, 1910

4

The discrimination against blacks in the labour market before the 1960s, Ogbu argued, was responsible for the creation of a disconnection between educational success and their treatment in the labour market (Ogbu 1978, 1997b, 2003). This had two consequences. Firstly, blacks spent little time thinking about how to prepare for specific jobs whilst in education. Secondly, the uncertainty about whether education would pay off encouraged them to leave school at the earliest opportunity. Thirdly, it meant that blacks were often unable to see the connection between success at different levels of schooling. For example, they might not have a strong grasp of the way in which their effort and performance at elementary school affects that at middle school, and so on. Ogbu’s framework has been termed the “oppositional culture” explanation for academic performance.

Oppositional Culture and Cultural Inversion Ogbu’s theory, it is argued here, has four main facets. Firstly, in an abstract sense, black children and parents recognise the value of education. Ogbu (1978) documented the deep and abiding belief of blacks in education as a crucial means to attain upward mobility. Secondly, this belief is not translated into better academic performance as blacks are aware that they face a job ceiling. The discordance between abstract beliefs in the importance of education and actual achievement has been termed the “attitudeachievement paradox” (Mickelson 1990). Ogbu (1974, 1997b, 2003) argued that black families may initially have high aspirations for their children, but as they are made aware of barriers to opportunities in further education and the labour market they develop a bleak view of their children’s prospects. Thirdly, the children themselves choose not to compete for scholastic rewards. The responses of their parents are reinforced by the child’s observation of the community surrounding them. Siblings, relatives and other adults who have left school may have had difficulty getting jobs. The children themselves may have experienced barriers to securing summer and part time jobs. They may have watched or participated in demonstrations for more jobs, or read reports in the media discussing the problem of black unemployment. Finally, in consequence, they adopt alternative strategies within a limited opportunity structure to defend against “failure” in mainstream schools and jobs. An important feature of these strategies is that they are created in opposition to the culture of white Americans, contributing to the formation of an oppositional cultural identity. Ogbu later refers to this as “cultural inversion”: Cultural inversion is the tendency for involuntary minorities to regard certain forms of behavior, events, symbols and meanings as inappropriate for them because these are characteristic of white Americans. At the same time the minorities value other forms of behavior, events, symbols and meanings, often the opposite, as more appropriate for themselves (Ogbu 1992: 8).

Fordham and Ogbu (1986) developed this idea further by claiming that this form of identity was reinforced by a sense of group loyalty among black Americans. Through this, members of the group are sanctioned for violating the established norms of the dominant culture. Fordham and Ogbu argued that black Americans face the “burden of ‘acting white’” if they aim for success academically, as this is seen to be the role of white Americans. In addition to seeing academic success as the territory of whites, they may see sporting success as their own terrain:

5

You know, we’re not as smart as they [ie. White students] are. They don’t have any muscles, we have muscles [for sports and athletics] (Black student. Quoted in Ogbu 2003: 89).

Fordham and Ogbu (1986) noted the tendency of “good” students to cope with the “burden of ‘acting white’” by increasing their participation in extra-curricular activities such as sport and cheerleading. There was, however, a recognition that not all sports were “black”. “White sports” included golf, lacrosse, tennis and skiing. Those who played these sports were often subject to criticism (Ogbu 2003). A typology of minorities Ogbu accounted for the fact that many ethnic minorities succeed at school, and resist the adoption of an oppositional cultural identity, by differentiating between three groups of minorities (Ogbu 1987, see also Ogbu 1974, 1978, 1981a, 1986, 1991a, 1991b, 1997a, 1997b, 1999, 2003). 1. Autonomous minorities: Minorities who possess a distinctive religious, ethnic, linguistic or cultural identity, but that are not socially, politically and economically subordinated are termed “autonomous minorities” (Ogbu 1986, 1987). Jews and Mormons fall into this group. Ogbu alludes to this group in few of his publications. 2. Immigrant minorities: Groups that migrated freely to the host country, for example East Asians, compare their condition to that of relatives in their homelands. This comparison is generally favourable so they develop optimistic attitudes regarding both their chances for success in the new country and reward for their efforts. Voluntary minorities see cultural differences as barriers to be overcome in order to achieve their long-range goals of future employment and assimilation into the host society. These groups also have a “symbolic option” of returning to their homeland, unless they have emigrated for political reasons (Ogbu 1986). 3. Involuntary minorities: Individuals from historically oppressed minorities such as black Americans, who have been conquered or relocated against their will and lack an identifiable foreign reference group, do not hold such positive expectations for their future. For this group, the central problem is the nature of the history, subordination, and exploitation of the minorities, and the nature of the minorities’ own instrumental and expressive responses to their treatment (Ogbu 1987: 317).

They learn from those around them that they have limited job opportunities and place little emphasis on success at school, developing a pattern of linguistic, cognitive, motivational, and other skills related to school which is at odds with the ethos of the American public school system. They do not accept their ascribed menial position and blame the “system” for their subordination. Involuntary minorities see cultural differences between them and the dominant group as barriers to be retained, and thus avoid complete assimilation into mainstream society. These minorities have also been described by Ogbu as “castelike” minorities (Ogbu 1978, 1981a, 1986, 1987, 1991a, 1991b, 1997a, 1997b, 2003). Ogbu (1978, 1981a, 1986) argued that the United States has a “caste system” as well as a class system because it satisfied Berreman’s (1967, 1972) definition of a caste system as a hierarchy of endogamous groups whose membership was determined permanently at 6

the time of birth. He contended that while attributes that determine class position can be acquired over the life course, the membership of a caste cannot: Class is a matter of acquired status rather than of birth-ascription, and is in this respect distinct from race, caste, and ethnic stratification… In a class system, one is ranked in accord with his behaviour and attributes (income, occupation, education, lifestyle etc.). In a birth-ascribed system… one behaves and exhibits attributes in accord with his rank. (Berreman 1972: 398-399).

Although he acknowledged that within the black and white castes there are class systems, these are not equal because the two groups’ access to education, jobs, income and other resources that enable class mobility are not uniform. Blacks experience a “job ceiling”; lower class whites do not. Berreman saw racism as “the basis of American caste” (Berreman 1972: 399). Middle class blacks In his final work, Ogbu (2003) focused on the question of whether the experiences of middle class and working class blacks had attitudinal commonalities through examining views about academic success among the more affluent black youth of Shaker Heights, Ohio. In the book that resulted, he claimed that there are a great many similarities between the attitudes of poor urban blacks and their counterparts in wealthier communities. These included disengagement with academic work, an inability to focus on the task at hand, low expectations for themselves, and a tendency to place the blame for their poor performance on their teachers. Although Ogbu acknowledged that negative stereotyping and racism may have played a part, he did not believe that these factors fully explained the black-white achievement gap.

Research in support of Ogbu’s framework The work referred to below is all American research; no substantive work on this framework has been carried out in Britain to date. In an overview of American research on ethnicity and schooling, Majors and Billson (1992) argued that “the American virtues of thrift, perseverance, and hard work” did not generate the same rewards for blacks as for whites (Majors and Billson 1992: 1). To cope with this, they argued, black males tended to adopt a “cool pose”. This notion is comparable to Ogbu’s “acting white” hypothesis. Black males distanced themselves from “uncool” activities such as studying, going on field trips to museums and relating positively to teachers. Fischer et al. (1996) echoed Ogbu’s work in attributing the low performance of blacks on IQ tests to low ethnic or caste status rather than to any biological factor. They directed a great deal of criticism at Herrnstein and Murray’s The Bell Curve, which argued that the genetic make-up of an individual determined their intelligence and in turn their performance on IQ tests (Herrnstein and Murray 1994). Fischer et al. asserted that IQ tests are simply “a test of what people have been taught… how much they recall, and … how much effort they make in the test” (Fischer et al.: 17). Like Ogbu, they argued that it is the fact that involuntary minorities are “rationally pessimistic” that breeds a resistance to schoolwork as the economic advantages of staying on in education are small (Fischer et al: 186). Neither Major and Billson’s or Fisher et al.’s research is based on empirical evidence. 7

Graham et al (1998) investigated the achievement values of middle school students. They examined two schools; one predominantly black, the other ethnically diverse. They used peer nomination procedures to test Ogbu’s theories. Respondents nominated peers whom they admired, respected and wanted to be like. They were also asked questions about which pupils tried hard and received good grades, which did not try hard and received poor grades, which followed and did not follow school rules, which were good at sport and which wore nice clothes. Finally, they collected data on actual academic achievement. They found that girls of all ethnicities valued high achieving classmates of the same gender. White boys valued boys of the same ethnicity who were high achievers. Ethnic minority boys, by contrast, least valued high achieving male students. This finding clearly supported Ogbu’s thesis. Farkas et al. (2002) use the National Assessment of Educational Progress (NAEP) dataset to explore the evidence for an oppositional culture among black students. They assessed responses to the question “my friends make fun of people who do well” to determine which gender/ethnic groups were subjected to the greatest peer opposition to school. They found that males experienced more peer opposition than females, and that blacks and Hispanics were subject to this to a greater extent than whites.

Research questioning the validity of Ogbu’s framework Erickson (1987) attacked Ogbu’s claims from three angles. Firstly, he pointed to the fact that Ogbu did not account for the success of some individuals from domestic minorities, including those individuals from castelike groups who exhibited high academic performance. Secondly, he saw Ogbu’s framework as economically deterministic and mechanical. This view of society involves a number of tight and immovable causal connections which drive the actions, perceptions and sentiments of the populations involved. When this type of social theory is applied to education, teachers and minority students are reduced to impotence; neither group is able to do anything positive to improve the performance of individuals or groups. Indeed, Ogbu’s theory portrayed a very bleak vision of the future for the education of blacks in America. Finally, Erickson questioned the causal validity of Ogbu’s work. Although Ogbu asserted causal relationships, he did not demonstrate them directly. Where there is empirical evidence, it provided confirmation of correlation only; no amount of correlational evidence can demonstrate cause. Trueba (1988) viewed Ogbu’s framework as overly structuralist and deterministic and argued that his classification of minority groups exhibited “overwhelming generalization” (Trueba 1988: 276). He saw the role of culture as important at the individual as well as the group level, and criticised Ogbu for his failure to explain why some minorities succeed. He notes, for example, that the castelike group of Mexicans in the United States exhibits “educational progress, an increase in English language proficiency, and upward economic mobility” (Trueba 1988: 277-8). Indeed, one weakness of Ogbu’s work is that he looks at “success” in absolute terms, rather than taking into account the relative success of minority groups when compared to their performance at an earlier date. Jaynes and Williams (1989) pointed to Ogbu’s failure to carry out comparative research on white students, arguing that unless he could establish whether the educational aspirations of black students were developed within a significantly more negative or uncertain framework than that of whites, his research could not be validated. MacLeod’s

8

(1987, 1995) research on two groups of teenagers in a high school, the “Hallway Hangers” and the “Brothers” supported this view. The social class consciousness of the predominantly white and poor “Hallway Hangers” resulted in them making less effort at school than their similarly impoverished and mainly black “Brothers” who were far less conscious of social class inequalities. O’Connor (1997) focused on 6 African American students from low income families. She highlighted the varied nature of black responses to education, arguing that the work of Ogbu and his collaborators was too preoccupied with explaining the academic failure of “marginal populations” and gave insufficient attention to students who are able to adapt to the norms and expectations of the school. She also questioned the assumption that those that succeed disaffiliate from their ethnic group. Her students were high achieving and expected to realise their ambitions. They all aspired to college and some to graduate study and aimed to assume middle class professions. Their academic histories suggested that these were realistic aims. They did not operate within the context of “acting white”; they embraced black culture and articulated a sense of affiliation to that culture. Nor did they display a lack of awareness of the disadvantages that they faced: The white people do have a hold on us, but the hold ain’t full. You know, they own most everything, control most everything, and they got more money. But we can break that hold – we can fight it. If we don’t at least try to break that hold, we never going nowhere. (Sharon, aged 16. Quoted in O’Connor 1997: 594).

O’Connor recognised that the selective nature of her sample did not permit generalisation and pointed to some counter-examples which indicated that these attitudes were not universal. Her key point was that the accounts of Ogbu and his collaborators failed to recognise that black responses to “white” education were not uniform. Cook and Ludwig (1998) described Ogbu’s oppositional culture theory as “one of the ‘chestnut’ explanations for black-white differences in educational outcomes” (Cook and Ludwig 1998: 375-6). They used the National Education Longitudinal Study (NLES) to assess the theory’s validity. Firstly, they looked at whether African American adolescents reported greater alienation from school than non-Hispanic whites. Secondly, they questioned whether academic success led to greater social ostracism among this group. They measured alienation from school by looking at the educational expectations of blacks, drop-out rates, effort and parental involvement in the school. Their conclusion was that blacks did not exhibit greater alienation than whites; their desire to stay on in education was high, their drop-out rates low, their effort levels impressive and parental involvement high. To assess whether there were greater social costs of academic success for blacks than for whites they looked at respondents’ experience of threat and whether they were “popular” or “part of the leading crowd”. Here they found that high-achieving blacks were more popular than low-achievers and that black honor society members were substantially more popular. One weakness of their paper is the absence of any serious measure of peer pressure or norms. Since the behaviours of whites and blacks were found to hardly differ, they concluded that peer norms regarding academic performance did not differ either. In addition, they do not consider the fact that any avoidance of “acting white” might not necessarily lead to lower attachment to school. Blacks might moderate their speech or try to “act ghetto” to assert their racial status, but such behaviours might not be correlated with attendance rates, homework hours and so on. In an important paper, Ainsworth-Darnell and Downey (1998) used the NLES to assess the claims of the oppositional culture thesis. They focused on African American, Asian American and non-Hispanic white students in order to represent, respectively,

9

involuntary, immigrant and dominant groups in the United States.2 They found that African American students were more likely to emphasise the importance of education in getting a job and to have optimistic educational expectations. In terms of “skills, habits and styles”, teachers’ evaluations and self-reports from respondents, their findings suggested that African Americans made less effort and were more likely to be disruptive than white students. Despite this, however, they exhibited more pro-school attitudes than whites. They reported good treatment from teachers to a greater extent than whites and were less likely to agree that it is acceptable to break school rules. They were more likely to get satisfaction from doing what they were supposed to do in class and were more likely to report that they tried hard. Rather than academic success being equated with unpopularity (as predicted by the “acting white” hypothesis), African American students were found to be especially popular when they were viewed as a “good student”. Ainsworth-Darnell and Downey found that differences between blacks and whites were substantially reduced once social class was controlled for, and became insignificant when skills, habits and styles were added to the model. They suggested that the low performance of African Americans at school could be largely attributed to their material circumstances; the communities within which many of them lived did not foster the development of skills, habits and styles that were rewarded by teachers. For AinsworthDarnell and Downey, however, these circumstances did not inevitably lead to the development of an oppositional culture to school. Using the NAEP dataset, Downey and Ainworth-Darnell (2002) found that black students were more likely to report that doing well at school was the aim of most of the students in their class. They were also more likely to report that they were comfortable going to friends for help with homework. They found little support for the “attitudeachievement paradox”; overall, there was a greater correlation between attitudes and attainment for blacks than for whites. Cross (2003) disputed Ogbu’s historical analysis, arguing that the creation of a culture fostering low achievement was far more contemporary than Ogbu supposes. He asserted that blacks exited slavery with all of the necessary attributes to succeed educationally: high social capital, inclusive and proactive family attitudes and patterns and a high motivation to achieve educationally (Cross 2003, see also Fischer et al. 1996). Immediately following the Civil War, the ex-slaves themselves organised and carried out educational ventures which aimed at spreading literacy among blacks. Later, blacks took advantage of the support of the Federal Freedman’s Funding Bureau, Northern benevolent societies and white teachers from the North to expand their educational system further. Cross argued that whites could have taken advantage of the high achievement motivation of blacks. Instead, however, they under-funded the black schooling system when blacks threatened to outperform their own offspring. Although unable to pinpoint the point when the positive motivation of blacks was eroded, Cross was certain that “low achievement motivation and undervaluing the role of education [were] not cultural themes carried over from slavery” (Cross 2003: 76).

“Castelike” minorities in Britain To date, empirical work testing the validity of Ogbu’s thesis has been carried out in the United States. The emphasis has been predominantly on African Americans, although Ogbu did make reference to other minority groups in connection with involuntary 2

The Asian American group includes Chinese, Filipino, Japanese and Korean students. It excludes SouthEast Asians.

10

minorities, including Mexicans, Native Hawaiians, Native Americans and Puerto Ricans (Ogbu 1997). In the British context, which will be the focus here, Ogbu is inconsistent in his identification of castelike minorities. In The Next Generation Ogbu singled out AfroCaribbeans and Pakistanis as falling into this group (Ogbu 1974). In Minority Education and Caste he expanded the group to include all “colored” minorities (Afro-Caribbeans, East Indians, Pakistanis and Bangladeshis), asserting that these groups face a level of discrimination not experienced by white immigrants (Ogbu 1978). In “The consequences of the American caste system” he identified Afro-Caribbeans as being a castelike minority group (Ogbu 1986). Throughout, however, Ogbu’s primary emphasis is on the black population. He argues that they have a more serious language barrier that Pakistanis and Bangladeshis because it is hidden, “an English dialect” (Ogbu 1978: 251). He states that the disenchantment of blacks is greater than other minority groups, and that they experience greater residential segregation as “their communities in Britain are neither cohesive nor exclusive like those of the Indians and Pakistanis”. As discussed below, Ogbu’s identification of these groups as castelike minorities, and of blacks as the most classic case of castelike status is at times questionable. Although it is without doubt that such groups face high levels of discrimination in Britain, they do not possess all of the attributes required to be defined as castelike, or involuntary, minorities. Firstly, one of the central features of the castelike minorities that he identified in the United States is that they came involuntarily to the country. The same cannot be said of British ethnic minority groups. Afro-Caribbean migration was one of the first waves of immigration following the Second World War. They were attracted by a strong demand for labour in Britain as a result of post-war reconstruction and the expansion of Western European economies which has been attributed, in part, to the Marshall Plan (LaytonHenry 2002). Access to the United States also became harder after the passage of the McCarran-Walter Act of 1952. Ogbu himself asserted that Afro-Caribbeans came to Britain of their own volition, with the majority financing their own passage. He also pointed to the economic motivation of Pakistanis to emigrate to Britain. Their immigration was largely dependent on a system of sponsorship and patronage whereby an individual (usually male) who was already working in Britain would pay to bring a relative to Britain to work. The new immigrant would then be required to settle his debt with his sponsor and to send some money home to his family (Ogbu 1978). Secondly, it is not necessarily the case that ethnic minority groups in Britain have no “homeland” to return to. Many ethnic minorities retain strong links with their country of origin. As Layton-Henry (2002) points out, international migrants are no longer as isolated from their countries of origin as they were in the past. Modern communication through the internet, telephone, satellite television and cheaper air travel means that migrants can maintain contact with the people and institutions of their country of origin. They can communicate regularly, sometimes daily, if their means permit it. In LaytonHenry’s Birmingham sample 94 per cent of Afro-Caribbeans had family members in the Caribbean with whom they were in contact. The remaining 6 per cent had no family, but did have friends in their country of origin. 83 per cent of the sample stated that they tried to keep informed about events in the Caribbean through reading Caribbean newspapers published in Britain, through news broadcasts and through contact with family and friends. However, while the second generation are able to maintain such contact through their parents, it may be that ties weaken by the third generation. Thirdly, Ogbu (1986) argued that legal devices in the United States had been instrumental in relegating blacks to menial positions. Britain has never had a formal law limiting the activities of ethnic minorities in Britain. Ogbu noted the absence of such a 11

law, but asserted that this did not affect the validity of his case for denoting them “castelike” minorities: Although the inferior position of the colored minorities is not maintained by a formal law, the stratification system contains many features associated with the color-caste system found in America (Ogbu 1978: 245).

This suggests that the way in which Ogbu developed his theory was by creating a post hoc explanation for sociological phenomena, rather than formulating a robust theoretical framework to be rigorously tested. Finally, it is not clear that blacks receive substantially more discrimination than other ethnic minorities in Britain. Ogbu (1978) argued that although all minorities were accorded lower class status, it was the Afro-Caribbeans who were most like blacks in the United States. Ogbu asserted that disillusionment and disenchantment were greater among blacks than Indians and Pakistanis. They were, he argued, linguistically and culturally closer to British whites through their practice of Christianity, their “strong” identification with the British and their tendency to regard themselves as British subjects rather than as foreigners. They therefore expected to achieve full and immediate acceptance by the British and were disappointed when this failed to materialise.3 Thomas Hope (1981) claimed that, compared to Afro-Caribbean immigrants to Canada and the USA, Afro-Caribbean immigrants to Britain were significantly less satisfied with their decision to move to a new country due to high levels of discrimination. However, it has recently been questioned whether blacks stand out from other ethnic minority groups in terms of the amount of discrimination that they face, specifically on the labour market. Owen et al. (2000) found that Caribbeans displayed higher rates of economic activity than other ethnic groups and that minority unemployment rates were highest for black-Africans (as opposed to Caribbeans), Pakistanis and Bangladeshis. Male earnings were lowest for the Pakistani/Bangladeshi ethnic group. These findings were not formally tested however; it is difficult to establish whether they were the result of differences in educational attainment or greater discrimination. Berthoud (2000), using evidence from the Labour Force Survey (LFS), found that blacks were not the only group to suffer notable discrimination. Although Indians and Chinese had employment rates and earnings levels similar to those of the white majority, Caribbeans, Africans, Pakistanis and Bangladeshis fared worse. There is also evidence of greater discrimination against the Muslim population, which is largely made up of Pakistanis and Bangladeshis. Modood (1997) found a consensus among Caribbeans, Indians, African Asians, Pakistanis, Bangladeshis and whites that “Asians” suffered the greatest level of discrimination in Britain. Indians, African Asians and Pakistanis felt that within the Asian group, Muslims faced the most prejudice. Blackaby et al. (1999) found that the predominantly Muslim Pakistani and Bangladeshi communities were discriminated against to a greater extent than non-Muslims. Lindley (2002) found that religious affiliation was an important determinant of earnings and employment. Her study found evidence of a considerable disadvantage to Muslims, relative to all other non-whites. Approximately half of this was explained by socio-economic characteristics; the residual was a pure Islamic penalty. Model and Lin (2002) also found substantial evidence of religious discrimination against British Muslims. 3

Richmond (1993) offered an alternative typology of voluntary and involuntary minorities. He argued that those that migrate voluntarily do so as the result of a combination of “push” and “pull” factors while involuntary immigrants do so because of structural, global inequalities and the violence associated with nationalism and independence movements.

12

The British Cohort Study The British Cohort Study sampled over 17,000 babies born in Britain in the week 5th-11th April 1970. Surveys in 1975, 1980, 1986 and 1996 have been carried out to follow up the 1970 sample. The 1986 survey, taken when cohort members were 16, is utilised here. This dataset contains a wealth of information on attitudes to school and work, and on perceived job opportunities and barriers to success. Unfortunately, there is not a more recent dataset with the quality of measures present in the 1986 BCS. The results of the analysis that follows remain important in that they represent the attitudes of the earliest of the second generation of ethnic minorities in Britain. The data also corresponds to the period in which Ogbu’s theory was attracting a great deal of academic focus. Ethnic minorities in the BCS The 1986 BCS contains only small numbers of ethnic minority respondents. The frequencies are detailed below. Ethnic group

N

White Afro-Caribbean Indian Pakistani/Bangladeshi Mixed/other No ethnic group given

9006 74 133 87 85 2230

Table 1 Ethnicity in the BCS Source: British Cohort Study 1986. An investigation into the extent to which each ethnic group evinces an oppositional culture would not be satisfactory if the analysis were simply carried out on the raw data above. The Ns for the majority of the groups are too small for any results to be reliable. Various solutions to this problem have been proposed, and their merits will be discussed below. Case-control methods In a case-control study respondents who exhibit a given trait resulting from the operation of a number of independent variables (here, the possession of an oppositional stance towards school) are compared with controls who do not display the trait. This enables estimation of odds ratios. Allowance is made for potential confounding factors through measuring them and making appropriate adjustments in the analysis. For greater efficiency, this statistical adjustment may match cases and controls for exposure to confounding factors. This may be done on an individual basis, for example by pairing each case with a control of the same age and gender, or in groups, for example by choosing a control group with an overall age and gender distribution similar to that of the cases. Case-control methods have been used effectively in a number of sociological studies (Lacy 1997). The most important feature defining a case-control study is that the sample

13

is stratified on a dependent variable, rather than a set of one or more independent variables. This makes the method unsuitable for use here, because the desired groups for comparison are based on an independent variable, ethnicity. Propensity score matching The propensity score method compares the responses given by a “treatment” group, to those of a group that has not been treated. For example, it might be used to compare the performance of recipients of a training programme with those who had not participated in the event. The propensity score approach is based on the idea that “matching” individuals in the groups should be compared. It emphasises the problems associated with selection bias. In the example of a training programme, for example, it may be that brighter people are opting for participation. Cochran (1965, 1968) advocated matching individuals using all of the covariates available and then estimating the treatment effect by averaging over the set of differences. However, proponents of the propensity score method argue that this method may meet difficulties, as the number of matching cells will increase exponentially as more covariates are used and cells can quickly become devoid of treatment individuals or control cases. Advocates of the approach propose combining all covariates into a “balancing score”, the propensity score (Rosenbaum and Rubin 1983). For an individual this is the probability that they are in the treatment group. If an individual in the treatment group and an individual in the control group have matching propensity scores, their allocation to treatment or control can be considered random. The effect of having the treatment can be calculated by averaging over the set of differences. The key problem with this approach is that it is conceptually questionable whether ethnicity can be regarded as a “treatment”, as membership of an ethnic group is defined at birth, rather than acquired. A further problem is evident in the fact that propensity scores do not “match” exact equivalents; the same propensity score may be allocated to two individuals who exhibit very different background characteristics. Matching cases Given the drawbacks of the methods discussed above, an alternative approach is utilised here. The ethnic group for which the number of cases is smallest, blacks, were matched with whites, Indians and the Pakistani/Bangladeshi group on variables for gender, income and region.4 Ideally, more variables would have been used, but the item response rate on other suitable background variables was too low to match cases efficiently. A new dataset was then created, containing matched cases only. This method has clear advantages over standard regression models. The relatively large size of the white sample in the dataset, and the relatively small sample of ethnic minorities means that the relationships modelled would be almost wholly driven by the responses for whites. It is also likely that the model would be mis-specified; regression analysis assumes that each group is well distributed. In this case it is doubtful whether more than one black, if any, would fall into the cell for the highest income level, for example. This means that it would be impossible to test for all of the interaction effects.

4

See Table A1 in the appendix for full details

14

Missing data On a number of items in the dataset, the item response rate was low. This was particularly problematic given the small number of cases in the dataset. The ExpectationMaximisation (EM) algorithm was utilised in order to impute the missing values. The EM algorithm has maximum-likelihood estimation at its foundation. The maximum likelihood estimate is the value of the parameter that is most consistent with the observed data. In other words, if the parameter equals that number, the pattern observed will have a greater chance of occurring. The EM algorithm is a method of calculating maximum-likelihood estimates through a two step iteration: E (expectation) and M (maximisation). If D denotes the data, it is assumed that there is a parameterised joint distribution on the observed and hidden values (θ). The objective is to maximize the likelihood log P(D‫׀‬θ). As it is hard to optimise the log of P directly, the EM algorithm optimises the expectation E(log P(D‫׀‬θ)). The E-step computes the expected values for missing data given the observed data and the current estimate of θ. During the M-step the parameters are updated to their best possible values given their presumed distribution on the hidden variables. The process is repeated until the frequencies converge. The calculation of missing values in this case is based on the responses given to the 11 items below:

15

Item Qualified people more chance to get job

Coding 1. Agree fully 2. Agree partly 3. Disagree

In comparison my job prospects are

1. 2. 3. 4. 5.

Much less

Possible to get job if really determined

1. 2. 3.

Agree fully Agree partly Disagree

Think homework is a bore

1. 2. 3.

Very true Partly true Not true at all

Feel school is largely a waste of time

1. 2. 3.

Very true Partly true Not true at all

Do not like school

1. 2. 3.

Very true Partly true Not true at all

Never take work seriously

1. 2. 3.

Very true Partly true Not true at all

How often get into trouble at school?

1. 2. 3. 4. 5.

More than others About the same as others Less than others No one in class does Other answer

Removed from class due to behaviour problem?

1. 2.

No need to remove Removed

Pupil ever suspended from school?

1. 2.

Yes No

Do teachers pick on you more than others?

1. 2. 3. 4.

Yes No No one in class does Other answer

About the same Much more

Table 2

Variables employed in calculation of missing values using the EM algorithm Source: British Cohort Study, 1986

An empirical examination of Ogbu’s theory of Oppositional Culture Theoretical dimensions

16

The thesis now moves to an empirical assessment of Ogbu’s theory of oppositional culture. Having presented an analysis of the literature, it is argued here that the theory makes four key points: 1. Black parents and children recognise the value of education in an abstract sense. 2. As they are made aware of barriers to opportunities in further education and the labour market, however, they develop a bleak view of their future prospects. 3. This bleak view translates into a rejection of competition for scholastic rewards; black children adopt anti-achievement values in order to reduce the discrepancy between work put in and rewards received on the labour market. 4. Anti-achievement norms are reflected in the behaviour of black children at school. Each dimension of the theory will be interrogated below. While the caveats of Ogbu’s delineation of castelike minorities in Britain must be borne in mind, it seems logical to test his theory of cultural inversion by utilising his own framework. Blacks will therefore be treated as the “primary” castelike minority, with Pakistanis and Bangladeshis forming the “secondary” group. Although Ogbu alludes to Indians as a castelike minority on one occasion (Ogbu 1978), elsewhere he refers to them as a voluntary minority and they will be treated as such here. Whites are defined as the dominant group. The mixed/other group are excluded from the analysis.5 Abstract and concrete attitudes Ogbu’s emphasis on the distinction between attitudes towards the general value of education and attitudes towards the probability of academic success at the individual level has been supported by other researchers. Mickelson (1990) emphasised the importance of recognising the two types of attitudes which students hold about the role of education. The first of these she called “abstract attitudes”, general ideological beliefs which “embody the Protestant ethic’s promise of schooling as a vehicle for success and upward mobility” (Mickelson 1990: 45). An example of such a belief might be “education is the key to success in the future”. “Concrete attitudes”, on the other hand: reflect the diverse empirical realities that people experience with respect to returns on education from the opportunity structure… Students’ concrete attitudes vary in accordance with their perception and understanding of how adults who are significant in their lives receive more equitable or less equitable wages, jobs, and promotions relative to their educational credentials. Adults whose job returns are commensurate with their education… generally have children who hold very positive beliefs about education and demonstrate high achievement to match. Conversely, children of working-class or minority adults have more pessimistic concrete attitudes toward schooling and earn lower grades (Mickelson 1990: 45-46).

Mickelson gives the following examples of concrete and abstract attitudes: Abstract attitudes 1. Education is the key to success in the future. 2. If everyone in America gets a good education, we can end poverty. 3. Achievement and effort in school lead to job success later on. 5

See Table A2 in the appendix for details of recoding of the ethnicity variable

17

4. The way for poor people to become middle class is for them to get a good education. 5. School success is not necessarily a clear path to a better life. 6. Getting a good education is a practical road to success for a young black [white] man [woman] like me. 7. Young white [black] women [men] like me have a chance of making it if we do well in school. 8. Education really pays off in the future for young black [white] men [women] like me. Concrete attitudes 1. Based on their experiences, my parents say people like us are not always paid or promoted according to our education. 2. All I need to learn for my future is to read, write and make change. 3. Although my parents tell me to get a good education in order to get a good job, they face barriers to job success. 4. When our teachers give us homework, my friends never think of doing it. 5. People in my family haven’t been treated fairly at work no matter how much education they have. 6. Studying in school rarely pays off later with good jobs. Ainsworth-Downey and Darnell (1998) also support this distinction. The tables below display the chi-square statistic (χ2) for the table and adjusted standardised residuals for each cell. The χ2 test calculates what distribution of responses would be expected (fe) if there was no association between the variables and compares these to the observed distribution of responses (fo). It equals χ2 = Σ (f o - f e)2 fe For each cell, the difference between the observed and expected frequencies is squared and then divided by the expected frequency. The terms are then summed to create the χ2 statistic. If there is no relationship between the variables, f o and f e tend to be close so χ2 is quite small. The larger the χ2, the greater the evidence that there is a relationship between the variables. If the χ2 is large, it indicates that somewhere in the table the data depart from what independence (no association between the variables) predicts. It does not indicate which cells of the table show the greatest departure from independence. The difference between f o and f e is called a residual. To test whether a residual is large enough to signify a departure from independence, the adjusted standardised residuals are calculated as follows: _____________ f o - f e__________________ ______________________________ √f e(1-row proportion)(1-column proportion) An adjusted residual greater than 1.96 provides evidence against independence in that cell. An examination of the abstract attitudes of each ethnic grouping, as measured by the percentage fully agreeing that qualified people have more of a chance of getting a job,

18

reveals some support for Ogbu’s thesis. As shown in the table below, the primary castelike minority, Afro-Caribbeans, yields a positive response, with 100 per cent of their number agreeing with the statement that qualified people have more chance of getting a job. However, the secondary ethnic minority groups, Pakistanis and Bangladeshis, also show a high regard for the value of education in this respect, with 99 per cent answering positively. Indians evince the same regard for education. Slightly fewer whites agree that qualifications increase the chances getting a job. It is possible that ethnic minority groups attach greater importance to qualifications as they see them as a tool to compensate for the greater discrimination that they receive on the labour market. As shown by the chisquare figure, the differences between the groups are not significant overall. Only the adjusted standardised residual for whites indicates that the difference between them and the ethnic minority groups is. For the second item, which asks the respondent whether or not they agree that it is possible to get a job if they are really determined, a similar pattern emerges. Once again every group evinces a positive response, with no significant differences between the cells (this time including the white group). On the second item, all four ethnic groups respond very positively. This suggests that although ethnic minorities recognise that there are barriers to their employment, they perceive these as obstacles to be overcome. However, care must be taken not to overinterpret the figures, as the adjusted standardised residuals and the chi-square figure are both insignificant.

19

India n

Pakistani/ Bangladeshi

χ2

P

Qualified people more 95 100 chance to get job (agree) (-2.4)6 (1.4)

99 (0.5)

99 (0.5)

6.12

0.11

Possible to get job if 99 really determined (agree) (0.8)

97 (0.0)

95 (-1.7)

3.08

0.38

White

AfroCaribbean

99 (0.8)

Table 3 Percentage agreeing with items measuring abstract attitudes Source: British Cohort Study, 1986 In the table that follows, the responses are broken down further to look at the percentage of each ethnic group that agrees fully that qualified people have more chance of getting a job and that it is possible to get a job if one is really determined. AfroCaribbean

India n

Pakistani/ Bangladeshi

χ2

P

Qualified people more 33 chance to get job (agree (0.9) fully)

22 (-1.5)

24 (-0.9)

35 (1.5)

11.28

0.08

Possible to get job if 61 really determined (agree (-1.3) fully)

64 (-0.7)

74 (1.6)

69 (0.4)

7.97

0.24

Average

43

49

52

White

47

Table 4 Percentage agreeing fully with items measuring abstract attitudes Source: British Cohort Study, 1986 There is mixed support for Ogbu’s thesis here. On the first item, the response of whites and the Pakistani/Bangladeshi group is more positive than that of Afro-Caribbeans and Indians. Of the ethnic minority groups, Afro-Caribbeans are the least likely to agree fully that qualified people have more chance of getting a job and that it is possible to get a job if one is really determined. This casts some doubt on Ogbu’s view that blacks maintain a view of the value of education comparable to that of other ethnic groups, at least in abstract terms. Indians have a slightly higher percentage agreeing fully, although they too are some distance from whites and the Pakistani/Bangladeshi group. On the second item, all ethnic minority groups are more likely to agree fully that it is possible to get a job if one is really determined. In fact, it is whites that are the most circumspect about this. However, blacks do not display a significantly dissimilar pattern of response to other ethnic groups; the χ 2 figure and the adjusted standardised residuals indicate that the variation by ethnicity is not significant. When concrete expectations about the future are turned to, the picture is somewhat different. 6

Adjusted standardised residuals are shown in parentheses

20

White AfroCaribbean In comparison, my job 45 35 prospects are (greater (2.8) (0.8) than others)

India n 24 (-1.5)

Pakistani/ Bangladeshi 22 (-2.1)

χ2

p

11.46 0.01

Table 5 Percentage agreeing that their job prospects are greater than those of others (a concrete attitude) Source: British Cohort Study, 1986 Ethnic minority groups are significantly less positive than whites about their future prospects. Pakistanis and Bangladeshis are most negative about getting a job on completing school, possibly because they have an expectation of high levels of discrimination. Afro-Caribbeans are actually the most positive ethnic minority group; this is not a characteristic that would be expected of the primary castelike group. It questions Ogbu’s perception that blacks, more than any other ethnic group, learn from those around them that their prospects are poor, and have low expectations for their future. On this measure, the chi-square statistic is significant. It may be questioned whether the pattern of response reflects reality. To a certain extent it does. Whites have the highest level of employment for both males, at 80 per cent, and females, at 71 per cent. Pakistanis and Bangladeshis have the lowest levels. For Pakistani men the figure is 61 per cent for men and 24 per cent for women. For Bangladeshis the rate of employment is 55 per cent for males and 17 per cent for females. The level of employment for Afro-Caribbeans falls between the white group and the Pakistani/Bangladeshi group at 67 per cent for men and 65 per cent for women. The Indian group is the anomaly here. 73 per cent of Indian men and 58 per cent of Indian women are in employment; for males, this represents the highest level of employment for any ethnic minority group. For females, the figure is lower than that for Afro-Caribbeans, but significantly higher than that for Pakistanis and Bangladeshis. Yet Indians have very negative expectations as far as their job prospects are concerned.

21

60

50

40

Abstract Concrete

30

20

10

0 White

Afro-Caribbean

Indian

Pakistani/Bangla

Figure I Positive response to abstract items (agree fully) against positive response to concrete item Source: British Cohort Study, 1986 Figure I displays the results graphically, demonstrating clearly that there is a significant disjuncture between abstract and concrete attitudes for all ethnic minority groups. The abstract measure used is an average of the percentages of a given ethnic group fully agreeing that “qualified people have more of a chance of getting a job” and that it is “possible to get a job if really determined”. The concrete measure represents the percentage of each ethnic group responding that “in comparison [with others] my job prospects are…” greater than others. The ethnic minority groups display a far greater distance between the two measures, as Ogbu would predict. However, it is not the primary castelike minority, Afro-Caribbeans, for which the gap is greatest. Contrary to Ogbu’s theory, it is a secondary castelike group, Pakistanis and Bangladeshis, who exhibit the most significant divergence. In addition, the gap is greater for the Indian group, which Ogbu defines as a voluntary minority, than it is for Afro-Caribbeans. Anti-achievement norms and behaviour Ogbu argued that blacks were more likely to exhibit anti-achievement norms than any other ethnic group in an attempt to minimise the discrepancy between effort at school (or in the workplace) and rewards received. This view is supported by the data displayed in Table 6. Afro-Caribbeans score lower than any other ethnic group on all four items, although on the homework item their response corresponds with that of whites.

22

χ2

p

16 (0.9)

Pakistani/ All Bangladeshi 20 13 (2.1)

16.03

0.01

81 (-1.2)

87 (0.3)

88 (0.7)

86

4.06

0.67

Do not like 35 school (not (-0.8) true at all)

30 (-1.9)

41 (0.3)

50 (2.3)

39

9.33

0.16

Never take 64 work seriously (-0.9) (not true at all)

60 (-1.7)

68 (0.0)

80 (2.6)

68

9.01

0.17

Index value (0 0.74 – 2.06)

0.83

0.68

0.54

0.70

-

-

Deviation +0.04 from overall mean

+0.13

-0.02

-0.16

-

-

-

Variance

0.33

0.46

0.37

0.31

0.37

Standard deviation

0.57

0.68

0.61

0.56

0.61

-

-

Item code h

White AfroCaribbean Think 8 8 homework is a (-1.5) (-1.5) bore (not true at all)

Indian

w

Feel school is 87 largely a (0.3) waste of time (not true at all)

d

n

Table 6 Responses to measures of anti-achievement norms Source: British Cohort Study, 1986 It is interesting that whites also exhibit fairly negative attitudes to school and school work, given that they do not share the relatively low level of negativity as regards their future prospects. It might be that they are fairly complacent about their future, since they do not anticipate the discrimination that other groups may expect to be subject to. The South Asian groups show the lowest level of anti-achievement norms. Given the low expectations that they hold for their job prospects, this might appear counter-intuitive. However, as described earlier, they exhibit a strong sense of their ability to overcome these barriers if they are “really determined”. The adjusted standardised residual for the Pakistani/Bangladeshi group is significant on the “never take work seriously” and “think homework is a bore” items, indicating that their response is significantly different than would be expected if there had been no relationship between ethnicity and attitudes towards school. It may be that they are more willing than Afro-Caribbeans to put in a disproportionate amount of effort for lower returns than whites. It is important to note,

23

however, that the differences between the ethnic groups are only small; the results must not be over-interpreted. The only measure that is significant overall is that pertaining to homework, as measured by the chi-square statistic. The index is a summary measure of the responses of individuals in a given ethnic group to all of the questions relating to anti-achievement norms. It is created as follows: 1. Each variable is recoded as a binary variable, where 1 indicates a negative attitude to school and 0 designates a positive attitude.7 2. A value for the individual’s response to each item is calculated. The response code (r) for each binary variable is multiplied by the overall percentage giving a positive response. This weights the variables according to the rarity of possessing each anti-achievement value. The values for the items in the table would be calculated as follows: h = r*0.13 w = r*0.86 d = r*0.39 n = r*0.68 3. The index of anti-achievement norms for the individual (y i) is calculated through summing the score for each item: i = (h*0.13) + (s*0.86) + (d*0.39) + (n*0.68) 4. The index of anti-achievement norms for each ethnicity (y e) equals the mean (y) of the index values generated by that group: y e = (h*0.13) + (s*0.86) + (d*0.39) + (n*0.68) y Here, the deviation from the overall mean is the difference between y e and the overall mean (y). A positive figure indicates the existence anti-achievement norms which are more pronounced than the average. The variance (s2) is a measure of how spread out the distribution is. It is computed as the averaged squared deviation of each value from its mean. Here, the variance for each ethnic group is calculated as follows _______ s 2 = / Σ (y i – y ) √ n-1 where n equals the number of scores. The formula for the standard deviation is the square root of the variance: _______ s = / Σ(y e – y ) √ n–1 7

See table A3 in the appendix for full details

24

In a normal distribution about 68 per cent of the scores are within one standard deviation of the mean and about 95 per cent of the scores are within two standard deviations of the mean. Table 7 shows, as measured by the index value, Afro-Caribbeans exhibit the highest degree of anti-achievement norms overall by some distance. Whites score slightly above the average, Indians slightly below, and the Pakistani/Bangladeshi group some distance below. This lends further credence to the earlier findings; not only did Pakistanis and Bangladeshis exhibit more pro-school attitudes than other groups in terms of raw response rates, they were also the only group whose responses were significantly different from the other ethnic groups as measured by the adjusted standardised residuals. The variance and standard deviation figures are interesting in that they indicate that AfroCaribbeans make up the group that fluctuates to the greatest extent around the mean. This questions the determinism of Ogbu’s theoretical framework; if Afro-Caribbeans exhibit the greatest variance from the mean, it suggests that the development of a black oppositional culture brought about by the experiences of their co-ethnics is far from inevitable. Ogbu argued that anti-achievement norms are reflected in individuals exhibiting behaviour which is at odds with the expectations of the school. The responses to four items measuring behaviour are tabulated below. Indices are calculated in the same manner as above.

25

All

χ2

p

68 (0.6)

Pakistani/ Bangladeshi 76 (2.3)

65

23.58

0.02

88 (-1.8)

92 (-0.3)

95 (0.8)

95

4.13

0.25

ever 100 from (1.0)

97 (-1.7)

100 (1.0)

99 (-0.3)

99

3.70

0.30

Do teachers pick on 89 you more than (1.0) others? (no)

83 (-1.0)

85 (-0.2)

87 (0.2)

86

11.12

0.27

Index value (0 – 0.16 2.44)

0.36

0.18

0.19

0.22

-

-

Deviation overall mean

+0.14

-0.04

-0.03

-

-

-

White AfroCaribbean How often get into 60 55 trouble at school (-1.1) (-1.9) (less than others)

Indian

Removed from class 96 due to behaviour (1.3) problem (no) Pupil suspended school (no)

from -0.06

Variance

0.18

0.46

0.17

0.25

0.27

-

-

Standard deviation

0.43

0.68

0.41

0.50

0.52

-

-

Table 7 Responses to measures of behaviour Source: British Cohort Study, 1986 Once again, on all four items, Afro-Caribbeans exhibit the most “anti-school” responses. The differences on many items are very small, however. The only variable for which the chi-square is significant is “how often do you get into trouble at school?” As might be expected from the response pattern on the questions relating to anti-achievement norms, the South Asian groups are least likely to exhibit behaviour which is at odds with the school. Afro-Caribbeans are the only group whose score on the index is higher than the overall figure, supporting Ogbu’s thesis. As with the anti-achievement norms index, however, they are also the group to display the greatest variation around the mean. Once again, this brings into question the extent to which the development of an oppositional stance to school is universal among blacks. The adjusted standardised residuals indicate that the differences are particularly salient for the Pakistani/Bangladeshi group. On the items for which there are larger differences, the results support Ogbu’s thesis. Afro-Caribbeans report getting into trouble at school to a larger extent than all other ethnic groups. This is backed up by the teacher response to the item asking whether a given pupil had been removed from class due to a behaviour problem.

26

When looking at behavioural measures, it is important to retain a degree of circumspection. It has been suggested that black children are no more badly behaved than white children but that they are disciplined more often by teachers because of their “style”. A number of studies have focused on “style” to explain poor relations between Afro-Caribbean pupils and their teachers (Callender 1995). “Style” is defined by Kochman (1981) as: “An attitude that individuals within a culture express through their choice of cultural form… Black style is more self-conscious, more expressive, more expansive, more colourful, more intense, more assertive, more aggressive and more focused on the individual than is the style of the larger society of which they are part” (p. 130).

Kochman asserts that this “style” permeates every aspect of life, including language, body language and mode of dress. It is argued that one aspect of “style”, body language, is being interpreted by some teachers as expressing insolence and a reluctance to conform to classroom rules. It is also linked to the disproportionate numbers of Afro-Caribbean children, especially boys, being excluded from schools. An examination of the validity of this concept is beyond the scope of the thesis, but it is certainly worthy of further investigation.

Concluding remarks The results presented here manifest mixed support for Ogbu’s theory of oppositional culture. The primary castelike group in Britain, Afro-Caribbeans, demonstrate a high regard for education in the abstract sense, with 100 per cent agreeing that qualifications are important in getting a job. This positive response is not reflected in their response to more concrete questions about their future, in accordance with Ogbu’s framework. Although they appear slightly less positive when agreement is further broken down into “agree fully”, their responses are not significantly different to those of the other ethnic groups. On the second abstract measure, the majority of Afro-Caribbeans agree that it is possible to get a job if they are really determined. Once again, however, their pattern of response is not significantly different from that of the other groups. Afro-Caribbeans are likely to feel that their job prospects are low compared to others (although both Asian groups are even less likely to feel positive about their future prospects). Ogbu does appear to be correct on the direction in which anti-achievement norms are exhibited by Afro-Caribbeans. Afro-Caribbeans are most likely to think that homework is boring and that school is largely a waste of time. They also report disliking school and not taking work seriously more than any other ethnic group. The extent to which Ogbu reports these norms to operate must be questioned, however. It must be remembered that the ethnic differences on these items are extremely small, and the chi-square statistic is only significant for one item. Anti-achievement norms appear to be reflected in the behaviour of the Afro-Caribbean group. On all four behaviour items they report a higher level of poor conduct, although once more the differences are small and the chi-square figure is only significant for one of the variables. As regards the secondary castelike group, Pakistanis and Bangladeshis, Ogbu’s framework appears to be less valid, at least in the British context. Like the AfroCaribbeans, they have positive abstract attitudes towards education, but are negative in terms of their prediction of their job prospects. Their almost universal agreement that it is possible to get a job if one is really determined puts Ogbu’s theory into question, though. 27

Given that all minority ethnic groups are more likely to respond positively to this item than whites, however, it can be argued that ethnic minorities are simply demonstrating an awareness of the discrimination that they will face on the labour market and a consciousness of the effort that they will have to make in order to succeed in securing a job. On the anti-achievement items, the responses of the Pakistani/Bangladeshi group evince far less support for Ogbu’s theory of oppositional culture. Despite their low expectations for the future, Pakistanis and Bangladeshis exhibit a high regard for school. They are the group most likely to report a positive attitude towards the school as an institution and the value of homework. They display the lowest proportion reporting that they dislike school, and the highest proportion responding that they take their work seriously. They also report getting into trouble less often than any other ethnic group. For many of the items, it is only the cell for the Pakistanis and Bangladeshis that is significant, further suggesting that Ogbu’s delineation of this group as a castelike minority is inaccurate. It may also be noted that the Indian group, defined here as a voluntary minority, do not display a significantly different response pattern to the Pakistani/Bangladeshi group. Indians have a positive abstract view of education, manifest low expectations for their future job prospects, and exhibit a low degree of anti-achievement norms and behaviour. This suggests that the dichotomy between voluntary and involuntary minorities proposed by Ogbu is inaccurate. An important final point to be made is that it is notable that the behaviour of the white group does not sit well within Ogbu’s framework. As might be expected, they are positive about their future job prospects, possibly because of an awareness that they are less likely to be subject to discrimination on the labour market. However, the attitudes to school that they report are not significantly different to those reported by AfroCaribbeans. On the behaviour items the picture is more mixed, but overall there is little suggestion that they stand apart from the castelike minorities. There are a number of reasons why one would expect rigorous empirical research to fail to replicate Ogbu’s results. A central problem with his research is that he did not systematically compare black students with white students. Without some evidence that anti-achievement norms are stronger among blacks than whites, one cannot conclude that they contribute to racial differences in achievement. Indeed, it is possible that many school pupils share anti-achievement norms. While blacks might be accused of “acting white”, for example, whites may be called “swots”. Moreover, Ogbu focused almost exclusively on anti-achievement norms without comparing these to pro-achievement norms. This had the effect of making many of his conclusions overly deterministic. He also ignored high achieving groups: Much of the research on Black students has focused on those who have pronounced academic needs and who are perceived as academically less prepared than their White counterparts. The disproportionate focus on Black underachievement in the literature… creates, perhaps unintentionally, a lower set of expectations for Black student achievement (Fries-Britt 1998: 556).

It may be the case that Ogbu’s theory is more applicable to the United States than it is to Britain. As discussed earlier, his typology of minorities as regards Britain is at times inconsistent. It is not clear whether his definition of British minorities as voluntary or involuntary minorities is appropriate. In terms of migration patterns, it is highly 28

questionable whether any ethnic minority group can be defined as an involuntary minority in Britain.8 Ogbu is also erroneous in stating that the primary involuntary minority, blacks, have little contact with their homeland. As Layton-Henry (2002) demonstrated, the majority of blacks in Britain have some form of contact with family, and remain in touch with events in their country of origin, particularly through newspapers. The absence of legal devices restricting the activities of blacks since their migration to Britain further questions the suitability of Ogbu’s schema. It may also be questioned whether blacks in Britain are subject to substantially more discrimination than other ethnic minority groups. In addition, the current situation of blacks in Britain is somewhat different to that in the United States, particularly with regard to residential segregation. Ethnic minority groups in Britain tend to be concentrated in certain localities; in 1991 the four metropolitan counties of Greater London, the West Midlands (Birmingham), Greater Manchester and West Yorkshire (Bradford/Leeds) contained over two thirds of the ethnic minority population. The figure for the overall population was one fifth (Peach 1998). Compared to the United States, however, levels of housing segregation are low. In Britain in 1991 there was not a single town or city in which all ethnic minorities constituted a majority (Peach 1996). In the United States, 64 per cent of African Americans alone live in areas where they are make up the majority (Johnston et al. 2002). Of the four ethnic minority groups examined here, Afro-Caribbeans tend to have the lowest levels of residential segregation; Bangladeshis evince the highest level, followed by Pakistanis, then Indians (Peach 1996, 1998, Johnston et al. 2002). It has been estimated that Afro-Caribbean levels of segregation are about half the level of those of African Americans in the United States (Peach 1996). The Afro-Caribbean population in Britain not only has low, but decreasing levels of segregation with evidence of diffusion outwards (Western 1993, Peach 1996). This is a very different trend to that observed in the United States, where levels of black-white segregation are both high and steady (Massey and Denton 1987, 1989, Massey et al. 1994). Clearly this has implications for Ogbu’s theoretical framework. A key point that his thesis makes is that it is the observation of large numbers of co-ethnics around them that leads black parents and their children to develop a bleak view of their future prospects. Siblings, relatives and other adults who have left school may have had difficulty getting jobs. Students may have experienced barriers to securing summer and part time jobs, watched or participated in demonstrations for more jobs, or read reports in the media discussing the problem of black unemployment. It is this, Ogbu argues, that leads them to adopt alternative strategies to defend against “failure” in mainstream schools and jobs. If blacks live in communities where the majority white population dominates, as in Britain, it seems likely that a framework of oppositional culture influenced by the experiences of other blacks around them is less likely to apply. In addition to this, intermarriage is more common in Britain than in America, particularly between blacks and whites (Model and Fisher 2002). This may further mitigate the tendency for Afro-Caribbeans in Britain to adopt an oppositional culture.

8

Although it could be argued that this label might have been applicable to Catholics in Northern Ireland, who attained involuntary minority status through invasion

29

In contrast to Ogbu, it has been suggested that the high concentration of a given ethnic group may have significant benefits (Portes and Zhou 2001 [1993]). Those with strong ethnic networks, access to capital, and limited ties to disaffected youth from the host country experience a “linear” ethnicity that creates networks of social ties and may provide access to job opportunities while reinforcing the authority and values of parents. In view of the relatively high concentration of South Asian groups in Britain, particularly Pakistanis and Bangladeshis, this is certainly a path to pursue.

30

Appendix Cases were matched on gender, parental income and regional health authority where possible. In cases where one of these variables was not available, cases were matched on the variables present in the dataset. Gender Annual Parental Income9 F 8 F 4 M 8 M 3 M 1 F missing M missing F 1 M 3 M 1 M 1 F 3 M 1 F 3 F 8 F 2 F 1 F 2 M 3 F 2

Regional Health Authority 5 5 6 3 3 6 8 9 14 3 6 10 4 5 6 14 5 2 2 2

F M F F F F M M F F M F M

2 1 1 1 missing 8 9 8 2 8 missing 1 1

2 11 11 6 12 9 3 6 2 8 12 12 11

F F F M M

1 2 2 9 2

12 12 12 6 6

Serial number White AfroIndia Pakistani/ Caribbean n Bangladeshi 3088 3932 443 917 930 232 354 45013 30271 80068 12106 407 201 454 11190 41 516 272 5276 7001 8 30041 393 489 9433 13 614 12231 61783 15760 9200 338 20 1271 7 4066 19 292 14950 61

243 252 1135 1342 1357 1657 1932 2276 2824 3242 3693 3699 3764 3950 4890 5173 5187 5277 5293 5302

7020 4243 80075 1365 1326 61014 70218 8738 30249 70140 15422 10469 3384 10442 8747 5872 80020 60764 8772 66

4509 1684 3295 12332 302 9419 4697 9003 5189 863 40009 11582 10444 70337 6439 4487 5430 5357 5366 7895

5314 6762 6846 6865 7051 7130 7441 7533 7928 8285 8707 8708 8770

1375 80103 2079 60190 4048 70226 40007 1412 10449 80064 259 8787 557

60772 12155 5801 12167 12188 3423 5080 70118 14961 60488 8719 60898 6760

8785 8799 9022 9105 9112

15441 8801 8739 10812 81

70336 60665 80084 12806 61004

9

1= under £5200, 2= £5200-£10399, 3= £10400-£15599, 4= £15600-£20799, 5= £20800 and above, 8= uncertain, 9= refused to answer.

31

M F M M

missing 9 4 1

6 8 8 8

M M F M M F F F F F M M F M F M M F F F F M M F M M

3 2 3 5 2 3 1 1 3 1 1 2 1 1 1 1 missing 8 1 3 8 8 1 missing missing 9

8 6 7 7 5 5 5 12 12 3 12 12 12 6 6 12 12 3 3 7 12 12 12 4 12 12

F F M M F M

2 missings 5 4 2 1

7 8 1 10 12 5

Table A1

445 9197 4806 1014 8 4667 2188 1107 9980 77 4877 14273 80003 2246 171 12 4019 5664 811 12676 312 4130 10183 3111 2554 299 9645 4036 1396 5663 1178 8 151 550 11039 11373 8710 30265

9113 9162 9171 9220

12092 10439 7006 8806

60707 3617 8355 5733

9249 9859 10222 10390 10441 10466 10810 11522 11550 11566 11594 11776 11783 12085 12090 12173 12359 12363 12397 13926 14927 14993 15020 15369 15398 30162

60899 8762 12108 3093 40034 1376 148 7043 8709 80063 8737 7045 7923 7037 80137 9680 7055 80081 80076 80112 7061 6997 6958 70291 30160 11576

60629 80127 60571 60806 70136 60019 80054 60692 6428 60832 7053 60669 60795 60220 60971 80083 80085 60833 6432 7916 8369 70008 80088 60078 80089 12806

60354 70047 70154 70182 70192 70254

80032 60999 60937 80111 8720 60985

80079 60216 80124 40019 80060 80130

Case matching

Recode

BCS

White

UK Irish Other European

West Indian/Guyana

West Indian/Guyana

Indian

Indian

32

Pakistani/Bangladeshi

Pakistani Bangladeshi

Mixed/other

Mixed/other

Table A2

Recoding of BCS ethnicity

33

Item Think homework is a bore

Original values 1 Very true 2 Partly true 3 Not true at all

Recoded values 1 True 0 Not true

Feel school is largely a 1 Very true waste of time 2 Partly true 3 Not true at all

1 True

Do not like school

1 Very true 2 Partly true 3 Not true at all

1 True

1 Very true 2 Partly true 3 Not true at all

1 True

Never take work seriously

0 Not true

0 Not true

0 Not true

How often get into trouble 1 More than others at school? 2 About same as others 3 Less than others 4 No one in class does 5 Other answer

1 More than others 0 Other answer

Removed from class due to 1 No need to remove behaviour problem? 2 Removed

1 Yes 0 No

Pupil ever suspended from 1 Yes school? 2 No

1 Yes 0 No

Do teachers pick on you 1Yes more than others? 2 No 3 No one in class does 4 Other answer

1 Yes 0 Other answer

Table A3

Recoding of anti-achievement and behaviour items

34

References Ainsworth-Darnell, J.W., Downey, D.B. 1998. ‘Assessing the oppositional culture explanation for racial/ethnic differences in school performance’. American Sociological Review 63:536-553. Aragon, J. 1973. ‘An impediment to cultural pluralism: culturally different educators attempting to teach culturally different children’. Pp. ?? in M. Stent, W. Hazard and H. Rivlin (eds.) Cultural Pluralism in Education. New York: Appleton-Century-Crofts. Berreman, G. 1967. ‘Caste in cross-cultural perspective: organisational components’. Pp. 275-307 in G. DeVos and H. Wagatsuma (eds.) Japan’s Invisible Race: Caste in Culture and Personality. Berkeley: University of California. Berreman, G. 1972. ‘Race, caste and other invidious distinctions in social stratification’. Race 23(4). Berthoud, R. 2000. ‘Ethnic employment penalties in Britain’. Journal of Ethnic and Migration Studies 26(3): 389-416. Blackaby, D., Leslie, D., Murphy, P. and O’Leary, N. 1999. ‘Unemployment among Britain’s ethnic minorities’. Manchester School 67(1): 1-20. Callender, C. 1995. ‘A question of “style”: black teachers and pupils in multi-ethnic schools’. Language and Education 9(3): 145-59. Coard, B. 1971. How the West Indian Child is Made Educationally Sub-Normal in the British School System. London: New Beacon Books. Cochran, W. 1965. ‘The planning of observational studies of human populations’. Journal of the Royal Statistical Society A 128: 234-255. Cochran, W. 1968. ‘The effectiveness of adjustment by subclassification in removing bias in observational studies’. Biometrics 24: 205-213. Cook, P.J. and Ludwig, J. 1998. ‘The burden of “acting White”: Do Black adolescents disparage academic achievement?’ Pp.375-400 in C. Jencks and M. Phillips (eds.) The Black-White Test Score Gap. Washington, DC: The Brookings Institution. Cross, W. 2003. ‘Tracing the historical origins of youth delinquency and violence: myths and realities about Black culture’. Journal of Social Issues 59(1): 67-82. Crow, J. 1969. ‘Genetic theories and influences: comments on the value of diversity’. Harvard Educational Review 39(2): 301-309. Deutsch, M. 1967. The Disadvantaged Child: Selected Papers of Martin Deutsch and His Associates. New York: Basic Books. Downey, D. and Ainsworth-Darnell, J. 2002. ‘The search for oppositional culture among black students’. American Sociological Review 67(1): 156-164.

35

Erickson, F. 1987. ‘Transformation and school success: the politics of culture of educational achievement.’ Anthropology and Education Quarterly 18: 336-355. Farkas, G., Lleras, C. and Maczuga, S. 2002. ‘Comment: does oppositional culture exist in minority poverty peer groups?’ American Sociological Review 67(1):148-155. Fischer, C., Hout, M., Sanchez Jankwoski, M., Lucas, S., Swidler, A and Voss, K. 1996. Inequality by Design: Cracking the Bell Curve Myth. Princeton, NJ: Princeton University Press. Fordham, S. and Ogbu, J.U. 1986. ‘Black students’ success: coping with the “burden of acting White”’. Urban Review 18: 176-206. Fries-Britt, S. 1998. ‘Moving beyond Black achiever isolation: experiences of gifted black collegians. The Journal of Higher Education 69: 556-576. Graham, S., Taylor, A. and Hudley, C. 1998. ‘Exploring achievement values among ethnic minority early adolescents’. Journal of Educational Psychology 90: 606-620. Herrnstein, R. 1971. ‘IQ’. The Atlantic Monthly. September 1971: 43-64. Herrnstein, R. and Murray, C. 1994. The Bell Curve: Intelligence and Class Structure in American Life. New York: Free Press. Holzer, H.J. 1991. ‘The spatial mismatch hypothesis: what evidence has it shown?’ Urban Studies 28: 105-122. Hunt, J. 1967. ‘Environment, development and scholastic achievement’. Pp. 293-336 in M. Deutsch, I. Katz and A. Jensen (eds.) Social Class, Race and Psychological Development. New York: Holt. Hunt, J. 1969. ‘Has compensatory education failed? Has it been tried?’ Harvard Educational Review 39: 278-300. Inkeles, A. 1966. ‘A note on social structure and the socialization of competence’. Harvard Educational Review 36: 265-283. Jaynes, G. and Williams, R. 1989. A Common Destiny: Blacks and American Society. Washington, DC: National Academy Press. Jensen, A. 1969. ‘How much can we boost IQ and scholastic achievement’. Harvard Educational Review 39(1): 1-122. Johnston, R., Forrest, J. and Poulsen, M. 2002. ‘The ethnic geography of ethnicities: the “American Model” and residential concentration in London’. Ethnicities 2(2): 209-235. Kochman, T. 1981. Black and White Styles in Conflict. Chicago: University of Chicago Press. Lacy, M. 1997. ‘Efficiently studying rare events: case-control methods for sociologists’. Sociological Perspectives 40(1): 129-154.

36

Layton-Henry, Z. 2002. ‘Transnational communities, citizenship and African-Caribbeans in Birmingham’. Working Paper WPTC-02-07, Institute of Social and Cultural Anthropology, University of Oxford. Lindley, J. 2002. ‘Race or religion? The impact of religion on the employment and earnings of Britain’s ethnic communities’. Journal of Ethnic and Migration Studies 28(3): 427-442. MacLeod, J. 1987. Ain’t No Makin’ It: Aspirations and Attainment in a Low Income Neighborhood. Boulder, CO: Westview. MacLeod, J. 1995. Ain’t No Makin’ It: Aspirations and Attainment in a Low Income Neighborhood. Boulder, CO: Westview. Majors, R. and Billson, J. 1992. Cool Pose: The Dilemmas of Black Manhood in America. New York: Lexington Books. Massey, D. and Denton, N. 1987. ‘Trends in residential segregation of blacks, Hispanics and Asians’. American Sociological Review 52: 802-25. Massey, D. and Denton, N. 1989. ‘Hypersegregation in US Metropolitan areas: black and Hispanic segregation along five dimensions’. Demography 26: 373-93. Massey, D., Gross, A. and Shibuya, K. 1994. ‘Migration, segregation, and the geographic concentration of poverty’. American Sociological Review 59(3): 425-445. Mickelson, R. 1990. ‘The attitude-achievement paradox among black adolescents’. Sociology of Education 63: 44-61. Miller, L. 1995. An American Imperative: Accelerating Minority Educational Advancement. New Haven, CT: Yale University Press. Model, S. and Fisher, G. 2002. ‘Unions between blacks and whites: England and the US compared’. Ethnic and Racial Studies 25: 728-754. Model, S. and Lin, L. 2002. ‘The cost of not being Christian: Hindus, Sikhs and Muslims in Britain and Canada’. International Migration Review 36(4): 1061-1092. Modood, T. 1997. ‘Employment’. Pp. 83-149 in T. Modood, R. Berthoud, J. Lakey, J. Nazroo, P. Smith, S. Virdee and S. Beishon Ethnic Minorities in Britain. London: Policy Studies Institute. O’Connor, C. 1997. ‘Dispositions toward (collective) struggle and educational resilience in the inner city: a case analysis of six African-American High School Students’. American Educational Research Journal 34: 593-629. Ogbu, J.U. 1974. The Next Generation: An Ethnography of Education in an Urban Neighborhood. New York and London: Academic Press. Ogbu, J.U. 1978. Minority Education and Caste: The American System in Cross-Cultural Perspective. New York: Academic Press.

37

Ogbu, J.U. 1981a. ‘Education, clientage, and social mobility: caste and social change in the United States and Nigeria’. Pp. 277-306 In G. Berreman (ed.) Social Inequality: Comparative and Developmental Approaches. New York: Academic Press. Ogbu, J.U. 1986. ‘The consequences of the American caste system’. Pp. 19-56 in U. Neisser (ed.) The School Achievements of Minority Children: New Perspectives. Hillsdale, NJ: Erlbaum. Ogbu, J.U. 1987. ‘Variability in minority school performance: a problem in search of an explanation’. Anthropology and Education Quarterly 18: 312-334. Ogbu, J.U. 1991a. ‘Low Performance as an Adaptation: The Case of Blacks in Stockton, California.’ In M.A. Gibson and J.U. Ogbu (eds) Minority Status and Schooling: A Comparative Study of Immigrant and Involuntary Minorities. New York: Grand Publishing . Ogbu, J. 1991b. ‘Minority coping responses and school experiences’. The Journal of Psychohistory 18: 443-56. Ogbu, J.U. 1992. ‘Understanding cultural diversity and learning’. Educational Researcher 21(8): 5-14. Ogbu, J.U. 1997a. ‘Racial Stratification and Education in the United States: Why Inequality Persists.’ Pp. 765-778 in A.H. Halsey, H. Lauder, P. Brown and A. Stuart Wells (eds) Education: Culture, Economy, Society Oxford: Oxford University Press. Ogbu, J.U. 1997b. ‘Understanding the school performance of urban blacks: some essential background knowledge’. Pp. 190-221 in H. Walberg, O. Reyes and R. Weissberg (eds.) Children and Youth: Interdisciplinary Perspectives. London: Sage. Ogbu, J.U. 1999. ‘A cultural context of children’s development’. Pp. 73-92 in H. Fitzgerald, B. Lister and B. Zuckerman (eds.) Children of Color: Research, Health, and Policy Issues. New York: Garland. Ogbu, J.U. 2003. Black American Students in an Affluent Suburb: A Study of Academic Disengagement. Mahway, NJ: Lawrence Erlbaum Associates Publishers. Owen, D., Green, A., Pitcher, J. and Maguire, M. 2000. Minority Ethnic Participation and Achievements in Education, Training and the Labour Market. Nottingham: DfEE Publications. Peach, C. 1996. ‘Does Britain have ghettos?’ Transactions of the Institute of British Geographers 21: 216-235. Peach, C. 1998. ‘South Asian and Caribbean ethnic minority housing choice in Britain’. Urban Studies 35(10): 1657-1680. Portes, A. and Zhou, M. 2001 [1993] ‘The new second generation: segmented assimilation and its variants’. Pp. 597-608 in: D.B. Grusky (Ed) Social Stratification: Class, Race and Gender. Oxford: Westview Press. Pumphrey, P.D. and Verna, G.K. 1990. Race Relations and Urban Education. London: Falmer Press.

38

Richmond, A. 1993. ‘Reactive migration: sociological perspectives on refugee movements’. Journal of Refugee Studies 6(1): 7-24. Riessman, F. 1962. The Culturally Deprived Child. New York: Harper. Rosenbaum, P. and Rubin, D. 1983. ‘The central role of the propensity score in observational studies for causal effects’. Biometrika 70: 41-55. Rosenthal, R. and Jacobson, L. 1967. ‘Self-fulfilling prophecies in the classroom: teachers’ expectations as unintended determinants of pupils’ intellectual competence’. Pp. 219-253 in M. Deutsch, I. Katz and A. Jensen (eds.) Social Class, Race and Psychological Development. New York: Holt. Rosenthal, R. and Jacobson, L. 1968. Pygmalion in the Classroom. New York: Holt. Trueba, H. 1988. ‘Culturally based explanations of minority students’ academic achievement.’ Anthropology and Education Quarterly 19(3): 270-287. Western, J. 1992. A Passage to England. London: University College of London Press. Willis, P.E. 1977. Learning to Labour. Farnborough: Saxon House Wilson, W.J. 1987. The Truly Disadvantaged. Chicago: University of Chicago Press

39