Help or Hindrance? Higher Education and the Route to Ethnic Equality

Upward social mobility has been evident among British ethnic minority communities since the 1960s, and education appears to have had a key role in this process. Despite this, social scientists have been slow to consider the link between education, ethnicity and social stratiŽcation. The role of higher education has been particularly neglected. Although there has been some suggestion of an ethnic bias in the allocation of university places, previous work in this area has been limited by the nature of the data that have been available and by the types of analysis that have been conducted. This article includes detailed consideration of the key stages of the university application procedure, and particular attention is given to the role of candidates’ predicted and actual A-level grades. Although young people from ethnic minority backgrounds are admitted into university in large numbers, it is suggested that higher education has an ambivalent role in relation to ethnic equality. Institutional biases mean that ethnic minority candidates are Žltered into the new university sector, and it is concluded that biases in education and the labour market combine to create a cumulative pattern of ethnic disadvantage. Introduction and Background Debates about ethnic inequality and disadvantage have historically focused on employment and the labour market. While this clearly re ects the importance of occupation as ‘a signiŽ cant attribute in all the dimensions of stratiŽ cation, [which] possesses connotations of power and prestige relationships’ (Kelsall et al., 1972, p. 18), it has tended to mean that other potentially important areas of inquiry have been neglected. Relatively little attention has, for example, been given to the link between education, ethnicity and social stratiŽ cation. This is a particularly important gap given the strong ideological and empirical links that exist between occupational status and education in industrial societies. Ideologically, occupational status is tied to education by the notion of meritocracy. This concept is often used to justify social stratiŽ cation on the basis that individuals’ positions within society are determined by merit (often deŽ ned in terms of educational attainment) rather than ascribed social characteristics (such as ethnicity). Empirically, the ISSN 0142-5692 (print)/ISSN 1465-3346 (online)/02/020209-24 Ó 2002 Taylor & Francis Ltd DOI: 10.1080/0142569022013772 9 D ow nl oa de d by [ U ni ve rs ity o f W ar w ic k] a t 1 4: 43 3 1 M ay 2 01 2 210 M. Shiner & T. Modood link between education and occupational attainment in ‘advanced industrial’ countries is relatively close (Cheng & Heath, 1993, p. 152). In such societies, higher education is often viewed as a ‘stepping stone to higher level occupations’ (Cheng & Heath, 1993, p. 151) and graduates enjoy semi-elite status in the form of high incomes and access to high-status professions (Kelsall et al., 1972; Dolton et al., 1990). The notion of a meritocracy is evident in the suggestion that some minority groups are consciously using higher education to alter their own class composition. An ‘ethnic minority drive for qualiŽ cations’ has been attributed to a certain ‘mentality’ associated with economic migrants that includes an over-riding ambition to better oneself and one’s family (Modood, 1993, 1998; Modood et al., 1997). Such is the strength of this drive that, while ethnic minority communities account for 8% of 18–24 year olds in Britain, they make up almost twice this proportion of university entrants. This level of representation confounds general social-class patterns as it is achieved from a situation of relative disadvantage. Thus, for example, while two-thirds of white university entrants are from non-manual backgrounds, this compares with slightly more than one-third of Pakistanis and Bangladeshis (Ballard, 1999). This, in part, re ects the extent to which working-class ethnic minority groups achieve better examination results than their white working-class peers (Modood, 1993). Although education may provide the basis for upward social mobility and has considerable potential as a force for increasing ethnic equality, there is nothing inevitable about this. Thus, for example, Cheng & Heath (1993, p. 152) have suggested that education may simply serve to reinforce broader patterns of social inequality: at each stage of their educational and occupational career the members of some ethnic minorities might experience discrimination leading to a cumulative pattern of disadvantage. The analysis presented in this article is speciŽ cally concerned with entry into higher education, as this constitutes a key moment in many people’s educational careers and provides the foundations for access to well-paid, high-status occupations. Relatively little attention has been given to issues of racism and ethnicity in higher education. Until recently very little data had been published in this area, and a comprehensive process of ethnic monitoring was only introduced during the late 1980s (Modood, 1993, p. 167). According to Law (1996, p. 179), the ‘belated’ nature of this focus re ects ‘the insularity of universities from local intervention, the myths of academic liberalism, hostility to prescription and arrogance in the face of inequality’. Before reviewing the research evidence in this area, it is important to be clear about the process by which higher education places are allocated. Applications to university typically involve the following stages. (i) Candidates make up to six initial applications through the Universities and Colleges Admissions Service (UCAS) [1]. (ii) Institutions decide whether or not to make an ‘initial’ offer. Typically, at this stage, applicants have not completed their A levels and offers are based on predicted results (as estimated by teachers) and are conditional on candidates gaining certain grades. (iii) Candidates may select one offer as a ‘Ž rm’ offer and another as an ‘insurance’ offer. (iv) These offers are automatically conŽ rmed if the conditions are fulŽ lled and, while candidates are committed to accepting them, Ž rm offers over-ride insurance offers. If a candidate does not meet the conditions of an offer, their application may be D ow nl oa de d by [ U ni ve rs ity o f W ar w ic k] a t 1 4: 43 3 1 M ay 2 01 2 Higher Education and Ethnic Equality 211 rejected. Even in these circumstances, however, an offer may still be conŽ rmed and, even if it is not, the institution may offer a place on a different course. (v) Candidates who fail to gain a place through the main application procedure may do so subsequently through a process known as clearing. Existing research has highlighted the informal nature of procedures by which applicants are admitted into higher education. A case study of 10 degree schemes drawn from a range of faculties at the University of Leeds identiŽ ed a set of widely differing practices and subjective perceptions that had signiŽ cant implications for ethnic minority applicants (Robinson et al., 1992). This study was based on quantitative and qualitative data, and highlighted the considerable scope that exists for individual ofŽ cers to exercise discretion and the ‘colour-blind’ nature of admissions procedures. Discretion is limited by a range of factors including the balance between supply and demand, departmental rules, and agreed criteria relating to candidates’ qualiŽ cations and grades. Despite this, a recent reviewer noted that: ‘The impression is often of admissions as a rather private process, where staff handle business using whatever methods meet immediate needs’ (Law, 1996, p. 184). The Leeds case study found that, even in the same department, admissions tutors had quite different and often contradictory judgements about how to assess factors such as age, social background and re-sits, and were given very little guidance by their departments. It also highlighted the way in which admissions tutors drew on ‘soft’ data on a range of non-academic issues including applicants’ pastimes, ‘articulacy’ and character (Robinson et al., 1992). In the current context, the role of discretion is particularly important because of the suggestion that where there is scope for subjective assessment in higher education, bias against some or all ethnic minority groups is a likely outcome (see, for example, Esmail & Dewart, 1998). The Leeds case study also revealed a striking absence of departmental policies relating to ethnicity. Little, if any, consideration had been given to targets, quotas and ethnic monitoring. Furthermore, while attempts had been made to make publicity more attractive to women, no such efforts had been made to attract ethnic minority applicants. While commentators have highlighted the ‘colour-blind’ nature of the admissions process, they have also noted that such an approach places tremendous faith on a wide range of unmonitored discretionary evaluations by individuals acting with little external guidance (Robinson et al., 1992; Law, 1996). That such faith may be misplaced is suggested by a growing body of, primarily statistical, evidence. It has already been noted that comprehensive ethnic monitoring of applications and admissions to higher education was introduced during the late 1980s. Although the results of this exercise were quickly used to refute the long-standing claim of ethnic minority under-representation, they also revealed important differences between groups and types of institution (Modood, 1993). Compared with the general population, ethnic minority groups were over-represented within new universities [2]. They were, however, less well represented in old universities where evidence of black Caribbean, Bangladeshi and Pakistani under-representation led Modood (1993) to suggest that there was a deŽ nite ethnic hierarchy within this sector. Data from subsequent years conŽ rmed this pattern (Modood, 1998). Although these patterns of ethnic differences are important, they do not necessarily constitute evidence of discrimination. They may, for example, simply re ect differences between candidates that may be regarded as providing a legitimate basis for selection. Thus, for example, having noted that minority candidates tend to gain lower average A-level scores than whites, the Universities Central Council on Admissions (UCCA) went D ow nl oa de d by [ U ni ve rs ity o f W ar w ic k] a t 1 4: 43 3 1 M ay 2 01 2 212 M. Shiner & T. Modood on to highlight a range of factors that might offer some explanation for ‘apparent’ ethnic differences in rates of admission (Universities Central Council on Admissions, 1991, 1993). (a) Applicants from minority groups are more likely to apply for subjects with high entrance requirements, such as medicine and law, and less likely to apply for subjects like teacher training that have low entrance requirements. (b) By favouring institutions in their home region to a greater extent than white applicants, those from ethnic minority groups limit their choice and may compromise their chances of securing a place [3]. (c) Selectors tend to give less weight to qualiŽ cations obtained after more than one sitting. This particularly affects minority applicants, as they are more likely than whites to have re-taken one or more subject. In those studies that have taken account of such factors, however, ethnic differences have persisted and this has strengthened the suggestion that some groups are discriminated against in the way that university places are allocated. A study of medical schools found that applicants from ethnic minority groups were 1.46 times less likely to be accepted even when qualiŽ cations and other factors were taken into account (McManus et al., 1995). High predicted grades were given less weight for ethnic minority candidates than for whites, and particularly low rates of success were evident in relation to candidates with ‘non-European surnames’, thus pointing towards direct discrimination (see also McManus, 1998): Having a European surname predicted acceptance better than ethnic origin itself, implying direct discrimination rather than disadvantage secondary to other possible differences between white and non-white applicants. (McManus et al., 1995, p. 496) Similarly, Modood & Shiner (1994) showed that, although the factors highlighted by UCCA are important, they do not wholly explain ethnic differences in admissions. This work also conŽ rmed the importance of distinctions between minority groups and between types of institution. Even when a range of academic and socio-demographic differences had been allowed for, black Caribbean and Pakistani applicants were less likely than whites to have gained admission to an old university, although Chinese candidates and those classiŽ ed as Asian other were more likely to have done so. Black Africans, Black ‘Others’, Indians, Bangladeshis and those classiŽ ed as being Other were no more or less likely than whites to have gained admission to an old university. A very different pattern was evident in relation to new universities: black Caribbeans and Indians were more likely than whites to have gained admission to such institutions, although Bangladeshis, Chinese and those classiŽ ed as Asian ‘other’ were less likely to have done so. Black Africans, Pakistanis and those classiŽ ed as ‘other’ were no more or less likely than whites to have been admitted to a new university. While there has been growing academic interest in the possible role of racial bias in the allocation of higher education places, a small number of studies have started to consider the experiences of ethnic minority students once they start to study at university. These studies have highlighted ways in which the experiences of ethnic minority students differ from those of whites and are, in some respects, shaped by racism. A recent qualitative study noted that some ethnic minority students reported insensitive comments from staff that made them feel different and unwanted (Acland & Azmi, 1998). Another study found that ethnic minority students felt alienated from aspects of, what they D ow nl oa de d by [ U ni ve rs ity o f W ar w ic k] a t 1 4: 43 3 1 M ay 2 01 2 Higher Education and Ethnic Equality 213 perceived to be, a ‘white’ syllabus, and complained of the lack of attention given to issues of racism and the achievements of ‘black’ people [4] (Allen, 1998). Further criticisms from students have been identiŽ ed in relation to the under-representation of ethnic minority academic staff (Carter et al., 1999). The possible role of racial bias in assessment procedures has also emerged as an important cause for concern within higher education. This possibility was considered explicitly by the Barrow Inquiry into Equal Opportunities at the Inns of Court School of Law (Barrow et al., 1994). More recently, written examinations have been found to yield high scores for Asian students and low scores for Caribbean students in a London University (van Dyke, 1998), and research at Manchester University Medical School has suggested that racial bias in face-to-face clinical assessments may help to explain the extremely high failure rate of Asian Ž nalists (Esmail & Dewart, 1998). Data and Methodology Previous attempts to identify the possible role of racial bias in the allocation of higher education places have been limited in a number of important ways. They have often focused on a narrow range of courses offered at a small number of institutions, have failed to take account of other factors that may help to explain success and/or have focused on admissions rather than offers. Admissions are less appropriate than offers as the basis for assessing discrimination because they con ate the decisions taken by institutions with those taken by candidates. As such, differences in patterns of admission may re ect the decisions made by candidates rather than institutions: it may be, for example, that applicants from some ethnic minority groups favour new universities over old universities. Methodologically, our analysis compares favourably with previous work in this area. It was based on a representative sample of applicants drawn from the full range of courses offered by universities in the UK; it took account of a range of factors that have been put forward in attempts to explain ethnic differences in rates of admission; and it focused on offers rather than admissions. As such, we were able to isolate the decisions taken by institutions from those made by candidates. Furthermore, in contrast to previous work in this area, we were also able to consider the role of predicted grades in the allocation of places. Applications to university for the academic year 1996–97 provided the basis for analysis. We were speciŽ cally concerned with the conventional route into higher education and thus focused on applications made by candidates who were 20 years old or younger, who were resident in the UK and for whom A-levels constituted their main qualiŽ cation. UCAS provided detailed information about the social/demographic characteristics, academic performance (actual and predicted) and applications of 7383 candidates who fulŽ lled these criteria. In addition, for each course provided at each institution, it provided the following information: (i) the number of initial applications received; (ii) the total A-level points gained by applicants; (iii) the number of admissions, including those resulting from clearing; and (iv) the total A-level points gained by admitted candidates. [5] Our sample of candidates was randomly selected although it was constructed in such a way as to provide approximately equal numbers of white, Black Caribbean, Black African, Indian, Pakistani, Bangladeshi and Chinese candidates [6] (around 1000 candidates were included from each group). The number of candidates with relatively poor A-level grades was disproportionately large for some minority groups and care was D ow nl oa de d by [ U ni ve rs ity o f W ar w ic k] a t 1 4: 43 3 1 M ay 2 01 2 214 M. Shiner & T. Modood thus taken to include a sufŽ cient number of similarly qualiŽ ed whites to permit meaningful comparisons. A system of weighting was developed to correct for the differential sampling fractions that were used and, while statistical signiŽ cance was assessed on the basis of unweighted data (using the 0.01 cut off), percentages and averages were estimated on the basis of weighted data (Skinner, unpublished, 1994) [7]. Much of our analysis rested on statistical tests that assume cases are independent of one another. This assumption was potentially problematic in relation to initial applications. While candidates may make up to six initial applications, those made by the same candidate may not be considered to be independent of one another. Consequently, for the purposes of analysis, one initial application was selected at random for each candidate. The proŽ le of these selected applications was almost identical to those included in the overall sample. Discussion and Analysis While previous research has established that rates of admission into higher education vary between ethnic groups, the analysis described here focused on the extent to which these differences re ect bias in the allocation of places. In particular, it sought to (i) establish the extent to which differences in rates of admission are evident at earlier stages of the applications procedure; (ii) consider how patterns of success vary between old and new universities; (iii) identify key differences between ethnic groups, such as those relating to academic proŽ le and patterns of application, which may help to explain the different rates at which offers are made; and (iv) assess the degree to which such differences account for the rates at which ethnic groups successfully negotiate the various stages of the applications procedure.