In an edited letter (TES, May 3) Herman Ouseley argues strongly in favour of ethnic monitoring in schools. This proposal is not new. It was recommended 11 years ago in the Swann Report. Indeed, in the late 1980s the Government appeared to support the idea, and Department of Education and Science's Circular 1689 required local education authorities and grant-maintained schools to collect information on pupils' ethnic backgrounds on entry.
This commitment was abandoned last year. The Department for Education and Employment has been extremely reticent about the reasons for the change, but it seems the data has been difficult to collect and returns from LEAs have been inconsistent and unreliable. I also suspect that certain ministers have been reluctant to become associated with a racial equality policy reminiscent of certain left-wing councils of the 1980s. However, continued evidence from sample surveys and LEAs of differentials in attainment between ethnic groups has raised again the question of whether data on ethnic origin should be routinely collected. It might be helpful, therefore, to review the arguments.
First, what benefits might such a policy bring? There is no doubt it would allow much more valid comparisons of the educational achievement levels of pupils from different ethnic groups. Currently it is difficult to judge the representativeness of information from sample surveys and individual LEAs, and therefore to form a broader national picture.
Large-scale data collection would also permit more differentiated comparisons, looking, for example, at ethnic groups which are numerically small, or at sub-groups within ethnic groups, such as gender or social class.
Data on a national scale would also facilitate more sophisticated comparisons, not only of average levels of achievement, but also of the distribution of individual achievements within groups. Such within-group differences are frequently neglected in the analysis of current data. If data were collected when pupils enter school it would also allow schools, LEAs and the DFEE to compare not only educational achievements, but also the educational progress of students from different ethnic groups.
Finally, ethnic data would enable administrators to identify the location of pupils from different ethnic groups more accurately, and to target resources should this be desired.
But there are also significant limitations in what can be done with ethnic data, and problems in its collection. Sometimes it is argued that the collection of ethnic data will enable the identification of racial discrimination in education . Unfortunately, this is not the case. Inequalities in educational outcomes (achievement or progress) between ethnic groups at school, LEA or national level may be the result of discrimination, but equally they may be the result of factors external to the school or system. We cannot assume racial discrimination from unequal outcomes.
Ethnic data tells us little about the causes of any inequality between groups, and thus little about what might be done to rectify it. To argue, as Herman Ouseley does, that where an inequality is discovered "resources can be directed and strategies for improvement implemented", is simplistic. This suggests that inequality is necessarily iniquitous, and that remedial resources or strategies can easily be identified and uncontroversially put in place by policy-makers and teachers. If the continued presence, despite educational reform, of social class inequalities in education has taught us anything, it is that things are not so simple.
Another limitation in the use of ethnic data is that it tends to encourage the view that the education system is working as it should as long as the average achievement levels of pupils from different ethnic groups are the same. But this, too, may not be the case. Students from one ethnic group may achieve on a par with those in another even though they have been treated less favourably, because they have been able to overcome such treatment by commitment and hard work or as a result of a better ability profile.
Moreover, equal ethnic group outcomes may occur even though individuals and sub-groups within the broader group do badly. In short, broad ethnic group comparisons may mask significant inequities and we would be foolish therefore to rely on them.
Another problem relates to people's perceptions of the collection of ethnic data. Despite the efforts of the Commission for Racial Equality and some LEAs to convince parents and teachers that such data will be used benignly, some remain suspicious. Ethnic minority parents sometimes feel that the information may be used to stereotype and discriminate against their children.
Some white parents are also sceptical because they think that information may be used to favour ethnic minority children. Some teachers are wary because they feel (correctly, in my view) that placing children into broad ethnic groups yields little additional information to enable them to cater more effectively for individual pupils' needs. These worries mean that the collection of ethnic data at school level is by no means straightforward or uncontroversial.
It also has costs. Reliable and consistent systems must be developed; they must be explained to schools and parents; the data must be requested from parents; if forthcoming it must be compiled by schools and supplied to LEAs, the DFEE and perhaps examination boards; these bodies in turn will have to conduct analyses, and supply composite information to policy-makers and practitioners.
Whilst these costs are not enormous we must decide whether they are justified given the benefits which ethnic data could bring. In my view the answer is not so clear-cut as Herman Ouseley suggests.
Peter Foster is senior lecturer in education at Manchester Metropolitan University.