Measuring and trying to understand the reasons for changes in school segregation in England is central to the evaluation of policies designed to increase choice and competition both in and since the 1988 Education Reform Act. Many argued these policies would have unintended sorting consequences in terms of changes in the distribution of pupils from low income families across schools.
The central hypothesis is that greater school choice will lead to higher socio-economic status (SES) parents being more successful than those from lower socio-economic groups in choosing the higher performing schools, either because higher SES parents are more active in exercising choice or because some schools ‘cream-skim’ the higher ability and easier to teach pupils. This will cause these high performing schools to improve still further due to positive peer effects from their advantaged intake. This so called virtuous cycle would, it is suggested, lead to increasing polarisation between schools in terms of the ability and socio-economic background of their intakes.
However, this argument for increasing polarisation of English schools needs to be set against the alternative proposition that the principal beneficiaries of a school choice system would be those who were previously the most constrained by a neighbourhood allocation system, i.e. families on low incomes who were unable to move house to access the best schools.
Whether this increased polarisation is actually happening is an empirical question and a sizeable body of evidence has been accumulated on this issue. Until relatively recently in England, the only data collected on the demographic composition of schools was via an Annual Schools Census, where schools declared the number of pupils known to be eligible for free school meals (FSM). This was used as a proxy for social disadvantage to compare the distribution of pupils who were eligible for FSM, versus those who were not, across state schools.
The first major research programme using large-scale longitudinal quantitative data measured secondary school segregation in the years 1989 to 1999. It claimed to show that, contrary to popular opinion, school segregation actually fell in the period immediately following the Act and has risen slightly since then (Gorard et al. 2002, 2003).
Was the large fall in segregation between 1991 and 1993 attributable to choice policies ‘unlocking the iron cage’ of neighbourhood catchment areas, or is there another explanation for this fall? Finding an answer to this question is somewhat linked to the fierce (and continuing) debate about how best to measure school segregation. For example, Allen and Vignoles (2007) argue that Gorard’s choice of segregation index overstates the magnitude of the fall in segregation by 100%, compared to a relative index such as the Dissimilarity index. However, on most indices segregation did fall over this two year period.
An important aspect of the pattern of the fall in segregation is that two years of falling segregation up to 1993 is followed by a rapid levelling out of the level of segregation. This particular pattern of data seems inconsistent with an explanation whereby a policy change produces a change in the sorting of pupils into secondary schools because secondary schools have five cohorts of pupils at any one time. Therefore, a one-off policy change should produce five years of falling segregation (at roughly the same rate) as the cohorts of pupils who entered the school under the previous regime gradually reach the age of 16 and leave the Census.
The large fall in FSM segregation between schools between 1991 and 1993 coincided with a serious recession in England, and most researchers now agree that this was responsible for the changes in segregation in schools. The problem with the FSM measure is that FSM eligibility is not a fixed attribute of the child, but changes as the economic circumstances of the family change. So, it is possible that segregation can change from one year to the next as pupils switch their status from NONFSM to FSM, and vice versa, even if there is no change in the actual composition of the school. A recession would produce falling school segregation, for example, if the incidence of unemployment disproportionately affected schools with low pre-existing unemployment levels.
All researchers, using very different measures of segregation, agree that secondary school segregation, measured using FSM, has been increasing very slowly from 1994 to 2004 (Allen and Vignoles, 2007; Goldstein and Noden, 2003; Noden, 2000). However, given the susceptibility of the FSM status indicator to changes in economic circumstances it is impossible for us to attribute this slight rise in school segregation to any education policy changes. Just as the recession lowered school segregation, it is possible that an improving economic environment raises segregation.
The new Pupil-Level Annual Schools Census makes it possible to overcome the problems with the FSM measure. Gibbons and Telhaj (2006) used this data to show that there was very little change in ability segregation between 1996 and 2002. Allen (2007) showed that school segregation is significantly higher today than it would be if all pupils were reallocated to their nearest school (via a simulation that mimics the implementation of catchment areas), but that grammar schools and voluntary-aided faith schools are responsible for much of this ‘post-residential sorting’: these school types existed in England prior to 1988. Both these studies lend weight to the argument that school segregation in England has changed very little recently.
So, should the English experience of school choice reforms be used to argue that choice is possible without increasing school segregation? This would be too strong a claim from the data we currently have since there are alternative explanations for the absence of a recent rise in segregation. Most importantly, it may be that de facto school choice did not in fact increase during this period. This would be the case for example, if school choice was already being exercised through parents’ choices of residential location and if capacity constraints prevented the further exercise of choice. An important part of this alternative explanation is the fact that the secondary school population rose significantly from a deep trough in 1988 to a peak in 2005. This meant that schools lower down the LEA league tables were protected from a serious deterioration in their intake as popular schools quickly became full to capacity and mostly chose not to expand. Legislation giving parents a say over which schools their child attended was in place in England as part of the 1944 and 1980 Education Acts, though choice was difficult in many parts of the country. With this in mind, it is even possible that the dip in the secondary school population in the late 1980s allowed parents of children entering secondary school in 1988 to have more genuine and achievable choice of secondary school than those in 2005!
Secondary school rolls in England are set to fall from 2005 onwards, thereby freeing up spare capacity. Thus, the era of schools admissions we are currently in may actually be the most interesting period to look at the effects of increased parental choice on school segregation.
Allen, R. (2007) ‘Allocating Pupils to their Nearest Secondary School: The Consequences for Social and Ability Stratification’. Urban Studies 44(4) pp. 751-770.
Allen, R. and Vignoles, A. (Nov 2007, forthcoming) ‘What should an index of school segregation measure?. Oxford Review of Education 33(5).
Gibbons, S., & Telhaj, S. (2006) Peer effects and pupil attainment: evidence from secondary school transition. Centre for the Economics of Education Discussion Paper.
Goldstein, H. & Noden, P. (2003) Modelling Social Segregation. Oxford Review of Education 29, pp. 225-237.
Gorard, S., Taylor, C. & Fitz, J. (2003) Schools, Markets and Choice Policies, London: RoutledgeFalmer
Gorard, S., Taylor, C. and Fitz, J. (2002) Markets in public policy: the case of the United Kingdom Education Reform Act 1988, International Studies in Sociology of Education, 12, pp. 23-41.
Noden, P. (2000) Rediscovering the impact of marketisation: dimensions of social segregation in England’s secondary schools, 1994-99, British Journal of Sociology of Education, 21, pp. 371-390.