Size matters: does school choice lead to 'spirals of decline'?
Stephen Gorard, Chris Taylor and John Fitz
School of Social Sciences
King Edward VII Avenue
CF10 3TW, UK
Paper presented at the British Educational Research Association Annual Conference, Cardiff University, September 7-10 2000
In a sense, both the advocates and the opponents of the greater use of market forces in compulsory education have predicted that some schools will enter what is termed here a 'spiral of decline'. This is a condition in which a school both loses market share and increases the proportion of socio-economic disadvantage in its intake. The spiral stems from the relationship between these two characteristics in a market driven by pupil-funding and raw-score performance indicators. As schools become more socially disadvantaged their 'league table' position declines, so more families prefer to use alternative schools. The losing school loses both numbers and probably more of the relatively socially advantaged families in its potential catchment, since the latter are deemed more likely to be the 'alert clients' using their powers of 'exit' (Hirschman 1970). This leads to even poorer league table results since there is a clear relationship at an aggregate level between socio-economic status and raw-score school outcomes, and therefore fewer pupils, smaller budget and so to a spiral of decline. Advocates of market forces see this as a temporary but probably necessary stage in systematic improvement based upon a mechanism where good schools will be popular and bad schools will either reform or eventually close through lack of numbers (Friedman and Friedman 1980). Opponents see this as a crucial component of their opposition to the concept of allowing families the freedom to choose schools for their children, since the system will penalise those who do not, or cannot, make 'good' choices. Our research set out to test this idea, and determine whether such declines have taken place in secondary schools in England and Wales and, if so, whether they were related to the process of school choice. The answer would appear to be that such declines do not occur, or at least no more so than they have always done. Schools have changed in size within limitations since 1944, and some schools have closed, merged, split and started up. There is no reason to believe that the pattern has changed as a result of the 1988 Education Reform Act, rather than being based more generally on demographic change for example. The reason why this is so, if one needs to provide a reason for something not occurring, appears to be that LEAs are preventing it. Different LEAs may use different mechanisms which have the effect of maintaining school rolls within certain boundaries, but most see this as a continued and valuable role for their school planning sections.
We have detailed elsewhere the policy provision in England and Wales designed to promote competition between schools, and so only a summary is given here. The 1988 Education Reform Act and associated legislation, among other objectives, set out to improve the social mix in schools and the standard of education provided in them. This was to be achieved by downgrading the notion of catchment areas, based all too often on residentially-segregated areas, allowing children from the most disadvantaged areas the choice of attending schools in the most advantaged areas. The policies of per capita funding, open enrolment, performance indicators (or benchmarks) and funding following pupils would supposedly drive up standards since poor schools would be unpopular. Poor schools would lose pupils until they either closed or improved. Good schools would be popular and would grow. Critics have suggested that such a model is unlikely to work. What is more likely is that those families who are already advantaged in educational terms will now be even more likely to gain places at desirable schools. Schools will therefore tend to become socially segregated, reinforcing existing polarisation in terms of raw-score results, leading to further loss of pupils in less successful schools, and so on. Two related factors would be chiefly to blame for this. Those families with knowledge of the system, confidence, leisure time and, above all, the ability to transport children to non-adjacent schools would be more likely to look for places in popular schools. Popular schools would be over-subscribed and in the allocation of contested places may, at least inadvertently, show preference to pupils likely to boost their raw-scores.
According much of the evidence presented in the UK, as elsewhere, the second scenario is more likely. Reay (1998), for example, claims that the 'market system of education provides the middle-classes with a competitive edge, of which they will increasingly take advantage' (p.1). Theoretical models have generally predicted a growth in social stratification between schools as a result of increased market forces in school placements (Bourdieu and Passeron 1992, Bowe et al. 1994). The findings of small-scale empirical studies of school choice in urban areas of England have reported finding evidence that supports these predictions (Blair 1994, Gewirtz et al. 1995), and the results from studies of school choice in England, Scotland, Israel and New Zealand have provided apparent confirmation (Willms and Echols 1992, Goldring 1995, Ambler 1997, Glatter et al. 1997, Waslander and Thrupp 1995, Woods et al. 1997, Lauder et al. 1999, also see Gorard 1999b for fuller discussion of this issue). Levacic and Hardman (1998) suggest that within a system of choice schools with high levels of students from poor families tend to lose student numbers, and therefore budget share, over time. Bagley and Woods (1998) report that families in their study were avoiding schools on the basis of current student intake characteristics such as race, religion and ability, suggesting that socio-economic segregation is linked to segregation in terms of other indicators as well. Hook (1999) describes how schools with low pass rates in examinations gain poor local reputations, which then have a strong deterrent effect on many residents. These so-called sink schools also have a high proportion of transient students, who may be both partly the cause and partly a symptom of the problem (Berki 1999). Families in these areas who have high aspirations therefore tend to move away (or use alternative schools), leading to a cycle of decline in inner cities, and an ever-increasing gap between the schools servicing the rich and those used by the poor. Worpole (1999) observes that the average length of trips to and from schools has increased from 2.1 to 2.7 miles over the last decade, and that the increasing use of family cars further exacerbates the educational divide between have and the have-nots. 'Schools which are left behind can get trapped into a vicious circle of decline' (p.17). In effect, social segregation between schools is increasing, leading some disadvantaged schools into a 'spiral of decline', and creating a clear system of winners and losers.
The recent media stories of high-profile government ministers, and even radical left-wing politicians, seeking to avoid their local comprehensive and using more distant grant-maintained (GM) schools can be seen as illustrations of this much larger trend. The policy of allowing Grant-maintained schools (as they then were) to opt out of LEA control has supposedly increased polarisation between institutions (TES 1999), simply because they exist or because they 'covertly select pupils by ability' according to Levacic and Hardman (1999). Much academic writing is therefore based on the social science 'fact' that markets in education have an increasingly stratifying impact on the makeup of schools (Conway 1997). Waslander and Thrupp state that 'those endowed with material and cultural capital will simply add to their existing advantages through choice policies' (1995, p. 21). Similarly, Gipps (1993) states that 'the concept of market choice allows the articulate middle and educated classes to exert their privilege, whilst not appearing to' (p.35). Commenting on experiments from the USA (where 'choice' has generally involved schemes providing free places at private schools for poor students), Powers and Cookson (1999) suggest that 'perhaps the most consistent effect of market-driven choice programs across the studies ... is that choice programs tend to have the effect of increasing stratification to one degree or another within school districts' (p.109). According to a study from Exeter University, 'within local markets, the evidence is clear that high-performing schools both improve their GCSE performance fastest and draw to themselves the most socially-advantaged pupils' (in Budge 1999, p.3).
The ensuing movement away from particular schools has two suggested impacts. One is that the number of students on roll falls, leading to a decline in the level of resources that those schools obtain (Whitty et al. 1998). It has been estimated that 75% of funding under Local Management of Schools (LMS) is based on numbers on roll within a school (Congdon and McCallum 1992). In addition, the social mix of such schools becomes increasingly problematic such that the prevalence of less able students, i.e. 'at-risk' students (Tomlinson 1997) or the 'wrong' students (Lauder et al. 1999), places extra pressure on their already declining resources. Critics of the new education market argue that by giving parents the opportunity to state a preference for particular schools will throw some schools increasingly into spirals of decline.
On the other hand, some commentators see less evidence of change over time as a result of choice policies, and for two main reasons. First: because of the importance of geographical location and local factors in the implementation of any national policy (Herbert 2000). Second: because much of the work cited above has a missing comparator in that no data is provided from before the onset of choice policies. In England and Wales different social classes have long been substantially segregated from each other by residence, which made any attempt to create a good social mix in 'local' comprehensive schools very difficult (Dore and Flowerdew 1981), and the situation does not seem to be improving. In fact, residential segregation may itself be reinforced by the rising cost of property in desirable catchment areas leading to selection by postcode and the continuance of educational 'ghettoisation' (Association of Teachers and Lecturers 2000), the so-called "Belfast model' (see Gorard 2000a). Advocates of increased school choice have suggested choice as a partial antidote to this reinforcing cycle of residential segregation, and there is some, albeit limited, evidence that this is possible. For example, Parsons et al. (2000) found that while there has been a progressive rise in the use of schools further away from home, this has not had quite the polarising effect predicted above. Out-of-catchment schools have been chosen by more children from 'struggling' neighbourhoods than 'prosperous' ones, and this is likely to reflect a greater dissatisfaction with their local school among those living in poorer areas (see also Gorard and Fitz 1998a). Since they are using post-code data Parsons et al. have no way of pursuing this matter, but it is possible that those leaving their struggling catchment areas to go to a different school represent a somehow privileged sub-group among those living in poorer areas. Whether this is in fact so is one of the things this paper sets out to discover.
Stillman (1990) suggests that there has not been much increase in active choice of schools or use of out-of-catchment schools since the reforms of the 1980s. This may be partly due to the number of school closures stemming from surplus places in the system (see below), and partly because some significant elements of choice already existed. For example, Education Ministry Circular 83 (1946) stated that Section 76 of the 1994 Education Act which permitted families to express a preference for a school, and to appeal against non-placement, was not limited to choice on religious (Anglican or Catholic) grounds. As the use of school allocation procedures involving an 11+ examination declined from 1968 to 1977, so the number of LEAs allocating places via choice schemes increased from 20% in 1968 to 27% in 1977, while the number using a catchment or feeder school system rose from 53% to 70% (Dore and Flowerdew 1981). According to Forrest (1996), in 1985 61% of LEAs operated catchment area systems, and 39% used a system of open preference. By 1996, despite the interim 1988 Education Reform Act, the number of LEAs using catchment areas had only dropped to 41%. This kind of evidence, of stability over time and the continued use of neighbourhood schools, suggest that contrary to the crisis account, spirals of decline will actually be relatively rare.
Despite the number of relatively small-scale, theoretical or qualitative studies cited, there has not been a recent definitive study of the impact of choice on the size of schools. 'The debate over school choice is rich in rhetoric but dismally poor when it comes to hard evidence' (Fuller 1996, p.11). Hardman and Levacic (1997) analysed the change in recruitment of 276 secondary schools from across six Local Education Authorities between 1990-91 and 1993-4. They found that of these 276 schools, 100 increased the size of their intake, 145 remained relatively constant and 31 saw a decrease in their intake. This, it was argued, 'suggests that the redistribution of the annual intake cohort amongst groups of competing schools reflects the differential popularity of those schools' (p. 123). This paper extends their analysis by examining the change in recruitment of secondary schools over an eleven-year period, from 1989 to 1999. It is almost assumed that with open enrolment, as long as there are enough surplus places in the education system, 'popular' schools will increase the size of their intakes year-on-year and those that are 'unpopular' will see a fall in numbers year-on-year. Popular schools can only keep increasing as long as there are places available in the school and generally therefore more popular schools will eventually become over-subscribed. It should be noted, however, that there are exceptions to this where schools have successfully admitted students over their planned admission number (PAN), and in Wales where the Popular Schools Initiative has allowed a few schools to expand in face of growing demand for their places. Unpopular schools will only see a fall in the size of their intakes as long as there are places available in other schools. If schools closures, rising birth-rates or other factors combine to keep school rolls high then, by definition, spirals of decline in terms of simple numbers cannot take place.
Consideration such as these lead us to five chief research questions about size matters:
• Have some schools consistently lost market share in the period since 1988?
• Are the changes noted larger than in previous periods?
• Are these same schools becoming relatively more socially disadvantaged in socio-economic terms?
• Are these schools therefore in a spiral of decline?
• Are spirals of decline caused by the process of school choice?
The data used in this paper comes from two large-scale datasets. The first is a database of the size and socio-economic composition of all secondary schools in England and Wales provided by the DfEE, the Welsh Office, and the LEAs themselves. It contains a record for every year from 1989-1999 of the number of pupils on roll, the number eligible for and taking free school meals, the number from each ethnic group, first language and special educational need. This data is used to decide whether schools are getting larger or smaller over time, and to document when they close, split, merge, or open. It is also used to calculate the proportion of 'disadvantaged' pupils in each school. The second dataset is a collection of interviews with a variety of LEA officials from the 40 LEAs selected as a sub-sample for more detailed study (the full list can be seen in White et al. 1999). The 30 of these LEAsbased in England are used here for the more detailed analyis. They are used to help explain the patterns of change in focus schools and the way in which LEA procedures for school allocation and these changes are related. Incidentally we generally preserve the anonymity of schools still in operation but feel that it is appropriate to use the proper names of schools that no longer exist and those that have already been publicly declared as being in Special Measures.
One of the first, and perhaps surprising, problems we had to face was how to measure the size of a school. The number of children in English secondary schools has increased from 2,944,722 in 1989 to 3,122,622 in 1999 (Figure 1). This increase has come alongside a reduction in the number of secondary schools in England; hence, the potential number of students in each school has steadily increased over the 11-year period from a mean of 730 pupils per school in 1989 to 876 pupils per school in 1999 (Figure 2).
Figure 1 Number of full-time secondary school pupils and number of secondary schools in England
Figure 2 Average number of pupils per secondary school in England
These average increases are clearly not evenly distributed across all LEAs, and Figure 3 shows the variation over 30 authorities in England (these 30 are all of the English LEAs in our stratified sub-sample, selected on the basis of location, size, poverty, segregation and school type). Changes in size cannot be assessed simply in raw figures, nor as percentage point differences for many of the reasons that we have discussed elsewhere (not least because urban and rural schools tend to be of different orders of magnitude, see also Gorard 1999a). In order to produce an estimate of the change in size in each area (or school) which is comparable between areas, we find the difference between the number on roll in 1989 and in 1999, and divide this by the original size of the school (Gorard 1999b). This yields a range of -4% to +78%. The largest increase in school sizes from this sample of LEAs was in Brent where the average school increased from 605 full-time students in 1989 to 1075 full-time students in 1999. A programme of school closures and amalgamations lieds behind this very large increase. For example, in 1989 there were 18 schools but by 1999 this had fallen to 13 schools even though the number of students in full-time secondary education actually increased from 10,891 students to 13,984 students in the 11-year period. In searching for spirals of decline it is therefore important to take these results into account. A decision to close schools on other grounds should not be confused with spirals of decline. Not all closed schools were in decline, and some schools only went into decline when closure was announced. Closing schools are generally phased out by age cohort so producing an ongoing drop in numbers. In areas such as the City of Bristol where the demographic trend has been down and where insufficient schools have been closed to reduce spare places, even a school facing a drop in numbers is not necessarily in decline. What we are searching for are schools which have no clear demographic reason for a drop in numbers, but which show such a drop in conjunction with an increasingly disadvantaged intake.
Figure 3 Change in average school intakes for a sample of 30 LEAs 1989-99
A related problem to the measurement of size is the assessment of relative disadvantage. We have written a considerable amount on this topic elsewhere (Gorard and Fitz 1998a, 1998b, Gorard and Fitz 2000a, 2000b, Gorard 2000b, 2000c, Taylor et al. 2000). The issue is not whether schools in potential spirals of decline are socially disadvantaged (they probably are), nor whether their percentage of children from families in poverty is increasing (as this is a national trend), but whether they are taking an increasing proportion of such disadvantaged pupils. In summary, we use here our segregation ratio for individual schools based upon eligibility for free school meals.
Because of the difficulty in interpreting changes in intake patterns without accurate knowledge of the underlying demographic changes affecting each school's intake, we also attempt to identify schools in 'spirals of decline' from a different perspective. We concentrate our search on two not wholly distinct groups of schools; those under Special Measures ('failing' schools as defined by OFSTED inspections) and those that have been closed during the last four years. These, along with those simply losing numbers over time, are likely candidates to be schools in a spiral of decline.
Changes in size
Figure 4 shows the proportionate change in individual school rolls between 1989 and 1999 for all English secondary schools. The degree of school size change reflects a normal distribution with most schools increasing between 0 and 10% over the 11 years, some by considerably more, and some declining in size. Only the 28% of schools showing a decrease in size, plus those closing or merging in the interim, are candidates for spirals of decline (although it should be noted that schools closed in the period do not appear in Figure 4). The majority of these experienced a fall in numbers of between 0 and 10%. There were a very few schools that saw much larger large falls in their school rolls, although Fryern Community College (Essex) declined by 63% and Kingsland Community College (Bedfordshire) fell by 43% for example. Both of these schools, and others, are considered in more detail below.
Figure 4 Frequency of school size changes 1989-99
There has been little research on school rolls prior to the expansion of market forces in education but one study by Butel (1988), who examined the changing rolls of 45 secondary schools in one LEA, calculated that the average school size in 1985-6 had fallen to 820 pupils from an average in 1978-9 of 861 pupils. This is an overall drop of 4.8%. Butel also analysed the differential impact of change across the LEA and observed a pattern of declining urban school intakes and growing rural school intakes. The magnitude of change in this one LEA before the introduction of the 1998 Education Reform Act makes it difficult to interpret many of the changes presented here throughout the 1990s as of great significance. From the 1970s to the 1980s the number on roll of most schools declined, leading to the surplus places that motivated at least part of the reforms in the 1980s. The decline was demographic, and as far as we can tell at this stage, unrelated to school allocation procedures. From the 1980s to the 1990s the number on roll of most schools has increased (by 20% from 1989-99). Again this has a demographic cause, exacerbated by the school closures stemming from the earlier decline.
There is considerable variation between LEAs in the relative size of their schools over time. Nearly all schools show some variation, and most of those with a lower number on roll in 1999 than in 1989 have intermediate years with a higher figure. The difficulty, then, lies in distinguishing a spiral of decline (in numbers) from the usual flux in school sizes both before and after 1988. In Hounslow LEA, for example, and several others no schools ended the period with fewer pupils (Figure 5), and given the average increase in school sizes this is not a very unusual outcome. Some schools show considerable growth over a short period, and others have a relatively flat 'trajectory'. Of course, those schools which remain similar in size have lost 'market share' to the growing schools, but in our operational definition of a school in decline it must be losing raw numbers, and therefore facing a lower level of per capita resources.
Figure 5 Trends in school rolls 1989-99, Hounslow LEA
The picture for Slough LEA is slightly different (Figure 6). Again there is variation between schools with some increasing at a faster rate than others, but there is also one school which ends the period with a smaller number on roll. This school dropped from 650 pupils in 1989 to under 500 by 1992. It would therefore be a candidate for a school in a terminal spin, losing both market share and raw numbers of pupils to other local schools. The timing is also interesting, taking place in the years immediately following the Education Reform Act 1988. It is important to note however, that the school is no longer in a terminal spin since it recovered in size to 570 by 1999. We shall return to this school for a consideration of changes in the character of its intake over the same period.
Figure 6 Trends in school rolls 1989-99, Slough LEA
A similar picture appears in both Reading, and Hammersmith and Fulham for example (see Figure 7). Reading LEA is particularly significant here as it lies at the opposite end of Figure 3 to Hounslow, meaning that growth of the average roll is quite small. Nevertheless, all except one school end the period with larger numbers than at the start. Even, the exception maintained numbers for several years after 1988, actually growing in size to above its 1989 base, before declined consistently from 1992.
Figure 7 Trends in school rolls 1989-99, Reading LEA
A very different picture emerges in Hillingdon LEA and a few others (Figure 8). Some schools increase in size and some decline, and the most likely reason is demographic since the total number of pupils has not grown as quickly as in Slough or Hounslow. Despite these differences none of the schools open in 1999 is obviously in a terminal spin. The range of school rolls in 1989 was from 200 to 1100, and by 1999 the range was from 400 to 1550. While individual schools have changed order in rank size, those schools surviving from 1989 to 1999 are around or above their original size. What has happened is that three schools have closed, and these were already three of the four smallest schools in 1989, and three schools have started out. The new schools are still much larger than those which closed, but it is interesting that two of them have dropped in size. The drop in size from 1989 in the closing schools was planned by the LEA, and stems from seeing out school careers for some existing pupils while not allowing further allocations to that school. Again, none of these patterns seems to fit that of a spiral of decline, but it is important to look also at the relative socio-economic composition of these schools (see below).
Figure 8 Trends in school rolls 1989-99, Hillingdon LEA
It is also interesting to observer what has happened in Bristol LEA (Figure 9), since Bristol was the only LEA with apparently falling rolls in Figure 3. Despite this most 'promising' of scenarios for the appearance of spirals of decline, very few candidates are apparent. The school with the largest consistent drop only opened in 1996. Only one other school shows a consistent drop over time. Most schools drop in numbers slightly in line with the overall demographic trend, while generally maintaining their market share.
Figure 9 Trends in school rolls 1989-99, Bristol LEA
There were 49 fewer secondary schools in England in 1999 than in 1995 (which is as far back as we have accurate identification for every single school - due to problems with reorganisation, loss of archives, official transcription errors, changes in DfEE numbering etc.). The combined effect of closing some schools, opening new schools and amalgamating others means that it is possible to identify a total of 90 school closures over the same time period. This represents just over 2% of the total number of secondary schools that were open in 1989. Around 40% of these schools faced a significant drop in size in the final year(s) before closure (Figure 10). Once an LEA has decided to close schools they often incrementally reduce a school's intake by not admitting a new year of pupils. The effect of this is to make the school fall in size irrespective of any other factors. However, our interviews with those responsible in the LEAs also show that any school considered for closure generally goes out for consultation with the local community as well as to Secretary of State, making closure a protracted process (Pollock 2000), and in many cases the closure does not take place as a result. Consequently, declines in recruitment in the year or two proceeding closure but after publication could be due to a lack of confidence in the school being open for the duration of the child's full schooling. While clearly an outcome of choice, the consequent drop does not seem to us to be a symptom of a spiral of decline. Although the school loses numbers and is closed, the notification of closure precedes the loss of numbers. What we are seeking, rather, is a school closed due to lack of numbers. The schools in Figure 8 do not meet this criterion. Five of the six schools in this sub-sample actually increased in size in the mid-1990s before their planned and phased closures. Only St Mark's Church of England School stands out from this analysis as a clear candidate for a spiral of decline, showing a year-on-year decline in numbers from 1989, although we shall also look in more detail at Margaret Danyers College. It is unfortunate that the DfEE are not able to provide us with the numbers on roll before 1989 since we do not know whether the school was already in spin before the ERA 1988, or whether we have at last identified a school in a market-driven spiral of decline.
Figure 10 - Trends in schools rolls of schools closed 1995-99
If post-1988 signalled the start of new trends in school recruitment then the relative size of schools in 1989 are likely to be in a different order of size by 1999. To make this comparison the size of schools in 1989 were correlated against their respective sizes in 1999 using Spearman's rank correlation. Across all the LEAs in this analysis the relative size of the schools in 1989 was related to their relative size in 1999 (R = 0.864), but this varied by LEA and in some LEAs nothing has changed (Figure 11). A similar conclusion was reached by Butel (1988) in an earlier study in one LEA, where the correlation between school sizes in 1978-9 and 1985-6 was R = 0.92. The size of schools in 1989 is significantly related to their relative sizes in 1999. Consequently, any change seen in their rolls between 1989 and 1999 might simply be an extension of demographic changes that began to affect school recruitment prior to marketisation. From, admittedly limited, previous studies the levels of change seen in the 1990s is not too dissimilar to changes seen in the early 1980s. It is also noteworthy that LEAs in which a significant number of school closures took place ended up with a lower level of socio-economic segregation than previously (see Gorard and Fitz 2000a).
Figure 11 Spearman's rank correlation of schools sizes 1989-99 by LEA
The notion of Special Measures was first introduced in the 1993 Education Act and reinforced in the School Inspections Act 1996. It is meant to highlight schools in which 'Special Measures are required to be taken in relation to a school if the schools is failing or likely to fail to give its pupils an acceptable standard of education' (Section 13,9). Riley (1996) claimed that schools deemed as 'failing' were affected by the measures introduced in the 1988 Education Reform Act, and argued that such schools tended to lose 'able' students to other schools, who in turn attempted to exclude 'troublesome' students from attending their school. The analysis here is based on two sets of schools placed under Special Measures by OFSTED and the DfEE Inspection Team from two different time periods. The first set of 20 secondary schools were those under Special Measures as at June 1996. The second set of 61 secondary schools were those under Special Measures further into the inspection cycle, as at August 1999.
Schools in Special Measures have both increased and decreased in size since their announcement, and the tendency to decrease was more marked in the first batch of 1996 (Figures 12 and 13 show the trends in school rolls for a sub-sample of such schools). This is because several schools in the first batch were closed down (including St Richard of Chichester and the Waterfield School). All comments made above about the procedure for closing schools apply here. The possibility that these schools are in a spiral of decline is considered below. In the second batch, no schools were closed but the greatest declines were observed in Abbotswood and Merryfield Schools, which were only started in the mid-1990s.
Figure 12 Trends in schools rolls of schools under Special Measures June 1996
Figure 13 Trends in schools rolls of schools under Special Measures August 1999
It should also be noted that the greatest fall in rolls, even among those schools falling consistently throughout this period, came at around the time of the OFSTED inspections, and particularly for schools that were eventually closed. This is complicated by the process of school closure, which would typically involve consultation with the local community and school governors, leading to a lack of confidence in these schools. Also, the gradual removal of pupils out of these schools inevitably makes these schools appear to lose allocated places where they are actually losing entire year cohorts. If the two closing schools are discounted for the present, the changes in size of schools in Special Measures is no different from changes in schools more generally. Only in a short period at the time of the OFSTED inspections does there appear to be any effect on school rolls. This suggests that the market may not lead to 'spirals of decline' but in fact that the announcement of the label of 'failing' is a more likely cause of falling numbers.
Any distinction between the earlier and later schools found 'failing' could also be explained by a shift in the nature of the OFSTED inspections. The later inspections overlapped with the second cycle of secondary school inspections such that, according the 1998-99 Annual Report of Her Majesty's Chief Inspector of Schools, Chris Woodhead, 'Secondary schools that were placed in special measures this year had made too little progress on the key issues from their first inspection'. Consequently, these schools were placed under Special Measures because of their lack of improvement rather than their failure to deliver an adequate education. This subtle shift in the balance of inspections is less likely to be related to falling rolls.
Changes in composition
Having identified a total of 12 schools as likely candidates to be in spirals of decline on the basis of size alone - whether falling rolls, closed schools, or those in special measures - we turn our attention to the second component of the spiral. Have the candidate schools also increased in the proportion of pupils with socio-economic disadvantage? According to the patterns shown in Figures 14 and 15 the answer is no. Of these 12 schools, half increase and half decrease their segregation ratios calculated for eligibility for free-school meals over the period, and many of these also show a 'roller-coaster' pattern of progress. Rather than spirals of decline, schools in trouble financially or in terms of academic reports appear to have volatile rolls. One school, Fryerns Community, stands out as the only school in our dataset with a profile of a spiral of decline - losing both numbers (and therefore budget) and having a growing proportion of pupils from families in poverty. Even with Fryerns, two caveats should be noted. The school showed a decrease in segregation from 1991 to 1992 and again from 1994 to 1998 which is hardly a symptom of a spiral of decline. Finally, finding one school from a close search of around 1,000 schools (a quarter of all secondary schools in England) which has some of the characteristics of this decline is less than convincing. There is no reason to assume that such rare patterns of change have not always occurred, and therefore no reason to attribute it to an increase in market forces.
Figure 14 - Segregation ratios (LEA weighted) of schools in decline, decreasing segregation
Figure 15 - Segregation ratios (LEA weighted) of schools in decline, increasing segregation
The role of the LEA
This paper has begun to outline the changes over time of school recruitment for all secondary schools in England between 1989, the introduction of open enrolment, and 1999, the most recent complete figures available. Our first observation is that the rise in the secondary school age cohort has meant that the actual number of students on roll has generally increased in most schools. This has been reinforced by a programme of school closure and amalgamation during the first half of the 1990s. The circumstances are therefore not propitious for schools to face a drop in numbers over time. Given the very limited number of schools actually decreasing in size, and the even smaller number of these who also increased their share of pupils facing socio-economic disadvantage, the notion of spirals of decline resulting from increased market forces since 1988 can be dismissed. There is no empirical evidence to support the notion.
Even where we focus only on those schools, such as 'failing' schools, likely to face such market pressures there is no evidence. Schools placed under Special Measures are more likely to have declining enrolment, but t is argued that such changes are closely related to the impact of the OFSTED inspections and not necessarily the effects of the market. Similarly, schools that were closed between 1995 and 1999 were not apparently in long-terrm decline before closure was proposed and implemented. Only with a long-term examination of school enrolment patterns, such as here, does a clearer understanding of the market upon intakes emerge.
Whatever the impact of the introduction of market principles into the provision of education in the UK has been, it is clear from this analysis that the role of the Local Education Authorities, OFSTED inspectors and natural population changes are key, if not paramount, in determining changes to school recruitment. Local Education Authorities (and even central government) still have a large influence on maintaining school numbers via a process of school closure, amalgamation or admissions arrangements. We hope that our ongoing research in 40 LEAs will provide a greater understanding of their current role in maintaining schools recruitment.
To some extent the lack of change as a result of introducing a policy of increased school choice is hardly surprising. Some observers had already suggested that this was a likely outcome (Levin and Riffel 1997, Gorard 1997), and others in the US are beginning to discover that competitive schemes such as the Cleveland Voucher Experiment (McGuinn and Hess 2000) and the Milwaukee Charter Scheme (Hess 2000) have produced no obvious reaction (other than symbolic ones by marketers and teachers' unions) and little discernible change in the administration or leadership of publicly-funded schools. It is easy to exagerate the significance of national policies in education, and important to retain a model in which such changes are heavily mediated by the actions of local agencies.
The data used here is part of an ESRC funded project (no. R000238031) entitled 'Measuring Markets' which studies the impact of open enrolment on schools intakes and performance. The authors would like to thank the DfEE, and all of the co-operating LEAs.
Ambler, J. (1997) Who benefits from educational choice? Some evidence from Europe, in Cohn, E. (Ed.) Market approaches to education, Oxford: Elsevier Science
Association of Teachers and Lecturers (2000) Social selection, Report, 22, 5
Bagley, C. and Woods, P. (1998) Rejecting schools: towards a fuller understanding of the process of parental choice, presentation at the BERA Annual Conference, Belfast
Berki, B. (1999) Parental choice in danger, Times Educational Supplement, 23/4/99, p. 2
Blair, M. (1994) Black Teachers, Black Students and Education Markets, Cambridge Journal of Education, 24, 2, 277-291
Bourdieu, P. and Passeron, C. (1992) Reproduction in Education, Society and Culture, London: Sage
Bowe, R., Gerwirtz, S. and Ball, S. (1994) Captured by the Discourse? Issues and concerns in researching 'parental choice', British Journal of Sociology of Education, 15, 1, 63-78
Budge, D. (1999) Gulf separating weak and strong increases, Times Educational Supplement, 30/4/99, p. 3
Butel , J. (1988) Falling school rolls in one LEA, Educational Studies, 14(2):176-186
Congdon, P. and McCallum, I. (1992) A demo-educational model for forecasting school rolls for localities, The Statistician, 41:573-590
Conway, S. (1997) The reproduction of exclusion and disadvantage: symbolic violence and social class inequalities in 'parental choice' of secondary education, Sociological Research On-line, 2, 4
Dore, C. and Flowerdew, R. (1981) Allocation procedures and the social composition of comprehensive schools, Manchester Geographer, 2, 1, 47-55
Forrest, K. (1996) Catchment 22, Education, March 1996, p.8
Friedman, M. and Friedman, R. (1980) Free to choose, Harmondsworth: Penguin
Fuller, B., Elmore, R. and Orfield, G. (1996) Policy-making in the dark, in Fuller, B. and Elmore, R. (Eds.) Who chooses? Who loses?, New York: Teachers College Press
Gewirtz, S., Ball, S. and Bowe, R. (1995) Markets, Choice and Equity in Education, Buckingham: Open University Press
Gipps, C. (1993) Policy-making and the use and misuse of evidence, pp. 31-44 in C. Chitty and B. Simon (Eds.) Education answers back, London: Lawrence and Wishart
Glatter, R., Woods, P. and Bagley, C. (1997) Diversity, differentiation and hierarchy. School choice and parental preferences, in: Glatter, R., Woods, P. and Bagley, C. (Eds.) Choice and Diversity in Schooling. Perspectives and Prospects, London: Routledge
Goldring, E. (1995) Communal Opportunities to Learn, Parental Involvement and School Choice: Israel and the United States, paper presented at ESRC/CEPAM Invitation Seminar, Milton Keynes 1995
Gorard, S. (1997) Market forces, choice and diversity in education: the early impact, Sociological Research Online, 2, 3
Gorard, S. (1999) Keeping a sense of proportion: the "politician's error" in analysing school outcomes, British Journal of Educational Studies , 47, 3, 235-246
Gorard, S. (1999a) Examining the paradox of achievement gaps, Social Research Update, 26, 1-4
Gorard, S. (2000a) Education and Social Justice, Cardiff: University of Wales Press
Gorard, S. (2000b) One of us cannot be wrong: the paradox of achievement gaps, British Journal of Sociology of Education, 21, 3
Gorard, S. (2000c) Questioning the crisis account: a review of evidence for increasing polarisation in schools, Educational Research, 42, 3
Gorard, S. and Fitz, J. (1998a) The more things change.... the missing impact of marketisation, British Journal of Sociology of Education, 19, 3, 365-376
Gorard, S. and Fitz, J. (1998b) Under starters orders: The established market, the Cardiff study and the Smithfield project, International Studies in Sociology of Education, 8, 3, 299-314
Gorard, S. and Fitz, J. (2000a) Investigating the determinants of segregation between schools, Research Papers in Education , 15, 2, 115-132
Gorard, S. and Fitz, J. (2000b) Markets and stratification: A view from England and Wales, Educational Policy , 14, 3
Hardman, J. and Levacic, R. (1997) The impact of competition on secondary schools, in Glatter, R., Woods, P. and Bagley, C. (Eds.) Choice and Diversity in Schooling, Routledge: London
Herbert, D. (2000) School choice in the local environment: headteachers as gatekeepers on an uneven playing field, School Leadership and Management, 20, 1, 79-98
Hess, F. (2000) Hints of the pick-axe: the impact of competition on public schooling in Milwaukee, presentation at AERA, New Orleans, April
Hirschman, A. (1970) Exit, Voice, and Loyalty: Responses to decline in firms, organizations, and states, Cambridge: Harvard University Press
Hook, S. (1999) Failing schools and dying cities, Times Educational Supplement, 7/5/99, p.5
Lauder, H., Hughes, D., Watson, S., Waslander, S., Thrupp, M., Strathdee, R., Simiyu, I., Dupuis, A., McGlinn, J. and Hamlin, J. (1999) Trading in futures: Why markets in education don't work, Buckingham: Open University Press
Levacic, R. and Hardman, J. (1998) Competing for resources: the impact of social disadvantage and other factors on English secondary schools' financial performance, Oxford Review of Education, 24, 3, 303-328
Levacic, R. and Hardman, J. (1999) The performance of grant-maintained schools in England: an experiment in autonomy, Journal of Education Policy , 14, 2, 185-212
Levin, B. and Riffel, J. (1997), School System Responses to External Change: implications for school choice, in Glatter, R., Woods, P. and Bagley, C. (Eds.) Choice and Diversity in Schooling. Perspectives and Prospects, London: Routledge
McGuinn, P. and Hess, F. (2000) Business as usual: the minimal competitive effects of the Cleveland Voucher Program, presentation at AERA, New Orleans, April
Parsons, E., Chalkley, B. and Jones, A. (2000) School catchments and pupil movements: a case study in parental choice, Educational Studies, 26, 1, 33-48
Pollock, L. (2000) SOC it to them, Times Educational Supplement, 30/6/00, p.26
Powers, J. and Cookson, P. (1999) The politics of school choice research: fact fiction and statistics, Educational Policy, 13, 1, 104-122
Reay, D. (1998) Engendering social reproduction: mothers in the educational marketplace, British Journal of Sociology of Education, 19, 2, 195-209
Riley, K. (1996) From intensive care to recovery, conference proceedings, London: London Borough of Haringey, November.
Stillman, A. (1990) Legislating for choice, in Flude, M. and Hammer, M. (Eds.) The Education Reform Act 1988, London: Falmer Press
Taylor, C., Gorard, S. and Fitz, J. (2000) A re-examination of segregation indices in terms of compositional invariance, Social Research Update , 30
TES (1999) GM could not fix class, Times Educational Supplement, 24/4/99, p.16
Tomlinson, S. (1997) Sociological perspectives on falling rolls, International Studies in Sociology of Education, 7(1):81-98
Waslander, S. and Thrupp, M. (1995) Choice, competition, and segregation: An empirical analysis of a New Zealand secondary school market, 1990-93, Journal of Education Policy, 10, 1, 1-26
White, P., Gorard, S. and Fitz, J. (1999) An analysis of local school admissions arrangements: Working Paper 35, Cardiff: School of Social Sciences
Whitty, G., Power, S. and Halpin, D. (1998) Devolution and Choice in Education: the school, the state and the market, Open University Press: Buckingham
Willms, J. and Echols, F. (1992) Alert and Inert Clients: The Scottish experience of parental choice of schools, Economics of Education Review, 11, 4, 339-350
Woods, P., Bagley, C. and Glatter, R. (1997) The public market in England: School responsiveness in a competitive climate. Occasional Research Paper 1, School of Education: Open University
Worpole, K. (1999) Driving forces, The Guardian, 8/6/99, p.17
This document was added to the Education-line database on 26 September 2000