Supplementary memorandum submitted by
the National Audit Office
1. At the deliberative session held on 14
June on the Chairman's draft Report Improving poorly performing
schools in England, the Committee raised some queries about
the performance of faith schools. I agreed to provide further
analysis, which is provided by this memorandum and covers:
— Ofsted inspection results;
— applications for places;
— admissions arrangements; and
— the Department for Education and
Skills' position on faith schools.
A. PUPIL ATTAINMENT
i) Absolute attainment
2. Perhaps the most important measure of
school performance comes from the levels of attainment of its
pupils. The Department publishes attainment data, and the simplest
datasets are for the absolute attainment of pupils: the National
Curriculum test results (for primary school pupils at the end
of Key Stage 2) and GCSEs.
3. We examined the religious character of
the highest performing schools, in absolute terms, in 2005. Figure
1 shows that faith-based primary schools accounted for two-thirds
of the top 229 schools, although faith schools represent only
around 36% of all primary schools. Faith-based secondary schools
accounted for nearly half of the top 100 schools, which is again
much higher than the 18% of all secondary schools that are faith-based.
Figure 1
HIGHEST PERFORMING SCHOOLS (ABSOLUTE TERMS),
2005
| |
| |
| Faith schools |
Non faith schools | Total
|
| |
| |
| Highest performing primary schools | 66%
| 34% | 100% |
All primary schools | 36%
| 64% | 100% |
Highest performing secondary schools |
45% | 55% | 100%
|
All secondary schools | 18%
| 82% | 100% |
| |
| |
| |
| |
Source: NAO analysis of DfES data
Notes:
1. The highest performing primary schools are the 229
schools that had the maximum possible score of 300 in the National
Curriculum tests.
2. The highest performing secondary schools excludes
selective schools, and are the 100 schools with the highest percentages
of pupils achieving five or more A*-C grade GCSEs or equivalent.
ii) Value added performance
4. We next examined performance in terms of the value
that schools added to the attainment of pupils on average in 2005—ie
pupils' attainment as they leave the school compared with their
attainment four years (primary schools) or five years (secondary
schools) earlier. The value added measure is generally considered
to be a more robust measure of a school's performance than absolute
attainment, because it measures the progress made by pupils.
5. Figure 2 shows that faith-based primary schools accounted
for nearly half of the top 100 schools, which is higher than the
36% of all primary schools that are faith schools. Faith-based
secondary schools accounted for around one-third of the top 100
schools, which is much higher than the 18% of all schools that
are faith schools. As with absolute attainment, faith schools
are more likely to be among the top schools. However, the value
added performance of faith schools is not as strong as for absolute
attainment.
Figure 2
HIGHEST PERFORMING SCHOOLS (VALUE ADDED), 2005
| |
| |
| Faith schools |
Non faith schools | Total
|
| |
| |
| Highest performing primary schools | 45%
| 55% | 100% |
All primary schools | 36%
| 64% | 100% |
Highest performing secondary schools |
34% | 66% | 100%
|
All secondary schools | 18%
| 82% | 100% |
| |
| |
| |
| |
Source: NAO analysis of DfES data
Notes:
1. The highest performing primary schools are the 100
schools with the highest "value added" measures, comparing
the performance of Year 6 pupils in the National Curriculum tests
with the middle performance of other pupils nationally who had
similar attainment four years earlier.
2. The highest performing secondary schools are the 100
(non-selective) schools with the highest "value added"
measures, comparing the performance of Year 11 pupils in GCSEs
and equivalents with the middle performance of other pupils nationally
who had similar attainment five years earlier.
iii) Value added performance, adjusted for external factors
6. The NAO's 2003 report Making a difference: Performance
of maintained secondary schools in England included an analysis
of school performance at Key Stage 3 (for 13-year olds) and GCSE,
adjusted for pupils' prior attainment and external factors such
as the percentage of pupils on free school meals. [7]This
is a more sophisticated analysis of school performance, which
the Department will be rolling out across all schools in 2006.
Our analysis[8] showed
that:
— at Key Stage 3, faith schools are associated with
very slightly higher results than non faith schools; and
— at GCSE, there is no statistically significant
difference between faith schools and non faith schools after taking
account of prior attainment and external factors that influence
attainment.
7. The NAO report commented that other research suggested
that many faith schools, and the communities from which they draw
their pupils, have a clear ethos and values that might lead to
better than average performance overall. [9]Research
also suggests that many high performing schools have developed
their own distinct identity. They have a positive, definable and
recognisable ethos that permeates the entire school, and is evident
in good pupil-teacher relations, shared vision, cooperative working
and common goals. [10]
B. PUPIL ATTENDANCE
8. The NAO has reported on attendance in schools (Improving
school attendance in England, (HC 212, Session 2004-05). The
report included a statistical analysis of the variations in absence
rates between different types of schools, after adjusting for
all other measurable factors that were associated with absence
(such as the incidence of free school meals).
9. Given that the great majority of voluntary aided and
voluntary controlled primary schools are also faith schools, the
analysis showed that faith-based primary schools are associated
with slightly lower levels of total absence compared with other
schools. Regarding secondary schools, the great majority of voluntary
aided and the majority of voluntary controlled schools are faith
schools. Faith-based secondary schools that are voluntary aided
are associated with lower absence (nearly 1.5 days per pupil on
average), but there is no relationship for voluntary controlled
schools. The detailed statistics are shown in Annex 1.
10. In response to a request from Richard Bacon MP, we
prepared a supplementary memorandum that compared faith schools'
total absence rates with those of other schools, without adjustment
for any other factors that are known to affect absence rates.
The memorandum was published with the Committee's 18th Report
of 2005-06. The comparison showed that, for secondary schools,
a much higher proportion of faith schools than non faith schools
performed better than average (Figure 3). The results of this
analysis are broadly consistent with the more sophisticated analysis
explained above (paragraphs 8 and 9).
Figure 3
PUPIL ATTENDANCE ANALYSIS (SECONDARY SCHOOLS), 2002-03
|
| Attendance level | Faith schools
| Non faith schools |
|
| Better than average | 63% |
39% |
| Average | 16% | 20%
|
| Worse than average | 21% |
41% |
| Total | 100% | 100%
|
|
| |
|
Source: NAO; data unadjusted for factors known to affect
absence rates.
11. In conclusion, faith schools tend to have better
pupil attendance, particularly at secondary level. Their performance
is likely to be related to ethos, parental support for the school
and to the types of children who are pupils, although a number
of other contextual factors are likely to be at play.
C. OFSTED INSPECTION
RESULTS
12. The NAO's 2006 report (Improving poorly performing
schools in England, HC 679, Session 2005-06) included a statistical
analysis of the relationship between 8 key factors and whether
a school was in an Ofsted category (Special Measures or Serious
Weaknesses). We found that there was no statistically significant
relationship between whether a school was a faith school and whether
it was in a category.
13. We have now carried out an additional analysis of
whether a relationship existed in London schools only between
the religious character of a school and whether it was in an Ofsted
category in July 2005. Again, we found that there was no statistically
significant relationship. Although 3.3% of faith schools compared
with 2.6% of non faith schools were in a category, this difference
is not statistically significant.
14. Our survey of schools that had recovered from being
in an Ofsted category found that of the 36 faith schools that
responded, 21 (58%) considered that they were helped to improve
by the guidance or support provided by a church or faith organisation.
15. We have carried out some additional research on primary
and secondary schools that were assessed by Ofsted in the 2004-05
school year as "outstanding" overall. Our analysis shows
that the proportion of outstanding primary schools that are faith-based
is similar to the proportion of faith-based primary schools across
all primary schools (Figure 4). Of the outstanding secondary schools,
a higher than average percentage are faith-based.
Figure 4
OUTSTANDING SCHOOLS, 2005
| |
| |
| Faith schools |
Non faith schools | Total
|
| |
| |
| Outstanding primary schools | 37%
| 63% | 100% |
All primary schools | 36%
| 64% | 100% |
Outstanding secondary schools | 25%
| 75% | 100% |
All secondary schools | 18%
| 82% | 100% |
| |
| |
| |
| |
Source: NAO analysis of list of 348 outstanding schools
published by Ofsted, 2005.
16. In conclusion, there is no clear evidence that being
a faith-school makes it easier to avoid poor performance. However,
Ofsted inspection results indicate that faith-based secondary
schools are more likely than non faith schools to be very effective
overall.
D. APPLICATIONS FOR
PLACES
17. Public awareness of the performance of schools, through
published materials and word of mouth, is likely to result in
schools perceived to be good schools being more popular among
parents applying for places for their children and thus over-subscribed.
Other factors exist that can cause a school to be over-subscribed:
in particular where there are insufficient school places in the
area, where a school markets itself well, and where other local
schools are known to be in difficulty.
18. The Department does not collect national data on
applications for individual schools. However, we expect that most,
if not all, local authorities hold such data. Some of them publish
it. We have analysed the published data for a sample of seven
local authorities, covering 423 schools (Figure 5). Our sample
indicates that faith-based primary schools are no more likely
to be over-subscribed than non faith schools, while faith-based
secondary schools are more likely to be over-subscribed than non
faith schools.
Figure 5
OVER-SUBSCRIPTION OF SCHOOLS, 2005
| |
| | |
| Faith schools over-subscribed
| Non faith schools over-subscribed
|
| Number |
Percentage of total
faith schools |
Number | Percentage of total non
faith schools
|
| |
| | |
| Primary | 46 | 36%
| 65 | 39% |
| Secondary | 21 | 62%
| 46 | 49% |
| |
| | |
| |
| | |
Source: NAO analysis of applications data for 423 schools
in seven local authorities, 2005.
19. In conclusion, at secondary level faith schools are
more likely to be very popular than non faith schools, and there
is little difference at primary level. This provides an indication
that more parents prefer that their secondary school aged children
be educated at a faith school. However, this is based on a small
number of authorities and caution should be exercised because
they may not be representative and other factors could be at play,
such as there being fewer places at faith-based secondary schools
than faith-based primary schools in England. It would require
more time and work to prepare a robust analysis across the country.
E. ADMISSIONS
20. Some commentators of faith schools consider that
they are able to improve their performance through exercising
a degree of selection over the pupils that they take in. Where
a school is voluntary aided (or foundation, academy or city technology
college), the school's governing body acts as its admissions authority.
5[11] Where a school
is voluntary controlled or community, the local authority acts
as the admissions authority. The commentators claim that some
schools that are their own admissions authorities select pupils
who are likely to be more able or to receive more parental support.
Although some non faith schools control their own admissions,
much higher proportions of faith schools are able to do so. Figure
6 gives our estimates of the percentages of schools that control
their own admissions. It shows that most faith schools control
their admissions, particularly in the secondary sector.
Figure 6
SCHOOLS THAT ARE ADMISSIONS AUTHORITIES
|
| Percentage of schools that are admissions authorities
|
| Faith schools |
Non faith schools |
|
| Primary | 60% | 3%
|
| Secondary | 88% | 19%
|
|
| | |
Source: NAO analysis of DfES data, 2003.
21. We examined how many of the top secondary schools
are admissions authorities, compared with the 88% of faith-based
secondary schools overall:
— of the 45 faith schools in the top 100 schools
for absolute GCSE performance, 44 (98%) controlled their own admissions;
— of the 34 faith schools in the top 100 schools
for value added, 32 (94%) controlled their own admissions; and
— of the 18 faith-based secondary schools assessed
as outstanding by Ofsted in 2004-05, all 18 (100%) controlled
their own admissions.
22. Therefore, top performing faith schools are more
likely to be their own admissions authority than other faith schools.
Although the numbers of schools included in these analyses are
not high, the analysis indicates that being an admissions authority
could contribute to the good academic performance of faith schools.
23. The Sutton Trust[12]
found that faith schools accounted for 42% of the top 200 comprehensives
on the basis of the percentage of pupils achieving five or more
GCSEs at A*-C grade (a similar percentage to our analysis of the
top 100). The Sutton Trust examined whether the pupil intake of
the top secondary schools, faith and non faith, reflected the
characteristics of their local neighbourhood (Figure 7). It found
that, on average, the faith schools were located in more deprived
neighbourhoods (based on the percentage of pupils eligible for
free school meals) than non faith schools. However, on average,
the pupils at the faith schools and non faith schools came from
similar backgrounds, because the faith-based secondary schools
were less reflective of their neighbourhoods than non faith schools.
Figure 7
TOP SECONDARY SCHOOLS AND THEIR NEIGHBOURHOODS
| |
| |
| Religious character of school | Average percentage of pupils eligible for free school meals
|
| In their
neighbourhood
| In school | Gap between school and
neighbourhood
|
| |
| |
| Faith | 15.2% | 5.9%
| 9.3% |
| Non faith | 8.6% | 5.3%
| 3.3% |
| |
| |
| |
| |
Notes:
1. The Sutton Trust defined a school's "neighbourhood"
as the postcode sector in which the school is situated. It is
not necessarily the same as the school's catchment area, which
may be larger, and a school may not be situated in the middle
of its postcode sector.
2. The average percentage eligibility for free school
meals at secondary schools is 17.9%.
24. In conclusion, many faith schools have scope to exercise
a degree of selection of their pupils. We do not know the extent
to which this occurs in practice.
F. THE DEPARTMENT'S
POSITION ON
FAITH SCHOOLS
25. The Committee's 19th Report of 2003-04 concluded
that the Department should identify and promote the strengths
of faith schools across the school sector. (The conclusion and
an extract of the response is at Annex 2). The Treasury Minute
responded that the Government supported more faith schools as
part of its policy to increase diversity and raise standards.
It stated that the case of faith schools did not rest on superior
performance, although many faith schools did perform well which
may reflect the following factors:
— strong values and unique ethos;
— they draw from communities that value education;
and
— parents that seek out faith schools may provide
their children with more support.
26. The 2005 Schools White Paper reiterated the need
for a diversity of providers, including faith-based groups.
OVERALL CONCLUSIONS
27. Faith schools are a very important part of the maintained
schools sector in England, particularly well regarded for having
strong ethos and values. High proportions of faith schools are
among the best performing schools in the country according to
measures of pupil attainment and Ofsted's assessments. Faith-based
secondary schools in particular are impressive performers, and
they are often over-subscribed. At the other end of the scale,
faith-based schools are just as likely non faith schools to get
into difficulty.
28. There are indications that some faith-based secondary
schools may owe their good performance at least in part to their
intake. The top schools do not fully reflect their neighbourhoods,
indicating that their ability to control their admissions could
be a factor in their good performance. The Department will be
publishing "contextually value added" performance information
for all schools on the 2006 exam and test results. Analysis of
this information would further inform the debate about whether
faith schools owe their good performance to factors other than
the overall effectiveness of the school.
Annex 1
EFFECT OF SCHOOL TYPE ON ABSENCE IN PRIMARY AND SECONDARY
SCHOOLS
| Effect on absence (days per pupil per year)
|
| Primary schools
| Secondary schools |
School type | Total absence
| Unauthorised
absence | Total absence
| Unauthorised
absence |
| |
| | |
| Foundation | -0.30 |
| -0.65 | -0.40 |
| Voluntary aided | | -0.35
| -1.41 | -0.77 |
| Voluntary controlled | -0.15
| +0.10 | |
|
| Faith (Christian) | | -0.16
| | |
Notes:
1. Figures are shown in comparison to all other schools.
2. No entry in this table means that the factor has no
statistically significant relationship with absence.
Source: Improving school attendance in England,
C&AG's Report (HC 212, Session 2004-05), paras 2.20-2.21 and
Appendix 2 (Figure 11).
3. Where faith (Christian) schools are voluntary aided,
voluntary controlled or foundation, the two effects should be
added.
Annex 2
EXTRACT FROM THE TREASURY MINUTE ON THE 19TH REPORT FROM
THE COMMITTEE OF PUBLIC ACCOUNTS, SESSION 2003-04
PAC conclusion (vi): Adjusted performance measures also
show that specialist schools, faith schools, beacon schools and
single sex schools do better than average. The strengths of these
schools, such as a strong set of values and ethos, should be identified
by the Department and promoted across the school sector.
21. The Department accepts this recommendation|
25. |The Government supports more faith schools in the
maintained sector as part of its policy on increasing diversity
and raising standards. While the case for faith schools has never
rested on superior performance, many faith schools do perform
well and the Department has identified some of the factors that
may be relevant in explaining why this is so:
— they tend to have strong values and a unique ethos;
— many draw from communities that particularly value
education—Jewish, Muslim and Sikh schools for example; and
— parents who seek out faith schools may provide
their children with a high level of support.
26. Social background may also be a consideration. On
average faith schools have fewer pupils with free school meals
than other schools.
27. The Department is actively encouraging faith schools
to work in partnership with others in the local family of schools
and to share values, ethos and best practice. It is also planning
to introduce parental involvement networks to promote and change
the culture around parental involvement in their children's education|
29. |The Department has introduced new "competition"
arrangements for providing additional secondary schools in order
to encourage a range of school providers, including faith groups,
to put forward proposals and increase the diversity of provision.
The Secretary of State will decide such proposals and contribution
to raising standards will be key consideration.
7
1 The proportion of pupils who are eligible for free school meals
is an indicator of socio-economic deprivation. Back
8
2 C&AG's Report, Making a difference: Performance of maintained
secondary schools in England (HC 1332, Session 2002-03), Figure
9, p 20. Back
9
3 The impact of specialist and faith schools, National
Foundation for Educational Research, 2002. Back
10
4 High performing specialist schools: What makes the difference,
National Foundation for Educational Research, 2002. Back
11
5 Admissions authorities have responsibility for deciding arrangements
for admitting pupils. All admissions authorities must have regard
to the statutory guidance within the School Admissions Code of
Practice and the School Admission Appeals Code of Practice. The
admissions code of practice provides guidance on acceptable oversubscription
criteria. Back
12
6 The Sutton Trust is a charity, established in 1997 with the
aim of challenging educational inequality. The evidence used in
this memorandum comes from The social composition of top comprehensive
schools, 2006. Back
| 
