Fair funding and levelling up for pupils with special educational needs and disability in England 2014 to 2023

This study explores the variations in funding by relative need, allocated to English local authorities (LAs) for special educational needs and disability, by analysis of the schools National Funding Formula (NFF). The ﬁndings demonstrate continuing and signiﬁcant diﬀerences in high needs (HN) funding between demographically similar LAs, despite a £ 3.5 billion (67%) increase in the HN budget since 2014-15. The funding variations range from £ 1,066 to £ 537 per pupil. The diﬀerences are largely the result of historical spending arrangements together with the application of a funding ﬂoor increase of 8% for all LAs and a funding ceiling of 11%. There is also a signiﬁcant link between the level of LA funding and their use of specialist provision and statutory assessments. The discussion explores the Government’s policy of fair funding and the NFF implementation, and concludes that any mechanisms which are applied to protect funding stability should not undermine levelling up.


Introduction
A new national funding formula (NFF) was introduced in England from 2018-19 to calculate the amount of core revenue funding that mainstream schools will attract, together with a revised formula for funding high needs pupils with special educational needs and disability (SEND).In explaining the rationale for the changes Justine Greening the then Secretary of State for Education said: This is an historic reform.It means, for the first time, the resources that the government is investing in our schools will be distributed according to a formula based on the individual needs and characteristics of every school in the country…….I want every child in England to have the opportunity to achieve their full potential for themselves and for our country.Too often it is someone's background that determines what they can achieve rather than their talent or hard work.(House of Commons, 2017;Department for Education, 2017a).
This study provides an update to research commissioned by the Special Educational Needs Policy Research Forum (SENPRF), which used data from 2018-19 (Marsh et al., 2019).The SENPRF research focused on the high needs SEND budget and showed continuing and significant differences in funding allocated to different parts of the country (TES, 2021).This current article will now use data from the 2022-23 NFF to examine the progress towards 'levelling up', which was a central Conservative manifesto pledge in 2019 (Institute for Government, 2021).The Prime Minister has returned to the theme of levelling up in July 2021 (Gov.UK, 2021a), within the context of the ongoing Covid-19 pandemic and as an introduction to a 2022 White paper (Gov.UK, 2022).
The main source of income to support a Local Authority's schools budget is from the Dedicated Schools Grant (DSG) (ESFA, 2021).The grant comprises of four funding blocks: schools, high needs, early years and central services.The case for a change in the system of schools funding was expressed in the consultation paper published by the Department for Education (DfE) in 2016.
The funding system for education should support the government's objective of enabling all schools to deliver excellence everywhere.Opportunity should be open to all, in all parts of the country: rural and urban, shire and metropolitan, north and south.It should be fair, with funding for schools based on the needs and characteristics of pupils, in a transparent, understandable way.It should be clear how much funding is following each pupil to their school, including disadvantaged pupils, and this should be the same everywhere, whilst reflecting higher costs in some areas.The allocation of high needs funding to local authorities should be equally rational (DfE, 2016).
Table 1 shows that in 2022-23 76% of the DSG is allocated to the schools block and 16% is allocated to High Needs.Table 2 displays some summary statistics to enable a comparison between the Schools block and High needs block for 2014-15 and 2022-23.The baseline of 2014-15 was chosen as the 2014 Children and Families Act and the associated SEND Code of Practice provided a reform of the system for identifying children and young people with special educational needs and disability (Gov.UK, 2014a).
Variations in funding have reduced in the Schools block in 2022-23, which are displayed in Figure 1.
Conversely the range has increased in the High Needs block over the eight year interval (see Figure 2).The revised High Needs (HN) funding budget for the NFF was derived from research and fair funding consultations since 2010 (DfE, 2017b;DfE, 2021a).The Government indicated its intention to review the existing basis for distribution of the high needs budget to local areas, when it introduced the National Funding Formula.Commissioned research was approved to recommend viable indicators of need.The research report concluded that the indicators of need could be a combination of prior attainment, social disadvantage measures, entitlement to Disability Living Allowance and measures of significant ill-health (DfE, 2015) (see Figure 3).2021).It is to enable Local Authorities and institutions to meet their statutory duties, but also to support alternative provision (AP) for pre-16 pupils who do not attend mainstream or special schools, due to exclusion, illness or other reasons.The HN funding block is for Local Authorities to use for place funding and top-up funding for institutions, and funding for high needs services delivered directly by the authority or under a separate funding agreement with institutions (including funding devolved to institutions), as permitted regulations.Second, there is a 'notional SEN' amount of up to £6,000 which comes directly from the schools' delegated budgets within the Schools Block.The notional SEN budget is not ringfenced and is subject to a formula calculated by individual LAs which can vary from £757 to £124 per 2-19 resident pupil (see Table 4).If a LA decides to proportion a higher level of resource into the notional SEND quantum, this does not affect the overall amount of school funding for the area, it just increases the amount that is linked to SEND related factors as opposed to the amount linked to the numbers on roll.There is also an age weighted pupil unit (AWPU) of at least £4,000 in the mainstream schools block of the dedicated schools grant (DSG) for each pupil on the school roll.Figure 4 shows an example of SEND funding within a mainstream school, whereby Pupil A with an education, health and care (EHC) plan (previously statement of SEN), is funded from both the mainstream schools block (£4,000), the notional SEN budget (£6,000) and the high needs block allocation (£5,000).Whereas Pupil C is funded by the mainstream schools block (£4,000) and from the notional SEN budget (£3,000).Pupil C is at the SEN support stage of the SEND Code of Practice (Gov.UK, 2014a), where support is provided without the need for an EHC plan.Each year LAs are required to detail their schools block funding formula using the authority proforma tool (APT) (Gov.UK, 2021b).In 2022-23 £12.8 billion (24% of the DSG) is specifically allocated to support pupils with SEND on top of the mainstream allocation of £4,000 per pupil.The funding comprises £4.1 billion of 'notional SEN' from within the schools block and £8.7 billion of 'high needs block' funding.
Table 1 shows that an additional £3.5 billion has been budgeted for high needs since 2014-15, which is a 67% increase (Gov.UK, 2013;2021d).
The Government decided to maintain allocations to higher-funded local authorities by basing half the formula on historic spend, which in 2022-23 has reduced to 32%, and by applying a 'funding floor' of 0.5% and 1% ('no area loses') for the first two years 2018-20, and with increases of 8% for 2020-23.
The growth to lower-funded LAs was capped at 3% per year for the first year 2018-19, then boosted to 17% in 2020-21, but reduced to 12% and 11% in 2021-23 (see Table 3).The impact of the funding floor and funding ceilings is highly significant and will be visited again later in this article.In this study there is an examination of factors that could be contributing to a higher spend in LAs, such as demographic factors and the use of specialist provision.The research will also aim to examine the association between HN funding/spend with the notional SEN budget, the percentage of 0-19 year olds with Education Health and Care plans and appeals to the SEND Tribunal.

RQ1
Are there differences in the High Needs budget allocations for demographically similar LAs?

RQ2
What is the impact of historic funding/spend to the HN budget allocations?RQ3 Do LAs with higher HN budget allocations make greater use of specialist provision?

RQ4
Is there a link between HN budget allocations and the notional SEN budget?

Methods
The project involved desktop research of the relationship between different variables through correlational and partial correlational analysis based on the methods used by Marsh (2017).The databases that are readily available are listed below.The DfE guidance stated that the Department had : updated the distribution for 2017-18 to reflect 2016-17 spending levels, but this remains directly linked to spending levels rather than to any estimate of levels of need [ … ] We are committed to moving to a more rational basis for distributing funding for children and young people with high needs, taking into account an up-to-date assessment of the level of need in each area (DfE, 2017).
The House of Commons (HoC) Education Committee were critical that the high needs funding formula was not sufficiently responsive to changing needs.
The factors and weightings in the formula should be amended to develop a more forwardlooking approach that is less reliant on historical factors, and takes greater account of projected trends and requirements for financial flexibility.(HoC Education Committee, 2019, para. 115).
For the 2022-23 budget the historic spend factor was amended to include the actual spend of LAs taken from the 2017-18 outturn financial statement.The historic spend factor is now 32% of the total HN budget compared to 44% in the initial HN formula from 2018-19.

Data and analysis relevant to RQ3:
The latest data from the Centre for Studies for Inclusive Education (CSIE) was for 2017 and had used the 0-19 population to calculate the percentage of pupils attending special schools, special academies or attending separate units, classes or resource bases within ordinary schools (Black and Norwich, 2019).
The CSIE range was from 2.1% in Torbay to 0.3% in Newham (England average 1.2%).This current article has used the 2022-23 High Needs National Funding Formula (HN NFF) (Gov.UK, 2021b) and by using the DfE's 'explore data' tool to calculate the number of pupils with EHC plans attending Alternate Provision (AP), that is PRUs, resourced provision and SEN units in mainstream schools (Gov.UK, 2021e).The AP data has then been aggregated with the HN NFF school census data which includes the number of pupils in maintained special schools, special academy, special free schools, non-maintained special schools, special post-16 institutions (SPI) and centrally funded institutions (CFI).The pupil numbers have been calculated by the LA in which the pupil is resident (Gov.UK, 2021c, import/export adjustment data).
The 2017 CSIE data is more comprehensive as it includes high needs pupils without EHC plans attending mainstream units, classes and resource bases as well as pupils attending special schools.
Nevertheless there is still a high correlation between the CSIE data and the 2022-23 HN NFF school census and the Alternative Provision data (r= 0.86).Correlation analyses have also been used to explore the links between specialist provision and historic high needs spend, the basic entitlement factor, HNB per pupil, proxy factors and EHC plans % (0-19).

Data and analysis relevant to RQ4:
Funding for notional SEN is not a separate formula factor in the NFF, as LAs specify what percentage of funding is allocated through each factor contributing to their notional SEN budget.In 2020-2021 £4.1 billion was allocated to notional SEN, the latest available data.The average notional SEN across all LAs is £358 per 2-18 pupil with a range from £124 to £757 (see Table 4).

Findings:
RQ1 Are there differences in the High Needs budget allocations for demographically similar LAs?
Table 4, Table 5 and Figure 5 illustrate the continuation of large differences between high funded and low funded demographically similar Local Authorities (LAs) under the HN NFF, after deductions for hospital education.The range is from £1066 to £537 per 2-18 resident with an England mean average of £754 (standard deviation £111).For example, Bury's total HNB £935 per pupil is 45% greater than Calderdale (£647), and Wandsworth's (£928) is 33% greater than Barnet (£713).Alternatively, Figure 5 shows that LA E1 would need an additional £32m to match LA A1; LA J1 would need £21m to match LA F1; LA O1 would need £11m to match LA K1 and LA T1 would need an additional £22m to receive the same level of funding as LA P1.
Figure 6 shows the highest LA quintile HN budget average in 2022-23 is £923 per head compared to a lowest LA quintile average of £612 per head.There continues to be a significant difference (51%) between the highest and lowest quintiles.

Low LAs mean £612
RQ2 What is the impact of historic funding/spend to the HN budget allocations?
Table 6 shows the significant correlations between the HN budget allocation with other NFF factors.
There is still a very strong correlation (r=0.87) between the historic spend factor and the final HN budget allocation per 2-18 resident at the LA level, highlighting the continuing importance of the 32% weighting given to the historic spend, even when reduced in 2022-23 from the initial 50% weighting.There is no significant correlation between the HN budget and appeals to the SEND Tribunal.
Using the multiple regression stepwise method with the 2022-23 data, a HN budget model based on only three factors: historic spend, free school meals (FSM) and HN budget allocation 2014-15, explains 88% of the variance (Adjusted R 2 =0.881).This finding highlights the power of these three factors over the rest.In a separate calculation, there is a very strong correlation between FSM with the whole basket of seven proxy factors (r=0.89),suggesting that the other six proxy factors have minimal impact on the formula.

13
RQ3 Do LAs with high HN budget make greater use of specialist provision?
Figure 8 shows the 2022-23 LA HN budget quintiles for the percentage of pupils with a EHC plan attending specialist provision by the resident LA.This figure shows that the LAs with higher HNBs also make greater use of special provision.Specialist provision embraces maintained special schools, special academies, special free schools, non-maintained special schools (NMSS), special post-16 institutions (SPI), centrally funded institutions (CFI) and alternative provision (AP) of pupil referral units, resourced provision or SEN units in academy, free schools or LA maintained schools.Table 7 illustrates the LAs with the highest and lowest percentage of pupils attending specialist provision with their corresponding HN budget per head.The table demonstrates that LAs with an higher HN budget make an overall greater use of specialist provision across all the LA types.Table 8 shows the significant correlation coefficients between specialist provision and the High Needs factors.The correlation between the percentage of pupils attending specialist provision and HN budget per head is r= 0.47.Table 8 also displays a link between specialist provision with EHC plans (r=0.62),Free School Meals (FSM) (r= 0.44) and Historic spend (r=0.32).There is no significant correlation between appeals to the SEND tribunal rate and the use of specialist provision.Table 9 shows the relationship between HN budget and the percentage of pupils attending specialist provision, when calculated for different types of LA.The correlations are higher (0.66 to 0.53).Note: ** all correlation coefficients significant at the 0.01 level Figure 9 displays specialist provision by type of LA, illustrating distinct differences between LA types.
The range is from 1.65% in the Metropolitan LAs to 1.30% in the Upper Tier (County) LAs. Figure 10 presents a scattergram that shows the relationship between the HN budget per pupil and the percentage in specialist provision for one type of LA (Unitary).The plots for the other LA types would have a similar distribution as all the correlations are significant in Table 9.

RQ4 Is there a link between HN budget allocations and the notional SEN budget?
Table 6 has already shown a small but significant link between the High Needs budget and the notional SEN budget (r=0.38).Figure 11 demonstrates the average notional SEN per pupil for each of the HN budget quintile groups.There is a range of £433 for the highest HN budget quintile to £305 for the lowest quintile.The combined notional SEN and HNB budget per pupil is from £1,356 for the highest quintile to £917 for the lowest quintile.

Discussion
The findings from this study demonstrate significant variations in the 2022-23 provisional HN block allocations to LAs, which undermine the principles of fair funding and levelling up.The differences in the level of budget allocations to English LAs range from £1,066 to £537 per 2-18 year old resident, with a mean average of £754.For example, the lowest funded LA would need an additional £11 million to receive the same levels of funding as one of its benchmarking neighbours.In another example, an LA E1 would need £32 million in funding in order to match the highest funded Authority A1, which is also one of its benchmarking neighbours (see Figure 5).
There is a high correlation between the 2022-23 HN allocations with the historic spend in 2017-18 (r=0.87) and with the previous pre-NFF HN budget in 2014-15 (r=0.80)(see Table 6).Additionally this research has confirmed a significant link (r=0.47) between the level of HN funding and the use of specialist provision for pupils (see Table 8) and also a relationship between the HN budget and the 'notional SEN' budget from within the schools block (r=0.38)(see Table 6).
One of the key policy aims of the NFF reforms is to address unintended variations in the amount of funding received by schools and pupils with similar characteristics (House of Commons, 2017, p3).The original proposal was to operate the NFF as a 'soft' formula in the first year before moving to a 'hard' formula in subsequent years.The schools and HN 'top up' budgets would be aggregated and allocated at a Local Authority level, then for local areas and schools forums to determine how to share out the funding.The financial year 2022-23 sees the fifth iteration of the National Funding Formula with limited indications so far of the transition to a 'hard' formula (see Belfield and Sibieta, 2017).
Despite an increase of £14.5 billion (37%) to the Dedicated Schools Grant and £3.5 billion to the HN block since 2014-15 (see  While the Department drags its feet, more and more children with special educational needs and disabilities are progressing through the school system without the support they need (HoC Public Accounts Committee, 2021).
Reports by the Comptroller and Auditor General (NAO) and the Institute for Fiscal Studies (IFS) have also both highlighted general concerns about the NFF.The NAO verified that there is still variation in funding between LAs with similar characteristics and that there has been a re-balancing of funding away from more deprived LAs towards less deprived LAs (NAO, 2021, paras 2.21-2.22).The IFS warns that the increase to the DSG needs to be evaluated against the fall in school spending per pupil in England during the last decade.There has been a 9% decrease in real terms between 2009-10 and 2019-20, which represents the largest cut in schools spending for over 40 years (IFS, 2021).The NAO have also warned that the government's Department for Levelling Up, Housing and Communities is not monitoring closely enough how it is spending the resources for levelling up schemes and whether its approach in different areas is working or not (NAO, 2022).
The increase in the HN budget allocation since 2014-15 has been approved by the government in response to a significant number of Local Authorities who have experienced budgetary pressures and overspending (ISOS Partnership, 2019;National Audit Office, 2019;DfE, 2020).The overspending has been attributed partly to an increase in numbers of pupils with complex needs and partly to increased demands for additional, alternative or specialist provision arising from budget pressures on mainstream schools, narrower school accountability and curriculum constraints (LGA, 2019).The National Audit Office (2019) showed that in 2017-18, 122 LAs (81%) overspent their schools high-needs budgets, compared to 2013-14 when 71 LAs (47%) overspent.More recent data was reported from 2019-20, that the number of overspending LAs had increased to 132 (87%) (TES, 2020) which has steered the policy towards a 'minimum funding guarantee' of a funding floor and a reduction of the funding ceiling (see Table 3).The DfE has required Local Authorities that have an overspend of 1% or more (on overall DSG) to produce a 'deficit recovery plan' (Gov.UK, 2019).This guidance appears to take little account of the level of funding that LAs were receiving in the first place.
One hypothesis explored in this study is that HN budget allocations are determined by the extent to which LAs have relied historically on specialist provision.There is a significant though modest correlation (r=0.47) between HN budget allocations and percentage of 2-18 pupils in special schools and Alternative Provision, including special units/resource bases.However, when the data is analysed by type of LA, correlations were higher for all four types of LAs (r= 0.66 to 0.53).This evidence is consistent with the suggestion that the greater historical use of specialist provision leads to the higher level of HN funding.

Limitations
This research has involved quantitative desktop analysis of the relationship between different variables which impact on the allocation of the High Needs budget of the National Funding Formula.Further qualitative research would widen the evidence base by interviewing stakeholders who 1) have a track record for successful budgetary control, for their insights of how to achieve good standards in financial health and resource management; and 2) have experienced cost pressures such that a very high deficit intervention was necessary from the Department for Education (Gov.UK, 2021h).

Conclusion
This research confirms there is continuing and significant inequity in the relative levels of HN budget funding allocated to English LAs, which cannot be explained by differences in need, as measured by indicators of social disadvantage (TES, 2021).Secondly the link, between the amount of funding to a local area and the amount of specialist provision being allocated to pupils in those areas, is also confirmed.LA systems that make greater use of specialist provision are generally more costly than an emphasis on inclusion (LGA, 2019; NAO, 2019; Gray et al., 2021).Specialist provision has now increased by 50% since 2014 (Gov.UK, 2014b; Gov.UK, 2021e), so if it is to be a preferable policy option, then an even higher level of resources will be required in the long run.
Funding according to relative needs has to be balanced against decisions about funding stability and responsiveness (NAO, 2011).Nevertheless if the primary objective of the NFF HN budget is to provide a fairer and more transparent funding system, then it is important that any mechanisms, which are applied to protect funding stability, do not undermine levelling up.

Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author received no financial support for the research, authorship, and/or publication of this article.

Introduction
A new national funding formula (NFF) was introduced in England from 2018-19 to calculate the amount of core revenue funding that mainstream schools will attract, together with a revised formula for funding high needs pupils with special educational needs and disability (SEND).In explaining the rationale for the changes Justine Greening the then Secretary of State for Education said: This is an historic reform.It means, for the first time, the resources that the government is investing in our schools will be distributed according to a formula based on the individual needs and characteristics of every school in the country…….I want every child in England to have the opportunity to achieve their full potential for themselves and for our country.Too often it is someone's background that determines what they can achieve rather than their talent or hard work.(House of Commons, 2017;Department for Education, 2017a).
This study provides an update to research commissioned by the Special Educational Needs Policy Research Forum (SENPRF), which used data from 2018-19 (Marsh et al., 2019).The SENPRF research focused on the high needs SEND budget and showed continuing and significant differences in funding allocated to different parts of the country (TES, 2021).This current article will now use data from the The funding system for education should support the government's objective of enabling all schools to deliver excellence everywhere.Opportunity should be open to all, in all parts of the country: rural and urban, shire and metropolitan, north and south.It should be fair, with funding for schools based on the needs and characteristics of pupils, in a transparent, understandable way.It should be clear how much funding is following each pupil to their school, including disadvantaged pupils, and this should be the same everywhere, whilst reflecting higher costs in some areas.The allocation of high needs funding to local authorities should be equally rational (DfE, 2016).
Table 1 shows that in 2022-23 76% of the DSG is allocated to the schools block and 16% is allocated to High Needs.Children and Families Act and the associated SEND Code of Practice provided a reform of the system for identifying children and young people with special educational needs and disability (Gov.UK, 2014a).
Variations in funding have reduced in the Schools block in 2022-23, which are displayed in Figure 1.
Conversely the range has increased in the High Needs block over the eight year interval (see Figure 2).The revised High Needs (HN) funding budget for the NFF was derived from research and fair funding consultations since 2010 (DfE, 2017b;DfE, 2021a).The Government indicated its intention to review the existing basis for distribution of the high needs budget to local areas, when it introduced the National Funding Formula.Commissioned research was approved to recommend viable indicators of need.The research report concluded that the indicators of need could be a combination of prior attainment, social disadvantage measures, entitlement to Disability Living Allowance and measures of significant ill-health (DfE, 2015) (see Figure 3).4).If a LA decides to proportion a higher level of resource into the notional SEND quantum, this does not affect the overall amount of school funding for the area, it just increases the amount that is linked to SEND related factors as opposed to the amount linked to the numbers on roll.There is also an age weighted pupil unit (AWPU) of at least £4,000 in the mainstream schools block of the dedicated schools grant (DSG) for each pupil on the school roll.Figure 4 shows an example of SEND funding within a mainstream school, whereby Pupil A with an education, health and care (EHC) plan (previously statement of SEN), is funded from both the mainstream schools block (£4,000), the notional SEN budget (£6,000) and the high needs block allocation (£5,000).Whereas Pupil C is funded by the mainstream schools block (£4,000) and from the notional SEN budget (£3,000).Pupil C is at the SEN support stage of the SEND Code of Practice (Gov.UK, 2014a), where support is provided without the need for an EHC plan.6 Each year LAs are required to detail their schools block funding formula using the authority proforma tool (APT) (Gov.UK, 2021b).In 2022-23 £12.8 billion (24% of the DSG) is specifically allocated to support pupils with SEND on top of the mainstream allocation of £4,000 per pupil.The funding comprises £4.1 billion of 'notional SEN' from within the schools block and £8.7 billion of 'high needs block' funding.
Table 1 shows that an additional £3.5 billion has been budgeted for high needs since 2014-15, which is a 67% increase (Gov.UK, 2013;2021d).
The Government decided to maintain allocations to higher-funded local authorities by basing half the formula on historic spend, which in 2022-23 has reduced to 32%, and by applying a 'funding floor' of 0.5% and 1% ('no area loses') for the first two years 2018-20, and with increases of 8% for 2020-23.
The growth to lower-funded LAs was capped at 3% per year for the first year 2018-19, then boosted to 17% in 2020-21, but reduced to 12% and 11% in 2021-23 (see Table 3).The impact of the funding floor and funding ceilings is highly significant and will be visited again later in this article.In this study there is an examination of factors that could be contributing to a higher spend in LAs, such as demographic factors and the use of specialist provision.The research will also aim to examine the association between HN funding/spend with the notional SEN budget, the percentage of 0-19 year olds with Education Health and Care plans and appeals to the SEND Tribunal.

RQ1
Are there differences in the High Needs budget allocations for demographically similar LAs?

RQ2
What is the impact of historic funding/spend to the HN budget allocations?RQ3 Do LAs with higher HN budget allocations make greater use of specialist provision?

RQ4
Is there a link between HN budget allocations and the notional SEN budget?

Methods
The project involved desktop research of the relationship between different variables through correlational and partial correlational analysis based on the methods used by Marsh (2017).The databases that are readily available are listed below.4. Percentage of pupils (0-19) with EHC plans for 2021 (Gov.UK, 2021e).As the number of EHC plans are very small for children aged below 24 months, the ONS (2022) age 2-18 population estimates have been used for the denominator and to be comparable with the NFF.

Data and analysis relevant to RQ1:
The total 2022-23 HN budget quantum to be modelled is £8.6 billion which excludes hospital education funding (£125m) but does include adjustments for import-export of pupils who reside in a different LA to the school they attend.The 2022-23 HN budget has been divided by the projected mid-2022 2-18 ONS population for each LA to calculate the HN budget per resident pupil (or per head rate) (ONS, 2022).
The City of London and the Isles of Scilly have been excluded as there is only one state school in each LA.The 'high LAs' in Table 4 are within the highest quintiles for the HN budget and the low LAs are within the lowest quintiles.LAs have been matched by the DfE High Needs Benchmarking Tool (Gov.UK, 2020), the free school meals entitlement percentage and by the 2-18 population.
the links between specialist provision and historic high needs spend, the basic entitlement factor, HNB per pupil, proxy factors and EHC plans % (0-19).

Data and analysis relevant to RQ4:
Funding for notional SEN is not a separate formula factor in the NFF, as LAs specify what percentage of funding is allocated through each factor contributing to their notional SEN budget.In 2020-2021 £4.1 billion was allocated to notional SEN, the latest available data.The average notional SEN across all LAs is £358 per 2-18 pupil with a range from £757 to £124 (see Table 4).

Findings:
RQ1 Are there differences in the High Needs budget allocations for demographically similar LAs?
Table 4, Table 5 and Figure 5 illustrate the continuation of large differences between high funded and low funded demographically similar Local Authorities (LAs) under the HN NFF, after deductions for hospital education.The range is from £1066 to £537 per 2-18 resident with an England mean average of £754 (standard deviation £111).For example, Bury's total HNB £935 per pupil is 45% greater than Calderdale (£647), and Wandsworth's (£928) is 33% greater than Barnet (£713).Alternatively, Figure 5 shows that LA E1 would need an additional £32m to match LA A1; LA J1 would need £21m to match LA F1; LA O1 would need £11m to match LA K1 and LA T1 would need an additional £22m to receive the same level of funding as LA P1.
Figure 6 shows the highest LA quintile HN budget average in 2022-23 is £923 per head compared to a lowest LA quintile average of £612 per head.There continues to be a significant difference (51%) between the highest and lowest quintiles.Figure 7 displays the HN budget allocation by type of LA.The range is from £859 per head in London to £652 in the upper tier County shires.Figure 6 and Figure 7 demonstrate a lack of levelling up when compared with 2014-15, despite the additional £3.5 billion increase to the HN budget.The budget difference remains the same between the London LAs and the upper tier LAs (Figure 7), although there has been a relative improvement in the average HN funding per head for Metropolitan and Unitary LAs.RQ2 What is the impact of historic funding/spend to the HN budget allocations?
Table 6 shows the significant correlations between the HN budget allocation with other NFF factors.
There is still a very strong correlation (r=0.87) between the historic spend factor and the final HN budget allocation per 2-18 resident at the LA level, highlighting the continuing importance of the 32% weighting given to the historic spend, even when reduced in 2022-23 from the initial 50% weighting.There is no significant correlation between the HN budget and appeals to the SEND Tribunal.
Using the multiple regression stepwise method with the 2022-23 data, a HN budget model based on only three factors: historic spend, free school meals (FSM) and HN budget allocation 2014-15, explains 88% of the variance (Adjusted R 2 =0.881).This finding highlights the power of these three factors over the rest.In a separate calculation, there is a very strong correlation between FSM with the whole basket of seven proxy factors (r=0.89),suggesting that the other six proxy factors have minimal impact on the formula.

13
RQ3 Do LAs with high HN budget make greater use of specialist provision?
Figure 8 shows the 2022-23 LA HN budget quintiles for the percentage of pupils with a EHC plan attending specialist provision by the resident LA.This figure shows that the LAs with higher HNBs also make greater use of special provision.Specialist provision embraces maintained special schools, special academies, special free schools, non-maintained special schools (NMSS), special post-16 institutions (SPI), centrally funded institutions (CFI) and alternative provision (AP) of pupil referral units, resourced provision or SEN units in academy, free schools or LA maintained schools.Table 7 illustrates the LAs with the highest and lowest percentage of pupils attending specialist provision with their corresponding HN budget per head.The table demonstrates that LAs with an higher HN budget make an overall greater use of specialist provision across all the LA types.Table 8 shows the significant correlation coefficients between specialist provision and the High Needs factors.The correlation between the percentage of pupils attending specialist provision and HN budget per head is r= 0.47.Table 8 also displays a link between specialist provision with EHC plans (r=0.62),Free School Meals (FSM) (r= 0.44) and Historic spend (r=0.32).There is no significant correlation between appeals to the SEND tribunal rate and the use of specialist provision.Table 9 shows the relationship between HN budget and the percentage of pupils attending specialist provision, when calculated for different types of LA.The correlations are higher (0.66 to 0.53).Note: ** all correlation coefficients significant at the 0.01 level Figure 9 displays specialist provision by type of LA, illustrating distinct differences between LA types.
The range is from 1.65% in the Metropolitan LAs to 1.30% in the Upper Tier (County) LAs. Figure 10 presents a scattergram that shows the relationship between the HN budget per pupil and the percentage in specialist provision for one type of LA (Unitary).The plots for the other LA types would have a similar distribution as all the correlations are significant in Table 9.

RQ4 Is there a link between HN budget allocations and the notional SEN budget?
Table 6 has already shown a small but significant link between the High Needs budget and the notional SEN budget (r=0.38).Figure 11 demonstrates the average notional SEN per pupil for each of the HN budget quintile groups.There is a range of £433 for the highest HN budget quintile to £305 for the lowest quintile.The combined notional SEN and HNB budget per pupil is from £1,356 for the highest quintile to £917 for the lowest quintile.

Discussion
The findings from this study demonstrate significant variations in the 2022-23 provisional HN block allocations to LAs, which undermine the principles of fair funding and levelling up.The differences in the level of budget allocations to English LAs range from £1,066 to £537 per 2-18 year old resident, with a mean average of £754.For example, the lowest funded LA would need an additional £11 million to receive the same levels of funding as one of its benchmarking neighbours.In another example, an LA E1 would need £32 million in funding in order to match the highest funded Authority A1, which is also one of its benchmarking neighbours (see Figure 5).
There is a high correlation between the 2022-23 HN allocations with the historic spend in 2017-18 (r=0.87) and with the previous pre-NFF HN budget in 2014-15 (r=0.80)(see Table 6).Additionally this research has confirmed a significant link (r=0.47) between the level of HN funding and the use of specialist provision for pupils (see Table 8) and also a relationship between the HN budget and the 'notional SEN' budget from within the schools block (r=0.38)(see Table 6).
One of the key policy aims of the NFF reforms is to address unintended variations in the amount of funding received by schools and pupils with similar characteristics (House of Commons, 2017, p3).The original proposal was to operate the NFF as a 'soft' formula in the first year before moving to a 'hard' formula in subsequent years.The schools and HN 'top up' budgets would be aggregated and allocated at a Local Authority level, then for local areas and schools forums to determine how to share out the funding.The financial year 2022-23 sees the fifth iteration of the National Funding Formula with limited indications so far of the transition to a 'hard' formula (see Belfield and Sibieta, 2017).
Despite an increase of £14.5 billion (37%) to the Dedicated Schools Grant and £3.5 billion to the HN block since 2014-15 (see  of the Department for Education's lack of urgency in making improvements to the school funding system.Their report on School Funding censored the DfE for further delays in the SEND review which had been announced in September 2019 (DfE, 2019).
While the Department drags its more and more children with special educational needs and disabilities are progressing through the school system without the support they need (HoC Public Accounts Committee, 2021).
Reports by the Comptroller and Auditor General (NAO) and the Institute for Fiscal Studies (IFS) have also both highlighted general concerns about the NFF.The NAO verified that there is still variation in funding between LAs with similar characteristics and that there has been a re-balancing of funding away from more deprived LAs towards less deprived LAs (NAO, 2021, paras 2.21-2.22).The IFS warns that the increase to the DSG needs to be evaluated against the fall in school spending per pupil in England during the last decade.There has been a 9% decrease in real terms between 2009-10 and 2019-20, which represents the largest cut in schools spending for over 40 years (IFS, 2021).The NAO have also warned that the government's Department for Levelling Up, Housing and Communities is not monitoring closely enough how it is spending the resources for levelling up schemes and whether its approach in different areas is working or not (NAO, 2022).
The increase in the HN budget allocation since 2014-15 has been approved by the government in response to a significant number of Local Authorities who have experienced budgetary pressures and overspending (ISOS Partnership, 2019;National Audit Office, 2019;DfE, 2020).The overspending has been attributed partly to an increase in numbers of pupils with complex needs and partly to increased demands for additional, alternative or specialist provision arising from budget pressures on mainstream schools, narrower school accountability and curriculum constraints (LGA, 2019).The National Audit Office (2019) showed that in 2017-18, 122 LAs (81%) overspent their schools high-needs budgets, compared to 2013-14 when 71 LAs (47%) overspent.More recent data was reported from 2019-20, that the number of overspending LAs had increased to 132 (87%) (TES, 2020) which has steered the policy towards a 'minimum funding guarantee' of a funding floor and a reduction of the funding ceiling (see Table 3).The DfE has required Local Authorities that have an overspend of 1% or more (on overall DSG) to produce a 'deficit recovery plan' (Gov.UK, 2019).This guidance appears to take little account of the level of funding that LAs were receiving in the first place.
One hypothesis explored in this study is that HN budget allocations are determined by the extent to which LAs have relied historically on specialist provision.There is a significant though modest correlation (r=0.47) between HN budget allocations and percentage of 2-18 pupils in special schools and Alternative Provision, including special units/resource bases.However, when the data is analysed by type of LA, correlations were higher for all four types of LAs (r= 0.66 to 0.53).This evidence is consistent with the suggestion that the greater historical use of specialist provision leads to the higher level of HN funding.

Limitations
This research has involved quantitative desktop analysis of the relationship between different variables which impact on the allocation of the High Needs budget of the National Funding Formula.Further qualitative research would widen the evidence base by interviewing stakeholders who 1) have a track record for successful budgetary control, for their insights of how to achieve good standards in financial health and resource management; and 2) have cost pressures such that a very high deficit intervention was necessary from the Department for Education (Gov.UK, 2021h).

Conclusion
This research confirms there is continuing and significant inequity in the relative levels of HN budget funding allocated to English LAs, which cannot be explained by differences in need, as measured by indicators of social disadvantage (TES, 2021).Secondly the link, between the amount of funding to a local area and the amount of specialist provision being allocated to pupils in those areas, is also confirmed.Local Authority systems that make greater use of specialist provision, which have now increased by 50% since 2014 (Gov.UK, 2014b;Gov.UK, 2021e), are due in part to the establishment of special and alternative provision free schools (Black et al., 2019).Specialist provision is also generally more costly than an emphasis on inclusion (LGA, 2019;NAO, 2019;Gray et al., 2021).So, if specialist provision and academisation of special schools are to be the preferable policy option, then an even higher level of resources will be required in the long run.
Funding according to relative needs has to be balanced against decisions about funding stability and responsiveness (NAO, 2011).Nevertheless, if the primary objective of the NFF HN budget is to provide a fairer and more transparent funding system, then it is important that any mechanisms applied to protect funding stability, do not undermine levelling up.

Figure 1 .
Figure 1.Distribution of funding per pupil for all LAs in the Schools block for 2022-23.

Figure 2 .
Figure 2. Distribution of funding per pupil for all LAs in the High Needs block for 2022-23.

Figure 3 .
Figure 3. Basic building blocks of the High Needs NFF 2022-23 to show that the factors will be added together to give the formula allocation, with an area cost adjustment applied to the proxy factors and basic entitlement (DfE, 2021b).

Figure 4 .
Figure 4. Examples of funding for SEND pupils (National Audit Office, 2019).
Figure7displays the HN budget allocation by type of LA.The range is from £859 per head in London to £652 in the upper tier County shires.Figure6and Figure7demonstrate a lack of levelling up when compared with 2014-15, despite the additional £3.5 billion increase to the HN budget.The budget difference remains the same between the London LAs and the upper tier LAs (Figure7), although there has been a relative improvement in the average HN funding per head for Metropolitan and Unitary LAs.

Figure 5 .Figure 6 .
Figure 5.The amount of funding for 2022-23 needed to match the highest funded Authority (A1, F1, K1 and P1) from within a set of five demographically comparable Statistical Neighbours.

Figure 9 .
Figure 9. Specialist provision 2021 by type of LA.

Figure 10 .
Figure 10.Scattergram of specialist provision 2021 in Unitary LAs against HN budget per pupil.
for Education's lack of urgency in making improvements to the school funding system.Their report on School Funding censored the DfE for further delays in the SEND review which had been announced in September 2019(DfE, 2019).
2022-23 NFF to examine the progress towards 'levelling up', which was a central Conservative manifesto pledge in 2019 (Institute for Government, 2021).The Prime Minister has returned to the theme of levelling up in July 2021 (Gov.UK, 2021a), within the context of the ongoing Covid-19 pandemic and as an introduction to a 2022 White paper (Gov.UK, 2022).The main source of income to support a Local Authority's schools budget is from the Dedicated Schools Grant (DSG)(ESFA, 2021).The grant comprises of four funding blocks: schools, high needs, early years and central services.The case for a change in the system of schools funding was expressed in the consultation paper published by the Department for Education (DfE) in 2016.

Figure 1 .
Figure 1.Distribution of funding per pupil for all LAs in the Schools block for 2022-23.

Figure 2 .
Figure 2. Distribution of funding per pupil for all LAs in the High Needs block for 2022-23.

Figure 3 .
Figure 3. Basic building blocks of the High Needs NFF 2022-23 to show that the factors will be added together to give the formula allocation, with an area cost adjustment applied to the proxy factors and basic entitlement (DfE, 2021b).

Figure 4 .
Figure 4. Examples of funding for SEND pupils (National Audit Office, 2019).
1. HN budget allocations to Local Authorities (LAs) for financial years2014-15 and 2022-23, with    deductions for hospital education(Gov.UK, 2013; 2021a).2. Gov.UK school census 2021 data for percentage of pupils (2-18) by the resident LA, in maintained special schools, special academy, special free schools and non-maintained special schools (NMSS), special post-16 institutions (SPI) and centrally funded institutions (CFI) by the resident LA (Gov.UK, 2021c, import/export adjustment data).3. Gov.UK SEN2 2021 Establishment type caseload data by the resident LA, for Alternative Provision -Pupil Referral Units (PRUs) in Academy, Free Schools and LA maintained schools, and Mainstream school -Resourced Provision and SEN units in Academy, Free Schools and LA maintained (Gov.UK, 2021e).

Figure 5 .Figure 6 .
Figure 5.The amount of funding for 2022-23 needed to match the highest funded Authority (A1, F1, K1 and P1) from within a set of five demographically comparable Statistical Neighbours.

Figure 9 .
Figure 9. Specialist provision 2021 by type of LA.

Figure 10 .
Figure 10.Scattergram of specialist provision 2021 in Unitary LAs against HN budget per pupil.

Table 2 .
Comparison of Schools block and High needs block allocations per pupil 2014-15 with 2022-23.

Table 3 .
National Funding Formula High Needs funding floor and funding ceiling 2018-19 to 2022-23.

Table 4 .
Summary statistics for Local Authorities from 2022-23 HN budget.

Table 5 .
The 2022-23 provisional HNB allocations per head for high and low funded LAs including additional funding for new and growing special free schools, adjustments for import/export factors and with deductions for hospital education.

Table 6 .
Correlations between the HN budget allocation per pupil 2022-23 and other factors.

Table 7 .
The use of specialist provision by LA type showing 2022-23 HN budget allocations per 2-18 resident.

Table 8 .
Correlations between the specialist provision in 2021 and other factors.

Table 9 .
Correlations between HN budget per head 2022-23 and percentage in specialist provision 2021 by types of LA.

Table 1
), the House of Commons Public Accounts Committee has been critical Table 2 displays some summary statistics to enable a comparison between the Schools block and High needs block for 2014-15 and 2022-23.The baseline of 2014-15 was chosen as the 2014

Table 2 .
Comparison of Schools block and High needs block allocations per pupil 2014-15 with 2022-23.

Table 3 .
National Funding Formula High Needs funding floor and funding ceiling 2018-19 to 2022-23.

Table 4 .
Summary statistics for Local Authorities from 2022-23 HN budget.

Table 5 .
The 2022-23 provisional HNB allocations per head for high and low funded LAs including additional funding for new and growing special free schools, adjustments for import/export factors and with deductions for hospital education.

Table 6 .
Correlations between the HN budget allocation per pupil 2022-23 and other factors.

Table 7 .
The use of specialist provision by LA type showing 2022-23 HN budget allocations per 2-18 resident.

Table 8 .
Correlations between the specialist provision in 2021 and other factors.

Table 9 .
Correlations between HN budget per head 2022-23 and percentage in specialist provision 2021 by types of LA.

Table 1 )
, the House of Commons Public Accounts Committee has been critical