Open consultation

G: Technical Annex on the Foundation Formula (FF)

Updated 20 June 2025

Applies to England

1. Introduction

The government proposes to introduce a Foundation Formula that assesses the relative need for services provided by local authorities that are not captured by more service-specific Relative Need Formulae (RNFs). The formula would thus cover a wide range of service areas such as central services, waste services, planning, leisure and sports, cultural services, environmental health services, public transport (including bus services), concessionary travel, flood defence, coastal protection.

The current structure of local government is based on a two-tier system, with lower and upper tier authorities being responsible for different service areas. There also exist unitary areas, where a single authority delivers both lower and upper tier services. the Foundation Formula will consist of two separate formulas for upper and lower tier authorities to reflect this structure. Shire Councils would thus be in scope of the Upper Tier Foundation Formula, while Shire Districts would be in scope of the Lower Tier Foundation Formula. Metropolitan Districts, London Boroughs (including City of London), and Unitary Authorities would be in scope of both formulae.

2. Proposed Foundation Formula methodology overview

The government proposes that the Foundation Formula is constructed using the ‘client group + need drivers’ approach that was first proposed in the 2017 Fair Funding Review. A ‘client group’ of people eligible for the services covered by the formula is first identified, and used to estimate the per-capita spend on these services for each authority. A set of potential drivers of this ‘need-to-spend’ are identified, with a linear regression being used to derive the weights of these drivers.

3. Proposed client group

Non-social-care services provided by local authorities are used by the widest range of people out of all the service areas in scope of reform. The government proposes that the client group of the Foundation Formula should be comprised of the following population groups: total local authority residents, commuters, domestic day visitors, and domestic overnight visitors. Table 1 below provides a summary of how non-resident population sources might increase need for non-social-care services within an authority. It is assumed that the components making up client population are the same for both the lower tier and upper tier formulae.

Table 1: Potential impact of local authority non-residential population on non-social care service need

Non-social-care service area	Commuters	Domestic and Foreign Tourists
Waste Services	Additional waste generated by offices and workplaces; additional street cleaning required in commercial areas.	Additional waste generated by e.g. day trippers and hotel guests. Additional street cleaning e.g. in food and entertainment areas.
Planning	Indirectly. Increasing commuter levels would increase demand for new offices and workplaces.	Indirectly. Increasing tourist levels would increase demand for new tourist accommodation and holiday homes.
Environmental Health	Additional noise pollution, air quality control from additional traffic, pressure on food safety standards through additional demand	Additional noise pollution, air quality control from additional traffic, pressure on food safety standards through additional demand
Parking Services	Need for on-street and off-street parking for commuter vehicles.	Need for on-street and off-street parking for e.g. day visitor vehicles and tour buses.
Transport Planning	Areas with higher commuter levels may face additional logistical challenges that require more extensive planning	Areas with higher tourism levels may face additional logistical challenges that require more extensive planning
Leisure and Cultural Services	Areas with higher commuter levels may see additional demand for authority-funded sports centres.	Additional demand for authority-funded museums, art galleries, and LA-funded events.

A summary of the data sources considered for estimating the size of each of these population groups is set out below.

Residential population

The government proposes to use estimates of authority total residential population drawn from the Office for National Statistics (ONS) mid-year population estimates. These are estimated by “rolling forward” authority populations recorded during the 2021 Census, by adjusting for births, deaths, and internal migration.

Commuters

Census 2021 origin-destination data on place of work are used to estimate the total number of commuters for each authority. Those aged 16 and over either in employment or temporarily away from work the week before Census Day were asked where they mainly work. As Census Day occurred during a national lockdown in England, home workers were not asked to provide a usual workplace address. In addition, when completing the questionnaire, those on the furlough scheme were asked to select the “temporarily away from work” option, as were those quarantining and self-isolating. This means that although this is the latest commuter data available, it may poorly reflect post-pandemic commuting patterns. The government is currently exploring whether it would be possible to uplift the data to better reflect  current commuting patterns. 

There have been a number of structural changes to local government since the 2021 Census, with new unitary authorities formed from the merger of non-metropolitan districts. The origin-destination data was transformed to reflect these changes, with commuter flow between two shire districts that were subsequently merged together being recalculated as flow within the same authority.

The origin-destination data was used to calculated three variables: gross out-commuters, gross in-commuters, and net in-commuters (the latter minus the former). There is a choice between using the gross or net in-commuter figure as the definition of commuter flow. Using the net figure would account for the reduced need authorities face during the workday from residents commuting out for work. Conversely, given that authorities do have some control over commuter flow, there was a concern that this might encourage them to boost their net figures by reducing out-commuting as opposed to promoting in-commuting. On balance, the government proposes to use gross in-commuter when accounting for commuting inflows.

Domestic tourists

Data on the volume of domestic day and overnight visits by residents in England was taken from the result of the annual Great Britain Tourism Survey (GBTS) run by VisitBritain.

VisitBritain has in the past published total day and overnight visits for each local authority, based on three-year averages. The most recent data is for the three-year period from 2017 to 2019. There has been a substantial change in the GBTS methodology from 2022 onward, which means that the results from after that year cannot be compared to data collected prior in a straightforward way. The change in methodology in 2022 means that the 3 full years of data required to produce the newest estimates of authority-level visits and trips would only be available later in 2025.

VisitBritain had provisionally produced 2-year average county-level data, covering the period from 2022 to 2023 for overnight trips and from October 2021 to September 2023 for day visits. This was the latest data available when the estimates of domestic tourism were made. The government proposes to use the latest county-level data to estimate the level of domestic tourism, apportioning the county-level totals based on 2017-2019 within-county shares. The government is considering switching to using the newest authority-level estimates when these are made available.

The GBTS uses 3 main definitions of domestic day visits: ‘3 hour+ leisure day visits’, ‘tourism day visits’, and ‘tourism day visits - activities core to tourism’^{[footnote 1]}. To qualify as an eligible day visit, the following criteria must be met:

• 3 hour+ Leisure Day Visits (LDV): visits that last 3 hours or more (including travel time), involve undertaking 1 or more eligible leisure activity, and must not be overnight.
• Tourism Day Visit (TDV): in addition to fulfilling criteria for a LDV, the visit must be undertaken less often than once a week, and include a visit to a place outside of the local authority where the trip started, with the exception where the main activity is a visitor attraction, attending a public event or watching live sport.
• Tourism day visits – Activities Core to Tourism: in addition to fulfilling criteria for a TDV, the visit must have included activities such as going to a visitor attraction, going sightseeing, visiting museums or art galleries, among others.

For the purposes of this formula, the government adopts the TDV as its definition of domestic day visits. The aim is to exclude as much as possible visits by residents within the same authority, to avoid double-counting members of the client group already captured by residential population. However, it is not possible to exclude all such visits due to the use of the TDV definition.

The GBTS measures the volume of overnight visits to an authority in two ways: trips and bed nights^{[footnote 2]}. A trip is defined as consisting of at least one night spent away from home by an adult or a child, while bed nights are defined as the total duration of a trip for each person. The government uses total bed nights across all journey purposes (holidays, visiting friends and relatives, and business) to capture need generated by overnight visitors across their entire stay in the authority. Similarly to day visitors, we would ideally exclude overnight visits done by residents within their own authority to avoid double-counting members of the client group already captured by residential population. However, we are currently unable to account for proportion of total trip volume such trips make up.

The data on day visitors and visitors nights is presented in terms of annual volume. In order to make the units of measurement consistent with the other two components of the client group, the annual volumes were divided by 365 (the usual number of days in a year) to obtain an estimate of the average daily volume of both measures.

Daytime Population

The government proposes for the client group to be defined as the “daytime population” of an authority, which would reflect the average daily amount of people of all ages in an authority. The government proposes to estimate daytime population by taking the sum of its components, as set out below:

The non-residential components of daytime population are held constant when calculating daytime population for different years, as only residential population is estimated annually.

The government has also considered accounting for the number of foreign visitor nights to an authority as part of the definition of daytime population. There is currently no official publication of authority-level estimates of foreign visitors, although VisitBritain does produce annual county-level estimates. This means that there would be less confidence in the volume of foreign visitors per authority compared to domestic visitors. Moreover, it was found that including an estimated foreign visitor value into the Foundation Formula model did not have a significant impact on the distribution of total need shares. As such, it was ultimately decided to not include foreign visitors in the definition of daytime population.

Non-residential population weights

Commuters and tourists likely use non-social-care services public services like waste services or cultural facilities at different levels of intensity compared to residents. Given this, the components of daytime population should be weighted to reflect this difference in per-capita need relative to that of residents. The government is currently exploring if these weights could be derived through statistical methods or using expert and/or ministerial judgement. No weights were applied to the non-residential components of daytime population when calculating the indicative Foundation Formula shares for this consultation.

4. Proposed needs drivers

The government identified a number of potential need drivers of non-social-care need and tested their inclusion in the Foundation Formula. The choice of the variables to test were based on conversations with local government stakeholders and experts.

When identifying potential need drivers, variables that are in direct control of local authorities were excluded. This was done to avoid any perverse incentives that could arise if such a driver were included in the formula. An illustrative example would be including total waste collected per person as a driver of waste services. This would lead to authorities with higher levels of waste to be assessed as having more need. Local authorities generally charge fees for specific waste services, such as for commercial waste collection or DIY waste disposal. Including this driver may encourage authorities to lower their fees in anticipation that the reduced fee income would be compensated by higher demand for waste collection, which would in turn increase their formula funding.

Given the wide range of services that the Foundation Formula covers, as well as a lack of granular data below the authority level, no need driver was found to have strong explanatory power for per-capita need. The lack of granular data also means that we would be unable control for differences in local preferences and quality of service provision, which may lead to different per-capita spending in two authorities with the same level of need.

As such, the final choice of need drivers and how these are weighted was based on a combination of statistical evidence, evidence gathered from stakeholders, as well as judgment. In line with this, the government proposes to include the Index of Multiple Deprivation (IMD) as the only need driver for both the Lower Tier and Upper Tier Foundation Formula.

The IMD was chosen over other potential measures of relative deprivation (such as household income) as it accounts for the broadest range of factors that impact the level of deprivation of an area. These include levels of income, employment, educational attainment, health, crime, barriers to housing, and the living environment. The government has aimed to capture the distribution of deprivation within authorities by estimating a population-weighted average of their overall score, at the highest level of granularity available (the Lower Super Output Area, which captures approximately 400 to 1,200 households).

An alternative measure of deprivation that was considered was that of Census 2021 household deprivation. This data is available at a more granular Output Area level (around 40 to 250 households) and thus may be able to capture even more local pockets of deprivation. However, unlike the IMD, the Census data does not capture either income or employment and thus is not able to capture quite important dimensions of deprivation.

The latest release of England IMD scores was in 2019, which is the version that is used to produce the indicative authority-level need shares in this consultation. MHCLG has commissioned an updated iteration of the index that is due to be published later in 2025. The government will be updating our formula to reflect the most up-to-date IMD data when it is made available.

5. Expenditure-based regression

The government proposes to use an expenditure-based regression approach to derive the RNF shares for the Foundation Formula.

Dependent variable estimation

The government proposes to calculate the dependent variable in the regression as set out below.

The average net current expenditure (NCE) on services in their scope is first calculated, using the last 3 years of available data. This is done to smooth out any annual volatility in spending. The most recent years this data is available are the 2021/22 to 2023/24 financial years. The NCE values are derived from Revenue Outturn (RO) data of local authorities using the same approach as the Control Totals for the Foundation Formula. More details can be found in the Control Total technical annex.

This average annual spend is then divided by the Area Cost Adjustment (ACA) to adjust for differences in input costs. A separate ACA is applied to the Lower Tier and Upper Tier Foundation Formulae.

Following this, the ACA-adjusted annual spend is then divided by a 3-year average of daytime population to arrive at the final dependent variable of the regression.

Where there is missing expenditure data for the full three years, the government proposes excluding these authorities from the regression estimation. Any authorities affected would still get a value from the model given their values for the independent variables but the coefficients in the model would be estimated without them. For the Lower Tier Foundation Formula, 17 such authorities were excluded: Babergh, Birmingham, Brentwood, Bromsgrove, Cambridge, Cannock Chase, Castle Point, Colchester, High Peak, Mid Suffolk, North Norfolk, North West Leicestershire, Redditch, Stafford, Westmorland & Furness, and Woking. For the upper tier Foundation Formula, only Birmingham and Westmorland & Furness were excluded.

Regression results for the proposed formulae

The proposed regression specification for the Lower Tier Foundation Formula is as follows:

The regression results of this specification are summarised in Table 2 below.

Table 2: Lower Tier Foundation Formula regression table

	Est.	S.E.	t val.	p
(Intercept)	88.27	4.17	21.14	0.00
IMD Score	1.28	0.19	6.58	0.00

Standard errors: OLS

R² (Adjusted R²): 0.14 (0.13)

Observations: 278 (17 missing)

The proposed regression specification for the Upper Tier Foundation Formula is as follows:

The regression results of this specification are summarised in Table 3 below.

Table 3: Upper Tier Foundation Formula regression table

	Est.	S.E.	t val.	p
(Intercept)	102.06	6.23	16.38	0.00
IMD Score	1.46	0.26	5.68	0.00

Standard errors: OLS

R² (Adjusted R²): 0.18 (0.17)

Observations: 150 (2 missing)

As mentioned in the previous section, the models suffer from a relatively low fit, with just 14% of the variation of ACA-adjusted per-capita spending of lower tier authorities being explained by the proposed model, and just 18% of the variation for upper tier authorities. This may be due to a combination of factors, including the wide range of services that the Foundation Formula covers, as well as a lack of granular data below the authority level.

6. Calculating proposed Foundation Formula shares

The proposed regression models are used to create the final Foundation Formula relative need shares through the following steps:

Lower Tier Foundation Formula

The share for each lower tier authority of the proposed Lower Tier Foundation Formula is calculated as follows:

i. 88.27 plus
ii. 1.28 multiplied by the population-weighted IMD score.

The result of the above is multiplied by:

i. ACA for the Lower Tier Foundation Formula, multiplied by
ii. Size of daytime population.

The results of the above for all lower tier authorities are then added together. The final RNF share of each authority would thus be the share of each authority’s contribution to that sum.

Upper Tier Foundation Formula

The share for each upper tier authority of the proposed Upper Tier Foundation Formula share is calculated as follows:

iii. 102.06 plus
iv. 1.46 multiplied by the population-weighted IMD score.

The result of the above multiplied by:

iii. ACA for the Upper Tier Foundation Formula, multiplied by
iv. Size of daytime population.

The results of the above for all upper tier authorities are then added together. The final RNF share of each authority would thus be the share of each authority’s contribution to that sum.

The size of daytime population for both formulae is determined as the latest available estimate of daytime population, that is the one based on the 2023 mid-year population estimates.  The government is considering switching to the ONS 2022-based subnational population projections for this measure of daytime population when these are made available, in keeping with the consultation principle of dynamism.

1. GB Day Visitor Annual Report 2024 ↩

2. GB Tourist Report 2024 ↩