Open consultation

I: Technical Annex on Highways Maintenance (HM) Relative Needs Formula (RNF)

Updated 20 June 2025

Applies to England

1. Introduction

Upper tier authorities (London Boroughs, Metropolitan Districts, Shire Counties and Unitary Authorities) have a statutory duty to maintain the roads in their area. These services include environmental, safety and routine road maintenance, structural maintenance, street lighting, and winter services.

The government intends to include a separate formula for Highways Maintenance. It does not consider the Foundation Formula to be an appropriate method of assessing demand for these services. Its drivers of need do not include spatial variables such as road length, which are particularly relevant in some transportation services.

The design of the updated Highways Maintenance formula is substantially similar to the previous 2013-14 formula for the service. The number of variables has been reduced to achieve a simpler model structure.

2. Proposed HM RNF methodology overview

The government proposes to use an expenditure-based regression to derive the RNF shares for this formula.

The government proposes that the dependent variable in the regression is the average of the three most recent years’ total net expenditure on the following revenue outturn (RO2) lines by each local authority:

• 11 - Highways maintenance planning, policy and strategy
• 31 - Structural maintenance (principal roads)
• 32 - Structural maintenance (other LA roads)
• 33 - Structural maintenance (bridges)
• 41 - Environmental, safety and routine maintenance (principal roads)
• 44 - Environmental, safety and routine maintenance (other LA roads)
• 48 - Winter service
• 49 - Streetlighting inc. energy costs

The total of this spend would then be divided by the road length in kilometres multiplied by the service-specific Area Cost Adjustment (ACA) to derive the dependent variable.

The government proposes that the independent variable in the regression is the average of the three most recent years’ data for traffic flow^{[footnote 1]} divided by the same for road length^{[footnote 2]} by each local authority.

The government proposes to make some adjustments to the data where appropriate to account for restructures in local government organisation. For expenditure data, this will follow the method set out in the Control Total technical annex. For the traffic flow and road length data, known splits in data will be used to create splits of previous years. For example, Cumbria County Council (Cumbria CC) and its 6 constituent district councils were abolished from the 2023/24 financial year onward and two unitary authorities (UAs), Cumberland Council and Westmorland and Furness Council, were created in their place. In this case, the known split between the two UAs will be applied to Cumbria CC’s values for the previous two years across the new unitary authorities.

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. This currently impacts on Birmingham and Westmorland and Furness.

The government proposes to exclude the City of London in the regression estimation due to it being a statistical outlier. As mentioned above, the authority would still get a value from the resulting model given its independent variable values.

Currently the government proposes estimating a regression model using the natural logarithm of both the dependent and independent variables. This is due to the slightly better fit this model results in for its fit to actual expenditure when compared to other specifications of the same model, such as log-linear and linear-linear. However, this may be subject to change depending on the nature of the data for the most recent three years at the time of estimation.

The government also proposes exploring the impact of using suitably granular climate/weather related data at local authority level in developing this formula.

3. Expenditure-based regression

The proposed regression specification for the Highways Maintenance RNF is as follows:

The regression results of this specification are summarised in Table 1 below.

Table 1: Highways Maintenance RNF regression table

	Est.	S.E.	t val.	p
(Intercept)	-5.09	0.04	-120.73	0.00
ln (Traffic flow / road length)	1.29	0.10	12.92	0.00

Standard errors: OLS

Adjusted R-squared: 0.53

Observations: 149 (3 missing)

The proposed independent variable on its own explains just over half of the variation in the dependent variable. It also intuitively has a positive coefficient value (which is statistically significant and indicates this is significantly different from zero) on the independent variable suggesting that on average higher levels of traffic flow per kilometre of road tend to be associated with higher spend per kilometre of road (after adjusting spend for variations in input costs through the Area Cost Adjustment).

4. Calculating proposed HM RNF shares

The proposed Highways Maintenance RNF share is calculated as follows for each upper tier authority:

i. -5.09 plus
ii. 1.29 multiplied by natural log of Traffic Flow divided by Road Length.

The result of the above is exponentiated and then multiplied by:

i. ACA for the Highways Maintenance RNF, multiplied by
ii. Road Length.

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.

1. TRA8906: Motor Vehicle traffic (vehicle kilometres) excluding trunk roads by local authority in England ↩

2. RDL0202: Road length (kilometres) by road type and local authority in Great Britain. Total of Major Principal Roads and All Minor Roads. ↩