Revaluation and reform of council tax in Scotland: design considerations and potential impacts

Footnotes

1 For example, we estimate that in Edinburgh City, among properties currently in Band D, 10% are worth £207,000 or less while another 10% are worth £410,000 or more. In Aberdeen City the respective figures are £113,000 and £199,000. Thus, there is wide variation in estimated values for properties in a given band both in a given council area and between council areas.

2 1991 values are not available for individual properties (only tax bands are). However, we estimate that based on Q3 2024 property values and the updated bandings we model later in this report, the average Band H property would be worth 15.5 times the average Band A property. We would expect a similar figure for existing bands based on 1991 values too.

3 Underlying this are several further implicit assumptions. In particular, we assume that councils collectively receive the same total amount in grant funding from the Scottish Government and spend the same total amount on local public services regardless of the structure of the council tax system. That means that in aggregate they must raise the same revenue from council tax regardless of the structure of the council tax system, thus implying that the average council tax bill across Scotland as a whole is unaffected by the structure of the council tax system.

4 Specifically, our approach is similar to what is termed ‘hedonic regression analysis’ in the academic literature. It differs from true hedonic regression analysis because the aim of that method is to estimate the true causal effects of different property and local area characteristics on property values. We are not interested in causal effects per se – just being able to predict property values for properties in Scotland. If a variable we include does not itself directly affect property values but is correlated with a variable we do observe which does, that variable is still useful as an explanatory variable in our property value equation.

5 Waves 11 to 13 were adversely affected by the Covid-19 pandemic and have therefore been omitted.

6 The methodology appendix (Appendix A) explains why our modelling is approximately but not precisely consistent with this assumption.

7 We define this as Glasgow City plus East Dunbartonshire, East Renfrewshire, Inverclyde, North Lanarkshire, Renfrewshire and West Dunbartonshire.

8 Equivalisation adjusts incomes to account for differences in the number and ages of household members to better reflect the living standards of different households.

9 Other health-related benefits we count are incapacity benefit, severe disablement allowance, war disablement allowance, industrial injury disablement allowance and other disability-related benefit or pay.

10 Note that the points in the property value distribution given here don’t precisely match those in the accompanying Excel workbook, as the property value distributions estimated in the geographical and household-level modelling do not fully align.

11 The estimated average (mean) property value for households in the lowest fifth of the income distribution is £156,000, compared to £297,000 for households in the top fifth of the income distribution.

12 See Ministry of Housing, Communities and Local Government (2016) for the scheme in place for the 2017 non-domestic rates revaluation in England. Note that the scheme in place for the 2023 revaluation differs, phasing in bill increases but not decreases, as explained in Department for Levelling Up, Housing and Communities (2022).

13 Formally, the process works differently. The amount of non-domestic rates revenues that councils are required to transfer to the Scottish Government via the ‘contributable amount’ (for redistribution as part of the Local Government Finance System) is reduced by the forecast cost of transitional relief. However, economically, this is equivalent to the contributable amount being based on non-domestic rates revenues before transitional relief is accounted for, and the Scottish Government providing a ring-fenced grant equal to the forecast cost of transitional relief, for councils to use to cover the cost of transitional relief. See Scottish Government (2023).

14 In particular, our geographical analysis provides the best estimates for the proportion of properties potentially eligible for transitional relief, but does not account for exemptions, discounts and the CTRS, which mean some potentially eligible properties pay a zero or reduced net bill. Therefore, to approximate the net costs of transitional relief we combine estimates of the share of properties potentially eligible and the gross cost of transitional relief from our geographical analysis, with an estimate of the ratio between net and gross cost of relief from our household-level analysis. We also model this assuming all councils set their Band D tax rate at the Scottish average of £1,543 (Scottish Government, 2025b).

15 Forecast council tax revenues taken from Scottish Government (2025e). Overall funding figure includes forecast council tax revenue plus planned funding from the Scottish Government set out in Scottish Government (2024b).

16 Unlike in England, where, following localisation of the CTRS in April 2013, many councils introduced minimum council tax payments for even the lowest-income working-age CTRS claimants (Adam et al, 2019), full CTRS is still available in Scotland.

17 Note that any of the council tax reform options we model that involve making the system less regressive would reduce households’ need for CTRS (because lower-income households tend to live in lower-value properties whose bills would be reduced on average, leaving them with lower tax bills for CTRS to cover) and therefore the cost of the scheme. Any policy changes to increase the scope and cost of CTRS would therefore be offsetting this prior reduction.

18 See Appendix A for details of how we apply the scheme to the 12-band, 14-band and continuous proportional systems.

19 There is a separate scheme for those in personal insolvency and a discretionary scheme for those facing excessive financial hardship.

20 Grant allocations to Scottish councils account for how much they could raise themselves if they set a Band D rate of £911 (Scottish Government, 2025f), while the average Band D rate charged by councils was £1,543 (Scottish Government, 2025c). This means that the notional rate used for grant allocation was 59% of the average rate actually set.

21 Rurality is measured using Scottish Government Urban Rural Classification (Scottish Government, 2022).

22 An exception is the two-child limit on benefit entitlements, for which we model the policy as it is in 2025–26.

23 Based on deciles of specific components of the IMD: income, employment, housing, education and health.

24 These include whether the household contains a couple, the number of adults, the number of children in different age groups, the highest qualification in the household, the age of the oldest household member and whether anyone in the household is in receipt of disability benefits or reports having a long-standing illness or disability.

25 If instead these properties are systematically more desirable, then we would understate the values of rented properties and hence underestimate the council tax liabilities of renter households.