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

2. Reforms modelled and summary of methodology

This section sets out the indicative reform options we model, and our approach to modelling their potential impacts.

2.1 The reforms modelled

Based on engagement with the Scottish Government and COSLA, this report analyses five different hypothetical reforms. Full information is provided in the Table 2.1, but in summary, the reforms analysed are:

• A pure revaluation. All properties are revalued based on up-to-date (Q3 2024) values, and assigned to one of eight bands. Band thresholds are set so that the same fraction of properties are in each band as now across Scotland as a whole.
• A revalued more differentiated 12-band system. All properties are revalued based on up-to-date values, and assigned to one of 12 bands. This allows for finer gradation in bills, and for a slight reduction in regressivity at the very bottom and very top of the property value distribution.
• A revalued less regressive 12-band system, which uses the same revalued bands to differentiate bills more, but goes further to reduce the regressivity of the current tax rate structure.
• A revalued less regressive 14-band system, which further differentiates tax bills while reducing the regressivity of the current tax rate structure.
• A revalued continuous, proportional system. All properties are revalued based on up-to-date values, and face a gross tax bill (before any exemptions, discounts and premiums) that is proportional to their property values. For example, a property with an estimated value of £300,000 faces a gross bill 3 times as high as a property with an estimated valued of £100,000.

Table 2.1 shows (for the banded systems) the band thresholds, the relative tax rate we assume applies to each band (where the relative tax applied to a Band D property is 1), the tax rate (gross tax bill) in each band if a council set its Band D rate at the current Scottish average (£1,543), and the share of properties that would be in each band across Scotland as a whole.

The table shows that a pure revaluation would keep the current tax structure as now, except that the tax bands’ upper thresholds would increase: for example, approximately 3.6-fold for Band A (from £27,000 to around £97,000), 4.4-fold for Band D (from £58,000 to around £255,000), and 4.2-fold for Band G (from £212,000 to around £892,000).

The table shows that the 12-band differentiated system introduces a new band at the bottom (A1), splits up bands E and G into more bands (E1, E2, G1, G2) and adds a new band at the top (I). Splitting up bands E and G allows for smaller jumps in tax rates than currently apply between bands D and G. The new Band A1 and the slightly lower tax rates for bands A2 to C, together with slightly higher tax rates on bands G2 and H and a new Band I, slightly reduce the regressivity of the current council tax system. If a council set its Band D rate at the current Scottish average, compared with a pure revaluation the lowest-value properties (in Band A1) would see a reduction in their standard gross bill of around £170 per year (from £1,029 to £857), with those in bands A2 to C seeing a reduction of around £50 a year compared to the equivalent bands in the current system. The very highest-value properties (in the new Band I) would see an increase in bill of £850 per year.

The 12-band less regressive system uses the same bands but further reduces the tax rates on bands A2-C and increases them on bands E and above, in order to do more to address the regressivity of the council tax system. If a council set its Band D rate at the current Scottish average, compared with a pure revaluation the lowest-value properties in Band A1 would again see a reduction in gross bill of around £170 a year (from £1,029 to £857), but those in bands A2 to C would now see reductions of £80 to £170 compared to the current equivalent bands (A to C). The very highest-value properties in the new Band I would see an increase in bill of around £1,600 a year.

The 14-band less regressive system adds two bands at the bottom rather than one (splitting B into B1 and B2 as well as A into A1 and A2) and further differentiates bills at the very top. If a council set its Band D rate at the Scottish average, compared with a pure revaluation the lowest-value properties in Band A1 would see a reduction in bill of around £260 (from £1,029 to £772), with those in bands A2 to C seeing reductions of around £130 to £200 compared to the current equivalent bands (A to C). Band K properties would see an increase in bill of around £2,750 per year.

Note: % of properties excludes garages that are on the existing council tax band list but for which values cannot be estimated using housing transactions values, and which are generally exempt from council tax. Tax rate shows the tax rate if this system applied this financial year and a council set the Scottish average Band D tax rate. The Band D rates councils would set in practice would vary from this and depend, for instance, on how grant funding was adjusted alongside council tax revaluation and reform, and on councils’ responses to both the new tax structure and grant allocations.

Source: Authors’ calculations using data from the Scottish Assessors, Registers of Scotland, and Energy Performance Certificates.

Note: See above. Source: See above

As we shall see below, though, its overall distributional effects are very similar to the 12-band less regressive system.

The continuous proportional system we model is not shown in Table 2.1 as it has no bands. A property with a value of £207,000 would face the same gross tax bill as a current Band D property, which we again refer to as a relative tax rate of 1. That means a property worth half of that, £103,500, would face a relative tax rate of 0.5, compared with 0.778 under a pure revaluation, 0.75 under the 12-band differentiated system, 0.667 under the 12-band less regressive system and 0.639 under the 14-band less regressive system. Conversely a property worth £414,000 would face a relative tax rate of 2, compared with 1.625, 1.725, 1.85, and 2 under the other systems. And a property worth £2,070,000 would face a relative tax rate of 10, compared with 2.45, 3, 3.5 and 4.222, respectively. This shows that the less regressive 12-band and 14-band systems would move closer to a proportional system for the bulk of the property value distribution, but would still be very far from the proportional system for the highest value properties.

When analysing the impact of these reforms across council areas and neighbourhoods, we focus on changes in bands and the relative tax rates applied to different bands (where a Band D property has a relative tax rate of 1 as in Table 2.1), as opposed to changes in tax bills. For our household-level analysis, we model the impact of each reform on tax bills on a revenue-neutral basis: that is, after accounting for any changes in the cost of exemptions, discounts, premiums and the means-tested CTRS, net council tax revenue and the average net bill across Scotland as a whole under the reform systems is the same as under the current council tax system.^[3] This allows us to examine the effect of revaluation and reform separately from any decisions over how much or how little to raise from council tax overall. Our remit in this report was to model the reforms on a revenue-neutral basis across Scotland as a whole – although reforms would not be revenue-neutral for each individual council area. Note that it is the tax rates set by local councils that ultimately determine council tax revenue, and would therefore determine whether reforms would be revenue-neutral in practice. But if grant funding levels were maintained, it would be possible for councils to maintain spending while setting tax rates that are revenue-neutral across Scotland as a whole.

Transition and mitigation measures

In addition to modelling reforms to the council tax rate structure, we also consider a number of transition and mitigation measures, which are discussed in detail in Section 6 of this report. We are able to model the following quantitatively:

• Transitional arrangements that phase in sizeable increases in bills gradually.
• An enhanced council tax reduction scheme for ‘cash-poor, asset-poor’ households whose incomes are too high for them to qualify for CTRS under current rules, and who face increased bills as a result of increases in tax rates applied to bands.

A lack of suitable data prevents us from accurately modelling a deferral scheme (allowing ‘cash-poor, asset-rich’ households facing large increases in bills to defer paying them until, for example, a property is sold or its owner dies). But key considerations for such a scheme are discussed in Section 6.

2.2 Approach and methodology summary

Our report considers the potential impacts of council tax reform both across different places and across different types of households.

Geographical analysis

Our place-based analysis considers the impact of the reform options on the share of properties in different tax bands and facing different relative tax rates in different council areas and different small neighbourhoods (technically ‘data zones’).

In order to estimate these impacts, the key step is to estimate an up-to-date property value for each property in Scotland. Full methodological information is provided in a separate methodology appendix (Appendix A), but in summary:

• We utilise data from three main property-level data sources: (1) Scottish Assessors data on all residential properties in Scotland, which tells us each property’s existing council tax band as well as some characteristics; (2) energy performance certificate (EPC) data, which provides further information on property characteristics (such as size, build date and energy efficiency) for a (large) sub-set of properties; and (3) Registers of Scotland data on the date and purchase price for properties that have transacted. We also use data on a range of local area characteristics, including deprivation and rural/urban classification, as well as the neighbourhood (‘data zone’ or ‘intermediate zone’ depending on sample size) a property is located in.
• Using those properties that have transacted and that we are able to match between all of the above datasets, we use regression analysis to estimate the relationship between property transaction price and various explanatory variables such as the date of transaction, property characteristics, local area characteristics, and specific geographic location.^[4]
• Using the estimated relationships, we then predict a Q3 2024 property value for all properties we are able match between the Scottish Assessors and EPC data – approximately 1.45 million properties, or 56% of all properties in the Assessors data (excluding garages). To make this fully representative, we then weight this data so that the number of properties by tax band and property type in each post code district (e.g. EH1) and council area matches the totals reported in Scottish Assessors data.
• Once we have an estimated up-to-date value for each property, we assign properties to the chosen tax bands and corresponding tax rates described in Section 2.1. We can then estimate the number of properties in each area that would fall in each band, and face the associated relative tax rate, under each alternative council tax structure.

It is important to note that we do not estimate the changes in net council tax bills that households in different councils or neighbourhoods would face. In order to do that accurately, the analysis would need to account for which properties are subject to different discounts, exemptions, premiums and the CTRS, and to account for how grant funding would be adjusted alongside any updated council tax system.

In particular, information on discounts, exemptions, premiums and the CTRS would be needed because they affect households’ net tax bills, and in turn the tax rates councils need to set to raise a given amount of net revenue. We do not currently have data on these for all properties in Scotland – only the survey sample used for the household-level analysis discussed below.

It would also be important to account for grant adjustments because, as highlighted by Adam et al. (2020a, 2020b), it is this (as well as councils’ tax-and-spending preferences), rather than the council tax system itself, which is the main determinant of how average council tax bills vary across councils. There are a range of different options for adjusting grants to account for council tax revenue-raising capacity (and for the impact of revaluation and reform on that capacity), but a concrete example can help to explain its importance for determining the impact of revaluation and reform. Suppose for instance, that councils wanted to maintain spending following reform. In that case, if grant funding were not adjusted, then each council would need to raise the same amount via council tax as in the absence of reform and so charge the same average council tax bill as in the absence of reform. Thus, in such circumstances, a council seeing a fall in its tax base as a result of reform would have to set a higher Band D rate to offset this fall and leave average tax bills unchanged, and vice versa. As a result, revaluation and reform would still lead to a big redistribution of tax bills across individual households within a council area (e.g. within Edinburgh City), but it would not redistribute tax bills across council areas (e.g. between Edinburgh City and Glasgow City). The more grant funding was adjusted to reflect changes in different councils’ tax bases as a result of revaluation and reform, the more the reforms would redistribute tax bills across council areas as well as within them. In reality, councils may or may not wish to maintain spending if the council tax system were reformed. Moreover, while the grant funding the Scottish Government provides to councils currently takes partial account of their existing council tax bases other approaches would be possible (Scottish Government, 2025f). But this example illustrates the importance of grant funding decisions to the impact of revaluation and reform of council tax across places.

Collating disaggregated data on discounts etc. and making decisions on grant funding are beyond the scope of this project. Our geographical analysis therefore focuses on the share of properties in each band by council area and local area, and the average relative tax rates applied to properties by council area and local area given these tax bands, rather than on council tax bills. Further analysis that accounted for discounts etc. and grant funding decisions could be undertaken in future to allow estimation of impacts of reforms on average council tax bills by councils and local areas.

Household-level analysis

Our household-level analysis considers the impact of the reform options on the gross and net council tax bills of a survey sample of Scottish households, weighted to be representative of Scotland as a whole. From this we calculate the proportion of households that would see bills increase or decrease by different amounts as a result of the different reform options, as well as average changes in bills for different household types. This includes groups defined based on household income, the age of the oldest adult in a household, housing tenure, and so on.

Again, the key step is to estimate an up-to-date value for each property. Full methodological information is provided in a separate methodology appendix (Appendix A), but in summary:

• Our main source of data is from waves 8 to 10 and 14 of the Understanding Society household panel survey, covering households interviewed between 2016 and 2019, and between 2022 and 2023.^[5] This contains information on household demographics and incomes, various property characteristics, existing council tax band and, for those properties which are owner-occupied, the owner’s self-reported estimate of their property’s value. We combine this with publicly available data on a range of local area characteristics, including deprivation and rural/urban classification, from the Scottish Index of Multiple Deprivation (SIMD) 2020 and the Scottish Government Urban Rural Classification 2020 (Scottish Government, 2020; Scottish Government, 2022).
• For owner-occupied properties, we uprate self-reported property values to Q3 2024 values using council- and property-type-level changes in property values according to the UK house price index.
• We use regression analysis to estimate the relationship between (uprated) self-reported property value and various explanatory variables such as the survey date, property characteristics, local area characteristics and specific geographic location.
• Using that estimated relationship, we then predict a Q3 2024 property value for all other properties (which are mainly rental properties).
• Once we have an estimated up-to-date value for each property, we can assign properties to the chosen tax band and rate structures described in Section 2.1. Given the information on other household characteristics in the survey data, we then use the IFS tax and benefit microsimulation model, TAXBEN, to estimate each household’s net council tax liability (accounting for single-person discounts, CTRS entitlements, etc.) under both current and alternative council tax systems and look at the resulting patterns of bill changes by household characteristics.

Our household-level modelling assumes that, in approximate terms, grant funding would be fully redistributed between councils to account for the fact that the amounts they could raise under the reformed systems for a given tax rate (their ‘tax base’) would change in different ways.^[6] If the grant funding was only partially adjusted, or not adjusted at all, this would change the impacts across groups, but to a lesser extent than for the geographical analysis.