How to pay for a Minimum Income Guarantee

Annex II: Modelling explanatory note

IPPR Tax Benefit Model

The IPPR tax benefit model is owned by the Institute for Public Policy Research and is maintained by researchers at Manchester Metropolitan University. Analysis using the IPPR tax benefit model is completed by WPI Economics on behalf of the Centre for Social Policy Studies (CSPS).

This IPPR tax benefit model is a static microsimulation model that, working off household-level survey microdata, can calculate the tax liability for each household, benefit unit, and family. These estimates are then grossed to give population level estimates of tax liability. The model allows the user to change the details of the UK’s tax and benefit system. By comparing these scenarios to a baseline, calculated under the existing system of taxes and benefits, the fiscal and distributional impacts of policy reforms can be estimated.

Modelling of scenarios with projections of future financial years (NICs and council tax) are based on the IPPR tax-benefit model uprating of the 2022-23 Family Resources Survey microdata. This uprating is completed using a variety of uprating series: for example, wages, rents and prices. These are derived from Office for Budget Responsibility and Bank of England macroeconomic forecasts.

Differences between FRS and LCFS

We have drawn on two different survey datasets for our policy modelling. For the modelling of income-based taxes (Scottish Income Tax, NICs and council tax) we have made use of the Family Resources Survey (FRS), which provides data on the incomes and living circumstances of households and families in the UK. For the modelling of expenditure-based taxes (VAT and excise duties) we have used the Living Costs and Food Survey (LCFS), which provides data on household-level expenditure across the UK.

Some differences in these surveys have been reflected in our distributional modelling. For example, FRS samples more Scottish households than LCFS, and thus distributional modelling using FRS records more sample households in each decile than distributional modelling using LCFS for expenditure taxes.

It is also important to note that the LCFS data is explicitly at the household level, while FRS contains both population and household-level data. Because of this, while we have provided the number of the population in each decile for the FRS-based modelling, we have only provided the number of households in each decile for the LCFS-based modelling.

Income deciles

Income deciles have been determined by dividing all households into equally-sized groups based on total household disposable income before housing costs. Income deciles are calculated based on Scottish households only. The deciles therefore contain roughly the same number of Scottish households, by design.

The slight differences in the number of households in each decile reflect technicalities with how different elements of the population are weighted in the LCF and FRS surveys, but they remain the closest approximations available in the survey data to the actual income deciles, where in reality each of these will contain the exact same number of households.

Council tax modelling decisions

There were some occasions where we have taken modelling decisions beyond the detail of the specific policies in order to enhance the quality of the data.

In the case of the council tax modelling, we made the decision to assume an increase in the average Band D tax rate in accordance with projected inflation in the baseline data (in other words, where no additional council tax policy is implemented), and have assumed this inflation also carries over into the counterfactual data (namely, where the new council tax policies have been modelled), alongside the new increases in council tax. This results in data which is more likely to reflect what would actually occur if this policy were implemented.

The council tax policy being modelled proposes real-terms increases in council tax rates as specified. In order to model these increases within the current council tax system, we have in each case calculated the council tax rate in real-terms for each band. In the current system, council tax band rates are represented as ratios with respect to the Band D tax rate. We have uprated the rate according to the specified policy, and then re-represented it as a ratio in terms of a new uprated Band D tax rate. In each case, we are also uprating the tax bands in line with inflation, to make inflation increases consistent across all tax bands.