Spot the difference: why do JRF and OBR forecasts for incomes differ?
We explain why our approach to income forecasts is more comprehensive and will more closely reflect the actual experience of households’ living standards.
- Written by:
-
Peter Matejic
- Date published:
- Reading time:
- 9 minutes
1. Introduction
JRF analysis suggests incomes in quarter 4 of 2029 will be lower than in quarter 4 of 2024, while the Office for Budget Responsibility (OBR) analysis suggests incomes will be higher. This briefing reconciles JRF and OBR methodologies. It shows that when JRF’s After Housing Costs (AHC) methodology is adjusted to a Before Housing Costs (BHC) version, it is broadly consistent with an adjusted version of OBR’s published analysis. JRF’s approach is more comprehensive and will more closely reflect the actual experience of households’ living standards.
2. The importance of understanding the living standards outlook
Assessing the prospects for living standards is an important part of every fiscal event, but will be increasingly important going forward, given the Prime Minister’s commitment to ‘higher living standards in every region of the country’ as a milestone for this Parliament. One of the most commonly used measures of living standards is household income.
JRF analysis following the Autumn Budget, which uses a more comprehensive measure of household income than the Office for Budget Responsibility (OBR), shows the Government is not on track to meet this milestone. JRF usually uses a measure of income after housing costs have been deducted, as rising housing costs will affect household living standards. This briefing shows how looking at the JRF analysis on a BHC basis is broadly consistent with OBR analysis using the National Accounts’ Real Household Disposable Income measure (RHDI), when a series of adjustments are made to that measure to bring the 2 series into as close alignment as possible.
3. Key differences between JRF and OBR analysis
There are 5 key differences in the methodology used to construct the JRF and OBR household income projections, which are summarised in the table below:
| Area | OBR | JRF |
|---|---|---|
| Housing costs | Before Housing Costs only. | Analysis is on an After Housing Costs basis, but Before Housing Costs analysis is also possible. |
| Denominator | Two series are published: the first is income per person and the second is per person aged 16+. | Per household. |
| Deflator | National Accounts household consumption deflator. | Consumer Price Index. |
| Income concept | Resources-based disposable income for households and non-profit institutions serving households (Before Housing Costs). | After Housing Costs household income. |
| Data source | National Accounts. | Family Resources Survey. |
4. Aligning the 2 methodologies
The aim of this analysis is to produce as comparable series from the National Accounts and FRS data as we can. We know the Family Resources Survey suffers from problems common to social surveys, such as non-response and undercounting of different income sources. The IPPR model compensates for some of these issues, but we would not expect total agreement given the different data sources, even after adjusting to align concepts.
Our approach for each of the first 4 issues are as follows:
- Housing costs: because the OBR analysis is on a Before Housing Costs basis, we have looked at a Before Housing Costs version of the JRF series for comparability.
- Denominator: there is a reasonable difference between the 2 series the OBR publish, which are real disposable household income per person and real disposable household income per person aged 16 or over. This is because there is a forecast reduction in the number of children in the population. Updated household projections are not yet available, and the latest available household projections are based on the situation in 2018 when migration estimates would have been much lower, so converting OBR data to a per household basis using these would not be correct. The income per individual aged 16 or over is probably the closest to the JRF analysis, given the falling number of children is unlikely to lead to a proportionate reduction in the number of households, so we look at the OBR RHDI per individual aged 16 or over series.
- Deflator: the Consumer Price Index (CPI) deflator is much more commonly used in living standards than the household consumption deflator in the National Accounts, so adjust the OBR series from using this to instead be CPI deflated.
- Income concept: we look at income components in both adjusted measures to see what makes up incomes using the OBR and JRF series.
The analysis looks at the picture in quarter 4 for each year, comparing changes in different periods to quarter 4 2024.
5. How do adjusted estimates compare?
Directly comparing OBR and JRF estimates without accounting for any differences gives an apparently contradictory picture, with the OBR (using a Before Housing Cost measure) predicting income growth of 1.8% over the period Q4 2024 to Q4 2029 and JRF (using an After Housing Cost measure) showing income falling by 1.8% over the period.
Looking at the adjusted series, we get a much more similar picture of a 0.6% increase in income over the period quarter 4 2024 to quarter 4 2029 according to OBR, when we look at incomes Before Housing Costs (BHC) and deflate using CPI, compared to a 0.4% reduction using JRF’s BHC series, meaning the adjustments explain around 3 quarters of the difference. Further analysis shows the growth in the OBR series is driven by non-Labour market income, which is mainly from property and imputed rents, which won’t fully be captured as income in the JRF analysis.
The remainder of this briefing goes through each issue to look at the difference being made.
Adjustment 1: the effect of housing costs on the assessment of changes to average incomes
Moving the JRF estimates to a Before Housing Cost basis shows that much of the reductions in income in the JRF analysis are due to housing costs, with only a 0.4% reduction on a Before Housing Cost basis, explaining around 40% of the difference between the series. Given the importance of housing costs for household finances, it is important to account for the effects of rising housing costs on living standards for a comprehensive assessment of changing living standards, which is why we choose to look at incomes after housing costs have been subtracted.
Adjustment 2: the effect of looking at income per person aged 16 or over, compared to per individual
To be entirely comparable with JRF analysis, we would want the OBR series to be on a per household basis, but for that we need a household population projection. The OBR does not produce this, and the latest official household population projections are based on the situation in 2018, which will not be an accurate reflection of the current position, given increased migration and other changes since then.
Moving the OBR series from an income per person to a per 16+ individual reduces the growth shown by the OBR analysis between 2024 and 2019 from 1.8% to 0.7%. The difference is because, according to OBR projections, the population of under 16s will fall from 13.9 million in Q4 2024 to 13.6 million in Q4 2029, at the same time as the population of people aged 16 and over will rise from 55.4 million in Q4 2024 to 57.5 million in Q4 2029. This means the growth in income per person aged 16 or over is lower than the growth per person, because of the difference in the denominator, with a 2.6% growth in the overall population compared to a 3.8% growth in the 16+ population over the period Q4 2024 to Q4 2029.
Adjustment 3: the effect of changing from a National Accounts derived deflator to using the Consumer Price Index
The OBR use a household consumption deflator based on National Accounts data, while JRF uses the much more commonly used Consumer Price Index. Moving the OBR series to be CPI-deflated only has a small impact on the change between 2024 and 2029, reducing the growth from 0.7% to 0.6%, although it does have a bigger effect on the period from 2019 to 2024, reducing the differences between the OBR and JRF series over that period.
Chart 2 which compares a JRF-based analysis with the OBR-based analysis on as consistent a basis shows a broadly consistent picture between the two series, with incomes falling between 2019 and 2022, before rising to around the year 2024 or 2025, then falling for a period to around the year 2027 or 2028, and rising at the end of the forecast.
Income growth after accounting for inflation is anaemic on either basis, growing just 0.6% over the 5 years from Q4 2024 to Q4 2029 according to OBR (close to just 0.1% a year) and actually being slightly lower at the end of the period than at the start according to JRF analysis. Bringing in rising housing costs over the period would mean both series would show falling living standards.
Remaining differences: which components of income are driving income growth between Q4 2024 and Q4 2029
We can look at what broad components of income are driving changes, to see whether JRF and the OBR agree what is acting to increase or decrease incomes. The 2 data sources do not fully agree on the importance of different income sources, with labour-market income more important and other income less important in the JRF series compared to the OBR. Taxes exceed benefits on average in the JRF series, but it is the other way round in the OBR series.
The 2 series do however agree that labour-market income changes are not important as a driver of income growth until 2029, and that the net effect of taxes and benefits acts as a drag on household income, with the big difference between the 2 series coming from ‘Other income’, which is driving the increase in the OBR series. This includes imputed rents and property income, which won’t fully be captured as income in the JRF series, and increases here will reduce some households’ incomes After Housing Costs if they are reflective of rising rents and mortgages.
6. Conclusion
By making a series of adjustments to the National Accounts-based OBR Real Household Disposable Income to bring it as close as possible to JRF’s Before Housing Costs income analysis, we have shown both methodologies agree on at best very anaemic income growth between Q4 2024 and Q4 2029 after adjusting for inflation, with incomes falling after rising housing costs are taken into account. The fact that the more comprehensive measure that JRF uses, which is a household-level analysis after taking account of rising housing costs and excludes imputed rents shows an even worse picture, makes this even more worrying.
Note
The JRF modelling in this report uses the Family Resources Survey (2022/23) and the IPPR Tax-Benefit Model (version v02_72) to project household income, housing, and tax expenditure for October of each forecast year. The model takes the base survey data to a given month (October in this instance) and then applies inflation and the known or anticipated tax and benefit policy regimes to project household incomes and tax liabilities in future years.
How to cite this briefing
If you are using this document in your own writing, our preferred citation is:
Matejic, P (2025) Spot the difference: why do JRF and OBR forecasts for household incomes differ? York: Joseph Rowntree Foundation.
This briefing is part of the cost of living topic.
Find out more about our work in this area.