Introduction

This publication provides estimates of the effect of some of the factors contributing to growth in the number of claimants in the Employment and Support Allowance Support Group (ESA SG), or receiving Universal Credit (UC) who are deemed to have Limited Capability for Work and Work-Related Activity (UC LCWRA).

Background

The number of working age people in receipt of incapacity benefits – Employment and Support Allowance (ESA) or Universal Credit (UC) with a health condition – increased from 2.6 million in 2019 to 2020 to 3.3 million in 2023 to 2024^{[footnote 1]}. The increase in those with UC LCWRA or in the ESA SG over the same period was from 1.85 million to 2.5 million.

The Office for Budget Responsibility in its October 2024 Welfare Trends Report^{[footnote 2]} discusses the recent growth in these benefits and the factors behind that growth.

The analysis presented in this document is an initial attempt to quantify the contribution of some of the factors behind that growth. Specifically, the contribution of demographic change, changes in state pension age, and certain structural features of the UC system relative to the system it replaces. The analysis focuses on those UC health claimants in the LCWRA group, and their equivalents in ESA. These accounted for 76 per cent of UC health / ESA claimants in 2023 to 2024.

As a foundation for the analysis, this document also presents for the first time UC health caseloads for May 2018; the published statistics starting from April 2019, by which time all claimants had been moved onto the online full-service system.

This analysis is likely to develop further in the future, for example if it is possible to narrow down the extent of unattributed change, as well as taking into account more recent data. This means that estimates presented here are subject to revision in the future.

Methodology

The modelling is based on case-level UC and ESA data, extracted from the DWP’s computer systems. This does not include any information that would enable individuals to be identified. The analysis here requires age, sex and stage and type of benefit, for both benefits, and information on how the UC award is calculated.

As noted above, the focus here is on UC health claimants with Limited Capability for Work and Work-Related Activity, along with ESA Support Group claimants. It covers the period May 2018 to May 2023. Note that in May 2018 not all Jobcentres were able to offer the full range of claimants access to UC, though this is not a relevant consideration for this analysis.

During this period the number of ESA claimants with means-tested benefit declined, while UC health numbers increased, through a combination of natural turnover of the caseload, and changes of circumstances that moved people from ESA to UC. Very few ESA claimants were sent migration notices under the Move to UC process during this period.

The modelling of the various factors is undertaken sequentially. This avoids double-counting claimants that receive UC, but would not have received ESA, for more than one reason, but it does mean that the number of claimants attributable to a particular factor is sensitive to the order chosen. The overall combined effect of the different factors is, however, not affected by the order in which the modelling is undertaken.

Structural changes

The effects of the following structural changes have been estimated, in the order stated:

• claimants can receive both UC and contributory New-Style ESA, so this double counting is removed in the model, accounting for 102,000 claimants in May 2023. Published headline figures on incapacity benefit caseloads, such as in the DWP’s Benefit Expenditure Tables, already take account of the receipt of both benefits and avoid double counting

• UC is available in work, whereas income-related ESA was only available to those working fewer than 16 hours a week, so we estimate the additional claimants who are on the health journey, in work, who would not have been eligible for ESA. This is done by removing those who have earnings – corresponding to around 56,000 claimants in this group by May 2023. This will include some recipients of Statutory Sick Pay, which is indistinguishable from other earnings paid by the employer. This will slightly overestimate the effect of structural change, as some ESA claimants will also have had earnings, but there is no data readily available to take account of them

• individual claimants in a couple are counted separately in UC, so where there are two UC health claimants they are counted twice, but under means-tested ESA would usually have been counted once^{[footnote 3]}. This accounts for 17,000 claimants in May 2023 who are counted as UC health, but who would not have been counted under ESA

• UC brings in housing support, which in the legacy system is covered by Housing Benefit. A disabled claimant in a household with an income too high to qualify for means-tested ESA (for example due to having a working partner or pension income) might, if their household income was low enough, still qualify for UC once their housing entitlement is added in. Such households would be counted as receiving UC health now, but wouldn’t have been counted on ESA in the legacy regime. We exclude these cases by removing those whose UC award is lower than their housing element. This corresponds to around 11,000 claimants by May 2023

• UC also includes child elements, which were part of tax credits in the legacy system. Similar to housing elements, there is a range of income where a household would qualify for UC – and therefore be counted in UC health if they were ill or disabled, but would not have qualified for ESA. We exclude these cases by removing those whose UC award is lower than their combined housing and child element. This accounts for 4,000 claimants in May 2023

• where the ages of a couple straddle state pension age, they now claim UC until both reach state pension age, when they can claim Pension Credit. Under the ESA system, claimants could move on to Pension Credit when the older member of the couple reached state pension age. Claimants aged over pension age have therefore been removed, accounting for 9,000 in May 2023^{[footnote 4]}

As the decomposition compares May 2023 with May 2018, the 6,000 claimants across these categories in 2018 are netted off the figures shown above to determine the contribution of structural factors to growth over this period.

Structural factors we have not modelled here are:

• the switch to UC also simplified the amounts paid to health claimants. Claimants with the LCWRA element in UC receive an amount that is more than a long-term ESA SG claimant that does not have a Severe Disability Premium, but an amount that is less than ESA SG claimants that do have a Severe Disability Premium. It is possible that a household with income within a certain range could move on or off UC health, compared to ESA, as a result of this change

• any behavioural effects resulting from the different structure of UC compared to ESA. For example, the Office for Budget Responsibility (OBR) Welfare Trends Report discussed potential behavioural responses to the change in the differential between being in the LCWRA group and being in Intensive Work Search, as well as the removal of the element for people with Limited Capability for Work

State Pension Age

State Pension age increased from 65 to 66 between November 2018 and October 2020. We estimate the impact of this change as being the increase in the number of 65-year-olds over the full period, after the modelled structural changes above. This adds 78,000 to the caseload in May 2023.

In May 2018 women’s state pension age was around 64 years and 8 months, as equalisation of men’s and women’s pension ages was not complete. In May 2018 there were just over 20,000 64-year-old women receiving ESA SG / UC LCWRA, rising to 31,000 in November 2018, when equalisation completed. So, equalisation adds 11,000 to the caseload in each year from 2019 onwards.

Demographic changes

This stage takes the individual-level data from the structural modelling and disaggregates it by single year of age and sex. Using Office for National Statistics (ONS) population estimates, the prevalence of benefit claiming is calculated for 2018. The timing gap between May 2018 benefit data and June 2018 population estimates is unlikely to materially affect our findings.

These benefit prevalence estimates are then applied to the relevant population^{[footnote 5]}, by single year of age and sex, to estimate the growth in claimants we would expect simply from population change.

Results

The table shows the results of the modelling – decomposing the growth in the UC LCWRA and ESA SG caseload to the various factors identified above.

Factors contributing to the increase in Universal Credit Limited Capability for Work and Work-Related Activity and Employment and Support Allowance Support Group caseloads, May 2018 to May 2023 (caseload in May of year in thousands)

	2018	2019	2020	2021	2022	2023
Employment and Support Allowance – Support Group	1,589	1,574	1,544	1,480	1,405	1,338
Universal Credit – Limited Capability for Work and Work-Related Activity	29	157	392	596	864	1,179
Less dual ESA / UC claims	4	17	34	56	78	102
Total ESA SG / UC LCWRA	1,614	1,714	1,903	2,021	2,191	2,415
Growth compared to 2018	Not applicable	100	289	407	576	801
Effect of UC structural change compared to ESA (a-e)	0	6	19	30	66	97
a. Claimants in employment	0	4	11	14	38	56
b. Two claimant couples	0	1	4	6	11	17
c. Housing costs	0	1	2	3	7	11
d. Child elements	0	1	1	1	3	4
e. Mixed-age couples	0	0	1	5	6	9
Effect of State pension age increases	0	32	64	79	84	89
ESA SG / UC LCWRA net of above consequences of policy change	1,614	1,676	1,820	1,912	2,041	2,230
Effect of demographic change	0	9	19	32	44	57
ESA SG / UC LCWRA net of policy and demographic change	1,614	1,667	1,800	1,880	1,997	2,173
Growth compared to 2018	Not Applicable	52	186	266	383	559
Growth due to identified policy and demographic change	Not Applicable	47	102	141	193	243
Share of total growth	Not Applicable	47%	35%	35%	34%	30%

The table shows:

• between May 2018 and May 2023 the caseload on the higher rates of UC health and ESA increased by just over 800,000, of this:
  • 97,000 (12%) is because UC rolls six benefits into one. People who were not eligible for ESA, but might have received another benefit that UC is replacing, are now counted in UC health
  • 89,000 (11%) is because of changes to state pension age
  • 57,000 (7%) is because of demographic change, with the population getting older, on average

• therefore, 30% of the increase in caseload on the higher rates of UC health and ESA can be attributed to these factors, which were predictable in 2018 and before, and included in forecasts

• in the earlier years of the analysis the proportion of the increase in caseload that can be attributed to these factors is somewhat higher, as the changes in state pension age were complete by the end of 2020, but the overall growth in the caseload has been more steady

• however, even after taking out the effects of all these factors, the number claiming the higher rates of UC health and ESA increased by almost 560,000 over those 5 years – or 70% of the overall observed increase

The factors underlying the 560,000 increase have been covered by many different organisations’ publications, in particular the OBR’s October 2024 Welfare Trends Report, but quantifying the impact of each of these different factors will be more complex, if it can be done at all, and is not undertaken in this analysis.

About this analysis

The figures presented in this release are based on internal analysis and modelling using Universal Credit and Employment and Support Allowance case-level administrative data, combined with ONS mid-year population estimates. As these are outturn figures, using 100 per cent of claimants in receipt of these benefits, the level of uncertainty in the analysis is low.

Statement of Compliance with the Code of Practice for Statistics

The Code of Practice for Statistics (the Code) is built around 3 main concepts, or pillars:

• trustworthiness – is about having confidence in the people and organisations that publish statistics
• quality – is about using data and methods that produce statistics
• value – is about publishing statistics that support society’s needs

The following explains how we have applied the pillars of the Code in a proportionate way.

Trustworthiness

DWP analysts work to a professional competency framework and Civil Service core values of integrity, honesty, objectivity, and impartiality. The analysis in this release has been scrutinised and received sign off by the expert lead analyst.

We protect the security of our data in order to maintain the privacy of the citizen, fulfil relevant legal obligations and uphold our obligation that no statistics will be produced that are likely to identify an individual, while at the same time taking account of our obligation to obtain maximum value from the data we hold for statistical purposes. All analysts are given security training and the majority of data accessed by analysts is obfuscated and access is business case controlled based to the minimum data required.

The figures have been seen in advance by Ministers and officials, in line with the Code, where pre-release access does not apply for an ad hoc analysis release.

Quality

The data which underpins this information is taken directly from departmental systems, which are relied upon by DWP for the administration and operation of Employment and Support Allowance and Universal Credit.

Quality assurance has taken place in line with the standards usually applied to DWP ad hoc releases, with an internal check that the results shown are robust, and checked against published data.

Value

Releasing this information serves the public interest regarding the growth in receipt of incapacity benefits, separating the known factors arising from policy and demographic change, which have been known about and included in forecasts for several years, from the less predictable factors. This helps ensure that the interpretation of published statistics is appropriate and the limitations of those figures are understood.

Contact information

For press enquiries contact DWP Press Office: 0203 267 5125