Headline evidence · Benefits · Great Britain · 1994–95 to 2024–25 · includes the State Pension

Welfare spending rose 76% over 30 years, after inflation. Where does the money go?

Benefits cost £289bn in 2024–25, about 10% of national income. Since 2019–20, most of the rise has been in support for working-age people, including disability and health benefits. Up to £2.4bn of Pension Credit also went unclaimed in 2024–25.

Proposed outcome: Help people get the support they need, weighing living costs, work and what the public can afford. What could change →

Geography
Great Britain
Evidence period
1994–95 to 2024–25
Evidence status
Site calculation from published data
Editorial review
Reviewed 22 October 2026

The short version

How much has spending changed?

76% more welfare spending after inflation

Great Britain welfare spending in real terms: total, pensioners, and working-age people and children.

Great Britain · 1994–95 to 2024–25

1994–95 to 2024–25 · £bn, 2026–27 prices · observed figures; gaps and definition breaks retained.

£bn, 2026–27 prices£0£100£200£300£400199520052015£320.8bn£173.7bn£147.1bn
Spending rose from £182.0bn to £320.8bn in the same prices. The total includes both groups; do not add the three lines. More residents, changing needs, eligibility and payment rates can all affect spending. This does not measure the buying power of an individual benefit.

Great Britain · 1994–95 to 2024–25 · Official statistics; observed figures only

Show the workings — 76% more welfare spending after inflation
Geography:
Great Britain
Measure:
£ billion, 2026–27 prices
Period:
1994–95 to 2024–25
Status:
Official statistics; observed figures only
Calculation:
DWP GB welfare real-terms total; growth = (2024–25 ÷ 1994–95 − 1) × 100. GDP deflator, 2026–27 prices.
Not included / breaks:
Northern Ireland excluded. DWP GB welfare definition includes HMRC welfare. Separate temporary Cost of Living payments are excluded. Forecasts are excluded.

Source: DWP, Benefit expenditure and caseload tables 2026, GB welfare; State Pension (released 14 April 2026).

View the data as a table
Financial yearTotal (£bn)Pensioners (£bn)Working age and children (£bn)
1994–95182.0291.7590.27
1995–96184.4492.5591.89
1996–97185.4694.3691.10
1997–98187.2597.7989.46
1998–99189.04100.4788.57
1999–00195.60105.3390.26
2000–01204.65110.5294.13
2001–02212.95116.2796.69
2002–03217.10117.8299.28
2003–04232.06120.50111.57
2004–05239.34125.42113.92
2005–06244.42129.35115.07
2006–07245.44129.83115.61
2007–08255.07135.99119.08
2008–09268.22142.38125.83
2009–10290.49150.25140.24
2010–11296.07153.43142.64
2011–12299.13156.02143.11
2012–13305.09161.70143.39
2013–14294.14159.58134.55
2014–15296.17162.23133.94
2015–16297.67164.08133.59
2016–17293.13162.96130.18
2017–18293.20163.15130.05
2018–19290.34162.50127.84
2019–20288.06161.23126.82
2020–21296.08155.67140.41
2021–22295.45159.06136.39
2022–23282.85153.64129.22
2023–24303.77165.30138.47
2024–25320.82173.69147.13

Scroll the table sideways to see every column.

Beside the rise: the share of the economy and who receives it

The same spending was 10.3% of UK GDP in 1994–95 and 10.4% in 2024–25: about the same share of the economy, a change of +0.1 percentage points, while the real-terms bill rose 76%. Pensioners’ spending rose from £91.8bn to £173.7bn (+89%) and spending on working-age people and children from £90.3bn to £147.1bn (+63%), all in 2026–27 prices. A rise or fall in spending is not good or bad in itself: it reflects how many people are eligible, payment rates and the size of the economy.

How large is the bill relative to the economy?

As a share of the economy, spending is about where it was 30 years ago

As a share of the economy, spending is about where it was 30 years ago

Great Britain · 1994–95 to 2024–25

1994–95 to 2024–25 · % of UK GDP · observed figures; gaps and definition breaks retained.

% of UK GDP0.0%5.0%10.0%15.0%19952005201510.4%
Great Britain welfare spending was 10.3% of UK GDP in 1994–95 and 10.4% in 2024–25. A larger economy can support a larger cash bill. This is the same GB spending definition as the headline, not a UK-wide benefits total.

Great Britain · 1994–95 to 2024–25 · Official statistics; observed figures only

Show the workings — As a share of the economy, spending is about where it was 30 years ago
Geography:
Great Britain
Measure:
% of UK GDP; GB spending numerator
Period:
1994–95 to 2024–25
Status:
Official statistics; observed figures only
Calculation:
As published, with units converted for display.
Not included / breaks:
The same Great Britain welfare definition as the headline (includes HMRC welfare and the State Pension; Northern Ireland and forecasts excluded). DWP also publishes a narrower benefit-expenditure share of GDP; it is on a different basis, so it is not placed beside this headline. A change in a share of GDP is in percentage points.

Source: DWP, Benefit expenditure and caseload tables 2026, GB welfare; State Pension (released 14 April 2026).

View the data as a table
Financial yearGB welfare as % of UK GDP
1994–9510.30
1995–9610.24
1996–979.95
1997–989.62
1998–999.42
1999–009.44
2000–019.46
2001–029.67
2002–039.64
2003–0410.00
2004–0510.08
2005–0610.01
2006–079.86
2007–089.93
2008–0910.69
2009–1011.92
2010–1111.86
2011–1211.89
2012–1311.96
2013–1411.26
2014–1511.03
2015–1610.85
2016–1710.45
2017–1810.14
2018–199.92
2019–209.78
2020–2111.34
2021–229.97
2022–239.31
2023–249.94
2024–2510.38

Where has the money gone?

Pensioners received just over half in both years

Pensioners received just over half in both years

Great Britain · 1994–95 and 2024–25

1994–95 · outturn · total £84.9bn

  • Pensioners: 50% (£42.8bn)
  • Working age and children: 50% (£42.1bn)

2024–25 · outturn · total £304.6bn

  • Pensioners: 54% (£164.9bn)
  • Working age and children: 46% (£139.7bn)
Pension-age benefits received 50.4% of spending in 1994–95 and 54.1% in 2024–25. These age groups divide the money, not the people. Disability, housing and in-work support overlap with the age groups and cannot be added as separate slices.

Great Britain · 1994–95 and 2024–25 · Official statistics; observed figures only

Show the workings — Pensioners received just over half in both years
Geography:
Great Britain
Measure:
% of GB welfare spending
Period:
1994–95 and 2024–25
Status:
Official statistics; observed figures only
Calculation:
As published, with units converted for display.

Source: DWP, Benefit expenditure and caseload tables 2026, GB welfare; State Pension (released 14 April 2026).

View the data as a table
Financial yearPensioners (%)Working age and children (%)
1994–9550.449.6
2024–2554.145.9

Three key measures

  • How much has welfare spending grown?

    76%higher after inflation

    The government paid £289bn in benefits in 2024–25, about a tenth of national income. Most of the rise since 2019–20 is support for working-age people, which includes disability and health benefits.

    Great Britain · 1994–95 to 2024–25 · Published statistics

    Definition, limits and source: How much has welfare spending grown?

    £182.0bn in 1994–95 to £320.8bn in 2024–25, in 2026–27 prices. Growth = (latest ÷ baseline − 1) × 100. Includes DWP and HMRC welfare; excludes Northern Ireland and separate temporary Cost of Living payments.

    Published statistics · DWP, Benefit expenditure and caseload tables 2026, GB welfare; State Pension

    Thirty years of welfare spending →
  • Survey-based evidence
    Family Resources Survey linked to administrative recordsFieldwork: Financial years ending 2013–2025; actual coverage variesEntitlement is modelled. Estimates have sampling ranges, method breaks and a missing year during the pandemic. Working-age estimates are Official Statistics in Development.

    How much Pension Credit went unclaimed?

    up to £2.4bnin 2024–25

    An upper estimate of Pension Credit left unclaimed by eligible families in 2024–25. The estimate has a range; working-age support is measured separately.

    Great Britain · 2024–25 (financial year ending 2025)

    Definition, limits and source: How much Pension Credit went unclaimed?

    DWP estimates that up to 900,000 pensioner families were entitled to Pension Credit but did not claim it in 2024–25; 62% of those entitled claimed (62% the year before). One year’s survey-based upper estimate. New working-age take-up estimates cover Universal Credit or equivalent benefits and are published separately.

    Survey-based · DWP, Income-related benefits: estimates of take-up, financial year ending 2025, PC1, PC2 and UC1

    Take-up, the benefit cap and hardship →

  • What share of welfare spending goes to pensioners?

    54%of spending

    The share of welfare money going to pension-age benefits, not the share of people receiving support.

    Great Britain · 2024–25 (outturn)

    Definition, limits and source: What share of welfare spending goes to pensioners?

    £164.9bn of £304.6bn in 2024–25 — a share of spending, not of claimants. Pensioner spending is DWP’s classification of benefits paid to people over State Pension age: the State Pension, Pension Credit, pension-age Housing Benefit, Attendance Allowance and other pension-age benefits. DWP's forecast for 2025–26 is 55%. The State Pension alone was £136.6bn.

    Published statistics · DWP, Benefit expenditure and caseload tables 2026 (Spring Forecast 2026), GB welfare summary table

    Where the money goes →
Different needs, overlapping groups

Support covers retirement, low income, disability, housing, caring and children. Universal Credit now records 8.4 million people, but most are not job-seekers: half have no work requirements — for example because of ill health, caring or a very young child, and 3.1 million are in paid work. Disability support is separate again — PIP covers extra costs and is not a work-status category.

Budget 2026 · 28 October 2026

What the Budget means for benefits

5

Budget measures on this topic

£195m

Net saving or gain, 2026–27

£10m

Net saving or gain, 2029–30

Biggest single-year effect of each measure · red costs money, teal saves or raises it

  • DWP Fraud and Error2031–32+£240m
  • Universal Credit2027–28−£140m
  • Carer’s Allowance2031–32−£80m
  • Parental Pay2029–30−£70m
  • Welfare Reform2027–28+£5m
All 5 measures on this topic
The Budget in 2 minutes →Is it being delivered? →

Treasury costings certified by the OBR, not measured results. Topic links are UK Facts’ classification.

The evidence

How the benefits bill has changed

The bill has grown after inflation. The economy has grown too. Pensioners received just over half the money at both ends of the 30-year comparison. Within those broad groups, different benefits have changed in different ways.

Spending as a share of the economy and the pensioner and working-age split are at the top of this page, beside the headline.

See spending by individual benefit

Which benefits does the money go on?

The State Pension is 47% of DWP benefit spending; Universal Credit 23%

Great Britain · 2024–25 (outturn)
  • State PensionRetirement£136.6bn
  • Universal CreditLow income, housing, children, unemployment, ill health£66.7bn
  • Personal Independence PaymentExtra costs of disability (working age)£25.9bn
  • Housing BenefitRent (mostly pensioners and remaining legacy cases)£15.5bn
  • Employment and Support AllowanceIllness or disability that limits work£12.3bn
  • Attendance AllowanceExtra costs of disability (pension age)£7.8bn
  • Disability Living AllowanceExtra costs of disability (children; older legacy cases)£7.7bn
  • Pension CreditLow income in retirement£6.0bn
  • Carer's AllowanceCaring£4.2bn
  • All other DWP benefitsJobseeker’s Allowance, Winter Fuel Payments, maternity, bereavement and others£6.6bn

The State Pension is by far the largest benefit. Support for disability is spread across several payments, so its total depends on what is included.

Great Britain · 2024–25 (outturn) · Accredited official statistics, outturn

Show the workings — The State Pension is 47% of DWP benefit spending; Universal Credit 23%

Spending on the largest individual DWP benefits, Great Britain, 2024–25, £ billion, cash terms. The bars are separate benefits and, with the remainder, add up to the £289.3bn Table 1a total.

One benefit, the State Pension, accounts for £136.6bn — more than Universal Credit, PIP and Housing Benefit combined. Universal Credit is the second largest and is still growing as people move across from older benefits. Disability-related payments are spread across several lines (PIP, Attendance Allowance, DLA, ESA and the health element of Universal Credit), which is why “disability spending” totals differ depending on what is counted.

Geography:
Great Britain — DWP benefits only (excludes HMRC Child Benefit and tax credits)
Measure:
£ billion, cash terms
Period:
2024–25 (outturn)
Status:
Accredited official statistics, outturn
Calculation:
As published (£ million ÷ 1,000). “All other DWP benefits” = the Table 1a total minus the nine named benefits (£6.6bn).
Compare with:
Total DWP benefit expenditure £289.3bn
Not included / breaks:
This total (£289.3bn) is smaller than the GB welfare total in BEN-01 (£304.6bn) because it leaves out welfare paid by HMRC — Child Benefit and the last tax credits. Universal Credit covers several needs in one payment (living costs, housing, children, ill health), so it cannot be split by need from this table.

Source: DWP, Benefit expenditure and caseload tables 2026 (Spring Forecast 2026), Table 1a, expenditure by benefit (released April 2026).

View the data as a table
BenefitMain need it meets2024–25 (£bn)Share of total
State PensionRetirement136.647.2%
Universal CreditLow income, housing, children, unemployment, ill health66.723.1%
Personal Independence PaymentExtra costs of disability (working age)25.98.9%
Housing BenefitRent (mostly pensioners and remaining legacy cases)15.55.3%
Employment and Support AllowanceIllness or disability that limits work12.34.3%
Attendance AllowanceExtra costs of disability (pension age)7.82.7%
Disability Living AllowanceExtra costs of disability (children; older legacy cases)7.72.7%
Pension CreditLow income in retirement6.02.1%
Carer's AllowanceCaring4.21.5%
All other DWP benefitsVarious6.62.3%
Total (Table 1a)289.3100%

Scroll the table sideways to see every column.

Explore the spending breakdown →

Millions of eligible families miss out on support

In 2024–25, an estimated 2.42 million families entitled to Universal Credit or equivalent benefits did not claim. The published range is 2.26–2.59 million. These new survey-based estimates are still in development. They include older benefits and tax credits, not Universal Credit alone.

There is no continuous 30-year total for everyone missing benefits. The two charts below show the longer Pension Credit history. Working-age estimates cover only 2023–24 and 2024–25, and changes in the eligible group prevent a direct comparison between those years.

Survey-based evidence
Family Resources Survey linked to administrative recordsFieldwork: Financial years ending 2013–2025; actual coverage variesEntitlement is modelled. Estimates have sampling ranges, method breaks and a missing year during the pandemic. Working-age estimates are Official Statistics in Development.

How many eligible pensioner families miss out?

Around 820,000 eligible families missed Pension Credit in 2024–25

Around 820,000 eligible families missed Pension Credit in 2024–25

Great Britain · 2012–13 to 2024–25; 2020–21 unavailable

2012–13 to 2024–25 · Million pensioner families · observed figures; gaps and definition breaks retained.

Million pensioner families0.00m0.50m1.00m1.50m2.00m20122016202020240.82m
A family is one adult or a couple, plus any dependent children. The line shows central estimates. The latest published range is 730,000–900,000 families. A smaller count can reflect fewer eligible families as well as better take-up; read it alongside the percentage chart.

Great Britain · 2012–13 to 2024–25; 2020–21 unavailable · Survey-based

Survey and entitlement-model estimates. Latest 95% range: 730,000–900,000. All published annual ranges are in the table.

Show the workings — Around 820,000 eligible families missed Pension Credit in 2024–25
Geography:
Great Britain
Measure:
Million families; central estimates
Period:
2012–13 to 2024–25; 2020–21 unavailable
Status:
Survey-based
Calculation:
As published, with units converted for display.
Uncertainty:
Survey and entitlement-model estimates. Latest 95% range: 730,000–900,000. All published annual ranges are in the table.
Not included / breaks:
No all-benefits total. Method refinements in 2016–17 and improved linking in 2023–24 are line breaks. 2020–21 is missing; do not interpolate. Eligibility and pension-age rules changed.

Source: DWP, Income-related benefits: estimates of take-up, financial year ending 2025, PC1, PC2 and UC1 (released 22 October 2026).

View the data as a table
Financial yearMissing: centralLowerUpperEligible families missing (%)
2012–1314400001340000154000038
2013–1413400001240000143000037
2014–1512700001170000137000038
2015–1612500001150000135000039
2016–1711500001050000125000039
2017–181060000970000115000039
2018–19920000840000100000037
2019–2077000069000085000034
2020–21UnavailableUnavailableUnavailableUnavailable
2021–2279000071000087000037
2022–2370000064000076000035
2023–2482000074000091000038
2024–2582000073000090000038

Scroll the table sideways to see every column.

Survey-based evidence
Family Resources Survey linked to administrative recordsFieldwork: Financial years ending 2013–2025; actual coverage variesEntitlement is modelled. Estimates have sampling ranges, method breaks and a missing year during the pandemic. Working-age estimates are Official Statistics in Development.

What share of eligible families miss out?

About 38% of eligible pensioner families did not claim

Around 820,000 eligible families missed Pension Credit in 2024–25

Great Britain · 2012–13 to 2024–25; 2020–21 unavailable

2012–13 to 2024–25 · % of eligible pensioner families · observed figures; gaps and definition breaks retained.

% of eligible pensioner families0.0%20.0%40.0%201220162020202438.0%
This percentage accounts for the changing number of eligible families. It is distinct from the share of benefit money left unclaimed. Method breaks and the missing pandemic year are retained.

Great Britain · 2012–13 to 2024–25; 2020–21 unavailable · Survey-based

Survey and entitlement-model estimates. Latest 95% range: 730,000–900,000. All published annual ranges are in the table.

Show the workings — About 38% of eligible pensioner families did not claim
Geography:
Great Britain
Measure:
% of entitled pensioner families
Period:
2012–13 to 2024–25; 2020–21 unavailable
Status:
Survey-based
Calculation:
100 − published caseload take-up percentage.
Uncertainty:
Survey and entitlement-model estimates. Latest 95% range: 730,000–900,000. All published annual ranges are in the table.
Not included / breaks:
No all-benefits total. Method refinements in 2016–17 and improved linking in 2023–24 are line breaks. 2020–21 is missing; do not interpolate. Eligibility and pension-age rules changed.

Source: DWP, Income-related benefits: estimates of take-up, financial year ending 2025, PC1, PC2 and UC1 (released 22 October 2026).

View the data as a table
Financial yearMissing: centralLowerUpperEligible families missing (%)
2012–1314400001340000154000038
2013–1413400001240000143000037
2014–1512700001170000137000038
2015–1612500001150000135000039
2016–1711500001050000125000039
2017–181060000970000115000039
2018–19920000840000100000037
2019–2077000069000085000034
2020–21UnavailableUnavailableUnavailableUnavailable
2021–2279000071000087000037
2022–2370000064000076000035
2023–2482000074000091000038
2024–2582000073000090000038

Scroll the table sideways to see every column.

Explore take-up and hardship →

An ageing population and the pension bill

There are more older residents. Pension spending also reflects payment rates, contribution histories and changes to State Pension age. These aligned charts show population and cost on their own scales; they do not establish how much of the increase ageing caused.

The population chart uses a fixed age of 65 for a consistent comparison. It is not a count of people above the changing State Pension age, or of pension recipients.

How has the older population changed?

More residents are aged 65 and over

More residents are aged 65 and over

Great Britain · Mid-1994 to mid-2024

Mid-1994 to Mid-2024 · Million GB residents aged 65+ · observed figures; gaps and definition breaks retained.

Million GB residents aged 65+0.00m5.00m10.00m15.00m19952005201512.85m
The number aged 65+ rose from 8.9m to 12.9m. Their population share rose from 15.9% to 19.1%. Age 65 is a fixed comparison, not the State Pension age, which changed during this period.

Great Britain · Mid-1994 to mid-2024 · Official statistics; observed figures only

Show the workings — More residents are aged 65 and over
Geography:
Great Britain
Measure:
Million residents aged 65+
Period:
Mid-1994 to mid-2024
Status:
Official statistics; observed figures only
Calculation:
Sum ONS single-year ages 65 and above, Persons, Great Britain, from one revised historical edition.
Not included / breaks:
Residents only. A fixed age threshold does not count State Pension recipients. Calendar mid-year population and financial-year spending are different periods.

Source: ONS, Population estimates: historical time series, mid-2025 edition, Table 6, Great Britain, persons by single year of age (released 1 October 2026).

View the data as a table
Mid-yearAged 65+All residentsAged 65+ (%)
199489408785621843815.90
199589808805637566815.93
199690064165650262315.94
199790256015664298815.93
199890506105679717415.93
199990573665700542115.89
200090865495720312115.88
200191487485742417815.93
200292030085766814315.96
200392601375793173815.98
200493174565823632216.00
200593800795868554315.98
200694161985908385415.94
200795010715955739215.95
200896461436004462016.06
200998184596046715316.24
2010100030926095462316.41
2011101924246147082716.58
2012105580846188594017.06
2013108308636230654417.38
2014110861656277550717.66
2015112664946323201517.82
2016114464506373979917.96
2017116013666408911418.10
2018117538016440046818.25
2019119416956472898818.45
2020120530596483934418.59
2021122145066507342418.77
2022124136106576943118.87
2023126127526666766618.92
2024128537066732841919.09

Scroll the table sideways to see every column.

How has the pension bill changed?

Pension costs reflect numbers and payments

Pension costs reflect numbers and payments

Great Britain · 1994–95 to 2024–25

1994–95 to 2024–25 · £bn, 2026–27 prices · observed figures; gaps and definition breaks retained.

£bn, 2026–27 prices£0£50£100£150£200199520052015£143.8bn
The State Pension caseload rose from 10.1m to 13.0m, averaged over each year. Spending also depends on payment rates, contribution histories and pension-age rules. About 1.1m recipients were paid abroad; they are not in the GB resident population chart. These charts do not isolate the cost caused by ageing.

Great Britain · 1994–95 to 2024–25 · Official statistics; observed figures only

Show the workings — Pension costs reflect numbers and payments
Geography:
Great Britain
Measure:
£ billion, 2026–27 prices
Period:
1994–95 to 2024–25
Status:
Official statistics; observed figures only
Calculation:
As published, with units converted for display.

Source: DWP, Benefit expenditure and caseload tables 2026, GB welfare; State Pension (released 14 April 2026).

View the data as a table
Financial yearState Pension (£bn, real)Recipients: annual averageOf whom paid abroad
1994–9561.73100970000
1995–9662.3710384000721000
1996–9764.4010536000752000
1997–9867.3710680000781000
1998–9970.4310782000805000
1999–0073.8610936000833000
2000–0175.0411004000864000
2001–0279.5911095000889000
2002–0382.5411195000923000
2003–0484.4511320000957000
2004–0586.0511477000991000
2005–0688.25115850001015000
2006–0789.41117150001046000
2007–0894.11119380001079000
2008–0997.02121600001109000
2009–10104.04124100001139000
2010–11106.79125660001165000
2011–12111.02126670001180000
2012–13117.46128100001195000
2013–14119.82128880001209000
2014–15123.00129580001219000
2015–16126.17129570001219000
2016–17126.76129300001216000
2017–18128.21128380001200000
2018–19129.28127240001187000
2019–20128.65125560001167000
2020–21126.25123790001153000
2021–22129.23124570001137000
2022–23127.48125970001120000
2023–24135.96127760001112000
2024–25143.84129870001102000

Scroll the table sideways to see every column.

Explore the State Pension →

Millions receiving support are already working

Universal Credit can top up income and help with rent, children and childcare. Employment does not mean a full-time job or enough earnings to meet a household’s costs.

Does receiving support mean being out of work?

3.1 million people on Universal Credit were working

3.1 million people on Universal Credit were working

Great Britain · April 2025 and April 2026

April 2025 to April 2026 · Million people in any employment · observed figures; gaps and definition breaks retained.

Million people in any employment0.00m1.00m2.00m3.00m4.00m20263.10m
The employed share was 41.9% in April 2025 and 37.7% in April 2026. Any employment includes employee earnings and reported self-employment. Movement from older benefits changes the UC caseload, so this is not a measure of changes in all working benefit recipients.

Great Britain · April 2025 and April 2026 · Official statistics; observed figures only

Show the workings — 3.1 million people on Universal Credit were working
Geography:
Great Britain
Measure:
Million people; any employment
Period:
April 2025 and April 2026
Status:
Official statistics; observed figures only
Calculation:
As published, with units converted for display.
Not included / breaks:
Only these two comparable annual snapshots are retained from this bulletin. Not a 30-year series. Not all open UC claims receive a payment.

Source: DWP, Universal Credit quarterly statistics to 14 May 2026, Main stories; any employment indicator (released 18 August 2026).

View the data as a table
MonthEmployed people (million)Share of UC claimants (%)
April 20253.241.9
April 20263.137.7

What does this support cost? Universal Credit awards belong to households, which can contain working and non-working adults. DWP says a spending breakdown for people in paid employment is not readily available. Total Universal Credit spending would answer a different question, so it is not overlaid here. Read DWP’s explanation.

Explore support in and out of work →

How disability support has changed

Disability payments help with extra costs and can be paid to people who work. They overlap with the other groups on this page. This chart combines Personal Independence Payment and the Disability Living Allowance it replaced for working-age people.

DWP-administered benefits only: excludes Universal Credit health payments and devolved Scottish replacements. Transfers to Scotland affect the later figures.

How has disability support changed?

Spending on extra disability costs has grown

Spending on extra disability costs has grown

Great Britain · 1992–93 to 2024–25

1994–95 to 2024–25 · £bn, 2026–27 prices · observed figures; gaps and definition breaks retained.

£bn, 2026–27 prices£0£10£20£30199520052015£23.8bn
Personal Independence Payment and Disability Living Allowance help with extra disability costs and can be paid in or out of work. Combining them avoids treating the transfer from DLA to PIP as entirely new support. This is a narrower measure than all health-related benefits.

Great Britain · 1992–93 to 2024–25 · Official statistics; observed figures only

Show the workings — Spending on extra disability costs has grown
Geography:
Great Britain
Measure:
£ billion, 2026–27 prices
Period:
1992–93 to 2024–25
Status:
Official statistics; observed figures only
Calculation:
As published, with units converted for display.
Not included / breaks:
DWP-administered PIP and DLA only. Excludes UC health payments and devolved Scottish replacement benefits, including Adult Disability Payment. Transfers to Scotland affect later comparisons. A rising caseload does not by itself establish its causes.

Source: DWP, Benefit expenditure and caseload tables 2026, Table 2b; working-age PIP and DLA, real terms (released 14 April 2026).

View the data as a table
Financial yearWorking-age PIP and DLA (£bn, real)
1992–932.77
1993–943.79
1994–954.16
1995–964.9
1996–975.72
1997–986.2
1998–996.45
1999–006.66
2000–016.95
2001–027.33
2002–037.64
2003–047.97
2004–058.16
2005–068.36
2006–078.54
2007–088.94
2008–099.14
2009–109.76
2010–119.87
2011–1210.33
2012–1310.92
2013–1411.06
2014–1512.09
2015–1613.43
2016–1713.66
2017–1815.44
2018–1916.03
2019–2016.18
2020–2115.7
2021–2217
2022–2318.13
2023–2420.88
2024–2523.81
Explore disability support →

What people think about benefits

Survey-based evidence
British Social Attitudes 2024Fieldwork: September to October 2024Push-to-web survey of adults in Great Britain, fieldwork autumn 2024, weighted. Estimates have sampling uncertainty; each item carries its own base.

Attitudes are mixed, not one-sided. More people say cutting benefits would damage too many lives than disagree, and more say most unemployed people could find work than disagree. Fewer than three in ten want more spending on benefits for the poor if taxes rise.

Agree includes strongly agree; disagree includes strongly disagree; the rest neither or don’t know. Different survey modes make comparisons with earlier years unreliable, so no trend is shown.

  • Cutting benefits would damage too many people’s lives: agree52%
  • Cutting benefits would damage too many people’s lives: disagree17%
  • Most unemployed people could find a job if they really wanted one: agree54%
  • Most unemployed people could find a job if they really wanted one: disagree19%
  • If benefits weren’t so generous, people would stand on their own feet: agree42%
  • If benefits weren’t so generous, people would stand on their own feet: disagree32%
  • Government should redistribute from better-off to less well-off: agree42%
  • Government should redistribute from better-off to less well-off: disagree30%
  • Government should spend more on benefits for the poor, even if taxes rise: agree28%
  • Government should spend more on benefits for the poor, even if taxes rise: disagree39%
  • Most people on the dole are fiddling in one way or another: agree29%
  • Most people on the dole are fiddling in one way or another: disagree32%

Fraud, error and viral claims

DWP estimates 3.2% of benefit spending was overpaid in 2025–26: fraud 2.2%, claimant error 0.6%, official error 0.4%. Separately, 0.4% was underpaid. These are sample estimates, not counts of convictions.

Fraud and error, by benefit, with claim checks →

Why it is happening

Our analysis

Benefits serve different needs: retirement, low pay, disability and caring. Changing the rules affects both household incomes and public spending; many eligible people also miss out on help.

Read the options, evidence and limits below.

Who is affected most

Who feels this most?

Does receiving Universal Credit mean someone is out of work?

1.6 million of 8.4 million people on Universal Credit must look for work

Great Britain · May 2026

People on Universal Credit, May 2026 · outturn · total 8.4m

  • No work requirements: 51% (4.3m)
  • Searching for work: 19% (1.6m)
  • Four other groups: 30% (2.5m)

Receiving Universal Credit does not mean someone is unemployed. Many recipients work; others are too ill to work, care for someone or have very young children.

Great Britain · May 2026 · Official statistics (administrative data)

Show the workings — 1.6 million of 8.4 million people on Universal Credit must look for work

People on Universal Credit by DWP conditionality regime (the work-related rules attached to their claim), Great Britain, May 2026. Millions of people.

Half of everyone on Universal Credit (51%) has no work requirements — for example people assessed as too ill to work, carers and parents of very young children. That group has grown as 1.7 million people moved across from older benefits, many from Employment and Support Allowance. The group required to look for work is 1.6 million. Separately, 3.1 million people on Universal Credit (37.7%) were in paid work in April 2026: receiving support does not tell you whether someone works.

Geography:
Great Britain — People on Universal Credit (claimants and partners)
Measure:
Millions of people
Period:
May 2026
Status:
Official statistics (administrative data)
Calculation:
“No work requirements” (4.3m) and “searching for work” (1.6m) as published; “four other groups” (working with requirements, working with no requirements, planning for work, preparing for work) = 8.4m − 4.3m − 1.6m, a UK Facts calculation from rounded totals (±0.1m).
Compare with:
“Searching for work” peaked at 2.4 million in March 2021
Not included / breaks:
A benefit rule category is not the ILO unemployment measure, and Universal Credit is not all benefit recipients (State Pension, PIP and others are separate). 35% of people in “searching for work” had some earnings in April 2026, because low earners can be placed there. PIP is not a work-status category at all.

Source: DWP, Universal Credit quarterly statistics, 29 April 2013 to 14 May 2026, Figures 5, 8 and 9 (people by conditionality regime; employment indicator) (released 18 August 2026).

View the data as a table
GroupPeople (millions)Share
No work requirements4.351%
Searching for work1.619%
Four other groups (calculated)2.530%
All people on Universal Credit8.4100%
In any paid employment (April 2026; overlaps the groups above)3.137.7%

The five numbers most often swapped for one another — and what each counts: support in work and out of work.

What could change

Proposals to consider. The bodies named below would need to agree to deliver them; responsibility is not a recorded commitment. Follow the links for costs and evidence.

  • Claim what people are owed

    Help people likely to qualify for Pension Credit apply for it.

    Who could deliver it: the Department for Work and Pensions, councils and advice providers.

    Main trade-off and delivery needs

    Privacy, mistaken advice and processing delays; fiscal cost is not equivalent to social loss.

    Specify eligible cohort, trusted outreach, application support and comparison group; arrange benefits-processing capacity.

    Cost: Only part of the cost has been estimated.

    Costs, drawbacks and evidence →

  • A child-poverty target and essentials test

    Set a child-poverty target and test whether families can afford essentials.

    Who could deliver it: Treasury, the Department for Work and Pensions and relevant service departments/devolved governments.

    Main trade-off and delivery needs

    Unfunded targets; narrow metrics; inadequate treatment of disability and housing costs.

    Define poverty/essentials measures, then compare specific benefit, housing-cost and service changes; assign accountability.

    Cost: The extra cost is not yet known.

    Costs, drawbacks and evidence →

  • Offer evidence-based employment support alongside healthcare

    Link employment support with healthcare and test whether it helps people into work.

    Who could deliver it: the Department for Work and Pensions, health commissioners and local employment partners.

    Main trade-off and delivery needs

    Clinical displacement; unsuitable pressure to work; variable fidelity.

    Specify cohort and voluntary offer; secure clinical referrals, trained employment specialists, employer links and linked-data evaluation.

    Cost: Only part of the cost has been estimated.

    Costs, drawbacks and evidence →

  • Prevent avoidable loss of rented homes

    Test whether early help with rent arrears and legal advice prevents people losing their homes.

    Who could deliver it: councils, housing/advice providers and landlords.

    Main trade-off and delivery needs

    The main risks still need to be assessed.

    Define arrears/risk trigger, legal advice and assistance design; cost staff, grants and sustained tenancies. Assess additional prevention against existing duties. Proposed owners: councils, housing/advice providers and landlords.

    Cost: The extra cost is not yet known.

    Costs, drawbacks and evidence →

See all options and what is still unknown →

Some problems have no assessed solution yet: see coverage and gaps.

Help and rights

These links can help with an individual situation. They do not resolve the wider pressures described above.

  • Check possible benefit entitlement

    GOV.UK links to independent calculators and lists circumstances in which their estimates are unsuitable. An estimate is not an award decision.

Domain SWOT

Recorded

Show the findings
Strengths · 1 finding

  • An established payment route: Universal Credit recorded 8.4 million people in May 2026; 3.1 million were in work in April 2026. Being on the caseload does not necessarily mean receiving a payment.

    Great Britain · May 2026 (people); April 2026 (in work)

Weaknesses · 4 findings

  • PAC reported that incorrect earnings-averaging guidance had led to an estimated 26,000 carers being wrongly recorded as overpaid. DWP began reassessing affected cases in April 2026; that start does not establish completed redress.

    Great Britain · PAC February 2026; DWP April 2026

  • NAO found progress in reducing overpayments but called for an implementation plan that prevents errors before debt builds up and monitors adverse effects on claimants. Savings from checks alone do not demonstrate an accurate or fair service.

    Great Britain · NAO, 22 October 2025

  • Only 62% of people entitled to Pension Credit claimed it; up to 900,000 entitled families did not, leaving up to £2.4bn unclaimed in 2024–25 (one year’s estimate).

    Great Britain · 2024–25 (FYE 2025)

  • Payments are wrong in both directions: 3.2% of benefit spending (£9.9bn) was overpaid and 0.4% (£1.2bn) underpaid in 2025–26.

    Great Britain · 2025–26

Opportunities · 1 finding

  • Raising take-up of existing entitlements, testing whether the basic rate covers essentials and offering suitable employment support address different failures in the same system.

    Depends on: The higher legitimate spending that successful take-up brings is budgeted, any rate change is costed and funded explicitly, and employment support stays voluntary and is judged on sustained work.

    our synthesis

Threats · 1 finding

  • Hardship and error can rise when savings drive design: for most new claims from April 2026 the Universal Credit health element is about half the previous rate, and DWP’s impact assessment calls the behavioural effects the most uncertain part of its estimates.

    risk

See the whole-UK SWOT →

Domain action plan

Proposed

Possible steps for government and the bodies responsible. These describe the order of work, not promised dates: starting an action and seeing results take different amounts of time.

  1. Short term · prepare

    Check support gaps and the barriers facing claimants.

  2. Medium term · deliver

    Assess the linked support and employment proposals, including possible harm.

  3. Long term · judge results

    Track living standards and sustainable work, alongside public costs.

Delivery steps and owners (4)
  1. Prerequisite: Identify likely unmet Pension Credit entitlement without treating a survey estimate as individual eligibility · DWP with councils · proposed
  2. Decision: Specify the essentials basket and options to cost; budget for extra entitlement from take-up · DWP with HM Treasury · proposed
  3. Delivery: Test outreach in defined areas; keep employment support voluntary and at fidelity · DWP, councils, advice services and Connect to Work providers · proposed
  4. Result: Higher take-up, smaller payment errors both ways, more people in sustained work · DWP statistics · proposed

Sources, method and changes

Shareable visual summary

Visual summary

Most welfare spending goes to pensioners.

Pensioners received 54% of Great Britain’s £304.6bn welfare spending in 2024–25. Separately, 3.1 million Universal Credit claimants were in paid work in April 2026. A share of spending, not of claimants.

Pensioners’ share of Great Britain welfare spending, 2024–25

54%

Share of money, not of people: State Pension, Pension Credit, pension-age Housing Benefit, Attendance Allowance and other pension-age benefits.

Separately: Universal Credit claimants in work, April 2026

37.7%

Text and data version

Most welfare spending goes to pensioners. Pensioners received 54% of Great Britain’s £304.6bn welfare spending in 2024–25. Separately, 3.1 million Universal Credit claimants were in paid work in April 2026. A share of spending, not of claimants.

Records behind this visual summary
MeasureValuePeriodGeographyTypeSource
Pensioners’ share of Great Britain welfare spendingCalculation: Pensioner welfare spending ÷ total GB welfare spending × 100 (cash, outturn)54%2024–25Great Britainderived calculationDWP, Benefit expenditure and caseload tables 2026, GB welfare summary table
People on Universal Credit in employment3.1mApril 2026Great Britainadministrative count (provisional)DWP, Universal Credit statistics
Share of Universal Credit claimants in employment37.7%April 2026Great Britainadministrative count (provisional)DWP, Universal Credit statistics

Available images and their source records are listed on Shareable visuals.

Sources and methods (9)

Method notes

  • Benefit spending covers Great Britain: Northern Ireland runs its own social security system. Spending by Social Security Scotland on devolved benefits (such as Adult Disability Payment) is not in the DWP figures.
  • The financial year that has just ended is still a DWP forecast in the Spring Forecast tables; the page leads with the latest outturn year and labels the forecast separately.
  • Shares of spending are shares of money, not of people: one person can receive several benefits, so recipient counts are never added together here.
  • Universal Credit counts are administrative and published quarterly; the employment indicator is a month behind the caseload.