← Cost of living

Cost of living · Food insecurity · UK · 2015 to 2026

Food banks gave out 2.6 million parcels in 2025. Why do people need them, and what would reduce it?

Food banks in the Trussell network gave out 2,644,835 emergency parcels in 2025, 45% more than in 2019 but 12% fewer than in 2024. About 9% of UK households were food insecure in 2024/25. Food prices have risen 38% in five years and average pay 30%. The strongest lever is income for the households that have least; no study compares it with cheaper food.

Evidence last reviewed 21 October 2026

The headline numbers

  • 2,644,835

    emergency food parcels in 2025

    Trussell network, UK. A count of parcels, not people.

  • +45%

    on 2019

    Down 12% on 2024 and 18% below the 2023 peak.

  • 9%

    of households food insecure

    2024/25, UK. Government survey; flat since 2022/23.

  • 38% v 30%

    food prices against average pay

    Five years to September (prices) and August (pay) 2026.

How many parcels are food banks giving out?

Parcels rose from 1,107,806 in 2015 to a peak of 3,237,114 in 2023, with a dip in 2021, then fell. 2025 was still 45% above 2019. Most of the rise since 2019 is in England and Northern Ireland; Scotland is below its 2019 level.

Million parcels0.0m1.0m2.0m3.0m4.0m201620202024All 2.64mChildren 0.91m
Parcels by nation, 2025, and change since 2019
NationParcels, 2025Change since 2019
England2,204,106+56%
Scotland218,838−8%
Wales154,040+19%
Northern Ireland67,851+60%

Trussell counts parcels (volume), not people: a family referred twice is two parcels per person. It covers food banks in its own network only, not the independent food banks. Its series moved from financial years to calendar years in March 2026, so the 2024/25 figure (2,885,086) and calendar 2024 (2,996,653) are different periods and are never joined. Local figures are by the location of the food bank and should not be compared between councils. Source: End of Year Stats 2025: parcel statistics workbook, 11 March 2026.

How many households cannot reliably afford food?

Survey-based evidence
Family Resources Survey 2024/25Fieldwork: April 2024 to March 2025DWP household survey, UK, face to face; households classed food insecure on a short scale of questions about the last 12 months. Estimates have sampling uncertainty.

Around one household in eleven was food insecure in 2024/25: 9%. It was 8% in 2019/20, rose to 10% in 2022/23 and has not fallen since. About 3.5% of households said they had used a food bank in the last 12 months.

Other surveys give higher figures because they ask different questions of different people. They are not in conflict and should not be averaged.

Three measures of food insecurity and how they differ
MeasureResultWho, when and how
DWP Family Resources Survey9% of householdsAll UK households, April 2024 to March 2025, face to face, last 12 months.
Trussell and Ipsos, Hunger in the UK16% of householdsUK, 2024, online panel, broader definition (running out or cutting meal size).
Food Foundation tracker12.2% of householdsUK adults, 25 to 30 June 2026, online panel, last month, three questions.

Who is most affected

Share of households that were food insecure, 2024/25. Receipt of a benefit marks low income or disability; it does not show the benefit is the cause.

  • Lone adult with three or more children39% of households food insecuresmall sample (134 households)
  • Households on Universal Credit32% of households food insecurea marker of low income, not a cause
  • Social renters28% of households food insecure
  • Lone adult with children28% of households food insecure
  • Households on Employment and Support Allowance27% of households food insecure
  • Households on Carer’s Allowance26% of households food insecure
  • Households with a disabled working-age adult20% of households food insecure
  • Private renters12% of households food insecure
  • All households9% of households food insecure
  • Households owning their home3% of households food insecure
  • Pensioner couples (both over State Pension age)1% of households food insecure

Source: Family Resources Survey FYE 2025, chapter 9: household food security tables, 26 March 2026. Percentage of households, UK. Food insecure = low or very low food security. FYE 2021 had a smaller, less representative sample (COVID collection). Single-year changes are rarely statistically significant.

Why do people use food banks?

The evidence points to incomes that do not cover essentials. Since August 2021, average weekly pay has risen 30% (£584 to £757) while food prices have risen 38%. In food terms, average pay is about 6% lower than five years ago. Average pay hides the lowest paid, and many people referred to food banks are not in work.

Change over five years to 2026: food prices (CPI, September), all prices (September) and average weekly pay (August). Different months because of publication dates.

  • Food and non-alcoholic drink prices+38.5% rise in five years
  • Average weekly pay+29.6% rise in five years
  • All prices (CPI)+28.2% rise in five years

Going cold and skipping meals

Survey-based evidence
DESNZ Public Attitudes TrackerFieldwork: Spring 2026UK adults 16+ (3,380 in spring 2026). Asked which actions they had taken in recent months because of the cost of their energy bills. Estimates have sampling uncertainty (about 1 point).

In spring 2026, about one adult in five (22%) said they had cut electricity or heating to an uncomfortable level because of the cost of their energy bills. About one in ten (10%) had skipped meals. One in twenty (5%) had done both. The survey records what people did, not a choice between the two, so we do not describe it as people “choosing” between heating and eating.

Share of adults taking each action because of energy-bill costs, Spring 2023 and Spring 2026. The question wording changed slightly in 2026, so read the change as small.

  • Cut heating, Spring 202622% of adults
  • Cut heating, Spring 202327% of adults
  • Skipped meals, Spring 202610% of adults
  • Skipped meals, Spring 202311% of adults
  • Both, Spring 20265% of adults
  • Both, Spring 20237% of adults
  • Used a food bank, Spring 20263% of adults
  • Used a food bank, Spring 20234% of adults

DESNZ Public Attitudes Tracker, 2021-2026, UK Data Service study 9568. Actions the respondent links to the cost of energy bills; not a general food-insecurity measure. The wave 16 question stem omits the words "the increase in". See Energy for prices and the energy mix.

What could reduce it: cost, who benefits, trade-offs

Is the better lever cheaper food or higher incomes? No study we found compares the cost per household moved out of food insecurity for income support, price caps and wage rises, so we do not rank them. What the evidence does show: the households most at risk are on benefits, disabled, renting socially or lone parents, and many have no earner, so a wage rise does not reach them. A price cap helps every shopper at the covered retailers, not those in need.

Options to reduce food insecurity, with the do-nothing baseline first
OptionWhat it isCost, as publishedWho benefitsEvidenceTrade-off
Do nothing more (baseline)DfE and DWP, Our Children, Our FutureCurrent policy continues: the two-child limit removed from April 2026, the Universal Credit standard allowance rising above inflation to 2029/30, free school meals for children in Universal Credit households from September 2026 and a £1bn Crisis and Resilience Fund for councils in England.Already committed in current plans.The government’s own target is 550,000 fewer children in relative low income by the end of the Parliament. 4.5 million children (31%) were in relative low income after housing costs in 2023/24.Food bank parcels are 45% above 2019 and falling; the Family Resources Survey shows 9% of households food insecure and flat.Nothing new to fund, but the measured problem is not falling quickly.
Remove the two-child limit in Universal CreditHM Treasury, Budget 2025; DWP, 4 December 2025Already legislated for England, Scotland and Wales from April 2026. Northern Ireland is a devolved choice.£2.4bn in 2026-27 rising to £3.1bn in 2029-30 (HM Treasury).DWP estimates 450,000 fewer children in relative low income by FYE 2030 and 2 million children in households that gain. Families with three or more children have higher food insecurity (14% of couples, 39% of lone adults).Official modelling, static (no behavioural response). DWP did not model the effect on food insecurity.A cost to the Exchequer; weakens the incentive effect the limit was designed to have.
Raise the Universal Credit standard allowance furtherJRF, 29 June 2026It rose 6.2% in April 2026 to £424.90 a month for a single person aged 25 or over. JRF proposes a further £7 (single) or £14 (couple) a week and a floor that stops deductions pulling support below a set level.JRF costs the uplift at £2.9bn in 2029/30 and the floor at £1.7bn (advocacy costings, 2026/27 prices).Households on Universal Credit: 32% are food insecure. JRF counts 3.8 million families gaining from the April 2026 rise.JRF’s costs are an advocacy organisation’s; there is no independent costing.Reaches households on benefits, not low-paid households outside Universal Credit.
An Essentials Guarantee in Universal CreditJRF and Trussell, Guarantee our Essentials, 30 January 2026A legal minimum so the allowance covers an independently set level of essentials: £120 a week for a single adult and £205 for a couple in 2026/27. The current single allowance is about £98 a week, so the gap is about £22.Not costed in the JRF pages we read. Earlier JRF costings (about £19bn to £22bn a year) are secondary and for earlier years.JRF counts 8.8 million low-income families, including 3.9 million with children. Trussell estimates about £17.6bn a year in economic and fiscal benefits; that is not a net figure and has no cost beside it.Strong on the size of the gap; moderate on the effect on food insecurity (Trussell modelling, an advocate).The largest of the income options; a cost to the Exchequer.
Free school meals for all children in Universal Credit households (England)DfE, 4 June 2025; HM Treasury, Budget 2025From September 2026.Sources disagree and are not reconciled: £410m a year by 2028-29 (Budget 2025), "over £1bn" (DWP, August 2026), about £1bn long-run (IFS, secondary).Over 500,000 additional children; government estimates 100,000 fewer children in poverty and about £500 a year per family.Official estimates for England. The effect on measured food insecurity was not found.In-kind, with learning and health benefits beyond food. Scotland and Wales have their own policies.
Local crisis support and cash-first helpDfE and DWP, Our Children, Our FutureA £1bn Crisis and Resilience Fund for councils in England, plus £600m for holiday activities and food and £80m for breakfast clubs in 2026-27.£1bn fund; start date and whether it replaces the Household Support Fund are not stated on the pages we read.Households in crisis, as local councils decide.No evidence found on the effect on food bank use.Flexible and local, but short-term and uneven between councils.
Extend Healthy Start to all Universal Credit familiesFood FoundationHealthy Start gives about £4.65 a week (£9.30 for under-1s) for food and milk to some pregnant women and young children (rates as reported by advocacy and council sources; the NHS page could not be read).About £150m a year, per the Food Foundation (an advocate).Young children and pregnant women in low-income families. About 352,000 receive it now.A University of Birmingham analysis commissioned by advocates claims £2.36 back per £1; its inputs are not shown. Not an official appraisal.Small and targeted; helps with food specifically.
Higher wages (the National Living Wage)HM Treasury, Budget 2025; Trussell and Ipsos, Hunger in the UK 2025The National Living Wage is £12.71 an hour from April 2026 (+4.1%).Falls on employers, not the Exchequer; employment effects are for the Low Pay Commission and the OBR. No costing found.Workers. 30% of people referred to Trussell food banks live in working households, but most referred households have no earner, and a pay rise does not reach them.Effect on food insecurity not found. Average pay rose 29.6% in five years while food prices rose 38.5%.Does not reach households with no earner, including many disabled people and carers.
Price caps on essential foodsScottish Government consultation, 1 September 2026; NIESR, 12 June 2023The Scottish Government is consulting on legal price caps on essential items at the largest retailers (to 24 November 2026). Not proposed for the UK.Not quantified in the consultation sections we read.All shoppers at covered retailers, not targeted; the poorest fifth spend 15% of income on food, the richest 8%.Only Croatia has a quantified effect, from one report. Hungary’s price cap was ruled incompatible with EU law in 2024. The CMA says recent food inflation reflected supply-chain costs. NIESR favours targeted income support.Retailers or other customers and suppliers pay; the cost is not quantified.

Costs are each source’s own, for its own year and basis, and are not added. They are not net present values: none has been appraised to Green Book standard, and “not costed” means we found no published figure. Trussell, JRF and the Food Foundation are advocacy organisations; their costings and benefit estimates are shown with their names. The cost of the problem itself is estimated by Trussell at £75.6bn a year for severe hardship in 2022/23, which is wider than food insecurity and an advocate’s figure.

What this does not show

Sources and methods (6)

Method notes

  • Trussell parcels are calendar years from the March 2026 release; they are never joined to earlier financial-year figures. Each release recalculates earlier years for late data.
  • Family Resources Survey food insecurity means low or very low food security. The 2020/21 sample was smaller (COVID collection), and single-year changes are rarely statistically significant.
  • Tracker figures are weighted with finalweight; the base and wave are shown. They record actions the respondent links to energy-bill costs and are not a general measure of food insecurity. Microdata is licensed by the UK Data Service and is not published here.
  • Prices and pay: five-year changes use September (CPI) and August (earnings) because of publication dates. Average earnings are for all employees; low-paid workers are not separated.
  • Where the 2024-25 and 2025-26 policy details come from advocacy bodies or secondary reports, the table says so. Not costed means no published figure was found; no cost has been estimated here.

Corrections and changes

  • 21 October 2026: New dev-preview page.