What the evidence shows
CSJ’s 1.85 million estimate for 2025 includes pensions and adjusts for overlapping benefits. In a separate April 2026 Universal Credit snapshot, 1.29 million of 8.32 million people were in recorded status groups outside the combined British, Irish and right-of-abode category. About half of that subgroup were recorded employed or self-employed. The full CSJ model remains unreproduced, and these records do not provide an exact count of current foreign citizens.
- “Almost two million” is a reasonable rounding of CSJ’s estimated 1.85 million adult benefit or pension recipients in 2025.
- In separate April 2026 official Universal Credit records, the relevant status groups account for about 15 in every 100 people on that benefit, not 15% of all UK benefit and pension recipients.
- About half of that defined Universal Credit subgroup were recorded employed or self-employed. Receiving Universal Credit does not mean someone is out of work.
- CSJ estimates roughly 400,000 State Pension recipients within its total, about one in five of its estimated recipients. That is a ratio within CSJ’s model, not independent verification of it.
- A headcount alone does not establish unlawful residence, fraud, no tax contribution or an individual legal basis for removal.
What Sammy Woodhouse posted
Woodhouse shared an image of a Daily Mail headline saying almost two million foreign nationals now receive benefits or the State Pension in Britain. Her caption said 1.3 million receive Universal Credit, including 745,000 Europeans who stayed after Brexit, and ended with “Deport them”.
That last phrase is a political demand. The caption does not explicitly say that every recipient is unemployed or living here illegally. We are not attributing those extra statements to Woodhouse. We are checking what a reader can and cannot learn from the figures used to support the demand.
The original Instagram post is our claim source. The CSJ report and government records are the evidence sources. We could inspect the complete Instagram caption, but did not recover the full Daily Mail article through normal access.
Sources: Sammy Woodhouse, @officialsammyukCentre for Social Justice
The near-two-million estimate includes pensions
The Centre for Social Justice estimates that 1.85 million non-British adults received benefits or the State Pension in 2025. Calling that “almost two million” is defensible rounding. We have not demonstrated a false headline total.
CSJ describes adjustments for people who receive more than one benefit and for the adults covered by household claims. It is therefore wrong to assume it simply added all benefit totals together without allowing for overlap.
The estimate combines administrative records, survey estimates and population assumptions. It is not a government register that directly identifies 1.85 million current foreign citizens receiving a payment on a single verified date. We have not recovered enough of the underlying calculation to reproduce the complete model.
The basket excludes some forms of support, including Child Benefit, free school meals and funded childcare. The exact basket and how adults are counted matter when looking for a fair comparison.
How big is the Universal Credit part?
A large count becomes easier to understand when we can compare it with a matching total. The Department for Work and Pensions (DWP) publishes Universal Credit records by immigration status and employment status. We used April 2026 for both comparisons below, so they describe the same population and month.
| What is counted | People | Meaning |
|---|---|---|
| All people in the table | 8,316,509 | The matching Universal Credit total |
| Outside the combined British, Irish and right-of-abode group | 1,287,180 | 15.5% of that total |
| Recorded employed or self-employed within that subgroup | 634,761 | 49.3% of the 1,287,180 subgroup |
In Great Britain, Table 2 contains 8,316,509 people on Universal Credit. Of these, 1,287,180 are recorded outside DWP’s combined British, Irish and “right of abode” group. Dividing those two counts gives 15.5%, or about 15 in every 100 people on Universal Credit.
Right of abode means an unrestricted right to live and work in the UK. All British citizens have it, and some Commonwealth citizens can also qualify. This is the name of a specific legal category, not a label we have applied to everyone allowed to settle here.
The 15.5% describes the specified administrative groups within Universal Credit. It is not the percentage of all benefit recipients who are foreign citizens. It is also not a check of CSJ’s broader 2025 total, which covers pensions and other benefits.
Sources: Department for Work and PensionsDepartment for Work and PensionsDepartment for Work and PensionsGOV.UK
About half of that subgroup were recorded in employment
Of the 1,287,180 people in those status groups, 634,761 were recorded employed or self-employed. That is 49.3%, reasonably described as about half.
This does not mean half of all foreign nationals, half of all benefit recipients or half of all Universal Credit claimants. The denominator is the same defined 1.29 million-person subgroup in the April table.
Universal Credit supports people on low incomes, including some people in work. Its recipient count cannot be read as a count of people who are unemployed or refuse to work.
The employment fields have limits too. Employment reflects the assessment-period records. The self-employment classification includes people required to report self-employed earnings; it does not guarantee positive earnings, a particular number of hours or financial independence. People not recorded employed or self-employed cannot automatically be described as unwilling to work.
Sources: Department for Work and PensionsDepartment for Work and PensionsGOV.UK
About one in five of CSJ’s estimated recipients receive the State Pension
CSJ estimates roughly 400,000 State Pension recipients within its 1.85 million adult total for 2025. That is approximately one in five of the people in its estimate.
Both figures are rounded outputs from the same model. This ratio helps explain its composition; it is not an independently verified rate from a complete nationality register. Pension recipients can also receive other benefits, so these categories must not be added as if they never overlap.
State Pension eligibility depends on National Insurance records. Qualifying years can come from contributions, credits or voluntary payments, with additional rules for some circumstances. It would be wrong to say every estimated recipient personally paid the same number of years of UK tax.
In a parliamentary answer dated 3 February 2026, DWP said nationality is not recorded in State Pension claims. The CSJ pension number is therefore an estimate, not a direct official count of claimants by current citizenship.
Sources: Centre for Social JusticeGOV.UKUK Parliament, answer from Department for Work and Pensions
Why these records are not an exact citizenship count
Immigration status and nationality answer different questions. A status may explain a person’s permission to remain or their potential eligibility for support. Nationality describes citizenship. Neither should silently stand in for the other.
The 1.29 million comparison includes “Other” and missing-status records. The official people-on-Universal-Credit series can include some people without a payment and some ineligible partners. It should not be described as an exact count of current payments to foreign citizens.
DWP also warns that some indefinite-leave records can remain unchanged after someone becomes British if the department is not notified. We do not know the size of that issue. It does not justify assuming that all such classifications are outdated or calculating an unsupported reduction in CSJ’s estimate.
Woodhouse’s 745,000 line is close to the official May 2026 EU Settlement Scheme count of 744,679. But that status group can include eligible family members and some later joiners. “Europeans who stayed after Brexit” is not a precise description of every person in it.
May is used only to trace that caption line. Our scale and employment comparisons use April throughout. Similar rounded counts from different dates are not proof that they describe exactly the same people.
Sources: Department for Work and PensionsDepartment for Work and PensionsHome Office, GOV.UKSammy Woodhouse, @officialsammyuk
What the total does and does not tell us
The estimate covers different kinds of support and different circumstances. Some recipients are pensioners. A separate official Universal Credit subgroup includes many people recorded in employment. Being included in either count is not proof that a person is jobless or lives here unlawfully.
The figures do not establish benefit fraud, willingness to work, past tax contributions or an individual legal basis for removal. They also do not calculate immigration’s net cost: benefit spending alone leaves out taxes, other services, time periods and wider economic effects.
A large number can matter, and a smaller share can still matter. Neither decides the policy question by itself. The useful conclusion here is narrower: the near-two-million estimate is not a single category of unemployed people, and the total alone cannot tell us each person’s circumstances.
Sources: Centre for Social JusticeDepartment for Work and PensionsGOV.UKGOV.UK
What we still need to know
Checked: the original post, what CSJ says its estimate includes, and the separate official Universal Credit scale and recorded-employment comparisons.
Still unclear: the full CSJ model and a compatible all-recipient total for its 2025 basket. We cannot honestly turn 1.85 million into a verified percentage of everyone receiving all those benefits and pensions.
Next update: the complete underlying calculations, corrected official records or a material source revision could allow a fuller check. A model enquiry has already been prepared for the related investigation, but remains unsent. We have not contacted CSJ for this new article and no reminder has been created for it.
The preserved Universal Credit workbook lists 17 November 2026 as its next publication. A new release could improve the Universal Credit context, but would not by itself validate CSJ’s complete model. This is a source-review trigger, not a promise that an automatic watch is running.
Sources: Centre for Social JusticeDepartment for Work and Pensions
Technical notes for checking the arithmetic
The Universal Credit calculation uses the official ODS released on 18 August 2026. In Table_2, the April month header is F8:I8 and the employment headings are F10:I10. Row 12 is the combined CTA / UK, Ireland, Right of Abode category. Rows 13 to 31 cover the remaining status and nationality rows, including other or unknown records.
SUM(F12:I31) gives 8,316,509. SUM(F13:I31) gives 1,287,180. SUM(G13:I31) gives 634,761 for PAYE employment only, self-employment only and both combined. Do not add column F to that employment numerator, because F means neither recorded employment nor self-employment.
The calculations are 1,287,180 / 8,316,509 × 100 = 15.4774%, and 634,761 / 1,287,180 × 100 = 49.3141%. The carousel rounds these to about 15 in 100 and about half. Counts round to 1.29 million, 8.32 million and about 635,000.
Use the same table on both sides. Statistical disclosure control means displayed component sums can differ slightly between tables. The April Table_1 sum is 21 higher than our Table_2 total. Mixing the two would needlessly change the denominator.
The within-model pension ratio is 400,000 / 1,850,000 × 100 = 21.6216%. Because both inputs are rounded estimates, “about one in five” is more honest than presenting a highly precise observed rate.
We did not divide the 1.85 million model total by DWP’s different benefit-combination total or by the UK population. Those are not verified matching denominators for this basket, geography, period and adult-counting method. Our downloadable selected figures preserve these distinctions.
Sources: Department for Work and PensionsDepartment for Work and PensionsCentre for Social Justice
What is established
- The original post and its full author caption were inspected and preserved, separately from reader comments.
- CSJ reports an overlap-adjusted estimate of 1.85 million adults for 2025 and explicitly includes the State Pension.
- The separate Universal Credit scale and recorded-employment comparisons can be reproduced from the same official table and month.
What remains unknown
- The complete CSJ underlying calculation, including its precise common observation period and reconciliation of all population, nationality and geography assumptions.
- A compatible all-recipient total for the full 2025 basket, so we cannot calculate a verified percentage for the whole 1.85 million estimate.
- Each recipient’s circumstances and the size of any effect from outdated nationality records. We have not established numerical inflation or deceptive intent.
The story, in order
Year covered by the CSJ estimate
The estimated 1.85 million adults combine benefit and pension categories, with overlap adjustments.
Official Universal Credit comparison
A separate Great Britain snapshot supplies matching total, status-group and employment counts.
Official workbook released
The DWP ODS contains April employment tables and May headline caseloads. Those months are kept separate.
CSJ State of Dependency report
The report is followed by CSJ’s near-two-million announcement on 14 October.
Woodhouse post and evidence check
The original post is inspected and the new article is prepared. This is not a TALLYCAT01 publication date.
YOUR TURN TO CHECK
Reproduce this investigation
Check the original claim, keep the 2025 estimate separate from the April 2026 official comparison, and then repeat the arithmetic using the same rows on both sides.
You will need
A browser, PDF reader and spreadsheet app that opens ODS and CSV files, such as LibreOffice Calc or Microsoft Excel
What you can and cannot repeat
The complete CSJ model is unavailable, so this route reproduces the published model ratios and the separate official table calculations, not CSJ’s full unique-recipient estimate. Instagram may require sign-in. The selected-figures CSV contains public aggregate figures only, not personal records.
Read the actual post before interpreting it
Open Woodhouse’s original post. Read the headline and full author caption, including the 1.3 million and 745,000 lines and the closing “Deport them”. Separate her words from reader comments. Record 15 October 2026 as the post date.
Keep in mind The full Daily Mail article was not recovered. The photo screenshot is claim evidence, not a licensed new illustration for redistribution.
Identify what the 2025 estimate includes
Read the CSJ announcement’s opening, Notes to editors and Table 1. Find 1.85 million estimated adults and 400,000 estimated State Pension recipients for 2025. Check the stated allowance for overlapping benefits and household adults. In the report, inspect PDF pages 191 to 192, printed pages 189 to 190.
Keep in mind The published explanation does not provide the complete underlying model. Do not sum overlapping caseloads into a new unique-person total.
Open the correct official workbook and month
Download the official ODS. Check Contents for the 18 August 2026 release. Open Table_2. Confirm F8:I8 says April 2026 and read employment headings F10:I10. Row 12 is the combined CTA / UK, Ireland, Right of Abode group. Rows 13 to 31 contain the other status and nationality rows, including other or unknown records.
Keep in mind Our preserved original has SHA-256 22c9f2f3aa448af7f5ea4458cb2286e494a79f69ac31a7fa0b7381e86d2f29e4. A future download can differ after revision. Keep the original notes and do not substitute Table_1 totals.
Calculate the matching Universal Credit total
In a new spreadsheet cell, sum Table_2 F12:I31. Expect 8,316,509 people. Then sum F13:I31. Expect 1,287,180 people outside the combined row-12 group. Divide the second result by the first and multiply by 100.
SUM(F13:I31) / SUM(F12:I31) × 100 = 1,287,180 / 8,316,509 × 100 = 15.4774%, about 15.5%Keep in mind Both numbers are people/status counts for the same table, geography and month. The result is not a citizenship census or the share for the full CSJ 2025 estimate.
Calculate recorded employment within the same subgroup
Sum G13:I31 in Table_2. Expect 634,761 people. Those three columns cover PAYE employment only, self-employment only and both. Divide by the same 1,287,180 subgroup total. Do not include F in the employment numerator.
SUM(G13:I31) / SUM(F13:I31) × 100 = 634,761 / 1,287,180 × 100 = 49.3141%, about halfKeep in mind Self-employment classification is based on a reporting requirement, not proof of positive earnings. The result is about half of the defined subgroup, not half of all Universal Credit claimants.
Check the pension comparison within CSJ’s own model
Divide the announcement’s estimated 400,000 pension recipients by its estimated 1,850,000 adult recipients. Read the National Insurance guidance and parliamentary answer separately to understand why pension eligibility is not the same question as nationality.
400,000 / 1,850,000 × 100 = 21.6216%, approximately one in fiveKeep in mind This is a ratio of rounded model outputs. It is not the share of all UK pensioners, nor independent verification of CSJ’s model.
Trace the caption’s 745,000 line without mixing dates
In Table_1, D8 identifies May 2026. D11 gives 744,679 for the EU Settlement Scheme. Round to the nearest thousand to obtain 745,000. Then read the settlement-scheme guidance on family members and later joiners.
744,679 rounded to the nearest 1,000 = 745,000Keep in mind This is a separate May status count. It does not establish that every person is European or has lived continuously in the UK since Brexit. Do not combine May numerators with April employment denominators.
Write the conclusion and preserve the missing denominator
Record what each count measures, its period and population. State that no compatible all-recipient total for CSJ’s entire 2025 basket was recovered. Keep the Universal Credit context separate and avoid claims about individual joblessness, unlawful residence, fraud or net tax contribution.
Keep in mind The existing CSJ model enquiry remains unsent. No new enquiry or automatic watch has been created. The workbook’s next listed release is 17 November 2026; that release alone cannot validate the full model.
Sources you can check
- Original near-two-million Instagram postSammy Woodhouse, @officialsammyuk / 15 October 2026The headline, 1.3 million and 745,000 caption lines, and “Deport them” demand assessed here
- Revealed: almost two million foreign nationals receiving benefits or State PensionCentre for Social Justice / 14 October 2026Opening, Notes to editors and Table 1: 1.85 million adult model estimate, overlap method and 400,000 pension estimate for 2025
- State of DependencyCentre for Social Justice / 13 October 2026Model discussion on PDF pages 191 to 192, printed pages 189 to 190
- Universal Credit immigration status and nationality tables to May 2026Department for Work and Pensions / 18 August 2026Table_2 April 2026 F12:I31, F13:I31 and G13:I31; Table_1 D11 for the separate May EUSS line; Contents for next publication
- Universal Credit quarterly statistics, 29 April 2013 to 14 May 2026Department for Work and Pensions / 18 August 2026Section 3, People on Universal Credit; immigration-status table and employment chart
- Universal Credit statistics: background information and methodologyDepartment for Work and PensionsEmployment, people-on-UC, immigration-status and nationality definitions; naturalisation and zero-payment caveats
- Universal Credit: eligibilityGOV.UKUniversal Credit can support working, out-of-work and unable-to-work people
- Prove you have right of abode in the UKGOV.UKUnrestricted residence and work; British citizenship and some Commonwealth eligibility
- The new State Pension: eligibilityGOV.UKNational Insurance qualifying records, contributions and credits
- State Retirement Pensions: Foreign Nationals, written question 108263UK Parliament, answer from Department for Work and Pensions / 3 February 2026DWP says nationality is not recorded in State Pension claims and refers to National Insurance records
- Apply to the EU Settlement Scheme: who can applyHome Office, GOV.UKEligible family members and later joining routes qualify the simple “Europeans who stayed” description
We assessed Woodhouse’s complete original Instagram caption, the CSJ announcement and report, and the official DWP definitions and tables. The CSJ report and official ODS are reused from preserved originals by matching file hashes. We independently checked the April 2026 table sums and ratios. We did not recover the full Daily Mail article, CSJ’s underlying model or a matching denominator for its entire 2025 estimate. The model enquiry already prepared for the related investigation remains unsent; no new correspondence or reminder is claimed.
Inspect the evidence
Source notes include the links and verification limits. The timeline opens in Excel or any spreadsheet app.
Photo credit and licence
British pound coins and banknotes. Mehaniq/Shutterstock.com. Shutterstock Standard Image License. Capture date not supplied. Resized and cropped for display. The homepage also uses a dark gradient and text overlay. The original stock file is not offered as a download.
Carousel photograph credits and licences
- UK currency illustration
Mehaniq/Shutterstock.com. Close macro photograph of British banknotes and pound coins. Financial-support illustration, not an actual benefit payment or a total of two million pounds.
Shutterstock Standard. Cropped, resized and darkened for the carousel. - Instagram illustration
Yalcin Sonat/Shutterstock.com. Instagram app icon on a phone screen, Antalya, Turkey, 3 March 2025. Platform illustration, not Woodhouse’s phone or her original post.
Shutterstock Standard, main-library Editorial Use Only. Cropped, resized and darkened for the carousel. - Calculation illustration
NMK-Studio/Shutterstock.com. Illustrative financial calculation sheet, pen and calculator. Not actual CSJ data, model, benefit records or recipient totals.
Shutterstock Standard. Cropped, resized and darkened for the carousel. - DWP offices, Leeds 2008
Chemical Engineer. Archive Quarry House DWP and Department of Health offices in Leeds,20June2008. Source-institution context, not a UC-office visit,2026 condition or count evidence.
Public domain dedication by copyright holder. Cropped, resized and darkened for the carousel. - Work illustration
DimaBerlin/Shutterstock.com. Anonymous carpenter hands marking wood. General work illustration, not identified claimant, nationality, location or income status.
Shutterstock Standard. Cropped, resized and darkened for the carousel. - National Insurance illustration
bangoland/Shutterstock.com. Handwritten National Insurance concept note beside orange highlighter and calculator. Illustration, not an official contribution record, actual claimant record or underlying CSJ model.
Shutterstock Standard. Cropped, resized and darkened for the carousel. - Advice sign, 2018
Hogyn Lleol. Archive bilingual Citizens Advice/Cyngor ar Bopeth window sign,23October2018. Advice-service illustration, not a photograph of any claimant, legal-status decision or our contact. Exact town not established.
CC BY-SA 4.0. Cropped, resized and darkened for the carousel. Original and photographic adaptation under CC BY-SA 4.0. - Research illustration
New Africa/Shutterstock.com. Magnifying glass over binder-clipped paper stacks. Research illustration, not the actual CSJ source documents, recovered model or enquiry response.
Shutterstock Standard. Cropped, resized and darkened for the carousel. - Reading illustration, 2004
Bobulous (Robert Downes). British Museum Reading Room, London,8March2004. Reading illustration; no claim these are the actual sources or current research location.
CC BY-SA 4.0. Cropped, resized and darkened for the carousel. Original and photographic adaptation under CC BY-SA 4.0.
These credits cover the carousel photographs. Original stock files are not distributed here. The logo remains separate brand artwork.
Research and production assisted by AI. Read our approach
Have evidence that could change this finding? Contact TALLYCAT01