In November 2025, the Centre for Social Justice (CSJ) published a report titled The Benefits Budget. One claim from that report was widely shared:

“An out-of-work family on combined benefits will now receive £18,000 more than the post-tax earnings of a family on the living wage.”¹

It is a striking figure. It suggests a welfare system so distorted that families are significantly better off not working than working. It also feeds directly into a familiar narrative about a “welfare crisis”, perverse incentives, and social breakdown.

Because of the weight such claims carry particularly for disabled people and families in poverty, I read the report in full and raised concerns about how these figures were constructed and communicated.

This post explains why I did so, what the report actually shows, and what an independent review later confirmed.

What the report actually does and where

The central comparison appears in the section “The abolition of the two-child limit” (pp. 9–11 of the report).

On page 10, CSJ states that:

“An out-of-work family with three children receiving the average Universal Credit housing element, health benefits and PIP is projected to take home around £46,000 a year… For families with five children, that figure rises to £55,000.”²

They then compare this to a working household:

“A working family… would take home roughly £28,000 after tax – £18,000 less than the benefit income now available to an equivalent three-child family outside work.”²

These figures are illustrated visually in Figures 4 and 5 on pages 11–12, which explicitly stack:

• Universal Credit standard allowance

• Universal Credit child elements

• Universal Credit housing element

• UC LCWRA (Limited Capability for Work and Related Activity) element

• Personal Independence Payment (PIP)³

What matters is not only what is included, but what is not explained.

Why I raised concerns

1. The report relies on maximum theoretical benefit combinations

Across pages 10–12, the report models households receiving multiple high-value benefit components simultaneously: housing support, LCWRA, and PIP, alongside full child elements and exemption from the benefit cap.

Nowhere in the report (pp. 1–19) does CSJ provide:

• the number of real households who receive all these components together

• median or typical combined benefit awards

• distributional data showing how common these scenarios are

The examples therefore represent theoretical maxima, not representative cases — but this distinction is not made explicit in public-facing claims.²³

2. Housing assumptions are not disclosed

The report refers to the “average Universal Credit housing element” (p. 10) but does not specify:

• which region this average applies to

• whether it assumes social or private rented housing

• the Local Housing Allowance band used

• household size assumptions for rent calculations

Given the scale of regional variation in housing costs, these omissions materially affect the headline figures.²

3. Disability benefits are assumed without justification

Both PIP and LCWRA are included by default in the example households (pp. 10–12; Figures 4 and 5).³

However, the report does not explain:

• why it assumes a disabled adult in every example household

• whether larger families are more likely to include PIP claimants

• how common it is for households to receive both PIP and LCWRA

The inclusion of disability benefits significantly inflates totals, yet the assumptions behind this choice are not set out anywhere in the document.

4. The comparison itself is asymmetrical

This is the most important issue.

On one side, the report presents:

• a stacked, high-end out-of-work benefit package

On the other, it compares this to:

• the pre-tax earnings required to reach the same income without any state support (p. 10)

Throughout the report, there is no acknowledgement that:

• Universal Credit continues when someone enters work

• UC is tapered, not withdrawn immediately

• PIP is unaffected by employment status

The comparison made is therefore not:

out-of-work income vs in-work income

but:

out-of-work income vs earnings stripped of in-work support

That distinction is central to any honest discussion of work incentives.²

5. The “5,000 people a day” claim lacks context in the report

The report repeatedly states that:

“5,000 people a day are being signed off work and onto long-term sickness benefits”⁴

This appears in the Foreword (p. 2) and again in the main analysis (p. 12).

No source, dataset, or methodology is provided in the report itself explaining:

• what “signed off” refers to

• whether this concerns new claims, assessments, or administrative decisions

• whether it reflects people permanently leaving work

For a statistic used to frame a “welfare crisis”, this lack of context is significant.⁴

What an independent review later confirmed

After raising these concerns, I received a response from an independent body that reviews the public use of statistics.

On the “5,000 a day” figure

They confirmed that the figure can be derived from Department for Work and Pensions (DWP) StatXplore data by:

• combining Universal Credit and ESA Work Capability Assessment decisions

• filtering to initial assessments only

• dividing monthly totals by the number of working days⁵

However, they also agreed that as presented publicly, the statistic lacked sufficient context to be understood accurately, because none of this methodology was explained.

On the £18,000 / £55,000 comparison

Crucially, the independent response stated that:

comparing an out-of-work household receiving housing support, LCWRA and PIP with the pre-tax earnings required to match that income without any state support “had the potential to mislead”.⁵

They explained that if someone in such a household moved into work at 35 hours per week on the national living wage:

• they would retain their PIP

• continue to receive Universal Credit, tapered as earnings rise

• and be almost £12,000 better off overall⁵

This directly undermines the impression created by the headline claim that families are financially better off remaining out of work.

Why this matters

This is not a technical disagreement. It affects how disabled people, families in poverty, and the welfare system are understood in public life.

When extreme or unrepresentative scenarios are presented as typical, and when key assumptions are left unstated; the result is:

• distorted public understanding

• increased stigma towards disabled people

• political pressure for punitive welfare reforms

• reduced focus on low pay, insecure work, housing costs, and health inequalities

Transparency is not optional when statistics are used to shape policy. It is essential.

The issue is not whether certain benefit combinations are technically possible. It is whether they are presented honestly, proportionately, and with sufficient context. Comparing maximum theoretical benefit packages with wages stripped of in-work support does not explain how the welfare system functions in practice. It obscures it. That is why I raised concerns, and why accuracy and transparency in welfare statistics matter, not only for policy, but for the dignity and lives of the people affected.

Endnotes

1. Centre for Social Justice, The Benefits Budget (November 2025), Executive Summary / public communications.

2. Ibid., p. 10.

3. Ibid., Figures 4–5, pp. 11–12.

4. Ibid., Foreword p. 2; main analysis p. 12.

5. Independent correspondence reviewing the public presentation of these statistics, summarising DWP StatXplore methodology and comparison effects (2025).