Executive summary

The government is in the process of conducting a review of universal credit (UC), the main means-tested benefit for working-age people, received by 6.3 million families at any one time. This report discusses some of the challenges UC faces, and outlines potential directions for reform.

1. Introduction

Universal credit (UC) first began to be rolled out in 2013, replacing a patchwork of six previous benefits. Thirteen years later, that roll-out is essentially complete, and UC now provides means-tested support for families who are unemployed, with caring responsibilities, in work on a low income, or unable to work because of ill health.1 In August 2025, 6.3 million families received at least some UC, representing 22% of all working-age families. Many more will get it at some point in their life. Given how many families receive UC, getting its design right is quite reasonably going to be a government priority.

The Labour manifesto committed to ‘reviewing Universal Credit so that it makes work pay and tackles poverty’. There is little publicly available about this review – the government has stated that it has begun, though as things stand there are no terms of reference.2 Any review of a benefit will naturally want to cover the big-picture questions about the overall generosity of the system and whether it provides the right work incentives, but – for any given amount of spending – the government should want to make the most of the resources used and ensure it is designed to work as well as possible for claimants. With that in mind, this short report considers a range of challenges that UC faces, and reviews the available policy options.

2. Universal credit and council tax support

Council tax support (CTS) is a benefit targeted at poorer households which offsets some or all of their council tax liability. It is separate from UC and, since 2013–14, has been ‘localised’ – English councils and the Scottish and Welsh national governments both administer the scheme and have broad latitude to design it as they wish.

A detailed investigation of CTS and its interaction with UC is available in Oulton and Wernham (2026). Here I summarise some of the key findings.

First, localisation has resulted in a complex patchwork of choices made by different local authorities. The schemes vary in terms of what proportion of bills they support, how much income claimants can have before support begins to be withdrawn, how this varies between different types of family, and even how to measure income. It is a significant administrative burden for local authorities to design and administer these schemes.

Second, the fact that CTS operates separately from UC means claimants on low earnings can see both reduced simultaneously, weakening work incentives. A typical case is that someone earning an extra £100 loses £55 in UC and £9 in CTS. Particularly weak work incentives to increase earnings by a small amount are sometimes observed in ‘banded schemes’, a way that CTS is designed in many local authorities which bases entitlements on which income band the claimant is in. This means that CTS receipt jumps down when income passes over certain thresholds. As a result, earning a little more can make a claimant worse off – something UC was set up specifically to avoid.

The case in favour of localisation is that it gives local authorities the opportunity to design schemes in view of local political preferences or constraints. But this needs to be set against the disadvantages. The existing system creates significant bureaucracy for local authorities, which could be removed if CTS were integrated into UC. Integration would also simplify the system for claimants in terms of reducing the number of applications they need to make to get their full entitlement. And it would ensure that they face a single taper rate out of benefits. Moreover, as Oulton and Wernham (2026) illustrate, there would be scope for maintaining a degree of local decision-making even if CTS were integrated with UC. Taken as a whole, integration offers a compelling case for reform.

3. Indefinite cash freezes

Several parts of the UC system are indefinitely frozen in cash terms, meaning that in real terms they are steadily eroded by inflation. This means both that the system becomes less generous without any ‘intentional’ or active decision by policymakers, and that the speed of that decline in generosity is up to the unpredictable undulations in inflation.

The most consequential example concerns housing support for private renters. The maximum housing support available through UC (and, for pensioners, housing benefit) is tied to the claimant’s local housing allowance (LHA) rates. Prior to 2013–14, LHA rates were set such that a claimant with no other private income could get support up to the 30th percentile of local rents (i.e. they could have all their rent covered if they rented one of the cheapest 30% of properties). Since then, LHA rates have been delinked from local rents, being uprated at different points by CPI inflation, by 1% per year or – as is now the case – frozen in cash terms. This has led to the generosity of housing support tending to fall relative to rents, and the government has responded to this by twice resetting LHA rates back to the 30th percentile (in 2020 and 2024) – only to begin the freeze again and restart the clock on this process.

Since the most recent reset (which set LHA rates to the 30th percentiles as they stood in the year to September 2023), keeping LHA rates frozen has resulted in an average loss of almost £1,500 per year for over a million households (Michael and Wernham, 2025). But behind those averages are arbitrary geographic differences – those areas where rents are rising fastest see a larger gap open up between actual rents and the maximum support.

UC also has various other thresholds and values that have been left unchanged for long periods, including the benefit cap (discussed in more detail below), the extra support claimants get for first children,3 maximum childcare support, and ‘asset limits’ which govern how much in assets families can have before their UC starts to be withdrawn or ended entirely. For example, the existing asset limits prevent families with more than £16,000 in savings from receiving any UC. This £16,000 figure was based upon the benefits that UC replaces and has not been changed since April 2006. If it had kept pace with inflation, today it would be more than £28,000.

This approach is hard to defend as a coherent way to run a safety net. An effective benefit system should not ‘accidentally’ drift into being less generous over time by freezing benefit values, any more than a good tax system should become more burdensome over time through freezing tax thresholds (‘fiscal drag’). If the government believes it is appropriate for support to fall, it should say so and specify in what way. If it believes support should maintain its value, it should build that into the rules by uprating thresholds and values by a well-chosen index (such as local rents for LHA rates, or CPI inflation or earnings growth for most other parameters).

4. The benefit cap

The benefit cap is an upper limit on the total amount of benefit income (in practice, UC plus child benefit) a household can receive. Different caps are applied to claimants who live alone and those with children or a partner, and to those who live in Greater London and those living elsewhere in the country.4 For example, the benefit cap for couples or those with children outside of London is £22,020 per year. The cap is not applied to those in paid work earning above a low threshold or those in receipt of disability, incapacity or carer’s benefits. It is also not applied for the first nine months of unemployment.

In practice, the cap primarily affects those who have been out of paid work for some time, who are not disabled, and who have several children and relatively high rent.

Presumably, the goal of the benefit cap is to eliminate benefit entitlements that the government considers to be too high. But it is worth seeing this in the broader context. UC is designed to provide support for different needs: rent, children, number of adults in the household and so on. If policymakers look at the total support a family receives and judge it to be too high, the most natural interpretation is that support for at least one component is too high. In that case, the right response would be to change the relevant part of the system – for example, reducing support for housing costs or reducing the child element. Applying a simple cash limit does not map onto a view about the minimum living standards the government wants to guarantee, because the living standards a given cash amount affords vary by the needs of the household (such as the number of adults, number of children and possibly the level of rent5).

One potential justification is that some needs are at least partly chosen. The cap can be viewed as saying, in effect, that the state will not fully support every combination of choices – claimants can choose to have a larger family or live in a high-rent area and receive full support for these needs, but not both. But even if one accepts that logic, it sits uneasily with the fact that the cap is higher for those living in London – which is itself at least in part something claimants can choose. It is not clear that there is a coherent view of the benefit system and what it is trying to do behind this design. The UC review should seek some clarity on a basic question: ‘What problem is the cap intended to solve, and is it the best tool for solving it?’.

These issues matter even more given the planned removal of the two-child limit in UC from April. 141,000 children (in 38,200 families) are affected by the two-child limit and the benefit cap.6 These families will not gain from the removal of the two-child limit since their overall entitlements will remain capped – and these children are especially likely to be in ‘deep poverty’ (i.e. live in households with income some distance below the official poverty line). Thus, while the two-child limit removal is a well-targeted way to reduce headline poverty statistics, the existence of the benefit cap means it is considerably less effective at reducing deep poverty. Measuring deep poverty is difficult for data quality reasons, but the existing evidence we have suggests that removing both the two-child limit and the benefit cap would reduce deep poverty by about twice as much as removing the two-child limit alone, at a roughly one-third increase in cost (Henry and Wernham, 2024). 

5. Payment timing: the five-week wait, advances and earnings volatility

Unlike the system it replaced, UC assesses entitlement on a monthly basis in arrears. In other words, the government waits until a month-long ‘assessment period’ has passed, and then pays claimants their UC on the basis of their earnings in that month. This design has clear advantages. In particular, it reduces the risk of large underpayments and overpayments – common in the tax credits system – that later have to be corrected, which can create hardship and unpredictability.

However, another consequence of this choice is the well-known ‘five-week wait’ for a first payment – an inevitable consequence of UC being assessed over a monthly period and paid shortly after that period ends. That naturally creates difficulty for new claimants without savings, access to cheap credit, earnings, or informal support from family.

The system’s main mitigation is advance payments – effectively a zero-interest loan, generally repaid through lower future UC receipt. Since the advent of UC, these advances have become more generous – both through increases to the maximum amount one can get and through decreases in the speed at which the advance is recouped through lower UC (around a third of families on UC are now repaying advances at any point in time7). Clegg et al. (2026) calculate from administrative data that 56% of new UC claimants took out an advance in 2024–25. Given that the advance is a zero-interest loan, one would typically think it worth taking (at a minimum, simply putting the money in a savings account would make the claimant better off relative to not taking the loan at all). It is difficult to believe that 44% of new claimants would be worse off from taking an advance.

However, qualitative evidence in Crossfield and Pinakova (2024) suggests that some claimants are not informed about the option to get the advance, or do not take it because of an aversion to debt – in some cases going into arrears on other payments or using credit (potentially with interest), creating a risk of longer-term financial difficulties. Policymakers should ensure that claimants are given sufficient information and a clear choice about advances. Again, given that the advance is zero-interest, it would likely be in the claimant’s interest to receive it. One option to consider is to learn from the experience of pensions auto-enrolment and similar policies and provide the advance by default with an option to opt out (though this would come at some cost to the government). 

A second issue is earnings volatility – for example, through individuals working different numbers of hours in different months. The most immediate impact of UC in this regard is to cushion such volatility – in periods when earnings are high, UC is low, and vice versa – which is an attractive feature of the system. However, in some cases, volatility can result in higher or lower amounts of UC, compared with having the same total earnings with no volatility.8 This is an undesirable aspect of UC because it treats similar people differently and so is arguably unfair. But it is an inevitable consequence of having finite assessment periods: the same thing happens in the National Insurance contributions system (assessed weekly or monthly) and, at an annual level, in the income tax system.9

This could be mitigated by allowing UC claimants to opt into a system that assesses them based on their average earnings across the last few months, as suggested by Clegg et al. (2026). However, it would make the system less responsive to changes in earnings – for those who had opted into averaging, their UC would take much longer to rise at job loss, for example. And those whose earnings fluctuate in a way that allows them to gain from volatility would be unlikely to opt in, thereby not doing anything for the unfairness among that group.10 

A third, related issue concerns workers paid every four weeks. Because there are 13 four-week pay periods in a year, once a year such workers have an assessment period in which they receive two pay cheques, resulting in their earnings appearing unusually high. In that month, UC entitlement falls sharply, sometimes to zero. This creates practical problems – including the risk of losing other support that is linked to UC receipt (as discussed below). This is not a minor issue – 10% of employees on UC are paid every four weeks.11 The difficulty is that this volatility is, to some degree, hardwired into UC’s design. Fixing it would likely require averaging earnings across months (as discussed above), or redefining assessment periods, or introducing smoothing mechanisms – all of which bring their own complexity and may undermine some of the very features (such as responsiveness to real-time income changes) that UC was designed to deliver.12 A less comprehensive but perhaps more feasible option, suggested by Clegg et al. (2026), would be for the Department for Work and Pensions (DWP) to include a flag on the claimant’s UC journal (the main mechanism for communication between the department and the claimant) when their next UC payment is likely to be low due to two pay cheques falling in the same assessment period.

6. Cliff edges in benefit entitlement

One of the fundamental goals of UC at its inception was, as the then Work and Pensions Secretary put it, ‘to make work pay, even for the poorest. This will finally make it easier for people to see they will be consistently and transparently better off for each hour they work and every pound they earn’ (Department for Work and Pensions, 2010). This goal – for UC to ‘make work pay’ – is repeated in the Labour manifesto as referenced above.

But so far, it has not been achieved. This is because there are numerous strands of support, outside the UC system itself, which are tied to getting at least some UC, or getting some UC and having earnings below a particular threshold – generating a ‘cliff edge’ in support. The ‘banded schemes’ in council tax support described above are one example. Another is free school meals, which are available to children in families whose parents are on UC and earn under £7,400 per year.13 For a family with two children, that means earning £7,399 and keeping free school meals could be financially preferable14 to earning anything up to £9,400 – equivalent to turning down an extra four hours of work a week at the National Living Wage (Cribb et al., 2023). Likewise, receipt of Scottish child payment requires getting at least some UC. Someone entitled to a tiny amount of UC can get £1,410 per year per child in Scottish child payment; if they increase their earnings slightly, they lose all of that. Ray-Chaudhuri and Waters (2025) show that, for an example worker in this situation with three children, this can mean that raising working hours by anything up to nine hours per week at the National Living Wage can leave them worse off. There are many cases of this – Griffiths and Wood (2024) list a huge range of strands of support that depend upon receiving UC and include a cliff edge, from Healthy Start vouchers to free NHS prescriptions to the Warm Home discount. This is not a minor issue: Elsom (2026) calculates that the total value of support that is tied to UC receipt is well into the billions of pounds per year.15

In many (though not quite all) cases, these cliff-edged strands of support are administered by a department other than DWP, making integration with the UC system – and avoidance of a cliff edge – more challenging. Some forms of support are not even provided by government – for example, some charities tie support to receipt of UC.16 And in some cases, the support being provided is an in-kind good (such as free school uniforms), making it difficult to ‘taper away’, as one can with cash.

Despite these difficulties, more can probably be done in at least some instances. For example, Cribb et al. (2023) argue that in the case of free school meals, cash card systems would allow the government to reduce the share of the school meal price that the government covers as parental earnings increase. If DWP administered the Scottish child payment on behalf of the Scottish Government, it could essentially incorporate it into the UC calculation as a higher child element – and hence allow it to be smoothly tapered away with the rest of UC.

It may not be possible to remove cliff edges in all cases. But their existence is an unattractive part of UC that undermines its original goals, so it is well worth looking for creative solutions. 

7. The balance of support for unemployment and incapacity

Over the past few decades, the gap between the level of support provided to someone who is unemployed and the level of support provided to someone deemed unable to work because of ill health has widened considerably (Henry et al., 2025), leading the government to raise the concern that this may push people to claim health-related rather than unemployment support.17 Today, the basic support provided for an unemployed individual (before additions for children and rent) is £400 per month – a relatively low level compared with most high-income countries (Mikloš and Xu, 2025). If that person is assessed as being unable to work because of ill health, they receive an extra £423 (the ‘health element’), giving them a total of £823 per month, and are exempt from requirements to search for a job (they may also receive personal independence payment, providing a further £127–812 per month).

The government has announced reforms to reduce support for ill health and increase it for unemployment (and in-work claiming), with the goal of improving the ‘basic adequacy’ of UC and reducing incentives to claim the health element (Department for Work and Pensions, 2025). The basic level of UC is set to rise by 5% in real terms over the coming years, while, from April, the extra support for those in ill health provided to new claimants will halve – though those assessed as having particularly severe disabilities will continue to receive the existing higher rate.18 By 2029–30, the basic support for an unemployed individual will, in today’s prices, be £427 per month (up from £400 now) and £628 if they are in ill health (down from £823 now).

These decisions to rebalance away from ill health and towards unemployment embody choices across a number of trade-offs. On the one hand, low levels of basic out-of-work support and higher levels of health-related support increase the incentives to claim those health-related benefits. Cribb et al. (2025) study four separate cuts to non-health-related benefits implemented in the past two decades and finds that in each case the cut resulted in more claims to disability benefits (such as personal independence payment). In addition, Delestre and Waters (2026) find evidence of significant falls in living standards and rises in financial distress among newly unemployed benefit claimants, especially for those who see little of their income replaced with higher benefits. Increasing basic UC goes some way to addressing this issue.

On the other hand, if someone’s health limits their ability to work, they may be less able to respond to financial incentives. All else equal, that counts in favour of tilting support towards disabled claimants versus non-disabled ones, since it will have less of a distortionary impact on their work choices. In addition, those in ill health are more dependent on benefits for their living standards as they are more likely to be out of work for long periods of time (and hence cannot simply rely on savings to tide them over a short period of unemployment), making them a particularly needy group.

There are no easy answers here, and in some sense the government is hoping to plot a path through these trade-offs by reserving higher amounts of health-related support for those assessed as having the most severe disabilities. Whether this captures the intended group, ends up being narrower than that, or is applied more widely in practice, is yet to be seen.

8. Conclusion

Universal credit was designed with laudable aims – to simplify a fragmented system, make it easier to claim one’s full entitlement and strengthen work incentives. It has in many ways made progress on these goals – not to mention being able to process a huge number of claims during the pandemic. Yet, as this report has shown, significant challenges remain. Some of these are difficult to make progress on. But in other areas, a reform-minded policymaker could do more. Integrating council tax support into UC, ending indefinite freezes, and tackling the cliff edges that undermine work incentives are all areas where sensible reforms are available and the case for action is strong. The UC review represents an opportunity to make progress on these issues. 

References

Clegg, A., Judge, L., Light, M. and Patrick, R., 2026. Listen and learn: improving the way that universal credit works. Resolution Foundation report, https://www.resolutionfoundation.org/publications/listen-and-learn/.

Cribb, J., Farquharson, C., McKendrick, A. and Waters, T., 2023. The policy menu for school lunches: options and trade-offs in expanding free school meals in England. IFS report, https://ifs.org.uk/publications/policy-menu-school-lunches-options-and-trade-offs-expanding-free-school-meals-england.

Cribb, J., Karjalainen, H., Latimer, E., Ray-Chaudhuri, S. and Waters, T., 2025. Do disability benefit claims rise when other benefits are cut? IFS report, https://ifs.org.uk/publications/do-disability-benefit-claims-rise-when-other-benefits-are-cut.

Cribb, J., Wernham, T. and Xu, X., 2023. Housing costs and income inequality in the UK. IFS report, https://ifs.org.uk/inequality/wp-content/uploads/2024/01/IFS-Report-Housing-costs-and-income-inequality-in-the-UK-edited.pdf.

Crossfield, J. and Pinakova, P., 2024. Take up and use of the universal credit advance payment. DWP Research Report RR 1076, https://www.gov.uk/government/publications/take-up-and-use-of-the-universal-credit-advance-payment/take-up-and-use-of-the-universal-credit-advance-payment.

Delestre, I. and Waters, T., 2026. Unemployment, benefits and household spending: new evidence from UK bank account data. IFS report, https://ifs.org.uk/publications/unemployment-benefits-and-household-spending-new-evidence-uk-bank-account-data.

Department for Work and Pensions, 2010. Welfare Reform White Paper: universal credit to make work pay: radical welfare reforms bring an end to complex system. Press Release, https://www.gov.uk/government/news/welfare-reform-white-paper-universal-credit-to-make-work-pay-radical-welfare-reforms-bring-an-end-to-complex-system.

Department for Work and Pensions, 2025. Pathways to work: reforming benefits and support to get Britain working Green Paper. https://www.gov.uk/government/consultations/pathways-to-work-reforming-benefits-and-support-to-get-britain-working-green-paper/pathways-to-work-reforming-benefits-and-support-to-get-britain-working-green-paper.

Elsom, C., 2026. The hidden benefits bill: how universal credit claimants get £10 billion in extra benefits. Onward report, https://ukonward.com/reports/the-hidden-benefits-bill-how-universal-credit-claimants-get-10-billion-in-extra-benefits/.

Full Fact, 2026. Is the government on track with its review of universal credit? https://fullfact.org/government-tracker/universal-credit-review/.

Griffiths, R. and Wood, M., 2024. Cliff edges and precipitous inclines: the interaction between universal credit and additional means-tested help for working claimants. Institute for Policy Research report, https://www.bath.ac.uk/publications/cliff-edges-and-precipitous-inclines-the-interaction-between-uc-and-additional-means-tested-help/attachments/cliff-edges-and-precipitous-inclines-research-report.pdf.

Henry, A., Latimer, E., Oulton, M., Ray-Chaudhuri, S. and Waters, T., 2025. IFS response to announced reforms to disability and incapacity benefits. IFS comment, https://ifs.org.uk/articles/ifs-response-announced-reforms-disability-and-incapacity-benefits.

Henry, A. and Wernham, T., 2024. Child poverty: trends and policy options. In C. Emmerson, P. Johnson and K. Ogden (eds), The IFS Green Budget: October 2024. https://ifs.org.uk/publications/child-poverty-trends-and-policy-options.

Labour Party, 2024. Change: Labour Party manifesto. https://labour.org.uk/wp-content/uploads/2024/06/Labour-Party-manifesto-2024.pdf.

Michael, J. and Wernham, T., 2025. Freezes in housing support once again widen geographic disparities for low-income renters. IFS comment, https://ifs.org.uk/articles/freezes-housing-support-once-again-widen-geographic-disparities-low-income-renters.

Mikloš, M. and Xu, X., 2025. Options for unemployment insurance. In C. Emmerson, K. Ogden and B. Zaranko (eds), The IFS Green Budget: October 2025. https://ifs.org.uk/publications/options-unemployment-insurance.

Oulton, M. and Wernham, T., 2026. How does council tax support shape household incomes and work incentives? IFS report, https://ifs.org.uk/publications/how-does-council-tax-support-shape-household-incomes-and-work-incentives.

Ray-Chaudhuri, S. and Waters, T., 2024. Universal credit: incomes, incentives and the remaining roll-out. IFS report, https://ifs.org.uk/publications/universal-credit-incomes-incentives-and-remaining-roll-out.

Ray-Chaudhuri, S. and Waters, T., 2025. Two-child limit mitigation in Scotland would help larger poor families but policy design could harm work incentives. IFS comment, https://ifs.org.uk/articles/two-child-limit-mitigation-scotland-would-help-larger-poor-families-policy-design-could.

Acknowledgements

I am grateful for feedback on earlier drafts from Jonathan Cribb, Eduin Latimer, Helen Miller and Tom Wernham. I gratefully acknowledge the support of the ESRC Centre for the Microeconomic Analysis of Public Policy (ES/Z504634/1).