The Long Squeeze

To offset the impact of tax cuts on the public finances, the Government is considering how it might cut spending. One option that has been discussed is the possibility of raising some benefits in line with earnings rather than inflation next April. This paper explores what this might entail, the potential savings and impacts, and the broader context of how such a change would affect the benefit system and the living standards of those on lower incomes.

Uprating in line with recent earnings growth (5.5 per cent) would permanently cut the real value of affected benefits by an estimated 4 per cent, and could save around £5 billion a year if applied broadly to all benefits except the State Pension and Pension Credit. However, some other benefits are protected in legislation. Excluding these, the potential savings from working-age benefits would be around £3 billion a year – and even this saving can only be achieved by freezing some forms of support that were exempt from earlier benefit freezes. If the cuts affected working-age benefits excluding protected benefits, they would affect 9 million households (7 million of which have someone in work) containing 30 million people, including a majority of working-age households containing children or people with disabilities. Losses would range from around £50 a year for one-child families getting only Child Benefit, to £1,000 or more for some parents on UC, with those in-work losing more in cash terms as UC’s work allowances would (presumably) also be cut.

Coming on top of substantial cuts in the 2010s, a new cut next year would leave the real value of basic out-of-work support 16 per cent lower in 2023-24 than in 2010-11 (and slightly lower than 40 years earlier in 1983-84). And 2023 is a uniquely bad time to be cutting incomes. On our latest projections, even if all benefits are uprated in line with inflation, the real disposable incomes of the poorest may fall by 11 per cent next year, the biggest drop on record and wiping out all income growth accrued over the past twenty years. If some were indexed only to earnings, the drop in income would be 14 per cent. The number of people living in absolute poverty is now projected to rise by 2.9 million between 2021-22 and 2023-24 and a real benefit cut would add another 600,000 people to this rise, including 300,000 children.

In May 2022, Rishi Sunak as Chancellor said that “I can reassure the House that next year, subject to the Secretary of State’s review, benefits will be uprated by this September’s CPI”; this is the usual custom and practice, and it is what is assumed by the Office for Budget Responsibility (OBR) when it makes its fiscal projections. Since then, the announcement of £45 billion a year of tax cuts in The Growth Plan (£2 billion of which have been cancelled), and the associated increase in government borrowing costs, have significantly worsened the fiscal outlook. In response, the Government is due to publish a ‘Medium-Term Fiscal Plan’ on 31 October that is expected to set out how they will get debt falling as a percentage of GDP. Achieving this – in effect, to pay for those tax cuts – will require a significant fiscal tightening. To contribute to this, it has been widely reported that the Government is considering making some savings by uprating some working-age benefits by less than September’s rate of inflation, perhaps by the growth rate of average earnings. Although average earnings usually grow faster than inflation, that is not the case at the moment, with the September rate of CPI inflation expected to be around 10 per cent (we use a figure of 10.1 per cent, from RF’s inflation calculations), whereas average earnings growth in the year to May to July was 5.5 per cent (this is the statistic used for the State Pension triple lock).

The Government might be considering cutting some but not all working-age benefits

The rules governing the uprating of benefits are complex, but it is helpful to think of there being four main categories:

1. The Basic State Pension (old and new) represents a large part of benefit spending (around £120 billion in 2026-27, the final year in the OBR’s current forecast horizon). The Government has made clear that it does not intend to deviate next year from the ‘triple lock’ principle. Given that the September rate of inflation will be higher than the relevant earnings figure, this means that the State Pension will almost certainly be increased in line with inflation in April. (It is not entirely clear what the Government intends to do with the Pension Credit (a £5 billion benefit), but we assume that this will follow the Basic State Pension and be increased in line with earnings.)
2. A specific set of benefits, mostly relating to people with long-term health conditions or disabilities – Personal Independence Payments, Disability Living Allowance, Attendance Allowance, Incapacity Benefit, Severe Disablement Allowance, Industrial Injuries Benefit – but also including Carer’s Allowance, Additional State Pension and Guardian’s Allowance, have to rise with prices at least. The Government could seek to alter this, but doing so would require primary legislation, and the view of commentators has been that this would be unlikely to pass through parliament.
3. Some parts of the benefit system are already permanently frozen by default, chiefly: Local Housing Allowances for Housing Benefit and Universal Credit (UC), the level of the benefit cap(s), and Winter Fuel Payments.
4. The remainder – chiefly UC and its predecessors and Child Benefit – which do not have to be increased by law, and where it would be harder for Parliament to prevent a cut, including:
   • UC (around £76 billion of spending in 2026-27, but this includes housing support which will be frozen, as mentioned above);
   • Child Benefit (around £12 billion);
   • Contributory Employment and Support Allowance (ESA; £5 billion) and contributory JSA;
   • Statutory Maternity/Paternity Pay and Maternity Allowance (around £4 billion);
   • Legacy benefits (which will be largely replaced by UC by 2026-27, but remain significant in the short-term) – tax credits, non-contributory ESA and – to a lesser extent – JSA and Income Support.

It seems likely that the Government is considering what to do with the fourth category above (the benefit freeze of 2016-17 to 2019-20 broadly covered the fourth category, but excluded parental pay, Maternity Allowance, bereavement benefits and benefit elements related to disability and care).

The most likely option is that, instead of increasing those benefits in line with inflation, the Government would use the rate of growth of average earnings (5.5 per cent).[1] If the State Pension, Pension Credit and legislatively protected benefits were increased in line with prices, but others downgraded to earnings uprating, that could raise around £3 billion a year in 2026-27 (following the pattern of earlier benefit freezes and not cutting parts of the benefit system such as Statutory Maternity Pay and disability-related elements of UC would reduce that saving). If all working-age benefits were to be uprated with earnings instead of inflation – i.e. by also freezing those in the second category above – the Government could make savings of around £5 billion a year in 2026-27; as we note above, this would require primary legislation, which is not guaranteed to pass through Parliament.

It is useful to compare this potential £3 billion a year saving to some of the tax cuts that were announced in the Growth Plan: reforms to tax-free shopping will cost £2 billion a year, a lack of enforcement of IR35 rules will cost £2 billion a year, the basic rate cut will cost £6 billion a year, and the Corporation Tax U-turn costs £3 billion a year for each percentage point of the previously announced rise that is cancelled.

Table 1: A one-year change to uprating for ‘unprotected’ non-pensioner benefits may only raise £3 billion a year

For a larger saving, the Government could announce that the same approach will be taken in April 2024. Inflation up to September 2023 is currently expected to be around 7 per cent, whereas earnings growth may again be lower (e.g. 6 or 7 per cent), though these projections – and therefore the possible savings and impacts – are highly uncertain. Here we focus on the decision for 2023, which has to be made shortly.

This change would affect a large number of people

Uprating some benefits by earnings at 5.5 per cent, rather than by CPI inflation at around 10.1 per cent means that they would be 4.2 per cent lower in 2023-24 than they would have been otherwise, and would remain permanently lower in future years.

Figure 1 shows the practical difference that would make for different family types. Compared to inflation uprating, Child Benefit would be cut by £52 a year for the first child and £34 for any subsequent children. The basic rate of UC for someone unemployed over 25 would be cut by £185 a year, while a single disabled adult on UC could lose £380. We assume that UC’s work allowances would also be under-indexed (which also forms part of our costings above) and therefore people in work would lose even more: a single parent in work with one child would lose £478, while a working couple with three children could lose £978 a year.

Figure 1: Depending on family circumstances, losses from this potential policy change range from £50 to £1,000 a year or more

Overall, 9 million households (45 per cent of working-age households) would face an income loss, containing 30 million people. Of those, 7 million households contain someone in work. Even among recipients of UC, many are in work: among those receiving UC in September 2022, 2.2 million are in work, 1.4 million are unemployed, and 2.2 million are not required to search for work. Furthermore, a significant number of households would lose a particularly significant amount of money, with 3 million households (15 per cent of the working-age total) facing an annual loss of over £500. Figure 2 shows the proportion of households in different groups that will face income falls.

Furthermore, some groups are more likely to be affected by the potential changes to benefit uprating than others are. Families with children are most likely be affected by these changes (due to the high proportion of families that receive Child Benefit), with 74 per cent of couples with children and 95 per cent of single parents facing income losses. Although social renters are the housing tenure group most likely to be affected by these changes (with 79 per cent at risk), it is still the case that 38 per cent of households with mortgages are set to see income falls if benefit uprating is reduced – adding to their challenge of rapidly climbing interest rates. And, even though we have assumed that most disability benefits would be inflation-uprated, 63 per cent of households containing someone with a disability are set to face falling incomes. Half of working-age households in the North East, North West, Yorkshire and West Midlands would lose out.

Figure 2: Families with children, workless families, disabled families, and social renters are the most likely to lose income from uprating benefits using inflation instead of earnings

Single-parent and larger families are the most likely to face a large income fall (of over £500) as a result of changes to uprating, with 62 per cent of single-parent families and 55 per cent of families with three or more children expected to see an income fall of at least £500.

Under-indexing benefits has become an unwelcome norm in recent years

Using earnings to uprate benefits instead of inflation in 2023-24 would continue a long-run trend of benefits either failing to grow in real terms or being actively cut. Figure 3 shows that an increase in line with average earnings would take basic levels of unemployment support down to its lowest real level since 1982-83, while using the triple lock to uprate the Basic State Pension would take it to its highest real value ever. Even a full uprating using inflation in 2023-24 would leave benefits temporarily 6 per cent below 2019-20 real-term levels.

In complete contrast to the triple lock that has been in operation for the State Pension, switching to earnings-based uprating for some working-age benefits this year would suggest a ‘triple knock’ policy, whereby benefits are uprated by the lower of earnings or inflation, or by some arbitrary value such as 1 per cent or zero. Since 2011, Child Benefit, for example, has had nine years of freezes or 1 per cent uprating, rose in line with CPI for the last three years, and is now at risk of a below-inflation earnings uprating.

Figure 3: Further real-term cuts would take unemployment benefits back to 1980s levels

Expressed as a fraction of earnings, a below-inflation uprating in April 2023 would reduce basic unemployment support to 13 per cent of Average Weekly Earnings (the same as in 2022-23), which is the lowest level since the post-war welfare state was invented.

The outlook for living standards and poverty is already dire, making this a particularly difficult time to cut social security

A below-inflation uprating in April 2023 would represent a permanent benefit cut, but the seriousness of the short-term context should be stressed, with a terrible outlook for 2023-24 and the next few years beyond. Overall prices may be another 7 per cent higher in 2023-24 than in 2022-23. The welcome energy price cap of £2,500 (for typical usage) is nonetheless 27 per cent higher than the April 2022 price cap of £1,971. The £400 Energy Bills Support Scheme will end in March; the £150 Council Tax Rebate will also not be repeated in 2023-24; and nor will the £650 Cost of Living Payments. Although some taxes are being cut, average mortgage interest payments may rise by around 50 per cent in 2023-24.

As a result, our latest projection, on the assumption that all working-age benefits are indexed in line with inflation, is that the typical real non-pensioner household income will fall by 7 per cent in 2023-24 (on top of a 4 per cent drop in 2022-23), and that falls will be even larger for lower-income households. For the typical non-pensioner in the bottom fifth of the income distribution (i.e. someone at the 10^th percentile), we project an 11 per cent real income fall in 2023-24. This would be the worst drop on record (back to 1962), even assuming that all working-age benefits are indexed in line with inflation; increasing the core means-tested benefits and Child Benefit in line with earnings would mean that the income fall next year for the typical non-pensioner in the bottom fifth of the income distribution would be 14 per cent. This would be twice as large as the previous record in our data: a 7 per cent fall in 1981 (a year when unemployment rocketed). With the benefit cut, the incomes of the poorest may be 20 per cent lower next year than they were in 2020-21. (Note that none of these calculations accounts for the fact that inflation is running higher for low-income families: in the year to August, inflation was an estimated 10.6 per cent for the poorest, compared to 9.9 per cent overall and 9.0 per cent for the richest tenth of the population.)

Figure 4: Even with 10 per cent uprating, 2023-24 is projected to be a catastrophically bad year for household incomes, and benefit cuts would make it worse

Not only is our projection for 2023-24 currently terrible, but the years beyond are also forecast to bring falling disposable incomes (in part due to rising mortgage interest payments). With the cut, typical real incomes for the poorest fifth of households would fall back next year to levels last seen over twenty years’ ago.

Figure 5: For poorer households, 2023-24 may bring the biggest real income fall on record – wiping out the small gains that occurred over the twenty years up to 2022-23

These prospects are also reflected in the outlook for absolute poverty, defined as the number of people in a household with less than 60 percent of the median income in 2010-11. The proportion of people living in absolute poverty is now projected to rise from 17 per cent to 21 per cent between 2021-22 and 2023-24 (an increase of 2.9 million people) and a real benefit cut would add another 600,000 people to this rise (including 300,000 children). Overall, under-indexing the core means-tested benefits and Child Benefit would mean that the number of people in absolute poverty in 2023-24 could top 15 million. Partly due to increasing mortgage costs, absolute child poverty is now projected to rise to 34 per cent of children by 2026-27 even without any new benefit cuts, higher than any year since 2000-01.

Figure 6:  Benefit cuts would further worsen prospects for absolute poverty, particularly for children

Conclusion

It cannot be argued that basic out-of-work support or Child Benefit, for example, are generous by historical or international standards. As we have set out in other work, the UK has a problem of unusually high inequality as well as poor growth. But beyond these long-term concerns, 2023 is clearly among the worst times possible to further lower the incomes of the poorest. Cutting working-age benefits by around 4 per cent, permanently, on top of the benefit cuts of the 2010s, and in the middle of an enormous cost of living crisis, would be a very contentious decision. Doing so to pay for tax cuts that particularly benefit those on higher incomes would be all the more so. The fact that the Government is having to consider these options – and perhaps also other, more painful, cuts to social security spending – so as to repair the damage to the public finances illustrates the risks the Government took when jettisoning fiscal conservatism alongside its initial Growth Plan.

[1] We note that the rate of inflation is likely to rise higher than that recorded in September, which will not capture the rise in energy costs on 1 October.