In this comment, Kate Ogden and David Phillips draw on their new report for the IFS Deaton Review of Inequalities to summarise long-term changes in the distribution of public service spending in the UK – and to discuss how recent and planned policy reforms are changing things.
On the eve of the COVID-19 pandemic in 2019–20, the UK government spent a total of £509 billion on providing public services such as healthcare, education, defence, policing and social care – equivalent to an average of £7,600 per person. This was more than twice as much as was spent on the benefits, tax credits and state pensions that provide cash directly to households, and equivalent to 22.5% of GDP. Increases in public spending and the economic challenges seen over subsequent years will push public service spending towards a quarter of national income this year.
Because this spending is funded via taxation, it would redistribute significantly from richer to poorer people, even if the same amount was spent on providing public services to everyone. In fact, more is spent on poorer individuals and families for a range of public services, including the two largest: health and schools.
Those in the lowest-income tenth of the population report spending almost twice as many days as an inpatient in NHS hospitals, and almost 50% more GP appointments than those in the highest-income tenth. Research suggests residents of the most-deprived fifth of neighbourhoods receive around one-fifth more healthcare spending than those in the least-deprived fifth, at the same age – with the differences based on family income almost certainly much higher again. This higher usage reflects the substantially worse health, on average, of low-income people compared with high-income people.
Funding per pupil in England was around 23% higher in 2019 for the fifth of schools with the highest share of pupils in receipt of free school meals than the fifth with the lowest share. The 7% of English pupils in private schools, most of whom are from families towards the top of the income distribution, save the government around £3.5 billion a year, while enjoying about twice as much spending on their education as their state school counterparts.
Long-term changes in the distribution of public service spending
Spending is progressive in absolute terms if more is spent on services for the poor than for the rich. It is progressive in relative terms if spending represents a higher fraction of the income of the poor than the rich. The impact of public service spending on the distribution of resources across households (which we term its redistributive impact) depends on both the progressivity and the level of spending – £1 spent on only the poor would be extremely progressive, but effectively meaningless in terms of redistribution.
Both the progressivity and amount of redistribution undertaken by public service spending has changed over time – driven by changes in the level and composition of spending, how spending is allocated between people, and changing patterns of service usage.
In a book published in 1987, Robert Goodin and Julian Le Grand argued that the biggest beneficiaries of spending on many public services were the middle class. This was particularly true of education, where the fact that children from high-income backgrounds were more likely to stay in education post-16, and the high levels of funding per pupil for further and, particularly, higher education, more than offset the higher funding per pupil for schools serving poorer families.
In the years since, public service spending has become more progressive and its redistributive impact greater, at least when measured relative to families’ and individuals’ long-term incomes.
Health spending increased over three-fold in real terms between 1988–89 and 2019–20, growing from 3.9% of GDP to 7.3%, and from under a fifth to almost a third of public service spending. As discussed above, healthcare use and spending is much higher for lower income households, reflecting their poorer health, so the growth in this component of public spending at the expense of things such as defence spending, will have increased how redistributive health spending is.
A number of factors have contributed to education spending becoming more progressive over recent decades. First, is the increase in participation of children from poorer families in further education (for those aged 16–18). In the 1970s, children of ‘professionals, employers and managers’ were 3.5 times more likely to attend further education than children of those with ‘semi and unskilled manual’ jobs. By 2010, the gap between the share of children from the richest fifth and poorest fifth of families who had undertaken at least some further education studies had narrowed to just 3 percentage points (99% versus 96%). Second are changes in education funding policy. Spending per pupil has increased by much more for school-aged children, than for those in further and higher education. Further education spending, long more generous for those in school sixth forms than in the further education colleges where poorer students often study, has seen this pattern flip since 2010. And the introduction of tuition fees, repaid via income-contingent loans has seen high-earning graduates, who often come from higher-income backgrounds, largely pay for their studies themselves, while low earners have continued to see significant government funding via loan write-offs.
The paring back of some areas of public service provision has also seen provision focused on the poorest and neediest. This includes social housing, the tenants of which have gone from having incomes averaging 85% of the population-wide median in 1979 to 70% in 2019, and social care, for which financial means-tests and care needs-tests have become more stringent, particularly during the 2010s. However, while more progressive, the increased rationing of these services means they have become harder to access, even among the poor.
The impact of recent and planned changes in policy
Even as the coalition government enacted regressive tax and benefit policies in the early 2010s, and cut public service spending overall, it was making much of that spending more progressive. For example, health spending in England continued to increase, albeit very slowly, and the introduction of the pupil premium led to spending per-pupil in the fifth of schools with the most pupils on free school meals rising to 35% higher than those with the least by 2013.
More recently though, policy has been working in the other direction, undoing some of the previous increases in progressivity. After 2013, spending per pupil fell significantly among the English schools with the most deprived pupils, but increased slightly among those with the least – driven in part by the design of a new national funding formula, and minimum funding guarantees that shifted funding from poorer to richer areas. Reforms to the English student loan system are set to see low- and middle-earning graduates repay substantially more of their loans – more than trebling expected repayments for the lowest-earning 30%, for example – while reducing them for the highest-earning graduates. And the planned expansion of free childcare for working families in England is set to benefit families in the upper-middle part of the income distribution far more than middle-income and, especially, low-income families.
Despite much noise and fanfare, there is also little sign of ‘levelling up’ in public service spending. Funding for local government, the police and public health is based on increasingly out-of-date estimates of local areas’ spending needs, and these estimates play an increasingly limited role in funding allocations.
Concluding thoughts
The most recent of these changes – to higher education and childcare spending policies – are taking place at a time when policy on taxation and cash benefits has actually moved in a somewhat more progressive direction overall. These complex trends mean it is important to look at spending on both cash benefits and public services, as well as taxation, to understand how the extent of redistribution being undertaken by the public sector as a whole is changing over time.
As our report for the IFS Deaton Review of Inequalities explains, this is a tricky task due to a range of conceptual and practical issues, which no single model or piece of empirical analysis is currently up to the task of addressing. And ultimately what matters is not how much is spent on public services but how much those receiving them value those services, and the outcomes they deliver for different groups of people. Further research on public service spending and associated outcomes would therefore be valuable, and this is something we hope to contribute to in the coming years.