How should we interpret parties’ public spending pledges this election?

Introduction

Over the last two weeks, major political parties in the UK have published their manifestos for the upcoming general election. These manifestos have featured a range of ‘fully costed’ spending policies for public services, such as hiring more teachers or delivering more appointments in the NHS. 

But it is essentially impossible to judge exactly what the additional spending presented in the manifestos means for the actual funding individual public services might receive. This is because we do not have detailed plans for departmental spending after this year, and so cannot tell what parties’ ‘additional’ funding commitments are relative to. We also cannot tell what these additional commitments mean for spending on other services where no plans are included in the manifesto. That makes it very challenging to interpret the plausibility of broader commitments to improve public services.

In this comment, we outline the challenges of interpreting the spending pledges recently made by political parties given the lack of department-by-department spending baselines (i.e. pre-existing spending plans). We first discuss why there are no departmental spending baselines available for parties to use. Next, we consider why this lack of baselines makes it hard to interpret pledges for spending on public services in manifestos. We go through three examples of spending – on health, defence and ‘unprotected’ departments – each illustrating a separate challenge in interpreting spending commitments made in manifestos. This election, ‘additional’ spending commitments provide no information about the actual funding public services might receive. But that does not have to be the case for future elections. We conclude by briefly discussing potential solutions to make public spending pledges both easier to interpret and more meaningful for public services at future elections.

What we know and what we don’t

The current government has set out plans for how total spending on public services will evolve over the next few years to enable the Office for Budget Responsibility (OBR) to forecast borrowing and debt and to evaluate performance against the government’s fiscal targets. Plans for two spending totals (or ‘envelopes’) – for total day-to-day public service spending and for total investment spending – are pencilled in after this year: total day-to-day spending on public services is set to grow by 1% in real terms each year between 2024–25 and 2028–29, and total investment spending on public services is set to be held flat in cash terms (implying cuts both in real terms and as a share of national income).¹  These spending envelopes act as baselines for total public service spending over the next few years, so are in a sense the ‘default’. 

While we know how overall public service spending is set to evolve over the next few years, we do not have similar baselines for department-by-department spending. Baselines for government departments are set at Spending Reviews, events at which total spending envelopes are allocated between departments, typically over a multi-year period.²  This means each department is given limits for its planned day-to-day and investment spending in each year – baselines – and knows how these limits will change between years. The last Spending Review was in 2021 and covered the three years 2022–23, 2023–24 and 2024–25. We therefore do not have totals for each department beyond March 2025. 

The next Spending Review will be undertaken by the new government – likely in the autumn – to allocate total spending from April 2025 between different departments. As discussed in more detail later, this lack of department-by-department detail means that the government can pencil in tight spending envelopes without actually laying out in detail where any cuts might fall. 

So while promises for public spending in manifestos come relative to known spending envelopes for total public service spending – meaning we can make statements about how parties have promised to change overall spending levels and the size of the state – we have no department-by-department plans against which more specific promises for individual public services can be compared.

What does this mean for manifestos?

During this election campaign, parties have mostly refrained from telling us about the broader paths of departmental spending over the next parliament in their manifestos. There have been some exceptions: the Conservative manifesto commits to increasing ‘NATO-qualifying defence spending’ to 2.5% of GDP by 2030. Additionally, the Conservative manifesto commits to increasing NHS spending at least in line with inflation, and to increasing per-pupil schools funding in England at least in line with inflation. Other parties have made similar commitments for some areas of spending, primarily defence and international aid.³

But in most cases, instead of committing to broader departmental spending plans, manifestos have set out specific commitments for particular public services. Many of these commitments appear to have been precisely costed. For example, the Conservative manifesto committed to delivering 250 new or modernised GP surgeries, at a cost of £183 million in 2028–29, while the Labour manifesto committed to free breakfast clubs in every primary school, at a cost of £315 million in 2028–29. 

Without pre-existing departmental baselines, it is impossible to interpret what these ‘additional’ spending commitments mean because we do not know what they are relative to. Parties’ commitments to ‘additional’ spending are therefore of limited use. There are some exceptions: they tell us about changes to overall spending envelopes, where we do have baselines, and they provide some information about the broad priorities of each party. But they do not tell us anything about overall spending on each public service. This matters for three reasons.

First, without overall spending numbers, we cannot evaluate whether the – often ambitious – commitments for public service provision are at all plausible. Second, if different parties use different baselines when calculating ‘additional’ spending, the same policy could be presented as costing different amounts by different parties. Equivalently, the same ‘additional’ spending by different parties could ultimately have very different impacts on total spending. Third, this allows parties to make fiscal plans that are almost certain to imply real-terms cuts to some ‘unprotected’ public services without being transparent about this. This is the basis of our repeated warnings about a ‘conspiracy of silence’ by major political parties. 

We now go on to discuss each of these issues in more detail.

An example: Labour and Conservative health spending commitments 

To illustrate the challenge of interpreting spending pledges without baselines, we consider in more detail the health spending commitments made in the Labour and Conservative manifestos. The Labour manifesto set out £1.8 billion (cash-terms) additional NHS spending in 2028–29, while the Conservative manifesto set out £1.0 billion (cash-terms) net additional NHS spending in 2028–29.⁴  But the ‘additional’ spending set out by both parties is small in the context of the overall health and social care budget: the size of Labour additions is just 0.9% of the current health and social care budget (£192 billion), and the size of Conservative additions 0.5%. 

As discussed earlier, we cannot say what these additional spending figures mean for the overall health and social care budget in 2028–29 because we do not have baseline spending numbers. Figure 1 illustrates a (non-exhaustive) range of potential baselines for total health and social care departmental spending. In each case, the figure also shows what the Conservative commitments and Labour commitments would mean for the overall health and social care budget in 2028–29. The different baselines make a vast difference to the eventual level of health and social care spending. In comparison, the details of promised additions make very little difference.

We start with two baselines that hold spending constant in different ways. The ‘flat cash’ line shows the baseline that spending remains constant in cash terms, implying (after accounting for inflation) a real-terms fall prior to any additional spending. This is very unlikely to happen in practice but is one potential baseline for judging ‘additional’ spending against. Allowing the health and social care budget to fall in real terms was explicitly ruled out by the Conservative party, but not by the Labour party. Under this baseline, the additional spending announced by Labour would not be enough to maintain spending in real terms. The ‘flat real’ line instead shows the baseline that spending is held constant in real terms. Under this baseline, the manifesto commitments would imply that health and social care spending reaches £193–194 billion in 2028–29 (in 2024–25 prices throughout).

The ‘0.8% real’ line instead shows the baseline that spending grows in line with the overall departmental spending envelope, i.e. by 0.8% per year in real terms (accounting for a 1% real growth rate in day-to-day spending and flat nominal capital budgets). This is the baseline used by the Nuffield Trust in its analysis. The manifesto commitments would, using this baseline, imply health and social care spending reaches £200 billion in 2028–29. In practice, this is also a highly unlikely scenario: since 1986, UK health spending has always grown at a faster rate than average public spending over the same horizon.⁵

The ‘3.6% real’ line shows the baseline that spending grows by 3.6% per year, in line with our central estimate of the cost of delivering the NHS workforce plan (to which both the Conservative and Labour parties are committed).⁶  This is what we have previously assumed when calculating the potential cuts for ‘unprotected’ departments, which we discuss in detail below. In this case, the manifesto commitments would imply that health and social care spending reaches £222–223 billion in 2028–29. Finally, the ‘4.4% real’ line takes our upper estimate of the cost of delivering the NHS workforce plan as the baseline, assuming that spending grows by 4.4% per year in real terms. The manifesto commitments in this case would imply that health and social care spending reaches £229–230 billion in 2028–29.

The point here is not that any of these spending outcomes is necessarily going to happen, or even that they are plausible. Either party holding the health budget constant – either in cash terms or in real terms – is incredibly unlikely, and even growth only in line with average public spending is far from probable. The point instead is to illustrate the extent to which the additional health spending promised in the Labour and Conservative manifestos tells us essentially nothing about the path of health spending in the coming years. The gap between promised additional spending in the Conservative and Labour manifestos is around £1 billion per year, while the gap between different baselines can be upwards of £40 billion per year. And without overall spending paths, it is not possible to evaluate whether ambitious party commitments to improve NHS performance, such as commitments to cut waiting times, are plausible.

An example: the Conservative defence spending commitment

The lack of a department-by-department spending baseline matters even when manifestos set a commitment for total spending on a department or service. This is because without a baseline, we still do not know what these new spending totals are relative to, and therefore how much ‘extra’ funding is being promised. 

This problem is well illustrated by the Conservative commitment to get ‘NATO-qualifying’ defence spending to 2.5% of GDP by 2030. This commitment gives us a promised path and level of defence spending – shown by the blue line in Figure 2 – as set out in the policy document launching the commitment. 

But the amount that this pledge ‘costs’ depends, fundamentally, on what you assume defence spending would do in the absence of the pledge, i.e. the baseline. Figure 2 sets out a (non-exhaustive) range of baselines: that defence spending stays constant in cash terms, that defence spending stays constant in real terms, and that defence spending stays constant as a share of GDP. The additional spending required to achieve the 2.5% commitment under each of these baselines differs substantially.

When launching its top-up to defence spending in April 2024, the government claimed it represented a £75 billion increase over six years (in cash terms). This figure was calculated by assuming that defence spending would, without this pledge, have stayed frozen in cash terms (i.e. the gap between the ‘flat cash’ line and the blue line in Figure 2), and adding together the additional money represented by the new plans each year.

But in the Conservative manifesto, the same policy was costed using the baseline that defence spending would otherwise have stayed at 2.3% of GDP (i.e. the gap between the ‘flat at 2.3% of GDP’ line and the blue line in Figure 2). This assumed baseline gives a much lower cumulative cost of the policy: £24 billion over six years, a third of the £75 billion required if assuming spending would otherwise have been constant in cash terms. A lower baseline is useful in this case, meaning that the manifesto has less to find in order to fund this ‘extra’ defence money.

As with health spending, other baselines are plausible. Comparing the Conservative plans and the ‘flat real’ line in Figure 2 shows the savings if defence spending would otherwise have been held constant in real terms, which fall somewhere in between the savings from the two baselines described above. In cumulative terms, additional costs implied by the three sets of baselines are tens of billion pounds apart.

This illustrates the broader problem: costings for ‘additional’ spending in manifestos mean little without a baseline to which they are additional. Currently, it is far too easy for parties to choose the baseline that serves them best. We have used the Conservatives’ defence pledge here as an example, since they are the only major party to have announced and costed a path of planned spending over the next parliament, but the broader problem is the same across parties.

An example: cuts to ‘unprotected’ areas

As discussed above, at the time of the March 2024 Budget, the baseline day-to-day resource spending envelope for all government departments was growing at 1% in real terms per year after this year. Neither main party has changed overall resource spending plans in significant ways with their manifestos: Labour’s £5 billion top-up in 2028–29 means real-terms resource spending will now grow at 1.2%, rather than 1%, on average per year. The Conservatives left total spending plans virtually unchanged, topping up total departmental spending in 2029–30 by around £500 million.

We have already discussed the fact that the lack of department-by-department plans after this year means that we are uncertain about the path of spending on particular public services, and that we are unable to evaluate the ‘cost’ of committing to a given path of spending. But the lack of department-by-department plans also means that parties can commit in their manifestos to overall spending plans that imply sharp real-terms cuts to a range of areas, without spelling out where those cuts will fall or how they are to be achieved.

The real growth rate pencilled in for day-to-day spending – whether 1% or 1.2% –- will not be equally allocated between departments. Certain areas of spending will almost certainly be ‘protected’, in line with commitments made by parties. Previous work at IFS has made assumptions on the future path of spending by department, based on government commitments. These assumptions are that the English NHS budget grows by 3.6% in real terms each year, consistent with meeting the NHS workforce plan, defence and aid spending grow with GDP in line with our international obligations, English schools funding is kept flat in real terms, and the childcare commitments made in last year’s Budget are funded. 

These areas represented around two-thirds of total (non-devolved) day-to-day spending this year. If they are protected in this way, then all other areas of spending – including areas such as courts, prisons and further education – would face cuts of between 1.9% and 3.5% each year, according to ‘baseline’ pre-manifesto spending totals. These cuts would require a top-up to spending in 2028–29 of £10–20 billion to be avoided. This is of course uncertain, being based on broad assumptions over spending on different departments: the problem is that because there are no baselines, we do not know where, or exactly how, these cuts will fall.

The two main parties’ manifestos did not top up spending sufficiently to avoid these cuts. The Conservatives’ commitment to higher defence spending without overall top-ups in fact means unprotected areas are squeezed by more. Labour’s £5 billion top-up to day-to-day spending might ease some pressures, but is still not on the right scale – particularly since the money came with new ambitions for areas such as schools, rather than flowing to unprotected areas. Because we have no baselines, we do not know where these cuts will fall, or how they will be implemented in a context where many services are in difficult positions already. The lack of baselines therefore allows parties to remain silent on the inevitability of cuts, instead blithely promising improvements across essentially all areas of public services.

What could be done in future?

The lack of department-by-department spending plans after this year leaves parties in a difficult position. It is not immediately clear what the best response within manifestos would have been: we are not proposing that parties lay out firm and fixed spending plans for every department for every year of a potential five-year parliament in a manifesto document. The best place for this is a Spending Review. But a situation where parties barely touch the topic of planned paths of spending – particularly, as now, when spending totals imply large cuts to some areas – could have been mitigated. 

More idea of parties’ priorities, and rough minimums or totals – subject to change – for different areas of spending, could help give a sense of what we can realistically expect from them in the next parliament. This is not a radical proposal – as discussed earlier, some parties are already doing this for some public services this election, such as schools and the NHS. But these current pledges are for departments that are typically ‘protected’. The failure in manifestos to grapple with implied cuts to other public services could, and should, have been avoided, and parties should have been clear about where the almost-inevitable cuts would fall in the absence of additional money – even if only in broad terms. This would not necessarily mean putting precise numbers on how spending in each department would change, but giving a sense of where or how the scope of the state might be reduced given no top-ups to overall spending.

Of course, there are serious political challenges here, and very few incentives to engage with the implied plans and cuts. Beyond thinking about what parties should – and plausibly would – do, there are broader institutional choices that could be made to avoid this kind of situation.

The Institute for Government (IfG) has suggested that Spending Reviews could cover a longer period, and be reviewed every three years, ensuring that department-by-department spending plans always exist for at least the next two years. This would make it harder for governments to pencil in unrealistically low spending totals in order to meet fiscal rules, by forcing them to confront cuts that would be implied for at least the next two years explicitly. It would also mean there were always department-by-department baselines for at least two years in the future when manifestos were being produced, avoiding the no-baselines problem.

The IfG has also recommended that the OBR tackle the implications of spending plans in the future in more detail. The OBR already does this to an extent, through scenarios analysis similar to ours described above. It could also set out implied cuts for each individual unprotected department, giving the Treasury an opportunity to change its assumptions on a department-by-department basis. The OBR could even go further, laying out whether plans would be consistent with services’ current level of performance, as it does in its analysis of the longer-run sustainability of public finances. At present, there is a rather odd change in approach taken by the OBR for spending five years out and spending six years out, which is entirely driven by the fact that the former sits inside the standard five-year forecasting horizon whereas the latter does not. 

The point here is not to set out a definitive view on these proposals, but rather to say that there are options on the table to avoid running into the series of problems we have laid out in this comment. The current situation of manifestos setting out ‘additional’ spending that is almost meaningless in terms of the actual level of spending on services – and accepting spending totals that almost certainly imply real-terms cuts while remaining silent as to how these are to be implemented – is undesirable but not inevitable.