Introduction
Rachel Reeves’s recent audit of public spending identified a range of in-year spending pressures facing the new government. One of the largest pressures was for costs relating to asylum and illegal migration, with an estimated £6.4 billion of pressures on day-to-day spending in 2024–25. A key plank of the new Chancellor’s argument was that these costs were genuinely unexpected and could not be met from within the departmental budgets she inherited from her predecessor. In response, Jeremy Hunt argued that civil servants had signed off the Home Office budget (its ‘Main Estimates’ in the jargon) just weeks earlier and suggested that these Main Estimates contradicted the figures presented by Ms Reeves.
In this comment, we explain this seeming contradiction and document the deeply unsatisfactory way that the Home Office and HM Treasury have budgeted for asylum costs in recent years.
The crux is that between 2021–22 and 2023–24, the Home Office planned at the start of each year to spend an average of £110 million on asylum, border, visa and passport operations (the relevant spending category). But spending turned out much higher: an average of £2.6 billion per year. It was clear that the amounts being budgeted were insufficient – the Home Office acknowledged as much. But every year, the Home Office would submit unrealistically low Main Estimates, and the Treasury would top them up later in the year (at the ‘Supplementary Estimates’) from the Reserve (a contingency fund supposedly set aside for ‘unforeseen, unaffordable and unavoidable’ spending items).
The problem for the new government is that it inherited a position where asylum costs this year appear to be on a similar scale to those seen in previous years, but the Reserve has already been spent elsewhere. This was to some degree predictable but is no less of a fiscal headache for that. We conclude with some thoughts on what this means for the spending framework.
How has spending differed from plans in recent years?
Figure 1 shows how much the Home Office planned to spend on asylum, border, visa and passport operations at the start of the financial year (its ‘Main Estimates’) and the final budget it was allocated towards the end of the financial year (its ‘Supplementary Estimates’) for each of the last six years. This approximates the amount the Home Office was budgeting at the start of the year, and the amount it ended up spending at the end of the year. The Home Office raises money from visa and passport fees, and so this shows net rather than gross spending (i.e. the net amount spent after taking account of income from fees) – in other words, the amount of funding the Home Office needs from the Treasury. Last year and this year, this amount was estimated to be negative (i.e. the Home Office expected its income from fees to exceed its spending).
Figure 1. Change between planned and actual Home Office day-to-day net spending on asylum, border, visa and passport operations
Note: Planned spending is from each year’s Home Office Main Estimates. Actual spending is from each year’s Home Office Supplementary Estimates up to 2023–24. For 2024–25, actual spending is not yet confirmed. As an estimate, we use the planned spending plus the estimate of pressures for ‘asylum and illegal migration’ from HM Treasury’s spending audit, removing the £1.5 billion Reserve claim for asylum spending we understand to be included in both the Main Estimates and the £6.4 billion ‘spending pressure’, and removing the £800 million the new government expects to save from ending the Rwanda migration partnership and removing retrospection of the Illegal Migration Act. That is, minus £0.2 billion (the starting surplus) plus £6.4 billion (the gross spending pressure) minus £1.5 billion (already in the Main Estimates) minus £0.8 billion (of savings from scrapping the Rwanda scheme) gives net spending of £3.9 billion.
Source: Home Office, Supplementary Estimates memorandum (various), https://committees.parliament.uk/committee/83/home-affairs-committee/publications/10/estimate-memoranda/; Home Office, Main Estimates memorandum 2024 to 2025, https://www.gov.uk/government/publications/home-office-main-estimates-memorandum-2024-to-2025; HM Treasury, ‘Fixing the foundations: public spending audit 2024-25’, https://www.gov.uk/government/publications/fixing-the-foundations-public-spending-audit-2024-25.
In 2019–20, the Home Office planned to spend £430 million on asylum, border, visa and passport operations (according to its budget submitted to parliament at the start of the year), but ultimately spent £230 million, £200 million less than planned. From 2020–21 onwards, this reversed, and the Home Office spent much more than originally planned each year. In both 2020–21 and 2021–22, the Home Office spent £1.5 billion more than planned (that is, its budget was topped up by £1.5 billion at the Supplementary Estimates). In 2022–23, it spent £1.9 billion more than planned. In 2023–24, it spent £4.2 billion more than planned, with a final budget of £4.0 billion, having originally planned to make a surplus of £230 million.
Taking the three years between 2021–22 and 2023–24 together (ignoring 2020–21, which was subject mostly to COVID-specific spending pressures), the Home Office planned to spend a net total of £320 million on asylum, border, visa and passport operations. In the end, it spent far more than budgeted – a total of £7.9 billion, i.e. £7.6 billion more than budgeted.
The major reason for the higher-than-planned spending by the Home Office in recent years is asylum costs. Higher spending partly reflects the costs of a number of special visa, humanitarian protection and resettlement schemes, most notably for Afghanistan and Ukraine. But it also reflects a broader upwards trend in both the number of people seeking asylum and the costs per asylum seeker, a trend that was not obvious at the time of the last Spending Review in October 2021. In 2023, there were 67,337 asylum applications (for 84,425 people). These numbers are slightly lower than in 2022, but levels remain much higher than in 2019, when there were 35,737 applications (for 45,537 people). Importantly, the number of asylum seekers in ‘contingency, initial and other types of accommodation’ (a category that includes hotels) was 16 times higher at the end of March 2024 than at the end of March 2020 – a major contributor to higher costs (and to inefficient management of the government’s aid budget).
The key point is that asylum costs are much higher now than pre-pandemic and have been elevated for several years, and so higher spending is entirely foreseeable (if difficult to predict with precision) – but this has not been reflected in the Main Estimates presented by the Home Office each year.
We do not yet know what the final amount spent on asylum, border, visa and passport operations in 2024–25 will be. But we can be near certain that it will be more than indicated in the Home Office’s Main Estimates submitted to parliament. The plan, similar to the previous year, is to achieve a £240 million surplus (though it is not possible to tell from the Main Estimates how much the Home Office is budgeting for asylum costs specifically, and how this compares with expected income). We understand these Main Estimates to include £1.5 billion of additional funding already provided from the Reserve (implying a much larger surplus in the absence of this top-up). Even so, the Home Office warned that ‘this is not sufficient to support those currently in the asylum system’ and that ‘we … will be requesting more [funding] from the Reserve through the Supplementary Estimates process’.
In Ms Reeves’s public spending audit, published last month, the Treasury warned that additional spending pressures for asylum and illegal migration amounted to £6.4 billion this year. We understand this figure to also include the £1.5 billion top-up mentioned above. The new government plans to make an immediate saving of £800 million in 2024–25 from ending the Rwanda migration deal and removing the retrospection of the Illegal Migration Act. That leaves a ‘net’ outstanding asylum pressure of £4.1 billion (£6.4 billion less £1.5 billion less £800 million) for the remainder of the year. Combined with the planned surplus of £240 million asylum, border, visa and passport operations, this points to eventual spending in this category of a little less than £4 billion in 2024–25 (as shown in Figure 1), with considerable uncertainty around this number.
Bringing all of this together, two things are true at once. One, the in-year spending pressures on asylum spending are high. Two, this is at odds with the amount budgeted for in the Home Office Main Estimates, which is insufficient and acknowledged as such. In other words, both Ms Reeves and Mr Hunt in some ways have a point. Ms Reeves is right to say that there is a genuinely big in-year asylum spending pressure that has not been budgeted for, though she perhaps overstates the extent to which this was unforeseeable and unexpected. Jeremy Hunt is right to say that this seemingly contradicts the Main Estimates put before parliament – but only because the Home Office and HM Treasury are continuing the poor budgeting practice of recent years, including the period when Mr Hunt was Chancellor. We turn next to an examination of this poor practice and potential solutions.
Why is this a problem? What are potential solutions?
This is clearly not a sensible way to plan public spending. When there is a one-off and unexpected spike in costs or demand, an overspend is entirely understandable and defensible. But if this overspend is occurring year after year, for the same reason, it starts to look like a bad habit. It also appears to be at odds with Treasury guidance, which states that departments must ensure ‘that Estimates [plans] are consistent with their best forecasts of requirements’. That certainly does not appear to be the case here, given that the Home Office has itself acknowledged that its budget for 2024–25 will be insufficient. The Home Affairs Select Committee has repeatedly flagged this issue.
Rather than adequately budget for asylum costs ahead of time, the previous government got into the habit of funding asylum costs out of the Reserve each year (to the tune of £4.0 billion in 2023–24). In doing so, it has avoided the need to top up overall spending plans and the baseline for future years. This itself raises questions. The Reserve can be loosely thought of as a contingency fund set aside for emergencies and unexpected spending items. Specifically, the Treasury guidance states that the Reserve should be used for ‘genuinely unforeseen contingencies that departments cannot absorb within their [budgets]’, as well as ‘certain special cases of expenditure that would otherwise be difficult to manage, as agreed with the Chief Secretary’.
Does spending on asylum costs fall into either of these two categories? One could argue that the scale of the spike in asylum seeker numbers was unforeseeable at the time of the October 2021 Spending Review, when the Home Office budget was set for 2022–23, 2023–24 and 2024–25. There is some truth to this, but asylum costs were also elevated in 2020–21 and 2021–22, and so higher spending in subsequent years ought not to have been a total surprise. And given that the Home Office can see this coming at the Main Estimates each year, it is difficult to see how spending on asylum costs classifies as ‘genuinely unforeseen’.
It is clearly true that asylum spending is unpredictable and difficult to manage, given that it depends on the number of asylum claimants, which itself depends on unpredictable global events. So it may fall into the second category. But in that case, given its repeat nature, a better option might be to change how it is classified, rather than rely on the Reserve for such a large spending item, and to leave the Reserve for genuinely unforeseen emergencies.
In the Treasury spending framework, asylum spending currently sits within the departmental expenditure limits (DEL), the spending limits that departments have for many types of routine spending which they cannot (in theory) exceed. But types of spending that are ‘demand-led but also exceptionally volatile in a way that could not be controlled by the department’ and ‘so large that departments could not be expected to absorb the effects of volatility in their DELs’, such as spending on working-age benefits or debt interest, are instead classified as annually managed expenditure (AME). To an extent, this description would seem to apply to asylum costs, and one option would therefore be to move asylum spending from DEL to AME. But while the number of asylum seekers in undoubtedly volatile and difficult to predict, one issue in recent years has been the failure to control costs per asylum seeker (the most obvious manifestation being the growth in the bill for hotel accommodation). The risk, from a Treasury perspective, might be that switching asylum spending to AME would (further) reduce the Home Office’s incentives to control and keep a lid on unit costs within the asylum system.
Conclusion
In recent years, the Home Office has repeatedly spent far more than it had budgeted for asylum, border, visa and passport operations. The precise scale of the pressures each year is difficult to predict – we never know exactly how many asylum seekers will arrive in a given year, for example – and the spending pressures this year do appear to be even greater than in the past. But the problem here is not that the spending pressures were entirely unexpected (they were not). It is that they have not been budgeted for.
The Home Office has got into the bad habit of each year submitting Main Estimates to parliament that it knew would be insufficient, relying on a top-up from the Treasury Reserve later in the year. Despite repeated reprimands from the Home Affairs Committee, and the fact that this behaviour seemingly contravenes Treasury guidance, it has been the practice of the past few years. The plan was to repeat this again in 2024–25. When Rachel Reeves pointed to a £6.4 billion in-year spending pressure on the asylum system, she was in effect pre-recognising the top-up that the Home Office was banking on receiving later in the year. The problem is that this time around, with major spending pressures elsewhere, there does not seem to be enough left in the Reserve for such a top-up.
This is not a desirable state of affairs. There is a strong case that the current budgeting process is failing for asylum spending. One obvious solution for the upcoming Spending Review would be to set higher initial budgets, based on the experience of previous years, to recognise up front the likely amount of asylum spending. If the Home Office is budgeting a smaller amount than in the previous year, this should come with a plausible explanation for why demand and/or spending would fall so quickly. The Reserve should be left for genuine emergencies, not spending that the government expects to do but does not want to budget for. An alternative solution would be to move asylum spending into AME, a category of spending designed precisely for things that are hard to budget for and control. That might reduce incentives to keep spending under control, but this risk needs to be weighed against the poor outcomes delivered by the existing process.