Written evidence submitted by The Northern Ireland Fiscal Council, relating to the funding and delivery of public services in Northern Ireland inquiry (FPC0028)

Foreword

The Northern Ireland Fiscal Council was established in 2021. It is chaired by Sir Robert Chote and the other members are Maureen O’Reilly, Professor Alan Barrett and Dr Esmond Birnie. Our aim is to bring greater transparency and independent scrutiny to the region’s public finances, focusing on the finances of the NI Executive. In doing so we hope to inform both public debate and policy decisions to the benefit of everyone in NI.

On 14 February 2023 the NI Affairs Committee (NIAC) of the UK House of Commons launched an inquiry into the funding and delivery of public services in NI.[1] According to the Committee: “The inquiry will look at the financial situation facing Northern Ireland’s public services, including health, education and the police, and how the lack of a functioning Executive has impacted on budgetary management.

The Committee will also take the opportunity to explore the effectiveness of the Barnett formula and examine the viability of other options for increasing revenue to put Northern Ireland’s finances on a more sustainable footing.”

This submission to the inquiry is structured along the four themes on which the Committee is gathering evidence:

• the impact of the lack of a functioning Executive on budgetary management and strategic decision-making across NI Departments;
• the financial situation facing NI’s public services. Given our remit we focus on the current and near-term outlook for NI departments’ spending rather than on the consequences for the quality of individual services;
• the operation of the Barnett formula in calculating the amount of money the UK Government makes available to NI for providing services; and
• options for increasing revenue which could be open to the NI Executive or UK Government to bring NI’s finances onto a more sustainable footing.

1              The lack of a functioning Executive

Even in normal times the NI Department of Finance is in a less powerful position than the UK Treasury to control departmental spending.[2] This in part reflects the greater formal autonomy of NI Departments, relative to Whitehall ones, as well as the fact that the Executive is designed to be a multi-party coalition in which Ministers do not take collective responsibility for its decisions and actions. But a strong case can be made for saying that spending control is even weaker in the absence of an Executive and Assembly, for reasons we discuss below.

The absence of a Budget

During the 2022-23 financial year, NI departments indicated that they expected to overspend relative to available funding and a £297 million call on the Treasury Reserve was eventually required to balance the books. Commenting on this situation last year, we noted that unexpectedly high inflation and upward pressure on pay were important factors, but common to the whole UK public sector. Interestingly, they did not lead to equivalent overspending by Whitehall departments, which suggested some weaknesses in financial control and management unique to NI.[3] These specific challenges of last year also took place against the background of the ongoing ‘Barnett squeeze’ on NI’s Block Grant funding (discussed below).[4]

The prospects for good financial management last year were not helped by the fact that the outgoing Executive failed to agree even a Draft Budget for 2022-23, with the Department of Finance publishing its own version purely for consultation. When the Stormont institutions collapsed in February 2022, responsibility for departmental spending decisions passed temporarily to Caretaker Ministers under the NI (Ministers, Elections and Petitions of Concern) Act 2022[5] (‘MEPOC’) acting alongside Permanent Secretaries in their capacity as departmental Accounting Officers.

The Budget, and the Budget Acts that give it statutory effect, are the key control mechanisms through which the Executive and Department of Finance allocate and constrain individual departmental spending. But the NI Act 1998[6] makes it clear that no Budget can be agreed in the absence of a functioning Executive. Compounded by uncertainty surrounding the length of time for which the institutions would be suspended, this necessitated an ad hoc approach with various drawbacks:

•                  Departments had no firm budget settlements from which to plan and manage their spending at the beginning of the year. They had only informal indicative allocations from the Department of Finance based on a Draft Budget that had failed to secure broad-based political endorsement, as well as the legislative limits on spending from the Budget Act 2022.

•                  Departments were told at the outset that their initial indicative allocation was to be treated as a minimum, and that they would probably be able to spend more, because some funding was being held centrally for distribution later in the year. (This included some funding that formally could not be distributed in the absence of an Executive.) Without central coordination, the obvious risk was that Ministers would have unrealistic and mutually inconsistent expectations of their eventual shares of this cake.

•                  Without an Executive in place, the Department of Finance could not operate the In-Year Monitoring Rounds at which departmental allocations are normally revised to reflect changes in the aggregate funding envelope and the emergence of unexpected pressures or reduced demands from individual departments. Under normal circumstances, these Rounds make the reallocation of spending within-year more transparent than at the UK level.

•                  When the Caretaker Ministers left office on 28 October 2022, responsibility for spending decisions passed to the Secretary of State (SoS) for NI and the Northern Ireland Office (NIO). Working with the NI Civil Service, the NIO reduced the prospective overspend but in practice it is hard for the SoS to make significant (and therefore typically contentious) money-saving decisions knowing they lack political ownership in NI.

Autonomy of Caretaker Ministers

Financial management without an Executive was always going to be something of an experiment last year, given the decision not to return to direct rule from Westminster (as between 2002 and 2007) or to constrained decision-making by the NI Civil Service (as between 2017 and 2020). As noted above, the MEPOC left Caretaker Ministers in charge of departments for 24 weeks, working alongside Permanent Secretaries in their capacity as departmental Accounting Officers.

Managing Public Money Northern Ireland (MPMNI)[7] – which “sets out the main principles for dealing with resources used by public sector organisations” – requires Accounting Officers to ensure that departmental spending (and other) decisions are not irregular, improper or offer poor value for money. It says that Accounting Officers should bring any conflict or concerns that arise between their Minister’s instructions and their own duties to the attention of their Minister.[8]

But ultimately Ministers are in charge (at least as regards decisions that do not require a functioning Executive to be in place). If, despite the advice of his or her Accounting Officer, a Minister decides to continue with a course of action then he or she can issue a formal Ministerial Direction instructing them to proceed. (That said, Accounting Officers – not just in NI – sometimes seek Directions for decisions that they themselves regard as sensible and appropriate but want formal cover for.)

By the time the Caretaker Ministers left office, Ministerial Directions had been issued for spending decisions worth £345.4 million, the vast majority by the Minister of Health – who presides over by far the biggest spending department. This

The lack of a functioning Executive

was equivalent to around half the projected overspend against available funding at the time. Normally approval for Directions is sought from the Executive or the Finance Minister, but last year these went to the Finance Minister alone and he had approved only £22.5 million by the time he and the other Ministers left office.

Although the decisions made by Caretaker Ministers were reflected in the projected levels of overspend by departments, it is notable that the largest single additional cost pressure, the NHS pay uplift, was only approved after the Caretaker Minister left office and follow the setting of a budget by the Secretary of State and the setting of a public sector pay policy for 2022-23 by the Department of Finance.

Lack of Assembly scrutiny

One of the consequences of the parties not electing a Speaker to the Assembly is that Assembly Committees could not be formed. The Department of Finance normally presents forecasts of departmental spending (and thus any under- or overspend against available funding) to the Finance Committee on a monthly basis and these are published by the NI Assembly on its website. Ministerial Directions are also normally brought to the attention of (and sometimes discussed by) the Public Accounts Committee before being published. Individual departmental committees also regularly review the financial position for each department. The absence of these scrutiny channels may at the margin help explain how the projected overspend escalated to the surprising levels that it did last year. Alternatively, the Committees might have challenged potential savings and made things worse.

The current position

Moving into the new financial year, the Secretary of State for now remains in charge of departmental spending allocations and the NI Civil Service has been given some limited additional room for maneuver in decision-making until 6 June 2023.[9] But several of the weaknesses in spending control and management remain – an absence of ministerial leadership, no agreed Budget to work from at the start of the year and no scope for In-Year Monitoring Rounds or Assembly scrutiny. As we shall see below, this is combined with a particularly challenging financial position as the call on the Treasury Reserve necessary to balance the books last year has to be repaid.

It should be noted that the failure once again to pass a Budget ahead of the new financial year – either at Stormont or at Westminster – does not mean that the NI public sector is going to grind to a halt. The ‘Vote on Account’ included in the NI Budget Act 2023 passed at Westminster provides NI departments with legal authority to spend up to 65 per cent of their 2022-23 allocations this year. This is higher than the normal 45 per cent, giving more time for the Stormont institutions to be restored before additional legislative authority is required. And even if this does not happen, emergency powers can be invoked as they were last year.

2              The financial situation for public services

The NI Budget[10] has three main components that are required to ‘balance’, so that planned spending does not exceed available financing:

•                  Resource spending (on public services, administration and debt interest), financed from the Block Grant plus revenue from the Regional Rates;

•                  Conventional capital spending (e.g., on infrastructure), financed from the Block Grant plus borrowing under the Reinvestment and Reform Initiative (RRI); and

•                  Financial Transactions Capital spending (on loans to or equity injections in private entities, including universities), financed from the Block Grant.

Resource spending is by far the largest component and the most directly related to public services as they are consumed. But capital spending contributes to the quality and quantity of services that can be delivered over the longer term.

Resource spending

Under the NIO Budget finalised towards the end of the last financial year (following the Westminster Spring Supplementary Estimates), NI departments had scope to spend £14,345 million in 2022-23 (around £18,700 per household) on public services and administration. This is 1 per cent less than in 2021-22 but 6 per cent (£819 million) more than anticipated in the unendorsed 2021 Draft Budget.

Most of that increase reflects additional Block Grant income, largely arising from increases in planned UK Government spending outside NI via the operation of the Barnett formula (see below). But it also includes the £297 million claim on the Treasury Reserve requested by the Secretary of State to balance the books – in effect a loan that will have to be repaid from the Block Grant next year.

Table 2.1 shows the how the total Resource Departmental Expenditure Limit (DEL) was distributed across departments, as well as the proportion of the prospective overspend that departments were forecasting at the point when Caretaker Ministers stepped down which was subsequently funded in the NIO Budget. It shows that Health is by far the largest spending department in resource terms and also that it had the highest proportion of its overspend covered.[11] This reflects the fact that prioritising Health was one of the few things that the outgoing Executive agreed on when trying to draw up a Draft Budget – and that Health tends to be the most successful bidder for additional resources (at the UK as well as the NI level).

Table 2.2 shows that even after the call on the Treasury Reserve, most NI departments will have seen a smaller increase in real terms spending between

2019-20 and 2022-23 than their nearest Whitehall equivalents. (These comparisons are only roughly indicative, as the functions of some NI departments vary significantly from those in Whitehall, but the broad trends are informative.) Among the larger spenders, Education fares relatively well[12] in NI compared to its Whitehall counterpart while Justice and Health fare relatively badly. The uplift for Agriculture is due to former EU Funding now being provided by the Treasury.

Looking ahead, Table 2.3 summarises the resource funding position in cash terms from last year to 2024-25. When the UK Government first set out its spending plans for these years in October 2021, the Block Grant settlement implied by the Barnett formula showed funding for NI rising by 1.7 per cent this year and a further 1.5 per cent next year. Subsequent increases in UK Government spending pushed up the Block Grant in each year, but the frontloading of these increases – along with the need to repay the Treasury Reserve claim next year – have created a V-shaped funding pattern with the Block Grant set to fall by 3.4 per cent this year and then rise by 3.6 per cent next year. Taking the latest Land and Property Services forecasts for Regional Rates revenue into account, the profile for total funding looks broadly similar, with a decline of 3.1 per cent in 2023-24 followed by an increase of  3.5 per cent in 2024-25. In real terms, adjusted for inflation, this corresponds of a drop of 5.5 per cent this year followed by an increase of 1.9 per cent in 2024-25.[13]

This funding profile confronts the NIO (or a restored Executive) with a very tricky challenge in setting a Budget for this year (or for this year and next if it were to decide to do so). In order to discourage overspending, ex ante, it is understandable why the Treasury demands that any Budget overruns in one year should be clawed back in the next. But, ex post, once that overspend has happened, setting a Budget to deliver a sharp real-terms cuts in one year followed by an increase in the next makes little sense from the perspective of efficient workforce or service delivery planning. It would not be surprising to see the NIO or a restored Executive arguing for some phasing of the repayment, especially as the Executive has limited capacity to smooth the impact itself as the Treasury does not permit it to run a Reserve or to borrow for resource spending except to manage its cash flow within a given year.

If the Reserve claim was clawed back over two years rather than one there would still be a significant real terms reduction in funding of 4.5 per cent in 2023-24 followed by an additional reduction of 0.3 per cent in 2024-25. And if it was not clawed back at all there would still be a real terms reduction in funding of 3.5 per cent in 2023-24. UK Government departmental resource spending is expected to rise in real terms by 1.2 per cent between 2024-25 and 2027-28. This implies that there will be limited respite for the NI public finances in the medium term.

Capital spending

Under the NIO final Budget, conventional capital spending was expected to total

£2,010 million last year – a much smaller increase of 1 per cent since the Draft Budget than for resource spending. But comparisons with the Draft Budget and earlier years are complicated by the reclassification of R&D funding for universities from resource to capital, which is proportionately much greater for the latter.

Capital spending is largely financed from the Block Grant, but the Draft Budget also assumed £140 million of capital borrowing under the Reinvestment and Reform Initiative (RRI) and a £7 million contribution to the cross-border A5 road project from the Irish Government. The NIO Budget increased borrowing to the £200 million annual maximum agreed with the Treasury in 2002 and the A5 has not progressed sufficiently to draw on the funds from Dublin. The Department of Finance held a negative reserve of £32 million centrally to balance this budget, but it expected departments to underspend their allocations sufficiently to offset that.

Financial Transactions Capital spending

Block Grant funding for Financial Transactions Capital spending arises largely from the ‘Barnett consequentials’ of UK Government action to support the housing market. It is almost always underspent as neither the Executive nor the other Devolved Administrations find loans to or equity injections in the private sector particularly convenient ways to address their investment priorities. (Universities are classified as private sector entities, however, and FTC funding has been used to help finance the new Ulster University campus in Belfast.)

Block grant funding for last year increased from £163 million at the Draft Budget to

£304 million at the NIO Budget, of which only £84 million was allocated to departments (specifically to Economy and Communities). £220 million continued to be held centrally and on past performance will not have been spent. Only a very small proportion of any underspend can be carried forward to future years.

3              The Barnett formula

As Table 3.1 illustrates, a very large proportion of spending within the Departmental Expenditure Limit (DEL) that the Treasury sets for NI Departments is financed from the Block Grant paid by the UK Government. This comprises a core element based on historical spending and adjusted at each fiscal event to reflect changes in UK Government spending according to the Barnett formula. It also includes ‘non- Barnett additions’, among them funding for political agreements like New Decade, New Approach, City and Growth Deals and farm and fisheries support to replace funding previously received from the European Union (EU).

Beyond the Block Grant, the Executive raises additional finance from the Regional Rates and capital borrowing under the Reinvestment and Reform Initiative (RRI), both of which are separately identified in the Executive’s Budget. Departments also raise money from fees and charges and some residual EU funding (even after Brexit). But these are treated as negative spending by the recipient departments and are not shown in the Budget, which does not aid transparency. In 2020-21 fees and charges raised around £700 million[14] and EU funding £180 million.

Application and history of the formula

The Barnett formula ensures that when the UK Government increases (or cuts) funding for public services delivered by Whitehall departments in England, it also increases (or cuts) the funding it gives Devolved Administrations that deliver those services in their own region by broadly the same amount in pounds per person. This ensures that if the UK Government uses UK-wide tax revenues to finance spending only in one part of the country, other regions do not lose out.

Devised by the Treasury mandarin Sir Leo Pliatsky, the Barnett formula is an administrative convention rather than having any legal status. It was designed as a temporary measure in the late 1970s when the Labour Government was legislating (unsuccessfully) for devolution in Scotland and Wales. It was applied first in Scotland in the 1978 Public Expenditure Survey and then in NI in 1979. It came to be named after Joel (later Lord) Barnett, then Chief Secretary to the Treasury (who ironically would later describe it as “grossly unfair” and a “terrible mistake”).[15]

When the Barnett formula is used at UK Government Spending Reviews[16], the changes in each Whitehall department’s resource and capital DEL settlements (compared to their baselines) are multiplied by three numbers to determine the corresponding change in the Executive’s Block Grant:

• A ‘comparability factor’ – the percentage of the department’s services that are delivered by the Executive rather than the UK Government in NI.
• A ‘population proportion’ – NI’s population as a percentage of England’s (or England & Wales’s or Great Britain’s if that is where the department delivers its services). At the 2021 Spending Review, the data used from the Office for National Statistics gave NI a population 3.35 per cent of England’s.
• A ‘VAT abatement factor’ – increases in the NI Block Grant are multiplied by 97.5 per cent (a 2.5 per cent reduction) to reflect – imprecisely – the fact that the Executive can reclaim VAT it pays from HM Revenue and Customs, unlike Whitehall departments or the other Devolved Administrations. This is because the Executive fulfils many of the responsibilities that fall to local councils elsewhere, and councils can reclaim VAT under the VAT Act 1994.

By way of illustration, take the hypothetical example of a 10 per cent increase in spending on Education in England from its 2019-20 level of just over £74 billion. All the Department’s services are delivered by the Executive in NI, so a 10 per cent increase of £7.4 billion in English education spending would be expected to result in £243 million of additional funding for the NI Executive as shown in Figure 3.1. But the Executive is not constrained to spend the funding increase on education.

The formula applies in the same way in the other Devolved Administrations, although – as we describe below – the UK Government has agreed to give the Welsh Government a 5 per cent addition to any ‘Barnett consequentials’ it receives to slow the rate at which funding-per-head there approaches an estimate of relative need.

The Barnett formula removes much of the need for negotiation between the Devolved Administrations and the Treasury over the size of their core Block Grants, but there have been arguments over which spending the formula should apply to.

These centre around whether the UK Government is responsible for delivering services only in England (in which case it is ‘Barnettable’) or to the whole of the UK (in which case not). To take two contentious examples, spending on Crossrail and capital spending on the London Olympics were deemed Barnettable because they were judged only of benefit to people in England. But current spending on the Olympics was judged to benefit the whole UK, so was not deemed Barnettable.

The ‘Barnett squeeze’

When the Barnett formula was first introduced, public spending per head was much higher in NI, Scotland and Wales than in England. As a result, increasing funding per head for the Devolved Administrations in line with increases in equivalent UK Government spending in England in cash (rather than percentage) terms results in smaller percentage increases in the Devolved Administrations than in England and reduces their percentage premium in spending per head over England.

To illustrate this, imagine that spending per head is £100 in England and spending / Block Grant funding per head is £200 in NI, a premium of 100 per cent. If the UK Government were to increase spending per head in England by £10 to £110, under the Barnett formula it would also increase the NI Block Grant by £10 per head,[17] to £210. This means that spending rises by 10 per cent in England but only 5 per cent in NI and that the NI spending premium over England falls from 100 to 91 per cent.

The higher the initial percentage spending premium is, the bigger the gap between the percentage increase in spending in England and the percentage increase in the Devolved Administrations’ funding. We can illustrate this by returning to our earlier example of a 10 per cent increase in UK Government spending on education.

Table 3.2 shows the increases in education spending that could be financed in each of the Devolved Administrations in 2020-21 as a result of a 10 per cent increase in the budget of the UK Department of Education. The UK Government has responsibility for schools, further education colleges and universities in England, but this has been devolved in Scotland, Wales and NI. Spending per head on education thus defined was 29 per cent higher in Scotland, 16 per cent higher in NI and 6 per cent higher in Wales than in England.[18] This reflects the proportions of the population that are of school age in each, as well as differences in spending per pupil and policies towards the funding of university education. Reflecting these baselines, the 10 per cent increase in spending in England would finance a 7.8 percent increase in Scotland, 8.4 per cent in NI and 9.9 per cent in Wales (boosted by the 5 per cent uplift referred to above). The NI Executive would need to find £46 million from other sources in order to achieve the same 10 per cent rate of growth in education spending as in England.

The formula causes particular challenges when the UK Government increases spending to finance pay awards. In the first example above, imagine that pay initially accounts for half of public spending in both England and NI and that the UK Government has increased spending by £10 per head in order to finance a 20 per cent pay award for public sector workers in England (lifting the pay-bill from £50 to £60 per head). The consequent £10 increase in the Block Grant is insufficient to finance the same 20 per cent pay award in NI (which would increase the pay-bill there from £100 to £120 and therefore cost £20). The Executive would have to choose whether to match the pay award in England, in which case it will need to cut other spending or raise more revenue, or increase pay in NI by proportionately less.

A similar issue arises at times of relatively high inflation, as is currently the case. NI Departments presumably face roughly the same rates of increase in their non-pay costs as their Whitehall counterparts. But if the UK Government provides departments delivering public services in England with a budget increase sufficient to cover an x per cent increase in those costs, the Barnett consequentials will not cover the equivalent percentage increase in costs in NI, leaving a funding gap.

The Barnett formula has the paradoxical result that the faster spending per head rises in England, the faster it rises in absolute terms in NI but the faster it shrinks in relative terms compared to England. And if spending is cut in England, NI loses money but sees its percentage premium over spending in England increase. People may differ in which they think is more important. As spending in England increases more often than it falls, other things being equal the percentage spending premium in NI over England shrinks over time – the so-called ‘Barnett squeeze’.

We can see from outturn data how the Barnett formula has contributed to a reduction in spending per head in NI relative to England over the last 35 years. Chart 3.1 looks at the Treasury’s estimates of ‘identifiable’ spending, namely that which specifically benefits the residents of a given nation or region. This includes spending by the UK Government and local councils in each nation/region, as well as the Executive and Scottish and Welsh Governments. But it excludes spending on areas like foreign policy that benefit the nation/regions as part of the UK as a whole.

In the late 1980s the NI’s advantage over England in identifiable spending per head was on average 55 per cent, in part reflecting the security situation and a much higher rate of unemployment than in England This fell to 38 per cent in the 1990s, 31 per cent in the 2000s and 25 per cent in the 2010s. The change in Scotland and Wales was less dramatic, reflecting the fact that when the Barnett formula was first implemented the initial gap in spending per head with England was much smaller.

The relationship in Chart 3.1 is volatile from year to year, reflecting the fact that we are looking at total identifiable spending and not just devolved spending – the Block Grant financed 54 per cent of total identifiable spending in NI between 2010-11 and 2020-21.[19] In addition, as noted above, there are ‘non-Barnett additions’ to the Block Grant, reflecting technical adjustments and political agreements. Several packages of non-Barnett funding, or other forms of financial support, have been negotiated between the NI Executive and the UK Government over recent years, including as part of the St Andrews Agreement (2006), Hillsborough Castle Agreement (2010), Fresh Start Agreement (2015) and New Decade, New Approach (2020). These packages often involve ear-marked funding for specific purposes, such as health transformation, or tackling paramilitary activity. But the packages are typically one- off or time-limited, which means that the reduction in the spending premium is steeper when they are removed. They simply delay the point at which NI’s overall spending advantage is reduced – unless of course they are replaced by future packages that provide a succession of ‘temporary’ boosts.

Looking ahead, repeated application of the Barnett formula would be expected to extend and intensify the Barnett squeeze. Focusing on spending that has been devolved to the Executive – rather than identifiable spending more broadly – last year we produced a projection of the premium of NI Block Grant funding over equivalent UK Government over the next 50 years.[20] This is illustrated in Chart 3.2. It shows that the premium is anticipated to fall from around 40 per cent in 2018-19 to 25 per cent by the end of the current Spending Review period in 2024-25, partly reflecting the expiry of some non-Barnett addition funding, (If the premium had remained at 40 per cent, NI departments would likely have had £1.35 billion more to spend last year and an additional £1.8 billion in 2024-25.) Based on the Office for Budget Responsibility’s long-term spending projections at the UK-level, and assuming a stable relationship between DEL and AME funding, we projected that the premium would to below 20 per cent in the early 2030s and below 10 per cent in the late 2040s before dropping to just over 5 per cent by the end of the projection in the early 2070s.

The Treasury publishes a slightly different comparison of Block Grant per head of population relative to equivalent UK Government spending, which includes the impact of business rates and omits funding received subsequently, including the Reserve claim. On this basis the Treasury estimated that the NI Executive would have a 21 per cent funding premium over the 2021 Spending Review period (2022- 23 to 2024-25), which is in line with the figure that has been quoted by the Secretary of State for NI.[21] This compares with the previous Treasury estimate of a 27 per cent funding premium for the 2020 Spending Review period (2020-21). This 6 percentage point fall is double the reduction for both the Scottish (129 to 126 per cent) and Welsh Governments (123 to 120 per cent) and leaves the NI Executive funding premium now only just above that for the Welsh Government. As the Treasury estimates are based on the outcome of the Spending Review the funding premium for each year potentially changes at each subsequent fiscal event.

Source: OBR, NIFC calculations

Some alternative models

Many countries have systems of fiscal transfer between the national and sub- national government. Where the fiscal transfers aim to offset differences in revenue or in the cost of providing services they are generally termed ‘fiscal equalisation payments’.[22] The size of the transfers is typically based on estimates of needs, tax- capacity or both. If an equalisation system of this sort was applied within the UK it would likely limit the convergence of Devolved Administration funding towards per capita spending levels in England that arises from the Barnett formula.

Australia

Australia introduced a formal equalisation payment in 1933[23] to support states with lower ability to raise revenue by redistributing revenue from the Goods and Services Tax (analogous to VAT). The system is overseen by the Commonwealth Grants Commission (CGC), an independent statutory body. The CGC’s formal objective is to ensure that after the transfer, each state should be able to provide services and infrastructure at the same standard (assuming each made the same effort to raise revenue from its own sources and operated at the same level of efficiency).[24]

In setting the transfer levels, the CGC takes various material factors affecting revenues and expenditures into account. It assesses differences in revenue raising capacity including land, property and mining values. On the expenditure side it assesses differences in socio-demographic characteristics, wage pressures, population dispersion, population density and rates of population growth.[25] About two-thirds of the amounts allocated are determined by revenue-raising capacities, and one-third by factors relating to the costs of providing services (both demand for services and variations in the costs of providing them). [26]

As with the use of the Block Grant in NI and the other UK Devolved Administrations, the Australian states are largely free to determine how the transfer funding that they receive should be used. This means that the fiscal transfers do not necessarily result in the same level or mix of services or taxes in each state.

Canada

Canada has supported provinces with less revenue-raising capacity since 1957.[27] The Canadian Constitution includes “the principle of making equalization payments to ensure that provincial governments have sufficient revenues to provide reasonably comparable levels of public services at reasonably comparable levels of taxation.” [28]

The equalisation payments are financed by the Government of Canada, largely from taxes. Payments are based on the revenues a province could raise if it were to tax at the national average rate and they ensure that provinces have access to per capita revenues equal to the potential average of all ten provinces. As in Australia, the equalization payments can be spent in any way the provincial government desires.

In contrast to Australia, the formula only considers revenues and does not take account of the cost of providing services or the expenditure need of the provinces. This is also the case with equalisation systems in Spain and Switzerland.[29]

Canada uses a different formula for the funding of its three territorial governments – Yukon, Nunavut and the Northwest Territories.[30] These are large geographic areas with small populations and their governments have more limited powers delegated from the federal government rather than enshrined in the constitution. The transfer payments to the territories reflect the relative cost of providing public services in addition to their ability to raise revenue (excluding from natural resources).

Germany

Germany has one of the highest levels of fiscal decentralisation in the EU. The Federal Republic of Germany comprises the Federation and 16 federal states, called Länder. Local authorities, called municipalities, also form part of the Länder.

In 2018, 50 per cent of government expenditure was managed by the Länder, including 19 per cent managed by the municipalities. The aim of Germany’s equalisation mechanism is to guarantee each Land the means to cover necessary expenditures and “to ensure equivalent living conditions”.[31] In summary, the German equalisation model involves a number of steps:[32]

• A maximum of 25 per cent of the Länders’ share of VAT goes to those with below average revenue from income tax, corporation tax and Länder tax. This partially closes the gap between the tax revenue of the fiscally weak

Länder and the average. The remainder of the Länder share of VAT, at least 75 per cent, is distributed according to the number of inhabitants among all Länder. The distribution of VAT therefore has a partially equalising effect.

• A system of financial equalisation among the Länder further reduces the differential in financial resources between them. A Land with a revenue- raising capacity per capita equal to 70, 90, 100, 110 or 120 per cent of the

average before equalisation would have revenue equal to 91, 96, 100, 104 or 107 per cent of the average respectively after equalisation.

• Disadvantaged Länder also receive supplementary federal grants. Any that still have less than 99.5 per cent of average financial capacity per inhabitant after equalisation will receive a grant making up 77.5 per cent of the remaining shortfall. A Land initially having 70 or 90 per cent of the average ends up with 97.5 or 98.5 per cent of the average per capita financial capacity after equalisation and general supplementary federal grants.
• Within the Länder, a separate equalisation mechanism redistributes resources between municipalities. As between federation and Länder, equalisation within the Länder is not only vertical but also horizontal, with wealthier municipalities having to contribute. [33]

Local government / health authorities in England

Another redistribution mechanism closer to home is how funding is distributed between different local government areas and health authorities in England. Under these systems, resource allocations are determined by:[34]

         population;

• additional needs (for example, from deprivation);
• relative access to their own resources (such as the capacity of local authorities to raise income from council tax);
• differences in costs of provision (for example, rurality or labour costs);
• ensuring authorities can provide services to the same standard (for the health service this is known as the 'health inequalities' formula); and
• a damping mechanism to avoid significant year-to-year changes.

For local authority funding, the relative resources factor is a negative figure intended to take account of the fact that some areas can raise more income locally and so require less support from Government to provide services.[35]

These systems include adjustments for relative needs, relative capacity to raise revenues locally and damping to prevent large single-year variations in grant payments. None of these features are present in the Barnett formula.

Relative need for spending per head

As we have seen, the Barnett Formula gradually erodes the historic funding premium received by the Devolved Administrations over equivalent UK Government spending but takes no explicit account of the relative cost of delivering public services. But the UK Government has long accepted that there should be some premium. As the Treasury’s Statement of Funding Policy notes: “Spending per head

in Scotland, Wales and NI is higher than the UK average, broadly reflecting that the costs of providing public services are also higher in these regions than the UK average”.

The cost of delivering public services varies between areas due to differences in population characteristics and socio-economic conditions such as age structure, sparsity and deprivation. These affect the number of people receiving a public service and the cost of providing it to each person. For example, it costs more to deliver school services in an area where schools are smaller, transport costs are higher, and more children need extra support because of material deprivation.

Various studies have been undertaken to assess the relative need for spending on public services in the Devolved Administrations. The first official systematic Needs Assessment Study (NAS) was conducted in 1978-79 by an inter-departmental group co-ordinated by Treasury. They developed a methodology and applied it to six spending programmes: health and personal social services; education and libraries (excluding universities); housing; environmental services; roads and transport (excluding railways); and law, order and protective services (excluding police).[36]

The results were presented by showing the percentage (or proportion) of spending per head of population in England required in each region/nation to deliver the same quantity and quality of service, assuming similar policies. (A relative need indicator of 110 implies that a region requires 10 per cent more funding per head to deliver the same services as in England.) Under this exercise, the need indicator for NI was estimated as 131, compared with 116 for Scotland and 109 for Wales.

The Treasury updated the analysis in 1993 with the relative need in NI estimated to have fallen to 122 in part due to methodological changes. (Relative actual spending had fallen by a similar amount.) The Treasury’s estimates of relative need in NI were lower than some unofficial estimates which suggested that the level of public spending in NI was broadly in line with relative need.[37] For example, the Northern Ireland Economic Council estimated that unadjusted public spending in NI was 36 per cent higher than in Great Britain on average between 1980-81 and 1985-86 but that adjusting for need the average premium fell to 0.5 per cent.[38]

A 2003 paper by Professor Iain McLean proposed the use of the inverse of relative GDP per head as a way to set the Block Grant in line with relative need.[39] Over the past 20 years NI Gross Value Added per head of population in NI has ranged between 74 and 80 per cent of that in England which implies a relative need for public spending of 125-135[40], broadly in line with previous estimates.[41]

In 2001 and 2002 the NI Executive departments updated and reviewed the Treasury-led methodology as well as extending its coverage to include almost all public spending (apart from Justice, responsibility for which was not devolved at that time). This resulted in a higher overall need indicator for NI of 125. By far the highest level of relative need was estimated for Agriculture at 342, compared with, for example, 133 for Schools and 117 for Health & Personal Social Services.

In 2009, the House of Lords Select Committee on the Barnett Formula published its First Report. It presented several indicators of need and recommended that any assessment should consider the age structure of the population; low income; ill- health and disability; and economic weakness. It concluded that England had lower need for spending than Scotland, which in turn had lower need than Wales and NI.

In 2010, the Final Report from the Holtham Commission recommended that the Welsh Government should seek a simple needs-based formula as the means of determining changes to the Welsh Block Grant.[42] Although the Holtham Commission was focused on Wales, it estimated a relative need of 121 for NI compared with 105 for Scotland and 115 for Wales. The estimates for NI and Scotland were referred to as “broadly indicative” because the analysis did not include policing and justice which were not devolved to Wales and therefore not included in the analysis.

Chart 3.3 illustrates the outcomes of these previous needs assessments for NI as compared to England. In our first Sustainability Report[43] we used the (conservative) figure of 120 for NI’s relative need, but we have now carried out our own estimate of NI’s relative need based on the established Holtham methodology.

The Holtham methodology was based on principles of simplicity and completeness with a small number of needs indicators covering demographics, deprivation, and costs. These indicators were available on a consistent basis across the UK but were not under the influence of any Devolved Administration. The Commission was of the view that the indicators used should be measurable to a high degree of accuracy and subject to periodic review with census data preferred where possible. On this basis the indicators set out in Table 3.3 were chosen. The combined benefit rate indicator measure, as a proxy for deprivation, was based on the number of key working age benefit claimants plus children in out of work families claiming tax credits plus guarantee-only pension credit claimants.

Source: Independent Commission on Funding & Finance for Wales

To weight the six need indicators, regression analysis was used to estimate their relationship with past budget allocations. Although fewer in number than in the Treasury model, the six need indicators explained over 95 per cent of the variation in spending across the sub-regions of Wales, Scotland and England. The coefficients provided the weights for each indicator in calculating the overall level of need as set out below. The combined benefit rate and the age standardised percentage of the population with a limiting long-term illness have the largest weights.

Our updated estimate uses the most recent data for each of the indicators while using the same GB weights as deployed by the Commission (see Table 3.4).

Under 16 Dependency Ratio (Census 2021)

Between 2008 and 2021 the population aged 16 and under in England increased by 8 per cent compared with only 2 per cent in NI. As a result, the under 16 dependency ratio has increased at a faster rate in England reducing the NI value for this need indicator from 115 to 111.

Pensioner Dependency Ratio (Census 2021)

In contrast, the number of people aged 65 and over increased at a faster rate in NI (31 per cent) than England (25 per cent) between 2008 and 2021 with the indicator amended to exclude females aged between 60 and 64. The net result is that the value of this need indicator has increased from 87 to 94.

Combined benefit rate (DWP and DfC Benefit Statistics May 2022 and HMRC Tax Credit Statistics 2020-21)

The combined benefit rate indicator was updated to May 2022 for the working age benefits with Universal Credit claimants now included, representing 60 per cent of the total for England and 38 per cent for NI. Overall, the combined number of benefit claimants was equivalent to just under one fifth of the NI population, representing an increase on 2008. However, there was a larger increase for England reducing the need indicator from 141 to 135.

Ethnicity (Census 2021)

The proportion of the NI population stating that they were not white increased from

0.8 per cent in the 2001 Census to 3.4 per cent in the 2021 Census. However, this was still substantially lower than the rate for England (19.0 per cent) with the result that the associated need indicator increased from 8 in 2001 to a still low value of 18.

Sparsity (Census 2011)

The proportion of the NI population living in a rural area fell from 39.7 to 36.5 per cent between 2001 and 2011. However, there was a greater decline in England with the result that the associated need indicator increased from 205 to 213.

Limiting long term illness (Census 2011)

Respondents to the 2001 Census were asked if they had a limiting long-term illness, but in the 2011 Census they were asked if they had a long- term health problem which limited day to day activities a little or a lot. This change in terminology did not have a significant impact with the value for this need indicator, which was unchanged at 125. More recent figures on ill health are available from the 2021 Census, but unfortunately, they are not comparable between NI and England.

Table 3.4 shows that the overall need indicator for NI (excluding policing and justice) has fallen from 121 to 120 due to a reduction in the combined benefit rate indicator and the under 16 dependency ratio indicator. The former is primarily due to the inclusion of Universal Credit claimants, although the transition from other forms of benefit distorts the comparisons by individual benefit.

In presenting its findings the Holtham Commission warned that the overall estimates for Scotland and NI were only broadly indicative because they did not include policing and justice. It seems reasonable to assume that the distinct challenges associated with the security situation in NI require higher spending per head in this area than elsewhere in the UK, but this is hard to quantify.

We have used the average level of spending per head on Police Services, Prisons and Law Courts in NI between 2017-18 and 2021-22 (£636) to derive a relative need indicator of 149,[44] compared with a need indicator of 183 if the same figures for 2011-12 to 2015-16 were used instead.[45] Weighted by total DEL spending for the Home Office, Ministry of Justice and Law Officers’ Departments as a proportion of total UK Government equivalent spending over the same period (7.6 per cent[46]), this suggest that the inclusion of policing and justice powers in the Holtham analysis would increase the overall relative need for public spending in NI to 124.

All in all, this suggests that NI still has a significantly higher need for public spending per head to deliver the same quality and quantity of services as England. The precise estimate depends on data used and the weight given to different population characteristics, but it does not seem to have changed dramatically in the last decade or so and is probably still in the range of 20-25 per cent. As the previous section showed, the premium of NI Block Grant funding over equivalent UK Government spending in England has fallen to broadly this level in recent years and would be set to fall significantly below it as the Barnett squeeze tightens over time.

NI’s capacity to increase Rates revenue

The Holtham Commission argued that “the needs-based funding regime for the devolved administrations should assess variations in the ability of each nation to raise its own resources in a consistent way”[47]. As we have seen, this is also a feature of some equalisation systems for sub-national governments in other countries.

The Commission suggested that the ability of the Devolved Administrations to raise resources (i.e., taxable capacity) should be estimated based on relative property prices, which could be incorporated into a needs formula. It argued that adjusting for taxable capacity based on relative house prices would increase the relative level of need for NI from 121 to 124 in 2011. The Commission said that: “In our view, a strong case could be made that such an adjustment gives a purer estimate of relative need than is obtained by basing the analysis on unadjusted expenditure data”.

House prices in NI are currently 56-67 per cent of those in England[48], while Domestic (Regional and District) Rates revenue per head of population has averaged 71 per cent of the average Council Tax payment per head in England over the past 10 years[49]. This relationship suggests that households in NI could, under similar policies as in England, be expected to contribute roughly 85 per cent of the average payment per head in England, although the differences in the systems and the complexity of the property market means that any comparison between Council Tax in England and the Rates in NI should be treated with caution. That said, we estimate that adjusting for taxable capacity on this basis would increase the current NI relative need indicator from 124 to 127, in line with the 3-percentage point increase previously estimated by the Holtham Commission for 2011.[50]

It is worth noting that here is significant variation in the average Council Tax payment per dwelling between local authorities in England (Chart 3.4). Although the average level of the domestic rates bill in NI is towards the bottom of the range in England, the NI average of £1,036 is not significantly different from that in Liverpool (£1,051), Sunderland (£976) or Manchester (£958).

4              Bringing NI’s finances onto a more sustainable footing

Having inherited significant overspending by NI Departments under the Caretaker Ministers, the Secretary of State balanced the books in 2022-23 in effect by taking a

£297 million advance on this year’s Block Grant. To keep the Reserve claim as low as this, he has also implemented some measures that cannot be repeated – notably requiring arm’s length bodies like Translink to run down their reserves. Kicking the can down the road in this way is not a sustainable way to run the NI public finances.

The analysis set out in the previous sections indicates that the Block Grant per head of population in NI is around 25 per cent higher than UK Government spending in 2023-24. This is broadly in line with the relative need for public spending, which we have estimated to be 124 per cent. However, the impact of the Barnett formula means that the funding premium is expected to fall below relative need in the very near future. More immediately, the pace of the fall in the funding premium from 40 per cent in 2018-19 to 25 per cent in 2023-24 has contributed to well-documented financial challenges faced by NI departments in 2022-23 which are expected to continue this fiscal year. Although the NI Executive received funding above relative need in recent years, it is the rapid removal of the additional premium – faster than public services can easily readjust – which has serious consequences for the delivery of public services. Both long-term sustainability and the short-term pressure from the unprecedented drop in relative funding need to be considered carefully.

The Council has identified several interventions that could help bring NI’s finances onto a more sustainable footing. In no particular order, these could include:

•                            Setting a Budget in advance of the financial year. However tight the financial position, the longer departments have to plan, the wider the range of potential options they can deploy to address budget pressures.
• An increase in the Regional Rates. The average Regional Rates bill in NI is significantly lower than the average council tax bill in England, but to some degree this of course reflects the fact that incomes and house prices are lower in NI. Taking account of these factors, NI might be expected to be able to generate only around 85 per cent of the English council tax revenue. Any increase is likely to be politically challenging, especially given current pressures on the cost of living. It would also need to be large to make much of an impact – a 10 per cent increase in Regional Rates would yield only £70- £75 million based on the gross income projections in the Draft Budget.
•                            The introduction of explicit domestic water charges allied to an increase in the Regional Rate. Introducing a charge (of roughly £800 per household) to bring the average level of the charge plus Regional Rates into line with the average level of water charges and council tax in England would raise £615 million, based on there being 768,900 NI households. This would be sufficient to increase total Executive spending by around 4 per cent in 2024- 25 but no doubt at the cost of significant political controversy. The revenue might well also be reduced by the need for some sort of protection or discount for low-income households. The Secretary of State threatened domestic water charging when he set out the implications of 2022-23 for this year’s budget, but it would likely take more than 12 months to implement given the requirement to consult and put logistics in place.
•                            Public sector pay restraint could bear down on spending. Pay accounted for 57 per cent of resource spending in 2019-20 (prior to the distorting effects of Covid-19) and – as we noted earlier – the Barnett consequentials of the extra spending required in England to finance a given percentage pay award are typically insufficient to finance equivalent awards in NI. The Department of Finance public sector pay policy published on 6 December 2022 advised departments that awards for this year should be affordable within the allocations made by the Secretary of State for 2022-23. In response, the Department of Health announced that it was implementing the health service national pay body recommendations for 2022-23 for Agenda for Change staff (£1,400) as well as doctors and dentists at an in-year cost of £215 million. On 6 January 2023, the Department of Finance announced that it had made a pay offer to Northern Ireland Civil Service staff (£552) at a cost of £27 million. The main remaining public sector pay pressure is in respect of Education, reflecting the outstanding teachers’ pay settlements for 2021 and 2022, the teachers’ pay awards and a further settlement due September 2023. Implementing lower awards than at the UK level would have eased the pressure on future years as well as in 2022-23. Additionally, it could have reduced the scale of the claim on the Treasury Reserve and consequently, the amount to be repaid this year. But this would be politically difficult during a cost-of-living crisis with many staffing groups already participating in or considering strike action. This would be highly unpopular and would stir major political controversy, but at the same time certain trade-offs are perhaps inevitable. NI can opt to try to maintain as much pay parity with GB as possible but the consequence of that is, increasingly, likely to be some reduction in range of and/or quality of public services.
•                            Departmental savings from the cessation of what they see as lower priority areas of expenditure, as has already happened in some areas. The Department of Education has stopped its ‘holiday hunger’ scheme (providing payments during holidays to families entitled to free school meals) and has ceased funding for the ‘Book Start Baby’ programme (providing families with a new-born one or two books alongside information on reading). The Department for Health has suspended applications to the core grant funding scheme through which it supports not-for-profit organisations that provide services which improve public health and social wellbeing. And the Department for Infrastructure has announced it will no longer fund cycling proficiency courses in NI schools. But further savings of this type would need to be combined with other measures, otherwise departments would need to deliver very challenging cuts of around 5 per cent (or 10 per cent excluding Health) to fully meet the budget gap from reducing services only. One drawback of planning spending without long-term political leadership and ownership is that spending cuts are more likely to be delivered to opportunistic or salami-slice cuts rather than through a strategic zero-based assessment of what departments should or should not be trying to deliver.

Bringing NI’s finances onto a more sustainable footing

•                            Increasing the efficiency with which public services are delivered. The Council took a detailed look at the NI health service in our sustainability report last year, drawing on analysis prepared for us by the Nuffield Trust. That showed, for example, that hospital stays are significantly longer on average in NI than in England. Of course, not all efficiency improvements are straightforwardly cash releasing. Some may yield improvements in the quantity and quality of service, which is of course desirable too.
•                            Increasing or introducing other fees and charges. Regional Rates and water charges are probably the largest potential revenue raisers, but there are many other public services where fees could be increased (and in some cases, like rail and bus fares, this has happened during 2021-22), or where charges could be levied where costs are not currently recovered. In some cases low or absent fees and charges reflect a conscious desire for ‘super parity’ of provision relative to England, but some of these decisions were taken in the 2010s when NI’s funding was much higher relative to English funding and indeed estimates of relative need. The question should at least be asked: do different financial circumstances require different policies?

                              Simply to accept that the quantity and quantity of services will be lower than elsewhere in the UK.

•                            To seek more money from the Treasury, either through funding for political agreements as in the past or, more durably, by seeking changes to the fiscal framework to incorporate an assessment of need in the operation of the Barnett formula. The UK and Welsh Governments agreed in 2015 that the Welsh Block Grant premium over equivalent UK Government spending would not fall below the 15 per cent relative needs premium identified by the Holtham Commission. This guarantee was initially for the duration of what was then the current Parliament, after which the floor was intended to be reset. But, as part of broader negotiations on the Welsh fiscal framework negotiations, the two governments agreed in 2016 to implement a new funding mechanism from 2018-19 (also recommended by the Holtham Commission). In addition to maintaining the 115 per cent floor, this added an extra 5 per cent to future Barnett consequentials for Wales, thereby slowing the convergence of funding towards the needs floor. There would clearly be potential to put a similar floor in place for NI, and in some respects the argument for both the transitional uplift and a spending floor are greater for NI than they were for Wales in 2015/2016, given the scale of the recent fall in the funding premium and the gap between that and relative need.
•                            Seek to increase the Executive’s tax raising and/or borrowing powers. Interestingly, neither the Scottish nor the Welsh Government have to date used their greater tax-raising powers significantly to raise or lower revenue and spending per head relative to England, but rather to change the structure and re-distributiveness of the tax system.

April 2023