Written evidence submitted by the Fraser of Allander Institute (FSG0007)

Prepared by Dr João Sousa and submitted on 5 January 2025

Who we are

The Fraser of Allander Institute (FAI) is a leading independent economic research institute based at the University of Strathclyde, a registered charity in Scotland.

The FAI is different from traditional academic institutes in that it combines internationally renowned researchers with knowledge exchange specialists who have significant experience from the public and private sectors. This ensures that the Institute’s analysis is not only cutting edge, but delivered in a way which is accessible and relevant. Institute staff are regularly called upon to provide independent briefing and advice to government, parliament and industry.

We have an excellent reputation for independence and impartiality. This ensures that our studies carry significant weight and impact amongst the business and policy community.

Summary of our response

Our response covers the history of the Barnett formula and how it came to be arrived at. We discuss the fact that devolution inherently causes difficulties in terms of allocating funds, but that this is precisely the point of it – to be able to make different choices. There is also a need for some sort of mechanism which takes into account the comparable allocation of funding by the UK Government and devolved administration, but it is hard for it to be easily understandable and fully fair, as well as addressing relative need.

The Barnett formula has stuck around, and has stopped some of the divergence that the Goschen formula that preceded it had created – but it has not really led to the substantial convergence it promised over time, largely due to stronger population growth in England. There has been no significant convergence since the early 1990s.

The transparency of the Barnett formula is commendable, but the lack of recourse over some decisions is not. The Treasury is the ultimate arbiter of comparability and geographical scope of programmes, which is not an institutionally optimal outcome. As we highlight a number of times, having a third party arbitrate disputes would be preferable. But we do not see the Barnett formula being on a statutory footing as making any significant difference to the process, and one needs only look at the lack of impact of the legislation regarding the Sewel Convention to see that would be no silver bullet.

In terms of the Fiscal Framework, we see some of the updates as positive, particularly in terms of uprating limits, even if we think nominal GDP would be a better metric for that. But the Scottish Government is still limited in terms of how it can manage the economic cycle and promote short and medium-term outcomes, and we discuss ways that could be achieved by changing the framework. We also discuss the advantages of multi-year spending reviews to the Scottish Government’s planning, as well as the need to solve the poor incentives faced by the Scottish Government when there are unforeseen in-year changes in policy by the UK Government.

Our response in full

1. How effective is the Barnett formula in calculating the amount of money made available to Scotland for providing essential public services?

- What was the original design intention behind the Barnett Formula, and to what extent has it fulfilled this purpose?

- What potential reforms to the Barnett formula could improve its effectiveness?

We will answer these questions together. Note that these relatively complex questions to answer and depend on how we define effectiveness and what constitutes essential public services.

How to set devolved budget allocations has been an issue for Scotland since administrative devolution in 1885, when the Scottish Office create separate ministerial and official discharging of responsibilities in Scotland.[1] This meant that the subject responsibilities of HM Government department no longer apportioned funds on a whole of Great Britain basis,[2] but instead for England and Wales only.

There are multiple ways in which caused and continues to cause difficulties, but it also precisely the point of devolution – to allow different choices regarding priorities. But the main issue is one of how to account for changes in fiscal policy and in spending in particular. Given that it is the Exchequer’s responsibility to conduct fiscal policy – and ultimate responsibility for managing the UK’s debt and financial market commitments – it is imperative that there is some link between the Treasury’s plans for overall spending and how those are apportioned across both subject responsibility departments and (as were then called) territorial departments, which then came to be devolved administrations.

There is also a fairness point that interacts with this. For example, if the Treasury decided to significantly increase spending, this should spread to not only the departments it allocates spending to (e.g. Department of Health and Social Care) but also on some sort of comparable basis to the overall envelope for the Scottish equivalents (as well as Welsh and Northern Irish) so they can benefit too. Equally, if the Treasury decided to engage in fiscal consolidation, it would not be fair for this to rest solely on departments which have partial coverage of the UK – it should be spread across the whole of the state.

It therefore stands to reason that some sort of equivalence calculation should be put in place, and the question is then how to do so.

The first systematic attempt to achieve this apportionment was the Goschen Formula, put in place in 1888 and named after then-Chancellor of the Exchequer George Goschen.[3] This was based on population shares, and initially set at an 80:11:9 ratio for England and Wales, Scotland and Ireland (the whole island back then, as it was prior to the Irish Free State’s independence).

The 11/80ths principle remained, though its application was characterised as “astonishingly badly documented” by David Heald in 1980.[4] There is no agreement as to whether the 11/80ths were meant to represent population shares, contributions to the Exchequer or a different measure of proportionality. Nick Macpherson, then Permanent Secretary to the Treasury, said in a 2015 speech[5] that the formula was at the time fair to Scotland on both population and revenue, though only fair to Ireland in terms of revenue – and that it was meant to allocate the proceed of wheel tax, horse tax and half of probate duty.

Irrespective of this, it never served as the full extent of the allocation to the Scottish Office, instead serving as a floor on top of which negotiations occurred. One such case was the 1963 and 1965 White Papers detailing the poor quality of housing in Scotland’s Central Belt, which allowed Scottish Secretary Willie Ross to negotiate higher settlements. This helped increase government spending per capita in Scotland to 122% of England by 1976-77, from 114% in 1965-66.

The White Papers were followed by the Kilbrandon Report, the result of a Royal Commission on the Constitution set up in 1969 and which reported in 1973. It paved the way for devolution, but suggested no definitive way for fiscal resource allocation. This was discussed numerous times in the UK Parliament, and the Scottish Labour Parliamentary Group’s paper from 23 March 1976 specifically stated that it was “important to devise a method for arranging or negotiating the block grant that will exclude the possibility of detailed Treasury control”.[6]

The political necessity for an updated formula appeared in early 1979 in advance of the devolution referendums in Scotland and Wales that would have put into place the Kilbrandon Report’s recommendations. Joel Barnett, then Chief Secretary to the Treasury, devised it as a short-term measure, but it is one that has stuck in practice.

The Barnett formula is an improvement in the Goschen formula in several respects. It no longer held proportions constant, which mattered especially given the divergence in population growth between England and the rest of the UK. Instead, these proportions have been recalculated since its introduction on the basis of (mostly) the latest mid-year population estimates, particularly since the 1990s.

It also introduced comparability factors to account for the level of responsibilities in different areas, which meant that only relevant comparisons would lead to changes in allocations. Again, this makes sense – if changes in spending are on matters for which there is no comparable devolved responsibility and are therefore covering that geography already, then it would not seem fair to double count them.

The final area in which there was a change was the method of calculation. The Goschen formula was predicated on levels of spending, that is that Scottish spending should be 11/80ths of England and Wales spending. This essentially fixed overall spending differentials, but given England’s faster growing population, it essentially continued to widen the gap between spending per capita in Scotland and England.

The Barnett formula instead focusses on additions only and calculates the Scottish share as a proportion of the cash addition for England, which should have – in theory – led to some convergence in spending between England and the devolved administrations. In practice, this has not happened significantly as population growth in England has continued to outpace that in the devolved administrations.

Chart 1: Identifiable public spending per person as a share of the UK average since 1981-82

Source: HM Treasury, ONS, FAI calculations

We can see from the chart above that while there was some convergence between Scottish spending and the UK average until the early 1990s, there has been very little movement since then. Identifiable spending in Scotland in 2023-24 was 114% of the UK average – essentially the same as in 1991-92 and within the range since then.

Scotland does relatively well out of the Barnett formula. Of course, there are geographical and demographic factors why delivery of services and infrastructure will always be more costly in Scotland: the sparsity of the population in the Highlands and Islands and the need to provide services such as ferries are part of the reason. But all needs assessments studies – the Treasury’s in 1979 and 1993[7], and the Holtham Commission’s in 2010[8] – showed identifiable Scottish spending above the relative needs assessment, which was not the case for Wales. However, all of these are now long in the past, and we believe that the commissioning of an assessment of relative need is now long overdue.

- How transparent is the UK Government’s decision making regarding the application of the Barnett Formula?

On the whole, much of it is very transparent. The Statement of Funding Policy document is published regularly, and contains the comparability factors which allow for the calculation and checking of the application of the formula. Its simplicity and predictability is its greatest strength – we can calculate what the additions should be, and compare them with what is presented.

The main issue over lack of transparency and recourse is the comparability factors which underpin the Statement of Funding Policy. As discussed later in this response, this is an area in which the Treasury is currently the ultimate arbiter, but that is not necessarily a great set of institutional arrangements

- What impact could putting the Barnett formula on a statutory footing, or otherwise formalising it, have on its effectiveness?

It is not clear to us that statutory footing for the Barnett formula would have any significant impacts on its effectiveness.

The Barnett formula has remained in use since 1978, and is by now a well-established part of the constitutional settlement. Its lack of statutory footing has no practical implications – in that sense, it is no different to other tenets of the UK’s unwritten constitution.

Giving it statutory footing might increase the barriers to changing it formally – by requiring an Act of Parliament – but that is not a particularly high threshold in a majoritarian system like the UK’s Parliament, with parliamentary time being the largest constraint. Recent experience of placing conventions on devolved matters into law is not particularly encouraging in terms of its effectiveness either. The Sewel Convention has been incorporated into the Scotland Act 2016, but the wording’s ‘get-out clause’ – “not normally” – has done much of the heavy lifting. In practice, this further reflects the doctrine of Parliamentary Sovereignty – no Parliament can bind its own successors. We cannot see how legislation would in and of itself make a substantial difference in terms of its permanence.

There are two ways in which a move to legislate could have a positive effect. One might be to create a forum for discussion and potential changes to the formula, if it were felt that those were necessary. For example, a floor could be introduced for Scotland as it has been for Wales. However, this is a less pressing issues for Scotland as expenditure is already at a level which goes beyond what a relative needs assessment would grant.

Another way in which a positive effect on permanence of the formula might occur would be through including it in the work of a Royal Commission. Of course, even in that case the UK Parliament would be formally entitled to either ignore it or change it through its own means, but public clamour against it might be larger. But this might be too big a process for only the Barnett formula, which might instead sit alongside any other constitutional matters in such a process.

2. Following its review in 2023, to what extent does the current Fiscal Framework effectively govern the mechanisms for Block Grant Adjustments and fiscal flexibilities of the Scottish Government?

The review of the Fiscal Framework Agreement in 2023 was welcome and provided a significant improvement in a number of areas. The first one was the provision to uprate all limits in line with inflation, rather than keeping them fixed in cash terms. This includes limits on annual borrowing, as well as the Scotland Reserve. It was never sensible to have these set in cash terms, and so a measure of uprating – especially in light of the large inflation since they were first set out in 2016 – makes complete sense.

There is a debate to be had, however, as to what is the right way of uprating these. The 2023 review used the GDP deflator, which is a broad measure of inflation for the domestic economy. That is more appropriate than alternative measures of inflation, which focus on narrower and sometimes external factors (e.g. CPI).

But the GDP deflator is only one of the components of nominal GDP, the other being growth in real GDP. This means that uprating these thresholds by the GDP deflator will mean that they are falling as a share of GDP every year – just by less than would be the case if they were fixed in cash terms. There is no reason for this to be the case – in fact, a neutral uprating mechanism would be to increase limits by nominal GDP. This is the sort of neutral assumption made by the Office for Budget Responsibility or the Scottish Fiscal Commission, and one which make the Fiscal Framework more aligned with what we’d expect a government’s resources to be.

The increase in the amounts the Scottish Government can borrow for reconciliations was welcome, and makes it less likely that forecasting errors will eat into the amounts available for resource spending.

The permanent move towards the index per capita (IPC) method of indexing block grant adjustments has meant that the Scottish Government’s preferred method – which accounts fully for differential population growth, and which is likely to lead to a larger Scottish Budget than the alternatives – is now set to be the baseline going forward. As the Bell, Eiser and Phillips report[9] highlighted, there are advantages and disadvantages and different levels of emphasis of different principles depending on the method – but clarity going forward is certainly welcome.

The review did leave out a decision on VAT assignment, which has been provided for since the Scotland Act 2016 but has never been implemented. The idea was to assign half the VAT revenues raised in Scotland to the Scottish Budget – so if the Scottish economy grew faster, the Scottish Government would be rewarded with higher revenues, and vice-versa. But after almost ten years, it has become clear that there is no reliable way of estimating Scottish VAT revenues in a precise enough manner to link it to a budget, and therefore we would welcome an admission on all parts that this cannot work and therefore should be dropped.

- To what extent could changes to Scottish Government borrowing powers enable more effective fiscal management, and mitigate against the impact of inflation and economic shocks in Scotland?

Some of the changes to the Fiscal Framework have improved the Scottish Government’s ability to manage economic shocks and their budgetary implications. In particular, the larger limits on borrowing for reconciliations and the uprating of borrowing limits for capital purposes – both annual and as a whole – have improved the Scottish Government’s flexibility in terms of being able to effectively manage budgetary demands within Scotland.

However, at present Scottish Government fiscal policy is largely determined by the overall fiscal stance of the UK Government. The Scottish Government has limited borrowing powers for resource purposes in particular. The Finance Secretary has said multiple times that her government will “balance the budget each year - as we have done every year for 17 years and we will do so again this year”[10], although that is not quite true in a meaningful economic sense. All that means that is that she has met the net cash requirement for the Scottish Government, meaning that she has all the funds to pay for expenditure. But some of those funds are borrowed – for example, for reconciliations or for capital expenditure, or used from revenues from previous years such as in the use of the ScotWind proceeds.

Saying that is balancing the budget is pretty misleading – one could also say that that on that basis, the UK Government runs a balanced budget because it can meet the public sector net cash requirement through debt issuance by the DMO. What matters for the fiscal stance is whether fiscal revenues are enough to cover annual expenditure, and that is why we have recently released some analysis which shows that the Scottish Government has in most recent years run a small deficit.

The constraints on the Scottish Government’s fiscal stance are very real though, because it cannot run a deficit beyond this small level. It is therefore unable to use deficit spending to induce economic activity in case of a downturn (or vice-versa to control expansion beyond a sustainable level) unless that comes through the UK Government’s spending plans.

This means that the Scottish Government’s ability to stimulate the economy in the short-run is very limited; its levers in terms of spending its resources on policies that will enhance economic growth will likely only have effects in the long-run. That is a product of the current devolution settlement – the fiscal stance is reserved, and that means that it is the UK Government’s responsibility to choose the appropriate level of stimulative fiscal policy to run.

But that can have real negative consequences for those living in Scotland. For one, there may be a disagreement on views as to the appropriate level of stimulus necessary (in whichever direction). But there may also be differences in economic growth that would necessitate different levels of deficit spending in Scotland and the rest of the UK, and it is not currently an option to meaningfully diverge on that.

And lastly, the timing of fiscal events might be such that the Scottish Government would like to react differently to economic conditions but is constrained by the time at which the UK Government has set out its fiscal policy.

In all these cases, the hard limits on borrowing by the Scottish Government constrain its ability to manage fiscal shocks. At the more extreme end of the bargain, it may even induce the Scottish Government to raise taxes to meet short-term budget constraints, even if that further weakened the economy, because it must meet its funding requirement every year.

What would be a better solution? As with all these decisions, it entails a trade-off between autonomy and responsibility. Good governance arrangements mean that there should be some way of making a government accountable for its decisions so that it cannot simply run up deficits that are guaranteed by another party, and that clearly is – with good reason – the basis on which the Fiscal Framework Agreement is predicated. But at the moment there is a reasonable argument that the constraints are too tight on what the Scottish Government can do in terms of resource borrowing.

Looser constraints on borrowing would improve the Scottish Government’s options, though a fair challenge then would be how to ensure that there is some form of constraint on the Scottish Government’s behaviour. There are options – for example, municipal bonds without ultimate backing of the UK Government do exist, as do sub-sovereign bonds in the US both at state and local government levels.

The issue is that at the moment any borrowing done by the Scottish Government counts towards its cap within the Fiscal Framework, and given it can borrow at essentially the UK Government’s rate from the National Loans Fund, it makes no economic sense to borrow on the open market. But because of these favourable conditions, the Treasury feels the need to impose tight controls to prevent a build-up of debt.

A reimagination of the current Fiscal Framework in which there is a cap on borrowing from the National Loans Fund but where the Scottish Government can issue debt on the open market that is not subject to the cap could lead to both larger discretion in terms of fiscal policy from the Scottish Government and the use of market behaviour to create fiscal discipline. Of course, the issue there would be whether there would be recourse to the UK Government in a crisis setting – and how that would sit constitutionally.

- Does the current methodology for calculating Block Grant Adjustments effectively reflect the associated transfer of tax revenues or welfare spending?

We have already discussed the index per capita issue, which has now been made permanent, and so we will instead focus on the issues regarding the other parts of the calculation.

It is very difficult to come up with a methodology that is workable, transparent and able to fully reflect the intricacies of all the taxes and social security payments.

The Block Grant Adjustment (BGA) methodology as a whole is used to determine the appropriate deduction to approximate what would have happened in the absence of tax and social security devolution. But it is not a true counterfactual because it is based on a separate geography.

While that is somewhat unsatisfactory, it has the benefit of being based on actual data for which there is a defined outturn that is not debatable – which any counterfactual analysis would be. So there is a practical element which lends itself to the use of receipts in other parts of the UK.

With income tax in particular – as it is the largest source of revenue and also the one most linked to economic performance – there are further complications due to regional imbalances. For a number of years, the BGA had been growing much more quickly than tax receipts, in large part owing to faster growth in incomes in London and the South East of England. The issue is actually exacerbated by the fact that these high-income regions form a larger part of the comparable area (England and Northern Ireland). This has become even more prominent since income tax devolution to Wales, which restricted the geography of comparator areas and therefore increased the share taken up by London and the South East of England.

One option to combat some of this would be to apply separate block grant adjustments for each of the rUK income tax rates (basic, higher and additional) so that the BGA is set from a particular point in time and then grown in line with growth in the different bands. This is the approach used in the Welsh fiscal framework, and means that growth from more tax-rich portions of the income distribution do not result in as large a deduction from the block grant. Doing so would be increase the funding for Scotland, and we would therefore expect it to be in the Scottish Government’s interest.

In terms of social security BGAs, similar issues can occur if the drivers of some welfare payments change within (for example) England and Wales but not Scotland. This might mean – as is the case in the latest Scottish Fiscal Commission forecasts – that some welfare payments’ BGAs could grow faster or slower for reasons outwith Scotland and the Scottish Government’s control, while significantly affecting the marginal cost of the Scottish system.

This is made all the more salient by the fact that the Scottish Government runs an AME-like system of welfare payments in what is more akin to a DEL-type budget – that is, with hard limits that cannot be changed within year. While the types of social security payments run by Social Security Scotland are more long-term and therefore less likely to be linked to the economic cycle, it is still likely that costs might go up at a time of a downturn (if nothing else because take-up is likely to be higher during difficult times). This might mean that the Scottish Government could have to reduce spending in other areas at the same time unless the UK Government loosens fiscal policy, which is a perverse outcome in such a situation – the macroeconomic consensus is that the government should increase rather than reduce spending to get activity going in a crisis.

3. Within the existing devolution settlement, what steps could the UK Government take to offer Scotland more financial certainty?

With no comprehensive spending review since 2021 – prior to the large inflation of 2022 – and with several years just before of one-year rollovers, the Scottish Government’s ability to reliably plan at longer intervals than a year has been severely curtailed. This is a real problem, and one that is completely within the UK Government’s gift to solve.

The new UK Government has committed to holding rolling three-year spending reviews, which is a welcome development. This should help the Scottish Government be able to plan more effectively and conduct its own spending reviews and medium-term financial strategies.

- Is there an appropriate level of coordination between the UK and Scottish Governments regarding in-year fiscal changes?

We do not think the current level of coordination is sufficient. In particular, the Scottish Government’s incentives for spending in-year additions from Barnett consequentials resulting from UK Government policy changes are poor.

At the moment, there is a hard limit on the amounts stores in the Scotland Reserve, set at just over £700 million in 2024-25. We do not see any reason why the Scotland Reserve should have a limit set – it is a perfectly legitimate and understandable behaviour to trade off funds from one year to the next, and one that should be available to the Scottish Government without limits. At present, if the Scotland Reserve’s capacity were near its maximum and the Scottish Government received high Barnett consequentials at Supplementary Estimates, it would have to spend it all in a couple of months or lose it.[11] This makes no sense from a value-for-money perspective; taxpayers would surely prefer money to be spent wisely rather than disbursed just because it would otherwise be gone.

So far as we can tell – mostly from statements on both sides, but also based on Scottish Government behaviour in terms of in-year adjustments to their financial position – there has been more coordination than previously since the new UK Government took office. This is welcome, and continuing this trend should avoid the unnecessary see-sawing we saw at times with the Scottish Government conducting emergency reviews before the UK’s fiscal events – though the Scottish Government’s own actions were also a major factor in those reviews. The move to a single fiscal event taking place in the Autumn should also reduce the size of policy changes close to the end of the financial year, although Supplementary Estimates will remain (with good reason) and therefore some movement might still happen then.

- To what extent can HM Treasury’s decisions regarding the block grant be challenged by the Scottish Government?

There are provisions for dispute resolution between the UK Government and the Scottish Government in the case of the Fiscal Framework, and particular on spillover effects of policies. For example, there was long-running dispute regarding the spillover effects of changes to the personal allowance, which was resolved back in 2022.

But these mechanisms are not automatic nor independent, and rely instead on agreement between the parties. If no agreement is arrived at, then no transfers occur. This means that it is not a particularly appealing process – it can drag out even if should be straightforward (again, see the personal allowance dispute), which disincentivises the raising of disputes.

A jointly nominated dispute resolution body might be a better solution if the objective were to raise the likelihood of these disputes being solved, although it would still face some of the issues regarding its permanence that other aspects of the devolution funding framework such as the Barnett formula.

The Treasury is also the ultimate arbiter of what is included in Barnett and what the comparability factors are through its Statement of Funding Policy. This is more of an issue for large projects that might have large consequentials such as High Speed 2 or the Olympics[12], but might in principle apply to any other projects. This process would also benefit from a third-party dispute resolution mechanism, although some of the issues raise above would still apply in terms of permanence.

4. Are there any comparative perspectives that should be considered when assessing the effectiveness of fiscal devolution in Scotland?

- What learnings can be drawn from Wales and Northern Ireland’s funding settlement arrangements?

- Are there any learnings from international perspectives that should be considered?

We will answer these questions together.

Devolution in the asymmetric form it exists within the UK is not overly common – most countries have a more formal arrangement than exists in the UK, and federal systems are also common (e.g. US, Germany, Canada, Australia).

It is beyond our expertise to provide a comprehensive overview of these, but it is worth noting that financial arrangements can vary widely depending on the principles and emphases put on different objectives. For example, German länder operate an equalisation payment system in which the revenue generation potential is equivalised to some extent. Looser sovereign states like the US have much less of that revenue sharing – the Nixon Administration introduced a programme that shared its name, but was much more like the block grants seen in the UK.

One of the main learnings from the Scottish, Welsh and Northern Irish funding arrangements is how permanent and long-standing arrangements meant as temporary can turn out to be. The Barnett formula is an example of this, but so are the different levels of identifiable spending per person that have remained in place over time.

The UK’s lack of a written constitution has probably made this harder to tackle, as there is no document in which the principles that should underlie devolution and funding on a regional basis are outlined. This means that it is those that are most disadvantaged by the current state of affairs that have the incentive to highlight this – as happened, for example, with the Welsh Government’s establishment of the Holtham Commission – while other areas of the UK have little incentive to do so if they do relatively better out of the system than they would with a more needs-based formula. January 2025