How should the Barnett Formula be adjusted to reflect devolved taxes? A question of allocating risk (wonkish)

David Bell and David Eiser

The UK parties have committed to retaining the Barnett Formula to allocate grant to Scotland. The way in which the Barnett Formula is adjusted to reflect devolved taxes will have important implications for the budgetary risks facing the Scottish and UK Parliaments. This blog discusses two possible approaches to adjusting the Barnett Formula. It shows that the adjustment approach proposed in relation to the Scotland Act powers exposes Scotland to a different set of risks on the spending side relative to the revenue side which seems inconsistent. This approach also exposes Scotland to UK policy decisions, and these could affect Scotland positively or negatively. An alternative approach is discussed which overcomes these problems, but exposes Scotland to a higher rate of convergence in per capita spending with rUK.

Tax devolution and the Barnett Formula

The Smith Commission is expected to propose devolving a basket of tax powers to the Scottish Parliament. It seems extremely unlikely that the revenues from the devolved taxes will be sufficient to cover Scotland’s spending budget, and there will therefore be a continuing requirement for the Scottish Parliament to receive some form of block grant from Westminster.

The UK parties have committed to retaining the Barnett Formula to determine this grant. But the way in which the Barnett formula is adjusted to reflect the taxes devolved to Scotland will have important implications for the budgetary risks that the Scottish Government will be exposed to, and those from which it will be protected. How the Barnett Formula is adjusted to reflect devolved taxes is therefore important.

In the first year that a tax is devolved, the adjustment to the block grant is straightforward. It simply entails deducting from Scotland’s existing block grant the value of the devolved tax in Scotland. But how should the grant be adjusted in the future to take account of changes to the tax rates or base?

A key principle underpinning the grant adjustment is presumably to ensure that Scotland faces the budgetary consequences of its policy decisions. But concern has been raised that this principle may be difficult to realise within the context of the Barnett Formula.

As an illustration, consider the following scenario. Imagine that income tax has been wholly devolved to Scotland. What would happen if the UK Government increased the rate of income tax in the rest of the UK (rUK) to raise an additional £10bn of revenue, and used this £10bn additional revenue to fund increases in health or education spending in England? Under the existing Barnett Formula, Scotland would receive an increase in its block grant equal to a population share of £10bn (approximately £1bn). The fact that Scots had received a funding consequential without facing any of the tax consequences would strike many as being unfair[1].

How might the Barnett Formula be adapted to ensure that this apparently perverse situation does not arise? The obvious approach to dealing with this issue is to treat devolved taxes as being ‘comparable’ in the same way that devolved spending is considered ‘comparable’. But there are at least two ways of doing this, each with different consequences for the budgetary risks that Scotland faces.

Defining Comparable Taxes: the Indexed Deduction method

The first approach indexes changes in the devolved revenue in Scotland to ‘comparable’ revenues in rUK. This is the approach that has been proposed in relation to the Scottish Rate of Income Tax (SRIT) introduced by the Scotland Act, and is known as Indexed Deduction (ID). In the first year that the tax is devolved, the value of revenues from the devolved tax is deducted from Scotland’s block grant; this is known as the Block Grant Adjustment (BGA). In future years, Scotland’s revenues from the devolved tax are indexed to the growth rate of comparable revenues in rUK. For example, if the ‘comparable’ devolved revenues in rUK grow by 5% in year 1, then the value of Scotland’s block grant deduction grows by 5%. If Scotland’s revenues grow by more than 5% it will be better off than it would have been without the devolved tax, but if its revenues grow by less than 5% it will be worse off.

One of the implications of the ID approach is that Scotland is exposed to demographic risks. If its population (or working age population) grows more slowly than rUK’s, its tax revenues may grow more slowly as well, even if incomes per head in Scotland have grown relatively more rapidly. One might question whether it is fair for Scotland to face such demographic risks when policy levers on migration are not devolved.

Indeed, it should be noted that, on the spending side of the equation, the Barnett Formula protects Scotland from demographic risks. This is because Scotland receives a population share of spending increases in England. There is therefore something of an inconsistency in protecting Scotland from demographic risks on the spending side of the adjusted Barnett Formula, but exposing it to these risks on the revenue side.

What about the hypothetical scenario described above, where income tax is wholly devolved to Scotland, and taxes are raised in England to fund spending on health/education in England? Imagine that income tax had been devolved to Scotland in 2013/14. To keep things simple, assume that there has been no growth in population or incomes in England, but that in 2014/15 income tax rate is raised in England so as to raise tax revenue by £10bn. This £10bn is used to fund additional health/education spend in England. This provides Scotland a funding consequential of £1bn (its population share of the additional spend). The question that then arises is how much of this £1bn additional grant is clawed back through the block grant adjustment?

The £10bn increase in tax revenues in England equates to around 7% of England’s income tax revenues in 2013/14. So Scotland’s block grant adjustment (BGA) will also rise by 7%. Given that Scotland’s income tax revenues in 2013/14 were £11bn, this amount would be deducted from Scotland’s block grant in the first year that income tax is devolved. In the second year, this amount will rise by £11bn*7% which equals £0.8bn. So Scotland’s overall grant increases by £0.2bn (the £1bn increase in spending due to the Barnett consequential minus the £0.8bn increase in BGA).

The reason for this apparently bizarre effect (i.e. Scotland benefitting from a balanced budget increase in comparable English revenues and spending) arises because of the way that the Indexed Deduction method combines a population share of a levels change on the spending side with a rate change on the revenue side. Scotland will benefit from an increase in English spending if this is funded through a devolved tax which raises less per capita in Scotland than in England. This is because Scotland’s population share of the change in nominal spend will always be greater than the proportionate change in Scottish revenues required to finance this. Of course, the corollary that if England cuts the devolved income tax in order to cut spending on health and education, then Scotland will receive a reduction in its grant. Finally, note that if Scotland’s per capita revenues from the devolved tax are higher than England’s, then the effects are reversed: Scotland would lose from a balanced budget increase in English spending and gain from a fall.

Defining Comparable Taxes: the Levels Deduction method

An alternative method of adjusting the Barnett Formula to reflect devolved taxes is to adjust it in levels terms. As before, in the first year Scotland’s grant is reduced by the value of the devolved revenues. In future years, Scotland’s grant is equal to its grant the previous year, plus a population share of the change in comparable spending, minus a population share of the change in comparable revenues.

This approach has a couple of advantages. One is that of symmetry of risks: Scotland is protected from demographic risks on both the spending and revenue sides. The other advantage is that an increase in the devolved tax in England to fund devolved spending in England has no budgetary effect on Scotland. To see this, return to our earlier scenario in which income tax is devolved. If England raises an additional £10bn in revenue to fund devolved spending, Scotland receives a population share of the spending increase minus a population share of the tax revenue increase; these two are identical and thus cancel out – Scotland receives no change in its grant.

This approach is however likely to have a long-run cost for Scotland, and that is of an increase in the rate of convergence towards equal per capita spending in Scotland and England. Aficionados of the Barnett Formula will know that it should theoretically induce convergence in per capita spending between Scotland and England over time. This is because, for a given nominal increase in ‘English’ spending, the per capita spending increment is the same for Scotland, and thus the effect of Scotland’s higher initial per capita spending should become proportionately less over time. The levels deduction method also introduces this convergence property to the revenue side of the equation: assuming Scotland starts from a position of lower revenue per capita than England from the devolved tax, by subsequently assuming that Scotland’s revenues increase at the same per capita level as those in England, the initial divergence becomes less over time.


This discussion highlights some of the difficulties involved in adjusting the Barnett Formula to reflect devolved taxation in a fair and sensible way. On the one hand, major tax devolution would provide a good opportunity to reform Scotland’s grant mechanism more fundamentally, but the leaders of the main Westminster parties seem to be committed to maintaining the Barnett Formula. On the other hand, many of the tricky issues raised in this blog – which budgetary risks should Scotland be exposed to for example – would remain to be resolved under any alternative grant system. And in this blog we have not even considered how grant might be allocated to Scotland to reflect any benefit spending powers that might be devolved following Smith; this will be the subject of a future post.

This blog has discussed two ways of adjusting the Barnett Formula to take account of devolved taxation. One partially exposes Scotland to demographic risks, and results in spillover type effects from balanced budget increases in the devolved tax in England, which may or may not advantage Scotland. The other protects Scotland from demographic and spillover risks, but does imply a faster rate of convergence towards equal per capita spending. The approach that a Scottish Government would prefer will depend on factors including: Scotland’s per capita revenues relative to England’s for any tax that is devolved; relative rates of population growth in England and Scotland; and likely growth rates of comparable spending and revenues in England.

[1] Wales and Northern Ireland would also receive a funding consequential, but assuming income tax had not been devolved to Wales and Northern Ireland, taxpayers in these countries would pay a higher rate of tax.

About David Bell

David Bell FRSE is Professor of Economics at the University of Stirling. He graduated in economics and statistics at the University of Aberdeen and in econometrics at the London School of Economics. He has worked at the Universities of St Andrews, Strathclyde, Warwick and Glasgow. His research is mainly in labour economics, fiscal decentralization and the economics of long-term care. He has been budget adviser to the Scottish Parliament Finance Committee. He is PI for the Healthy AGing In Scotland project (HAGIS).
This entry was posted in Uncategorized. Bookmark the permalink.