brittaniaThis note reports on a paper of mine which has just been published in the Fraser of Allander (FoA) Economic Commentary, modelling how the Barnett formula will operate once it has been modified under the Smith reforms: (“The Barnett formula under the Smith Reforms”, J.R. Cuthbert: FOA EC, vol 39(1). As will be seen, the modelling has some pretty unpalatable implications.

After Smith has been implemented, Scotland will receive revenues from the taxes it will control, (principally income tax), or which are hypothecated to Scotland, (like about half of Scottish VAT receipts). There will be an abatement to the Scottish government’s block grant as calculated by the original Barnett formula, to compensate Westminster for the tax revenues it will have foregone. And this abatement will be increased each year by an appropriate form of indexation. For income tax, this indexation will be on the so-called “Holtham” basis – i.e., in line with the growth in the overall UK income tax base.

What the FoA paper does is to model the operation of the Barnett formula, as modified by these arrangements, and then to consider how relative levels of per capita public expenditure in Scotland and England are likely to evolve. The paper makes a number of simplifying assumptions: but nevertheless, it is of considerable interest, in that it reveals the underlying pressures which are likely to determine how the system evolves.

It turns out that there are three key parameters which will determine how things will change: namely, the rate of growth of public expenditure in England; the relative rate of growth in the tax base between England and Scotland; and the relative rate of population growth between England and Scotland.

The implication of the modelling is that, left to itself, the system could not go on indefinitely with fixed values of these parameters, (other than in a very unlikely special case). Ultimately, ( and possibly quite quickly), the ratio of per capita spending in Scotland relative to England would evolve to politically unacceptable levels: with the likelihood being that this would involve public expenditure per head in Scotland being pushed down well below levels in England. Moreover, it is probable that dynamic effects would set in, with the parameters themselves changing in ways which would make the adverse effects even worse.

In other words, the fiscal system being set up under Smith will be unstable, and will only operate tolerably under continued intervention by government. Moreover, it is unlikely that the powers which the Scottish government will possess, (which, in the economic sphere, are very limited), will be sufficient to enable it to stabilise the system by its actions. The implication is that, if the system is going to work at all satisfactorily, Westminster is going to have to be prepared to intervene actively, and to adjust fiscal transfers between Scotland and England on a regular basis.

This, however, poses a number of problems. On its past record, it is very doubtful if Westminster is equipped to discharge such a role effectively. And the whole thrust of Smith, that Scotland now has increased powers and should get on with it, runs quite counter to the role which Westminster would have to discharge.

It is essential that the inherent flaws in the fiscal system being set up post Smith should be exposed now. Otherwise, once things go badly wrong under Smith, (as they will), the Scots will get the blame: “You can’t even run Smith: what would you be like running an independent Scotland”.

But on the other hand, this puts the SNP in a difficult position. The SNP signed up to Smith: so it might appear to be very difficult for the SNP to come out now and say that the proposals are ill-conceived and potentially disastrous.

Overall, the current situation is an object lesson of the pitfalls of trying to do fundamental constitutional reform in seven weeks, without actually looking at the detailed implications of what is being agreed to. But this points to a way out. The SNP should argue that, while they were willing to sign up to the principles in Smith, this was of course conditional upon satisfactory arrangements being arrived at for implementation. And satisfactory arrangements are far from being arrived at yet. This is true both as regards the adjustments to the Barnett formula, (as demonstrated by the modelling paper): and also in relation to the critically important gearing problem, (which under laid an earlier Bella piece, of 25th January 2015).