Havering over Welsh taxation

While the Silk Commission carries on work on Part 2 of its inquiry, the UK Government has been deliberating slowly on the Part 1 report. Promises of an ‘early’ response vanished, as did the commitment to one in the ‘Spring’. The summer solstice came and went, with no response from the Secretary of State other than a note that ‘good, positive progress’ had been made, ‘many issues’ resolved, but some remained outstanding. The rumour mill abounds with explanations of what the unresolved issues might be. If what happens to the Welsh Government’s Block Grant is not one of them, it should be.

The approach for dealing with the reduction in the block grant recommended by the Silk Commission sounds comparatively straightforward in principle, though it is rather harder to apply in practice. In the first year the new arrangements are in operation, the block grant is cut by an amount corresponding to the yield of the devolved tax ‘space’ – 10 points of personal income tax in the case of Silk (and Calman/Scotland Act 2012) for Scotland.  That cut is then adjusted ‘proportionately’ in subsequent years. What ‘proportionately’ means here is not clear. The Holtham Commission did sterling work in identifying what that might mean in practical terms, recommending what it called the ‘indexed deduction’ approach for personal income tax.  The same approach applies in principle to other devolved taxes, but the yields of those are modest so the issue is not so vital there.

The ‘indexed deduction’ method would involve taking the Welsh proportion of the overall UK revenues from that tax, and reducing the block grant by that proportion.  So, if devolved income tax in Wales generates 1.75 per cent of total UK personal income tax revenues in year one, the reduction in the block grant would be 1.75 per cent of UK personal income tax in each subsequent year – whatever the change in overall UK personal income tax revenues. The amount of the deduction would go down if overall tax revenues went down, and be increased if revenues went up. The result would be that the Welsh Government would gain if its use of its powers increased tax revenues in Wales ahead of the UK as a whole, and lose out if they declined more than the UK as a whole.  This approach has been agreed between the UK and Scottish Governments for the working of the Scotland Act 2012, but work on what it means in practice is ongoing in the ‘Joint Exchequer Committee’ established by the two governments. There has still not been any published attempt to show what the impact of making the cut and adjusting it by that method would be.

Applying the ‘indexed deduction’ method is comparatively easy for Scotland. The Barnett formula means that the Scottish block grant is comparatively generous. One can argue about how generous it is, but it is clear that the Scottish Government’s block grant exceeds by some distance any reasonable estimate of Scottish relative needs. Holtham estimated Scottish needs at 104 or 105 per cent of English ones, but depending on how one cuts the numbers (which is tricky) Scotland gets around 118-120 per cent of English spending for services covered by the block grant.

A further quirk is that the public spending boom of the 2,000s should have led to quite rapid convergence in devolved spending on the ‘English’ level – but, for Scotland, it did not. It appears that Scotland’s declining population cancelled out the convergence effect in the block grant, since convergence relates to per capita levels of spending, while the block itself is calculated as a lump sum and updated population numbers only affect incremental changes to that. So if the application of the reduction in the block grant affects the overall resources available to Scotland, it will only eat into that ‘cushion’ of the Barnett bonus – and it will not make a difficult situation significantly worse as time goes by.

Wales would love to have Scotland’s problems. It is clear that Wales is somewhat ‘underfunded’ given its present relative needs. At present, the block grant provides 113 per cent of the English level of spending on devolved services – while Holtham found Wales’s relative needs were between 114 and 117 per cent. That creates a different set of difficulties. If the block grant fails to produce a ‘fair’ level of funding relative to need at the outset, any cut in that grant – however it is adjusted – will probably make matters worse, as convergence happens. As a result, it becomes very hard to reconcile devolved fiscal accountability with reasonable UK-wide equity in public spending.

Matters needs not necessarily get worse, if the grant were adjusted to compensate for unfairness in funding before any reduction is made to allow for devolved income tax. The demand for a ‘fair’, needs-based grant – as articulated by Holtham – would be the simplest and most effective way of doing that. But a needs-based grant looks to be pretty clearly off the cards at present. The effects of introducing that for Scotland, particularly in the run-up to the 2014 independence referendum, are frightening enough to send politicians running in the opposite direction.

By pursuing its bilateral discussions about the block grant with the UK Government – and excluding it from the Silk Commission’s remit – the Welsh Government minimised its influence over securing ‘fair funding’, as well as preventing the Silk Commission from taking a comprehensive view on Welsh devolved funding. What it got instead – the deal announced last October – was promise of some undefined action if convergence appeared to become a material issue, though it isn’t at present because of the restraints on public spending at Westminster.  (My discussion of that on Devolution Matters is here.)

How this would be resolved if or when convergence comes back on the agenda would involve a good deal of bargaining and haggling between the Treasury and Welsh Government, and a good deal of reliance on subjective assessments.  Although the Welsh Government seems to have a good deal of confidence in that deal, it is not so much a sticking plaster to help a broken leg, as a fig leaf.

Even then, a ‘fair’ grant would need an adjustment mechanism. You would need to be able to adjust the Welsh block (before the deduction for the share of devolved income tax) as spending changes in the reference point – so Wales gets a consequential change as spending on health or transport in England goes up (or down).  The simplest adjustment mechanism is that used for Barnett – allocating a population share of changes in spending on ‘comparable functions’ in England.  But any formula that works in that way will have a convergence element built into it.  So the problems caused by the Welsh block grant falling below Welsh relative need will not go away.

Indeed, it is made worse because the devolved tax power transfers a degree of volatility risk to the devolved level, while devolved public services are counter-cyclical or inflationary in their cost. A devolved government needs to know as accurately as it can how much money it will have for those services, and the starting point for that figure must deliver a comparable level of spending to that in England.  The more subjective the mechanism for adjusting the numbers, the less certainty and accuracy there is in the system.

Each of these problems is capable of being fixed. It would be quite possible to build into the mechanism for implementing Silk an adjustment to the block grant to avoid convergence, and another to cut the block grant to allow for partially devolved income tax. It would even be possible to establish a system that was also robust and predictable, and pretty stable, though HM Treasury would probably baulk at the loss of control over spending policy that would entail.

But the problem is that such mechanisms will need to be applied by the Treasury, and run on Treasury estimates which will necessarily have an element of subjective estimation built into them. By contrast, the day to day, year to year, operation of Barnett is pretty automatic and clear. The most serious problems arise when it is changed at a spending review.  So ironically, there is a real prospect that the overall effect of devolving income tax while making sure other changes do not damage Wales financially will increase the extent to which Welsh public spending depends on HM Treasury’s calculations, not reduce it. Ensuring a measure of fairness may mean less clarity about how financing works.

And that is the real problem.  The goal of the Silk recommendations is to increase the National Assembly and Welsh Government’s ‘fiscal accountability’. That means establishing clear lines between what is a devolved responsibility and what is a UK responsibility. There is little point in voters being able to hold the Assembly to account for increased (or reduced) income tax if there can then be arguments that this only happened because the Treasury has allowed it. That would not add to accountability. In fact, by creating scope for extra arguments between governments and blame-shifting, it would reduce it.

There are two points here that require further consideration. The first is that the detail of any response implementing Silk needs to be looked at carefully, to see how that mechanism will work.  Steering a course that delivers the benefits of Silk – in the form of increased autonomy and accountability – is difficult, and UK Government claims of success should be treated with scepticism given the difficulties of delivering these objectives.

Second, one has to ask how long the financial system for devolution can go on being amended and patched in this way. It is increasingly looking like one of Heath Robinson’s strange jerry-rigged machines, and increasingly incapable of actually doing what is demanded of it. These problems are much worse for Wales than for Scotland, but Scotland has them too. What look like bolder approaches – such as my proposals set out as part of the IPPR’s ‘Devo More’ project – in fact resolve them much more effectively, by trying to start with a clean slate rather than perpetuating the mess that has accumulated over decades. At some point, clarity and comprehensibility need to take priority over political or administrative convenience.

As part of that, the Treasury needs to be asked a hard question: why does the same framework for financing arrangements have to apply to Wales as to Scotland and Northern Ireland? While almost every part of the three sets of devolution arrangements varies a good deal, the Treasury has insisted on a measure of symmetricality in the block grant and the Barnett formula. It is rather a superficial form of symmetry, as when one digs down there are many substantial differences between each country’s arrangements. At present it is Wales alone that is underfunded relative to need by the block grant, and therefore only Wales that is exposed to the acute problems of convergence on an English level of public spending.  Symmetry causes problems for Wales in a way that it does not for Scotland or Northern Ireland.

In his recent speech at the Wales Governance Centre in Cardiff, Welsh Secretary David Jones lauded the virtues of ‘asymmetric devolution’. Asymmetry when it comes to the operation of financing would have a direct and tangible value for Wales.  It will be interesting to see whether that was merely an attempt to defend a messy status quo, or a preparatory step for an imaginative deal to make fiscal devolution for Wales work.