There has been an interesting debate between Patrick Dunleavy and Iain McLean about the costs of Scottish independence. Jim Gallagher adds his voice to the debate to outline the context of the different costs and highlights why these numbers matter.
It’s not surprising that the Scottish government tell voters in a referendum that an independent Scotland would be flowing not with milk and honey, but with low taxes and high public spending. Business tax would be reduced, and public spending increased to produce a more equal society. This particular argument about numbers, however, has its roots in a UK government publication that suggested Scotland’s fiscal position would be markedly less idyllic. One reason would be the additional costs arising directly from setting up a separate state. Iain is surely right to say that the £200 million suggested by Patrick won’t pay that bill.
Different kinds of independence costs
In principle it should be possible to split these state-building costs into 3 categories:
- One-off, non-recurring, costs of transition (this is where Patrick’s work was first referred to). These matter a lot in the short-term, but, amortised over decades, are not all that significant.
- Additional costs from setting up new Scottish institutions to discharge functions currently managed at a UK level. Obviously Scotland would need a Treasury, Foreign Office and other core national government functions. But as Iain rightly points out the big money is in the main non-devolved functions of government–tax collection and benefit payment. Patrick says these are not ‘set up’ costs. Iain disagrees. But regardless of what they are called, they will have to be paid.
- The costs that will follow from independent policy choices – to save money, say on defence, or to spend more – most likely driven by demographics. The Scottish government have made some proposals (quantifying those which save money); the demographic demands from an ageing population are predictable, and worrying for Scotland.
Non-recurring costs of transition
Reorganising things costs money. It was when the UK government referred to Patrick’s work suggesting a typical Whitehall departmental reorganisation costs £15-20 million, and aggregated it up, that he was moved to object. But as a cost of an individual organisational change the figure is certainly not excessive. To give a Scottish example, Audit Scotland described the cost of reorganising the Scottish police service as follows:
“The Financial Memorandum estimated the total cost of implementing police reform would be £137 million by 2016/17 with the majority of spending (£95 million, 69 per cent) expected to occur in 2013/14 and 2014/15. The legislation required the Scottish Government to outline the costs of restructuring. These are expected to account for about 20 per cent of total reform costs (£29 million) with the remaining costs expected to be incurred from investments to deliver wider aspects of reform.”
It might, I suppose, be possible to reorganise an entire government for another £60 million or so more than it cost to reform the Scottish police service, with its 20,000 or so employees. But the comparison suggests that an estimate of £200 million for non-recurring costs is on the low side.
Additional costs that necessarily flow from independence
The big money, however, is in costs that keep on running year after year. Of course it will cost money to have a head of state, a Treasury, a foreign service and so on. But a cursory glance at the UK’s accounts will confirm that these high profile government functions are not the most expensive. The heavy costs are to be found in running the big departments. In Scotland, many of these already devolved, so additional costs there should be minimal.
The two big government functions that would have to be set up for Scotland are tax collection and benefit payment. Iain Maclean’s estimates of these are, in my view, broadly plausible.
Here is why: both tax collection and benefit payment are excellent examples of organisations with a substantial fixed cost, and a relatively low marginal cost. One more taxpayer in the Pay As You Earn system, or one more old age pension to be paid, does not make a proportional addition to the overall cost of running the system. In other words there are economies of scale. No one has done the work to make any detailed estimate of the fixed and variable costs of each organisation. But it’s easy enough to make plausible estimates. Those chartered accountants whom Iain admires do this kind of arithmetic all the time.
In broad terms, the running costs of HMRC are just under £4 billion a year. Scotland’s per capita share of that (used today to make estimates of Scotland’s overall fiscal balance) is something over £300 million. But due to of the economies of scale, it cannot be expected that a separate Scottish tax system could collect all the taxes raised in Scotland today at that cost. If only 20% of the cost of the tax system is fixed rather than variable, then the total cost of running a tax system in Scotland would be just under £1 billion a year. Not very far away from what the accountants estimate and not wholly unlike what, as Patrick says, John Swinney privately thinks.
Similarly, the cost of running the Department of Work and Pensions is about £6.6 billion a year. Do the same arithmetic, and you get benefit administration costs plausibly at £1.5 billion a year. So running costs for these two new Scottish departments can plausibly be estimated at £2.5 billion a year, an addition of over £1.5bn.
Of course one can hope that Scotland might do this work more cheaply, and Patrick seems to think it would. But there is little basis for that hope, as Scotland will inherit UK tax obligations and benefit profiles.
Another even more significant set of costs has been estimated in detail, this time by the Treasury, based on work by the National Institute for Social and Economic Research: an independent Scotland as a new state would inevitably find its borrowing costs higher than the UK’s. There are several effects going on here. First there is the lack of a track record of repayment. Maybe that could be overcome in time, though starting off by refusing to take on a share of UK debt would not be the best way to create a positive image in the markets. Then there is a liquidity effect–there would be less Scottish debt, and it might not be so easily traded as UK gilts. Finally there is the overall fiscal position of an independent Scotland, which would be a heavy borrower. The Treasury estimates suggest that this effect would quickly produce unsustainably large deficits, unless immediate and substantial fiscal adjustments were made.
Policy choices and spending pressures
Governments always have choices, and an independent Scottish government could certainly choose to save money. The Scottish National Party (SNP) has suggested that savings could be made on defence. There are obvious economies of scale in defence, but in the end governments can choose to have less of it, and the SNP White Paper proposes to spend £500 million a year less on defence than a per capita share of the UK budget. You might not get very much for that. If Scotland insists on maintaining a big naval base at Faslane to replace the jobs lost by removing Trident and keeping the historic Scottish regiments, the Scottish Armed Forces are unlikely to be particularly well equipped. But the saving can indeed be made.
Other policy choices are likely offset it – for example the provision of free childcare. But in making choices about independence, these kinds of possibilities (either way) should really be disregarded. Independence isn’t just for the first term of a Scottish Parliament: so it is those costs, and benefits, which are intrinsic to the choice which really count.
What matters more than anything else on spending is demographic pressure. Here independence matters, as Scotland’s demographics are forecast to be more worrying that the UK. The key age dependency ratio which drives demand and affordability not just for old age pensions but other age related services like health would, over the first half century of independence, be worse than the UK’s, with consequences for the overall fiscal position of the new state, set out by IFS and others. Within the UK these pressures still have to be met but would be absorbed in the wider tax and economic base.
What does this tell us?
The one thing this tells us for sure is that if the First Minister says that independence will cost Scotland an additional £200 million, he is talking through his hat. Annual costs from necessary institutions can easily be estimated as between an additional £1 and £2 billion a year. Demographic pressures and interest costs are more significant. The UK deficit is admittedly still worryingly large. But even on today’s levels of spending, Scotland would set out with a deficit of about £5 billion a year more, i.e. 10% of Scottish tax revenue. So additional costs, particularly given the interest cost of financing such a deficit, would clearly require Scotland to make a substantial ‘fiscal adjustment’, that is raise taxes or cut spending. Neither milk nor honey is on the menu.
Jim Gallagher is Gwilym Gibbon research fellow at Nuffield College, Oxford, and visiting professor of government at Glasgow University. He advises the Better Together campaign.
*This article was initially submitted to Patrick Dunleavy’s British Politics and Policy blog but was not posted.
I agree diseconomies of scale are conceivable, and that some departments of Government may be simply too big to manage well – the NHS, say? But for transactional process like tax collection all the logic points to scale economies. And this is bolstered by the only estimates in circulation (ICAS and the Scottish government’s own) suggesting that’s how the numbers look.
What won’t do is to enter into planning for independence on the basis that somehow (because they are now Scottish?) these things will turn out to be cheap to run. I’d have a deal more respect for the position of Mr Salmond and those who support him if they acknowledged that and said “but it’s worth it for self determination”.
“No one has done the work to make any detailed estimate of the fixed and variable costs of each organisation. But it’s easy enough to make plausible estimates”. I think the economies-of-scale argument is plausible only up to a point, and is no substitute for evidence, perhaps from comparable large and small systems (nor is some speculative arithmetic from very clever people). What seems equally plausible is that large organisations (in this case, the one in charge of gathering the money) become inefficient and, at times, dysfunctional – producing diseconomies of scale. Certainly, in my formative years, studying large UK government departments, I rarely came across the argument that they were efficiently large.