I’ve taken quite some time to read the report from the SNP’s Growth Commission. Partly this is a function of the report’s size and complexity, partly the imposition of real life and partly apathy at having to do all this again.
I’ll save my general views on the report for another blog but, for now, I wanted to record those items which are not matters of opinion with which I may disagree but rather appear to be just plain wrong.
The report comes in 3 parts – policies for improving the “potential and performance of the Scottish economy” (part A), a framework for sustainable public finances with independence (part B) and lastly currency and monetary policy (part C). I have focused on Part B covering the framework for public finances, largely because it’s the data I am most familiar with but also because part A appears to a series of aspirations to “get better” which everyone would agree with and is otherwise primarily a list of reviews that will provide the actual policies.
Current v Fiscal Deficit
Let’s start with a simple one. Whilst discussing Scotland’s deficit right now within the UK (B4.29 onwards), the report includes a footnote on page 25 as follows:
It should be noted that Scotland’s current budget balance (excluding capital expenditure that would be expected to boost the economy’s productivity in the longer term) is smaller than the net fiscal deficit – over the last 10 years Scotland’s current budget deficit has typically been 2% lower than the net fiscal balance. The implication is that infrastructure spending and public investment have been running 2% lower in Scotland than the rest of the UK.
SGC Report, Part B, Footnote 20, p25
There really is no other way to put this – this is complete bollocks.
The first part is fine. Current budget balance does indeed exclude capital expenditure. It is, after all, a measure of current spending. And it does indeed give a deficit of typically 2% lower than the net fiscal deficit which, shock horror, includes said capital expenditure. So the 2% difference has absolutely nothing to do with spending anywhere else in the UK. It is simply adding on the 2% GDP capital investment spending that takes place in Scotland.
This should come as no surprise to anyone given that it’s a governmental target to spend 2% of GDP on capital expenditure. And when we compare the actual figures for current versus fiscal deficit for both UK and Scotland (with the difference being the implied capital investment), then it’s clear that they are largely similar. Which, of course, makes sense given the way the Barnett Formula works.
It is, frankly, bizarre that such an obviously false statement could make it’s way into what is supposed to be a peer-reviewed report that a commission, including Scotland’s Finance Minister, has taken two years to produce.
As an aside, it’s worth noting that the capital element of the block grant provided to the Scottish Government from the UK Government will, from 2015/16 to 2019/20, see a 20% real terms increase which, in the word’s of the SNP’s own report, will “be expected to boost the economy’s productivity in the longer term”.
During the first once-in-a-generation indyref, there was briefly some focus on the transition costs an independent Scotland would face in separating from the union, with Alex Salmond at one point claiming this would be as little as £200m. The SGC Report uses the same base data as the former First Minister, a report by Prof Dunleavy of the London School of Economics, but have now settled on £450m as the total cost of transition, spread over 5 years.
I don’t wish to revisit the debate over the veracity of the Dunleavy analysis (for that you may wish to read this [LINK]) but I do find it odd that the SGC have relied on a 4 year-old estimate rather than taken lessons of actual transition costs from their own government. If they had, they would see that the 2017/18 Scottish Government budget includes allowance for £92m to implement the tax and non-tax elements of the 2016 Scotland Act – and that’s just in 2017/18. The next year, they’ve allocated a further £83.6m.
Not sure how someone can argue that transition costs for full independence will be less than that for an extended devolution settlement but it doesn’t stop there. In February 2016, Nicola Sturgeon wrote to then PM David Cameron, complaining (no surprise there) about UK Government contributions to the devolution of Personal Independence Payment and Universal Credit. That letter read:
…we estimate ongoing administration costs to be approximately £200m annually, and set up costs to be between £400m-£600m.”
You’ll forgive me if I am somewhat more than sceptical that transitioning to full independence would cost less than transferring two benefits.
Reserved & Non-Identifiable Spending
Bizarrely, it appears as if the report has confused reserved spending with non-identifiable spending. Reserved spending is that conducted by the UK Government for reserved powers – a good example being pensions. Non-identifiable spending is that which cannot be directly attributed to a particular area of the UK and so responsibility for that spending is apportioned across the entire country according to agreed ratios – a good example being Defence. (I’ve covered this in some detail here [LINK] if you want to read more)
From B4.43 onwards, the SGC attempts to identify savings that an independent Scotland could achieve versus our position in the UK. Here it focuses on spending currently undertaken by the UK Government, i.e. reserved spending. The report, entirely fairly, suggests that an independent Scotland may make different choices and therefore face different, presuming lower, costs. The report first excludes welfare, pensions and economic affairs from any suggested savings – others may wish to note that replicating spending in these areas implies no increase in state pension (not even for WASPI claimants), no increase in benefits and no lifting of the 2-child cap on child benefit with its dreaded “rape clause”.
The report then provides a table of other UK Government spending which is assigned to Scotland in GERS, lifted from Table 3.8 in GERS 2016/17.
So far, so good. I’m sure, in an effort to reduce fiscal deficit, an independent Scotland would indeed look for savings here.
However, B4.58 then goes on to suggest £1bn worth of savings – £0.4bn from spending, £0.6bn in additional revenue. There are reasons to doubt the accuracy of both.
First, the revenue benefits. B4.58 suggests that £2.4bn of spending “would transfer to Scotland and so generate taxation”. They get this £2.4bn by taking the £4.7bn reserved spending figure in Table 4.2 (above), removing £1.5bn for science and innovation that they wish to keep as-is and excluding the £810m shown for International Affairs which the report rightly identifies as taking place overseas (foreign aid and embassies).
The problem is that large portions of reserved spending already take place in Scotland. They are not, as the SGC seems to think, non-identifiable (non-ID) spending elsewhere in the UK. This is a fundamental error that the most rudimentary of proof-reading should have spotted. I struggle to reconcile this with a report that is produced by a Commission including the Scottish Government’s current Cabinet Secretary for Finance.
Using one example to highlight the difference, let’s look at transport. The full database for 2016/17 GERS showing the non-ID / identifiable (ID) split is not yet available but using 2015/16, the total UK Government reserved spending on transport assigned to Scotland was £751m. The portion of this that was non-identifiable spending, i.e. not within Scotland, was £38.6m. Or 5% of what the SGC say would “transfer to Scotland”.
Although the database is not yet available, I had already estimated implied non-ID allocations for 2016/17 in a blog post responding to a similar argument by Prof Richard Murphy. Using the same methodology as in that blog [LINK], instead of the £2.4bn the SGC argue would be “returned to Scotland”, the absolute maximum returned spending from the categories in Table 4.2 is £1.15bn. In fact the sum is likely to be no more than half that when you consider it’s largest contributors are the cost of the BBC – with the majority of Scotland’s contribution already spent here – and nuclear decommissioning – for which actual spend in Scotland is double our contribution to the UK’s total cost.
That £1.15bn also includes £409m on “public and common services” which is dominated by Scotland’s contribution to HMRC, around £250m. Knowing this, we shouldn’t forget that 13% of HMRC employees are already located in Scotland, almost 5% more than our per capita contribution to their costs (8.2%)¹.
Using the erroneous £2.4bn reserved figure, the SGC report argues that Scotland’s fiscal position would immediately improve by £0.6bn. If we instead correctly use non-ID spending, assume an extremely generous reallocation of £1bn and employ the SGC’s methodology, this sum would fall to ~£260m. Using a more likely figure of £500m, the improvement would be just £130m.
Further, the other half of the suggested £1bn savings – on spending – includes “more than £100m for Whitehall Department running costs that will not need to be duplicated in Scotland”. It’s not at all clear from the report how this £100m figure is reached and so I will not try to quantify an alternative but, assuming it is based on the same misunderstanding of reserved and not non-ID spending, then this proportion would also have to come down. Taking a closer look at the proposed civil service staffing changes reveals further problems with the SGC calculation.
Civil Service Staffing Changes
Briefly, the SGC report argues that additional civil service employment in Scotland will result in economic benefit from additional GDP and tax revenue. I won’t challenge the numbers that are given, they seem fairly arbitrary with no provenance provided, but I will point out two inconsistencies.
First, there is an apparent disconnect between employing 4100 additional civil servants, a 75% increase on the current level of staff employed by the Scottish Govt², and the suggestion in B4.58 that an independent Scotland would save £100m versus our current contribution to Whitehall. Excluding the defence costs, Table 5-2 of the SGC report suggests the additional civil service employment salary cost would be £176m. Using the same methodology of salary cost versus total cost as the SGC (B4.58, footnote 24) this would imply additional costs for these new civil servants of £586m. This is a much larger sum than our current contributions to “Whitehall costs” – either the £484m the SGC wrongly suggest could be “transferred to Scotland”, or my own estimate of, at the very most, £409m.
Second, the calculation does not take account of the negative economic impact of civil service staff who will leave Scotland, or lose their jobs, on independence. As already covered above, Scotland enjoys the benefit of some civil service employment that is above our per capita share of costs – 13% of HMRC staff, for example, or 11.4% of DWP. It does not seem realistic to argue that Scotland, with 8.2% of the UK population, would retain these levels of staff to manage the affairs of an independent Scotland. It is therefore only logical that the economic drag from this lost GDP and revenue should be modelled. It is not.
Indeed, the SGC calculations appear to account for absolutely no negative impact whatsoever from independence. But that will perhaps have to wait for another time.
Asset Assumptions for Debt Allocation
The SGC argue that Scotland would start with zero debt but “would agree to contribute to debt interest payments” against a negotiated settlement that takes account of assets and liabilities that are transferred on separation. This contribution to debt interest would form part of an Annual Solidarity Payment which may also include contributions to ongoing shared services – but the debt portion would be fixed at £3bn per year in cash terms.
Other than implying a healthy dose of scepticism as to how likely such a negotiated settlement would be in reality, I’ll try to focus on how that £3bn is reached. Current Scottish contributions to UK-wide debt repayments are £3,249m, 2% of GDP, and projected to remain at the same ratio until 2020/21. The SGC argues that netting off the asset settlement would reduce this sum to £3000m in 2020/21, or 1.6% of GDP.
That asset settlement is reached by taking Scotland’s per capita share of total UK assets (£116bn) and subtracting the value of “Scottish Government, Agencies and Local Government Associated Assets” (£90bn). This leaves a £26bn delta which the SGC argue should be offset against the per capita allocation of debt, i.e. we get a refund commensurate to £26bn of assets.
There’s a slight problem here. The £90bn figure for Scottish assets is taken from the CIPFA report and, as the SGC itself states, includes only “total assets held by devolved Scottish government and local government” – i.e. no assets, whether in Scotland or elsewhere, held by the UK Government. If this is the sum used to negotiate a share of debt then it implies that an independent Scotland would seek to inherit absolutely no military equipment or infrastructure – including land and assets currently in Scotland, absolutely no overseas assets such as embassies, absolutely no share of Royal Bank of Scotland and absolutely no share of the United Kingdom Continental Shelf with all that oil and gas.
It is patently obvious that an independent Scotland would seek a share of these items, certainly the UKCS at the very least. Any addition of this value to the debt share agreement – and I won’t venture a number – would start to increase that Annual Solidarity Payment very quickly. Given the significant value of the offshore industry, it would be fair to suggest that adding just this to the calculation would see Scotland paying more than our current per capita share of UK debt.
After finding much to agree with in Part A, I’m now left with the uneasy feeling that a report being lauded for taking a more honest approach to the economics of independence than the 2014 white paper – something of a low threshold that one – contains such basic errors that will materially impact its conclusions. If this is the “clear and solid foundation” on which the SNP aim to build a “compelling case for independence” then I remain utterly unconvinced.
¹Civil service statistics can be found here [LINK] and, to answer the usual immediate objection, also include reference to the mean salary level of Scottish civil servants. At 92% of the UK average, there would not appear to be a huge boon to be found in returning high value jobs – although some small impact is entirely possible
²The SGC report provides a comparison against total public sector employment of 543,000 to argue it is only an increase of 1%. This seems disingenuous to me as this number would include NHS staff, teachers, municipal workers, etc.