Thursday 17 November 2016

Does Scotland Simply Spend Too Much?

Today saw the release of the latest HM Treasury Country and Regional Analysis (CRA) report which looks at public spending levels across the UK.

From a Scottish perspective the figures suggest a simple and powerful headline message: "Public spending in Scotland ‘£1500 higher than UK average’"

This is of course not news. In a post-indyref world, surely most Scots are aware that Scotland enjoys higher public spending than the average of the rest of the UK?

Regular readers of this blog will know there are nuances aplenty and some caveats needed (which we'll come to) - but the simple maths is easy to follow: £1,500 per person higher spending on 5m people = £7.5bn. That's the lion's share of the current c.£9bn effective fiscal transfer Scotland receives from the rest of the UK - so the transfer is largely explained by the fact that we spend more, not that we generate less.

Those who react to observations about the scale of Scotland's notional* higher deficit (and the resultant fiscal transfer from rUK) with "this just proves we suffer under mismanagement from Westminster" have overlooked this very simple fact: we have a bigger deficit because we enjoy higher levels of public spending. These are levels of public spending that (as I frequently tire of pointing out) would be simply unsustainable without either ongoing fiscal transfers from the rest of the UK or - were we to be independent - new windfall revenues (like another oil boom) or decades of spectacular economic growth.

* notional, because the whole point of being in the UK is that this deficit doesn't actually exist - Scotland's spending doesn't have to be met by Scotland's tax revenues and our population share of the UK debt alone, we get to pool and shares across the UK (if you like, this support for our higher spending now is merely fair reciprocation for sharing "our" oil revenues in the 1980s)

So few people should be surprised to hear that Scotland has relatively high levels of public spending. The absolute size of the difference may, however, cause raised eyebrows among that small cluster of people who are familiar with the Scottish Government's own Government Expenditure and Revenue Scotland (GERS) figures
  • GERS 2015-16 showed Scotland having £1,2141 per head higher Total Managed Expenditure (TME) than the UK average
  • This CRA report shows Scotland having £1,4602 per head higher Identifiable Expenditure
That's a difference of £250 per capita or £1.3bn - so not an insignificant amount.

There are a few factors that can explain the difference ;
  1. The CRA headlines are based on identifiable expenditure only - but given the vast majority of non-identifiable spend is allocated in GERS on a per capita basis, it seems unlikely that this would explain the difference3
  2. Timing: by publishing GERS earlier than in previous years, the Scottish Government had to rely on their own estimates and analysis rather than using the published CRA figures as they were able to in prior years4. We'll probably have to wait to see how the GERS 2015-16 figures are restated in the 2016-17 publication to find out the extent to which this led to inaccuracies.
  3. Methodology: the Scottish Government make various adjustments to the CRA spending data when compiling GERS - for a combination of technical and/or judgement-based reasons5.
I haven't fully reconciled the figures - life's too short - but overall it seems like overall identifiable expenditure matches reasonably well between Scottish Government GERS and the CRA report - with a net difference of only about £90 per capita (with higher spend in the CRA analysis)6. We'll use the detail of this analysis to apply caveats below - read the note if you really want to follow the data trail or just trust me; it's up to you.


The balance of the difference between GERS and CRA (£160 per capita) must be explained somewhere among the world of pain that is "accounting adjustments" - we'll leave those for another day8.


So armed with the above knowledge and caveats, here's what the CRA analysis shows us.

First of all there's the age-old question of whether we compare our spend per capita with UK average (including Scotland) or with the rest of the UK (rUK). Personally I find it easier to think in "vs rUK" terms, on which basis our (CRA analysed) spend is in fact about £1,600 higher.


As an aside: those who think that HM Treasury are always spinning against Scotland might ask themselves: if that's the case, why they don't use "versus rUK" comparisons for their press-releases?


Looking at the breakdown by spend type in the table below it's clear Scotland  spends more than the UK average in all spend categories (as we have aleady observed through the GERS figures in blogs passim).

Figures we can trust: materially consistent between CRA and GERS
The £ figures in these bullet points are how much higher Scottish annual spend per capita is than the rest of the UK, the percentage is that figure vs the rUK per capita spend
  • £331 (+8%) on Social Protection (mainly pensions & benefits) 
  • £195 (+15%) on Education & Training (primary, secondary, tertiary)
  • £149 (+7%) on Health 
  • £131 (+168%) on Enterprise & Economic Development
  • £99 (+105%) on Public & Common services (executive & legislative etc.)
  • £93 (+154%) on Agriculture, Forestry & Fisheries
  • £90 (+82%) on Recreation Culture & Religion (sports & culture)
  • £79 (+19%) on Public Order & Safety (police, fire, law courts, prisons)
Figures to treat with Caution as CRA vs GERS differences are material
  • £68 Transport (road, rail, other) - GERS suggest even higher?
  • £91 Environment Protection (mainly waste management) - GERS suggests not that high, (nuclear related?)
  • £239 on Housing & Community Amenities - but big caveat as GERS suggests a far smaller figure (Housing Association7 accounting treatment related?)
It's pretty straightforward really - even with caveats around methodology and even if we accept the (kinder on Scotland) Scottish Government GERS assumptions - Scotland, as it's governed today, is an expensive country to run.

The extent to which this relative high cost is a structural issue (remote and island communities, demographics that mean a relatively high dependency ratio, areas of chronic deprivation etc.) or are a result of a Scottish Government that's keen to spend the Barnett Formula money on freebies that win votes (no tuition fees, free prescriptions, scrapped tolls etc.) is a debate that will run and run.

This analysis is a timely reminder that there would be one easy way for the SNP to further the cause of Nationalism: they could ask for the fiscal framework to be scrapped so that Scots could get used to the lower public spending we'd need to endure while we waited for the miracle of independence to produce exceptional economic growth ...








NOTES

1. GERS 2015-16 page 4 table S.4 shows this rounded to £1,200 - precise figure from back-up data tables

2. CRA Report page 14 table A2 - per capita spend figures: UK £10,536 - Scotland £9,076 = £1,460

3. The three largest non-identifiable spend areas are public sector debt interest, defence and international services. Between them these account for more than 80% of non-identifiable expenditure. All of these (and many others) are allocated to Scotland in GERS on a per capita basis (they therefore can't explain any per capita spend difference between Scotland and the UK average)

4. GERS page 25:
"This edition of GERS sees a change in methodology to allow earlier estimates of expenditure to be produced. Whilst expenditure for years prior to 2014-15 continues to be based on the CRA, the 2015-16 estimate is primarily based on data from the Scottish Government accounts system combined with data from HM Treasury‟s Public Expenditure Statistical Analyses"

5. Detailed Expenditure Methodology Paper page 10
"A number of adjustments are then made to this spending estimate.
1. Public sector debt interestIn 2015-16, UK public sector debt interest includes expenditure associated with English Housing Associations. Scotland is allocated none of this expenditure in GERS. This decreases the Scottish share of HM Treasury current expenditure on public sector debt interest.
2. Network RailIn 2015-16, spending by Network Rail has been moved from an accounting adjustment into TES. The shares from GERS 2014-15 will not reflect this spending. Spending associated with Network Rail has therefore been added into Scottish spend estimates. This increases the Scottish share of Department for Transport current and capital spending on transport.
3. Social protection expenditureScottish social protection expenditure by the Department for Work and Pensions and HMRC is estimated directly, rather than apportioning shares of the UK total. Spending by the Department for Work and Pensions is estimated based on data from the tabulation tool and UK spending data. Spending by HMRC is based on HMRC spending data and HMRC geographical award statistics."
6.


7. GERS page 2.
"The ONS reclassified English Housing Associations (HAs) into the public sector on 30 October 2015. In 2015-16, this increased UK public sector revenue by £6.9 billion and UK expenditure by £10.8 billion, resulting in a £3.9 billion increase in the UK net fiscal deficit. A similar impact is seen in earlier years. Scotland is apportioned none of this additional revenue or expenditure in GERS. The ONS have not yet announced a decision on the classification of Scottish Housing Associations."

8. I confess to being flummoxed by the fact that the non-identifiable cost totals in table A.1 are not explained by summing the figures given by spend category - see figures highlighted amber


10 comments:

Unknown said...

Thank you once again Kevin. I'm tired of the argument that GERS tells us nothing about the finances of an Independent Scotland. What is inescapable is what GERS told us about our deficit on the day we were asked to vote and on the day we were told we would become Independent. Approximately £14bn on the 18/9/14 and £15bn 18 months later. We were asked to follow leaders with no plan. Let's face it, there is no credible even medium-term plan for whomever happened to be leading Scotland.

As a graphic reminder to the Independence fantasists I'd like to see a Scotland's "deficit ticker" published on the net on an appropriate website. Just like the one on Tines Square.

soccer doc said...

Kevin, there are other parts of this that I have reservations about, but do you really believe that "The CRA headlines are based on identifiable expenditure only - but given the vast majority of non-identifiable spend is allocated in GERS on a per capita basis, it seems unlikely that this would explain the difference" is really justifiable?
You say in your footnote that "The three largest non-identifiable spend areas are public sector debt interest, defence and international services", but that is not all there is to it, is it? How about the difference in spending on administration? George Rosie went into this at some length more than 25 years ago in "Scotching the Myth" where he shows that not only is there a disproportionate number of civil servants in and around London and the south east, but they tend to be the more highly paid civil servants.
Now it can be pointed out here that there is a functional need for things to be organized in this way. After all the Minister/ Secy of State does not really want his Permanent Secy in Newcastle or Cardiff or even Edinburgh, does he. But this does not alter the fact that such spending is great for one part of the country but does little for elsewhere.
Indeed GERS includes a cost attributed to Scotland "for services delivered to Scotland". But how many of these services would deliver the kind of outputs that Scotland values or wants? These will be outputs determined by a Conservative govt which we clearly did not vote for. Its a bit like when I go out with friends. I don't drink (nothing religious or moral, just never acquired the taste) and we have a kitty to which we all contribute equally. I know the cost of mineral water in some places can be pretty stupid, but its seldom as much as the booze. But, without the consideration of my friends who, recognising that I wont be drinking (so being able to drive some of them home!)look for a smaller contribution from me. Otherwise I would be paying for booze that I wouldn't be drinking because I don't like it. In the same way Scotland is paying for services that we would not necessarily want (or not in that form - so something to drink, not necessarily alcoholic to take it back to the analogy).
Moreover, does it not strike you as a bit odd that almost every single penny spent in Scotland can be identified, yet Rosie was able to show that the bulk of the unidentified spending was, if you drilled down, spent in/around the south east? At the time of his research some 35% of civil servants were in the south east.
In some ways this argument was replicated by Margaret Cuthbert in her "The UK Economy: a strong and secure UK?" where she argues there is "a vicious circle in which London sucks in people and investment and holds back growth elsewhere, which in turn makes London more attractive, feeding the problem".
So statically (the Rosie argument) and dynamically (Margaret Cuthbert's argument) I think your assumption that "The CRA headlines are based on identifiable expenditure only - but given the vast majority of non-identifiable spend is allocated in GERS on a per capita basis, it seems unlikely that this would explain the difference" is questionable. Highly questionable.

Kevin Hague said...

Soccer Doc

Look at note 6 above. The figures which (in GERS, according to the published methodology) are allocated purely on a population basis account for 99% of [outside UK + non-identifiable] spend - only three categories are allocated on any other basis (a mix of Pop and GVA) and these have very small [outside UK + non-identifiable] spend as you can see from note 8.

Given my point was very specifically about explaining the difference in per capita spend reported in the CRA figures vs those reported in GERS I am extremely confident to say that the allocation of these [outside UK + non-identifiable] figures in GERS is not what explains that.

The rest of your comment is arguing that GERS is wrong to allocate these costs on a per capita basis. That's off topic for this post, but these shared services are administered by a shared UK infrastructure - allocating on per capita basis therefore seems fair to me. All that really matters of course is what you conclude from the analysis in GERS - if you want to make an argument that an independent Scotland could be administered for a lower cost than our per capita share of UK administration costs, you're welcome to. But the challenge of replicating all shared UK administration functions and systems for a lower cost that our per capita UK share would be - how can i put this - non-trivial?



Anonymous said...

Soccer Doc
Just to back Kevin up here. I know the defence sector well. Scotland would have to create (from scratch) a Ministry of Defence (with knowledgeable civil servants), a procurement organisation, a recruitment set up, training, maintenance and repair, communications, munitions storage, an infrastructure support organisation, intelligence and security, testing and quality assurance, a military aviation authority and so on. Unless you are going for a very simple and ineffective territorial defence type set up with 3 or 3 static infantry battalions, half a dozen small offshore patrol vessels, and a handful of transport aircraft (basically a third world defence force), you rapidly find you are duplicating a host of existing UK organisations.

None of this will be cheap or quick and many of these functions simply do not currently exist in Scotland and will have to be created from scratch, assuming you can find individuals with the necessary expertise who want to work in a tiny structure with few promotion prospects. Many will drift south of the border and join the much larger UK MOD armed services, which will simply attract all your talent. This is a problem the Irish have faced for decades. irish personnel join the British forces because they have access to equipment and training that goes way outside any available back home. What is left is very much to second or third XI in terms of quality.

Even setting up a basic and small diplomatic service will not be easy unless you are happy to allow most of the world to intercept your communications, break your codes and infiltrate your security. But then Scotland may not have much worth knowing I suppose. You can bet though that the Five Eyes members like the UK and USA will have no interest in inviting Scotland into the club or involving Edinburgh in diplomatic initiatives. No doubt, however, that the French and Germans will be happy to take an inexperienced new state under their wings and shape it to their own complete satisfaction.

soccer doc said...

Kevin, as usual, you misstate the point. My point is not how these common services are charged out - population share is arguable for sure. But that is not my point and its not George Rosie's either. The point is not what is spent, nor is it how the costs are shared out, but the distribution of spend.
I would have thought though that I made that point perfectly clearly with even just this part of my post
"Now it can be pointed out here that there is a functional need for things to be organized in this way. After all the Minister/ Secy of State does not really want his Permanent Secy in Newcastle or Cardiff or even Edinburgh, does he. But this does not alter the fact that such spending is great for one part of the country but does little for elsewhere."
The charge - how the costs are allocated - is not what is at issue. Its where the money is spent, and you simply avoid that.I dont know if you have read Margaret Cuthbert's paper, but I would really recommend it to you. In some respects the problem with the UK is not that the English are nefarious and that there is a plot to do Scotland down - its simply how the UK works that is the problem.
As for whether Scotland could be administered for a lower cost, the point was made by Dunleavy that Scotland's administrative needs are more homogeneous and less complex. One only has to look at the struggle that IDS had with Universal Credit, chuck in the horror of the UK tax system to see that the UK is well into diseconomies of scale. If Scotland can get this back to where diseconomies kick in then we will have lower costs. But, one thing we should avoid is copying the UK, for we know where that takes us.

soccer doc said...

Anon, thanks for your post. But in the same ceteris paribus way that Kevin approaches a good deal of this, you, it seems to me, do the same. I dont think it will shock anyone on here to learn I support independence, but if all we do with our independence is to recreate the UK on a smaller scale then there is no point to it at all.
Applying that to the defence sector, I take all your points, but I think you need to appreciate that, I hope, the Scottish armed forces will be about the defence of Scotland and not about being able to go to distant parts of the world as America's bag carrier. Our needs will therefore differ from the needs of the UK. I think your "simple" force is pretty close to what we should be aiming for. I am not sure what you mean by ineffective though? If Norway, for instance, was being attacked by Russia then, to quote Sir Alex, its squeaky bum time all round If it all kicks off and we start lobbing nuclear weapons at each other then it gives me little comfort to think that we have Trident and we can kill our share of Russians.
As for Ireland, their armed forces are known as "the Defence Forces". According to their Wiki entry their duties are
* Preparation for the defence of the state against armed attack.
* Assisting the police force, the Garda Síochána, including the protection of the internal security of the state.
* Peacekeeping, crisis management and humanitarian relief operations in support of the United Nations.
* Policing the fisheries, in accordance with the state's obligations under European Union agreements.
* Miscellaneous civil contingency duties requested by the Government
I could live with that in an independent Scotland.
As for those who will "drift south" , for specialists - especially in areas the Scottish Defence Forces do not intend to replicate - this is pretty much inevitable, but I think it is very likely the UK (with or without Scotland) will continue its present defence policies. Fairly clearly the difference between you and I is not just that I would pursue independence and you would not, but that I - and a good many others - envision a different Scotland from the UK in terms of defence (and a good many other things).
As for diplomacy, I recall the issue of who would get the Security Council seat in the event of Scottish independence. I earnestly hope that no Scottish government is ever induced to even contemplate getting into that nest of snakes. A quiet life as a small social democracy in N/W Europe will do me fine. If we have little worth knowing, so much the better - lets leave these games to others. Nor would I imagine that Scotland is any more likely to get invited to "the Five Eyes" or any similar organization, any more than say Denmark or Norway, but that lack does not seem to have been all that much of a problem for them?

Kevin Hague said...

Soccer Doc - I don't misstate your point - you asked:

"do you really believe that "The CRA headlines are based on identifiable expenditure only - but given the vast majority of non-identifiable spend is allocated in GERS on a per capita basis, it seems unlikely that this would explain the difference" is really justifiable?"

I answered that I clearly do - you maybe need to re-read the blog again and pay attention to the point I was making which is simply analytically correct.

You want to make another point about *where* the money is spent. Feel free, fill you boots - but if you quote Dunleavy to support your argument you are in trouble - see my review of his *ahem* analysis here (and note how time proved me right): Dunleavy and the Costs of Independence.

If you think your point about where the non-identifiable costs are spent makes a material difference to the £9bn deficit gap between Scotland and the UK, do please share your analysis - I'd love to see it.

Drew said...

'The extent to which this relative high cost is a structural issue (remote and island communities, demographics that mean a relatively high dependency ratio, areas of chronic deprivation etc.) or are a result of a Scottish Government that's keen to spend the Barnett Formula money on freebies that win votes (no tuition fees, free prescriptions, scrapped tolls etc.) is a debate that will run and run.'

The 1st of the problem is Scotland's over reliance on defence, both as an employer and economic stimulant for local economies because it has become highly politicised by both Unionists and Nationalists alike.

The UK Government knows the armed forces have always been a source of local pride in Scotland, particularly the Highlands while areas of high unemployment and lack of job opportunities in rural communities have been an ideal recruiting ground through the centuries.

Even the SNP have adopted the ridiculous stance that they want to maintain the current defence estate and spending in Scotland.

From defence contracts on the Clyde, the Naval bases at Faslane and Coulport to remote communities in Moray, defence was a key argument for Unionists during the independence referendum.

The MOD is one of the biggest landowners in Scotland and owns land equivalent to the size some 8 times the area of Glasgow.

At least the UK Government is 'smelling the coffee' as it were and reducing the defence estates in Scotland by 20% over the 2 decades or so.

For a country of Scotland's size to spend £3-4 billion annually on defence is clearly unsustainable.

The 2nd part of the problem is generally as a nation we smoke, drink and now eating ourselves to death.

Much of the crime and violence is caused by alcohol-related incidents while jailing people for drug addiction is far more expensive but a voter friendly option rather than sending them to treatment.

Unless we have the tools to raise the price of alcohol and cigarettes through taxation and change the failed Misuse of Drugs Act, we are stuck as we are.


Anonymous said...

Looking at the data regarding GERS on Scottish government website.

It clearly states that the Scottish version of Country and Regional analysis SCRA (which to me looks like UK CRA and per capita non identified expenditure) as extracted from UK Country and Regional analysis , has the following expenditure totals for EU (ex vat)

2014-15 SCRA tab
£724m SCRA (Scottish version of Country and Regional analysis)

after adjustments
£485m GERS tab

How did they work that figure down by -£250m!

very easy

EU receipts
-£365m SCRA (which works out at 8.28% of the UK total)
adjusted to
-£606m GERS (which works out at 13.76% of the UK total)

my bet for the latest version 2015-16 of GERS is the exclusion of Scottish water in the expenditure figures of £571m.



Anonymous said...

Sorry I forgot to include my source for SCRA (Scottish version of Country and Regional analysis)for those not familiar with scot.gov website.

http://www.gov.scot/Resource/0049/00495354.xlsx

Interesting point (to me)

£68.846bn expenditure 2014-15 SCRA (column)
adjusted to
£67.164bn expenditure 2014-15 GERS (column)