Wednesday, 5 February 2020

Deficits, Deficit Gaps and Fiscal Transfers

To understand what is going on when we talk about implied fiscal transfers between different parts of the UK (as discussed here), it's perhaps easiest to think of what happens when we split the bill in a restaurant.

To know whether or not we benefit from splitting the bill, we only need to know two things:
  1. How much of the bill are we responsible for creating?
  2. How much of the bill do we actually have to pay?
If the first figure is greater than the second, we benefit from splitting the bill (we receive an implied transfer from the others we're splitting the bill with).

In the context of the debate around Scottish independence, the first of these questions is answered by the Scottish Government's own GERS report. This tells us, based on a series of explicit assumptions, how much of the UK's deficit (the bill) Scotland is responsbile for.

The second question is more contentious, as there are no "official figures" as to how responsibility for the national debt (the cumulation of annual deficits) is shared - so how much of the bill does Scotland have to pay?

Fortunately there is broad consensus around the view that the UK's debt should (or at the very least reasonably could) be shared on a population basis.
  1. The GERS figures include a population share of the interest charge generated by the UK's debt - given that debt is merely the accumulaton of the UK's deficits over time, that is effectively a population share of the UK's (cumulative) deficit
  2. The Independence White Paper in 2014 stated "Scotland and the rest of the UK will agree a share of the national debt. This could be by reference to the historical contribution made to the UK’s public finances by Scotland. An alternative approach would be to use our population share."
  3. The SNP's own Sustainable Growth Commission danced around this question, but eventually assumed a population share of UK debt interest within their proposed "solidarity payment"
So it's really pretty simple: the difference between the share of the UK's deficit Scotland is responsible for creating (see GERS) and the share of the UK's deficit Scotland pays for (assume population share) is the implied fiscal transfer.

As we'll come on to see, you can make different assumptions about the share of the UK's deficit Scotland will ultimately have to pay, and conclude a different figure for the implied fiscal transfer.


At the risk of labouring the restaurant analogy, let's run through an example with some illustrative figures to help us explain the differences between three terms that often get confused: the deficit, the deficit gap and the effective fiscal transfer:
  • Group A: 18 people go for a meal, the bill comes to £1,800 so they have spent £100/head
  • Group B: 2 different people go for a meal, their bill is £300 so they have spent £150/head
If Group A and Group B decide to get together and split the bill (to "pool & share the deficit"), what happens? The total bill would be £2,100 which split equally between 20 people would be £105/head. Here's a simple summary:

The spending gap between the groups is £50/head (Group B spent £50/head more than Group A), but the benefit of pooling and sharing - the effective transfer Group B receives - is £45/head2.

The total transfer from Group A to Group B is £45x2 = £90

Now let's replace the figures in our analogy with the fiscal reality (per GERS 2018-19) - the "bill" is the deficit, Group B is Scotland and Group A is the rest of the UK.

So Scotland's GERS deficit is £12.6bn, the deficit gap is £11.6bn and the effective fiscal transfer to Scotland is £10.7bn. 

A huge amount of confusion is caused by people failing to understand the conceptual the differences between these figures - if you've followed what's going on up to here, give yourself a pat on the back.

So armed with this understanding, let's take a look at the most common mistake made when people debate the "£10bn fiscal transfer". To illustrate, let me use the following screen-capture which (incredibly) is taken from Stuart Campbell's own "Wings Over Scotland" blog:

If you've been following this blog post so far, you will realise who the twit is in the exchange above (hint: it's not Paul). To walk through this carefully, per the figures above:
  • Scotland's deficit is £12.6bn
  • We assume Scotland bears a population share of the UK's deficit - so in this year Scotland takes on an additional "loan" of just £1.9bn
  • The difference of £10.7bn is the effective fiscal transfer Scotland receives - it's the amount over and above the "loan" Scotland takes on
Fun Fact: this means that those who argue Scotland should assume less than our population share of the UK's debt are - whether they realise it or not - arguing that the effective fiscal transfer in Scotland's favour is in fact larger than £10.7bn.


A common reaction to these figures is "how can Scotland's 8% of the UK population possibly be responsible for a third of the UK's deficit - that seems unbelievable". This is what is technically known as an "argument from incredulity" and is perhaps best summarised by this quote from Professor Richard Murphy:
"I have been continually bemused by the fact that GERS says that Scotland runs a deficit so  much larger in proportionate terms than that for the UK as a whole."
Here our restaurant bill analogy falls short, because what we're dealing with when we're sharing the deficit is not how much we've spent but the net effect of how much revenue we've generated less the amount we've spent. I've explained the dynamics involved here in this brief video (with apologies for my exasperated tone and the figures being a year out-of-date)

Another way to help understand this point is to look at fiscal transfers across the UK (including the English regions) as this blog has recently done here. There is nothing surprising or hard to fathom going on here - it's just simple fiscal arithemetic.

As I've pointed out before: it's not hard to imagine a situation where Scotland runs a small deficit while the the UK overall is in fiscal balance - in that scenario Scotland would be responsible for an infinite (or more accurately: a "divide by zero error") share of the UK's deficit. It's just maths.


When we use the GERS figures to scale the effective fiscal transfer, we have to recognise that these are only pro-forma figures, they represent what Scotland's stand-alone defict would be if we kept generating revenues and incurring spending as shown in GERS.

In case it's not already dead, let me flog the restaurant analogy one more time: "if we weren't sharing the bill, maybe we wouldn't have tipped the waiter 15% and perhaps we wouldn't have ordered the bottled water for the table."

This is a fair point. Even before we consider the likely economic shock impacts on revenue or spending that separation from the UK would cause (see Brexit), the scale of deficit that the GERS figures reveal means that current levels of spending would be unsustainable for a newly independent Scotland, particularly if trying to launch a new currency.

It's true that some of that spending in GERS is costs allocated from the rest of the UK on a simple population basis (defence, debt interest and international aid being the vast majority of these), so any case for independence needs to start by working out what an independent Scotland would replace these costs with. For reference: relative to that £10.7bn fiscal transfer, the notoriously optimistic White Paper on independence assumed a net saving of £0.6bn.

What typically happens at this point is that some of the more blindly-committed supporters of independence start suggesting that the GERS figures are all made up anyway as part of some vast conspiracy by which Westminster has managed to get the Scottish Government's own economists to pull the wool over the eyes of the SNP (and their Sustainable Growth Commission, their Fiscal Commission Working Group, the IFS, Fraser of Allandar, NIESR, UK Statistics Authority, etc. etc.).

This is of course a ridiculous position to adopt (which, to be fair, is why only those flakier members of the independence movement attempt to adopt it). Alex Salmond was certainly very clear about what the GERS figures told us when he thought he could spin them in his favour:

Salmond is the man who once proudly boasted of his ability to put “a gloss on statistics or any economic figure” to build a political case, and he certainly did his best to do that with the 2010-11 GERS figures. He made the highly dubious claim that they showed an independent Scotland could have been spending £2.7bn more and therefore should have been running an even higher deficit than that shown in GERS!

Still: desite the fact that he used a different method for "splitting the bill" (based on a GDP share not a population share), he was recognising the principle of the fiscal transfer3.

Unfortunately for independence supporters, taking the logic Salmond applied to the 2011-12 figures and applying them to the 2018-19 figures produces a massive fiscal transfer now in Scotland's favour - so by his own logic, an independent Scotland should now be spending £10bn less4.

At this point, most of those arguing for independence ignore how wedded they used to be to the figures and return to straight-froward "GERS denial" - fortunately this blog has already comprehensively dealt with those denials here > GERS Deniers.

Ah but wait: what about "this is just a snapshot"?

OK, well we can do this analysis over time and plot the size of the deficit gap5 for the last 21 years:

You can see why Alex Salmond was so excited about the 2008/09 to 2010/11 figures6.

The reason for the dramatic reversal and growth in that gap will be familiar to regular readers of Chokkablog - they are most easily summarised by this graph:

  • The gap closed when North Sea revenues boomed, but has grown massively as North Sea revenues have plummeted
  • Scotland has not only continued to spend more per head than the rest of the UK, that spending gap itself has actually grown (thanks to the Barnett Formula and low levels of absolute spending growth7
  • Scotland's onshore revenue performance has declined relative to rUK8
For completeness, we can plot the onshore deficit gap over time (i.e. to see what happens if we strip out North Sea revenue effects from these figures):

Without oil revenues, there would never have been a prime facie economic case for Scottish independence - and the vagaries of the Barnett Formula (plus perhaps the impact of the SNP's tax rises) have led to the scale of the fiscal transfer that Scotland benefits from within the UK actually increasing in recent years.



1. This very carefully worded FoI response is sometimes in debates around the fiscal transfer:
"Official figures for any fiscal transfer are not available.
The reason this information is not available is that such a figure requires a number of assumptions to be made. For example, as the UK as a whole spends more than is raised in revenue, an assumption would need to be made about which parts of the UK borrowing is undertaken for, or which types of public spending are financed by borrowing as opposed to taxation. This information is not available as, for example, some taxes are ringfenced to fund particular services; for example, some national insurance contributions are ring-fenced to fund the NHS. As such, any figure for a fiscal transfer from the rest of the UK to Scotland would rely on a number of assumptions."
this is entirely consistent with what this blog (and others) have always said - to calculate the implied fiscal transfer, we have to make some assumptions. In fact, argue we can calculate and implied fiscal transfer by only making one assumption: that the burden of the UK's deficit (and associated debt) is borne on a population share basis

2. People used to dealing with numbers will have spotted that the transfer = [(1-population share) x the gap] - something easily proved if you care for such things

This matters only insofar as we need to understand that, in the case of Scotland in the UK, the fiscal transfer is 92% of the deficit gap

3. The IFS implicitly use that same assumption when referring to the fiscal tranfer here
The most recent figures (2016–17) imply a budget deficit for Scotland of 8.3% of GDP. Managing this is the UK Government’s responsibility as it is part of the UK’s deficit, which was 2.3% of UK-wide GDP in the same year. Therefore there was a fiscal transfer from the rest of the UK to Scotland of about 6% of Scotland’s GDP (equivalent to around £1,750 per person in Scotland).
Because GDP/Capita is now about the same for Scotland and rUK, allocating the deficit on a per capita basis or per GDP basis makes no material difference - but I would still argue that per capta is the right way to do the analysis as long as GERS uses per capita allocations for all shared UK-wide costs
4. To be completely accurate: if we used his GDP share rather than population share method then the figure would be £9.8bn (rather than the £10.7bn we get using population share) - but the broader point stands

5. Remember: the implied fiscal transfer = [(1-population share) x the gap] = 92% of this figure

6. These are the latest available restated historical figures - when first released the figures showed a significantly more favourable position for Scotland, but later revisions lowered Scotland's apparent fiscal advantage vs rUK - covered in some here: The SNP: Living in the Past

7. A dynamic most easily understood if you imagine a scenario where UK spend (and therefore Scotland's spend) doesn't change, but Scotland's population grows more slowly than rUK's - under that scenario it is inevitable that the gap between Scotland's spend/head and rUK's must increase

8. Due to some combination of historically over-estimating the number of top-rate tax payers in Scotland and/or the increase in the Scotttish Rate of Income Tax causing some of those tax payers to redomicile

Sunday, 2 February 2020

Pooling and Sharing: The English Regions

In a few weeks time I'll be chairing a conference in Newcastle - These Islands: Our Past, Present & Future

The conference will feature an impressive array of speakers and panelists including: Douglas Alexander, Philippe Auclair, Gordon Brown, Andy Burnham, Frances Coppola, Sir John Curtice, Simon Evans, Sophia Gaston, Ayesha Hazarika, Gerry Hassan, Fiona Hill, Henry Hill, Colin Kidd, Carwyn Jones, David Lidington, Ian Murray, Baronesss Quin, Mark Reckless, Willie Rennie, Lord Salisbury and many, many more. If you're interested in coming along, you can find more details and ticket booking information here > eventbrite page.

Needless to say: just as Bob Geldof wasn't going to go to all the trouble of organising Live Aid and not get on stage to perform with the Boomtown Rats, so I will not be passing up the opportunity to put some graphs in front of this captive audience1 .

We're holding the conference in Newcastle to highlight the importance of the English regions in any debate about the future of the UK. To this end I've been doing fresh analysis on the fiscal economics of the English regions and - because I won't have time to present detailed analysis at the conference - I though I'd quickly blog about it here.

Chokkablog regulars will be familiar with the concept of the implied fiscal transfer, but to recap: if a devolved nation runs a deficit per head (aka "per capita deficit") higher than the UK average, then that nation is benefitting from an implied fiscal transfer from the rest of the UK2.

The same principle can be applied to the English regions. Fortunately the data now exists to allow us to calculate and understand these regional fiscal transfers just as the (notorious?) Government Expenditure and Revenue Scotland (aka GERS) figures do for Scotland3.

The source data is Country and Region Public Sector Finances analysis as produced by the ONS4. All the figures we use here are those that allocate a "geographical share" of oil & gas revenues (i.e. Scotland gets to keep the oil & gas revenues generated by oil in Scottish waters).

The only other thing we have to remember before diving into this analysis is that the per capita deficit difference to the UK average is made up of two distinct parts: the per capita revenue difference and the per capita spending difference. The former reflects the economic performance of the region in terms of tax revenue generation, the latter reflects the cost of delivering public services to that region5.

So let's look first at per capita revenue generation differences by region:

Remember that what we're seeing here is how well these regions generate tax revenue versus the UK average. There has been a lot of talk in recent weeks of "levelling up" - that's basically about getting these bars to shrink back towards zero, so the size of the red bars is a decent guide to which areas are in greatest need of "levelling up".

It should come as no surprise that London and the South East are the areas which "out-perform" and - unless you've been deceived by the SNP's grievance rehetoric - it should also be no surprise to see Scotland (like the East of England) performing as per the UK average and significantly out-performing Wales, Northern Ireland and all other English regions.

So what's Scotland's problem?

Well let's look at per capita spending per region4:

Here's where we see the areas that enjoy (or require) higher spending per capita: Northern Ireland and Scotland most significantly, with London, Wales and the North East as the other "relatively high spend" areas. The extent to which this is based on greater need (e.g. to deliver equivalent services in areas of lower population density and/or with remote/island communities and/or to reflecting higher social costs driven by demographic factors and/or due to areas of endemic poverty) or greater investment (for better services than the UK average or to stimulate economic development) is the subject of some debate.

If we combine the per capita revenue difference with the per capita spending difference, we get to the per capita fiscal balance difference (and hence the implied fiscal transfer):

It's quite a striking picture isn't it? The Devolved Administrations in Wales and Northern Ireland receive far greater per capita fiscal transfers than Scotland, as does the North East.  London, the South East and the East of England are responsible for generating fiscal transfers that go to the rest of the UK.

It's perhaps helpful to summarise all of this data on one exhibit:

What this shows us is the extent to which the fiscal transfer to Scotland (caused by the much debated higher notional Scottish deficit) is a function of higher spending, not lower revenue generation (i.e. not "weaker economic performance"). This contrasts dramatically with Wales the Midlands and the North of England, where relatively poor fiscal performance is explained by weaker revenue generation (i.e. "weaker economic performance") far more than by any spending differences.

It's clear that any debate about the future of the UK has to grapple with these two related questons
  • How are resources most fairly and efficiently distributed between and administered in the devolved nations and English regions?
  • What practical steps can be taken to "level up" economic performance across the UK?

For those who like to see the figures behind the pictures:

For those who wonder about regional difference within Scotland, I offer the following GDP/capita chart - suffice to say we could expect there to be similar fiscal transfers happening within Scotland, and the variance of economic performance within Scotland appears not dissimilar to the variance across the UK (or indeed in continental Europe):


1. Yes, I'm aware how ridiculous it is to imply that this conference is some sort of constitutional Live Aid

2. We share the burden of the UK's deficit UK-wide - both in these analyses (i.e. the cost of the total UK debt is allocated to the regions and devolved administrations on a per capita basis) and as widely assumed and accepted in the case of inherited liabilities (i.e. were Scotland to separate from the UK, it would inherit a population share of he UK's debt, as accepted by the Independence White Paper and the SNP's more recent Sustainable Growth Commission)

3. This figures differ from GERS, but not materially so at the deficit per head level - comparing 2018-19 ONS and GERS, it looks like there is a different approach to what is taken as revenue vs what is netted off against cost (but I'm guessing here) - all that really matters is that the figures used for the exhibits on this blog post are all compiled on a comparable basis
  • Spend/head: ONS = £14.5k; GERS = £13.9k
  • Revenue/head: ONS = £12.0k; GERS = £11.5k
  • Deficit per head: ONS = £2.5k; GERS = £2.6k
4. It is worth noting that these are not qualified as National Statistics, but rather Experimental Statistics

5. It's worth noting that those costs which are shared on a population basis (mainly debt interest, defence and international aid) have no impact on this analysis - there is by definition zero difference between per capita costs allocated on a per capita basis!