GERS: Location, location, location…

David Comerford, Fraser of Allander Institute

With the new GERS figures released (here), the debate around the usefulness for, and relevance of, this publication to the constitutional debate has kicked off again.

On the back of this, we have been asked about the significance of the ‘for’ vs. ‘in’ methodology GERS uses to calculate public spending.

Some have argued that Scotland is short-changed – be that via spending on Defence or Whitehall civil servants – and that too much of what is allocated to Scotland in GERS is actually spent elsewhere in the UK.

The purpose of this blog is not to address that particular argument directly but instead to simply ask – if true – what could the potential impacts be on the net Scottish fiscal position?

A quick recap – what is the difference between ‘in’ vs. ‘for

GERS attempts to calculate the public sector expenditure “incurred for the benefit of residents of Scotland”, and the public sector revenue generated “where a financial burden is imposed on residents and enterprises in Scotland”.

On this basis, GERS shows a net fiscal balance for Scotland of -£13.3bn or -8.3% of GDP. This compares with a net balance for the UK of -£46.2bn, or -2.4% of GDP.

GERS allocates spending to Scotland on the basis of spending ‘for the people of Scotland’ rather than on the basis of spending ‘in Scotland’.

So what’s the difference?

The ‘for’ methodology apportions spending to Scotland if the benefit is thought to accrue to residents living in Scotland.

The ‘in’ methodology would apportion spending – no matter who benefited – on the basis of whether or not the money was actually spent somewhere in Scotland.

For most expenditures, spending ‘for’ or ‘in’ Scotland is the same. For example, the vast majority of health expenditure by NHS Scotland occurs in Scotland and is for patients resident in Scotland.

But it does mean that under the ‘for’ approach, GERS assigns Scotland a share of some expenditure which does not take place ‘in’ Scotland. At the same time, the Scottish fiscal position is not burdened by the full cost of spending ‘in’ Scotland but which benefits the UK as a whole.

As an example, the new Queen Elizabeth aircraft carrier was built ‘in’ Scotland but its home port will be ‘in’ the South of England. In GERS, Scotland gets a per capita share of the cost of both building the ship as well as the running costs.

So what about the numbers?

Table 3.8 in GERS shows that total expenditure for Scotland in 2016-17 was £71.2 billion.

Of that, £42.1 billion (59.1% of spend) is estimated to have been spent by the Scottish Government, Scottish local authorities, or Scottish public corporations (such as Scottish Water). Aside from a relatively small amount of overseas activity – e.g. in Malawi or Brussels – this spending is not only ‘for’ the people of Scotland but takes place ‘in’ Scotland.

Social security spending by the UK Government amounts to a further £18.1 billion (25.4% of spend) Again, give or take a share of DWP administration, this is spending on people’s benefits – and so is spending ‘for’ the people of Scotland and spending ‘in’ Scotland.

What about the rest?

Therefore around 90% expenditure can be identified as ‘in’ Scotland as well as ‘for’ Scotland. And even then, this is likely to be a lower estimate, as GERS allocates Scotland a population share of spending on reserved functions which occurs in Scotland.

That leaves:

  • Scotland’s population share of UK foreign aid is around £600 million. This doesn’t occur in Scotland or in the UK. Additionally, spending on embassies, and (somewhat controversially!) contributions to the EU, does not occur in Scotland or in the UK.
  • Spending on debt interest and repayments is £3.2 billion. It is hard to say that this spending has any location. While the magnitude of these payments is an issue in the constitutional debate, it is not the subject of this blog post.
  • Other bits and pieces, including spending on embassies, counter-terrorism, and HMRC administration, make up the final £1.5 billion or so.

What will different assumptions on how much UK-wide spending Scotland is allocated have on the results?

Let’s take the £3.1 billion on Defence. Assume that the GERS methodology was changed to only allocate that spending to Scotland that occurred ‘in’ Scotland. As an illustration, let’s assume that there is Defence expenditure ‘in’ Scotland of around £1.5 billion. In that scenario, the current GERS expenditure would be reduced by £1.6 billion, and Scotland’s net fiscal balance would fall to -7.3%.

Let’s go further and assume that as well as Defence expenditure ‘in’ Scotland of £1.5 billion, there is no expenditure at all on debt interest. This would reduce the current GERS expenditure by £4.8 billion, and Scotland’s net fiscal balance would fall to -5.3%.

And so on.

So changing the spending lines does have an impact on the results, but given that the vast majority of spending occurs ‘in’ and ‘for’, any realistic assumption doesn’t really alter the conclusions.

Alternatively, what will different assumptions on where the money is spent have on the Scottish results?

Suppose that there are 10 senior civil servants employed in London, who work on behalf of the whole of the UK (so that one of them could notionally be said to work on behalf of Scotland).

In GERS approx. 1/10 of the cost of these civil servants would be assigned to Scottish expenditure, but none of their income tax or national insurance would be credited to Scottish revenues.

If one of these civil servants were to be moved to Scotland, then the GERS deficit would be lower not just by the amount of tax generated by this civil servant’s salary but also by the knock-on effects their spending had on the wider economy.

Of course it works both ways. A senior civil servant working for the UK in Scotland under the GERS methodology will have their salary cost shared across the UK as a whole, but their tax revenues (and the boost to the economy) allocated to Scotland.

Is this feature of the GERS methodology a driver of the higher deficit in Scotland compared with the UK as a whole?

To check we can make use of the Scottish input-output model. The IO model captures the structure of the Scottish economy, and thereby enables us to assess the impact of potential multiplier effects from more public spending in Scotland both through more Scottish procurement and through the impact of higher public sector wages being spent within the Scottish economy.

To test this hypothesis, let’s take an extreme example.

We’ve shown that at least 90% of the total Scottish GERS expenditure ‘for’ Scotland, is spent ‘in’ Scotland.

For illustrative purposes, let’s boost the Scottish Exchequer by the full tax and economic multipliers as if the remaining 10% (i.e. £7.1 billion) had been injected into the Scottish economy. (NB: to be clear this means that we’re effectively assuming none of this 10% of non-identifiable spending is currently spent in Scotland: on defence, on HMRC admin etc., not to mention the 43,000 people employed by UK departments etc. Further, this illustration assumes that money currently spent on debt repayments or on foreign aid or embassies is instead able to be spent on public services in Scotland.)

What’s the impact on the net fiscal position?

We add the £7.1 billion to government demand for the output of either (1) the “Public administration & defence” sector, or (2) the “Education” sector, in the Scottish Input-Output table. This boosts Scottish GDP by (1) 5.5%, or (2) 7.5%, once all the multiplier effects are included (the difference reflects the higher multipliers associated with Education spending than with spending on Public administration & defence). If tax revenues also grew by the same amount, the combination would reduce the GERS deficit from 8.3% currently to (1) 6.0%, or (2) 5.2%, of Scottish GDP.


Changing assumptions about how much spending is allocated ‘for‘ Scotland or spent ‘in’ Scotland in GERS will change the net fiscal position. But any revisions are relatively small.

The idea is an interesting nuance when linking the GERS figures to the constitutional debate. But the “on behalf of” vs “in the territory of” issue does not explain the Scottish fiscal position vis-à-vis that of the UK as a whole. Of course, it is possible to close this gap by explicitly reducing certain expenditures or by assuming higher tax revenues – either through increased rates or faster growth. Others will argue though, that in the context of independence, there may be additional costs. The debates will no doubt continue.


David Comerford

Post-doctoral researcher in economics