WHAT IS THE BARNETT SQUEEZE?

What exactly is the Barnett ‘squeeze’?

[This briefing builds upon and should be read in conjunction with “How Does the Barnett Formula Actually Work?”]

The Barnett Formula was only ever intended to be a short-term measure, put in place to reduce the need for - what by some accounts had become increasingly fraught - annual budget negotiations between the Treasury and the Scotland Office. It was, however, created with the idea of ‘convergence’ in mind; meaning that public spending per person would tend towards being the same in England, Scotland, Wales or Northern Ireland - a concept commonly referred to as the ‘Barnett Squeeze’.

This Briefing will explain how the Barnett Formula has worked in practice and explain why the “Barnett Squeeze” many people expected has not in fact occurred.

An issue since 1888

The reasons for differing public spending per person today can be traced back to 1888 when the then Chancellor of the Exchequer, George Goschen, devised a formula based on a fixed ratio in order to fund the newly established Scottish Office and public services in Ireland. He set a fixed formula of 80% for England & Wales, 11.0% for Scotland and 9.0% for Ireland.

The problem of the fixed ratio was that it did not allow for the long-term trend of Scotland and Ireland’s populations declining relative to England & Wales. Initially, the formula was not equitable for Scotland because Scotland’s population was slightly greater than 11% but, over time, it became relatively favourable.

This trend was further reinforced after the creation of the Irish Free State in 1922. Scottish politicians and civil servants successfully campaigned for Scotland to receive its ‘full Goschen share’, calculated as 11/80ths of England and Wales’s amount. The 11/80ths split became a point of principle for Scottish politicians and civil servants when arguing for public service spending.

Even after Goschen fell out of use in 1959, through the effective annual negotiations of Scottish Secretaries of State, public spending in Scotland remained higher per person than in England and Wales. By 1976, public spending person in Scotland was a full 22% higher than in England.

Joel Barnett’s dilemma

Following the IMF UK bailout in 1976 and amidst increasing Labour backbench agitation, the Treasury in 1978 needed a strategy for funding the planned new Scottish Assembly without the need for annual negotiations and in a way that would be seen as fair to the rest of the UK. The short-term solution was the Barnett Formula before a ‘needs’ assessment would take over.

The formula was named after the Chief Secretary of the Treasury, Joel Barnett. The Barnett Formula would not use a fixed ratio like Goschen, but rather it would allocate any budget changes (increases or cuts) on a per capita basis.^{1 For more detail on the Barnett Formula, see “How Does the Barnett Formula Actually Work”}

The failure of the devolution referendums in 1978, followed by Margaret Thatcher’s election victory killed the needs assessment model of funding. However, the Barnett Formula alone would cause public spending per person to gradually converge because a larger and larger proportion of the budget would be accounted for by nominal terms budget increases which had been allocated on a per capita basis.

This is not how things worked out.

The best-laid schemes o' mice an' men

Firstly - for unknown reasons - from 1978 to 1992, the formula was applied using out-of-date 1976 population data. This meant the problem of Goschen was replicated. Scotland’s relative population share continued to decline over this period, resulting in Scotland benefiting from an “incorrect” application of the Barnett Formula.

The dynamics of the Barnett Formula are hard to understand, but the effect can be illustrated with a simple model.^{2 For full explanation, see Chokkablog: Calling Time on the Barnett Formula}
 

The model above isolates the impact of relative population growth only, using realistic assumptions around absolute spending growth rates. This shows the impact of relative population dynamics – showing how Wales or NI would fare if they had started with the same per capita spending premium as Scotland.

It is clear that if “true” population data had been used prior to 1992, then there would have been some convergence, but “as applied” Scotland’s relative spend reduction was lower than would otherwise have been expected. The line for Wales up to 1992 shows the Barnett Squeeze working as expected when the relative population share remains static.

What stands out from the model is that even when the “true” Barnett Formula is applied, the rates of convergence differ based on each nations’ different population trends. Moreover, in recent years the actual population and spend growth dynamics in fact cause divergence (that is: causes per person spending to relatively increase in devolved nations).

The reason for these differences is not easy to intuitively understand, but comes from two factors: relative population growth and the rate of spending growth.

Consider the example where there is no nominal spending increase at all. If Scotland’s population share relative to England declines, then its spend per capita must relatively increase as the same amount of spending is applied to relatively fewer people. This effect works counter to the perhaps more intuitively obvious convergence dynamic,^{3 Another small factor is that prior year’s population figures have to be used for the calculation (the figures for the year in which the spending is to be incurred are of course not yet known), so if population is relatively declining there is another small lag effect.} so when the absolute amount of spending growth is low (as in recent years) application of the Barnett Formula actually causes divergence of spend per person (towards areas where population is relatively declining).

The bottom line is that even since the use of correct population figures from 1992, the Barnett Formula itself has in fact not created any significant “Barnett squeeze”. The model used here isolates the “pure” Barnett effect, but we should recognise that in practice other dynamics are in play.

‘Formula bypasses’ (i.e. extra money, from the Treasury) have regularly been negotiated with Scotland and Northern Ireland being more effective in securing these ‘bypasses’ than the Welsh. Indeed, concern that the Barnett formula was not meeting Wales’s ‘needs’ led to the creation of a ‘block grant floor’ in 2015.^{4 Wales’s ‘Block Grant’ will now not fall below 115% of comparable UK government spending per head. BBC News 19th December 2016.}Furthermore, tax raising powers are increasingly being devolved to Scotland which comes with associated Barnett adjustments.

Despite these other factors, there are two simple conclusions we can draw about the Barnett Formula itself and how it’s been applied in practice:

• Scotland has been relatively favoured over the other nations, before and after devolution
   
• With current rates of population and spending growth, there is no ongoing “Barnett Squeeze” effect

Whether further “squeezing” would be fair or desirable given the differing needs of the constituent nations and ongoing effects of devolution is a topic for another day.

NOTES

1. For more detail on the Barnett Formula, see “How Does the Barnett Formula Actually Work”
    
2. For full explanation, see Chokkablog: Calling Time on the Barnett Formula
    
3. Another small factor is that prior year’s population figures have to be used for the calculation (the figures for the year in which the spending is to be incurred are of course not yet known), so if population is relatively declining there is another small lag effect. 
    
4. Wales’s ‘Block Grant’ will now not fall below 115% of comparable UK government spending per head. BBC News 19th December 2016.

Additional Reading

Matthew Keep, ‘The Barnett Formula’, House of Commons Library Briefing Paper, No. 7386

Iain McLean & Alistair McMillan, ‘The Distribution of Public Expenditure across the UK Regions’, Fiscal Studies 24.1 (2003), pp. 45-71.

Iain McLean, ‘The No-men of England: the Geordie revolt that defeated the Scotland and Wales Bill in 1977’, The Political Quarterly 87 (2016), pp. 601-608.

Iain Mclean, Guy Lodge & Katie Schmuecker, ‘Fair shares? Barnett and the politics of public expenditure’, Institute for Public Policy Research