The Barnett Formula - Select Committee on the Barnett Formula Contents


Memorandum by Professor David King (Professor of Public Economics, University of Stirling)

  The Committee's Call for Evidence makes many references to needs assessment. I hope this paper will help the Committee's discussions of those issues. Points 1-5 below relate to the Call for Evidence paragraph 1a, which asks whether the current disparities in public spending between the countries of the UK relate to need.

    1. Currently, spending per head on public services in Scotland is about 20 per cent above that in England, and one cannot say definitively whether this difference is in excess of its relative needs, because there is no definitive way of estimating needs.

    2. However, it is most unlikely that any plausible needs assessment formula would treat Scotland so generously. Over half the relevant public service spending is accounted for by local authorities, and around 2004 I applied to Scottish local authorities the very complicated formulae used by Westminster to estimate the spending needs of English local authorities. Table 1 gives some results: eg spending needs per head on local education services were £505 in Scotland, slightly below the £509 in England.

    3. Table 1 shows that Scotland had higher needs for other service blocks: eg £84 per head as opposed to £75 for personal social services for children; £128 compared with £107 for personal social services for older people; and £44 compared with £40 for highway maintenance.

    4. Overall, the English formulae suggested that Scotland's needs for local authority services were around 6 per cent higher than England's needs.

    5. As noted in point 1, different formulae might give very different results. Holyrood also assesses the spending needs of its local authorities, and it gets very different figures for relative need. Assuming, purely for example, that the English formulae were right, then Figure 1 shows that the Scottish formulae errors in 2003-04 for education ranged from eight per cent in East Ayrshire to +19 per cent in Eilean Siar. Figure 2 shows that the Scottish formulae errors for personal social services in 2004-05 ranged from 23 per cent in Glasgow to +49 per cent in Orkney. Figure 3 shows that the Scottish formulae errors for highway maintenance in 2004-05 ranged from 41 per cent in Inverclyde to +239 per cent in Eilean Siar.

  Points 6-7 below relate to the Call for Evidence paragraph 4m, which considers the appropriate factors to allow for when estimating spending needs.

    6. Needs assessment should certainly allow for high spending needs in areas which need relatively more units of output per head to provide a given service level: eg areas with relatively high numbers of children need relatively more school places.

    7. Some people argue that allowance should also be made for areas which face higher costs per unit of output: eg London has high wage costs and so needs to spend more to provide each school place; and Scotland's sparse population leads to many small primary schools which cost more per pupil. But the case for allowing for these costs can be contested. At present, London receives generous grants to help it pay the high wages of its public sector workers, but these high wages arise because London is a popular place to live, and so has high property prices. Arguably, its high grants just make it a better place to live, and so attract migrants who drive up property prices further; the result is that public sector wages have to rise again, so costs rise still further, leading to yet more grants.

  Points 8-9 below relate to the Call for Evidence paragraph 4o which asks whether any needs assessment formulae should be based on a wide or small range of factors.

    8. The formulae used for local authority spending in England, Scotland and Wales are very complex and allow for many factors, most of them being demographic, social or economic. Spending needs per head probably vary much less per head between the four countries of the UK than between local authorities, so in principle it might be possible to use simpler formula that focused on a few key factors.

    9. Nevertheless, there are many factors that vary sharply even between the countries. Eg Table 1 shows that only 1 per cent of Scotland's schoolchildren come from low achieving ethnic groups, compared with 9 per cent in England; Scotland would get £88 per child for children living in flats, compared with £30 in England; and Scotland would get £88 per older adult to allow for those on low incomes, compared with £52 in England.

    10. Anyone who undertook the needs assessments would face great pressure to allow for many factors like these, if they could be shown to be relevant. Each factor might have only a marginal effect, but each country would have an incentive to seek factors that would give it a marginal advantage.

    11. Also, once any formula was announced, English regions could at once see whether the public spending they enjoyed was above or below that which the formula would allow, and in practice it might prove difficult not to apply the formula to them. But needs will vary greatly between regions within England, especially between London and other regions, and it may well require a complex formula to allow satisfactorily for this.

    12. In short, the prospects of getting wide approval for a simple formula may be slim.

  Points 13-19 below relate to the Call for Evidence paragraph 4p which asks who should carry out any needs-based assessment. I would tentatively suggest the following:

    13. The formula should be devised by an autonomous body such as Australia's Commonwealth Grants Commission. If the central government devised it, there would be sure to be accusations of bias.

    14. The body should update its assessments annually: large periodic assessments would lead to the same adjustment problems that arise with periodic property tax revaluations.

    15. The body would need people with links to each UK country. Ideally, many would have links to more than one country, to lessen the tendency for those from single countries simply to argue their corner. There should also be some overseas representatives.

    16. The body should include people with direct technical experience of needs assessment. It would be useful for technicians from each UK country to pool their experiences.

    17. The body should be willing to consider some specific grants as well as general grants.

    18. The body should try to make provisional announcements well ahead.

    19. Ideally, some members of the body would serve for fairly long periods of time. Currently, it is often hard to find anyone who knows why a particular indicator is used in the local government formulae, because it was already allowed for before the current team members became involved.

  Finally, if the Barnett formula is replaced by a needs assessment exercise, the transition from one to the other must be handled with great care. A major reason for the unpopularity of the poll tax stemmed from the very carelessly designed transitional safety net arrangements.

Table 1

RELATIVE SCOTTISH AND ENGLISH SPENDING NEEDS FOR SOME LOCAL AUTHORITY SERVICES, AS ASSESSED BY THE FORMULAE USED IN ENGLAND


SCOTLAND
ENGLAND

LOCAL AUTHORITY EDUCATION SERVICES, 2003-04
     % of children from low achieving ethnic groups
1%
9%
% of children who depend on an IS/IBJSA claimant
21%
20%
% of children who depend on a WFTC/DPTC claimant
20%
19%
% of children in sparse wards
18%
14%
% of children in supersparse wards
13%
3%
% of 3-15 year olds in high cost groups
4%
4%
Local authority education services, needs per pupil
£3,419
£3,319
% of population who are pupils aged 3-15
16.0%
16.7%
Local authority education services, needs per head
£505
£509
PERSONAL SOCIAL SERVICES FOR CHILDREN 2004-05
Basic sum per child
-£18
-£18
Sum per child for children living in flats
£88
£30
Sum per child for long-term ill children
£63
£60
Sum per child for children who depend on an IS/IBJSA claimant
£93
£92
Sum per child for children living in 1-adult homes
£140
£122
Sum per child for dense population
£7
£16
Sum per child based on factors affecting need for foster care
£13
£15
Sum per child for area cost adjustment
£4
£19
Personal social services for children, needs per child
£337
£391
Children as % of total population
21.5%
22.4%
Personal social services for children, needs per head
£84
£75
PERSONAL SOCIAL SERVICES FOR OLDER PEOPLE 2004-05
Basic sum per older adult
-£288
-£288
Sum per older adult for adults aged 75-84 
£116
£121
Sum per older adult for adults aged 85+
£120
£126
Sum per older adult for renting pensioners
£137
£92
Sum per older adult for those who are long-term ill
£156
£152
Sum per older adult for those who are on IS/IBJSA
£109
£92
Sum per older adult for pensioners living alone
£111
£102
Sum per older adult for those 65+ on AA/DLA
£197
£167
Sum per older adult for those not in couples
£49
£47
Sum per older adult to allow for those on low incomes
£88
£52
Sum per older adult to allow for sparsity
£3
£1
Sum per older adult for area cost adjustment
£9
£29
Personal social services for older people, needs per older adult
£807
£692
Older adults as % of total population
15.8%
15.5%
Personal social services for older people, needs per head
£128
£107
LOCAL AUTHORITYHIGHWAY MAINTENANCE 2004-05
Sum per head for weighted road length
£6
£4
Sum per head for traffic flow
£20
£20
Sum per head for daytime population
£10
£12
Sum per head for winter maintenance
£7
£3
Sum per head for area cost adjustment
£1
£1
Local authority highway maintenance, needs per head
£44
£40


Figure 1

PERCENTAGE ERRORS IN LOCAL AUTHORITY EDUCATION NEEDS IN 2003-04, AS ASSESSED BY THE SCOTTISH NEEDS ASSESSMENT FORMULA, IF RELATIVE NEEDS ARE ACCURATELY MEASURED BY THE ENGLISH FORMULA


  Source: D King, M Pashley and R Ball, 2004, "An English assessment of Scotland's education spending needs", Fiscal Studies 25:4.

Figure 2

PERCENTAGE ERRORS IN LOCAL AUTHORITY PERSONAL SOCIAL SERVICES NEEDS IN 2004-05, AS ASSESSED BY THE SCOTTISH NEEDS ASSESSMENT FORMULA, IF RELATIVE NEEDS ARE ACCURATELY MEASURED BY THE ENGLISH FORMULA.


  Source: D King, M Pashley and R Ball, 2007, "Scotland's social services spending needs: an English view," Government and Policy, 25:6.

Figure 3

PERCENTAGE ERRORS IN LOCAL AUTHORITY HIGHWAY MAINTENANCE NEEDS IN 2004-05, AS ASSESSED BY THE SCOTTISH NEEDS ASSESSMENT FORMULA, IF RELATIVE NEEDS ARE ACCURATELY MEASURED BY THE ENGLISH FORMULA.


  Source: D King, M Pashley and R Ball, 2007, "The environmental spending needs of Scotland's local authorities," Local Government Studies, 33:2.



 
previous page contents next page

House of Lords home page Parliament home page House of Commons home page search page enquiries index

© Parliamentary copyright 2009