The Barnett Formula - Select Committee on the Barnett Formula Contents

Memorandum by Professor Colin Thain



Q1.  What do you understand was the purpose of the Barnett Formula was when it was first introduced, and has its purpose changed over time? Was it designed to reduce tensions arising from disparities in public spending per head of population? Has it succeeded in resolving such tensions?


  1.  It is important to note that like so many developments in British public administration the Barnett formula evolved from rules and precedents going back well before the 1970s. There was a tradition of allocating resources to Scotland through the Goschen formula from 1891 giving Scotland 11/80th of British spending on comparable programmes, this continued long after population shifts meant this was generous to Scotland, and for education the formula remained in use until the end of the 1950s; in Northern Ireland from 1920 to 1938 special bilateral negotiations between the Treasury and the Northern Ireland Ministry of Finance enabled the Province to by-pass normal UK Department-HM Treasury discussions. By 1938 the two departments agreed the principle that Northern Ireland citizens should have parity of service provision with the rest of the UK. By 1942 it was recognised that this meant the province would require extra resources, and in 1950 a de-facto allocative arrangement based on local needs was established.1 Although during the 1960s formula-based systems of allocation lapsed, the principle of using allocations agreed for English departments was the benchmark for discussions between the three territorial departments (the Welsh Office having been established in 1964). Moreover, the principle evolved that Secretaries of State (and the Northern Ireland Government, pre-direct rule) would have discretion in dividing the pot of resources and in practice block grant arrangements were in place. Thus some of the building blocks of what was to become immortalised as the Barnett Formula were already part of the non-statutory rules-of-the-game agreed between HM Treasury and its counterparts in Edinburgh, Belfast and Cardiff.

  2.  The immediate antecedents to the Barnett formula were the debates prior to the plans to devolve legislative and executive power to Scotland and Wales during the last years of the 1974-79 Labour Government. The Treasury led an inter-departmental review to try and establish an "objective" measure for needs for each part of the UK. The Needs Assessment Study2 worked on the assumption that what each part of the UK received in identifiable expenditure should ensure "broadly the same level of public services" and be "allocated according to their relative needs". The report was replete with caveats about the difficulty of deciding on needs and then measuring these objectively rather than subjectively. It concluded that in 1976-77 Scotland should receive per capita spending of 116, Wales 109 and Northern Ireland 131 on the basis of spending in England of 100. The reality was that Scotland received 123, Northern Ireland 136 and Wales 101—suggesting the baseline benefitted Scotland and Northern Ireland but not Wales. The Barnett formula then introduced in 1978 was based on a population balance between England, Scotland and Wales of 85, 10 and 5. Northern Ireland was treated differently with spending based Northern Ireland's share (2.75 per cent) of UK population, rather than a share of GB population used for the rest of the UK. The NIO was also ring-fenced outside the process with separate discussions between the Treasury and NIO on spending related to security, law and order and constitutional matters. A significant element of Barnett was that the territories received increments based on spending increases (or decreases) in comparable English functional programmes. It was therefore based on marginal changes in resources not the overall level. Once the increments were added up a block of resources was then created with a great deal of discretion given to the respective Secretaries of State to allocate down to local programmes. Barnett implied that the Treasury accepted that the levels of spending were roughly acceptable enough to justify a formula based on population rather than some more complex needs formula. Wales lost out in the setting of the baseline based on the historic accumulation of resources up to 1978. Thereafter the intention of the Treasury was that over a considerable time there would be convergence of expenditure toward population share. As spending continued to rise over time, there would be relative gains to England and Wales and losses to Scotland and Northern Ireland (cuts in spending levels would reverse this). The problem for both Scotland and Northern Ireland was that the formula was not revised until 1993-94, and the Conservative Government failed to cut spending thus ensuring the dynamic of the formula would produce convergence with English spending levels.

  3.  There is no evidence that the Barnett formula was anything other than a relatively short-term fix to potential problems created by devolution proposals which in the end failed to pass referenda in Scotland and Wales. It survived long after those initial circumstances passed into political history. That it survived was the result of inertia, the administrative rationality of the process, and some inherent flexibility in the rules of the game. It cannot be stressed enough how important it was to finance officials in Scotland and Wales (and by implication Northern Ireland) that the Treasury (a) left them alone to determine the value and effectiveness of particular programmes and (b) allowed their Secretaries of State to determine the allocation within a block grant between local departments and agencies, without recourse to the Treasury in Whitehall. English departments thus bore the brunt of Treasury scrutiny. This was rational for both the Treasury and the territories. In addition, for most of the period after 1978 the Territories could bid "outside the block" in exceptional circumstances. A careful balancing act was maintained with the territories seeking to extend comparability where appropriate and ask for more than the increments received, but not to press too hard so that the Treasury would then seek to open up territorial spending to greater scrutiny. For the Treasury it was administratively easier to devote scarce staff resources to scrutinising English departments than to delve into all spending in the UK. Thus a typically British solution evolved which just about worked, which maintained the closed world of administrative discretion based on uncodified and non-statutory rules.


Q2.  Do you consider that convergence in per capita levels of public spending on the English level was an intentional feature of the formula, or merely an incidental one? Do you think it is overall a beneficial or harmful feature of the working of the Formula?


  4.  As I note above, convergence was an intentional feature of the formula although the Treasury did not expect the formula to last as long as it has.3 Given that more than 80 per cent of public spending is allocated to England, and that we still have what is in essence a unitary state (albeit with quasi-federal features), I would argue convergence in per capita levels of public spending is on balance a beneficial feature of the formula. Creating greater equity across the UK in terms of public policy outcomes should be the province of other public policies, many of which can be UK based and many determined according to local priorities and the operation of local politics, rather than the budgetary allocation system. The overall budgetary system needs to maintain a large element of convergence to ensure its acceptability to the majority of UK citizens. It is important to note that the current formula applies to DEL (departmental expenditure limits) or what might be called discretionary expenditure, rather than to AME (aggregate managed expenditure), which includes demand-led spending such as social security benefits and tax credits. Thus the total public spending allocation system does include elements outside Barnett which in effect adjust spending to some elements of local needs (higher unemployment in Northern Ireland or higher working tax credits as a result of poorer income distribution).


Q3.  What criteria do you consider to be important in assessing the success or otherwise of the present formula, and of any possible replacement to it? Would a fair or equitable allocation system necessarily be a needs-based system?


  5.  I would place a great deal of emphasis on the importance of the current formula in being administratively easy to apply, and the adjunct rules allowing a great deal of discretion to the territorial administrations in allocating resources to local priorities within the envelope of the block grant. If the Union is to survive it is important that the allocative mechanism used is in a rough and ready way based on population shares.

  6.  I would accept the tenor of the question that in practice there are many definitions of fairness when discussing budgetary policies. I am also sceptical that it is possible to arrive at a needs based allocation regarded as balanced and fair to all parts, regions and nations of the UK. As I note above, the division of spending into DEL and AME does allow spending based on needs to automatically feed through to deprived parts of the UK—assuming policies for ensuring take-up of benefits are applied equally across the UK.

Q4.  To what extent are there tensions between allocating expenditure according to such criteria as need, efficiency and effectiveness? How would you suggest those tensions might be resolved?


  7.  There clearly are tensions between the various criteria by which spending could be allocated. The current Barnett formula gives greater precedence to efficiency and effectiveness than to needs-based elements. I remain to be convinced that the way to deal with the political, social and economic inequalities in the UK is through wholesale changes to the Barnett formula. If the Union is to be maintained in more or less its current form, the onus should be on changing public policies, and changing the way in which those policies are delivered rather than through changes to the allocative mechanism. Greater emphasis on the demand (and by implication needs based) led element of spending outside the DEL process would be one way to help resolve tensions. If demand for social security payments rise differentially across regions, this is easier to "sell" as part of the social contract implied by the Union, than highly contentious debates about which parts of the UK have greater need.

Q5.  How effective would it be to use population or other proxy indicators of need, such as inverse GDP or perhaps social security spending per head, as alternatives to carrying out a detailed needs assessment? What would be the overall effect, in terms of the distribution of spending, of adopting those?


  8.  I can see many problems with using inverse GDP or social spending per head as proxy measures for need. GDP data is notoriously difficult to predict and is subject to considerable revision. Social security spending per head represents only one element of needs (and as I note above is already part of the non-Barnett element of public spending which acts as a social stabiliser). In Northern Ireland, for example there are needs which result from historic inequalities of opportunity in education and skills which would not be picked up through social security benefit indicators. The costs of separate schools and even health centres based on the sectarian divide cannot easily be calculated, but remain a significant element of any needs assessment for the Province (and could apply equally to parts of lowland Scotland) and other parts of the UK—the North East or North West of England—for example would not have particular elements of needs allowed for.

Q6.  Assuming there is to be a mechanism for distributing financial resources from the UK Government to the Devolved Administrations, as the main source of revenue for the Devolved Administrations, do you think that a needs-based formula is the only real alternative to the current Formula? What other alternatives might there be?


  9.  I would propose a Barnett-plus approach to reforming the current formula. Essentially the existing population based formula should be applied to comparable Departmental Spending, but a more explicit bidding process outside that allocation should be set up allowing the devolved administrations to bid for a Reserve of additional spending. This would be set as a proportion of total DEL and could also be open to the regions of England (via the Ministers for the regions) to bid for specific additional resources seeking to address quantifiable social and economic need. The bids would be made on the basis of programmes to be funded for the three years of the Survey cycle (assuming there is another CSR in the near future). Allocations would be made on the basis of submission to the Treasury and final adjudication would be before a Cabinet Committee made up of senior non-departmental ministers. I would be open to persuasion that adjudication could be through the Australian model (in Q9) below or the Canadian model, rather than via a Cabinet Committee.


Q7.  Are there still problems relating to the collection, quality or availability of data on the distribution of public spending and its effects? What issues are there on data about indicators of need and tax revenues?


  10.  We still have some way to go to have more reliable data on the distribution of spending and on need and tax revenues. The most pressing relate to assigning tax revenue where there is an HQ in one region for purely administrative purposes whilst there is a greater distribution of outlets in other parts of the UK—does this over-accentuate the dominance of London and the South East? How far can the administrative and HQ spending of UK departments be assigned more broadly when many departments (such as the Treasury and Cabinet Office) are effectively both English and UK departments?; how far should the MoD's central HQ spending be allocated outside London?. Much of this is the product of the failure by successive governments to move civil service jobs outside London. I am sure we could learn from comparative research on the way Canada, Australia or Germany arrive at data measures upon which to build fiscal transfer policies.


Q8.  Most writers consider that procedural fairness and transparency are important aspects of any system of financing the devolved administration, and that this an area in which the present arrangements are defective. Do you agree? What information should be published or other processes adopted to improve procedural transparency?


  11.  Transparency has been improved since the Treasury published the basic handbook on funding arrangements for the territories after each Comprehensive Spending Review.4 Prior to 1999 researchers relied on occasional morsels from Parliamentary Written Statements or off-the record discussions with officials. It is now possible to track programmes and sub-programmes for English/UK Departments deemed comparable with spending in the territories, and read explanations about how different elements of spending are treated using the Barnett Formula. However, this only goes part of the way to opening up the process. It would greatly help if the Treasury published the actual increments distributed to the territories in the construction of each of the three block grants—it would then be possible to see how the territories had decided to allocate more or less than comparable spending in England. It would also enhance procedural transparency if details of bids outside the Barnett formula made by the territories were published, and outcomes agreed with the Treasury clearly outlined. The machinery for operating the concordats between Whitehall, Westminster and the Territories needs to be dusted off, oiled and given more regular outings, and minutes of meetings between Ministers published.

Q9.  How workable would be a UK Territorial Grants Board given that its Australian prototype, the Commonwealth Grants Commission, operates in a symmetrical, federal system of government, with substantive fiscal autonomy for the States? Can a Territorial Grants Board improve procedural fairness or provide a system which is deemed legitimate?


  12.  There is a need to debate comparative experiences and see what can be learnt by the UK. Although beyond the scope of the Committee's inquiry, there is a prior issue to be addressed—creating more explicit rules for our quasi-federal polity. If a Territorial Grants Board were to be created, we need to have clearer rules about representation (should English Regions through RDAs or regional assemblies be part of the process?), dispute adjudication, and whether a statutory framework should replace the "gentlemen's agreements" of the Concordats (as they so anachronistically define themselves). As so often in the analysis of apparently technical budgetary procedures, the real political issues which lie beneath are about constitutions, rules and procedural norms.


  1  Discussed in Colin Thain and Maurice Wright, The Treasury and Whitehall: The Planning and Control of Public Expenditure, 1976-1993 (Oxford: Oxford University Press, 1995), ch 14.

  2  HM Treasury, Needs Assessment—Report, Report of an interdepartmental study coordinated by HM Treasury on the Relative Public Expenditure Needs in England, Wales and Northern Ireland (London: HMSO, 1979).

  3  In recent discussions with Treasury officials as part of a current research project—the Treasury under New Labour ( )—it was apparent that officials who deal with territorial expenditure are as surprised as outsiders that the system has survived more than 20 years. Lord Barnett is also on record as expressing surprise at the longevity of the formula to which his name was assigned.

  4  HM Treasury, Funding the Scottish Parliament, National Assembly for Wales and Northern Ireland Assembly: Statement of Funding Policy, 5th Edition (October 2007) [available from ]

22 March 2009

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