The University's Resource Allocation Method (RAM) (updated for 2003-4): Supplement (1) to Gazette No. 4676

  • Introduction
  • General principles
  • Details of the resource allocation method
    • Estimate total University income
    • Top-slice funds for specific purposes
    • Distribution of remaining income to divisions and services
    • Directly allocated income
    • Allocate the remaining funds by formula
    • Adjust for differential salary costs
    • Levy capital charge
    • Levy infrastructure charge for service budgets
    • Transitional arrangements
    • Supplementary funding
  • Further development work
  • Summary diagram (PDF file)

Introduction

1. In 2001–2, the University introduced a new Resource Allocation Method (RAM) to govern the allocation of university resources to divisions/OUDCE, and academic and non-academic services. Supplement (2) to Gazette No. 4629, Wednesday, 10 July 2002 (Vol. 132, pp. 1453–8), provided an updated account of this method for 2002–3. Since then, Council's Planning and Resource Allocation Committee (PRAC) has reviewed the method once more, in the light of experience of the first two years and taking into account comments which have been received within the University on its operation. As a result, some further revisions have been introduced into the allocation method for 2003–4. The purpose of this note is to provide an updated account of the Resource Allocation Method.

2. The paragraphs below set out the underlying principles, and main objectives, of the method, and give details of the formulaic and non-formulaic elements embodied within it. They highlight refinements to the method which have been made following the third year of its implementation.

3. Further details on any aspect of the method may be obtained from relevant staff of the Planning and Resource Allocation Section in the University Offices (and in particular Mr M.D. Sibly, telephone: (2)80303, e- mail: michael.sibly@admin.ox.ac.uk, or Mr P.F. Clark, telephone: (2)70078, e-mail: paul.clark@admin.ox.ac.uk). Details can also be found on the Section's Web site, at http://www/admin.ox.ac.uk/pra/.


General principles

4. The RAM was introduced to complement the University's new governance and planning structures. It was intended to provide a transparent and consistent means of allocating the University's income to divisions, to whom responsibility for preparing academic budgets was devolved. Thus, whilst Council through the PRAC remains ultimately responsible for approving proposed divisional budgets, and for monitoring activity against them, responsibility for determining detailed spending decisions now rests with divisions. Details of the application of the method within each division may be obtained from the relevant divisional board secretariat.

5. Among other things, the design of the RAM is intended to encourage increases in the University's income, and its most efficient use. This is achieved by allocating as much income as possible directly to divisions and OUDCE, and building appropriate drivers into the formulaic allocation of the remaining funds.

6. Implementation of the RAM is also designed to link the allocation of income directly to planned academic activity, so that money flows according to the strategic academic priorities set by the University and the divisions. Operation of the RAM should be governed by these priorities, and the academic and managerial principles established by the University, rather than by the contingencies of a given financial situation. The RAM is concerned with the most effective and fair distribution of the University's income. It is not concerned with monitoring expenditure or the costs of different activities, although this may of course inform the future development of the RAM.

7. The RAM takes account of the way in which income comes into the University, but it is not simply an `income as earned' model. The RAM thus also provides the University with a safeguard against fluctuations in the methods used externally to calculate the University's funding, and a means of mitigating the effects of external policies and methods of allocating funds which may not fit with its own strategic priorities. It gives the University the ability to move away, e.g., from the use of the HEFCE price bands and cost drivers in the distribution of income. It also allows the University to moderate the financial effects of external `market' forces, by redistributing funds towards activities which it deems to be academically desirable, but which may not be financially self-sustaining (e.g., the support of important minority subjects). Thus, the RAM is a flexible tool which the University can adapt to maximise the use of its resources in the future, and take best advantage of new opportunities and circumstances. It is up to the University to decide how best to use this tool.


Details of the resource allocation method

8. The basic method by which income is allocated to divisions is as follows:

(a) Estimate total University income.

(b) Top-slice funds for certain specific purposes.

(c) Directly allocate income where appropriate.

(d) Agree service budgets.

(e) Allocate the remaining funds by formula to the divisions/OUDCE.

(f) Adjust for differential academic salary costs.

(g) Levy capital charge, and redistribute funds to divisions.

( h) Levy infrastructure charge for service budgets.

(i) Apply transitional arrangements (moderation). Each stage is dealt with in more detail in the paragraphs below.


Estimate total University income

9. The starting point for resource allocation is to estimate total university income for the forthcoming year. For 2003–4, the estimated total is currently c.£398.8m.

10. This total is made up of income from a variety of sources. The two biggest elements are HEFCE grant (£132.2m in 2003–4), and income from research grants and contracts (estimated at £154.3m for 2003–4). Other significant components include income from university composition fees and from trust funds. It should be noted that whilst some elements in the total are known with a high degree of certainty (especially the HEFCE block grant), other elements are estimates, and the actual level of income is dependent on factors such as the level of student recruitment, and continued success in obtaining research grants and contracts.


Top-slice funds for specific purposes

11. The next step in the new method is to set aside from this sum the estimated total funds which are to be held centrally and are thus not available for distribution to divisions/OUDCE, or for the funding of services. There are three principal elements here: the transfer to colleges of the college fee replacement income (£37.6m in 2003–4); the provision for central contingencies, reserves, or special initiatives, amounting in 2003–4 to approximately £10.5m; and a contribution to the services budgets, amounting in 2003–4 to some £11.8m. This leaves remaining income of approximately £338.9m for allocation in 2003–4.


Distribution of remaining income to divisions and services

12. The method for distributing this sum comprises three distinct components:


Directly allocated income

13. The directly allocated income (bullet point 1 above) consists of three principal elements:

(a) All research grant and contract income, including 100 per cent of overhead payments on such grants and contracts (totalling an estimated £154.3m for 2003–4). It should be noted that whilst the direct income associated with research grants and contracts is clearly earmarked and not available for other purposes, the overhead payments on such contracts are not earmarked. Although initially allocated directly to divisions/OUDCE, these overheads are in part intended to support activities outside the departments or faculties which provide immediate support for research and therefore should be regarded as being available to contribute to the infrastructure charge for service budgets, dealt with below. Allocating all overheads to divisions/departments should act as a clear incentive to them to maximise overhead recovery rates.

(b) Various other earmarked or directly earned sums: these include, for example, income from a wide range of trust funds, earnings from trading activities, etc. A significant component of the funds under this heading consists of funds which are earmarked for or earned by academic services or other service units.

(c) Certain unearmarked general funds, which the University is free to allocate as it wishes, but which are nonetheless being directly allocated to academic units because it is felt that this is the most appropriate way to treat them. The most important elements under this heading consist of income from tuition fees paid by overseas students, or income from tuition fees which is for other reasons being allocated directly to the host department (and in particular OUDCE). The rationale here is that in those cases where the University is free to set the level of tuition fees itself, such fees should be directly allocated, so as to encourage divisions and departments to consider the level at which they are set, and to ensure all costs are recovered.

14. It is intended that the direct allocation to divisions, departments, and relevant services is not only simple, clear and transparent, but will also act as a strong incentive to the relevant units to increase such income wherever possible, and to consider carefully how their costs relate to the income generated.


Allocate the remaining funds by formula

15. The second main step in the new method is the formulaic allocation of remaining income to divisions/OUDCE. For 2003–4, the total available is approximately £108.1m. This figure includes approximately £5.8m supplementary funding transferred to the divisions from the University's Central Strategic Reserve (see paras. 38--9).

16. The formulaic allocations are based principally on (a) teaching-related criteria and (b) research- related criteria. Allocations of funds on this basis involve a series of linked steps. The first is to decide how to divide available resources between those to be allocated on teaching criteria and those to be allocated on research criteria. For 2003–4 the agreed ratio is 40:60 between teaching and research, this ratio broadly reflecting the basis on which the funds are received from HEFCE.

17. Teaching related funds (approximately £42.4m in 2003–4) are then allocated pro rata to weighted full-time equivalent home/EC student load. The weightings primarily reflect differences in costs between different subject areas, and in the first instance the basis for these are the weightings used by HEFCE in its own block grant allocation formula (these are: 4.5 for clinical subjects; 2.0 for laboratory-based subjects; 1.5 for subjects with a laboratory or fieldwork element; 1.0 for all other subjects). Oxford has also introduced a further weighting (of 1.25) to reflect the added costs of teaching postgraduate taught (PGT) students. It should also be noted that, because student load relating to postgraduate research students (PGR) in years 1–4 of their study are used both in the teaching component of the formula, and in the research component described below, there is de facto an additional weighting attributable to such students.

18. In previous years part-time undergraduate students have attracted a 5 per cent premium in the teaching- related element of the RAM. However, as HEFCE's 2003–4 funding for part-time students (£162k in 2003–4) was incorporated within the grant for widening participation and improving retention, this premium was removed from the 2003–4 RAM and the associated funding directly allocated to divisions pro rata to numbers of part-time undergraduate students (mainly but not exclusively to OUDCE).

19. The research related allocations (approximately £63.6m in 2003–4) are more complex, but are designed to reflect both the basis on which funds are received by the University, and also the quality and volume of research activity in a particular subject area.

20. The principal component of the research-related allocation distributes funds pro rata to a measure of the volume of research activity, weighted to take account of quality (as reflected in the most recent RAE), and a relative cost band weighting (derived from HEFCE's own research-related cost band weightings). The core volume measure comprises numbers of staff returned as research active (`category A' and `category A*' staff in post), in the most recent RAE, who are funded from general income or NHS-specific funds (this provision chiefly excluding those research-active staff funded from research grants and contracts, and some staff funded from endowments) and PGR load (weighted at x 0.15). These numbers are then weighted by a quality factor derived from the most recent RAE ratings for the relevant department or unit, and by an appropriate HEFCE cost weighting, designed to reflect the differential costs of undertaking research in different subjects. The quality-related weightings used in Oxford are: RAE grade 4 = 2.0; RAE grade 5 = 3.0; RAE grade 5* = 4.0. Thus, no credit is given for RAE grades 3a or below (Oxford no longer has any such departments). The cost weightings are: 1.6 for high-cost laboratory and clinical subjects; 1.3 for intermediate cost subjects; 1.0 for all other subjects. [1] In 2003–4 £46.4m was allocated by this method.

21. A significant proportion of the research-related block grant received from HEFCE is generated on the basis of the volume of external research grants and contract income received from UK- based charities: under the RAM, this funding (some £11.6m in 2003–4) is distributed pro- rata to a two year moving average of UK-based charity research income, weighted for RAE-derived quality ratings, and HEFCE research cost band weightings.

22. Two other smaller sums are distributed on the following basis:

(a) some £4.7m in 2003–4 is allocated pro rata to non-charitable UK research grant and contract income, again weighted for RAE-derived quality ratings and HEFCE research cost band weightings; and

(b) a further sum of some £945k in 2003–4 is distributed pro rata to the number of research fellows employed in a given department (using the same definition of what constitutes a research fellow as applied for the RAE 2001). This number is taken from the annual Research Activity Survey data return made to HEFCE, on the basis of which a part of the block grant for research is distributed. The number of research fellows is not weighted for either relative subject cost or quality.

23. The combination of these four components is designed to achieve an appropriate balance of incentives to generate additional research-related income, to reflect research quality, and to recognise differential costs. As with the teaching-related formula, use of externally derived quality and cost weightings is the subject of ongoing review.

24. For 2003–4, the University received £4.4m of research funding in respect of 5** Units of Assessment i.e. those units graded as 5* in both the 1996 and 2001 RAEs. This funding was topsliced at 25 per cent (£1.1m) in recognition of the fact that the additional funding resulted in an above inflation increase in the Quantum: these funds were used to notionally offset the topslice. The remaining 5** funding (£3.3m) was directly allocated to the earning divisions, outside of the formulaic allocation of research funding.


Adjust for differential salary costs

25. The basic formulaic allocations on teaching and research related criteria derived from the above approach are then subject to two further modifications. The first of these is an adjustment designed to reflect the differential costs to departments and faculties of certain types of academic appointment—and in particular the relative costs of CUF/faculty lecturers, and university lecturers. This adjustment is intended to be financially neutral, and any plans to move academic posts from one category of appointment to another will need to be approved by PRAC and by Council, after consultation with colleges, as part of the annual planning and budgeting process. For 2003–4 prospective data was used, and the £7.1m generated by this means used to offset the top-slice.


Levy capital charge

26. The second adjustment is made by the application of a capital charge. This is intended to provide an incentive to departments to be efficient in their use of space, and to ensure that they do not expand to larger premises without being exposed to the financial consequences which such expansion has for the University as a whole. It is also intended to put a price on the opportunity cost of using capital funds to invest in buildings rather than staff. The charge is currently set at £20 per square metre, and is to be levied on divisions pro rata to the area occupied, although some types of space are excluded from the charge (particularly recently constructed space funded from external donations, grants etc. from non-public sources).

27. The funds generated by the capital charge (about £3.4m for 2003–4) are redistributed to divisions formulaically on the basis of total research active staff numbers in 2001 RAE, weighted by research quality weightings, and student numbers (with 1.25 weighting for PGT students). This formulaic reallocation of funds shares some of the same features of the main allocation formula, but is deliberately different in certain respects, notably in that it includes all students (i.e., Home/EC and Overseas), adopts different factors for the research-related allocation, and excludes any subject weightings (teaching or research).


Levy infrastructure charge for service budgets

28. The approach to setting and providing for service budgets involves several related steps. The first is to agree the level of the budget for each particular service. This is founded on an activity- based analysis, looking carefully at the costs which need to be incurred in order to deliver a particular level of service. Scrutiny of proposed service budgets was undertaken for 2003–4 by a sub-group of PRAC chaired by the Pro-Vice-Chancellor for Planning and Resource Allocation, and discussions were held both with the Pro-Vice-Chancellor for Academic Services and University Collections, and with the heads of the principal services themselves.

29. Once service budgets have been settled, the next step is to take account of the direct allocations of resources made to services, referred to in paragraph 13 (b) above. In the case of a number of services, these directly attributable or directly earned allocations comprise a significant proportion of the relevant budget.

30. The next step is to calculate the proportion of the agreed service budgets which cannot be met by such direct allocations, and then to provide funds for these through an infrastructure charge levied on total divisional income. This formulaic charge is based on a number of factors which are designed, in very broad terms, to represent proxies of usage of particular services. The principal drivers being used for 2003–4 are student load, staff load, and area occupied or maintained. The total volume of funds being provided for in this way for 2003–4 is arrived at through the following process.

31. For 2003–4, the PRAC sub-group recommended a 3.5 per cent increase in the budgets of the services and £2.2m of funding for additional items bid for during the budgeting process. However, these recommendations did not take account of:

(a) a number of funding decisions (amounting to £3.2m) taken by PRAC during the course of 2002–3, which were not consolidated until after the recommendations had been made;

(b) £1.8m of deficit funding which had been allocated to the services for 2002–3. This funding had been incorporated into the service budgets baseline prior to the 2003–4 uplift.

32. As a result, when the 2003–4 RAM formulaic income was examined, it was generally acknowledged that an imbalance had developed between the budget settlement for the services and the provisional settlement for the divisions: the total RAM budget for the services having been calculated as £70.5m. In order to address this imbalance PRAC took the decision to recalculate the baseline for the service budgets: cutting £0.5m of 2002–3 deficit funding and all previously agreed additionally funded items from the baseline.

33. Once this new baseline had been calculated, PRAC also agreed that it should be inflated by 2.5 per cent and not the 3.5 per cent originally recommended by the sub-group—effectively a 1 per cent cut. However, PRAC also allocated a further £2.2m of non-recurrent supplementary funding to the services for those additional items that PRAC had already agreed to fund (£1.2m for 2003–4) and a revised list of new items which had been bid for during the budget settlement process (£1m for 2003–4). Thus the total RAM service budget for 2003–4 was £67.58m: £65.38m funded from the infrastructure charge and topslice; and £2.2m of supplementary funding.

34. A summary of all the steps described in paragraphs 9--33 above is provided in the flow chart [PDF file].


Transitional arrangements

35. The transitional adjustment (also referred to as moderation) is the means of moving divisions smoothly to their ultimate RAM positions. Funds are taken from `gaining' divisions and redistributed to `losing' divisions in a planned way, allowing all divisions to adjust their baselines over a period of time. In a change from 2002–3, it has been agreed by Council that gains and losses arising from the formula should be limited to plus or minus 5 per cent of gross formulaic income for 2003–4 (in 2002–3 gains and losses were limited to plus or minus 2 per cent).

36. In essence, the transitional funding distributed to divisions once the RAM has been run is calculated by comparing the expected financial position in the current year with the projected position of the year for which allocations are being made. The additional funding to `gaining' divisions is capped at plus 5 per cent, and the funds released as a result are used to support `losing' divisions.

37. Council has also agreed that the funding allocated to the Medical Sciences Division as a result of their successful bid to HEFCE for Additional Student Numbers (ASNs) should be excluded from moderation. Thus the 2003–4 RAM was run (a) including the ASNs and (b) excluding the ASNs. The cash difference between the two models was then allocated to the division as a lump sum.


Supplementary funding

38. For 2002–3, £8m of deficit funding had been provided to the divisions through the formulaic element of the RAM. For 2003–4, it was proposed that no general deficit funding should be provided i.e. through the formulaic element of the RAM, but that supplementary funding should be allocated outside of the RAM, targeted to the areas of greatest need.

39. In the event, it did not prove possible to implement the new system. Instead, the following funding decisions were taken:

(a) £4.4m of funding to moderate, and raise divisional allocations to 2002–3 cash levels;

(b) £1.4m of additional funding for 5** departments;

(c) £2.2m of supplementary funding to the Services.


Further development work

40. Analysis and review of the implementation of the RAM, and its strategic implications for the University, is a continuing process. Detailed scrutiny of its operation will continue, with proposals for refinement and development being brought to PRAC and Council for approval. This will also seek to ensure that further changes take full account of the University's emerging academic and financial strategies.


Footnote

[1] Following the RAE in 2001, HEFCE reduced the cost band weighting for high-cost laboratory subjects from 1.7 to 1.6. However, in order to maintain stability, the 1.7 cost weighting was retained for the 2002–3 RAM. After further analysis PRAC agreed that this weighting should be reduced to 1.6 for the 2003–4 RAM.

  • direct allocation of a substantial proportion of total income to divisions/OUDCE, and to services, on the basis that the income in question is clearly generated directly by the relevant unit;
  • the formulaic allocation of remaining income to divisions/OUDCE; and
  • a formulaic infrastructure charge, to fund those elements of service budgets which are not covered by direct allocations to the services.

ram 2003 4

RAM: Summary diagram