Academic pay policy
The pay policy seeks to set levels of remuneration at amounts which enable the University of Cape Town (the University) to attract, retain and motivate staff. It has been decided that the best way of achieving this aim is to benchmark salaries at the University relative to those at a selection of comparator universities, namely, those universities which share a similar research-led culture to that at the University. Arguably these universities constitute the University's major competitors for staff and face similar challenges in terms of spending demands and opportunities to earn income. In making this choice of comparator universities the University acknowledges its commitment to offering a quality educational experience to its students and to a flourishing research culture, and thus to maintaining the necessary staffing levels and infrastructural support for each objective.
The benchmark is to be fixed as follows: the Standard Academic Salary Package (SASP)1 at each rank is to be aligned with the 75th percentile of salaries at that rank at the comparator universities. The rationale for choosing this benchmark is that the University recognizes that, as confirmed by a number of objective measures, it has a pre-eminent status amongst the comparator universities (indeed amongst universities nationally) and primarily owes that status to the quality of its academic and research work. It is thus reasonable to assume that members of UCT's academic staff who are meeting the criteria for SASP are performing at least at the level of three quarters of staff in comparator universities and thus, independent of their other earnings in terms of scarcity, merit or excellence awards, should be paid accordingly. Notwithstanding this, movement away from the benchmark may arise as a result of constraints of sustainability and of attempting to match appropriate differentials between salaries at each rank. (This is dealt with in more detail below.)
The method of implementing the benchmark is complex given the nature and availability of data and is detailed in section 3. The broad thrust of the plan is to fix the benchmark using retrospective data of salary levels combined with an agreed method of projection. Recognising that even retrospective data may be misleading about the true level of the benchmark, this document details a set of triggers which will:
- suggest a need to treat the data with some caution; and
- suggest a judgement about the true level of the benchmark.
The effects of variations caused by inaccuracies in the data will be mitigated in the application of the benchmarking process over time.
Implementation of the benchmark is subject to constraints of being financially sustainable. This document attempts to frame a conception of financial sustainability which divorces that notion from the short-term spending priorities of the University; rather whether a particular salary rise is deemed affordable will depend on the fundamental financial health of the University measured by such things as the levels of its free cash reserves. This is dealt with in more detail below.
Finally, it is acknowledged that the salary structure reflects a career path for academics/researchers and thus maintenance of appropriate differentials between ranks also needs to be a goal of the policy. Differentials should be set at levels which, whilst not threatening solidarity within the staff body, should constitute an incentive to earn promotion. Here it is proposed that targeted differentials be fixed as bands and that these be implemented in informing the judgements with regard to salary increases arising out of the benchmarking process.
The salary negotiation process will proceed in the following steps:
- Parties (i.e. Union and Management) agree to the sources of the relevant data.
- The data is used to arrive at a judgement on the levels of the 75th percentile at comparator universities.
- The data is scrutinised by both parties relative to the agreed triggers2 to decide whether the data can be relied on in its raw form.
- Should it be agreed that a trigger is activated, a fresh judgement of the levels of the 75th percentile at comparator universities will be arrived at.
- If parties are in agreement at stage (iii) or at stage (iv), then the benchmark is submitted to the test of being sustainably affordable.
- If parties agree to a particular salary rise having considered the constraints of financial sustainability then it will be compared to desired differentials between ranks.
- If parties reach agreement about the salary rise taking into account the need to maintain appropriate differentials, then the negotiation process for salary rises is complete.
- A dispute may occur should agreement be failed to be reached on: what salary rise is determined by the benchmark (this may include an inability to agree on whether the data is reliable, i.e., whether a trigger should be activated); whether the salary rise determined by the benchmark is sustainably affordable; or what an appropriate rise across ranks is, given the desired differentials.
In order to retain its leading position as a research-led university, the University must be able to attract excellent staff from within and outside of South Africa while being attractive to young people as a good career option; it is thus imperative to have an appropriate benchmark against which to measure the University's academic salary levels.
The benchmark must be constructed on the basis that the University should pay its academic staff at the upper quartile of the tertiary market (as determined by an appropriate, agreed set of comparators)3 and that such a benchmark should:
- be implementable;
- maintain competitiveness in salaries; and
- be reasonably foreseen to be affordable in the medium term, i.e., sustainable.
In addition, the data must be reliable and comparable.
The benchmark for academic salaries is the 75th percentile of the market (as determined by comparison to the pay line of an agreed set of comparator universities)4. This benchmark is determined as follows:
- the retrospective data indicating the 75th percentile as at January of the year in which the salary negotiations are being undertaken (i.e. January of the year preceding the increase date) is taken as the basis;
- the year-on-year national CPI as at end May of the year in which the salary negotiations are being undertaken, with a further adjustment for the balance of the year based on estimates from recognised and agreed sources.
This method of establishing the benchmark will be reviewed every three years.
In this way the benchmark would be informed by historical data, thus obviating the need for debate on any forward projections of future increases to be made in the market or what CPI will be in the coming year.
Although the aim in this element of the negotiation process is predicated on determining the 75th percentile of academic salaries at comparator institutions, it is acknowledged that the best available data may be deceptive5. Applying the benchmarking process over a number of years will have the effect of mitigating some of the variation in the data. Nonetheless the parties agree to monitor a set of triggers used to indicate possible unreliability in the data. These triggers include:
- a detailed examination of the composition of the data set, searching for sources of incompleteness and inaccuracy;
- consideration of trends over the last number of years, searching for possible inexplicable movements away from the trends;
- a comparison of the salary rises suggested by the data with increases in CPI, searching for large discrepancies; and
- a comparison of the salary rises suggested by the data with increases in salaries in the general job market, searching for large discrepancies.
In any year in which there are grounds for treating the data with some suspicion, those grounds must be articulated and used to arrive at a judgement on a reasonable implementation of the benchmark.
In addition, a review of staff turnover and retention data should be considered in order to determine the adequacy of the benchmark. (See below.)
Financial sustainability is a necessary consideration in any pay policy and consequently a discussion of the sustainability of proposed increases is likely to form a part of annual salary negotiations. In order to link admissible salary rises to the underlying financial health of the University, it is agreed that the evaluation of sustainability be made against an agreed set of indicators.
The criteria to be used for discussions on financial sustainability should be a set of data using metrics looking back over a number of years. The indicators of financial sustainability are:
- the amount of University's free cash reserves as assessed6 by the March meeting of the UFC,
- total academic salaries as a percentage of agreed metrics (which could include operating surplus and some form of staff spend to student volume ratio FTE:SLE); and
- total academic salaries measured against subsidies and fees.
It is understood that any particular event that is likely to have a medium to long-term effect on the above metrics, can be raised by the University in any discussions on sustainability: for example, external events such as a change in subsidy.
Management will have the right to argue that considerations of sustainability may, in a particular year, justify a settlement below the figures dictated by the Benchmark. The Union will have the right to argue that in a particular year sustainability considerations allow and performance warrants a settlement above the Benchmark.
In line with the recommendations from the RFJ task team and national and international benchmarking, there should be appropriate differentials between ranks.
The pay policy accommodates a set of differentials between ranks, to provide incentives across the life-span of an academic career, yet maintaining solidarity across ranks.
The suggested differential target between ranks to strive for is 18-22%, with 20% being the ideal. The tolerance level of 2% above and below the ideal of 20% ensures that the target does not become excessively rigid and recognizes the fact that a choice to bring the differentials to the desired norm, may not be compatible with achieving the 75th percentile at a given rank. Nonetheless the rationale for benchmarking salaries requires that salaries are benchmarked at each rank; in addition, large movements away from these targets will frustrate efficient implementation of the policy in future years. Thus the benchmarking element of the policy has a priority over the targeting of differentials between ranks. This policy envisages a process in which benchmarking is carried out and then an attempt is made to impose a structure on salaries which approaches the targeted differentials between ranks; but in doing so a movement from the benchmark figures which exceeds 3.5% will not be permitted.
The targets for differentials between ranks are to be reviewed at least every three years.
The benchmark and the means of tracking it should be reviewed every three years as a matter of course; but may be reviewed at an earlier stage should circumstances require it. The policy has been written in a particular governmental and economic climate and with certain goals in mind. Should there be evidence of radical change in the climate or that the chosen benchmark and mechanism of tracking it are failing to meet the goals of the policy the parties agree to review the pertinent sections. The relevant items include:
- Large shifts in levels of inflation;
- Significant changes in the mandate of the reserve bank with regards to inflation targeting;
- Data indicating a failure to attract and retain quality academic staff;
- Large shifts in government policy on funding in higher education;
- Changes in the higher education sector indicating an inappropriateness in the chosen comparator universities.
- Sustained tension between the benchmark and the targeted differentials.
2 Here an indicator is a piece of data used to determine the level of another item (e.g. salaries or percentage rises); a trigger is a piece of data used to determine whether some further process is required (e.g. interrogation of a set of data). Some triggers, e.g those monitored by the UFC, will be known long in advance.
- incomplete data sets arising from failure of some institutions to upload their data; and
- inclusion of out of date data arising from some institutions failing to update their data sets timeously.
Approved: 23 November 2011
Page last updated: 12 June 2012