UCT’s muddled minimum wage

University of Cape Town. Photo from Wikipedia by Adrian Frith (CC-BY-SA 2.5).

Josh Budlender and Johan Lorenzen

8 December 2014

Josh Budlender and Johan Lorenzen argue that the reasons given by the University of Cape Town (UCT) for the minimum wage of outsourced workers in 2015 do not make sense.

In June 2013, the University of Cape Town instituted a review of outsourcing at the university. This review was completed earlier this year.

Its findings, as well as UCT management’s response, were circulated to students in early November after UCT had completed its consideration of the review’s recommendations, despite requests from the SRC, among others, to make the findings public while stakeholders could still make meaningful comments.

Amongst other things, the review advocated for meaningful increases in the minimum wage prescribed for outsourced workers at UCT. Yet the university’s management rejected this recommendation in favour of keeping the minimum wage at a level which is well below the poverty line that it uses as a benchmark.

This means that by UCT’s own measure it prescribes a poverty wage for the majority of the outsourced workers on its campus, though it seems management tries to obscure this fact.

UCT began outsourcing ‘non-core’ services in the 1990s under then Vice-Chancellor Dr Mamphela Ramphele. This effort to cut costs, or as Dr Ramphele called it, ‘pruning the tree of dead wood’, resulted in the dramatic reduction of wages and benefits for workers in these sectors, which were cleaning, security, catering, transport and grounds-keeping.

Recognising that outsourcing is perceived as exploitative of workers, and in an effort to implement ‘outsourcing with heart’, UCT in 2006 created an outsourcing Code of Conduct. Apart from other things, the Code set a minimum wage for outsourced workers on campus.

Out of UCT’s approximately 1,200 outsourced workers, between 813 and 904 of these workers earn exactly this minimum wage. This minimum is determined by a race-based, apartheid-era cost-of-living measure called the Supplemented Living Level (SLL).

The SLL is widely understood to be deeply problematic, and it was partly because of this that the outsourcing review was initiated. This measure, which is meant to specify a rand amount per household that will ensure a ‘modest low-level standard of living’, sets different amounts for different racial groups, though UCT does not use the racial differentiation – all workers are paid at the highest rate.

The SLL is based on inadequate and strange assumptions about people’s needs, such as allowing, for example, adult men two pairs of underwear to last for two years, and while provision is made for young girls’ school shirts, no provision is made for school skirts or pants.

In the 2011 version of the UCT SLL, R70 and R182 were allocated for households to spend per month on rent and transport respectively. However households which earned similar amounts to the SLL actually spent R291 and R513 per month on these items, on average.

The fact that so little provision is made for these and other costs means that households have to use their food budget for these non-food necessities. As a consequence of this, households earning the SLL do not, on average, consume adequate food.

The SLL is an outdated, discredited measure which UCT did well to review and it is unsurprising that the review recommended an increase in this minimum wage. The review recommended that within 5 years, ‘in the spirit of equity and other values propagated by UCT’, the basic remuneration of outsourced workers should be no less than 1 to 25 when compared to the guaranteed remuneration package of the vice-chancellor, which according to the outsourcing review was approximately R2,800,000 in 2014. This means workers’ monthly basic remuneration would amount to R9,333.

It is tricky to convert workers’ remuneration, which includes things like unemployment insurance, into a salary amount. But if we assume that total remuneration is about 30% greater than the basic salary, which is a generous assumption in favour of management, we arrive at a monthly wage of approximately R7 179 in 2014. For comparison, the 2014 SLL value is R4 704 per month. This is a shortfall of nearly R 30 000 per year.

Without any comment on the obvious conflict of interest, UCT management rejected this recommendation for a higher wage. In a document filled with flaws, they attempted to justify their proposal that UCT’s minimum wage be kept at the SLL level. Unintentionally, this document implicitly acknowledges that UCT pays a poverty wage.

The benchmark that management uses to assess the adequacy of its minimum wage is a Cost of Basic Needs (CoBN) measure developed for South Africa by two World Bank economists, Johannes Hoogeveen and Berk Ozler. Its use here is conceptually problematic, however, as UCT management does not recognise that the CoBN measure is a poverty line, and is not designed for wage-setting. Instead it gives lines which we can use to classify whether someone is poor. People who earn incomes less than the line are classified as income-poor, while people who earn above it are non-poor.

The CoBN measure, which is based on estimates of the amount of food people need, gives us two possible poverty lines: a lower bound and an upper bound. At the lower bound, where very little provision is made for non-food items, people will not consume ‘enough’ food, as they have non-food necessities and will use some of their food budget for these. At the upper bound we can be quite sure that people are on average getting ‘enough’ food in terms of calories consumed, though questions remain regarding non-food expenditure. You can read a more detailed explanation of the CoBN measure here and here.

While UCT management’s understanding of the upper and lower bounds is entirely incorrect, they correctly do not use the lower bound as any kind of benchmark. However they do use the upper bound as a measure of how to judge a sufficient wage, which is also wrong. While the SLL is a problematic measure and indeed should no longer be used for wage-setting in South Africa, it at least was developed for this purpose. If UCT sets wages according to the CoBN, it is essentially saying that its benchmark is a wage amount which puts UCT workers right on the precipice of poverty.

However even if this conceptual issue is ignored, there are technical errors in the document which significantly bias the results against an increased minimum wage. Perhaps the worst mistakes are in how UCT management adjusts the upper-bound CoBN measure for inflation.

Management uses what they call ‘reasonable inflation measures’ to transform their 2008 CoBN estimate into a 2014 value. However it is unclear what UCT management means by ‘reasonable inflation measures’. The inflation measure they use results in a 2014 CoBN estimate lower than the result you get using either StatsSA’s headline (overall) inflation or quintile-2 (‘low-income’) inflation.

The UCT Finance Department uses quintile-2 inflation when adjusting the SLL each year, as it better reflects the price changes that these workers face. Yet UCT management estimates the 2014 CoBN to be R5,203 per month instead of R5,552, which would be the more correct estimate using quintile-2 inflation. Table 1 compares results from UCT’s unidentified “reasonable inflation measures” to quintile-2 inflation.


Management then compares their CoBN estimate to the 2015 wage they decide on, which is the 2015 SLL (calculated using quintile-2 inflation), equal to R5,018 per month. UCT management concludes that the SLL amount (approximately R5,000) is not that different to the CoBN amount (approximately R5,200), and that therefore the SLL is an adequate wage.

However this is wrong on both conceptual and technical grounds. Conceptually, because the CoBN is a poverty line, paying a wage even R1 below this line means paying someone a poverty wage. R200 less than the poverty line is clearly not an adequate wage. However, to make things worse, the comparison between R5,000 and R5,200 is wrong. This is because while R5,000 is the wage for 2015, R5200 is the supposed CoBN measure for 2014. To make this a 2015 measure as well, we inflate it by management’s unidentified inflation measure, and find that a better comparison would be between the wage of R5,000 and a 2015 CoBN poverty line of R5,511, or approximately R5,500. This means UCT’s wage is approximately R500 less per month than their own poverty line.

But this is still not accurate. If we finally correct UCT’s estimates of the CoBN by using the more correct inflation measure (quintile-2 inflation), we find that the 2015 value for the CoBN measure is actually R5,923, or approximately R5,900. The end result is that UCT’s decision to stick with the SLL, and set a minimum wage of R5,000 per month, means that they are setting a minimum wage which is R900 less per month than their own poverty line. Put differently, the wage is 15% less than what is needed to make sure workers aren’t in poverty, by UCT management’s own standard. This makes it clear that UCT is prescribing a poverty wage.

It is important to briefly address one more distortion made by management, because it is a potentially significant issue. This concerns the effect that wage increases would have on students’ fees. In calculating this, management makes the extraordinary assumption that if the minimum wage increases, other outsourced workers at UCT currently earning above the minimum wage will also require an increase in their wages of the same proportionate magnitude.

What this means is that if the wage for a worker at G4S (UCT’s security company, which pays the minimum) increases from R4,700 to R5,200 (11% increase), then management assumes that another worker at Sibanye (the bus company, which pays above the minimum) will also increase her wage from R7,200 to just under R8,000 (11% increase).

This assumption exaggerates the effect that a minimum wage increase would have on fees. It is not plausible for a number of reasons, including the fact that some of the wages above the SLL at UCT (wages for Sibanye workers) are determined by National Bargaining Council and won’t change in reaction to UCT-specific policy.

Also it is generally the case that workers who earn above the minimum are mostly in different sectors to minimum wage-workers at UCT, and it is unreasonable to assume that these workers will automatically demand wage increases proportionate to the change in the UCT minimum wage when it hardly affects them. The figures that UCT provides on the effects of an increased minimum wage on student fees are very likely a significant exaggeration of reality.

The flaws in the section on fees are unfortunately part of a consistent pattern throughout management’s document. In almost every case where a judgement call must be made and some bias is unavoidable (or even where it is avoidable, such as in the inflation adjustments) the bias is against an increased minimum wage. It is disturbing that UCT will continue to prescribe a wage that ensures that its outsourced employees live in poverty according to UCT’s own measurements.

It is even more disappointing that UCT has attempted to brand this wage as aligned with its commitment to “promote social justice and equity.” If UCT management does not want to increase the wages of its outsourced workers then that is its prerogative. It is entitled to argue that workers’ wages are less important than other university priorities, and that paying poverty wages is necessary.

However, management is clearly not the only stakeholder with a legitimate interest in this issue. To enable all interested stakeholders to contribute to this debate meaningfully, management’s reasons must be made on sensible assumptions and must be presented openly and transparently. In the process by which it has set a poverty wage here, it has failed to do so.

Budlender is completing his Honours in Economics at UCT while Lorenzen is completing his LLB. Budlender wrote his Honours thesis on UCT’s use of the SLL to prescribe a minimum wage at UCT for outsourced workers. Lorenzen is co-founder of the Workers’ Referral and Advice Programme (WRAP), which provides free legal advice to workers at UCT.

The opinions expressed in this article are solely those of the authors. No inference should be made on whether these reflect the editorial position of GroundUp.