Deputy President Cyril Ramaphosa and members of the committee of principals received the report of the advisory panel on the national minimum wage at the National Economic Development and Labour Council last Sunday. The panel recommended that, subject to certain exclusions and tiers, SA’s first national minimum wage should be R20 an hour – about R3,500 per month for the average worker who, recent research shows, works about 41 hours per week.
It has been heartening for the panel that our proposal has received widespread support from all parts of society.
We have, however, received our fair share of criticism. Some have criticised us for setting a level that is too low, arguing it is below various poverty lines. For example, some estimates place the working poverty line at R4,317 per month. Arguments have been made for R4,500 and even R12,500 as the minimum level.
On the other hand, a number of commentators have argued that the proposed level is too high and is likely to lead to an increase in the already unacceptably high level of unemployment in SA. Criticism of this sort was amplified by the release of the quarterly labour force survey for the third quarter of 2016, which showed a rise in the unemployment rate to 27.1%.
Not untypically, commentary on the unemployment rate and minimum wages misses the fact that the increase in the third-quarter unemployment rate was not the result of a fall in employment.
Employment actually grew marginally compared to the same period last year, and by 288,000 compared to the previous quarter. The rise in the unemployment rate was as a result of a fall in the number of discouraged workers.
The dilemma the panel faced was unlike that in other areas of public policy, where the choices about winners and losers are choices about resource allocation between different income classes. For us, the potential winners and losers are all among the most vulnerable people and households in SA.
Set the level too high, and the policy could have the unintended effect of job destruction. Set the level too low, and we maintain the unacceptable levels of working poverty. The panel’s task was to find the right balance.
At the outset, we agreed that our proposals should be evidence-based and we focused on three types of evidence.
First, the panel looked at statistical evidence on poverty and earnings in SA. The average monthly income for quintile one and quintile two households is R1,671 and R3,125, respectively. At the other end of the distribution, the average monthly income for a quintile five household is R24,090.
Furthermore, having access to employment is no guarantee of escaping poverty. Almost 35% of workers in SA earn less than R2,500 per month, 47% earn less than R3,500 per month, and 51% earn less than R4,000 per month. These striking levels of working poverty demonstrate the urgency to tackle the issue.
Second, the panel looked at the international experience. We were fortunate to have an international adviser from the International Labour Office, Dr Patrick Belser, assisting us in our deliberations. A number of countries throughout the world including Chile, Costa Rica, Mexico, Turkey, Germany and the UK have successful minimum wage regimes.
From the international experience, we took three key lessons: minimum wage legislation does have a positive effect on poverty levels; minimum wage regimes do have a positive effect on inequality; and the overwhelming evidence is that minimum wages — set at the correct level — have a relatively benign consequence on employment.
Third, we carefully assessed South African research on the potential effect of a national minimum wage. It may be useful to consider this from a microeconomic perspective first. The question is: how might an employer respond to an increase in the wages that have to be paid to employees?
They could: absorb the costs if the increase is not too large and is correctly phased (a benevolent employer); accept the higher wage and work with employees to improve productivity (an efficiency-seeking employer); if they are in the correct market, gleefully prepare for an increase in business as more low-income workers buy more of their products (an employer in the right industry); pass the costs on to customers or even suppliers (an employer with market power); look to expand their business to fund the rise in wage costs (an entrepreneurial employer); cut employment; invest in technology, which could increase productivity, wages and employment or result in a fall in employment as machinery replaces workers.
I have not by any means exhausted all the possible responses, but the message should be clear: this is a complex process and we should carefully consider the matter before reaching any conclusions.
The panel was presented with the results of three modelling exercises that sought to estimate the possible consequence of a national minimum age in SA. It is worth mentioning that by their very nature, modelling exercises are inexact and the results should be treated with great caution.
They are attempts to predict the future and we should therefore see them for what they are: a simplification of the real world. Economists like their models to be parsimonious – as simple as possible and capturing only the most essential features of the issues under consideration.
So, what did the three models say? The first, by the Treasury, estimated that a national minimum wage of R3,189 could result in employment losses in the order of 715,000 jobs.
A second model, by the Development Policy Research Unit (DPRU), estimates job losses of 566,000 based on a minimum wage of R3,400 and an employment elasticity of -0.3 (a 1% wage increase results in a 0.3% decrease in employment).
On the other hand, the so-called ADRS model, used by the Wits National Minimum Wage Research Initiative, suggests that a national minimum wage in the region of R4,500 will lead to only a marginal decrease in employment, which will be significantly outweighed by a boost in consumption-led growth among low earners.
All of these results are driven by their structure and assumptions. The DPRU model, for example, is based on a theory that says an increase in the price of labour must result in a fall in the demand for labour, hence the negative consequence.
The ADRS model, on the other hand, sees a possible channel for productivity gains as a result of the minimum wage. This model allows for the theoretical argument that workers who are better paid will work more productively (the so-called efficiency wage hypothesis).
What did the panel make of all of this research? The panel drew three conclusions.
First, that the adjustment process by which the economy realigns to a national minimum wage will probably be highly complex and might include all of the economic effects captured by the models above (and some that have not been captured by these models).
Second, that there was no point in commissioning further research to generate yet more models — the results would be driven by the theoretical assumptions made and would merely result in a further set of results.
Third, the panel concluded that the realignment and adjustment process was the most critical issue. The panel had to recommend a process that would create a framework to take on the issue of low wages, while at the same time allowing for a responsible adjustment and realignment process to reduce the risk of job losses.
It is for this reason the panel recommends:
- In sectors where wages are lowest – agriculture and domestic work – the level of the wage should be set initially at 90% and 75%, respectively, of the minimum wage.
- A two-year adjustment process to allow enterprises ample time to adjust and realign.
- Because they are less able to adapt, small businesses be granted a three-year adjustment and realignment period.
- An efficient temporary exemption process be designed.
- Most important, any adjustments and revisions to the system should be evidence-based.
• Valodia is dean of commerce and law at the University of the Witwatersrand, and chaired the National Minimum Wage Advisory Panel.
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.