Johannesburg - The national minimum wage will cost some jobs no matter how low it is set, but new research this week suggests the indirect benefits will more than compensate for the loss.
With the announcement of some kind of minimum wage almost certainly scheduled for next year’s state of the nation address, more and more sophisticated research about what it might actually achieve is pouring out of think-tanks and universities.
Researchers are as divided on the wage as labour federation Cosatu and business groups at Nedlac, the national consensus-seeking body between government, labour, business and civil society, where talks have in effect reached a deadlock.
One of the centres dedicating time and resources to this research is the Corporate Strategy and Industrial Development centre at Wits University.
This week, it released a new modelling exercise predicting what would happen if the wage were set at a variety of different levels – from very low to very high.
The modelling was done by Applied Development Research Solutions, an economic consultancy. Its researcher Asghar Adelzadeh presented his results this week at a seminar at Wits.
Compared to a “base case” where the economy runs as it has, all the scenarios lead to lower employment by 2020 compared with not having a minimum wage.
The damage is, however, according to Adelzadeh’s modelling, going to be minimal – something in the region of 65 000 fewer jobs five years from now.
The point is to look beyond the initial direct “supply-side” effect, where inevitably some employers will cut jobs or hire fewer workers due to higher wages.
Instead, the economic modelling tries to predict the indirect effect of having workers paid better, having them spend more and claim less through social grants due to graduating out of the means test for welfare support.
Adelzadeh ran the numbers for a hypothetical “minimal” minimum wage of R2 250, as well as a “maximal” R6 000, and two in-between scenarios with large discounts for farm workers, domestic workers and public works programmes.
The maximal wage will increase unemployment to 25.9% compared with 25.7% in 2020, while a low wage effect will not even show up in the rate.
The losses would be mostly in the construction, retail and hospitality sectors, Adelzadeh told City Press.
The gains, however, include higher GDP growth, with especially the manufacturing sector potentially getting a boost from higher domestic demand.
Personal taxes will increase, social grants will fall and poverty rates will drop faster than otherwise, the modelling predicts.
Even a minimal minimum wage raises Adelzadeh’s growth expectation from an average of 2% up to 2020 to 2.2%.
Predictably, it is mostly the services sector that stands to suffer from a high national minimum wage.
In the manufacturing industry, its sectors of food processing, as well as leather and footwear, could take a knock from losing cheap labour if the wage is relatively high, the modelling shows.
Cosatu this week threatened a nationwide socioeconomic protest if the outstanding issues around the wage were not sorted out in Nedlac.
Deputy president Cyril Ramaphosa, in a speech last week, claimed the talks were at an “advanced stage”, aside from the fact that there was no agreement on the level of the minimum wage or the mechanism through which it was supposed to be set.
These two issues are really the only important ones, creating the impression that nothing of substance has really been agreed to yet.
Cosatu has pushed for a wage of at least R4 500 or up to R6 000 – a level far above what most workers are currently paid. It is also fighting the apparent plan to have the minimum wage set by a technocratic panel of experts.
Nicoli Nattrass and Jeremy Seekings, academics at the University of Cape Town (UCT) with years of experience in labour market research, this week argued that the Employment Conditions Commission, which sets the country’s existing sectoral determinations, should also set the minimum wage.
A mechanism for this is already written into labour law.
Divorcing the wage from the commission’s process from sectoral determinations creates the possibility of a powerful new tool for mobilising for wage increases on a national and cross-sectoral basis by unions and other groups.
This week, Nattrass and Seekings wrote on Groundup.org.za that R2 700 looked more or less like the minimum wage South Africa would adopt if it followed international practice in setting it as a ratio to the average and median wages already being paid in the economy.
R2 700 is also the level where South Africa’s existing patchwork of 124 sectoral minimum wages tend to converge.
They, however, added that there was a case to be made for setting it lower in the interest of job creation through lower wages.
Arguments for a high or a low minimum have only one thing in common – the assumption that people are actually paid the minimum wage.
Haroon Bhorat, another economist at UCT, has been researching the extent to which employers have evidently ignored minimum wages.
A recently published update of this research, for instance, shows that there is a tendency to “partially comply” with hikes in minimum wages. Wages go up, but stay below the actual legal minimum in many places.
More than half of farm workers were paid less than the then minimum in their sector in 2007, according to his analysis of official labour force surveys up to that point.
It is not hard to get away with it either, as the probability of Cape farmers being visited by a labour inspector that year was 11%, according to Bhorat.