OPINION | SA must face impact of minimum wage on jobs - News24

The number of people in a statistical sub-sample estimated to have lost their jobs due to the national minimum wage hike in 2024 is 86 000, while the best available estimate of total job losses due to the national minimum wage hike of 2024 is 228 000, and 430 000 since 2019. Facing these numbers should change how people talk about, research, and decide on the minimum wage, says Gabriel Crouse. 

Although there is clear evidence that the National Minimum Wage (NMW) has destroyed jobs, no one has bothered to report how many (until we did here, and here). Not reporting such estimates stifles public debate and increases the odds of bad decisions that lead to more job losses.

A telling example of how poor victims of the NMW policy get passed over is the recent News24 article titled A record hike in the minimum wage last year didn't trigger jobs bleed, study finds.

Carol Paton, one of the best analytical journalists in South Africa, reported that a recent paper by UCT's Development Policy Research Unit (DPRU), hired by the NMW Commission, found a 3.4% reduction in employment among low-paid workers.

But is "3.4%" very big, or very small? That should depend on what it really means.

The article in question called the 3.4% reduction small, noting that employment "reduced slightly" amongst a cohort of "low-paid workers", but did not report what the "3.4%" effect amounts to in absolute terms. It also did not include the number of low-paid workers so that readers could make that calculation themselves.

This approach followed the NMW Commission's lead. In its 39-page investigation report the commission repeatedly mentioned a "3.4%" reduction, which it described as "small", without giving a numerical sense of what "small" is really supposed to mean.

To fill the information gap, IRR Legal wrote to the DPRU and were sent back an "internal report" that was the basis of the NMW Commission's claims. This showed, alarmingly, that when the "3.4%" estimate is "[t]ranslated into an estimated number of jobs, this is equivalent to approximately 86 500 fewer people employed" out of a limited sub-sample.

To be clear, 86 000 is an eye-watering number of jobs to be destroyed in a single year by a single policy, namely the NMW. That is probably the biggest job loss tally from a single policy ever given to the state by experts it chooses to hire. Leaving that number out of its 39-page public report, and calling the vague "3.4%" figure "small", is a serious omission by the NMW Commission.

Worse still, the internal report indicates that 86 000 job losses is not even half the picture, since it only refers to 3.4% of 2.5 million "low-wage workers", whereas there are millions more "low-wage workers" than that.

To find out how many low-wage workers there are really estimated to be, and what the best estimate is for total job losses due to the NMW 2024 hike, IRR Legal had to correspond further with the DPRU asking if the 3.4% could be applied to the total universe of low-wage workers.

DPRU co-author Benjamin Stanwix wrote back: "One could reasonably extrapolate the findings (in percentage terms) to the larger group/s in each approach as you suggest, to obtain a more aggregate picture. This is a more speculative [sic], but we do try to show that our panel sample is a very good match to the broader (non-panel) population."

Stanwix further stated that there are "roughly 6.7 million" people in the "low-wage group in 2024".

In response, the following simple calculation was sent to check for accuracy, "So the disemployment effect for 2024 of 3.4% * 6.7 million = 227 800."

Stanwix replied, "Yes that sounds more or less right", reiterating the caution that this "reasonable approximation" only works to the extent that the DPRU's "panel subsample" "is generalise-able to workers not in the panel".

Happily, the "panel subsample" and those not in the panel were practically identical in all measured outcomes according to the DPRU internal report. As Stanwix said: "Our panel sample is a very good match to the broader (non-panel) population".

In summary, if anyone had bothered to ask what the "3.4%" applies to in terms of the most "aggregate picture" of reality that is currently to hand, then they would find themselves confronted by the figure of 6.7 million low-wage workers. 3.4% of that is roughly 228 000 estimated livelihoods (at least temporarily) destroyed due to the 2024 NMW hike alone.

The NMW Commission did not ask that question, and so never considered the total number of jobs lost due to the NMW, which is a direct violation of its duty under the NMW Act.

Even more brazenly, it withheld the 86 000 figure relating to the statistical subsample, leaving journalists who read all 39 pages of the report under the misimpression that 3.4% is "small".

But no one would call 86 000 job losses "small" or a "slight reduction", let alone 228 000.

The misimpression that NMW job losses are "small" is spreading at just the moment when the NMW Commissioners are deliberating about the annual adjustment of the NMW. The impact of a bad decision would be real, and likely harmful to tens or hundreds of thousands more low-wage workers, which can be multiplied further if one considers family dependents.

For some context on how job losses are usually covered, last week News24 ran the headline: "Steel industry fears catastrophic impact of ArcelorMittal plant closure".

In that case there is an estimated job loss of 3 500 directly impacted individuals and "more than 100 000 jobs at risk". This was correctly described as "catastrophic".

The estimated NMW job losses should be treated with the same direct appreciation of human loss. The 3.4% impact of 2024 is not even the worst of it, as DPRU's published estimates for the impact since 2019 range from 4.8% to 8.1% of low-wage workers, though the DPRU could not tell me how many low-wage workers that applies to, or how many job losses that amounts to.

However, the mean product of that range and 6.7 million is 430 000, an estimate for total job losses due to the NMW that is unfortunately sketchy, which must trigger more open consideration from experts and the public.

Such big numbers can be hard to think about. A total of 227 800 jobs destroyed in a year amounts to 4 380 jobs destroyed every week, or 624 destroyed every day, or 26 jobs destroyed every hour.

Take a minute to think about the poor workers and work seekers immiserated by the NMW.

Gabriel Crouse is executive director of IRR Legal

https://www.news24.com/fin24/opinion/opinion-sa-must-face-impact-of-minimum-wage-on-jobs-20250201