Job axe doesn’t fall equally
Results of the Nids-Cram survey show job losses have not been uniform: groups who have historically had worse labour market experiences have borne the brunt of the pandemic downturn
What happens in the SA labour market has an impact on a number of welfare measures, including poverty and inequality. This means that understanding the dynamics in the labour market is key to understanding how SA as a society is adapting and evolving in response to Covid-19.
What has been the impact of the pandemic and lockdown? At present, we simply don’t know with certainty, due to a lack of data. But we have analysed the first wave of data from the National Income Dynamics Study: Coronavirus Rapid Mobile Survey (Nids-Cram).
The survey asked respondents, aged 18 to 59, about their employment status in February and April, as well as related information about hours of work and earnings. Most of our analysis is thus focused on measuring changes between February and April — giving us a picture of how different things were just before lockdown, and during level 5 of lockdown.
Unsurprisingly, a very high proportion of those surveyed either lost their jobs or were furloughed in April.
In our sample, the proportion of adults who were employed decreased from 57% in February to 48% in April, using the conventional definition of employment.
Employed individuals in the black/African race group had a 43% chance of losing their job against 17% for white workers
However, the conventional definition is not adequate for the current moment, as it classifies people who are temporarily absent from work as employed. Under "normal" circumstance this is fine, as it affects only a tiny fraction of people. But circumstances are not normal.
We created an additional variable for employment, classifying furloughed workers — those who lost their income, but would still have a job to return to — as "not employed". Using this alternative definition, we estimate that the proportion of adults who were employed decreased from 57% in February to just 38% in April. Put simply: almost one in three of the adults in our sample who were employed in February were not working by April.
The impact of this substantial loss in employment was negative for everyone — but it was more acutely felt in groups who were already at a disadvantage in the labour market. Among those black African people who were employed in February, 43% had lost their job, using the alternative definition, by April — against 17% of employed whites. In contrast, 15% of white people found employment between February and April, against only 7% of black Africans.
Similar job-loss differentials are observed for women relative to men (44% vs 35%); for young people against prime-aged adults; and for people with at most a matric education against those with at least some tertiary education (44% vs 29%).
The overall point is that the pandemic and lockdown will have further widened the already substantial inequality in the local labour market.
We also investigated whether low-wage workers were especially affected by the unprecedented increase in unemployment. As is evident in the graph, low-wage earners were much more likely to lose their jobs compared with high-wage earners.
Of low-wage earners who earned less than R3,000 a month, 38% lost their jobs between February and April — about two in every five workers in this group. For those in the two middle-earning groups — R3,000-R6,000 and R6,000-R12,000 — this number was still high, at about 19%. In contrast, the higher-earnings groups experienced relatively low levels of job loss, at just over 5%.
When we consider the alternative definition of a job loss, which includes people who were temporarily absent from work, all of these proportions increase. Again, however, the increases are more prominent among low-wage earners.
What it means:
Almost one in three of the adults in the sample who were employed in February were not working by April
Combined, these dynamics have substantial implications for both poverty and inequality. They also greatly reinforce the argument that the state should provide rapid social assistance to families in need.
There are several important questions for which we still have no answers. These include questions about wage scarring, early retirement, the challenges of youth unemployment, returns to education, and which groups of people manage to exit unemployment and which seem to fall into unemployment traps. We also don’t yet know which sectors of the economy have been most affected by job losses, and, of these, the extent to which that loss is permanent or temporary.
Finally, until, and unless, a better data set is released, Nids-Cram is the only option available for researchers who use micro-level survey data to analyse the labour market. We are confident that policymakers will take these findings into account when providing relief to firms and families.
*Ranchhod is deputy director of the Southern Africa labour & development research unit and professor in the school of economics at the University of Cape Town; Daniels is an associate professor at UCT’s school of economics, and a former principal investigator of the National Income Dynamics Study. For more information on the Nids-Cram survey, visit http://www.cramsurvey.org
Would you like to comment on this article or view other readers' comments?
Register (it’s quick and free) or sign in now.
Please read our Comment Policy before commenting.