Unemployment statistics: StatsSA briefing

Meeting report

The Chairperson opened the virtual meeting, welcoming the Members, support staff as well as the delegation from Statistics SA (Stats SA). She also acknowledged the State of the Nation (SONA) that would be happening on the following day.

She acknowledged the recent passing of Mr N Hinana’s (DA) mother and gave her condolences. She encouraged other Committee Members to reach out to her via Whatsapp, text messages or phones call to convey their condolences too.

Mr Z Sakasa, Committee Secretary, took roll call, reporting the absentee apologies.

Briefing by Statistics SA: Quarterly Labour Force Survey (QLFS)
Mr Resinga Malukele, Statistician-General (SG) and Head of Statistics SA (Stats SA), greeted the Chairperson and Members. He indicated that he would be doing the presentation.

As a precursor, Mr Maluleke indicated that the lockdown restrictions necessary to combat COVID-19 created an obstacle to normal data collection approaches and operations, exactly when there was a massive increase in the demand for information. Stats SA changed the mode of collection for QLFS data to Computer Assisted Telephone Interviewing (CATI). To facilitate CATI, the sample that was used for QLFS during the first quarter (Q1) of 2020 was also used in second (Q2) and third (Q3) quarters. Not all dwelling units on the sample had contact numbers resulting in data being collected from part of the sample where contact numbers were available for QLFS of the third quarter. This introduced bias in the estimates. Details on how the bias adjustment was done were included in the report.

In Q2 of 2020, statistics showed that there were 14.1 million employed individuals, while about 4.3 million were unemployed, 2.5 million were discouraged work seekers and another 18.1 million were not economically active. There were large movements in all these categories during the third quarter: 14.7m were employed, 6.5m unemployed, 2.7m discouraged work seekers and 15.2 were still not economically active. Stats SA also showed that out of a total of 39.2 million within working age (15-64 year olds), the labour force consisted of 21.2m and the other 17.9m were economically inactive. However, there were 543 000 more people employed in Q3 than Q2 of 2020. South Africa's official unemployment rate stood at 30.8%, reflecting an increase by 7.5% between Q2 and Q3. The expanded definition, which includes discouraged work seekers and those having other reasons for not job searching (e.g. lockdown) showed a 43.1% unemployment rate, with a 1.1% increase Q3:2020 from Q2:2020. Eastern Cape recorded the highest official and expanded unemployment rates. KwaZulu-Natal, Northern Cape and Limpopo provinces have more than 20% difference between their expanded and official unemployment rates.

Unemployment among the black African population group remains higher than the national average and other population groups. Black African women are the most vulnerable with an unemployment rate above 36.0%. Irrespective of gender, the black African and coloured population groups remain vulnerable in the labour market.

Of the 14.7 million persons who were employed in Q3: 2020, 73.2% were expected to work during the national lockdown by the companies/organisations they work for. About 87.3% of the employed continued to receive pay during lockdown; 18.9% of those who received pay during lockdown were paid reduced salaries. Close to 90% of graduates received their full salary in Q3:2020. The share of those receiving full salary increased irrespective of level of education between Q2:2020 and Q3:2020. Almost 21% of those with less than matric received reduced pay.

See presentation document for more details

Discussion
The Chairperson commented that the presentation was interesting and challenging, before opening the floor for Members to engage it.

Mr M Bagraim (DA) thanked the Chairperson. He then thanked Stats SA, saying that it was a wonderful experience listening to them. Is there an explanation as to why the Western Cape is so much further ahead in employment than all the other provinces? The gap is so small. In other words, once people have found a job they do not give up looking or they find a job quite quickly. What is the Western Cape doing right and how can it be recreated in other provinces?

He also commended Stats SA for using the telephone as a way to get data, rather than throwing its hands up and saying: “well, gathering data is not possible.”

The figures are skewed quite radically because many people who are unemployed or who have given up looking for employment do not have telephones, data or any telephonic activity. Can Stats SA give the Committee an estimation of the percentage of those who are unemployed or have given up looking for employment, if people in villages or townships were actually visited?

Ms C Mkhonto (EFF) thanked the Chairperson and welcomed the presentation. She was concerned about what the presentation presented as ‘discouraged work seekers’. Maybe one should look deeper into why they are discouraged. It might be a matter of someone having R10 and preferring to buy bread over buying data. “Discouraged work seekers make up a high percentage, but we do not have all the reasons why people are in that category.” This presentation shows that there are many unemployed people, but the Department alone cannot be able to address all the challenges that are mentioned in the presentation. Maybe the Chairperson can advise the Committee on how to develop an inter-ministerial team with the President, after the COVID-19 resurgence. All relevant and affected departments can be included. This is a ticking time bomb, and an intervention is needed.

Dr M Cardo (DA) asked whether Stats SA has done any research into the effect of minimum wage increases on the employment profile in South Africa. In other words, has the increase affected jobs and job losses? Has Stats SA done any forecasting and projections around total job losses due to COVID-19? Has Stats SA given any thought or done any research on what it thinks the overall employment profile will look like at the end of this calendar year because of COVID-19 and the associated lockdowns?

Mr M Nontsele (ANC) asked whether Stats SA could talk more to underemployment. Underemployment talks to people who are employed but whose skills are not recognised or people who are highly skilled but get low wages.

Mr S Mdabe (ANC) asked if Stats SA has done a survey to find out whether the high rate of unemployment is because of population increase, retrenchments or an increase in the number of skills required in relation to growing the economy.

The Chairperson asked if Stats SA looks into the same kind of information that institutions look at – for example, how many people are trained in a particular skill or field, or which skills employers are interested in.

Does the category of ‘unemployed’ include people who are self-employed? Are these questions in the scope of Stats SA? Maybe it is important for these questions to become within the scope of Stats SA because of the difficult environment of COVID-19. It is the Committee’s responsibility to create a conducive environment. What is frustrating is that due to the pandemic, Stats SA could not interact with young people.

Responses
Ms Malerto Mosiane, Acting Chief Director: Labour Stats, Stats SA, first responded to Mr Nontsele’s question. She said that Stats SA does collect data on earnings. Information on which occupations people are in and how much they earn is also published. Stats SA only focuses on time-related underemployment in its reports – looking at people who work 35 hours or less but who wish to work for more hours and have taken steps to look for extra work. The proportion of these people is not large, at about 25% of time-related underemployment. When Stats SA looks at skills, it only uses occupation as an indication of skills. People that work in high-skilled occupations do earn more than those in low-skilled occupations.

Ms Mosiane responded to Mr Mdabe’s question, saying that Stats SA sees an increase in the population every quarter. Therefore, there is a quarterly increase of working-age people. There are also some people who leave the working-age population – primarily the people who turn 65 in a particular quarter. Stats SA sees a lot of new entrants into the labour market in the first quarter who were not in the labour market in the previous quarter, such as young people who have just graduated. There is a high percentage of people struggling to find employment, mainly women and young people.

Ms Mosiane then responded to Ms Dunjwa’s question. Stats SA does not go to institutions, but does go to where people live to ask questions about employment, such as if they have looked for work or what qualifications they have. Stats SA does have that information, but it gets it from the respondents themselves, not from institutions of higher learning.

People who are self-employed are included in Stats SA’s estimates of employment.

Mr Malukele also responded to the Chairperson’s question about institutions of higher learning. He said that these institutions could tell Stats SA how many people they trained in the previous year, but they could not know how many of those trained people got employment after graduating.

Stats SA also has the Quarterly Employment Survey, where Stats SA goes to the factory gates. The business register, which has all the businesses listed, is from the South African Revenue Services (SARS) and gets updated every month. Stats SA goes to factories and finds out how many people are employed and in which areas they are employed. This information also helps with understanding wages and salaries.

Mr Malukele then responded to Mr Bagraim’s questions. In all provinces, except the Western Cape, people say that reaching workplaces is difficult because they are far away. People in Limpopo say that work is in the city centres, but they run out of money and travelling into the city centres is thus not financially sustainable. Whereas in the Western Cape, people say that even if they are far away from urban centres, they are able to access farms and other workplaces where they can readily work.

With regards to telephonic interviews and potential bias, he indicated that bias was eliminated. The last time information was collected in person was in quarter 1 of 2020. After that, the only people who could be reached were those with telephones – about 65% of the number of people reached in person previously. The percentage was lowest in the Northern Cape at 43%, and the highest number was in Mpumalanga. Response rates ranged from 73% to 86% in different provinces. A process called Data Confrontation was used to eliminate bias.

Stats SA has been in correspondence with other countries, discussing different statistical methods. Mr Malukele said that later on in the day he would be presenting to the United Nations (UN) on the impact of COVID-19 on data collection. The International Labour Organisation (ILO) was very helpful in providing guidance in how to deal with the situation.

Mr Malukele responded to Ms Mkhonto’s questions. The numbers of discouraged work seekers keeps increasing. What are the reasons? This is similar to the situation that Mr Bagraim mentioned. There are fewer discouraged work seekers in the Western Cape because people can reach work centres more easily than people in other provinces. In Stats SA’s General Household Survey, it can be seen that the number of informal settlements is increasing. They are increasing the fastest in Gauteng, followed by the North West, followed by the Western Cape. These places are where households are increasing the most. People want to be nearer to workplaces. Every morning, people drive to their workplaces and they see people at street corners. Those people come to look for employment so that they do not become discouraged work seekers.

Mr Malukele responded to Dr Cardo’s questions, saying that Stats SA deals with outputs. Input indicators are budgets where people spend money. Process indicators are the rate at which objects are produced per hour or per day, for example. Output indicators are the number of people who access those objects. Stats SA works at the level of outputs – the number of people who say they can access jobs. Impact involves the changes in the life circumstances as a result of the outputs. That work is not what Stats SA focuses on. The Human Science Research Council (HSRC) does that kind of research; this is why Stats SA cannot and does not forecast.

Stats SA only forecasts on media estimates, with a five-year horizon. Populations have a particular causality that they follow. A pandemic will show over time. The other areas are not really suitable for forecasting. For example, unemployment was sitting at about 30.1% in quarter one. In quarter two, because of COVID-19, it dropped to 23.3%. No amount of forecasting could have seen that coming. People went and sat at home and did not look for employment. Commentators said that unemployment would be very high because of lockdown, so Stats SA said it would go and measure.

The other issue in relation to forecasting is that labour markets form part of the economy and impact what happens in the economy. The economy is affected by internal and external forces. If Stats SA were to forecast something about unemployment due to COVID-19 and something drastic happened in the economy, Stats SA would have missed it. This means that when the actual numbers come, Stats SA would shy away from making them available. If that were to happen, Stats SA would lose its independence. This must not happen. Iceland once lost its independence in the 1980’s, resulting in lots of challenges in its statistics system.

The Chairperson said that she is aware of some employers who approach institutions of higher learning and ask for people with specific skills. Would it not be advisable for Stats SA to devise another way, which is not via the telephone, to reach people? When Stats SA talks about the Western Cape, is it talking about the City of Cape Town or about the province at large? There are rural areas in the Western Cape. Are they included?

Mr Malukele responded that the approach is more about refining administrative records to assist Stats SA in making its official statistics much stronger and rigorous. Data gets put through the South African Quality Assessment Framework (SAQAF), which is housed in Stats SA’s agency. Stats SA graduates present poor statistics if they are poor. The stages of statistics are: poor, manageable, quality and official.

Stats SA does not collect information on cities alone. Information is collected from all South Africans. Stats SA’s samples must include all settlement types, such as villages, townships, suburbs, informal settlements and farms. If anyone was excluded, the data would be biased and would not be allowed to be published.

Mr Malukele thanked the Chairperson and the Committee for the opportunity.

The Chairperson thanked the Members, support staff and the Stats SA delegation for attending the meeting.

The meeting was adjourned.