Displaying more than a good dose of confirmation bias, Mike latches onto incompatible data sets that fail to correlate scientifically with Quarterly Labour Force Survey results to advance an inconsequential argument about minimum wages in South Africa.

It is not the first time that Mr Schussler has compared administrative (e.g. tax) and survey data (e.g. Labour Force Survey) collected by different agencies for different purposes.

In allegedly debunking what he calls the “South African wage myth,” Schussler resorts to, for example, extrapolating data on the number of cars moving around the country and the number of people working in the formal sector to make inaccurate inferences about averaged wages paid to working South Africans.

In his broader observation, Schussler questions what he regards as the silence of the informal sector and big business in the national discussion about the national minimum wage, ignoring the complexities of the informal sector in particular and the difficulties of measuring the sector.

Needless to say, the informal sector can be influenced by the minimum wage, particularly when unemployment levels are low. Informal firms have to compete with formal sector firms for labour – known as the “light house effect.” It is less clear whether or not the informal sector will experience upward movement in minimum wages when there are high levels of unemployment. This is something researchers are investigating in the South African context with history of exclusion and persisting trust deficit.

Back to the numbers: Schussler’s argument that median wages are around R5,100 – as opposed to R4,800 suggested by the Labour Force Survey results – does not explain what the cut-off point was for the calculation from Stats SA data or reveal the profile of those used in the calculations.

Methodology always influences the measurement, so we need to understand exactly how Schussler went about his assumptions.

While the formal sector may be paying certain classes of workers fairly well, sources in the comparisons may exclude those masses of South Africans in the informal sector that earn significantly less and are often referred to as the working poor – the class of workers targeted by national minimum wage policies.

The use of consumer credit data to compare with the Quarterly Labour Force Survey is highly precarious, as the decision to procure credit is often based on the level of confidence and earnings for those who ask for credit.

This leaves out many people who won’t even take a step to look for credit because low wages and high commitments result in low consumer confidence. Such data was never intended to measure wages while results of such data can be derived in relation to credit regulations.

Debt counsellors’ data cannot be compared to the Quarterly Labour Force Survey because the primary objectives and measurement indicators are totally different.

Throwing around – or together – different data sources to justify belief in some quarters of our economy that South African workers are doing much better than their counterparts elsewhere and that the cost of labour is an inhibitor of investment, adds no value or closure to the current discussion on the monetary level at which the national minimum wage should be pegged.

The problem with Schussler’s analysis is that he is comparing apples and oranges. There is certainly a possibility that some South African workers are earning more than they report. They may be better paid if for example they are not taking in-kind payments into account when reporting their earnings.

However, most academic work in this area has focused on relatively well paid workers.

The findings for SA are largely consistent with international evidence which finds that wealthier households are the ones most likely to under-report earnings (possibly because of the impact that pension contributions, garnishee orders or other factors on take home pay). Before the findings for wealthier households can be generalised to poorer households, it is critical to understand the reason for the under-reporting.

It is not clear that under-reporting for miners, who typically earn R7, 000 a month according to Stats SA data, would be the same as any under-reporting that might occur for domestic workers, whose minimum wage is R1,993.82 a month in rural areas.

To understand the impact of under-reporting, in-depth, rigorous analysis comparing the reported earnings of low paid workers with their actual take home pay is required.

The data sets Schussler uses to try make this comparison are not suitable for understanding wage income since they place a larger weight on wealthier households and may exclude a number of low-paid workers.
Another important point is that people who get credit (NCR data) are not the same group of people as those who work (QLFS data). Credit providers tend to give money to people who have regular jobs and prefer to give it to those with more money.

Furthermore, the NCR’s income estimates are based on individuals’ self-reported incomes and include not only wage data, but also other income sources (including the wages of a joint applicant), so estimates of income are likely to be higher than Stats SA’s data which only includes wages. Individuals’ incomes are reported within broad income bands, which make it impossible to calculate median or average wages from the NCR data, making it unsuitable for an analysis of the minimum wage.

The National Treasury Budget Review of 2015 notes there are 7,024 million individuals with taxable income of over R70,000 per annum. This group includes not only workers, but also pensioners. Again, these individuals’ income includes wages, as well as investment income, royalties and capital gains tax applicable to individuals, so the figures will be higher than wage data, and cannot be used to draw an estimate of median worker wages. Further, firms are obliged to include all employees regardless of whether they are in the tax paying bracket or not.

Many more numbers need to be crunched, with the necessary levels of transparency and integrity that will allow our nation to derive not just at an appropriate number, but at a credible socio-economic outcome that will benefit all sectors of our society and the economy in general.

So far, Mike Schussler’s reading of the status quo doesn’t help this quest in any way. The continuing consultation on the national minimum wage is very historic for South Africa and requires facts and scientific methods that can stand the test of time.

Even before this rebuttal was submitted to the editor, a Fin 24 (10 February 2016) headline dropped; “Is state delaying job stats to avoid Budget headache?” The story quotes Schussler arguing that Stats SA’s postponement in publishing the latest Labour Force Survey is suspect and maybe engineered by government in pursuit of predetermined political ends.

Most unfortunate; especially because Stats SA issued a statement citing data technicalities as reason behind the delay. The fact of the matter is that Stats SA always release reports on stipulated dates unless there are compelling reasons. If indeed the work of these renowned professionals was influenced by the political calendar, as Schussler insinuates, why would they have released the 2015 results a day before the State of the Nation Address? Again, Schussler’s calculus doesn’t add up. DM

Ngcaweni works in The Presidency.