Unless reliable, quantifiable data is publicly available, reports cannot be taken seriously.
The employment report by Mike Schussler (“The unemployed are the real poor”, Mail & Guardian, June 8 to 14) has sparked a heated debate about labour cost development and wage levels for workers. At the core of the debate has been which data to use when discussing these issues.
Everyone knows that there are many problems with the official statistics. One problem is the sudden big jumps, or “structural breaks”, in the number of reported employees. The jumps distort accounts for unit labour cost and productivity development if not addressed. But there is one basic advantage with all the data from the Reserve Bank and Statistics South Africa: the data is publicly available and transparent. Everyone can assess it critically, use it and revise it according to established methods. If such efforts in their turn are made public and are transparent, there is common ground for a debate, despite all other differences of opinion.
As for the Schussler report, important aspects of the data are not transparent and publicly available. It is private property. The Alternative Information and Development Centre requested the wage data, some of which was provided, but not all. When we tried to obtain the material from the remuneration consultant, 21st Century Pay Solutions, it was refused.
In the part of the debate over unit labour cost development, Dick Forslund mistakenly assumed that Schussler relied on the original, unrevised data series of Statistics South Africa without making the necessary revisions, but he had done so. Nevertheless, the fact remains that Schussler’s account of real unit labour cost development differs completely from the bank’s account.
The bank’s account for real unit labour costs stretches back to 1990, providing a sound basis for policy discussions. The bank registers a fall in real unit labour costs of 10% for the period, but Schussler reports a stunning increase of 16%. You will find no explanation in the Schussler report for this sensational difference. It is as if the official accounts for key indicators and policy discussions based on this account do not exist. Of course, an explanation is still wanting.
In the debate about wage levels, Neva Makgetla (M&G, June 1 to 7), Chris Malikane (PoliticsWeb, May 16) and Dick Forslund (Business Report, May 31) all referred to the report “Monthly Earnings of South Africans 2010”. It is published in Statistics South Africa’s Quarterly Labour Force Survey.
In Schussler’s response to Forslund (Business Report, June 4), he accounts for an incredible doubling in the average salaries from R6621 a month in 2010 to R13284 a month in 2011 as having to do with the difference between constant and current prices, “a fundamental economic concept”.
If that difference depends on constant or current prices, it would mean more than 100% inflation between 2010 and 2011. The average wage of R6621 came from the Statistics South Africa report and its wage data. The R13284 a month average comes from the Quarterly Employment Survey based on questions put to employers.
Unlike the monthly earnings report, it includes the salaries of people who are employers themselves. The methodology between the two reports also differs. But, seemingly, the apartheid wage gap remains so great that when you include the salaries of top management, you get a doubling of the average compared with when only the wages of ordinary employees are included.
It is important to know the difference between median and average wages. The average wage is what you get after dividing the whole wage sum equally between everyone with a wage. A minority with very high wages increases the average significantly. The picture of the situation for the majority then gets distorted. To address this problem the median should also be reported. This is the wage of the person in the middle, also called the typical wage. Half of the employed population earns more and half earns less than the person in the middle, the person with the median wage.
The median or typical wage in the sample is close to R11 000 a month, according to Schussler. If this reflects the situation in the non-agricultural formal sector, then half of the employees typically earned R11 000 or less a month in 2010. But this is three times the typical wage reported by Statistics South Africa. According to its report, half of the formally employed workers were paid R3 683 or less in 2010. This is an extraordinary discrepancy between Schussler’s report and official statistics. The reader is expected to take more seriously a report that is not based on official statistics, but rather on a private remuneration consultant whose data sample is not available for public scrutiny. It is bizarre and lacks credibility.
Finally, the South African Revenue Service does not take samples. The tax statistics cover everyone who is formally employed and should resolve this dispute, even if not to the political satisfaction of everyone.
In 2010, 10.3-million people were registered to pay personal income tax. There are about 9.6-million formally employed in the non-agricultural sector, according to the Quarterly Labour Force Survey. When 91% of the tax forms had been assessed, nearly 3.5-million had a taxable income of more than R5500 a month. Conclusion: about six million formal employees earned less than that. The allowed deductions from total personal incomes do not change this picture at all. We are referring to public and transparent data. Everyone can critically examine the basis of what we say and the facts speak for themselves.
Of course, other reports point in the same direction. In its diagnostic report, the National Planning Commission reported that half of the formally and informally employed people earned R2500 a month or less in 2008.
In a ground-breaking study, Bhorat, Kanbur and Mayet also established the minimum wage violations in South Africa – out of all workers covered by this legislation in 2007, 45% were paid less than the minimum wage and, on average, they were paid 36% less than the legislated minimum.
Beyond the fantasies entertained by anti-union economists, there is no high wage regime in South Africa. A low wage regime is still firmly in place. This breeds unemployment by strangling domestic economic demand and it makes life extremely difficult for the majority. It must be recognised and torn up by its roots. Forward to a wage-led and sustainable growth path.
Diagram 1: Two completely different accounts for real unit labour cost development (Sources: The SA Reserve Bank research department (for the SARB index series) & Mike Schüssler (for the Uasa index series) − Index set to 100 in 1990 by AIDC).
Brian Ashley is the director and Dick Forslund a researcher at the Alternative Information and Development Centre in Cape Town