The economist's employment thesis is based on blatantly inaccurate wage data, writes Neva Makgetla.
Anyone who has worked in wage negotiations will know that the figures given for pay in the “United Association of South Africa 11th Employment Report”, written by Mike Schussler, are simply ludicrous.
Schussler is an economist known more for his ability to get press coverage than for academic rigour, a quality only too visible in this report.
The core of his argument is that the average South African worker earns the equivalent of $30 000 a year, that is, R210 000 a year or almost R20 000 a month. On this basis he argues that high pay prevents employment creation, which in turn condemns South Africa to massive unemployment and real poverty among the unemployed.
But his figures for wages are not borne out by any reputable source. Compare his figures to the finding by Statistics South Africa that the median employee earned R2800 a month in 2010 and the median formal sector employee just R3683 a month. Similarly, surveys of actual wage settlements by the Labour Research Service found that minimum negotiated wages averaged R3162 a month in 2011, ranging with one exception from just less than R2000 to just less than R5000 a month.
So how did Schussler come up with such inflated figures? It is simple: he relied on private, for-profit sources designed not to analyse national wage levels, but to help human resource personnel in large organisations to design packages especially for top managers and professionals. These sources essentially rely on self-reporting by large companies and institutions to establish benchmark packages. Invariably, the result is heavily inflated figures that, even when they concern common occupational categories, focus on the best-paid personnel.
Specifically, Schussler relied on this kind of survey as conducted by two companies: 21st Century and Hays. Both obtain their information by asking participants to send in their remuneration details in return for access to a database that consolidates the returns of all the participants.
Hays does not specify in its publications how it gets its information and, in particular, how many organisations participate in South Africa. For its part, 21st Century said 700 organisations reported for its survey. This may sound substantial, but more than a million public, private and close corporations submitted reports to the company registrar, Cipro, in 2010-2011.
Moreover, participants in private, web-based surveys are neither re-presentative nor randomly sampled. They are overwhelmingly large and formal and therefore likely to pay more. And because most customers are interested primarily in setting pay for higher-level employees, the reliability of reporting on remuneration for ordinary workers is questionable.
In contrast, the official Quarterly Labour Force Survey covers workers in 30 000 households, chosen at random and weighted according to standard techniques. In other words, it is qualitatively more inclusive, representative and objective than personnel surveys conducted by remuneration companies.
Statistics South Africa also publishes the Quarterly Employment Survey, which has figures on employment numbers and costs outside the agricultural formal sector based on a survey of 20200 employers. Again, it makes a conscious effort to ensure a random, representative sample. It finds that the cost of employment, including pension and health benefits, averaged R13300 a month outside the agricultural formal sector in November 2011. Of this, about one-third is probably pension and health benefits.
Statistics South Africa’s employer survey clearly finds far higher median pay for formal workers than the labour force survey. In part, it is because the figure includes benefits and because where pay is deeply inequitable, as in South Africa, the median will be substantially below the average (the median is the point at which half of all workers earn less). In addition, the higher figures in the quarterly employment statistics may reflect problems in how employers report remuneration costs.
Still, even the higher figure for formal employees in the quarterly employment statistics is less than half of what Schussler claims as the average wage for all workers in South Africa. And he explicitly talks about wages, not the full package.
Schussler’s unfortunate choice of data sources also leads to his peculiar conclusion that pay inequalities in South Africa are not particularly high by global standards. In contrast, as the graph shows, using standard International Labour Organisation measures, South Africa has one of the most unequal pay scales in the world. The history of apartheid means it is unsurprising.
Schussler contends that public sector pay is higher than private pay. Again, the data he uses is suspect. But his analysis also ignores the fact that most public sector workers are professionals and semiprofessionals. Almost half of public servants are health and educational workers and another fifth are soldiers and police. This means that most public servants are relatively well educated and thus naturally earn more.
According to the Quarterly Labour Force Survey, government employees, including municipal workers and national and provincial public servants, constituted about one-fifth of total employment in 2010. More than one-quarter had a tertiary degree and more than half had completed secondary school. In contrast, only one private employee in seven had a tertiary degree and just less than a third had completed secondary school.
Public servants with a tertiary degree – mostly teachers and nurses – earned about half as much as private sector workers with degrees. In contrast, public servants who had not completed a degree earned 15% to 20% more than private sector workers with similar educational levels.
Finally, Schussler argues that employment in South Africa peaked in 1988 and then fell until 2004. He says this proves that the purported excessive pay for lower-skilled workers constrained employment creation in the 2000s. But he gives no sources at all for his employment data. Because official data under apartheid excluded blacks almost entirely, it is not simple to figure out long-term employment trends. An analysis of census data as well as detailed assessment by authors Jeremy Seekings and Nicoli Nattrass show that employment started falling in the late 1970s and only recovered with the transition to democracy in 1994.
Anyone reading the Schussler report would be forgiven for suffering from a sense of cognitive dissonance. Surely, no one can live in South Africa today and think that unskilled workers earn the equivalent of a lower-middle income-earner in the United States or Europe?
The explanation is simple. Schussler’s report is best understood not as a product of serious research, but rather as an interesting social artefact that demonstrates how deep divisions and inequalities still shape South Africa.
Yes, if you spend your days in Sandton City or Melrose Arch in Johannesburg (and you are colour-blind), you might believe that ordinary South Africans are part of the global middle class. But that illusion will not last long if you spend even a modicum of time in Soweto, the Cape Flats or the downtown area of any metro in the country.
The fact is that many employed people in South Africa barely earn enough to get by. For many, even a full-time job is not enough to lift them out of poverty. No amount of words will change this.
Neva Makgetla is the deputy director general for economic policy at the economic development department