The primary source of employment data in South Africa is Statistics SA’s Quarterly Labour Force Survey (QLFS, Statistical Release P0211), which is a survey of the labour market activities of a sample of individuals aged between 15 and 64 years. QLFS data is subject to many deficiencies. The survey results are released between 4 and 6 weeks after the end of each quarter, with the result that the data are a belated guide to underlying economic and employment trends.
The QLFS samples 30,000 households per quarter and extrapolates the results to an economy comprising 49 million people. Consequently some of the survey results are unreliable. For example, estimates of total employment have a statistical confidence value of around 34%, and four out of nine provinces have confidence values around 40%. Additionally, Statistics SA data do not distinguish between increasingly important non-traditional employment types, such as contract, temporary and agency employment amounting to approximately 3.9 million employees (30% of the workforce). Also, Statistics SA’s methodology is subject to more-or-less continuous revisions, which makes the identification of consistent long-term trends impossible. Information on household incomes was simply not collected between 2008 and 2009. Employment estimates exhibit inexplicable “jumps”, most recently in 2008 but also in 2004 and 2006, with the result that a consistent long-term picture of employment in South Africa is impossible.
Despite these differences, Adcorp’s and Statistics SA’s estimates of formal sector employment correspond fairly closely, with a correlation of 85%. However, Adcorp’s estimates of informal sector employment – based on an accredited, peer-reviewed methodology, the so-called “currency demand” method – are substantially higher than Statistics SA’s estimates. Statistics SA’s figures are loosely based on the original 2001 population census, which does not account for the estimated 2-3.5 million illegal immigrants since that census was conducted. Also, the figures do not fully account for the rapidly growing proportion of South African economic activity that is conducted in the “unofficial sector”, largely to evade income taxes and circumvent labour laws. Unfortunately, until the 2011 population census is completed, it is impossible to state definitively which figures are correct.
Adcorp, South Africa’s largest diversified employment services company, possesses many rich sources of fine-grained data, both within the group (permanent and temporary placements, job search times, work applications, etc.) and from industries and sectors in which the group operates (skills development levies, unemployment insurance claims, labour relations cases, etc.). Although Adcorp is not fully representative of the South African economy, Adcorp’s data is readily available (in most cases with no material delays in data collection) and reliable (based on more than 1 million labour market transactions per year).