Exploring Dynamics in South African Firms

South Africa has a huge unemployment problem and South African policymakers, as in many other countries, believe that the ability of small firms to create employment is a major part of the answer to this problem. But, in contrast to the situation in high and middle income countries, the reality is that nothing is known about the patterns of job creation and destruction in South Africa. What little we do know about firms is the result of small cross sectional surveys funded mainly by the World Bank.

Since 2005 Statistics South Africa has collected enterprise employment data using the Quarterly Employment Survey (QES) but no work on dynamics at the firm level has been published using this data. The researchers have obtained a version of the QES from Statistics South Africa and their first aim is to use this data to undertake an analysis of job creation and destruction in private enterprises, including through entry and exit of enterprises. This study would open up a whole new research area in South Africa and could potentially radically transform the way policymakers think about job creation in particular.

The team's second objective is more exploratory in nature. They aim to find older firm survey data on the archive servers of Statistics South Africa and the Human Sciences Research Council (HSRC). Statistics South Africa conducted detailed firm surveys until the mid- 1990s. These included manufacturing censuses conducted every three years from at least 1972 until 1996 that included detailed questions on employment, costs, inputs and outputs. These data, if found and retrieved, would provide a window onto manufacturing firm dynamics over a fascinating period in South Africa’s history that would also be relevant to many LICs today. The researchers will explore whether this older data is indeed available on the Statistics SA server and then retrieve it if it is indeed available. The HSRC has also given the team permission to explore their electronic archive to find the Manpower Survey data, which were undertaken from the 1960s until the early 1990s. Using this data it would be possible to explore how firms responded to changing macroeconomic conditions, describe patterns of job creation and destruction and explore how firms’ employment patterns, particularly their occupational employment breakdowns, altered as a result of macroeconomic shocks. If successful these efforts would open up a totally new field of investigation in South Africa, analysis of firm level dynamics. This is crucial to policy makers thinking about how to create jobs in a country with a massive unemployment rate.


Andrew Kerr

University of Cape Town

Martin Wittenberg

University of Cape Town