While South Africa enjoys a wealth of household and firm data that speaks to the evolution of the
labour market since the end of apartheid in 1994, the interpretation of these data is complicated by
a variety of measurement and fieldwork changes that have occurred over this time period. These
changes have been well documented by Wittenberg (2004, 2014), Casale, Muller, and Posel (2004),
and Yu (2007). One of the most dramatic changes that must be considered when examining
employment trends over this period is the apparent increase in self‐employment that took place
with the switch from the October Household Surveys (OHS) to the Labour Force Surveys (LFS). With
this change in survey instrument, there was a seeming increase in the number of self‐employed
agricultural workers from roughly 150 000 in the last wave of the OHS (October 1999) to more than
1.4 million in the first wave of the LFS (February 2000). The number of self‐employed agricultural
workers (SEAWs) drops somewhat after September 2000 but remains elevated throughout all waves
of the LFS compared to previous OHS waves and later Quarterly Labour Force Survey (QLFS) waves.
This series, calculated using the Post‐Apartheid Labour Market Series (PALMS) which combines all
three survey instruments—OHS, LFS, and QLFS—is shown in Figure 1.