Skip to content

Trends in The South African Labor Market

Literature › Works

Jesse Naidoo (2012). Trends in The South African Labor Market. Presentation slides, University of Chicago (dated February 10, 2012; footer cites 2011). University of Chicago.
Type conference paper
Scope Post-1994 South African economy
Method descriptive quantitative
Period studied 1970–2009
Regions South Africa
Themes labour markets wages, inequality poverty unemployment, methods data sources, demography urbanisation, education human capital, bantustans removals influx, origins segregation, transition post apartheid
Citation key naidoo2012

Argument

The slides give a brief history of racial labour-market restrictions and then describe trends in population, education, participation and unemployment from census data (1970-2001) and household surveys (OHS, LFS, QLFS). They report that African participation and unemployment rose sharply after apartheid. The author argues that part of the post-1999 participation spike may be a measurement artefact from the switch from the OHS to the LFS. To test this he calibrates a reclassification matrix between labour-market states, with zero restrictions and random starting points, and splits the observed change into 'real' and 'measurement' components. The data only partially identify this split, and the aggregate data are too noisy to show obvious policy effects.

Evidence and method

Apartheid-era censuses (1970, 1980, 1985, 1991) and the 1996 and 2001 censuses, tabulated by race, sex and education. Post-apartheid October Household Surveys (1994-1999, about 30,000 households), Labour Force Surveys (biannual to 2007) and the QLFS (2008 on), reweighted with the Branson and Wittenberg cross-entropy weights calibrated to the ASSA demographic model. Secondary literature is cited on bargaining councils, cash transfers and the 1995-2000 income declines.

Findings

  • The 1913 Natives' Land Act confined African land ownership to a small share of the country. (p. 6)

    1913 Natives’ Land Act: designated approximately 7.5% of the country’s area in which Africans were allowed to own land.

  • Verwoerd's statement on Bantu education shows its deliberate design to limit African occupational prospects. (p. 9)

    There is no place for [the Bantu] in the European community above the level of certain forms of labour

  • Population growth was rapid but slowed sharply by 2009. (p. 21)

    Average growth rate: 3.4%, though this has slowed to about 1% in 2009.

  • The 1980 census cannot be used to estimate unemployment because of how occupation was recorded. (p. 27)

    Thus estimating unemployment rates is impossible because employment and unemployment cannot be distinguished.

  • Census population counts for 1980-1991 show implausible swings even after adjustment for undercount. (p. 28)

    the 1985 Census, for example, appears to show a drop of 14% from 1980 to 1985 and then a rebound of 30% from 1985 to 1991.

  • The coverage of the 'independent' homelands varies across censuses, which compromises comparability over time. (p. 30)

    homelands not covered by 1985 Census; only some covered by 1970.

  • African participation rose strongly after apartheid, especially among women. (p. 36)

    African participation rates have risen dramatically, particularly for women

  • African unemployment peaked in the early 2000s at very high levels, higher for women than for men. (p. 37)

    African unemployment rates have also risen to a high of around 40% for women in the early 2000s, and around 30% for men in the same years.

  • The participation spike coincides with a change of survey series, which suggests a measurement problem. (p. 53)

    The spike coincides with the end of the OHS series and the beginning of the LFS series.

  • The total observed change in participation between the 1999 OHS and the 2000 LFS is large, and the exercise tries to split it into real and measurement components. (p. 75)

    Observed (total) changes: +8.3% (African men), +10.3% (African women)

  • African employment in the 1995 OHS looks anomalously high, so how much of the rise in unemployment is real remains an open question. (p. 80)

    African employment in the 1995 OHS seems unusually high