Skip to content

Analysing Wage Formation in the South African Labour Market: The Role of Bargaining Councils

Literature › Works

Haroon Bhorat; Carlene van der Westhuizen; Sumayya Goga (2007). Analysing Wage Formation in the South African Labour Market: The Role of Bargaining Councils. School of Economics, University of Cape Town (DPRU).
Type working paper
Scope Post-1994 South African economy
Method econometric
Period studied 1924–2005
Regions South Africa
Themes labour markets wages, unions strikes, inequality poverty unemployment, transition post apartheid, methods data sources, origins segregation
Citation key bhorat2007a

Argument

The paper estimates the wage premium associated with industrial/bargaining council coverage, distinct from union membership, in 1995 and 2005. It builds coverage estimates by matching council scope data to household surveys. Descriptively and in OLS and quantile regressions, it finds a persistent union premium of roughly 17-23 percent. The 1995 council premium was small and applied only to non-unionised members. By 2005 private-sector councils conferred no premium, while the newly created public-sector councils carried a premium of about 28 percent. Overall council coverage doubled but only through the public sector, while private-sector coverage stagnated. Section 2 also gives a short institutional history of the dual, race-based industrial relations system from 1924 to the 1990s.

Evidence and method

Godfrey's 1992 Industrial Council Digest and Government Gazette extension notices, the AWARD database and council websites, matched to the 1995 October Household Survey and the September 2005 Labour Force Survey. It also draws on secondary literature (Godfrey, SALDRU, Butcher & Rouse, Standing et al.) for the historical overview, and on Godfrey et al. (2006) for extension and exemption data.

Findings

  • Council coverage of formal employment doubled between 1995 and 2005, but private-sector coverage did not grow. (p. 23)

    in 1995, 15 percent of those in formal employment (almost 1,2 million workers) were covered by bargaining council agreements. This more than doubled to 2,5 million workers or 32 percent of the formally employed in 2005."

  • Over half of 2005 coverage was in government departments; excluding government, coverage was 13 percent. (p. 22)

    more than half of those covered by bargaining council agreements in 2005 were employed in national, provincial and local government departments. Excluding the three spheres of government, 13 percent of those in formal employment were covered

  • Extensions to non-parties cover a small share of workers, so the rigidity argument looks overstated. (p. 17)

    of the estimated 32,6 percent of formally employed workers (with the total excluding all Managers and Professionals in the private sector) covered by bargaining councils, only 4,6 percent were covered by extensions to agreements.

  • Most exemption applications are granted, weakening the claim that extension-exemption rules cause rigidity. (p. 18)

    The evidence here is that almost 80 percent exemptions are granted.

  • In 1995 the council premium accrued only to non-unionised members, while the union premium was far larger. (p. 43)

    bargaining council members were likely to earn almost seven percent more than those not covered by a wage agreement. Membership of a union in 1995 provided a much larger wage premium

  • In 2005 private-sector councils gave no premium but public-sector councils gave about 28 percent, with a union premium of about 23 percent. (p. 46)

    public sector workers did enjoy a wage premium due to their membership of the public sector bargaining councils, of about 28 percent. The union wage premium for 2005 according to this estimation is about 23 percent.

  • Council and union premia can combine to a large joint premium. (p. 47)

    a joint premium of as high as 51 percent may be present through these institutions of the labour market

  • Union premia fall across the wage distribution, with the largest gains at the bottom. (p. 51)

    relatively low-earning (and disproportionately unskilled) workers at the bottom end of the distribution benefited more from belonging to a union than better-earning, higher skilled workers.

  • By 2005 the public-sector council premium matched or exceeded the union premium at several points in the distribution. (p. 54)

    the public sector bargaining council wage premia is, in fact, higher than the union wage premia.

  • Historically, the 1924 Act excluded pass-bearing Africans from the definition of employee, creating a race-based dual system. (p. 5)

    it excluded pass-bearing Africans from the definition of an ‘employee’, meaning that these workers were excluded from representation on industrial councils and also excluded from the agreements reached by industrial councils.

  • Employer dominance and weak union organisation kept stipulated minimum wages low, especially for the less skilled. (p. 7)

    employers generally had the balance of power in industrial councils. They dictated the levels at which bargaining took place in the councils and this, for example, resulted in relatively low levels of stipulated wage minima