Affirmative action policies and South Africa’s racial wage gap

Racial wage inequality and discrimination have pervaded South African society for centuries. Apartheid legislation cemented these disparities by institutionalizing white job reservation and many other unfair practices. While racial wage gaps started to decline towards the end of apartheid in 1994, they increased (against all expectations) in the immediate post-transition period.

The rise in the racial wage gap only started to reverse in 2005. Existing studies therefore report limited effects of affirmative action policies, with highly skilled occupations still dominated by white men, and racial wage gaps remaining higher than in 1997. A new WIDER Working Paper, by Rulof Burger, Rachel Jafta, and Dieter von Fintel, addresses some limitations associated with these studies and argues that affirmative action policies have had an impact on reducing the racial wage gap.

The wage gap under apartheid

One study[1] in 1970 when apartheid laws regulated labour market activity, found that white males earned approximately 5.7 times more than their black counterparts. While this large difference was partly attributable to differences in education, white workers could still expect to have earned between 3.6 and 5 times more than equally educated black counterparts.

However a substantial narrowing of the racial wage gap was noted during the 1980s.[2] This was partly (but not completely) driven by the gradual increase in the educational attainment of black workers.

A widening wage-gap post-apartheid?

Perhaps surprisingly, several studies[3] find that the previous trend of decreasing discrimination was subsequently reversed in the immediate post-apartheid period and that there was a continued (but slight) widening of the racial wage gap. During this period racial employment discrimination decreased slightly, whereas racial occupational and wage discrimination increased marginally.

The effect of affirmative action

Affirmative action legislation in South Africa was enacted with a lag, first targeting employment equity and skills development in 1998 and then more extensive ‘black economic empowerment’ in 2003. The previous studies mentioned above showed a widening wage gap after the first wave of legislation, and did not cover extensively the period after the second wave.

In their WIDER Working Paper the authors address several methodological concerns[4] with previous studies, as well as extending the period studied. This new approach suggests that discrimination did start to decline — though modestly — soon after the implementation of employment equity legislation and that the implementation of black economic empowerment legislation in 2003 represented a more visible turning point. The decline in the discrimination did not reverse after 2003, even after the financial crisis. The authors suggest that after stripping out generational unobservables, and accounting for compositional changes, affirmative action legislation appears to have been effective in reducing the racial wage gap.

Are the effects likely to be long-lasting?

The authors show that the generations that entered the labour market at the height of apartheid experienced successively higher discrimination. The generational component of discrimination has not yet undergone a turning point, though its growth has slowed down. This portion of the total wage gap explains why raw figures measure growth in the wage gap — new entrants are facing higher discrimination than retirees. The exact reasons for this pattern are difficult to pinpoint, though one potential channel might be the persistent racial gap in education quality that continues to drive a wedge in labour market opportunities between the various groups.

To whom have the benefits from affirmative action accrued?

The authors show that among men, black wages became as responsive to economic growth as white wages after the imposition of affirmative action legislation. Reforms therefore appear to have removed some obstacles that prevented black men as a whole from enjoying the benefits of economic expansion.

However, the way education was rewarded plays a dominant role in the evolution of the wage gap. Returns to tertiary education have grown for all groups, but especially among black men; the trend for black men intensified even more after employment equity laws were enacted. It therefore appears that the legislation has been effective at drawing highly-qualified black men into well-remunerated positions, and that this drives much of the average effects observed by the authors.

This paper is a part of the project Discrimination & Affirmative Action: What have we learnt so far.

The views expressed in this piece are those of the author(s), and do not necessarily reflect the views of the Institute or the United Nations University, nor the programme/project donors.

Follow us on twitter @UNUWIDER. Follow James Stewart on twitter @deveconwatcher.


[1] Knight and McGrath (1977).

[2] Moll (2000).

[3] Sherer (2000), Erichsen and Wakeford (2001), Allanson, Atkins and Hinks (2002), Rospabé (2002), Brookes and Hinks (2004).

[4] This paper addresses methodological concerns in gauging the time evolution of discrimination; by controlling for cohort fixed effects, the bias in standard wage gap decompositions is reduced, and compositional differences between entrants and retirees are largely accounted for.