Why South African Incomes Declined

Featured in print Digest

Average incomes of South African men and women fell by about 40 percent between 1995 and 2000, and that there has been little improvement since then.

South Africans are worse off than they were before the end of apartheid, at least as measured by real incomes. In Incomes in South Africa Since the Fall of Apartheid (NBER Working Paper No. 11384), co-authors Murray Leibbrandt, James Levinsohn, and Justin McCrary document that decline and attempt to explain what has happened. They show that average incomes of South African men and women fell by about 40 percent between 1995 and 2000, and note that there has been little improvement since then. These researchers explore income patterns in the South African economy overall and in specific groups, such as men and women, older and younger workers, and whites and blacks. Their focus is on economic well being as measured by income, rather than on other ways of evaluating social welfare, including measurement of political freedom.

The change in income is most pronounced in the lower half of the income distribution and has disproportionately affected younger workers, women, and blacks. For men in the bottom 5 percent of the income distribution, total real income in 2000 was about half the level of 1995. In the tenth through the seventy-fifth percentile, incomes were about one third lower than in 1995. For those in the top 10 percent, incomes declined by about one-seventh. For women, the results are very similar, although above the ninetieth percentile, women fare slightly better than men, with roughly constant real incomes. In 1995, white South African men were paid 98 percent more than black men. By 2000, this discrepancy had grown to 118 percent -- a difference of 20 percentage points. Black and white women saw the analogous gap grow by 40 percentage points.

The decline in income may reflect a slack labor market and skill-biased technical change, the authors explain. Using a variety of statistical techniques and a range of data sources, they show that rather than a change in the "endowments" of workers, such as education and skills, it was the change in returns on these endowments that provides the underlying explanation for declining income.

Skill-biased technical change is generally thought to have contributed to stagnation of real wages for lower skilled workers in other countries around the world in the past decade. But a 40 percent decline in real incomes in such a short span of time is a highly unusual event. In South Africa's case, the effect has been exacerbated by the economy's poor economic performance and the large amount of slack in the labor market, which has put downward pressure on real wages. Unemployment was already high in South Africa in 1994. In 1995-2000, while the number of job seekers rose by 5-6 million, net job creation was only in the range of 1.5 to 2 million. Labor, and particularly lower-skilled labor, has been entering the market faster than it could be absorbed, putting downward pressure on wages. Blacks have suffered because their education levels are lower following the discrimination of the apartheid era.

South Africa's reengagement with the world economy, after relative isolation during the period of anti-apartheid trade sanctions, may have added to the downward pressure on the incomes of lower skilled workers. Younger workers - from 18 to their early 30s - and women have suffered because of lower skill levels and less-established positions in the workforce than older workers and men.

The researchers emphasize that, in spite of these plausible explanations, the drop in South African incomes is not fully understood. But there are some implausible explanations that they can dismiss. They include the notion that high-skilled and high-earning white workers left the country after the African National Congress came to power in 1994, or that the data are somehow faulty. Nor was it the case that the most able workers were no longer among those reporting positive incomes. The controversial nature of the paper's claims emphasizes the need for more research to better understand the declines documented, the authors stress

-- Andrew Balls