Working Paper

Sticking to the Facts: Official and Unofficial Stories about Poverty and Unemployment in South Africa

Poverty’s major cause is unemployment. In the unlikely event that unemployment is halved by 2014, there will still be millions of people in South Africa in workerless households. The only way to address this is to develop a truly comprehensive social protection system. This paper looks at aspects of the way government responds to claims that are made, chiefl y by academics, about poverty and unemployment. Bearers of good tidings are effusively welcomed while critics are dismissed with a hail of numbers about the achievements, actual or projected, of the democratic government. Offi cial statistics on poverty and unemployment enjoy little favour among senior politicians and civil servants. ‘Unoffi cial’ poverty and unemployment statistics, by contrast, are seized upon with enthusiasm if they contradict the gloomy picture created by numbers that suggest (with monotonous regularity) that improvements in the lives of the poor are not happening fast enough. The fi rst part of the paper explores possible explanations of government’s extreme sensitivity to criticism. The second part of the paper looks at an old chestnut, the repeated claim by government that the severity of the unemployment problem has (in part?) to do with the ‘fact’ that the number of economically active people has grown faster than the number of working age people. The claim is false, and may readily be seen to be so. It sits awkwardly with the equally frequently repeated claim that ‘we are on target to halve unemployment’. The usefulness of the (false) claim is presumably the sympathy and understanding it evokes for the plight of a government faced, in its attempts to solve an immensely diffi cult problem, by continually moving goalposts. The third part of the paper looks briefl y at the van der Berg et al (2005) poverty reduction estimates for the period 2000-2004 (they have the headcount falling from 18.5 to 15.4 million), before reproducing my estimates for the period 2001-2004 (the headcount falls from about 19.5 to somewhere in the region of 18 million). Extracted from two recent papers (Meth 2006a and 2006b), a brief description of the way these fi gures have been constructed, including a discussion of the many data diffi culties faced in doing so, is offered. The van der Berg et al fi nding of a headcount fall of three million, said largely to be the result of massive expansion of the social grant system (and possible improved job creation) is argued to be dubious. My results suggest that the grants lifted an additional 1.2-1.5 million people over the poverty line. They also suggest that job creation benefi ted the well-off rather than the poor. Figures from a paper by Burger and Yu (2006), two of van der Berg’s co-authors, suggest that earnings growth, the only other possible source of income for raising the poor out of poverty, cannot account for more than a trivial proportion of such movement out of poverty as did occur during the period. Suggestions are offered for the proper academic conduct upon releasing contentious results into a highly-charged political debate.