Independent study shows CSS's jobless rate is accurate
Government ministers have been questioning the official unemployment figures, writes CAROL PATON
The study found the CSS's October household survey, which is used as the basis for the unemployment figure, to be broadly accurate, despite frequent criticism by government ministers of the accuracy of the 30% rate.
However, the independent study heavily discredits the formal sector unemployment statistics compiled by the CSS and used by the Reserve Bank. These are the figures used in calculations on job creation in the government's growth, employment and redistribution policy.
The study, which is not yet officially available, is also believed to state that between 1993 and 1995, as many as 400 000 new jobs could have been created in trade and tourism as well as community, personal and social services.
However, the creation of new jobs did not affect the unemployment rate as an estimated 250 000 students a year entered the job market during the same period.
The study, which was commissioned in the light of government scepticism of the figures and criticism by the International Labour Organisation, was conducted by Stephan Klasen, a research fellow at Cambridge, and Ingrid Woolard of the University of Port Elizabeth.
The 1995 survey - the most recent survey for which figures are available - measured employment at between 16% and 29%, depending on the definition used. The narrow definition includes all those willing to accept a job if offered and who have actively sought employment in the last month.
The broader definition covers all those who are not working but who state they would be willing to accept a suitable job. This definition is considered broader as it includes "discouraged workseekers" who have lost hope of finding a job and as a result are not actively seeking work. It would also include people making a living in the informal sector.
But while endorsing the survey as a measure of unemployment, it also warns that until more accurate information from Census 96 is available, it is difficult to be certain about any of the figures. Updated census information could change the research sample.
The investigation into South Africa's employment statistics follows recent comments by a number of senior government ministers on the veracity of the CSS figures. It also follows a set of detailed criticisms from the International Labour Organisation made in its country review in 1996, which claimed that the poor quality of statistics and a range of methodological problems had exaggerated the unemployment problem.
Among the problems identified were: ý undercounting in mining; ý the omission of small-scale agriculture and small self-employed manufacturers; ý underestimating employment in a number of sectors such as construction, trade, transport and finance, where the trend has been to employ contract workers; ý the omission of new firms from the sample; ý undercounting of rural employment; and ý undercounting of the informal sector.
However, the study found that few of these criticisms applied to the Household Survey, although the CSS series on employment had undercounted employment in agriculture and failed to capture new forms: the expansion in trade, catering and hotels as well as a range of informal tourism-related activities.
In the case of the survey, a slight modification of the definitions of employment to bring these in line with other internationally accepted norms had altered the employment rate downwards by less than a percentage point.
The search for "hidden workers" on contract or in seasonal employment or subsistence agriculture also had little effect on the employment rate - at most 2% or 3%. The survey had undercounted employment in mining. Meanwhile, the CSS also said this week it was revamping its "entire suite" of labour statistics with the assistance of the Australian Bureau of Statistics.
The CSS employment series consisting of 17 separate monthly or quarterly surveys and an annual survey are to "reconceptualised and overhauled" this year and will be replaced by three new surveys. The informal sector will be a particular focus.