Abstract:
In the absence of South African longitudinal data for the ten years post apartheid, national cross-sectional household survey data is frequently used to analyse change over time. When these data are stacked side-by-side however, they reveal inconsistencies both in trends across time and between the household and person level data. These inconsistencies can introduce biases into research which analyse change. This study calculates a new set of person and household weights for the October Household Surveys between 1995 and 1999 and the Labour Force Surveys between 2000 and 2004. A cross entropy estimation approach is used. This approach is favoured because the calculated weights are similar to the initial sample weights (and hence retain the survey design benefits) while simultaneously being consistent with aggregate auxiliary data. A consistent series of aggregates from the Actuarial Society of South Africa (ASSA) model and the 1996 and 2001 South African Census data are used as benchmarks. The new weights result in consistent demographic and geographic trends over time and greater consistency between person and household level data.
