Abstract
The main objective of the Post-Apartheid Income and Expenditure Series (PIES) is to try and ameliorate some of the inconsistencies that have resulted from the alterations that occured in the Income and Expenditure Surveys (IES) released by Statistics South Africa in the past two decades. At the time of the production of PIES version 1, four IESs had been conducted - in 1995, 2000, 2005-2006, and 2010-2011, and the data published by Statistics South Africa. Two major methodological changes in the 2005-2006 IES have complicated comparisons of the IES data over time. PIES stacks these four IES surveys and standardises, matches, and adjusts the data and the sampling weights. This produces consolidated, time-consistent, and item-comparable data that provides more accurate and nuanced interpretations of the income and consumption profiles of South African households over a span of 15 years.
Three files are released in version 1 of PIES. The first is a nominal dataset called "zaf-dpru-pies-nominal-v1" where no price adjustments have been made. For the other two datasets, the expenditure, income and other monetary values are deflated to June 2016 prices. There are two "real price" datasets because the deflation methods were different, which is discussed briefly here. Users are advised to consult the documentation for a detailed discussion. In “zaf-dpru-pies-real1-v1”, the nominal amounts for expenditure, income and other monetary values are deflated with “total CPI” indices, which were made for each year of the IES (1995, 2000, 2005 and 2010). A “total CPI” index is calculated as the weighted sum of the (sub)category price indices. The “zaf-dpru-pies-real2-v2” data however differs in two ways. Firstly, its expenditure items are deflated using multiple deflators instead of a single "total CPI" deflator. Secondly, the deflators for income, debt and other unmatched items are determined by the household's particular expenditure pattern.
