An Important Factor in Translating Monetary Income

An important factor in translating monetary income into some measure of welfare is the set of prices facing households. Identical households with identical monetary incomes may appear to be in the same position in terms of welfare, but this would only be the case if they face the same set of prices. If one of these identical households is required to pay higher prices than the other, the former's welfare will be lower than the latter's. This may result in the misidentification of technically poor households as non-poor and/or vice versa. Thus, when estimating poverty lines and measuring poverty, the different sets of prices facing different households should ideally be taken into account. One of the important sources of price variations across households derives from geography: households in different regions may face different price sets because of different local tastes and preferences, transport costs, climatic and agricultural differences, and demand patterns. One important facet of spatial price differentials is the difference between urban and rural prices, although there is likely demand from policymakers and officials for poverty lines tailored to more detailed geographical units, such as for individual provinces, or for metropolitan, urban and rural areas within provinces. This is particularly relevant in South Africa given that inequalities inherited from the apartheid system have a clear spatial dimension.As noted above, one of the data issues that has arisen in the 2005 Income and Expenditure Survey is a probable underestimate of food expenditure. Such an underestimate would serve to bias the non-food component of the absolute poverty line upwards, resulting in an upwardly biased poverty line. Another issue that impacts on the food component of the derived absolute poverty line relates to consumption of home production.

The IES 2005, like its predecessor, is unable to accurately quantify consumption of home production (in quantity terms). Although the questionnaire makes provision for the consumption of home production, it only records a respondent-assigned value of the consumption, with no quantity

11

information, which would allow for direct estimation of calories, being included in the released dataset. The use of data on the average sales price per quantity is not a feasible option since there is no evidence to suggest that respondents both know ruling market prices per quantity and value the consumption of home production accordingly. Furthermore, it is highly likely that there exist regional price variations for which it is not possible to make adjustments. This is a critical defect in the design of the questionnaire for the 2005 IES as it relates to poverty work, impacting negatively on the ability to accurately estimate a consumption-based poverty line using the IES. Specifically, the analysis is forced to omit consumption of home production: the result will be