Poverty lines as context deflators in the DRC: a methodology to account for contextual differences

In this paper we present a specific methodology to make spatial well-being and poverty assessments based on expenditure data to some extent sensitive to contextual aspects other than price differentials. The rationale behind this method coincides with the view expressed by the advocates of human development pointing to the irrelevance of (real) income levels for well-being measurement compared to an analysis of people?s ability to deploy this purchasing power in a very specific time and setting. Yet, in order to operationalize this principle, we opted to employ the ordinary technique of deflating nominal incomes, but in such a way that genuine comparability over different geographical entities comes within reach. Of course, the extent to which our methodology is able to achieve this goal, largely depend on the exact content and construction of the underlying deflators used. Given our research agenda to analyze the distribution of poverty and well-being in the Democratic Republic of Congo (DRC), we decided to construct 56 regional poverty lines whose pair wise ratios in turn were used as a set of context deflators. The challenge of this exercise essentially boiled down to reconciling the two seemingly discordant –but highly appreciated– poverty line characteristics of „specificity? and „consistency?: i.e. how to give due attention to the myriad of local living conditions while still ensuring sufficient comparability. Although this theoretical discordance has been settled some time ago, methodological problems to align both features still remain cumbersome in practice. Therefore, the main contribution of this paper needs to be understood in addressing these methodological issues within the framework of a household and expenditure survey. As a natural starting point to deal with these issues, we screened the validity of the major criticisms raised against the Food Energy Intake (FEI) method. Indeed, this context-sensitive methodology for setting poverty lines is often condemned for generating inconsistent results, but –after closer scrutiny– not for every reason put forward in the literature. On the basis of these insights, we started to mould our own specific poverty line methodology in which we tried to accommodate the remaining pieces of critique. This resulted in a three-step FEI-like procedure where the anchoring device to ensure consistency between all 56 localities, was built upon the basic human functionings of being adequately nourished and sheltered. Finally, we applied this procedure to our data in order to discern the impact of our own methodology vis-à-vis other internally and externally computed approaches.