A Tale of Two Short Consumption Modules

Two short consumption modules were piloted in Bogra and Sirajganj (Bangladesh) in May-June 2012 as part of the Global MICS5 Pilot. This paper aims at validating this exercise and assessing the accuracy and reliability of the consumption estimates obtained. The use of a benchmark consumption module is essential in order to assess how well the two short options fare; the analysis therefore consists of a systematic comparison of both short modules with a benchmark. The attempt made is to isolate and test the impact of the length (degree of commodity) of the consumption questionnaire on the quality of consumption and poverty estimates as well as distributional measures obtained. We conclude that it is feasible to include a short consumption module in MICS (Multiple Indicator Cluster Surveys). The Bangladesh experience suggests that this module can give accurate predictions of aggregate consumption and poverty, allowing for the analysis of monetary and non-monetary dimensions of welfare together. However the module cannot be used to analyze individual consumption groups (like food, nonfoods, etc.) or consumption patterns.