The commodity chain of the household: from survey design to policy and practice

Data collection and analysis and policy formulation all require a social unit to be defined, generally called the household. Multidisciplinary evidence shows that households as defined by survey practitioners often bear little resemblance to lived socio-economic units. This study examines how a shared language, the 'household', can generate misunderstandings because different groups with distinctive understandings of the term 'household' are often unaware that others may be using ‘household’ differently. Results from 4 interlinked and iterative methods are presented: review of household survey documentation (1950s-present); ethnographic ground-truthing fieldwork; in-depth key informant interviews; and modelling. Results show that whereas data collectors have a clear idea of what a `household` is, data users are often unaware of the nuances of the constraints imposed by data collection. This has implications for policy planning and practice. What interviewees consider when they think of their household can differ systematically from data collectors' definitions.