The Analytical Returns to Measuring a Detailed Household Roster
Richard Akresh, Eric V. Edmonds
Households are dynamic while most surveys only collect information on individuals who are present at a single point in time. We exploit a unique and thorough household membership enumeration in Burkina Faso to consider the analytical costs of the typical static household roster. We document that households are extremely fluid with 10 percent of individuals spending sometime away over a three year period, averaging 16 of the 36 months away. The residency status of persons age 16 to 24 is most in flux. A more complete enumeration offers substantial analytical richness that is especially important for the analysis of issues that are intertwined with who is present in the household, such as the measurement of income inequality and the nature of sibling interactions in education decisions. We find that evidence of sibling rivalry in Burkina Faso appears to owe to the correlation between the presence of sisters in a household and nonagricultural income. We argue for more detailed and thorough measurement of household composition in future multi-purpose household surveys.
Text: See Discussion Paper No. 4759