Inter-Household Allocations within Extended Households: Evidence from the Indonesia Family Life Survey

摘要

Some longitudinal household surveys interview split-off households in addition to the original households during the follow-up surveys. For the purpose of analyzing household consumption changes between years, defining the panel households as only the panel of original households may potentially induce selectivity bias if households split non-randomly. One way to avoid this problem is to treat the original household and its split-off as a single extended household. However, this approach implicitly assumes that the extended household acts as if it is a single household. Using this approach to analyze changes in household consumption amounts to assuming income pooling within the extended household. We use the data from two waves of the Indonesia Family Life Survey to test whether households that originate from the same original household pool their income to smooth the consumption across the households. Our findings suggest that in contradiction to the null hypothesis of income pooling, the distribution of income between sub-households in extended household does affect the distribution of their consumption.