Why does the Engel method work? Food demand, economies of size and household survey methods

Estimates of household size economies are needed for the analysis of poverty and inequality. This paper shows that Engel estimates of size economies are large when household expenditures are obtained by respondent recall but small when expenditures are obtained by daily recording in diaries. Expenditure estimates from recall surveys appear to have measurement errors correlated with household size. As well as demonstrating the fragility of Engel estimates of size economies, these results help resolve a puzzle raised by Deaton and Paxson (1998) about differences between rich and poor countries in the effect of household size on food demand.