DP13469 A theory of structural change that can fit the data
We study structural change in historical consumption expenditure of the United States, the United Kingdom, Canada and Australia over more than a century. To identify preference parameters from aggregate data, we characterize the most general class of preferences in a time-additive setting that admits aggregation of the intertemporal saving decision. We parametrize and estimate such intertemporally aggregable (IA) preferences and discuss their properties in a dynamic general equilibrium framework with sustained growth. Our preferences class is considerably more flexible than the Gorman form or PIGL/PIGLOG, giving rise to a good fit of the non-monotonic pattern of structural change.