Internal U.S. Migration and Consumption Dynamics : A Panel Data Analysis

Using the Kilts Nielsen Consumer Panel (KNCP), this paper documents new stylized facts on expenditure and income growth upon migration. First, movers' KNCP expenditures relative to non-movers' decline by 6% during a year before the move, and then increase by 9% during a year after the move. Thus, there is a 3% net increase in relative expenditures from a year before to a year after the move. This net change in movers' relative expenditures is consistent with that in movers' relative earnings, which I estimate to increase by 3% over the same 2-year window around the move time. The V-shaped pattern in movers' relative expenditures departs from predictions of the standard permanent income hypothesis (PIH), which many migration models in the literature implicitly assume. If movers can expect an income increase after moving, the PIH predicts movers to raise their spending upon deciding to move, not subsequently. I see similar spending responses to moving for non-storable foods, implying that the V-shaped response does not reflect moving households' dissipating then accumulating inventories. Next, I document that movers' income growth displays higher variance than non-movers', both before and after moving. Movers' expenditure growth also exhibits higher variance than non-movers' before and after the move. These findings suggest that movers have higher precautionary savings motives than non-movers. Hence I explore whether the PIH combined with precautionary savings motives is consistent with the documented data moments. I find that a PIH model with a reasonable size of prudence, i.e., a relative risk aversion coefficient between 1 and 2, can match the delayed increase in movers' expenditures after the move