Consumption and aggregate constraints : evidence from U.S. states and Canadian provinces

State-level consumption exhibits excess sensitivity to lagged income to the same extent as US aggregate data, but state-specific (idiosyncratic) consumption exhibits substantially less sensitivity to lagged state-specific income---a result that also holds for Canadian provinces. We propose the following interpretation: borrowing and lending in response to changes in consumer demand is easier for an individual US state than it is for the US as a whole. The PIH may thus be a good model for describing the reaction of consumption to idiosyncratic disposable income shocks even if it fails at the aggregate US level. Further analysis, centered on the persistence of income shocks and on the consumption/income ratio, is consistent with this interpretation but suggests that the PIH still requires qualification. We contrast our results with tests of full inter-state risk sharing.