Using household survey data for the U.S. and the U.K., we show that the aggregate response of consumption to interest rate changes is driven by households with a mortgage. Outright home-owners do not adjust expenditure at all while renters change their spending but by less than mortgagors. Income rises for all households as interest rate cuts directly affect firm investment and household consumption, boosting aggregate demand. A crucial difference between the housing tenure groups is the composition of their balance sheets: mortgagors hold sizable illiquid assets but little liquid wealth. Our results reveal that general equilibrium effects on household income coupled with balance-sheet-driven heterogeneity in the marginal propensity to consume play a key role in the transmission of monetary policy.

By extending his data, we document the instability of low-frequency regression coefficients that Lucas (1980) used to express the quantity theory of money. We impute the differences in these regression coefficients to differences in monetary policies across periods. A DSGE model estimated over a subsample like Lucas's implies values of the regression coefficients that confirm Lucas's results for his sample period. But perturbing monetary policy rule parameters away from the values estimated over Lucas's subsample alters the regression coefficients in ways that reproduce their instability over our longer sample.