Abstract About half of the change in U.S. non-durable consumption expenditure is due to changes in the products entering households’ consumption basket (the extensive margin). Changes in the basket are driven by fluctuations in the rate at which households add products; removals fluctuate little. These patterns reflect that, in response to income increases, households adopt consumer products already available in the market. Household adoption amplifies the effects of fiscal transfers on consumption by more than 30%. Cyclical household adoption of products also implies that inflation measures based on a representative household consuming all varieties available in the market underestimate true household-level inflation by as much as 1% per year over the Great Recession in the consumption categories covered by our data.

We build a model of the mortgage market in which banks attain their optimal mortgage portfolio by setting rates and steering customers. Sophisticated households know which mortgage type is best for them; naive households are susceptible to banks’ steering. Using data on the universe of Italian mortgages, we estimate the model and quantify the welfare implications of steering. The average cost of the distortion is equivalent to 16% of the annual mortgage payment. A financial literacy campaign is beneficial for naive households, but hurts sophisticated ones. Since steering also conveys information about mortgages, restricting steering might result in significant welfare losses.