Rather than a shock specific to low-income borrowers, the results suggest that the housing cycle was driven by an aggregate factor. Although a drop in nationwide interest rates is one possibility, Glaeser, Gottlieb, and Gyourko (2013) find that neither theory nor data support the interest-rate hypothesis. Overly optimistic house-price expectations are a better explanation for the boom: borrowers would have wanted to buy houses that were rising in price and lenders would have been eager to write mortgages against this rapidly appreciating collateral. Some new papers in the macroeconomic literature are now exploring the relative roles of income shocks, looser credit constraints, and over-optimistic expectations in driving the mortgage boom. These papers typically find that looser credit constraints do not do a good job of capturing important features of the data. However, widely shared optimism about the path of future house prices is consistent with a broad-based expansion of mortgage debt—the key empirical finding of this paper.