Housing as a transmission channel : household responses to aggregate shocks and implications for asset prices

This dissertation studies housing as a transmission channel for aggregate shocks. Three essays show that the behavior of investors and households affects how house prices, expectations, and labor supply respond to interest rate changes, regional price booms, and the Covid-19 pandemic. Chapter 2 asks whether duration captures the true interest rate sensitivity of house prices. In fixed-income markets, long-duration assets are more sensitive to interest rate changes, and this principle is commonly assumed to extend to other asset classes. Using ZIP-code data from the American Community Survey, I construct a new measure of housing duration and show the opposite for housing: short-duration markets respond more strongly to rate changes than long-duration markets. A one-percentage-point cut in interest rates raises house prices by 1.86 percentage points over two years on average, but markets one standard deviation shorter in duration experience an additional 0.71-percentage-point increase. The mechanism is a discount-rate channel driven by "reaching-for-income'' buy-to-rent investors who shift toward high-yield, short-duration properties after rate cuts. Chapter 3 documents that homebuyers extrapolate across geographic markets. Using nearly three million U.S. out-of-town housing transactions, I show that buyers from origin ZIP codes with 50 percentage points higher house price growth over the prior five years pay roughly 2% more for otherwise comparable properties. Two complementary approaches identify extrapolative beliefs, not wealth effects alone, as the driver: I separate renters, migrants, and second-home buyers, and I build a belief-sensitivity instrumental variable. The chapter reveals a behavioral channel through which housing booms transmit across regions. Chapter 4, joint with Jack Favilukis, shows that the Covid-19 housing boom largely explains the Great Resignation among older U.S. workers. The labor force participation of older homeowners falls in metropolitan areas with stronger house price growth, while the response is muted or reversed for renters and younger workers. A counterfactual indicates that, absent the 2021 housing boom, the labor force participation of older Americans would have remained near its 2019 level. A life-cycle model with realistic wealth, income, and tenure heterogeneity replicates the empirical patterns.