Jinho Kim

Lower-income households are far less likely to capture the gains from refinancing when interest rates fall. Conditional on refinancing being financially advantageous, bottom-income-quintile borrowers refinance 43 percent less often than the top quintile. One in five borrowers refinance within four months of becoming in the money in the top quintile, compared with ten months in the bottom quintile. Yet the savings the bottom quintile leaves unrealized equal 6.7 percent of its monthly income, highest among all income groups. After a policy-induced 1 percentage point fall in mortgage rates, the top-bottom gap in cumulative refinancing reaches 4.3 percentage points at two quarters, against an average of 2.7. I explain these facts with a heterogeneous-agent model in which households choose how closely to monitor their mortgage, weighing the dollar gain from refinancing against an attention cost that falls with income. Small-balance borrowers monitor less even when their gains relative to income are large. A notification policy calibrated to field-experimental evidence raises the five-year consumption response to a rate cut by 10.8 percent, with gains concentrated among lower-middle-income households.

Demographic age structure systematically shapes housing-boom vulnerability. Using an OECD panel of 29 countries and a U.S. county panel, we show that locations with a higher young-to-old population ratio are more likely to experience large housing booms a one-unit increase in the ratio raises boom probability by 18.6 percentage points across countries and 4.3 points across counties, robust to location and time fixed effects, income, and housing-supply controls. Two household-level results connect this pattern to behavior. In mortgage application data, young applicants increase their application intensity 4.3 to 6.6 percent more than older applicants during local booms; and after an identified credit-supply easing, house prices rise more, and more persistently, in younger areas. We interpret these facts with a standard annual life-cycle housing model with tenure choice. Fed the same identified credit shock, a younger economy amplifies the house-price response and reproduces the impact age gradient. We also provide suggestive evidence that matching its persistence may require adaptive house-price expectations. Demographic composition thus emerges as a demand-side state variable for housing-boom risk.

This project empirically examines how wealthy hand-to-mouth households respond to monetary policy through consumption and how housing wealth affects spending. I exploit substantial cross-regional heterogeneity in mortgage-ownership ratios, which serve as a proxy for local mortgage burden and differential exposure to credit constraints. First, using this variation in a shift-share framework combined with high-frequency monetary policy shocks, I show that regions with higher mortgage exposure exhibit significantly stronger consumption responses to interest rate cuts, providing evidence of high marginal propensities to consume among wealthy hand-to-mouth households. Second, adapting the elasticity-based identification strategy of Guren et al. (2021), I construct a new instrumental variable to examine how house price increases affect consumption through wealth effects. While the total effect closely matches their findings, the consumption response to housing wealth is significantly weaker in regions with higher mortgage ownership, suggesting that mortgage-related liquidity constraints dampen the wealth effect of rising asset prices.