UCLA

This dissertation comprises three chapters that explore the role housing plays in household decision-making and its macroeconomic implications, aiming to improve our understanding of how macroeconomic shocks are amplified or muted due to financial frictions and the special properties of housing through the household side of the economy.

The first chapter investigates how households adjust their car spending in response to a housing wealth shock. Utilizing detailed Swedish administrative data, my co-authors and I study the economic consequences of the unexpected decision to continue operating Bromma Airport in Stockholm, Sweden, in 2007. The airport's continuation changed perceptions of housing values in areas exposed to its negative externalities, such as noise and accident risk, causing persistent and heterogeneous changes in housing wealth across households. By using transaction data on homes across differently exposed neighborhoods, we estimated household-level housing wealth losses. We also observe car purchases for all households and can match car purchases to housing wealth losses and other household characteristics which allows us to quantify the effect on car spending in a difference-in-difference setting. The key finding is that households exhibit a muted response in car purchases compared to prior studies. Using loan-level data, we show that the effect is concentrated among homeowners who rely more on mortgage borrowing to finance car purchases; as home prices unexpectedly change, their borrowing ability is impacted, suggesting a significant role for the collateral channel. By exploiting other dimensions of household heterogeneity, we also conclude that the pure wealth effect is weak.

The second chapter addresses the puzzle of low migration rates from areas experiencing economic decline. Focusing on the labor-market area around Stavanger, Norway, following the 2014 global oil price drop, I empirically document that diminishing home values are associated with homeowners reducing their probability to leave, while renters and homeowners with the least housing wealth exhibit an increase in their leaving probabilities. On net, the out-migration response is economically insignificant while the in-migration response exhibits a strong reduction—the net-migration falls because people stop moving to Stavanger. A life-cycle model with location, housing, and saving decisions explains these results: The reduction in home prices decreases the affordability of housing in potential destinations, thus making migration less attractive—a "housing wealth effect." This finding challenges the prevailing notion that cheaper local housing encourages workers to stay—i.e., benefits stayers—and instead suggests that falling home values act as a barrier to mobility. From a general equilibrium perspective, prices have to fall to clear the housing market. As prices fall, it becomes less attractive for homeowners to move. Potential immigrants are deterred by the fall in potential earnings, and current homeowners end up holding the stock of housing. While effective policy transfers welfare from those impacted by shocks, a policy experiment with moving vouchers shows that the beneficiaries are largely renters, who are already compensated by cheaper rents but still act as if they face lower moving costs.

The third chapter complements the analysis of the first, using the same empirical setting as Chapter 2 but narrowing the sample to government workers who did not suffer differently from the oil shock as a function of their location. However, changes in home prices accompanying the oil shock differed greatly depending on the location's exposure to the petroleum sector. Using Norwegian administrative data matched to high-frequency digital payment data, we document the dynamic response of expenditures by various goods and services as functions of lost housing wealth at the household level. The analysis provides a more detailed understanding of housing wealth shocks' spending dynamics and heterogeneity than previous work. We find that overall spending is affected less than in previous work, with a marginal propensity for expenditures close to 0.02 kroner per krone housing wealth change. Vehicles and furnishings—durable goods often financed using credit and whose purchases can be postponed—respond more strongly, while food and beverages do not change when contrasting between government workers in Stavanger versus the rest of Norway. We also find that more indebted households and those with less liquid assets respond about twice as strongly than the average. Our findings indicate that the potential spillovers of reduced consumption due to home price shocks can be limited due to the nature of goods more affected, and that the distribution of debt in the economy matters for the aggregate effect.

Collectively, these essays contribute to the fields of household finance and macroeconomics by leveraging comprehensive Scandinavian administrative data and quasi-experiments, offering insights that add to previous studies. The findings underscore the complexity of housing wealth effects on consumption and migration, with policy implications for addressing economic decline and household financial stability.