Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Stuart Rosenthal


This dissertation is comprised of two essays on the spatial patterns and economic consequences of volatility in two highly related markets. The first looks at spatial variation of home price volatility and related investor exposure to risk across and within metropolitan areas, finding that they contribute to sustained differences in housing returns at both levels of geography. The second turns to the labor market and returns to high-skilled married women’s investment in education. Here volatility is in terms of daily variation in travel time due to congestion. I find that commute volatility has an adverse effect on both the likelihood of labor market participation as well as the quality of job matches for those who are employed.The first chapter is motivated by a new stylized fact: in large urban areas, home prices appreciate faster in city centers and generally in more densely developed locations. This echoes a similar pattern across cities with more supply constrained locations experiencing higher returns. Supply constraints amplify unobserved shocks contributing to investor exposure to risk which in equilibrium should be met with higher expected returns. Using a housing-oriented Capital Asset Pricing Model housing risk is decomposed into a systematic and location specific component. We find that while differences in risk across locations contribute to differences in returns, the dominant source of risk varies across geographic levels. Within urban areas, location-specific risk is most important, while across cities, systematic risk dominates. Previous literature has focused on spatial variation in house price levels due to differences in local advantages. These findings indicate that spatial differences in home price appreciation can also persist in equilibrium. The second paper starts with a puzzle. While large urban areas should enhance labor market opportunities, married women’s labor force participation decreases with metropolitan size. Large urban areas are also subject to more congestion and related commute volatility which is particularly costly for married women who are disproportionately responsible for household tasks such as running errand and chauffeuring children. I consider the effect of commute volatility on two labor market outcomes, participation and quality of the job match where job match quality is measured as whether an individual with a professional degree (e.g. Doctor of Medicine, Juris Doctor, Doctor of Veterinary Medicine) is working in a professional occupation (e.g. physician, lawyer, veterinarian). Previous work on women’s labor force participation has focused on the role of commute time. New in this paper, I find that commute volatility more so than commute time discourages participation perhaps because long, but predictable commutes can be planned around while variable commutes are harder to manage. However, I find evidence that both commute time and volatility affect job choice as either may induce women to limit their search radius or favor job flexibility over match quality. Finally, while I cannot perfectly rule out the possibility of sorting, the direction of sorting incentives suggests that commute volatility adversely affects married women’s labor force outcomes either directly or by inducing them to locate in lower amenity but less-congested cities like Detroit, Pittsburgh, and Cleveland.


Open Access