There is growing acknowledgment of the need to restore degraded environments. This paper studies optimal investment in ecosystem restoration under environmental change. We develop an optimal control model of the restoration decision to explicitly characterize the optimal extent and timing of restoration given time-dependent marginal damages. We provide the first results on optimal dynamic investment in ecosystem restoration, highlighting the important role that growth in restored patches plays in shaping the time profile of investment. We then apply the model in a numerical simulation of coastal wetlands restoration in Huntington Beach, California, that accounts for projected sea-level rise, uncertainty over flooding severity, and the option to abandon damage properties. Our results show that early investments in restoration are optimal in order build up a wetlands stock that can mitigate future flooding damages exacerbated by sea-level rise. We find large option values associated with delaying irreversible decisions to abandon damaged properties.

After temporarily rising during the subprime boom, U.S. homeownership rates plunged for young households before slowly rising in the late 2010s. We identify four key drivers of the homeownership rate of the young. The first is the relative cost of owning versus renting. The second is the severity of mortgage credit constraints, which delay the transition to homeownership. The role of credit constraints predated the subprime boom but became prominent in the wake of the subprime boom and bust. The third factor encompasses secular changes in family structure, proxied by the age when mothers first give birth. A fourth factor is the rising burden of student debt that has delayed the accumulation of downpayments and the transition to homeownership. We create a novel dataset of the average debt service burden ratio from the early 1980s on for each age cohort. Using a cointegrating model framework, we find that all four factors are significant drivers of the homeownership rate of young households, and that not accounting for college debt gives rise to omitted variable bias. Our study makes two contributions. First, we provide new time series data on the evolution of college debt service burdens. Second, we develop time series models of homeownership that jointly analyze the impact of access to credit, student debt, family factors, and the cost of owning to renting.

Community Organization and Risk Management

The goal of this study is to examine the role of community organization in self-protection in a model of interdependent risk, with a focus on wildfire risk. We present the first theoretical model that incorporates both self-protection and community organization, and complement it with a spatial analysis to estimate the benefits within and beyond organized communities. This work contributes to understanding how collective action can reduce risk and generate spillover effects.