Risk is an important component of the decision-making process. Often time, risk assessment is associated with either space or time. How agents perceive risk and how they respond to risk can have significant policy implications, especially when government programs are designed to either incentivize the provision of environmental amenities or reduce the production of environmental disamenities. This dissertation features three chapters that examine the role of risk, time, and space in evaluating environmental disamenities and amenities in the context of climate adaptation and ecosystem goods and services.

The first chapter studies the spillover effects of levee building in response to rising flood risks in the U.S. Mississippi. Using newly digitized data on levee locations and elevations with the Great Mississippi Flood of 2011 as a natural experiment, I show that a 1% increase in the upstream levee elevation increased the downstream levee elevation by 0.7%. A back-of-the-envelope calculation suggests the external costs due to upstream levee building are at least $0.2 billion, reducing the net benefits of heightened levees by 55%. The results highlight the importance of regional coordination to manage large-scale natural disasters while mitigating inter-jurisdictional spillovers.

The second chapter uses a discrete choice experiment implemented in a farmer survey to elicit landowners’ willingness to enroll in long-term payments for ecosystem services programs in Maryland. We address the issue of serial non-participation (SNP) when landowners always choose the status quo option and examine the role of time and risk preferences in landowners’ enrollment decisions. We find that ceiling on program participation is evident when SNP is accounted for, pointing to an inherent limitation in voluntary programs. Failing to account for SNP can also lead to quantitatively different welfare measures. Landowners are responsive to program payments with low discount rates consistent with market interest rates. Risk-averse landowners are less likely to enroll in programs, suggesting that they perceive participation to increase income risk. The third chapter proposes a novel extension of existing semi-parametric approaches to examine spatial patterns of willingness to pay (WTP) and status quo effects, including tests for global spatial autocorrelation, spatial interpolation techniques, and local hotspot analysis. We incorporate the statistical precision of WTP values into the spatial analyses using a two-step methodology and demonstrate this method using data from a stated preference survey that elicited values for improvements in water quality in the Chesapeake Bay and lakes in the surrounding watershed. Our proposed methodology offers a flexible way to identify potential spatial patterns of welfare impacts and facilitates more accurate benefit-cost and distributional analyses.