The Role of Flood Insurance on Population and Demographic Changes

• In Event: Housing policy levers for equitable climate adaptation

Abstract

Introduction/Background
The National Flood Insurance Program (NFIP) provides federally subsidized flood insurance to improve access to financial protection against flood damages. However, by historically setting premiums below actuarially fair levels, the program raises concerns about moral hazard—specifically, whether it distorts residential location decisions by encouraging households to settle in flood-prone areas, as it lowers the private cost of flood damages. This issue is highly relevant to current public policy debates. With the NFIP’s recent transition to a new risk-based premium pricing methodology (Risk Rating 2.0), this study is especially timely for understanding baseline behavioral responses and assessing whether Risk Rating 2.0 will reverse the historical concentration of populations in risky areas.

Purpose/Research Question
This paper asks whether the availability of subsidized flood insurance causally affects where people choose to live. It addresses critical limitations in prior studies, which generally rely on county-level data that fails to align with the NFIP’s actual community-level implementation (e.g., cities and towns). Furthermore, this research seeks to answer how these spatial behavioral responses vary across different socioeconomic groups, an area where empirical evidence remains limited.

Data & Methods
To achieve precise measurement of population changes within communities, this study leverages highly granular population data from the Gridded Environmental Impact Frame (EIF) at the 0.01-degree grid level. The empirical strategy identifies the causal impact of NFIP availability using a stacked difference-in-differences design combined with an instrumental variables (IV) approach. This method exploits plausibly exogenous variation in the timing of FEMA’s Flood Insurance Rate Map (FIRM) issuance, which was distributed at the community level and served as a prerequisite for program participation. The validity of this instrument is supported by historical Government Accountability Office (GAO) reports indicating that the FIRM rollout lacked systematic prioritization.

Results/Findings
Contrary to prior county-level findings, the granular analysis reveals that NFIP availability does not induce population growth in high-risk areas, yielding no evidence of moral hazard. Instead, population increases are concentrated in lower-risk communities, where populations rise by approximately 20 percent. Additionally, the effects are highly heterogeneous across income levels: middle-income communities experience the largest gains (approximately 29 percent), while behavioral responses among low- and high-income groups remain small and statistically insignificant.

Conclusion/Implications
These findings indicate that subsidized flood insurance facilitates demographic growth in relatively safer locations rather than incentivizing settlement in the riskiest areas. The main mechanisms driving this result could be that participating communities prioritize inland community growth after NFIP participation through the implementation of building ordinances and enhanced local flood management. By aligning the empirical analysis with the NFIP's institutional structure, this study provides vital new evidence on the behavioral and spatial impacts of subsidized insurance. Future work will extend this analysis to examine heterogeneous effects across racial groups and assess the program's broader distributional impacts. Ultimately, these results provide crucial evidence for policymakers tasked with balancing risk-based pricing and equity considerations in the ongoing reform of federal disaster insurance programs.

Author

• Sungha Jo, Indiana University
  Presenting Author