Climate change presents new risks for property in the United States. Due to the high cost and sometimes unavailability of location-specific property risk data, home buyers can greatly benefit from acquiring knowledge about these risks. To explore this, a large-scale nationwide natural field experiment was conducted through Redfin to estimate the causal impact of providing home-specific flood risk information on the behavior of home buyers in terms of their search, bidding, and purchasing decisions. Redfin randomly assigned 17.5 million users to receive information detailing the flood risk associated with the properties they searched for on the platform. Our analysis reveals several key findings: (1) the flood risk information influences every stage of the house buying process, including the initial search, bidding activities, and final purchase; (2) individuals are willing to make trade-offs concerning property amenities in order to own a property with a lower flood risk; (3) the impact of the flood risk information on behavior is more pronounced for users conducting searches in high flood risk areas, but does not differ significantly between buyers in Republican and Democrat Counties; and (4) the information resulted in changes to property prices and altered the market's hedonic equilibrium, providing a new finding that climate adaptation can be forward-thinking and proactive.
Authors
Daryl Fairweather, Matthew E. Kahn, Robert D. Metcalfe, Sebastian Sandoval Olascoaga
Related Organizations
- Acknowledgements & Disclosure
- Kahn, Metcalfe, and Sandoval-Olascoaga did not receive any funding or payment for the research from Redfin or First Street Foundation. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research. Fairweather is a current employee of Redfin. Sandoval-Olascoaga is a former employee of and current contract worker for Redfin.
- DOI
- https://doi.org/10.3386/w33119
- Pages
- 114
- Published in
- United States of America
Table of Contents
- Introduction 1
- Primary Findings 6
- Relationship to Existing Literature 10
- The Field Experiment 13
- Background 14
- The Design of the Field Experiment 16
- Implementation of the Field Experiment 17
- The Unbiasedness of the Treatment Effect Due to Randomization 19
- Demand Side Hypotheses Through the Lens of a Hedonic Housing Assignment Model 20
- Data and Empirical Strategy 24
- Data 24
- The Empirical Strategy 25
- Pre-Experiment Balance 27
- Results 29
- Home Search Behavior Dynamics 29
- Nonparametric Conditional Average Treatment Effects 38
- Engagement 39
- Tours, Offers, and Closes 42
- Does Political Ideology or Recent Flooding Events Influence the Treatment's Effectiveness? 51
- Real Estate Market Price Responses to Property Specific Flood Risk Information 52
- Conclusion 58
- Appendix 73
- Theory extension 73
- The Treatments 73
- Additional figures and tables 75
- Nonparametric Heterogeneous Treatment Effects 94
- Six Categories 98
