Hurricane Maria and Housing Market in Puerto Rico

The deadliest hurricane ever recorded in Puerto Rico, Hurricane Maria, made landfall in 2017. Hundreds of thousands of homes were damaged, and millions lost power for days. This study seeks to investigate how that devastation affected the housing prices in Puerto Rico. We collected 1001 single-family house data from the Zillow website between 2018 to 2021. For the causal inference (treatment-effect), distance buffer from the tract, location (house is on the right or left side of the hurricane path), and flood zone (house located in a flood zone or not), etc., were used as part of the primary identification strategy. We also combined the traditional hedonic price model with Regression Discontinuity Design (RDD) to measure the hurricane's causal (treatment) effect on housing prices. First, findings from the basic difference-in-difference hedonic price models indicated a downward trend/pattern of housing prices in post-hurricane years. We then used sharp and fuzzy RDD models with single and multiple cut-offs to estimate similar specifications. The RDD results also confirm the negative trend of falling prices. Given that findings from traditional hedonic pricing models may be plagued by the small samples of recent sales information, we ran various specifications of the spatial hedonic models using Kriging to find the best predictor of housing prices in Puerto Rico. Identifying the best functional form of spatial hedonic models can help Puerto Rican policymakers, and future researchers analyze housing price fluctuation in Puerto Rico in the aftermath of a major natural disaster