Flood Risk Repricing and Housing Market Adjustment after Hurricane Harvey

• In Event: Student Research Panel: Housing Markets, Climate Adaptation, and Place-Based Investments

Abstract

Flooding induced by climate change affects housing markets through at least two channels: structural damage and changes in the valuation of risky locations after a major event. These channels matter for housing and urban policy because they shape not only property values, but also rebuilding, relocation, and the distribution of climate risk across neighborhoods. We study how Hurricane Harvey changed housing values and post-disaster adjustment in Harris County, Texas, and how these effects varied by realized flood severity, official flood-risk designation, neighborhood income, and pre-disaster flood insurance coverage.

We use parcel-level tax assessor data for single-family homes from 2014 to 2020 and decompose assessed property market values into building and land components. This allows us to distinguish structural losses from more persistent locational repricing. We combine these data with realized inundation depth, FEMA Special Flood Hazard Area (SFHA) designations, neighborhood National Flood Insurance Program (NFIP) take-up, and measures of post-disaster ownership turnover. We estimate difference-in-differences event-study models with parcel fixed effects and block-group-by-year fixed effects, comparing non-flooded homes to those that experienced different flood depths within the same local market and year, while accounting for submarket flood exposure and geographic proximity to water.

Three findings emerge. First, housing-market responses are strongly non-linear in flood depth. Flooding below 3 feet is associated with limited value change, while economically meaningful losses occur once inundation exceeds 3 feet. Moderate flooding is more strongly associated with declines in building values, reflecting structural damage that remains repairable, whereas more severe flooding is associated with larger and more persistent land-value declines, indicating locational risk repricing. Second, official flood-risk designation through the 100-year floodplain matters. Before Harvey, homes within the SFHA carried a premium, especially in land values, suggesting that homeowners placed greater value on water-adjacent amenities than on ex-ante flood risk. After Harvey, that premium reverses for heavily flooded homes, particularly in higher-income neighborhoods. Third, insurance preparedness is associated with both recovery and mobility. In moderately flooded SFHA areas, higher pre-disaster NFIP take-up is associated with smaller declines in building value, suggesting that insurance helps ease short-run liquidity constraints. Supplementary building permit evidence also indicates that moderate flooding is primarily linked to disaster-repair activity, and severe flooding is more closely associated with demolition. Outside FEMA-designated flood zones, higher NFIP take-up is associated with greater post-disaster ownership turnover, including a 2.7 percentage-point increase in the 3–6 foot flood bin.

These findings suggest that ex-ante floodplain designations and insurance take-up influence not only housing values, but also who can rebuild, who can relocate, and how climate risk is distributed across urban neighborhoods. For housing and community development policy, the results highlight the importance of pre-disaster insurance coverage and risk communication in shaping post-disaster recovery.

Author

• Hyunjung Cho, University of Illinois at Urbana Champaign
  Presenting Author