How flooding affects home values can determine the path of economic recovery for communities and have lasting impacts on national and global financial systems. Yet, our understanding of how flood insurance, community risk perception, and past flooding events shape future housing prices remains limited. To explore this, we used a Socio-Environmental (SE) model and studied the temporal impacts of flooding on mean housing values across 496 coastal census tracts of New York, Connecticut, and New Jersey, US, from 1970 to 2021. The modeling exercise demonstrated that the initial economic impact of Hurricane Sandy was largely absorbed by the National Flood Insurance Program (NFIP); however, the region then exhibited a long-term decline in home values, which was well described by an Interrupted Time Series (ITS) model. We found significant correlations between SE model parameters describing housing price change and those describing tract-scale behaviors and perceptions, suggesting that the salience of past flooding events and NFIP participation may be important regional drivers of housing prices. Tracts with greater post-flood change in active insurance policies exhibited larger decreases in mean home values than those with more stable NFIP participation. An improved understanding of relationships between housing prices, flood insurance, and community perceptions could support more equitable distributions of resources and improved policy interventions to reduce flooding risk.