<p dir="ltr">The National Flood Insurance Program (NFIP), which is run by the U.S. Federal Emergency Management Agency (FEMA), is presently under huge debt to the U.S. treasury. The debt is primarily caused by low flood insurance take-up rate, low willingness to pay for flood insurance, and large payouts after major disasters. Addressing this insolvency problem requires the NFIP to understand (1) what drives the demand for flood insurance so that it can be increased, (2) how risk factors contribute towards large flood insurance payouts so that effective risk reduction policies can be planned, and (3) how to predict the future flood insurance payouts so that the NFIP can be financially prepared. This research has answered these three fundamental questions by developing empirical models based on historical data. To answer the first question, this research has developed a propensity score-based causal model that analyzed one of the key components that influences the demand for flood insurance – the availability of post-disaster government assistance. It was found that the availability of the federal payout in a county in a year increased the number of flood insurance policies by 5.2% and the total insured value of the policies by 4.6% in the following year. Next, this research has developed Mixed Effects Regression model that quantified the causal relationships between the annual flood insurance payout in a county and flood related risk factors such as flood exposure, infrastructure vulnerability, social vulnerability, community resilience, and the number of mobile homes in the county. Based on the derived causal estimates, it was predicted that climate change, which is expected to increase flood exposure in coastal counties, will increase the annual NFIP payout in New Orleans, Louisiana by $2.04 billion in the next 30 years. Lastly, to make the NFIP financially prepared for future payouts, this research has developed a predictive model that can predict the annual NFIP payout in a county with adequate predictive accuracy. The predictive model was used to predict the NFIP payout for 2021 and it was able to predict that with a 9.8% prediction error. The outcomes of this research create new knowledge to inform policy decisions and strategies aimed at fortifying the NFIP. This includes strategies such as flood protection infrastructure, tailored disaster assistance, and other interventions that can bolster flood insurance uptake while mitigating the risk of substantial payouts. Ultimately, this research contributes to sustaining the NFIP's ability to provide vital flood insurance coverage to millions of Americans.</p>