TO THE DEATH and destruction left in Hurricane Sandy’s wake, add this concern: One of the federal government’s key programs for coping with such disaster is itself underwater.

We refer to the National Flood Insurance Program (NFIP), which protects 5.6 million customers in 21,000 communities across the country — both coastal and inland — against losses from inundation. The NFIP has $1.25 trillion of coverage in force, of which just under half is in seaside areas such as the Jersey Shore. In fact, there are a quarter-million policyholders in New Jersey, insured up to $350,000 each for homes and $1 million for businesses. At the moment, the NFIP has access to about $4 billion, plus a $20.8 billion credit line with the U.S. Treasury — of which it has already borrowed $18 billion.

How did the program reach this financial predicament? Congress established it in 1968, three years after Hurricane Betsy ravaged the Gulf Coast. Then, the private sector generally avoided flood insurance because the risks were too great and too unpredictable; but, as Betsy showed, without insurance, the federal government would be exposed to huge episodic demands for aid.

As with any insurance scheme, the NFIP faced moral hazard — the tendency of customers to take on more risk because they believe that someone else will pay. In other words, even a well-designed plan would have encouraged excessive development in coastal areas, putting more property and population at risk. And for most of its history, the NFIP has not been well-designed. The program required communities in flood-prone areas to mitigate hazards. It also tried to compel participation by linking federal backing for home loans in flood-prone areas to the purchase of flood insurance. Alas, government flood-zone maps were notoriously inexact and much-disputed by local governments and real estate interests. Premiums were set below the true costs of coverage — and many policyholders let their insurance lapse.

A disproportionate share of the benefits flowed to well-to-do vacation homeowners and developers. A disproportionate share of the expenditures covered repeat losses on older properties in flood plains — which, through one of the program’s perverse quirks, happened to pay the lowest premiums.

Congress never endowed the NFIP with a reserve fund or required it to build one. Pay-as-you-go worked — until 2005, when hurricanes Katrina, Rita and Wilma struck, and the NFIP had to borrow $17 billion from the Treasury. Hence the program’s massive and probably unpayable debt.

In July, Congress reauthorized the program for five years and included reforms such as the gradual set-aside of a reserve fund, the phase-out of subsidized insurance for second homes and repeatedly flooded properties, a phased-in premium increase and $400 million in annual budget authority for flood-map modernization. The program was also authorized for the first time to buy reinsurance from private companies. This is progress; whether the NFIP is sustainable post-Sandy is another question.

Congress must revisit the NFIP, perhaps with an eye toward such radical measures as writing off its debt, permitting flood-prone states to join forces and replace the federal government — and considering a greater role for the private sector. Forty years ago, insurance companies were smaller and less technically sophisticated than they are today. Now they might be able to hedge some of the huge flood risks that would otherwise fall entirely on taxpayers.