The National Flood Insurance Program: Drowning in Debt and Due for Phase-out

Report Government Regulation

The National Flood Insurance Program: Drowning in Debt and Due for Phase-out

June 22, 2017 23 min read Download Report
Diane Katz
Diane Katz
Former Senior Research Fellow in Regulatory Policy
Diane Katz was a research fellow in regulatory policy at The Heritage Foundation.


The federal government holds a monopoly on primary flood insurance for homeowners and businesses, and the program is debt-ridden and dysfunctional. A large proportion of the flood-risk maps are obsolete, and thus the premiums charged under the National Flood Insurance Program do not reflect actual risk. Because property owners do not bear the full cost of flood risk, they are more likely to locate in flood-prone areas and less likely to undertake preventive measures. The devastation of natural disasters is worsened as a result. Tinkering with operational reforms cannot remedy a program designed by Congress to be financially unsound. A private market in flood insurance is the ultimate solution.

Key Takeaways

The federal government holds a monopoly on primary flood insurance for homeowners and businesses, and the program is debt-ridden and dysfunctional.

The NFIP owes taxpayers nearly $25 billion as a result of borrowing from the U.S. Treasury to cover damage claims.

The best course of action is to phase out government flood insurance and enable private insurance to replace it.

The federal government issues virtually all primary flood insurance for homeowners and businesses, and the National Flood Insurance Program (NFIP) owes nearly $25 billion to the U.S. Treasury and taxpayers for covering claims. Beyond this alarming debt, the NFIP cannot keep up with risk assessments and its subsidies actually promote development in flood zones, which worsens the devastation of natural disasters. Congress is now debating reforms to the NFIP. The best course of action is to phase out the deeply flawed program and enable private insurance to replace it.

Background on the NFIP

The NFIP was established in 1968[REF] to provide flood insurance for at-risk properties and to mitigate flood risks through land-use regulation. Congress noted at the time that ad hoc disaster relief was placing “an increasing burden on the nation’s resources,”[REF] which could be alleviated by insurance coverage.

Authorization for the NFIP is set to expire on September 30, 2017, and discussion drafts of legislation to renew it are circulating in both the House and the Senate. Unless the NFIP is reauthorized, the Federal Emergency Management Agency (FEMA), which administers the program, will be barred from writing new policies (although existing ones would remain in force).

Congress last renewed the NFIP in 2012, taking steps to reduce subsidies and base rates on a property’s flood risk—an essential element of viable insurance. However, the congressional “flood caucus” in 2014 successfully reversed many of the reforms at the behest of local politicians and property owners who benefit from the subsidies.[REF]

Tinkering with operational reforms will not remedy the distortionary incentives inherent in a government insurance scheme—especially because the NFIP, as designed, is financially unsound. Problems include:

  • Wealth redistribution. The use of taxpayer funds to subsidize the lifestyle preferences of a select few who live in or near flood zones is inherently unjust.
  • Dysfunctional pricing. A large proportion of the FEMA risk maps are obsolete, and thus insurance rates do not reflect actual risk.
  • Moral hazard. Property owners expect the government to provide disaster assistance regardless of their insurance status. Consequently, NFIP enrollment is skewed to the most flood-prone properties.
  • Uncontrolled costs. More than one-third of premiums are paid to private insurers who sell and service the policies but hold no risk liability. With direct access to the U.S. Treasury, FEMA has little budgetary discipline.
  • Unintended negative consequences. Because property owners do not bear the full cost of flood risk, they are less likely to be dissuaded from locating in flood-prone areas and less likely to undertake preventive measures.

How the NFIP Works

Some five million properties are currently covered under the NFIP. Property owners are eligible if their community adopts and enforces floodplain management regulations that meet or exceed federal standards.[REF]

FEMA has little discretion in issuing policies, regardless of the degree of flood risk or repetitive claims.[REF] For example, property owners who receive financial assistance from the federal government following a presidentially declared disaster are generally required to maintain flood insurance coverage.

For purposes of rate-setting and mitigation planning, FEMA develops Flood Insurance Rate Maps (FIRMs) of flood-prone communities. Areas in which there is a 1 in 100 or greater risk of annual flooding are designated as Special Flood Hazard Areas (SFHA). Properties within these areas require flood insurance if the mortgage was issued by a federally regulated lender, a federal agency lender, or a government-sponsored enterprise.[REF]

NFIP policies have coverage limits:[REF]

  • Residential:
    • Structure Coverage: $250,000
    • Contents Coverage: $100,000
  • Non-residential:
    • Structure Coverage: $500,000
    • Contents Coverage: $500,000

Properties that have repeatedly claimed losses are serviced separately by FEMA’s Special Direct Facility. State and local officials are notified when such a designation is made to facilitate community sponsorship of mitigation. Property owners who decline an official mitigation offer are charged higher premiums.

For areas with moderate flood risk, the NFIP offers the less expensive Preferred Risk Policies (PRP). Whether they are within a hazard zone or in an area with only moderate risk, policyholders are required to purchase Cost of Compliance policies to cover the added expense of rebuilding to newer stricter construction codes than were originally in place.

To be eligible for participation in the NFIP, communities must enact floodplain management standards and regulate development to reduce flood risk. Policyholders may see reductions on insurance premiums if their communities undertake mitigation to improve their status under FEMA’s Community Rating System (CRS). The discount may range from as little as 5 percent to as much as 45 percent based on the degree to which mitigation actions exceed the minimum federal standards.[REF] Federal grants are available for projects that reduce the risk of flood damage to insured structures.[REF]

Under the Write Your Own (WYO) program, FEMA contracts with private property and casualty insurers to sell and administer identical flood insurance policies, although they bear none of the risk. Between 2012 and 2016, the insurers received a commission of 15 percent of net written premiums.[REF] Claim adjustment expenses are also reimbursed.

According to the Government Accountability Office (GAO), most of the payments are not reimbursements of actual expenses, but allowances on which the companies can either make a profit or incur a loss. FEMA lacks the information necessary to determine whether its compensation payments are appropriate.[REF]

Under the 2012 reauthorization, Congress directed FEMA to develop within one year a methodology for accurate payment of WYO expenses. As of August 12, 2016, however, FEMA had not revised its compensation structure.

Measuring Risk

A confluence of factors determines the risk designation of a property, including:

  • Elevation of the property,
  • Building structure,
  • Date of construction, and
  • Community’s flood history.

The accuracy of FEMA’s Flood Insurance Rate Maps also varies greatly depending upon the age of the map and the technology used to create it.[REF] FEMA is supposed to re-evaluate all floodplain areas and flood-risk zones at least once every five years, but the agency is unable to keep up—a consequence of budgetary and bureaucratic constraints.[REF]

Flood studies can take from three years to 10 years. According to a review of the agency’s mapping process by its Technical Mapping Advisory Council:[REF]

  • “FEMA continues to utilize the paper cartographic-driven process in constructing FIRMs, which is time consuming and adds expense to the process.”[REF]
  • The unduly prolonged process also “results in products that are outdated before they are completed.”[REF]
  • “In some cases, pertinent newer data that is developed during the mapping cycle is not incorporated into the final product, and it may be many years before updated information is published.”[REF]

The government’s monopoly on insurance coverage also politicizes the process of risk mapping, which now involves numerous special interests and various stakeholders.

The agency’s inventory of flood hazard maps encompasses more than one million “flood miles.”[REF] (Some analyses indicate that only one-third of the potential miles of floodplain in the nation have been studied and included in FEMA’s inventory.[REF]) However, of the mapping in FEMA’s inventory:

  • Only 49 percent is designated as “valid,” meaning that the map “adequately identifies the level of flood risk.”[REF]
  • Another 11 percent is designated as “unverified,” which is FEMA-speak for deficient.
  • In addition, 39 percent is “unknown,” as in yet to be validated.[REF]

In its review, the Technical Mapping Advisory Council concluded, “Many populations across the Nation are not covered in updated, valid flood studies and are therefore subject to unknown flood risk.”[REF]

One of the barriers to accurate mapping is a lack of the data necessary to calculate flood probabilities. In a 2016 review of FEMA’s rate-setting methods, agency staff reported that they began to gather information on water depths in 2010, but still lack enough data to conduct a statistically valid probability assessment.[REF] FEMA staff reported that they will have to gather information over the next five to 10 years before the data allows for a statistically valid assessment.

Another hurdle to risk mapping is the cost of obtaining accurate elevation certificates for more than one million properties, which will likely take years and cost several hundred million dollars to obtain if administered through the NFIP.[REF]

The accuracy of the risk maps is also undermined by FEMA’s practice of crediting an entire community for a single action that may impact only one structure.[REF] Moreover, once a mitigation action is completed, subsequent actions are essentially ignored.[REF] These practices distort the measurement of risk and thus thwart actuarial rate-setting.


NFIP coverage has two types of premiums: “full-risk” and “subsidized.”

  1. Full-risk rates are supposed to be actuarially sound, that is, cover anticipated losses and administrative expenses.[REF] According to FEMA, about 80 percent of policyholders pay full-risk rates. However, absent accurate risk mapping, the agency cannot determine a credible full-risk rate.

  2. Subsidized premiums, established by Congress to encourage enrollment in the NFIP, apply to properties built before a community’s flood-risk map was issued, or before January 1, 1974. Property owners who have maintained continuous coverage since originally enrolling in the NFIP, or whose property was built in compliance with NFIP standards, have the option of paying a “grandfathered” premium tied to a previous rate map.[REF]

    The financial impact of these subsidies on the program is largely unknown because FEMA failed to track grandfathered properties before 2010. However, in congressional testimony earlier this year, Roy E. Wright, FEMA’s deputy associate administrator, said the fiscal solvency of the NFIP depends on setting premiums that accurately reflect flood risk: “[T]he program needs to be on a course to eventually arrive at full risk rates for all policyholders. This includes addressing grandfathered and subsidized rates.”[REF]


NFIP funding is provided by congressional appropriation as well as premiums, fees, and surcharges paid by policyholders. The flood-risk mapping is largely supported by discretionary funds,[REF] while premium revenues cover most program operations and claims.

Under its 2012 authorization, the NFIP was granted authority to borrow up to $30.425 billion from the U.S. Treasury—a necessity because the program was designed with subsidies and would not generate the funds necessary to fulfill claims in the event of a major flood.

  • In 2005, after 15 years of lower-than-expected flooding, Hurricanes Katrina, Rita, and Wilma hit the U.S.[REF] To meet damage claims, the NFIP borrowed $17.5 billion from the Treasury.
  • In 2012, Hurricane Sandy hit the East Coast, and the NFIP borrowed an additional $6.25 billion.
  • In 2016, Hurricane Matthew brought on multiple floods, requiring the NFIP to borrow another $1.6 billion from the Treasury to cover claims and pay interest on the outstanding debt.

Interest on the debt costs $400 million annually, or about 11 percent of all premium dollars.[REF] To date, the NFIP has paid $2.9 billion of premium revenue to service the debt. FEMA’s Roy E. Wright acknowledged that the subsidies drive the debt, and there is no way to repay it under the current rate structure. “There is no practical way for us to pay nearly $25 billion in debt. Full stop,” he stated.[REF]

In 2012, Congress directed FEMA to establish a reserve fund to offset claim expenses in the event of a disaster that would otherwise deplete the insurance fund. (The assessment is calculated as 15 percent of a total premium.[REF]) In 2014, Congress added a surcharge of $25 for primary residences and $250 for all other properties for deposit in the Reserve Fund.

FEMA is required by statute to maintain a Reserve Fund balance of 1 percent of the prior year’s total loss exposure. Officials say it could take 20 years or more to reach the balance required by statute. The Reserve Fund was tapped to pay $1.3 billion in claims in 2016. A Federal Policy Fee of $50 is also charged on all new policies and renewals “to defray administrative expenses.”[REF]

House Proposals

Draft legislation circulating in the House proposes a five-year reauthorization of the NFIP. However, the need for immediate and far-reaching reforms argues for a shorter reauthorization in order to assess the progress of reforms.

Any reauthorization must avoid entrenching further the dysfunctional elements of the NFIP.

Among the reform proposals circulating in the House, a few are worth serious consideration, including:

  • Establish equivalency for private insurance. A market for flood insurance depends upon allowing private coverage to fulfill the mandatory purchase requirements for federally regulated lenders, a federal agency lender, or a government-sponsored enterprise. Public access to flood records and risk ratings is also essential (without personally identifiable information).
  • Disclose premium methodology. Accountability requires transparency. FEMA should be required to divulge its methodology for determining premiums.
  • Base rates on replacement cost. FEMA’s current practice of setting coverage based on a national average for various structures fails to account for vast differences in property values across local communities and states. Rates based on replacement cost, by structure, would help to achieve actuarial soundness.
  • Allow pro rata refunds. The NFIP should be required to provide refunds on a proportional basis to policyholders who cancel during a policy term to obtain private coverage.
  • Adjust risk mapping and premiums. FEMA should be required to adjust all NFIP premiums for actuarial risk within four years of enactment.
  • Expand rights of appeal. Policyholders should be allowed to appeal a full or partial denial of claim from their WYO insurer. States, local governments, and property owners should be allowed to appeal Special Flood Hazard Area designations and FEMA denials of requests to update flood maps.

Some proposals have merit, but need adjusting:

  • End NFIP coverage. The NFIP should cease underwriting new coverage for future structures built in Special Flood Hazard Areas four years after enactment. However, this provision should not be made contingent on “local availability of private flood insurance” or the “affordability” of private coverage. Such caveats would deter property owners from purchasing private policies on the assumption that the NFIP would remain available. Foregoing coverage should limit eligibility for other disaster assistance.
  • Transfer risk. Risk-transfer tools such as reinsurance, catastrophe bonds, and other insurance-related securities could help to reduce the NFIP’s debt burden on taxpayers until the program is phased out. However, FEMA is a service agency, not a financial brain trust, and authority for implementing risk transfers would be best handled by U.S. Treasury officials who already oversee the program’s debt.
  • Limit lifetime losses. A small number of “repetitive loss properties” are responsible for a disproportionate share of claims, and limits should be imposed on their coverage. It is reasonable to debate where to set the payment threshold, but waiting until 2021 to impose such a limit is not. It is also unreasonable to make such a limit contingent on the availability of private insurance. The point is that repetitive losses should result in either a substantial redesign of the dwelling or a move to safer ground (although a private insurer should be free to underwrite coverage at their discretion). The NFIP should not grant “prioritized mitigation assistance” or “100 percent cost share” to multiple-loss properties. Doing so would create an incentive to forestall repairs and remodeling. 
  • Prohibit coverage for excessive risks. The NFIP should be barred from insuring any property with lifetime losses that in the aggregate exceed twice the amount of the replacement value of the structure. However, the calculation should include payments made before enactment of the new rules.
  • Limit WYO compensation. Controlling the costs of servicing insurance policies is reasonable. However, as the GAO documented, FEMA lacks a rational compensation methodology.[REF] Establishing an arbitrary ceiling (in this case, 25 percent of a premium) hardly inspires thrift. The better approach would be to determine actual costs for various policy services, and allow insurers to bid for the business.
  • Allow community mapping. Allow localities to develop flood-mapping alternatives to the NFIP. FEMA should not be authorized to set mapping standards given its failure to maintain its own. The same mapping privilege should be extended to private insurers, who actually possess the most expertise in gauging risk.
  • Ensure better WYO oversight. Questionable handling of claims after Superstorm Sandy indicates a lack of oversight. But rather than expand FEMA’s oversight powers in light of its past failures, the agency’s inspector general should be directed to ensure that WYO litigation expenses are appropriate.

Several other proposals are unfeasible, including:

  • Conducting an independent actuarial study to determine long-term losses, and quarterly reports to Congress on the policyholder pool and risk profile. With more than half of FEMA’s land inventory lacking valid risk assessments, a credible actuarial study would take tens of millions of dollars and years to complete. Likewise, quarterly reports on policyholders would be inaccurate because flood risks are assessed at the community level, not the property level, which does not allow for valid profiles. Even with valid data, the unpredictability of disasters makes any forecasts of losses speculative.
  • Requiring FEMA to use other risk-assessment tools. FEMA does not possess accurate, standardized risk-assessment measures for individual structures, which limits their usefulness in rate-setting. Nor would a single federal point system accurately reflect the diverse local conditions across the U.S. A variety of other factors already do inform rates, including the flood zone, location, age of the building, building occupancy, and foundation type.
  • Basing premium rates on up-to-date replacement cost, by structure. Most premiums are currently based on the cash value of the structure and its contents, applying national averages. Replacement cost coverage is available for single-family homes. Configuring an accurate replacement cost structure for millions of homes and businesses in the program would increase administrative costs and increase premiums.

Several proposals are particularly flawed given the need for reform, including:

  • Increasing the “Affordability Surcharge.” Rather than require full-risk rates, Congress in 2014 approved a surcharge to compensate for the shortfall in NFIP revenue. To increase that surcharge for primary and secondary residences does not address the root cause of the problem: the continuation of subsidies and the lack of actuarially sound rates.
  • Ensuring solvency. The NFIP was designed to subsidize flood insurance, and lawmakers are considering a variety of other coverage-related entitlements. Therefore, it would be irrational for Congress to direct the FEMA administrator to ensure that the NFIP remains financially sound.
  • Imposing equivalency fees on private insurance. Property owners who purchase private insurance should not be forced to pay for FEMA’s dysfunctional mapping system. One of the primary benefits of an insurance market would be more accurate risk assessment.
  • Limiting annual rate increases to 15 percent. The NFIP’s debt spiral is largely a function of its actuarially unsound rates. Increases are currently capped at 18 percent annually (and a total premium of $10,000 annually for a single-family property). Lowering the cap to 15 percent would further delay sound underwriting and increase the burden on taxpayers to subsidize the lifestyle choices of a select group of homeowners.
  • Awarding community credits for environmental amenities. Current subsidies deprive the NFIP of the funds needed for disaster recovery as well as burden taxpayers with the program’s perpetual debt. Providing new subsidies for enhancing a floodplain’s natural benefits would only divert resources from more pressing needs.
  • Increasing the Reserve Fund assessment. A financial cushion for the NFIP would certainly be an improvement over the NFIP’s current debt, but the real remedy is to charge actuarial rates for all policyholders rather than increase fees.
  • Authorizing state affordability programs. Creating a new flood insurance entitlement for low-income homeowners would only worsen the NFIP’s fiscal woes and undermine the shift to a more beneficial private system. Adding yet another surcharge on other policyholders to pay for it all would make flood insurance all the more unaffordable—leaving more property owners without it (and at the mercy of ad hoc disaster assistance). Homeowners who cannot afford to insure their property must make choices based on their income, not subsidies.
  • Conducting a study on coverage for cooperative housing. FEMA should not be taking any steps toward new types of coverage. Private insurers can customize policies based on demand.

Privatization Is the Ultimate Solution

While a few of the proposals circulating in the House would facilitate the availability of private flood insurance, the ultimate solution is to eliminate the subsidies and other giveaways that secure the government’s flood insurance monopoly. Private insurers are interested in underwriting wide swaths of properties in flood zones. The benefits of phasing out the NFIP are reflected in the differences between the government-run program and the private sector.

The NFIP has fiscally irrational policy objectives, such as offering subsidized insurance premiums that do not reflect the full risk of flooding to encourage program participation and community-based floodplain management, and reducing the reliance on federal disaster assistance.

Private insurers focus on a different set of objectives. These include ensuring rate and capital adequacy, maintaining solvency, and producing a return on investment.

The differences between the NFIP and private insurers apply to rate-setting as well. For example, the NFIP generally accepts all applicants regardless of an individual’s property risk and sets rates across a smaller number of broad rate classes. Private insurers generally insure applicants based on individual property risks and a larger number of more specific rate classes.

Opponents predict that private insurers will cherry-pick customers, leaving behind property owners with the highest risk. But according to the Reinsurance Association of America,[REF] this fear has not borne out. Two data points in particular show this is not likely to be the case:

  1. Most private companies in Florida are writing insurance in higher hazard areas and are not choosing the least-risky properties; and
  2. Private insurance “take outs” led to a much smaller and stronger state-run insurance program.

The availability of customized options through private companies is likely to increase the number of homeowners who insure against flooding. Such an arrangement would certainly enhance the finances of the NFIP—and unburden taxpayers.

Diane Katz is a Senior Research Fellow for Regulatory Policy in the Thomas A. Roe Institute for Economic Policy Studies, of the Institute for Economic Freedom, at The Heritage Foundation.


Diane Katz
Diane Katz

Former Senior Research Fellow in Regulatory Policy