In the wake of Hurricane Harvey, Eileen and Jeff Swanson faced the unthinkable. They had just paid off the last of the mortgage on their $225,000 home in the Canyon Gate neighborhood of Houston, where they lived with two sons, one of whom is severely developmentally disabled. During the storm, a foot of water inundated their home, and in its wake, they faced $60,000 in costs to repair the damage. Like many Houston residents, the Swansons had no flood insurance.
The lesson one might be tempted to draw is the one Federal Emergency Management Agency (FEMA) Administrator Brock Long promoted in a recent congressional hearing: that reform of the National Flood Insurance Program (NFIP) should focus on ensuring that those at risk have flood insurance to speed up their recovery. While this is a desirable goal, it fails to address more fundamental structural problems with the NFIP that the Swansons' predicament reveals. Ensuring everyone at risk is covered will help people to recover after disaster strikes but at a high cost in suffering and dollars. Reform must also enlist the NFIP as a tool to help avoid predictable flooding of people’s homes, businesses, and communities.
That requires that we look at several root causes of the Swansons' problem: that the local government allowed construction in an area intended — not just expected, but intended — to flood in extreme storms, and no federal, state, or local law required either special measures to reduce the risk to homes or disclosure of this risk to prospective buyers. (The Barker Reservoir, built by the Army Corps of Engineers in the 1940s, was not an obvious source of risk to the Swansons or other purchasers. In dry weather, it is a park with trails and sports fields.)
How can we enhance the resilience of communities through an ounce of prevention, instead of simply paying for a pound of cure over and over after each new flood? The path to achieving exactly this was blazed when the NFIP was first enacted. To find the way forward requires that we look back to the program’s history.
The National Flood Insurance Program celebrated its 50th anniversary in 2018. At middle age, it is not in good shape. The program is some $20.5 billion in debt to the federal treasury as of February 2018 (even after the Treasury Department canceled an additional $16 billion of debt after Hurricanes Harvey, Irma, and Maria). At the same time, the cost and severity of flood damage have not diminished. What could account for the NFIP’s widely acknowledged failings?
As the program has been implemented by the federal government and amended by Congress over the past half century, it has strayed far from the original goals of the 1968 Congress that enacted the underlying legislation. Today, many probably believe that Congress intended to permanently subsidize federal insurance to protect property constructed in flood-prone areas. Nothing could be further from the truth.
Third time is the charm? Beyond levees and disaster relief
Congress did not turn to federal flood insurance until two other federal efforts to minimize flood damage had fallen short. Early in the 20th century, flood prevention and recovery had been the responsibility of state and local governments. When they proved unequal to the challenge, the federal government, primarily through the Army Corps of Engineers, gradually began to offer assistance. At first, the Corps followed a “levees only” policy, relying solely upon the construction of levees to contain floodwaters. Despite those early efforts, the Great Mississippi River Flood of 1927 caused about 500 deaths and left at least 700,000 people homeless. In response, Congress passed the Flood Control Act of 1928, through which the federal government took on a broader role in engineered flood control, authorizing the construction of more federal levees, but also spillways, floodways, reservoirs, and other structures to protect floodplain development.
As flood damage continued to accrue, Congress began to experiment with a second important response to flooding: disaster relief. Through the Disaster Relief Act of 1950, Congress authorized post-flood financial assistance from the federal government. But such relief, funded by taxpayers, also proved to be inadequate. This lesson hit home after Hurricane Betsy of 1965, a Category 3 storm, made landfall in Florida and Louisiana, killing 75 people and submerging tens of thousands of homes, some up to their rooftops. Betsy was the nation’s first “billion-dollar hurricane” in terms of flood damage (about $7.9 billion, adjusted for inflation).
Just three years later, Congress tried yet a third approach: flood insurance. When it passed the National Flood Insurance Act of 1968, Congress intended to defray the expense of after-the-fact disaster relief by encouraging floodplain occupants to pay insurance premiums into an insurance pool before disaster struck. As the House of Representatives explained in its report on the pending legislation, disaster relief from the federal government and voluntary relief agencies had proved inadequate, thereby “underlin[ing] the need for a program which will make insurance against flood damage available, encourage persons to become aware of the risk of occupying the flood plains, and reduce the mounting Federal expenditures for disaster relief assistance.” Congress was well aware that any insurance program — especially one with federal subsidies — could actually “aggravate rather than ameliorate” flood danger by giving floodplain occupants a false sense of security and creating what the insurance industry calls “moral hazard” — the propensity for excessive risk-taking by those who do not bear the full cost of risky actions such as floodplain development. To avoid such hazard, Congress incorporated three critical components into the National Flood Insurance Program.
Temporary federal subsidies
The task force report that gave rise to the NFIP in 1968 originally estimated that federal subsidization of the cost of flood premiums for existing high-risk properties would be required for a limited period of time only — approximately 25 years. As the House Report asserted, “Any Federal ‘subsidy’ which will accrue under the insurance program to the benefit of property owners now occupying the flood plain is defensible only as part of an interim solution to long-range readjustments in land use. . . .” Existing floodplain structures were grandfathered in and their insurance premiums available at federally subsidized rates, but the House Report explained that such a temporary partial subsidy for new properties “is not at all valid.” Instead, Congress assumed that after existing floodplain structures completed their useful lives, the program would be turned over to the private insurance industry, which would charge full actuarial rates that reflected the full measure of risk assumed by those who chose to build new structures within areas at high risk of flooding. At that time, the task force report explained, “private insurers would take over the bulk of the program, charging full, risk-based actuarial premiums, and the federal government would have no liability, except with possible reinsurance against catastrophic losses.”
State and local land use regulation
How would private insurers be able to provide economical insurance at some future date? The key lies in the state and local land use regulations that Congress envisioned as the centerpiece of the NFIP. In the statute’s statement of purpose in Section 1302(c), Congress found that “a program of flood insurance can promote the public interest by providing appropriate protection against the perils of flood losses and encouraging sound land use by minimizing exposure of property to flood losses. . . .” In fact, under Section 1305(c), federal insurance would be available only to participating communities that provided satisfactory assurances that they were adopting permanent land use and control measures, with effective enforcement mechanisms, in conformity with federal criteria to be developed by the Secretary of the Department of Housing and Urban Development. Further, the law made federal disaster assistance unavailable for losses covered by the flood insurance program, or that could have been so covered by landowners in participating communities, with exceptions for low-income individuals.
Partial floodplain retreat
Thus, state and local land use regulation was an essential cornerstone of the National Flood Insurance Program. Such regulation would perform at least two critical functions, as stated in the NFIP’s declaration of purpose. First, it would “constrict the development of land which is exposed to flood damage and minimize damage caused by flood losses.” Second, regulation would “guide the development of proposed future construction, where practicable, away from locations which are threatened by flood hazards.” If fully implemented, these “constrict” and “guide away” principles could have done much to protect lives and property from the ravages of floodwaters, as well as insulate the federal fisc from unsustainable costs. Instead, regulatory efforts were thwarted by many factors, including a growing antipathy toward regulation and the rise of the regulatory takings doctrine, as further explained elsewhere in this report.
The path forward calls for a return to the principles articulated in the National Flood Insurance Act of 1968 — providing only temporary subsidies, ensuring enactment of sound land use regulation, encouraging partial retreat, and advancing social equity. The repeatedly postponed reauthorization of the NFIP provides an ideal opportunity for Congress to focus on these goals. A February 2018 Public Opinion Strategies poll commissioned by the Pew Charitable Trusts shows strong public support across political parties for policies that more fairly allocate the costs of flood insurance and emphasize prevention, rather than just recovery, by incorporating sensible mitigation measures.
Phase out federal subsidies
An important concern of the 1968 legislators was to “encourage persons to become aware of the risk of occupying the flood plains.” Appropriate pricing of flood insurance is a critical way of accomplishing this goal. The Biggert-Waters Flood Insurance Reform Act of 2012 tried to phase out subsidies rapidly but was met with a severe backlash. The subsequent Homeowner Flood Insurance Affordability Act of 2014 also included a phase-out of subsidies, albeit on a more gradual schedule. Importantly, the 2014 legislation also called for measures to enhance affordability, hearkening back to an original component of the flood insurance program.
In some cases, the rich are able to remain in the floodplain and to elevate their structures so as to qualify for federal flood insurance, or to buy property without a federally backed mortgage and therefore escape the need for federal flood insurance. Care needs to be taken in subsidizing insurance policies — even for low-income individuals and families — because it may only work to keep people in harm’s way. Premium support must be coupled with steps that reduce the risk to vulnerable populations.
The Trump administration is proposing a funding mechanism that appears to be a bad idea: transfer NFIP risk to the capital markets. This partial privatization is superficially appealing and seemingly consistent with the original NFIP goal of moving from federal subsidies to private insurance. However, such reforms could easily end up allowing private insurers to “cherry pick” the properties with the lowest risk — similar to letting private health insurers take on only healthy people, leaving the government to pay for the rest.
Reinvigorate state and local land use regulation
It is time to live up to one of Congress’s original purposes in enacting the flood insurance program: to encourage strong state and local land use regulation. This may require action at the federal, state, and local levels, and by the judicial as well as legislative branches.
Congress should review and strengthen the incentives for local governments to adopt tough limits on new development in floodplains and areas subject to flooding in extreme events. In April 2018, Houston adopted building standards that, according to one report, “could have spared 84 percent of the buildings flooded by Hurricane Harvey.” The standards increased the elevation required for new buildings from one foot to two feet of elevation above the 500-year floodplain. The incentives provided by the NFIP to local governments were clearly inadequate to achieve the program’s goals. Allowing local governments to wait until after flooding strikes shifts the losses from developers — who are better able to assess and factor costs into their decisions — to residents and the public at large — who generally lack both the expertise and the basic facts to enable them to accurately assess their risks.
Congress should also strengthen the requirements of the NFIP to insist that flood maps on which the federal insurance program, local communities, and residents rely are updated to reflect the true risk presented in an era of climate change, as noted elsewhere in this report.
Federal judicial decisions also play a significant role in deterring local governments from adopting needed land use restrictions. Local officials fear they will incur liability for a regulatory taking if they adopt the minimum standards needed to protect the public’s health, safety and property. The U.S. Supreme Court should mitigate this as it elaborates how the takings clause applies in future cases where local governments are seeking to address the slow-moving emergency of flood risk. Other creative ideas like insurance for regulatory takings claims could also take the chill off needed local regulation.
Encourage partial floodplain retreat
Beyond phasing out subsidies for insurance premiums, it is important to encourage the removal of more buildings from the floodplain altogether through voluntary buy-out programs. This would help to solve the well-documented “repetitive loss” problem, under which a small number of high-risk properties take up a disproportionately large proportion of insurance payouts. Section 1323 of the National Flood Insurance Act, added in 2004, provides a repetitive flood claims grant program to mitigate structures, which includes acquisition or relocation of at-risk structures. Even before the most recent rounds of hurricanes, for example, Harris County, Texas, bought out more than 3,000 flood-prone properties between 1985 and 2015, using federal and local loans and funds. This amounted to a purchase of more than 1,000 acres that were restored as natural floodplains, which the county estimated saved at least 1,500 homes from flooding during one storm alone (the so-called “Tax Day Flood” in April 2015). This program could be expanded, perhaps partially funded through premium increases over time, emphasizing buyouts and retreats over the incomplete solution offered by elevation of structures.
Provide better signaling
There are many sources that document how inaccurate FEMA’s floodplain maps are, how they fail to take advantage of the best available data, and how they fail to take into account the reality of climate change. The 2014 Act requires the mapping program to use “the most accurate topography and elevation data available,” which would appear to require incorporation of the best available climate change projections where these affect elevation. In Houston, for example, many homeowners, like the Swansons, did not realize they were within identified “flood pools” where stored flood waters could be released periodically. This represents a failure of signaling, as well as an abdication of responsibility by local government by allowing homes to be built within the known flood pool.
Requiring disclosure of a property’s location within a flood zone or flood pool, as some states require, is only meaningful if the flood zones reflect risk accurately. Sellers may already be required to disclose past flooding under state statutory or common law standards, but this can be difficult and costly for a misled buyer to enforce. State legislatures should update their disclosure statutes to account for this.
Since its enactment, the NFIP has included a focus on supporting the most vulnerable in our communities. Reforms of all aspects of the NFIP should incorporate needs-based distinctions that provide relief to those who need it most. With the growing deficit in the NFIP and the prospect of ever more extreme storms, subsidizing those with adequate resources may not be a sustainable strategy. In addition, greater transparency by FEMA in reporting on the types of assistance provided, income levels of those receiving assistance, and overall cost could help ensure that support is directed where it is needed most.