Chairman Hensarling, Ranking Member Waters and members of the committee,
My name is R.J. Lehmann, and I am the editor-in-chief, co-founder and a senior fellow with the R Street Institute. I appreciate the opportunity to testify and greatly appreciate the work done by committee members and staff to draft legislation to reform and reauthorize the National Flood Insurance Program. We strongly support many aspects of the legislation set to come before the committee, including provisions to encourage more private sector competition and choice, to shift more risk to the private sector through reinsurance and to limit taxpayer exposure to NFIP losses.
R Street is a nonpartisan, nonprofit public policy research organization based here in Washington, D.C. Through our research and outreach, we seek to promote free markets and limited, effective government at both the state and federal level, with a particular focus on issues that might be considered relatively low salience and high complexity. We have a commitment to work with broad coalitions and, wherever possible, to build support for pragmatic market-oriented proposals that can earn bipartisan consensus. As one notable example of that commitment, like the National Wildlife Federation and Taxpayers for Common Sense, R Street is a part of the SmarterSafer coalition, a group of taxpayer advocates, conservation groups, insurance interests and housing advocates that supports NFIP reform.
R Street’s insurance project has been a core part of our mission since we opened our doors five years ago. In fact, the first white paper we ever published was a “report card” evaluating the regulatory environment for insurance in each of the 50 states, which we have continued to update annually. Our work seeks to highlight the crucial role that competitive private insurance markets play in sending price signals that allow society to better evaluate, mitigate and manage risk. It is in no small part thanks to risk-based insurance rates produced through the price discovery process of private-sector competition that America has significantly safer roads and workplaces today than it did 50 or 100 years ago, as the prospect of lower premiums have offered strong incentives for employers, automakers and drivers to opt for safer behavior and safer processes.
However, such price signals can be muted or deliberately distorted where underwriting and ratemaking decisions are subject to explicit government-imposed price controls, or where private insurance and reinsurance capacity is displaced by taxpayer-backed entities, such as the federal programs for crop insurance, trade credit insurance and terrorism reinsurance and state-backed residual markets for home and auto insurance. The National Flood Insurance Program is among the largest and most distortionary of these government insurance entities. In 1966, when Congress initially deliberated the legislation that would eventually authorize the NFIP, the Presidential Task Force on Federal Flood Control Policy warned that creating a federal program that provided “insurance in which premiums are not proportionate to risk would be to invite economic waste of great magnitude.” That prediction has, unfortunately, come to pass.
Over much of its history, the NFIP has subsidized irresponsible floodplain development at taxpayers’ expense. For example, a December 2014 report by the Government Accountability Office found that, thanks to subsidized policies, the program collected $11 to $17 billion less in premiums than was actuarially prudent over the dozen years from 2002 to 2013. Despite reforms passed by this committee and ultimately signed by President Obama in 2012, NFIP premiums still do not reflect the full risk of loss.
When property owners don’t bear the full cost of the risks they face, they are encouraged to take on more. Thus, it should not surprise anyone that, over its 50-year history, the NFIP has literally changed the landscape, allowing acres of lush river valleys and miles of coastal land to be transformed into manicured lawns and beachfront cottages. Today, more than half the U.S. population lives in coastal counties, up 45 percent from 1970 to 2010. It is no coincidence that, according to Federal Emergency Management Agency’s own statistics, more than 90 percent of all presidentially declared national disasters involve flooding.
Moreover, those development and population trends show no signs of slowing down, even as rising sea levels and other changes in climate patterns merit heightened concern about the risk of flooding. A 2013 report that appeared in the journal Nature projected that annual global flood losses would be expected to rise to between $60 and $63 billion by 2050, compared to the current $6 billion, even if governments made significant investments in flood defenses. But even more striking was the researchers’ projection that average global flood losses would rise to at least $52 billion a year by 2050 based solely on economic development and growth in coastal populations, ignoring the expected effects of climate change.
Given such projections, the NFIP is not sustainable in its current form, as evidenced by its $24.6 billion debt to federal taxpayers. To prepare for these shifting risks, to ensure that markets receive and incorporate appropriate price signals and to protect taxpayers from the exploding cost of disaster assistance, it is essential that Congress move, as Chairman Hensarling has proposed, to “take up legislation to transition to a private, innovative, competitive, sustainable flood insurance market.”
Transitioning to a private, risk-based insurance market for floods will not be easy, but it is a challenge private insurers and reinsurers can meet, just as they have done in countries such as the United Kingdom. Advances in mapping, risk modeling and the ability to spread risk across the globe means that many of the logistical problems the insurance industry once faced in attempting to underwrite flood risks have been significantly addressed, if not completely resolved.
Indeed, as this committee moves forward with reforms, I encourage members to recognize the burgeoning private flood insurance market for what it is: an example of entrepreneurial innovation. The insurance industry may have a public image as stodgy and old-fashioned, but that doesn’t mean it’s incapable of breaking new ground. I would argue that an insurance contract is a kind of technology. Some might quibble with that claim, but there can be no question that the catastrophe models that insurers use to project potential losses employ some of the most advanced computational algorithms in the world. Shifting flood insurance to the private sector will mean bringing those powers to bear to much more accurately segment and price property-level risks. It also will mean having companies’ marketing and underwriting teams compete to fashion products that are more attractive to policyholders and better meet their needs.
Requiring risk transfer
One area where considerable progress already has been made in transferring flood risk off the backs of taxpayers and onto the private market is in reinsurance. Even if all of the NFIP’s policies were charged full risk-based rates—a scenario that remains some years in the future—the nature of catastrophe risk dictates that the program will sometimes experience extraordinarily large events that exceed anticipated losses. Losses from such events, notably 2005’s Hurricane Katrina and 2012’s Superstorm Sandy, continue to account for the bulk of the program’s debts.
Private market insurers curtail their exposure to such catastrophes through the use of reinsurance, commonly described as “insurance for insurance companies.” The NFIP historically has instead relied on loans from the federal Treasury in years when losses exceed available resources. The Biggert-Waters Act clarified FEMA’s authority to execute reinsurance contracts for the NFIP, which can take the form of traditional facultative or treaty reinsurance or various forms of collateralized reinsurance, such as catastrophe bonds or other insurance-linked securities. Earlier this year, FEMA executed the first such risk transfer, ceding $1 billion of flood exposure risk to a consortium of 25 private reinsurers.
We are pleased to see the legislation set to come before the committee would incorporate Rep. Luetkemeyer’s Taxpayer Exposure Mitigation Act, which would make it a requirement that FEMA’s administrator use reinsurance and other risk-transfer tools to lower taxpayers’ direct exposure to catastrophic losses. As FEMA gains more experience buying reinsurance, and as reinsurers gain more experience absorbing risk from the NFIP, we anticipate that future risk transfers could be significantly larger.
The NFIP purchases reinsurance using the resources accumulated in its Reserve Fund, which also was established by the Biggert-Waters Act. The fund is required to hold at least 1 percent of the NFIP’s total potential loss exposure, which stood at $1.25 trillion at year-end 2016. It is capitalized by an assessment levied on all policies, which currently is calculated at 15 percent of total premium, as well as an additional surcharge of $25 for each residential policy and $250 for each nonresidential and nonprimary residential policy. By law, the annual combination of the assessments and surcharges are required to total at least 7.5 percent of the mandated Reserve Fund, or about $937.5 million annually, based on current NFIP exposures.
We welcome provisions of the draft legislation that would require FEMA’s administrator to raise the assessment rate by 1 percentage point each year until the program meets the minimum reserve ratio phase-in of 7.5 percent. We think the language could be further improved if, rather than either a flat percentage assessment (which serves actually to magnify inequities between policies that pay subsidized or grandfathered rates and those that pay full-risk rates) or flat fee surcharges (some of which would change under the current draft), the committee were instead to consider applying Reserve Fund charges that are based on the catastrophe risk posed by each individual property. Private insurers already use similar catastrophe loads to buy reinsurance and the NFIP should, as well.
Encouraging private flood insurance
While the private market for primary flood insurance remains comparatively small, it is not insubstantial and most sources agree that it is growing. While FEMA collected $3.3 billion in premiums last year from its 5.1 NFIP million policies, according to data from S&P Global Market Intelligence, there was $412.6 million of standalone private flood insurance written in the United States in 2016, the first year for which data are available. The total includes coverage written on an excess-of-loss basis, although it does not capture flood risks covered by umbrella policies or by multiperil commercial insurance policies.
The largest private flood markets can be found in California, with $48.8 million of direct premium, and Florida, with $47.8 million. Florida is notable not only in that it is the largest participating state in the NFIP—with 1.8 million policies, representing nearly 40 percent of the program—but also in that the state Legislature moved in 2014 to create a statutory framework to regulate four different kinds of private flood insurance policies, which has served as a legislative model adopted in a handful of other states.
But there remain a number of issues that serve to hinder growth of the private market, several of which are addressed in the draft legislation before the committee.
First and perhaps simplest to accomplish, the draft legislation would remove a longstanding restriction that prohibits insurers who participate in the Write Your Own program from selling standalone coverage outside of the NFIP. As companies with experience marketing flood insurance policies and adjusting flood claims, these are the insurers that arguably are best positioned to enter the market for private flood. The legislation includes transparency safeguards to ensure policyholders are clear as to whether a given policy is underwritten by the NFIP or a private company, which should address the only potential objection to removing this otherwise needless restriction.
We also are pleased that the legislation incorporates the Flood Insurance Market Parity and Modernization Act, sponsored by Reps. Ross and Castor, which previously passed the House unanimously. While the Biggert-Waters Act required regulated lending institutions to accept private flood insurance to satisfy the Flood Disaster Protection Act of 1973’s mandatory purchase requirement, the market still doesn’t have clarity from federal banking regulators about which private policies will be deemed compliant. This legislation would clarify that policyholders who leave the NFIP for private coverage and later return are considered to have continuous coverage and that both admitted market and surplus lines policies may qualify for the mandatory purchase requirement. We think the language could be further strengthened to be self-executing, so that lenders and borrowers could choose to accept private coverage today, without having to continue to wait for rulemaking from federal banking regulators, which could take years.
One area where the legislation may fall short of the mark is in the access it would grant insurers to NFIP claims data. While the draft includes language that would require FEMA to make information relevant to assessing flood risk and identifying and establishing flood elevations and premiums available to the public through an open source system, we are concerned that these data are not sufficiently granular to serve the purposes for which it would be needed. The legislation provides that data would be made available at the ZIP code or census-block level, while insurance underwriters would need claims and policy information at the individual property level. FEMA has raised the concern that it is barred by the Privacy Act of 1974 from releasing more specific information. This argument is lacking. For one, the most sensitive data that would actually interest insurers—the price the policyholder paid for the property—is information that now is readily available to the public through apps like Zillow and Trulia. Moreover, to the extent that other data could be considered personally identifiable, insurers and reinsurers routinely consent to nondisclosure agreements to address such concerns.
I also would like to address the concern—raised by some members and other observers—that a more active private market for flood insurance could destabilize the NFIP by cherry-picking low-risk policies until it is rendered an underfunded high-risk pool. Almost every piece of this charge misses the market. The program already is unstable and underfunded. Its $25 billion in debt is sufficient testimony to that reality. It already serves, in essence, as a high-risk pool. In the United Kingdom, where flood insurance is wrapped into all-risks homeowners coverage sold by private insurers, about 95 percent of homeowners have coverage for flood risks. In the United States, only about 14 percent of homeowners do.
The vast bulk of existing NFIP policyholders are required to purchase coverage, meaning they face at least a 1 percent chance of flooding in any given year. This is a high-risk cohort. There are, by and large, no cherries to pick. Reducing the size of the program will necessarily reduce its overall exposure and the potential burden it could place on taxpayers.
Further countering this concern are provisions in the proposed legislation that would ensure the NFIP’s highest-risk properties are either mitigated or forcefully pushed into the private sector. One such provision would bar NFIP coverage for any property whose lifetime claims, subsequent to the law’s passage, amount to more than twice its replacement value. Another would bar coverage for new construction in high-risk special flood hazard areas, as well as residential structures whose replacement costs top $1 million, where private flood insurance options are available. We support these provisions.
Rates and affordability
While the Ross-Castor language would help address technical issues that represent obstacles to further flowering of the private flood insurance market, the single biggest impediment remains the fact that the program does not charge sufficient risk-based rates, making it nearly impossible for private companies to compete for some NFIP business.
For roughly 1 million policies—representing 20 percent of the program—rates are explicitly “subsidized.” These are properties that joined the NFIP before the introduction of flood insurance rate maps in the mid-1970s and have never been charged a risk-based premium. Indeed, FEMA reported earlier this year that 97 percent of the program’s subsidized properties lack flood elevation certificates. This means that, effectively, not only do they not pay what actuarial guidelines would recommend, but neither FEMA, Congress nor the policyholders themselves have any solid information about what they should be paying or what sort of financial exposure they represent for the program. As others have suggested, we strongly recommend this committee require FEMA to follow the State of North Carolina’s lead and contract to use LIDAR or other modern technologies to obtain property-level elevation data, which would reduce the burden on policyholders to obtain elevation certificates.
The best available estimates from the Government Accountability Office are that subsidized properties currently pay about 35 percent to 40 percent of their full risk-based rates. Moreover, nearly 80 percent of subsidized properties are located in counties that are among the top 30 percent of the nation in home values, while only about 5 percent of subsidized properties are located in counties in the bottom half of the home-price distribution. While the 2012 and 2014 reform bills set nearly all of these policies on the path toward risk-based rates, we would urge the committee to revisit changes made in the 2014 bill to slow that progress.
Another set of policies—whose precise number FEMA has, to date, been unable to quantify—are “grandfathered” into rates associated with lower-risk flood zones even after changes in flood maps have reclassified them into higher-risk zones. While we do not know how many grandfathered properties there are, FEMA has asserted that, taken together with other properties within their risk classes, they produce enough premiums to cover expected losses. If this claim is true, by definition, the other properties within such risk classes are contributing more in premium than their actual risk characteristics would merit. This would make them natural targets for competition from the private sector. FEMA is currently engaged in a project to gather information on grandfathered properties, which it has said it expects to complete by September 2018.
We support moving to risk-based rates for all NFIP properties, with an understanding that, for some lower-income policyholders, assistance may be needed to ensure those rates are reasonably affordable. Such assistance must be targeted, limited, means-tested and executed outside the rate structure of the NFIP. We think the draft legislation’s proposal to authorize states to create affordability programs for lower-income policyholders, financed by surcharges on other NFIP policyholders within that state, represents an important step in the right direction. Not only does this comport with the principles of federalism, but states already collect income data for a variety of means-tested programs, which should ease what might otherwise be substantial compliance costs.
On the other hand, we oppose the legislation’s proposal to decrease the cap on annual rate increases from 18 to 15 percent and we strongly oppose the provision to impose a hard cap of $10,000 on the risk premium that can be charged to any single-family residential property. While this provision would affect very few properties—we understand that there are only 763 properties nationwide out of roughly 3.5 million residential policyholders that pay more than $10,000 annually in premium—we fear the precedent set by any such arbitrary rate cap. Once introduced as a statutory mechanism, it could be lowered by a future Congress or possibly even by executive order. Moreover, it ignores that any policy that would be charged an annual premium of more than $10,000 for what are maximum benefits of just $350,000 ($250,000 for the structure and $100,000 for contents) must, by definition, pose an extraordinarily high level of risk. Any offer of premium relief made to such policyholders must, at a minimum, be conditioned on agreeing to some level of pre-disaster mitigation.
In closing, I’d like to reiterate our support for the broad contours of the proposed legislation, which in many respects mirrors similar proposals the SmarterSafer coalition put forward earlier this year in our National Flood Insurance Program Reform Proposal. I urge committee members who haven’t done do to read that proposal, which also includes detailed analysis of such topics as mapping and mitigation, which I’ve only touched on briefly here.
Making the transition to private flood insurance will be complicated endeavor, although not nearly as complicated as continually rebuilding flood-prone communities that have no incentive to adapt and mitigate risk because that risk is borne by others. I would be glad to answer any questions the members might have.
Image by Juergen Faelchle