Can private insurance help the National Flood Insurance Program meet its congressional expectations?
There is nothing like a statutory deadline to focus legislators on updating programs. At a minimum, Congress is to vote on reauthorizing the National Flood Insurance Program (NFIP) by September 30, or it will lapse. The NFIP, part of the Federal Emergency Management Agency (FEMA), is the nation’s primary flood insurer for consumers and small businesses.
“There is huge pressure to resolve it in a timely manner and not let it expire,” says Jim MacGinnitie, senior property/casualty fellow at the American Academy of Actuaries (the Academy).
Beyond reauthorizing the NFIP, the direction of reform is significantly different than in the past. As lawmakers consider changes to address the agency’s multifold challenges, they are also contemplating provisions to encourage more private insurers to offer flood coverage. The NFIP currently sells up to $250,000 in property coverage to residential customers and $500,000 for businesses.
Many insurers are interested in offering flood insurance, says Nat Wienecke, senior vice president of federal government relations for the Property Casualty Insurers Association of America (PCI). “There are quite a few companies that feel they can offer not just equal but better coverage and can be price competitive,” he adds.
Private sector interest in flood insurance stems from a convergence of developments. Reinsurers, which have been backing private sector policies in other countries, possess a growing appetite for offering flood insurance in the United States. Excess capacity has been triggered by several factors, including a decade of low interest rates, which compels reinsurers to seek out new markets, said Rade Musulin, vice president of the Academy and chair of its Flood Insurance Work Group. In other countries, private insurers often profitably underwrite flood insurance.
Musulin explains that there is also greater data availability and technological innovations that are piquing insurer interest. (See sidebar.) “Modeling technologies similar to those used in other fields, such as earthquake and tropical cyclones, have helped improve many other markets including flood,” he adds.
Other stakeholders, such as the Center for Economic Justice, a consumer advocacy group, and state regulators through the National Association of Insurance Commissioners (NAIC), see expanding privatization as a way to better serve consumers. “The NAIC supports legislative efforts to facilitate the growth of a state-regulated private flood insurance market,” reads the organization’s letter to Congress dated May 2, 2017, “to help provide consumers with more choices and coverage potentially at more affordable prices.”
Strengthening the private market, however, could also weaken the NFIP’s position to meet its legislative mandates. This raises the public policy question: What is the best way to ensure that American lives and property will be protected from current and long-term losses? “The right way forward is to blend the best of both worlds to produce good public policy,” Musulin says.
Considering this question and related implications begins with understanding why the federal government became a public flood insurer in the first place.
Why the NFIP?
The NFIP was a long way in coming. By 1929, the private insurance industry effectively stopped offering coverage, according to “A Chronology of Major Events Affecting the National Flood Insurance Program,” published by FEMA in 2005. In 1952, President Harry S. Truman pushed for a federal flood insurance program, recommending $50 million to start it, the report notes. Congress passed the National Flood Insurance Act of 1968, which included creation of the NFIP.
Congress tasks FEMA with several roles to serve the public interest, with the purpose of reducing federal financial exposure to flood loss, improving community resilience to flooding, mitigating potential damage to properties and speeding recovery after a loss. This article focuses on the NFIP’s role as the nation’s primary insurer against weather-related flood loss, the private sector’s potential expansion into the flood insurance market and potential implications.
The conflicts, in short, are that the NFIP must collect sufficient premium to cover losses and provide subsidized rates to a significant number of policies and charge affordable premiums to expand participation and pay off its debt to the U.S. Treasury.
Since its inception, the NFIP’s insurance role has been fraught with conflicts stemming from various congressional mandates, according to The National Flood Insurance Program: Challenges and Solutions, written by the American Academy of Actuaries’ Flood Insurance Work Group and published in April. The conflicts, in short, are that the NFIP must collect sufficient premium to cover losses and provide subsidized rates to a significant number of policies and charge affordable premiums to expand participation and pay off its debt to the U.S. Treasury.
The NFIP started operations in 1969. Since that time, the agency has been able to borrow money from the U.S. Treasury as needed and pay it back, according to “Flood Insurance: Comprehensive Reform Could Improve Solvency and Enhance Resilience,” released by the U.S. General Accountability Office (GAO) in April.
That changed in 2005, when hurricanes Katrina and Rita — among the top 10 most costly U.S. storms — contributed to an additional $16.6 billion of debt to the U.S. Treasury, according to the GAO report. Hurricane Sandy added $6.25 billion more to the debt in 2012. As of March 2017, NFIP owed $24.6 billion to the Treasury.
The NFIP also faces program participation and retention issues. It is common for people to buy flood coverage after a major weather event only to drop it later, according to the NAIC’s Center for Insurance Policy and Research’s report, “Flood Risk and Insurance,” released in April. The NAIC report also cites research estimating that about half of those who live in a 100-year-flood plain, also known as Special Flood Hazard Area (SFHA), do not purchase coverage and that participation outside of SFHAs is low.
According to FEMA.gov, NFIP’s average flood claim payout is $43,000 and the average annual premium is $700. Flood insurance penetration is in the high single digits in the United States, indicating a large potential market, Musulin says. As of March 31, there were 4.99 million policies in force nationwide, generating $3.5 billion in premium (excluding surcharges) and $1.2 trillion in coverage.
This is a mere slice of the overall potential need for flood insurance. The potential total property exposure to flood loss, including buildings and contents, was estimated to be more than $90 trillion for the year 2014, assuming that all the nation’s properties need flood coverage, according to “Increasing Concentrations of Property Values and Catastrophe Risk in the US,” a report published in 2015 by Karen Clark & Company, a catastrophe modeling firm.
Of that $90 trillion, 45 percent is residential property and another 24 percent are industries with mostly small businesses, says Karen C. Clark, the company’s president and CEO. This means that the potential coverage need could be more than $62.1 trillion.
Unpacking Some NFIP Challenges
The NFIP faces several challenges in satisfying congressional mandates. Perhaps most notably, the agency does not collect sufficient revenue to pay its losses, especially those caused by catastrophic flood events. This is largely due to past congressional directives requiring coverage to be affordable and because catastrophe loss potential historically has not been fully factored into rates.
The NFIP looks at rating differently than private insurers would, Musulin says, because the agency was created to address many problems beyond offering affordable and available flood coverage. “The NFIP has a tool kit built to solve what Congress has asked it to solve and Congress has changed priorities over time,” he explains, “The tool kit is different than what insurance companies would use.”
It can be argued that underwriting flood risks is much more difficult than any other major peril. Coastal storm frequency tends to be low by geographical location and severity can be quite high. Add to this the many causes of flooding, including: rain, rapid snow and ice melt; failure of dam or levee; or just unusually high tides.
Further, robust flood insurance models have only recently become available. As result, they are relatively untested in the United States, according to the Academy report. While catastrophe hurricane models have been around for years, inland flood models were just released a couple years ago, the Academy report notes.
“Developing rates between areas with riverine flooding exposure and those on the coast is very different,” Musulin says. Catastrophic coastal floods have low frequency but extremely intense losses often driven by tropical cyclones,” he explains. Meanwhile, “riverine areas have more gradual losses that occur in a much wider area and generally affect the first floor as opposed to an entire structure,” he adds. Currently, the NFIP does not distinguish rating between these very different perils.
In the midst of this, the NFIP has a rating structure designed to meet the congressional mandate to make flood insurance more affordable. About 78.5 percent of NFIP policyholders pay “full-risk rates,” which differ from actuarial rates that insurers would charge, according to the Academy report. This majority of NFIP policyholders pay premiums to help subsidize the other 21.5 percent.
These subsidies, according to the Academy report, only apply to the $60,000 basic limit of coverage. However, while the average subsidy is about 50 to 55 percent of full-risk premium, the average rate per subsidized policy is nearly three times the average price of full-risk rated policies, the report notes. This does not include the NFIP’s additional fees, including the policy surcharge for mapping and paying off past debt. Musulin says that these considerations would not be included in a private sector rate base.
Homeowners are often not aware of their subsidies or their properties’ risk potential. Instead of baking subsidies into the rates, MacGinnitie says that the NFIP should transparently indicate discounts to policyholders. This will help insureds understand their degree of risk, which should encourage flood damage mitigation, he explains.
One key point Musulin makes is that FEMA and private insurers have significantly different mandates, which helps explain why its approaches to problems sometimes seem very different. A specific example he cites is that FEMA is interested in a forensic understanding of a neighborhood for flood mitigation, leading the federal agency to invest significant resources in mapping.
The NFIP relies on these same maps for rating, Musulin explains. The maps identify risks in SFHAs (also known as the 100-year-flood plain) and those that are not. This triggers many government requirements, including community mitigation plans and mandatory purchases of flood insurance to qualify for mortgages.
Location of property on FEMA maps oversimplifies the risk variation among properties. This can give consumers a false impression that coverage is not necessary, experts say. This approach is problematic for attracting and retaining customers, MacGinnitie says. “If people believe they are on the right side of the line they think they do not need coverage.”
This false impression unnecessarily exposes consumers to absorbing damages the NFIP was created to cover, sources say. This also leads rating inequities among policyholders. In contrast to the NFIP, MacGinnitie says, “The private market will have zones or individual risk rating and charge you less the further you get from the river.”
Musulin says that private insurers take a more risk-level view consistent with their use of house-specific pricing. The NFIP does consider specific property factors, including elevation, category of FEMA flood risk zone, occupancy type, number of floors and the nature of basement or crawl space, says Andy Neal, the agency’s chief actuary. In contrast, Musulin says, private insurers take a more robust individual property assessment for finer rating and include factors such as the property’s value, type of electrical wiring, power cut-off and contents.
“The real problem is that we are building things that will last 50 years (absent being washed away) based on today’s exposure to floods when in fact we need to be considering the conditions that will exist over the design life of the building.”
—Rade Musulin
The NFIP needs forward-looking maps to account for rising sea levels, Musulin says. Currently, he explains, the maps reflect today’s hydrology. FEMA has tools to forecast future conditions but these are generally limited to looking at the effect of development on drainage such as parking lots, though they could be used for examining sea-level rise. “The program’s large inventory of coastal properties makes consideration of rising sea levels critical, particularly when you consider the NFIP’s effect on building codes,” he adds.
While private insurers are considering entry into the flood market due to new tools to assess flood risk, the NFIP is in the process of deploying many of them. “We are looking to incorporate current practices industry uses,” Neal says. This includes moving to more granular rates, considering inland and coastal flood models and adopting latest actuarial science techniques to make sure NFIP rates are “credible and accurate,” he adds.
Making changes to the NFIP takes a significant amount of time, according to the Academy report, due to congressional oversight and because the federal rulemaking process can take up to two years. A cursory look at legislative changes in recent years also demonstrates how Congress can impede the NFIP’s financial progress. The Biggert-Waters Flood Insurance Reform Act of 2012 contained provisions to improve the NFIP’s financial strength, including phasing out discounted insurance premiums and moving towards full-risk rates, according to the GAO report. The Homeowner Flood Insurance Affordability Act of 2014 (HFIAA), however, decelerated movement toward implementing risk-based rates and renewed some subsidies, the report notes.
However, Neal says that evolving the rating approach does not depend on new congressional legislation.
What is unknown, MacGinnitie says, is “How long will rate changes take and what will they look like?”
The Case for Greater Privatization
Sources agree that technological improvements and the reinsurance appetite for covering flood risk largely explain why insurers’ interest is growing.
Private insurers can offer multifold benefits. Besides deploying finer-tuned rating structures, carriers could expand available products and coverage. For businesses, packages could include other types of insurance such as business interruption coverage. For residential properties, insurers could either fold coverage into homeowners and renters policies or offer it as an endorsement. “(It’s) more efficient to offer one product that covers fire, wind and flood,” says Birny Birnbaum, executive director of the Center for Economic Justice.
MacGinnitie points out, “If you can get this in a homeowners policy or as an endorsement, two good things happen.” First, more people will be covered for flood. Second, even with claim causation issues such as water vs. wind, policyholders would be covered.
Since people tend to automatically renew homeowners and renters policies, offering flood insurance could encourage retention. Experience is showing that the private sector already does a better job at retention, Wienecke says. “Even just having the ability to settle the claims using the company’s methods, rather than those mandated by the federal program, may lead to less disputes and better results for private sector policyholders,” he adds.
Right now, Musulin says, the NFIP primarily offers credit for raising a structure above the flood elevation. Private insurers often offer a range of mitigation credits, often costing much less than elevating a structure. Rewarding policyholders for more modest mitigation strategies would also improve risk, MacGinnitie says.
Insurer Interest
Private insurers are already involved in the flood market. According to the NAIC report, there are about 20 insurance products in the nation.
Also, private insurers participate in the NFIP’s Write Your Own (WYO) program. They act as third-party administrators and, along with insurance agents, sell the majority of policies (though coverage can be purchased directly from the NFIP). Currently, there are about 900 active home or farmowners insurers, with fewer than 8 percent of eligible WYO insurers choosing to participate, according to PCI.
“The number of WYOs in the program has dropped from 107 to 70 since 2004 — more than 35 percent,” Wienecke says. Several household name insurers, including State Farm, Travelers and Nationwide, have left the program in recent years, he adds. Among the reasons are the “narrow” average profit margin of 4 percent, the program’s growing complexity to administer and reputational issues. After Hurricane Sandy, he explains, the media portrayed insurers as being unfair to consumers when in most cases, this was due to NFIP rules.
Whether private insurers should begin selling small-risk flood insurance boils down to profitability. Before Biggert-Waters, a 2011 PCI study revealed that private insurers could not compete with the NFIP’s prices, Wienecke says. At that time, private insurers’ premiums would have averaged twice the NFIP’s low-risk policies and as much as four times more for higher risk properties, he explains. However, since the NFIP is moving slowly closer to government-determined risk-based rates, the picture is different today.
Private sector interest in offering flood coverage is not universal. While surplus lines and reinsurers are very much in favor of providing coverage, MacGinnitie says that there is reluctance among some of the household name insurers. “These companies are worried that if they get (involved) in flood insurance in a big way they might be forced to subsidize rates,” he explains, “just like what sometimes occurs in auto and hurricane coverage.”
Insurers are concerned about the degree of profit potential, the regulatory environment, and possible fees and subsidies, Musulin says. While flood insurance is widely available in other countries, he explains, rates are often unregulated. Therefore, insurers’ willingness to offer coverage will also depend on the flexibility of state regulators when private insurers gain new insight and need to adjust rates. Further, since there has been relatively little private flood insurance offered in the United States, models are largely untested and the market is immature, he explains.
While the details of weather models have greatly improved, rising sea levels and growing population density in coastal areas put more properties at risk. The Actuaries Climate Index and the National Oceanic and Atmospheric Administration show sea levels are clearly on the rise, according to the Academy report. When considering risk, Musulin says, the effect that maps have on building codes is critical. “The real problem is that we are building things that will last 50 years (absent being washed away) based on today’s exposure to floods when in fact we need to be considering the conditions that will exist over the design life of the building,” he adds.
Higher risk potential, however, is not a deterrent for insurers. The challenge is to find the “equilibrium price where insurers are willing to sell it and people willing to buy it,” Clark says. While models have come a long way, she adds, it is important to keep in mind that there will always be much uncertainty with flood perils because the nature of low frequency events means data is limited.
Congress also would have to remove a laundry list of impediments to encourage private insurance participation. The PCI would like to see greater lender acceptance of private flood policies, Wienecke says. The organization is also advocating for policyholders to be able to move between the private market and the NFIP without penalty, with the NFIP viewing these policyholders as having continuous flood coverage.
Legislation should also require FEMA to share its data, Wienecke says. The optimal approach, sources say, is for the NFIP and private insurers to share data through a third-party aggregator such as Insurance Services Office, Inc. or the Independent Statistical Service, Inc. Further, Wienecke adds, the WYO noncompete clause should be eliminated, with the understanding that a WYO’s propriety data will not be used to unfairly compete.
In a statement before the Housing and Insurance Subcommittee Financial Services Committee in March, Roy E. Wright, FEMA’s deputy associate administrator of federal insurance and mitigation administration, said that, for its part, FEMA supports private insurers offering flood coverage.
However, Wright also expressed concerns about adverse selection. “This could lower NFIP premium revenue while increasing potential claims payouts. Such actions would leave the program and taxpayers with even more financial risk,” he said. Already, private insurers are selectively picking specific properties in SFHAs that could be overpriced, according to the NAIC report.
Conclusion
Congress started the NFIP because private insurance was unprofitable. Thanks to finer rating techniques, greater data, recent technological innovation and reinsurance appetite, there are good reasons to believe private insurers will have the profit incentive necessary to reenter the flood market after a nearly century-old hiatus. After all, flood insurance is profitable in other nations.
While the NFIP is in the process of implementing many of the new developments that are enticing private insurers to offer flood coverage, it faces several issues, including financial debt, low consumer participation and inadequate maps for granular ratemaking.
Many of NFIP’s struggles stem from contradictory congressional directives. Sorting out an optimal public-private partnership for covering flood loss likely will take more time than the September 30 deadline allows.
Annmarie Geddes Baribeau has been covering actuarial topics for more than 25 years. Her blog can be found at http://insurancecommunicators.com.
Promising Developments
Until recently, the private sector lacked sufficient information and tools to confidently underwrite flood coverage in the United States. Thanks to higher quality data and technological innovation, the understanding of flood risk is improving, boosting the confidence of reinsurers, excess insurers and an increasing amount of private insurers.
The same big data revolution affecting other insurance lines is also providing more insight into flood coverage, says Rade Musulin, vice president of the American Academy of Actuaries and chair of its Flood Insurance Work Group. For example, insurers can glean more property specific information that was too costly to collect in the past.
Several new tools have recently become available, such as mobile LIDAR, which uses light detection and ranging to reveal house level elevation. “Mobile applications are being developed that can provide a quote from an iPhone and bind in two minutes,” he says. Even Google Street View will support underwriting from a remote location instead of requiring a costly physical inspection, he adds.
Greater storage capacity and computing power are also greatly improving flood models, says Karen C. Clark, president of Karen Clark & Co. Not only are property and hazard characteristics data fuller and more accurate, but also weather and flood models have grown in sophistication, she says. “There’s higher resolution data and more ability to analyze them,” she adds. Her company offers transparent models that show the underlying data and methodologies along with high-resolution interactive maps, instead of black box models that offer only a numerical value.
Standards for evaluating flood models are also underway, according to “The National Flood Insurance Program: Challenges and Solutions,” written by the American Academy of Actuaries’ Flood Insurance Work Group and published in April 2017. The Florida Commission on Hurricane Loss Projection Methodology, which is developing such standards, is considered to have the most complete set of requirements for development and use of hurricane models in the United States, the report notes.