DRAFT(Word Count: 11,186)
December 2007
Mercatus Policy Comment
Helping Homeowners in the Gulf post-Katrina: The Road Home
By
Eileen Norcross and Anthony Skriba
Introduction
The 2005 Gulf Hurricanes destroyed or damaged over 300,000 homes in five states.[1] Entire communities were deserted, and many homes left uninhabited and unsalvageable. Damage in Louisianawas particularly acute making helping homeowners an early recovery policy goal. Towards that end,Louisiana used $6.9 billion of their $10.4 billion Community Development Block Grant (CDBG) hurricane recovery allocation for this purpose. However, Road Home was intended as more than simply a disaster compensation program: it fashioned itself as both planning and housing policy. The program simultaneously tried to compensate victims and recreate pre-existing neighborhoods by giving larger awards to those choosing to stay in Louisiana. The state targeted a broad population, extending eligibility to those suffering wind damage – an event typically covered by homeowners insurance. It also devoted some of its resources to developing affordable housing.[2]
Road Home’s performance to date has been widely criticized. By Katrina’s second anniversary, only 23 percent of applicants had received grants. Applicants express frustration at a complex application process, the program’s inequitable design, confusing policies, erroneous calculations, and the slow pace of payouts. Road Home creators blame overly rigid federal regulations, insufficient Congressional allocations, mismanagement by the private firm hired to administer the program, and miscommunication with the federal government.
This Policy Comment examines the reasons for Road Home’s poor performance. We begin with background on Katrina’s effects, and role that insurance and government play ex-ante and ex-post disaster. We then describe Road Home’s policy goals and design and contrast these to Mississippi’s Homeowners Assistance Program, which gave funds directly to a narrowly defined set of homeowners with few strictures on how the money is spent. Though not without problems[3], Mississippi’s program moved at a much faster clip, even taking into account the fewer number of overall applicants. In Section IV, we analyze the most significant policy features of Road Home and their likely effects. We conclude with recommendations policymakers should consider when designing disaster compensation programs, and also specific recommendations for the Road Home program.
Specifically, we conclude that Mississippi’s clearer eligibility criteria and the prioritization of applicants allowed for faster progress. Louisiana’s decision to expand Road Home to cover wind damage was a major error, de-concentrating funds and contributing to moral hazard. Distributing funds widely (both geographically and temporally) and without regard to the severity of damage, has contributed to a slower recovery for neighborhoods hit with the full force of flooding from the compromised levees. We find that clearly defining culpability and determining eligibility is vital in structuring ex-post disaster assistance. Secondly, we find that Road Home’s goal: encourage homeowners to return by penalizing people who leave the state, was an attempt to engineer a “tipping point.” This goal, while understandable, given the large amount of out-migration prompted by Katrina, is more likely achieved by treating all eligible applicants equally without regard to the choices they make about where they intend to live in the future. Equally important to creating a “tipping point” in a post-disaster setting, is ensuring that grants are awarded quickly.
One result of Road Home’s design has been to slow payouts. In a post-disaster situation, the effect of these delays can be catastrophic for long-term recovery. This is because evacuees must coordinate their plans and expectations with other members of their community in order to make the most optimal decision. Early returnees faced extremely high first-mover costs, an element Louisiana’s Road Home program completely failed to take into account, an omission with severe ramifications in terms of the rate of rebuilding shattered communities.
- Katrina and its Effects
Though tragic, flooding in New Orleans and the nearby neighborhoods was not inevitable – but rather the result of extensive infrastructure failure[4] (see Textbox 1: The New Orleans Levees).Levees and floodwalls built to protect residents around Lake Pontchartrain and LakeBorgne were breached and overtopped in more than 50 locations,[5] covering 80 percent of New Orleans in up to ten feet of water. Nearly 800 of the 1300 people who died lost their lives in the flood.[6] Pumping stations designed to remove flood waters failed throughout the city, submerging entire communities for days.
The catastrophic failure of man-made systems (levees, floodwalls, and pumping stations) designed to protect the city contributed to two-thirds of the death toll and damage.[7] The American Society of Civil Engineers (ASCE) estimated that less than half of the actual property losses in New Orleans would have occurred had the levees and pumping stations not failed.
No single entity is responsible for the engineering-related failures. The levees, constructed by the U.S. Army Corps of Engineers with federal funds over a period of years were maintained and operated by four levee district boards, which included state and local appointees. Pumping stations are operated by the city’s water and sewer boards. Indeed, many levee fractures “resulted from unclear lines of authority and insufficient coordination amongst the various agencies having jurisdiction over the levee system.”[8] In this sense, Hurricane Katrina’s impact on New Orleans was both an Act of God (rains, high winds) and an Act of man (failure of infrastructure designed to protect against such an event.)
Katrina’s effects along Mississippi’s GulfCoast were total and devastating. The hurricane shoved barges, boats, and debris into neighborhoods, killing 236 people. Entire communities were leveled. Hancock, Harrison, and Jackson counties were those most affected.
Katrina was one of three major hurricanes to hit the GulfCoast in the fall of 2005. On September 24, Hurricane Rita hit southwestern Louisiana and Texas, producing rainfall that breached Katrina-damaged levees in New Orleans, flooding Gentilly and the Lower 9th Ward a second time. Rita’s direct hit in southwestern Louisiana led to the destruction of several communities in Cameron, Calcasieu, and Beauregard Parishes. Cameron parish was almost completely obliterated.
Textbox 1: The New Orleans Levees
The federal government assumed responsibility for levee and floodwall construction with the Flood Act of 1936 and the Flood Control Act of 1965. In partnership with state and local governments, the U.S. Army Corp of Engineers (USACE) designed and built most of the levees in New Orleans over a period of decades.[9] The local levee boards own and operate the levees and are responsible for maintenance.
Hurricane Betsy in 1965 prompted enhancements to the New Orleans levees that were eventually abandoned due to a court ruling against the USACE. In the mid-1980s another levee improvement project, the high level plan was started but incomplete and some argue, still inadequately designed, when Hurricane Katrina hit.[10] In addition to floodwalls and levees, New Orleans had developed and operated a series of pumping stations throughout the city beginning in the early 20th century to remove flood waters, as well as reclaim marshland.
The development of levees and draining of marshlands led developers to build right up to the levees. Paradoxically, flood protection systems designed to mitigate against disaster also encouraged development in high flood risk areas.
According to the American Society of Civil Engineers,[11]nearly 169 of 284 miles of federal levees and floodwalls were damaged in Katrina. In some places, levees collapsed due to their design. Engineers didn’t account for the soft soil, or the existence of a water-filled gap that developed behind the concrete walls (I-walls). In other areas, levees were overtopped. They were not protected against soil erosion, “an engineering choice of catastrophic consequence”, which allowed soil to be scored away and water to pour into the city. New Orleans hurricane protection system was piecemeal in design and built with incorrect elevation data that failed to take New Orleans rate of sinking into account (up to 1 inch a year).
Poor engineering choices were magnified by incompetent government management, Congressional pork-barrel spending politics that plague U.S. Army Corp of Engineers funds[12], and the diversion of millions of tax dollars over the years from public infrastructure improvements, on the part of local levee boards, for “bloated contracts and political patronage.”[13]
End Textbox 1
- Who Pays?
Ideally, individuals insure against disaster. In the case of Katrina, many people were uninsured or underinsured against flooding. Popular sentiment concerning disaster reliefsuggests that people shouldn’t be settling in areas located below sea level,[14]and if they fail to insure they should bear the cost of the decision to live in a high-risk area. Yet the underlying incentives of where to locate and whether or not to insurefacing residents of the GulfCoast were distorted by federal intervention begun decades ago as well as state and local policies.To a large extent there is a strong element of government responsibility for at least some of the storm-related damage.
Total damage from Hurricane Katrina is estimated at over $100 billion, making it the costliest Atlantic hurricane in history. In New Orleans 65 percent of its 147,000 residential properties were flooded, over half sustained severe damage.[15]
Many property owners in the Gulf held homeowners policies. Homeowners insurance covers wind-related damage, but generally excludes water-related damage from flooding.[16] Due to the lack of private providers of flood insurance, the federal government offers flood insurance through the National Flood Insurance Program (NFIP)[17]which subsidizes flood insurance to homeowners located in special flood areas.[18]
The NFIP uses Flood Insurance Rate Maps (FIRMs) produced by the Federal Emergency Management Agency (FEMA) to pricepremiums, which are then sold through local insurers separate from homeowners insurance. The 100-year flood, or a one-percent annual chance that flooding occurs is the standard used to map and manage flood hazards.[19](See Textbox 2: The National Flood Insurance Program)
With hindsight, it may seem obvious that communities flooded by Katrina were located in floodplains and residents should have bought flood insurance. Yet only 40 percent of residents in Orleans parish, and 57.7 of St. Bernard parish had flood insurance.[20]
There are several explanations for the relatively low rate of residents carrying flood insurance in the Gulf.
First, NFIP does not distinguish between 100-year flood protection provided by an artificial levee versus natural topography. NFIP implicitly assumes the levees will hold. This assumption, one also held by many residents in New Orleans results in fewer people purchasing flood insurance.[21] However, the presence of levees does not eliminate risk entirely – the levees are always at risk of breaching, just as happened in 2005. This “residual risk” is not taken into account by the NFIP or state and local policymakers.[22] Approximately 35,000 of the flooded homes in New Orleans did not have flood insurance,[23] many because they were told that they did not need it. FEMA-flood maps incorrectly made the assumption that levees and floodwalls would hold.
The scope of flooding in other locations indicates that FEMA’s flood maps did not correctly capture the actual chance of a floodoccurring. This is true in Mississippi as well. Some residents in both states were living in areas that were not designated as a 100-year floodplain, and thus were uninsured or underinsured against a flood. They made a decision not to insure based on inaccurate advice derived from government-generated maps.
Some residents had maximum coverage: both homeowners policies through private insurers and flood insurance through NFIP and were located in the path of the levee-produced floods. These individuals took every possible measure available to insure against hurricane-related damage, but in effect were not fully insured against the levees breaking because NFIP did not price this residual risk (i.e. the risk associated with the levees breaking) making it an incomplete policy.
Lastly, there are residents who were living in designated flood plains, who were advised to carry flood insurance, and who chose not to insure against flood. There are several reasons why people living in hazard-prone areas may choose to not insure: they erroneously believe homeowners insurance is sufficient, they don’t fully appreciate the risk, or they believe disaster relief will compensate them after the fact.[24]
Textbox 2: The National Flood Insurance Program (NFIP):
For a variety of reasons, beginning in the middle of the 20th century private insurers decided to not offer flood insurance in the United States, claiming they could not provide profitable coverage at an affordable price.[25] These reasons included the inability to accurately calculate risk, and a lack of adequate financial tools (e.g., portfolio diversification) to help replenish capital.[26]Another reason for insurers’ reluctance to offer flood policies is the high correlation of losses that follow a disaster. Damage suffered during natural disasters is generally geographically concentrated and results in a high number of claims, making it difficult to pool risk. Insurers face a greater risk of financial insolvency in disasters if that year’s premiums are not sufficient to cover a sudden spike in claims.[27]
After Hurricane Betsy flooded New Orleans in 1965, Congress created the National Flood Insurance Program (NFIP), administered by FEMA. To participate, a community must agree to undertake flood mitigation measures based, at minimum, on federal flood construction standards. As part of the program, FEMA developed Flood Insurance Rate Maps (FIRMs) establishing the boundaries of floodplains. To encourage communities to participate, NFIP initially offered subsidized rates to residents living in floodplains prior to the issuance of the flood maps. The number of subsidized residents has declined to 26 percent over time.[28]
Several criticisms of NFIP have been raised, first is the accuracy of the maps used to define “Special Flood Hazard Areas.” These are places that have a 1 percent chance of being flooded each year, known as the 100-year flood. This is the standard used to determine if a resident needs flood insurance.The 100-year flood standard is based on a judgment made by experts in the 1960s about what represented, “a reasonable probability of [flood] occurrence and loss worth protecting against.”[29] In this sense the 100-year flood is a vulnerability standard and not a risk standard. At the time of Katrina, the maps had not incorporated the latest information on risk in the region, including the fact that sea levels were rising, New Orleans was sinking (subsidence) at a rate of up to one inch a year (4 feet a century), and increased hurricane activity in the 1990s.[30] This is the reason why floods destroyed areas in the Gulf “well beyond” the designated 100 year flood plains, where flood insurance was required.
In addition to the inaccuracies of the underlying flood maps, NFIP fails to take into account the residual risk associated with infrastructure failure. As mentioned earlier, NFIP implicitly assumes the levees provide sufficient protection to residents located nearby. This assumption led NFIP to advise residents near the levees to not purchase flood coverage.
Another criticism of NFIP is that the premiums it charges are kept low through subsidies and also by an annual limit on premium increases enacted by Congress in 1994. Subsidized rates convey inaccurate information to policyholders about the real level of risk they face. Secondly, the subsidies contributed to the insufficient cash reserves held by NFIP and its inability to pay claims, money it must instead borrow from the U.S. Treasury and pay back with interest. The borrowing limit for NFIP was unchanged since 1968 and set at $1 billion. After Katrina, the limit was increased to $20.8 billion.[31]
Lastly, repetitive loss properties account for almost 30 percent of NFIP’s claims. These are properties that are repaired multiple times with insurance dollars. Homeowners are being subsidized through NFIP to rebuild in high-risk areas, while not bearing the true expense because premiums remain artificially low.
With advances in risk analysis to more accurately price policies, and in financial markets to allow insurers to quickly restock capital reserves, from an actuarial standpoint, flood (and other catastrophes), are insurable events.[32] Policy recommendations to improve NFIP include pricing policies at an actuarially fair level, (based on the best available risk information), and ensuring the program has sufficient reserves. These recommendations imply, “a well-designed public catastrophe insurance program mimics as far as possible the procedures of an equivalent competitive private market.”[33] In other words, if risks can be priced to yield a profit, and financial markets can provide sufficient capital to fund losses, “there is no obvious reason why private insurance markets should not be able to provide catastrophe insurance. [34] Further, the existence of the government-run NFIP acts to crowd out market alternatives.