Affordability of National Flood Insurance Program Premiums: Report 2 (2016)

4

Analytical Next Steps and Further Findings for Affordability Policy Options

This is the second of two reports from this National Academy of Sciences committee. Report 1 discussed, among other topics, how to identify when National Flood Insurance Program (NFIP) premiums would result in a cost burden on policyholders, decisions that must be made by policy makers when designing an assistance program, and policy options for delivering assistance or for reducing premiums for all policyholders. In this report—Report 2—the task was to propose alternative approaches for evaluating affordability policy options. This second report describes analytical methods to evaluate affordability policy options and how the Federal Emergency Management Agency (FEMA) might expand its analytical capabilities; discusses data issues, which include data needs and availability of data; draws examples from a proof-of-concept pilot analysis conducted for a state with relatively rich data; and discusses how the data needs for a national affordability study might be addressed.

The first section in this chapter suggests some near-term analyses FEMA might complete as it is building its analytical capacity. The content of the remaining sections is the result of committee discussions and insights gained in the process of preparing Report 2. Those next sections include findings that add to or refine those in Report 1, Chapters 3 through 7.

NEAR-TERM ANALYSIS OF POLICY OPTIONS

FEMA is required to propose an affordability framework to Congress 18 months after submitting the affordability study.¹ In doing so, it must choose among numerous possible policy options. Ideally, FEMA would formulate alternatives for consideration, conduct an evaluation of the alternative options, and propose a preferred alternative. For FEMA to conduct an affordability analysis, both supporting data and an analytical platform are needed. Chapters 2 and 3 discussed the design of analytical procedures and the necessary supporting data to conduct an affordability analysis. In those chapters, specific observations and findings were presented. As a general matter, the findings in Chapters 2 and 3 suggest that FEMA will not be able to conduct a comprehensive analysis in the near term. There are analyses, however, that FEMA can undertake in the near term while building its analytical capacity in the longer term. First, based on Reports 1 and 2, some of the questions likely to be posed can be answered in a nonquantitative way. For example, one question (raised previously in Report 1) is, “Who might administer the [assistance] program?” This question might raise additional questions, such as, What legal authorities would the agency need to implement an alternative policy option or what other agencies would need to be partners in executing the option? The answers to such questions might affect the decision to pursue or not pursue an alternative policy option. FEMA may choose to narrow the range of options as it prepares answers to such questions.

Second, some alternatives might be initially removed from consideration (maybe to be reintroduced at a later date) based on a deductive and conceptual argument. For example, some of the options discussed in Report 1, such as disaster savings accounts, tax credits and deductions, and capping the NFIP responsibility to pay claims in high-loss years, might be put aside as viable near-term alternatives if the alternative would have little applicability to low-income property owners, if the alternative would require specialized legislation and execution by another agency, or if FEMA concluded that the alternative would have limited political acceptability.

Third, after this initial screening, some alternatives will remain candidates for inclusion in an affordability framework and, by Section 100236 of the Biggert-Waters Flood Insurance Reform Act of 2012 (BW 2012) direction, could be subject to quantitative analysis. Given FEMA’s current

__________________

¹ The Homeowner Flood Insurance Affordability Act (HFIAA) 2014, Section 9, states that “Not later than 18 months after the date on which the Administrator submits the affordability study referred to in subsection (a), the Administrator shall submit to the full Committee on Banking, Housing, and Urban Affairs and the full Committee on Appropriations of the Senate and the full Committee on Financial Services and the full Committee on Appropriations of the House of Representatives the draft affordability framework required under subsection (a).”

analytical capabilities and available data, however, the immediate prospects for a quantitative analysis of formulated alternatives are limited. Nonetheless, NFIP operations are well understood and the policy questions FEMA is expected to answer are well understood. In fact, the North Carolina proof-of-concept pilot analysis illustrates that computational modules can be built to determine premiums, predict future claims, and make estimates of NFIP net revenues for a limited range of alternative policy options. Based on the North Carolina analysis, FEMA could begin the conceptual development of a microsimulation framework that represents the operation of the NFIP and that can be ready to evaluate affordability policy options as data gaps are filled (see Chapter 3).²

Fourth, some analyses can be completed with data now available, or with limited investments in database development. To illustrate:

• Some descriptive questions might be answered. As examples, how many policies will lose pre-flood insurance rate map (FIRM) subsidies under BW 2012; using algorithms developed for the North Carolina analysis, how many current NFIP polices are paying grandfathered rates; using American Community Survey (ACS) data, where are the census block groups that have a high percentage of policies losing pre-FIRM subsidized (PFS) rates on primary residences and have low median income or a high poverty rate relative to surrounding areas? This latter kind of analysis, perhaps with different criteria, might be a way to identify geographic areas of possible high cost burden.
• Some questions can be answered using existing or readily obtainable national data. For example, prior to BW 2012 about 10 percent of all policyholders (about 500,000) were paying pre-FIRM subsidized rates, and were also primary residences (see Report 1). Under BW 2012, as well as HFIAA 2014, these 500,000 properties over time will be required to pay NFIP risk-based rates. Some policyholders are not going to be cost burdened (whatever the definition) because (a) premiums will not rise substantially, (b) the policyholder has the ability to pay the higher premium, or both. However, FEMA has not been able to answer questions about how high these rates might go, how many of the approximately 500,000 policyholders will find the NFIP rates unaffordable, or how much an assistance program targeted to these policyholders might cost. Some initial analyses

__________________

² The North Carolina analysis found that the model-predicted premium for a given property often differed—sometimes substantially—from the premium being paid, as reported in the NFIP policy database. This demonstrates one challenge in developing a simulation model that can replicate how the NFIP operates in terms of calculating premiums and, in turn, for analyzing policy changes regarding how premiums are actually quoted by Write-Your-Own agents or how an assistance program might improve the affordability of flood insurance premiums.

A key data gap for such analysis is the absence of first-floor elevation data for pre-FIRM properties. Such data are now being requested for properties that were previously paying pre-FIRM subsidized rates. As authorized by the Homeowner Flood Insurance Affordability Act in 2014 (HFIAA 2014), FEMA is now increasing rates by 18 percent per year and will keep doing so until an elevation certificate is provided. FEMA might use the data from those certificates to impute first-floor elevations on structures for which elevations are not yet known.³ Then, one near-term analysis would use current rating tables, available information on the properties and policies, and recently acquired data on first-floor elevations to estimate the range of premium increases. Knowing the size distribution of increases can be the basis for a policy discussion of whether the increases are “large” and for those that are deemed significant whether those policies are concentrated in particular areas. Also, knowing the total increase for all policies provides an upper-bound estimate of the new revenues that would flow to the NFIP (if all policies remained in force) and at the same time is an upper-bound cost estimate for a premium assistance program that would fully offset the increased cost for all PFS policyholders.

• Some questions can be answered using North Carolina data, building on the North Carolina analysis. For example, the costs of simply designed premium and mitigation grant assistance programs within North Carolina were estimated. More complex program designs could be formulated and evaluated for North Carolina.⁴ Depending on study resources and schedule, the socioeconomic data gaps that now exist in North Carolina might be filled with a sample survey of flood-prone property owners in that state. Having such data would make it possible to compare the different measures of cost burden. Specifically, the cost burden definition in the North Carolina analysis was premium as a percent of coverage. The committee, in the following sections on cost burden, finds that this measure is not a measure of ability to pay and other measures of cost burden will need to be developed. If an income-referenced measure of cost burden is desired, imputing income characteristics of an aggregate

__________________

³ For such an analysis, it would be important to capture the error introduced by imputation.

⁴ Some options for the affordability framework might still be put aside for later study because the technical foundation for simulation of effects is lacking. For example, there are no available estimates of the relationship between premium levels and mitigation actions other than structure elevation in the NFIP rating tables (see Report 1).

• census unit (e.g., a census block group) to individual policyholders using ACS data is not defensible. Therefore, a sample survey or a match to administrative income records within North Carolina may be required to obtain income data, although it may be unlikely that such a survey or records match could be undertaken in the near term. Alternatively, an analysis could use already available data on assessed property values for measuring cost burden, a possibility discussed in the next section. Such an analysis could examine, for example, how the number of property holders eligible for assistance and the amount of assistance that they receive varies when different percentage thresholds (for premium to property value) are used to identify who is cost burdened.⁵ The results of any North Carolina analysis can be realized quickly and can make an important contribution to the design of an affordability framework. However, these results will be specific to the state and might have limited generalizability for making inferences to the nation, an important limitation that would have to be recognized.

• Fifth, the absence of some basic data for current policyholders (i.e., first-floor elevations of structures and household incomes of policyholders) and the lack of data for at-risk properties that are not currently covered by policies mean that the ability to use a microsimulation model to quantitatively analyze policy options presently will be limited in scope and prone to uncertainty. As FEMA uses the North Carolina proof-of-concept analysis to guide its model building it can, at the same time, use that effort to identify data needs, and strategies and priorities for data collection at the national level.

FURTHER THOUGHTS AFTER REPORT 1

As Report 2 was being prepared, the committee’s attention was on the analytical challenges to doing an affordability analysis. As ways to address those challenges were investigated, additional insights into the topics

__________________

⁵ In addition, an analysis could assess the sensitivity of results to different behavioral assumptions about which policyholders would drop coverage when rates go up and would resume coverage when provided with assistance.

covered in each chapter of Report 1 were gained. As a result, the committee developed additional findings or refinements to those findings. These further findings that are relevant to the statement of task for Report 1 are reported in this section. To make the relationship to Report 1 clear, its chapter titles are used as the section headings below.

National Flood Insurance Pricing, Policies, and Premiums

Grandfathering

Grandfathered properties are those that were built in compliance with the flood hazard map in effect at the time of building construction, and the properties are allowed to maintain a lower flood insurance premium rate if a new map moves the property into a higher flood-risk zone or new base flood elevation. These policyholders would face increases if, as BW 2012 specified, grandfathering was no longer available and NFIP risk-based rates were to be paid. In addition, properties newly mapped in the special flood hazard areas (SFHAs) would be subject to the mandatory purchase requirement. For future NFIP-proposed changes pertaining to grandfathering, see footnote below.⁶

HFIAA 2014 reinstated grandfathering; however, Report 1 found that there is no reliable way to estimate the number of currently grandfathered policies and FEMA currently has no mechanism to identify grandfathered properties going forward. In addition, as was the case prior to BW 2012, HFIAA 2014 allowed grandfathered rates to transfer with the property by documenting that the structure was grandfathered by one of two ways. This was done by demonstrating the structure was built in compliance with the map at the time of construction and that continuous coverage has been maintained since the map change. Preferred risk policy (PRP) rates, however, cannot be grandfathered. When newly mapped into an SFHA, PRPs can apply for the newly mapped procedure. This gives formerly PRP properties 1 year of a PRP rate plus a reserved fund assessment and a federal policy fee, to comply with HFIAA 2014. For properties eligible for grandfathering, after this year they will be transitioned to a grandfathered rate, which would be a zone X (areas outside the 500-year floodplain) rate, even though they have now been mapped into an SFHA.⁷

Prior to BW 2012 and HFIAA 2014, all the NFIP rating tables included

__________________

⁶ FEMA will be making program changes in early 2016. In doing so, they will be capturing additional information on grandfathered policies. A summary of program changes are available at http://nfipiservice.com/Stakeholder/FEMA7/W-15046.html.

⁷ Information sourced from and additional information available at https://www.fema.gov/media-library-data/1428947341380-23a056704409206c86cc89ac72f9f070/FEMA-HFIAA_NewlyMappedFS_041015.pdf.

an explicit addition to premiums to account for forgone revenues from grandfathering. This cross-subsidy is being maintained by the NFIP and does support the BW 2012 expectation that NFIP revenues through time cover claims paid plus expenses (see Chapters 2 and 3 of Report 1). However, this cross-subsidy violates the actuarial principle that each property pays rates commensurate with its flood risk. Looking ahead, climate change, land development, and improved flood mapping mean that, in the future, some properties will be mapped into SFHAs when they are not currently, or will see higher base flood elevations. The owners of those properties will have the opportunity to pay grandfathered rates under HFIAA 2014 (in addition to those paying grandfathered rates prior to HFIAA 2014). The NFIP practice of increasing rates for all policyholders to account for revenue loss from grandfathering (i.e., cross-subsidizing rates) may result in an ever-increasing violation of the actuarial principle that rates paid should be in relation to risk.⁸ Specifically, the result will be that for policies that are grandfathered, premiums will be too low, and for those who bear the cross subsidy, premiums will be too high.⁹

NFIP Risk-Based Premiums

Chapters 2 and 3 in Report 1 discussed NFIP rates and rate setting and the changes called for by BW 2012 and HFIAA 2014. With specific reference to rates and BW 2012, Congress instructed the NFIP to move toward flood insurance premiums that better reflected the full risks of flooding at a given location, following actuarial pricing principles. As noted, grandfathering was to be eliminated, but HFIAA reinstated that practice with the result that rates for those properties may not be risk based. However, PFS rates, even with HFIAA 2014, will be phased out. Another gap in attaining risk-based rates is pricing of policies that are outside of the SFHA. Outside

__________________

⁸ Through time, some properties that had been paying grandfathered policies will either drop the policy or move to NFIP risk-based rates (for example, the policy lapsed and the property was sold). However, if risks increase through time that number may be small relative to the increase in newly grandfathered policies.

⁹ One approach to making premiums affordable is to increase the number of communities participating in the Community Rating System (CRS). As discussed in Report 1, revenue losses from offering premium discounts in CRS communities are currently made up by cross-subsidization. If promoting CRS enrollment is an affordability option then, as with grandfathering, NFIP rates will increasingly violate actuarial pricing principles.

of the SFHA, FEMA does not rate based on elevation and, as a result, elevation data are not available for those properties. The result is that zone X and PRP rates are not risk based; specifically, those properties at much higher risk in these zones may be paying rates that are lower than their true risk and those at lower risk could be paying rates that are higher. BW 2012 does not specifically direct FEMA to review and modify PRP and zone X rates to make them risk based. One result is that the rates charged may continue to fail in providing accurate risk information on these properties to their owners.

For the NFIP to move toward risk-based rates for all policies, it will be necessary to take at least two actions:

• Obtaining first-floor elevation data for all insured properties including those outside the SFHA. New technologies can make collection of elevation data easier on a wide scale. In the future, communities or FEMA would be in a position to use a technique such as vehicle-based light detection and ranging (LiDAR), which is a remote sensing technology that measures distance, to obtain first-floor elevations for multiple properties instead of requiring each property owner to obtain an elevation certificate. This possibility was discussed in a recent National Research Council (NRC) report on tying flood insurance to flood risk for low-lying structures in the floodplain (NRC, 2015b) and has been used throughout the state of North Carolina.¹⁰
• Have flood maps depict the spectrum of risk that properties face rather than focusing only on the boundary of the SFHA. The existing focus on the boundary is an artifact of the program where mandatory insurance purchase requirements are defined by the location of the boundary. Digital flood insurance rate maps, however, have emphasized not only defining the boundaries of the SFHA, but also mapping risk zones within the SFHA. Delineating risk across the floodplain would enable FEMA to provide better information for local zoning and minimize possible neglect or misunderstanding of risk by property owners. Risk-based pricing will require maps to include new zones outside the SFHA that reflect the different likelihood and magnitude of flood insurance claims as a structure is further removed from the SFHA boundary. If rates changed continuously with flood risk across the landscape, there would not be a dramatic change in rates from just crossing the SFHA line.

__________________

¹⁰ In areas where basements are common, LiDAR data will need to be supplemented with building information data.

The Insurance Purchase Decision

Promoting Takeup Through Assistance Programs

Prior to BW 2012, the NFIP had about 5.3 million policies in force. Some number of those (that number cannot be determined) was the result of the mandatory flood insurance purchase requirement based on the policyholder having a federally backed mortgage and the insured property being located in a SFHA. Report 1 reviewed the limited evidence available, which suggested that some property owners were not purchasing insurance even when the purchase of flood insurance was mandatory. As rates increase, compliance with the mandatory purchase requirement may be further reduced, especially for those households where the costs of the premium exceed their ability to pay, given their income and other living expenses. To address both compliance and affordability concerns, an assistance program could focus on aid to policyholders who are required to purchase flood insurance. Such a focus could also include aid for currently uninsured households who would be required to pay NFIP risk-based premiums (or pay grandfathered premiums) in the future as a result of map changes.

Among the affordability concerns expressed in BW 2012 and HFIAA 2014 there was a desire to keep premiums affordable, not only to discourage dropping of mandatory coverage, but also to motivate the voluntary purchase of flood insurance. The rationale for having high takeup rates was discussed in Report 1. If the goal is to expand the number of policies in force both within and outside the SFHA, providing assistance may encourage voluntary purchase when the insurance premium exceeds a household’s ability to pay, using some chosen criterion for defining cost burden.¹¹

__________________

¹¹ Willingness to pay may be the barrier to purchase. See Chapter 4 of Report 1 on takeup rates for further discussion. Further, from October 1, 2013, through September 30, 2014, flood insurance policy growth across the nation decreased by about 4 percentage points (FEMA, 2015). Policies in force across the nation were 5,568,642 and 5,350,887 in 2013 and 2014, respectively. Contributing factors could include premium rate increases from reform legislation adopted by Congress in 2012 and 2014.

Information Dissemination

Congress, in Section 9 of HFIAA 2014, directed that FEMA give consumers accurate information about the flood risk associated with their properties. If property owners are not asked to pay NFIP risk-based rates (for example, through grandfathering or the offering of non-elevation-rated rates), FEMA might still make estimates of those rates and inform the property owner about the premium discount they are receiving. To make such a calculation, however, FEMA needs access to first-floor elevation data and detailed flood risk maps.

With respect to disseminating information to homeowners, appropriate messaging would be helpful and could be included along with the annual premium letter. Any messaging could be piloted in focus groups and should draw from the literature (Kousky and Shabman, 2015). Where the specific information is available, the message could include components of an illustrative example shown in Box 4-1. Should specific information needed to estimate NFIP risk-based risks be lacking, a more general statement would be included in the annual premium letter.

Affordability Concepts and a Framework for Assistance Program Design Decisions

The Ability to Pay Flood Insurance Premiums

BW 2012, Section 100236, states that FEMA

The phrase “cannot afford” can be understood as exceeding an individual’s ability to pay an NFIP risk-based premium. This focus on ability to pay requires FEMA to define when such premiums impose a cost burden on an individual.

The ability of a property owner to bear a particular cost, such as a flood insurance premium, is often described in terms of some measure of household gross or net income.

For implementing an assistance program, individuals can be required to submit tax returns, W-2s, or other documents as proof of having a qualifying income. For policy simulations and other analytical purposes, however, analysts must rely on survey or administrative data, which often do not have measures of the specific income concept used by the assistance program. Therefore, other measures of income or proxies are often employed.

An alternative approach to defining ability to pay is the use of household wealth, rather than income. Wealth consists of the tangible and intangible assets owned by a household. It may be closely tied to income in some cases, but it is not the same as income. It can be argued that it takes income to produce wealth (e.g., investments) and that wealth can create income (e.g., return on investments). Monetary measures of wealth are generally expected to be positively correlated with income (high-wealth households tend to have high income; low-income households tend to have low wealth), but the correlation may be only moderately strong because there are sufficiently many low-income/high-wealth and high-income/low-wealth households.

Because it can take multiple forms, wealth is not always directly observable. Some components of wealth may vary quite slowly over time (e.g., value of housing stock) but other components can be more volatile (e.g., unrealized capital gains on investments, etc.).¹³ For premium affordability, wealth can be seen as a proxy for income, or instead as a useful measure of ability to pay. There is a particular connection between household wealth and flood insurance that argues for the latter view. In the event of a flood, an uncompensated loss is, in the first instance, a loss of wealth. To the extent that a structure is repaired or rebuilt out of current income,

__________________

¹² Emphasis is not in the original text.

¹³ With respect to difficulties in measuring wealth, see, for example, Boskin (1988) and Stiglitz (2015); on the treatment of capital gains, see Armour et al. (2013).

the loss of wealth is converted to a loss of disposable income. But if that is not the case, all or part of the damage remains as a loss of wealth. The role of flood insurance is to help the household avoid possible flood-related wealth losses, converting them to a much smaller and stable reduction in disposable income (the insurance premium).

Attempts to measure wealth encounter many of the same problems associated with income measurement. Although household wealth may consist of many different components, the principal component for those low-wealth households that might be the target of an affordability policy is the equity in the residence. A more readily available metric is the assessed valuation of real estate property, which includes not only the household’s equity in the property, but also the debt-financed portion. The use of this metric as an affordability-related premium cap relies on the relatively strong correlation of home equity in lower-wealth households with household wealth¹⁴ and the less well understood correlation of assessed valuation with home equity. Therefore, at least for the lower-wealth households usually associated with affordability issues, property value may be a useful indicator of ability to pay, especially given that data on property values are much more readily available than data on income or broader measures of wealth.

Defining Cost Burden for Assistance Program Design

Report 1 noted that there were many possible measures of cost burden and discussed three specific measures, two of which were related to an individual’s income. Specifically, Report 1 discussed an income approach and a housing cost as percent of income approach to identify those who would be cost burdened by their NFIP premiums. Report 1 also discussed a “capped premium” approach based on the premium as a percent of flood insurance coverage, as suggested by HFIAA 2014, as a measure of cost burden. This report revisits the insurance coverage-based capped premium approach and, in light of the analytical needs and data gaps discussed in Chapters 2 and 3 of this report, introduces a new type of capped premium based on property value.

__________________

¹⁴ 2011 data published by the Bureau of the Census (available at http://www.census.gov/people/wealth, “Table 1. Median Value of Assets for Households, by Type of Asset Owned and Selected Characteristics: 2011”) showed that for all groupings of households with net worth (wealth) below $250,000, the median value of home equity was in the range of 60 to 74 percent of median net worth. Above $250,000, this share fell rapidly to less than 30 percent.

Cost Burden Measured by Premium as a Percent of Coverage

The capped premium approach based on the amount of flood insurance coverage purchased requires policy makers to specify a threshold percentage—premium relative to flood insurance coverage—at which the premium is judged to become cost burdensome.¹⁵ The HFIAA 2014 legislation suggested that this threshold could be 1 percent. If this approach is chosen, policy makers will have to select that value or some other value. Using this criterion, a household whose risk-based premium exceeds the specified percentage of coverage would be deemed cost burdened and might be provided with assistance, potentially enough assistance to bring the premium down to the chosen threshold percentage.¹⁶

Consider a household which has (or is considering the purchase of) $100,000 in flood insurance coverage requiring an NFIP risk-based premium of $3,000. Using the criterion of 1.0 percent of total value of coverage for determining whether the household is cost burdened, the capped premium would be $1,000 and, if assistance is provided to eliminate the entire cost burden, the amount of assistance provided to this household would be between $3,000 and $1,000, or $2,000.

A property owner’s income or wealth characteristics do not enter into this calculation. In the previous example ($100,000 of coverage for a premium of $3,000 and a cost burden threshold of 1 percent), a household with income of $500,000 would be considered just as cost burdened as a household with income of $50,000. Furthermore, if an assistance program provided assistance—a voucher, say—to eliminate the entire cost burden ($2,000 in this example), both households would receive the same amount of assistance. More generally, if assistance is provided to eliminate the entire cost burden measured by this particular approach, the amount of assistance is independent of income and only sometimes related to the wealth of a household.¹⁷ In addition, this definition would make most assistance available to the highest-risk properties, since these are the policyholders for which the premium as a percentage of coverage would be the greatest. High-risk property owners may or may not face challenges with ability to pay.

__________________

¹⁵ The committee was explicitly asked to evaluate and say if the approach (premium as a percent of flood insurance coverage) could be justified as a way to measure cost burden (affordability) in a letter to the committee from then-Senator Landrieu (see Appendix E for a copy of the letter).

¹⁶ The amount of assistance is a separate decision that must be made by policy makers, as discussed in Report 1 (Chapter 6).

¹⁷ Insurance coverage may reflect the asset value of the insured structure, a component of household wealth. However, coverage is limited to $250,000, so this association does not exist for more valuable structures. Also, some policyholders subject to the mandatory coverage provision may choose to purchase insurance for the (smaller) mortgage principal, rather than the asset value.

Cost Burden Measured by Premium as a Percent of Property Value

This approach defines the premium cap as a specified percentage of the assessed valuation of the insured real property.¹⁹ Property value, which is a substantial component of total wealth for many households, especially low-wealth households, is used as a proxy for wealth. Wealth, in turn, is employed as a metric for ability to pay for flood insurance. Consider a property with an assessed valuation of $50,000 for the land and $100,000 for improvements. The household purchases a flood insurance policy with coverage of $100,000, at a cost of $3,000. Suppose that households are considered cost burdened if the insurance premium exceeds 0.67 percent of the property value; in this case, $1,000.²⁰ Assistance needed to lower the premium to the cap would be between $3,000 and $1,000, or $2,000.

Now suppose that a different (and perhaps higher-income) household has a home on an identical tract of land, but the assessed value of the improvements is $750,000. Total assessed valuation is $750,000 plus $50,000 = $800,000 and the NFIP risk-based premium for $250,000 of coverage is $5,000. (NFIP coverage is capped at $250,000 although the homeowner may be able to buy excess coverage on the private market.) The land value is still $50,000. The premium cap for $250,000 coverage is now 0.67 percent times $800,000, or $5,360. This property owner is not considered to be cost burdened and would receive no assistance.²¹

A principal advantage of this approach is that data on property values

__________________

¹⁸ HFIAA, Sec. 9, (b) (2).

¹⁹ This approach was not presented in Report 1.

²⁰ Because the cap is based on land value as well as structure value, 0.67 percent is used for comparability with the coverage-based cap. The actual percentage would be determined by FEMA, consistent with its directions from Congress to base assistance on financial ability as well as thorough empirical analysis.

²¹ The alternative measure of cost burden discussed here takes the logic of the capped-premiums approach and uses property value rather than coverage for gauging the size of a premium. Yet another alternative would be to use property value like income in the income approach discussed in Report 1. For example, premiums could be deemed burdensome to those with relatively low property values (e.g., below some specified percentage of the median).

are relatively readily available.²² This availability of assessed valuations for all structures in a floodplain, presently insured or not, creates another advantage—a microsimulation database that includes properties that are not currently insured but might become insured if premiums became more affordable. In the event that FEMA implements an assistance program based on property values, it will have the ability to anticipate assistance needs that may be associated with increased takeup rates. Along with its advantages, this approach has the problem of potentially granting unneeded assistance to low-wealth/high-income households, just as income-based cost burden measurements can lead to unneeded assistance for high-wealth/ low-income households. Another problem with this wealth-based approach is not providing assistance to low-income households who have very limited wealth aside from the equity in their homes. This could be addressed through additional eligibility criteria or an appeals process.

Alternative Measures of Premium Cost Burden Compared

This report and the prior report (Report 1) present four different approaches to define the cost burden associated with NFIP premiums. Table 4-1 summarizes the four measures of cost burden, along with selected pros and cons for each.

Loss of Property Value from Eliminating Pre-FIRM Subsidized Rates

Prior to passage of BW 2012, 20 percent of policyholders were paying less than NFIP risk-based premiums for properties located in the SFHA. BW 2012, as modified by HFIAA 2014, will increase those premiums until they reach the NFIP risk-based rate. The result is to increase the annual cost of ownership for the affected properties, which in turn should reduce the market price of the property. Any loss in market value will be incurred by the property owner at the time of the announced rate change and realized when the property is sold.

__________________

²² In carrying out the planning process for enhancing its modeling capabilities, FEMA would still need to assess the resources required to obtain, prepare for use, and regularly update data on assessed property values.

TABLE 4-1 Advantages and Disadvantages of Four Measures of Cost Burden

Isolating premium increase effects from other determinants of market price will be difficult, even if the best data for making such a calculation were available. For example, suppose there is a rise in mortgage interest rates coincident with the rise in flood insurance premiums. Both would tend to reduce property values, but isolating the specific effect of either will be analytically difficult. Hedonic price analysis could, in principle, be employed to isolate the effect of subsidy removal on property prices. Some studies have documented a reduction in the price of property located in SFHAs after controlling for other determinants of housing prices (Harrison et al, 2001; Shilling et al., 1989). The property price effects of removing pre-FIRM subsidies may be harder to detect than those described in the literature, since data are not available on the specific properties that paid less than NFIP risk-based rates or how much the NFIP risk-based premium would have been. Indeed, the uncertainty about the level of the new rates will certainly influence the extent of capitalization of higher premiums that occurs.

One way to illustrate the effect is to use a simple capitalization calculation as might be done by a property assessor. For example, a $1,000-per-year increase in premiums from losing a pre-FIRM subsidized rate could result in a $20,000 reduction in property value (5 percent discount rate); a $2,000 increase might be $40,000. The loss in property value could be a small or large part of total asset value and could be a small or large part of the property owner’s total wealth, recognizing that the property owner may have other assets.

These effects can be mitigated. One option would be to cap rates at a level less than NFIP risk-based rates for all properties that had pre-FIRM subsidized rates and also allow that cap on the premium to transfer with the property to all future owners, without regard to the future owners’ ability to pay. On the other hand, if a property owner was allowed to pay a reduced annual premium, but not transfer that reduced premium to the next owner upon sale of the property, the value of the property will be reduced by the capitalized value of the increase in future NFIP premiums, as though the premium reduction had never occurred. This option would be contrary to the goals of BW 2012 to have property owners pay NFIP risk-based rates and would result in lost revenues to the NFIP unless offsetting increases in revenues were provided by the federal treasury or by cross subsidy. Another option would be for FEMA to buy the affected property at the market price that would be realized if the below NFIP risk-based rate was continued. FEMA could also provide a grant, allowing the property owner to implement flood damage reduction measures that would reduce NFIP risk-based rates to levels that would restore some or all of the property’s market value. Both of these policy options would be challenging to implement and could require significant budget expenditures.

Cost Burden and Multifamily Properties

Multifamily rental apartments are a business and premiums for the building are paid by the property owner (landlord). Based on the data used for Report 1, about 70,000 apartment buildings nationwide were paying PFS rates. Following BW 2012 and HFIAA 2014, FEMA chose to define multifamily buildings as primary residences for purposes of applying the rate increases under HFIAA 2014. As rates increase, one possibility is that landlords may pass on the cost of increased NFIP premiums to renters, some of whom may be low income. Passing on the premium increase might make rents less affordable. The amount of rent increase that may be passed on, however, will depend on the number of comparable rental units in the same market area that would not have such increases. If comparable rental units are few, then landlords may be able to pass on the increased insurance costs.

Many of these flood insurance policies are for buildings concentrated in urban areas and were constructed in ways that make flood mitigation through elevation impractical. In lieu of elevating structures to mitigate flood loss, abandonment of commercial or rental use of the current first floor might be a mitigation action (to reduce premiums). In theory, this would impose a cost in the form of forgone rental income that may not be justified by premium savings. And even if justified by premium savings, vacating retail space (if that was the use) may diminish the mix and pattern of retail and residential space that defines “neighborhood character.”

Policy Alternatives for an Affordability Strategy

Linking Mitigation with Premium Assistance

Report 1 recognized that linking mitigation with premium assistance can lead to property owners having a cost-effective combination of mitiga-

tion and insurance coverage. However, mitigation measures, particularly elevating structures, can be quite costly and a policyholder may not have the necessary up-front funds.²³ As a result, mitigation implementation costs might be initially paid for by taking a loan from a newly created mitigation loan program or by taking a commercial loan. Further, mitigation costs may also be paid for by a one-time governmental grant.

Report 1 also described how assistance might be offered for making premium payments, for paying for some or all of the mitigation that can lead to reduced premiums, or for a combination of the two. An annual assistance payment could be used for making an annual loan payment and paying some share of the NFIP risk-based premium (the financial logic was described in Report 1). Alternatively, it may be that the property owner receives a mitigation grant and then receives no further premium assistance, or minimal assistance, in future years.

One way to link premium assistance and mitigation is through provision of an annual payment that the property owner can use for paying the premium or for paying off the mitigation loan. If the property owner is to make this decision then the owner needs clear information on available options to reduce risk, how such measures could be financed, and the impact of adopting one or more measures on premium reductions. This type of information is not currently available to homeowners and would require FEMA to develop a new outreach and communication effort.

Another approach would be for the NFIP to make the calculation to determine the most cost-effective mitigation; NFIP does the calculation based on an analysis that the marginal dollar spent on mitigation justifies a reduction in premiums (and also restoration of property values). If assistance was offered, the NFIP could require that some level of mitigation be implemented. The assistance offered could be a combination of a mitigation grant, plus access to a loan with a commitment to an amount of assistance that would be used to make the annual loan payments and premium payment assistance for any remaining (after mitigation) cost burden imposed by paying the NFIP risk-based insurance premium. The argument for a mandatory linking, is that people lack the information to make a financial calculation about mitigation, so the assistance program makes it for them as an eligibility requirement.

Finally, there is a broader array of potential mitigation measures beyond elevation that merit consideration for reduced premiums. This includes measures for structures that cannot be elevated, such as row homes, as well as less costly measures that can still help lower flood losses. FEMA may wish to focus future studies on how to appropriately price a broader range

__________________

²³ Currently, policyholders can receive a reduction in flood insurance premiums for elevating their structure and a small set of other measures.

of flood mitigation measures. When there are no data to support premium reductions for measures across the country, a framework could be developed for communities to submit proof of the efficacy of such measures in lowering claims to obtain a flood insurance premium discount. This could be useful if such measures have the possibility of helping to lower premiums to more affordable levels while at the same time reducing flood losses.

SUMMARY

HFIAA 2014 directs FEMA to propose an affordability framework for the NFIP. In doing so, it must evaluate affordability policy options. To do this, both supporting data and an analytical approach are needed. In the near term, FEMA can undertake analyses while building its analytical capacity over the long term.

As the committee focused on analytical challenges during preparation of Report 2, additional findings and refinements to findings that were presented in Report 1, specific to each of the chapters in Report 1, were identified.

__________________

²⁴ HFIAA, Section 9(b)(2).

A FINAL REFLECTION

Floodplains and coastal areas across the United States will continue to be inhabited. These areas will sustain damages from future riverine floods and coastal storms. The costs of these losses will be borne in three possible ways, or in some combination. Individual NFIP policyholders will bear location cost in the form of insurance premiums paid and damages falling within policy deductible amounts. The federal taxpayer might bear floodplain location costs if the federal treasury develops a premium assistance program, makes up for NFIP premium revenue shortfalls, or makes post-flood disaster assistance payments to individual households. Property owners and other floodplain or coastal zone inhabitants will bear costs for the losses that are uninsured or otherwise uncompensated.

An original intent of the NFIP was to replace disaster aid payment with flood insurance purchase to the maximum extent possible, shifting the cost of floodplain location onto those persons who occupy such places (Report 1, Charter 2). If this goal is to be pursued, then requests for premium assistance or mitigation grants and loans may increase due to future possible premium increases and from changes in flood risk stemming from changes in climate and changes in watershed runoff due to development. As an affordability framework is developed for the NFIP, FEMA and Congress will confront the central question, “Who will bear the costs of floodplain occupancy in the future?” With specific reference to the goal of “affordable premiums,” that question will be answered in recognition of the available governmental budget for premium or mitigation assistance and the adherence to the actuarial principle of minimizing cross-subsidies within the NFIP.