Tax Deductions for Catastrophic Risk Insurance Reserves: Explanation and Economic Analysis

Tax Deductions for Catastrophic Risk Insurance
Reserves: Explanation and Economic Analysis
Updated June 5, 2008
Rawle O. King
Analyst in Industry Economics
Government and Finance Division



Tax Deductions for Catastrophic Risk Insurance
Reserves: Explanation and Economic Analysis
Summary
In the wake of the three destructive hurricanes that struck the Gulf Coast during
the 2005 hurricane season (Katrina, Rita, and Wilma), the attention of policymakers
in Congress and elsewhere has turned to the subject of insurance for large
catastrophic risks, including natural disasters such as hurricanes and earthquakes.
The generally perceived increase in the incidence of major catastrophes and their
increasingly costly nature has prompted some analysts to question whether the
economy’s market for catastrophe insurance is sufficient to meet the burdens of
major catastrophes: does the market provide a sufficient amount of insurance against
major catastrophes, or is there a shortage? And, to the extent that catastrophe
insurance exists, are the insuring firms sufficiently capitalized so that widespread
insolvencies would not occur? Some have suggested that federal action is advisable
to make sure that insurance industry resources are adequate to ensure the availability
and affordability of disaster insurance and payment of claims when disasters occur.
One widely discussed proposal would change the tax treatment of catastrophic
risk insurance by permitting insurance companies to establish tax-deductible reserve
funds for catastrophes. A version of the proposal was developed by a working group
of the National Association of Insurance Commissioners, and a similar plan was
introduced as legislation in the 110th Congress (H.R. 164/S. 926) and in the 109th
Congress (H.R. 2668, Representative Foley). The proposals’ supporters argue that
they would enhance the ability of insurance firms to meet the requirements of major
disasters without risking insolvency and would increase the availability of
catastrophe insurance.
Economic analysis suggests that provision of a tax-deductible insurance reserve
for catastrophes would constitute a preferential tax benefit in the form of a deferral
or postponement of taxes: tax rules for most other types of activities permit the
deduction of losses only when losses actually occur, not when reserves for losses are
established. According to theory, tax benefits that favor one type of activity over
another ordinarily hamper economic efficiency and reduce economic welfare by
diverting resources from their most productive use. In some cases, however, markets
may fail and government intervention through taxes or other policies may improve
efficiency. The question economic analysis asks of the tax-deferred reserve
deduction is thus whether market failures exist in the case of catastrophe insurance.
While some economic analysts have identified a number of market failures that might
reduce the volume of catastrophe insurance, others are skeptical.
This report will be updated as legislative developments occur.



Contents
The Market for Catastrophe Insurance..................................2
Development of the Proposal for Tax Deductible Reserves.................3
Outline of the Proposal.............................................5
Arguments For and Against Tax-Deductible Catastrophe Reserves...........6
Economic Analysis................................................7
Appendix: Section-by-Section Analysis of the Policyholder Disaster Protection
Act of 2007 (H.R. 164/S. 926)..................................12
Section 1: Short Title..........................................12
Section 2: Findings...........................................12
Section 3: Creation of Policyholder Disaster Protection Funds; Contributions
to and Distributions from Funds; Other Rules..................12



Tax Deductions for Catastrophic Risk
Insurance Reserves: Explanation and
Economic Analysis
The devastation caused by Hurricanes Katrina, Rita and Wilma, which struck
along the Gulf of Mexico and Atlantic coast during the 2005 hurricane season,
highlights the fact that the United States continues to be subject to natural hazard
risks, primarily weather-related risks such as hurricanes and windstorms, but also
seismic risk (earthquakes, tsunami, volcanic eruptions) driven by basic geophysics,
and inland flood hazard risks caused by increasing floodplain development and
climate-change-driven extreme weather events.1 According to the Insurance Services
Office, Inc., (ISO), the property/casualty (p/c) insurance industry paid $62.2 billion
in catastrophe losses from 24 disasters and more than 4.4 million claims in 2005,
making 2005 the most costly year for catastrophe losses. By contrast, the 2006 and
2007 hurricane seasons were much less active than predicted. The availability and
affordability of property insurance in coastal counties along the Gulf and Atlantic
coasts have become a concern to property owners, state legislators and policymakers
as insurers pull back from high-risk areas. Insurers insist they must take this action
to reduce future potential hurricane losses and minimize the potential for insolvency.
Federal outlays for disaster victims have been increasing, and the frequency of
weather-related natural disasters is generally perceived to be rising. The insurance
industry had the financial resources to pay the claims from Hurricane Katrina, Rita,
and Wilma and earn significant profits, but the devastation in the Gulf Coast and
New Orleans from Katrina and high federal and private costs have generated interest
in Congress and elsewhere in proposals designed to change the way individuals and
communities evaluate and protect themselves against the risk of natural disasters (i.e.,
financing risk with insurance). Given the increasing concentration of insured
property values and sophisticated computer models that suggest an increased
frequency of hurricanes and high probable maximum losses (PML) from catastrophic
earthquakes, respectively, there has been some sense of urgency in Congress, state
legislatures, and the private sector to address the nation’s financial exposure to
catastrophic risks.
One widely discussed proposal would change the tax treatment of catastrophic
risk insurance by permitting firms to establish tax-deductible reserve funds for
catastrophes. A version of the proposal was developed by a working group of the
National Association of Insurance Commissioners (NAIC). Along with providing
a tax deduction for additions to catastrophe reserves, the plan would make the


1 David L. Brumbaugh, Specialist in Public Finance, Government and Finance Division, was
the co-author of this report before retiring from the Congressional Research Service.

reserves mandatory and would permit the reserves to be accounted for on the balance
sheet of the insurers. A similar plan has been introduced as legislation in the 110th
Congress (H.R. 164/S. 926 — Policyholder Disaster Protection Act of 2007) and
previously in the 108th and 109th Congresses (H.R. 4186 and H.R. 2668,
Representative Foley). Unlike the NAIC plan, the legislation’s reserves would not
be mandatory nor directly change current accounting practices. The proposals’
supporters argue that they would enhance the ability of insurance firms to meet the
requirements of major disasters without risking insolvency and would increase the
availability of catastrophe insurance.
The discussion that follows begins by providing some background on the market
for catastrophe insurance. It continues by describing the proposal for tax-deductible
reserve accounts as set forth in H.R. 164/S. 926 of the 110th Congress, and concludes
by providing an economic analysis of the plan. The appendix contains a section-by-
section explanation of H.R. 164/S. 926, the version of the plan introduced in the
current Congress.
The Market for Catastrophe Insurance
In broad terms, catastrophic risk is distinct from other risks simply because of
its large size — catastrophes affect a large number of persons and firms
simultaneously, imposing huge losses. Traditional examples include natural disasters
such as Hurricanes Andrew (1992) and Katrina (2005), the Northridge earthquake
(1994), and the Midwest floods (1993). Another example is the risk of loss from the
World Trade Center terrorist attacks on September 11, 2001, although this particular
type of risk has certain unique characteristics that place it beyond the scope of the
analysis here.
In terms of economic analysis, the large size of catastrophes produces a
particular result for insurance: it makes the elimination of uncertainty by an
individual insurance firm difficult. The market for insurance is based on insurance
firms being able to pool risks faced by a large number of policyholders whose risks
are not related to each other. When a large number of uncorrelated risks are pooled,
insurance firms can predict with relative certainty the average occurrence of a
particular insured event (say, an accident) occurring among insured persons, and can
provide insurance based on the reduction of that uncertainty. But the ability of
pooled risks to reduce uncertainty diminishes when the risks of policyholders are
correlated, and in the case of catastrophes, the risk of large numbers of insurance
buyers is, indeed, related. (For example, large groups of homeowners may
simultaneously incur damage from a hurricane or earthquake.) Thus, to sell
catastrophe insurance, insurance firms must pool an exceedingly large group of risks
in order to avoid correlated risks and develop marketable insurance.2


2 For a good intuitive explanation of the basic economics of catastrophic risk, see Mario
Miranda and Dmitry V. Vedenov, “Innovations in Agricultural and Natural Disaster
Insurance,” American Journal of Agricultural Economics, August 2001, p. 650.

In keeping with this feature of catastrophe insurance, an important characteristic
of the market is its reliance on what is known as “reinsurance” in insurance parlance.
That is, a particular insurance firm, in writing policies for catastrophic risk, may not
be able to find a sufficiently large pool of uncorrelated risks in its own marketing
area. For example, a company whose market is limited to a particular coastal area
may have difficulty developing a pool of customers whose risk of facing hurricane
damage is unrelated. The company may accordingly sell hurricane insurance
policies, but may itself purchase insurance (“reinsurance”) against the catastrophe,
in effect expanding the pool to unrelated risks beyond its own market.3
Another technique insurance firms have used in recent years to provide
catastrophe insurance is “securitization.” Here, insurance firms turn to general
capital markets to diversify their risks by marketing securities (for example,
“catastrophe bonds,” whose interest rate is contingent on the likelihood of a
catastrophe’s occurrence) that are attractive to investors because market risk is not
related to catastrophe risk, and the catastrophe securities may thus provide a means
of portfolio diversification for investors.4
Notwithstanding these methods of financing catastrophe insurance, concern has
been expressed by some analysts and policymakers about the ability of the industry
to satisfactorily meet the challenges of a major catastrophe. Concern has focused on
two factors that are sometimes termed the industry’s “capacity” to face catastrophic
risk: whether the insurance industry provides a sufficient level of catastrophe
coverage; and whether a catastrophe would result in a large number of insolvencies
among insurance firms. Proposals for tax-deductible reserve accounts have been
advanced as a means of increasing the industry’s capacity.
Development of the Proposal for Tax
Deductible Reserves
Current U.S. tax law and accounting systems do not permit deducting reserves
for future catastrophe losses. Reportedly, this situation has discouraged insurers from
accumulating assets specifically to pay for future catastrophe losses and has limited
the industry’s ability to increase its capacity to underwrite catastrophe risks.5 Instead,
payments for catastrophe losses are made from unrestricted policyholders’ surplus
after the losses are incurred.


3 For a detailed description of reinsurance and its role in providing catastrophic loss
insurance, see George E. Rejda, Principles of Risk Management and Insurance, 8th ed.
(Boston: Addison Wesley, 2003), pp. 535-541.
4 For a description of how securitization works, See ibid., p. 541 and Miranda and Vedenov,
“Innovations in Agricultural and Natural Disaster Insurance,” p. 651.
5 Scott E. Harrington and Greg Niehaus, “Government Insurance, Tax Policy, and the
Affordability and Availability of Catastrophe Insurance,” Journal of Insurance Regulation,
Summer 2001, p. 594.

Under the current U.S. tax law, property and casualty insurer income in excess
of annual expenses is considered profit and is subject to federal income tax. In
addition, U.S. accounting principles (both Generally Accepted Accounting Principles
and Statutory Accounting Principles) applicable to property and casualty insurers
limit the recording of loss reserves to losses that have already occurred and require
the recognition of catastrophe premiums in periods prior to the period in which
catastrophe losses are incurred. Thus, both U.S. tax law and accounting standards
discourage insurers from setting aside money for natural disaster events that cause
unusually costly damages.
Under §832(b)(5) IRC — “Deduction of Unpaid Property Loss Reserves for
Property and Casualty Insurance Companies” — insurers are currently allowed to
“deduct the discounted value of estimated losses they will be required to pay in the
future under insurance policies currently in force, including claims in dispute.”6 This
allows insurers to deduct future expenses from current income, and thereby defer tax
liability. This unpaid property loss reserve is comparable to what is known in the
insurance industry as “Incurred But Not Reported” (IBNR) losses/reserve. It is an
estimate of losses that have been incurred, but not reported. In some cases, insurers
know IBNR losses and make preliminary loss estimates. In other cases, however,
IBNR losses show up years after the damage first occurs. Insurers will adjust IBNR
reserves for such losses as new information becomes available and deduct from
current taxes the discounted value of the estimated losses. Reserves set aside for the
IBNR losses are different from the reserves that insurers want to deduct for future
catastrophic losses in that those losses have not yet been incurred nor reported.
In 1995, the National Association of Insurance Commissioners (NAIC)
established a Catastrophe Reserve Subgroup of the Catastrophe Insurance Working
Group to work with the insurance industry and actuarial resources groups for the
development and implementation of a tax-deductible pre-event catastrophe reserve
for property and casualty insurers. By 1997, the NAIC had developed a working
model for the reserve which, if it were to receive the support of Congress, would
enable insurers to set aside funds earmarked for large catastrophe losses and thus
increase their capability to manage catastrophic losses. The model was revised
several times in the late 1990s, based on the results of economic evaluation and
public comments on the design of the reserve and the statutory accounting changes
needed to implement such a reserve.
In 2000, the NAIC’s Catastrophe Insurance Working Group made its proposal
to the NAIC Property and Casualty Committee, but the NAIC did not adopt the
proposal beyond the working group level.7 The NAIC maintains that it will adopt the
proposal only if Congress changes the tax law to allow insurers to establish reserves


6 U.S. Congress, Senate Committee on the Budget, Tax Expenditures: Compendium of
Background Material on Individual Provisions, report prepared by the Congressionalth
Research Service, 108 Cong., 2d sess., S. Prt. 108-54, (Washington: GPO, 2004), pp. 201-

203.


7 For more information on the NAIC proposal see, NAIC Catastrophe Working Group,
“Summary of the NAIC Catastrophe Proposal,” NAIC Research Quarterly 6, No. 2 (summer

2000).



for future catastrophic events on a tax-deductible basis — with no offsetting tax
increases to the insurance industry. Insurers also drafted an example of the kind of
federal tax legislation that they felt could be used to implement this tax-deductible,
pre-event catastrophe reserve proposal.
In 2001, the National Conference of Insurance Legislators (NCOIL) adopted a
resolution to support a tax-deductible, pre-event natural disaster reserve fund. The
resolution states that NCOIL will work with the NAIC and federal legislators to
develop tax-deductible, pre-event catastrophe reserve legislation.
Under the insurance industry’s tax-deductible catastrophe reserve proposal, state
insurance regulators would approve catastrophe insurance rates that insurers insist
must reflect past experience and projected exposures. During discussions in 2000
before the NAIC’s Catastrophe Insurance Working Group, regulators expressed an
understanding about the concerns of insurers regarding the current tax and accounting
treatment of catastrophe reserves, and have assured insurers that they would approve
catastrophe-related rates that reflect both past experience and projected exposures.
Insurers had insisted that property insurance rates in regions with significant
historical or projected catastrophe exposures should reflect this exposure.
Regulators, however, needed assurances by the industry that the portion of approved
catastrophe-related premiums set aside in a dedicated reserve account for catastrophe
exposures will be earmarked for financing catastrophe losses that are not expected
to occur on an annual basis, and that these funds will be used only to meet
obligations to policyholders. Both regulators and insurers agree that requiring a
catastrophe reserve without tax deductibility would not provide additional assets to
finance insured catastrophe claims. Hence, many see congressional action to change
federal tax law as a precondition for implementing the NAIC catastrophe reserve
proposal.
Outline of the Proposal
The tax-deduction proposal set forth by the NAIC and the approach specified
by H.R. 164/S. 926 (as well as legislation in prior congresses) are the same in their
broad outline; there are, however, some differences. For example, the NAIC plan
would make establishment of reserves mandatory, while H.R. 164/S. 926 would not.
And, the NAIC plan would have a specific dollar target for industry reserves ($40
billion after a 20-year build up), while H.R. 164/S. 926 would not. Each plan,
however, would allow insurance firms a tax deduction for amounts contributed to an
account designed to provide a reserve for disaster payments; distributions from the
accounts would be subject to tax. As discussed below, the result is a tax benefit in
the form of a deferral (postponement) of tax.
The plan specified by H.R. 164/S. 926 would permit firms to establish
“policyholder disaster protection funds” which would consist of assets to be used
“solely for the payment of qualified losses.” Contributions to the funds would be tax-
deductible as long as they do not exceed the difference between fund-caps specified
in the legislation and a fund’s balance. (The latter amount would generally be the
cumulative amount of previous contributions, less distributions generally linked to



a firm’s catastrophe losses.) In short, contributions would be deductible in a
particular year only to the extent there is room between a firm’s cap and its fund
balance.
Conversely, distributions from the funds would be included in gross income and
subject to tax. While a complex set of rules would determine the amount of
distributions in a given year, the amounts would generally be linked to catastrophe
losses. Catastrophes, in turn, would generally consist of disasters such as hurricanes,
cyclones, tornadoes, earthquakes, winter catastrophes, fires, tsunami, floods, volcanic
eruptions, or hail.
Arguments For and Against Tax-Deductible
Catastrophe Reserves
Calls for government and industry action in the area of catastrophic risk
insurance have stemmed from concerns about the insurance industry’s “capacity” in
this area — that is, the industry’s ability to weather a large catastrophe without a
large number of insolvencies or the inability to honor insured losses. Supporters of
the tax-deductible reserve proposal have argued that the plan — by encouraging
insurers to establish catastrophe reserves — will both help protect against
insolvencies and will increase the availability of catastrophe insurance. It is argued
by the plan’s proponents that a number of benefits would follow from such outcomes,
including:
!enhanced ability of stricken areas to recover from disasters;
!reduced necessity in the long run for other forms of federal
assistance;
!reduced flow of U.S. capital to offshore foreign reinsurance firms.
Critics of the plan have pointed out that it would reduce federal tax revenue in
an era of large budget deficits. They also question whether the proposal would
accomplish what is intended — that is, whether it would increase the availability of
catastrophe insurance and the financial soundness of firms providing the insurance.
Some analysts, for example, have suggested the tax-favored reserves may simply
become substitutes for the purchase of reinsurance, while others have suggested the
tax-favored reserves may shield unrelated activities of insurance firms from tax.8
Rather than assess these arguments on a point-by-point basis, the next section
provides an economic analysis of the proposal.


8 The skepticism of these analysts is cited in U.S. Government Accountability Office,
Catastrophe Risk: U.S. and European Approaches to Insure Natural Catastrophe and
Terrorism Risks, GAO-05-199, pp. 30, 25.

Economic Analysis
As the “dismal science,” economics begins with the recognition that economic
resources are fixed. Thus, even if the proposed tax deduction for catastrophe reserves
would increase the supply of catastrophe insurance or reduce insolvencies among
insurance firms, the added resources that enable these outcomes are necessarily
diverted from other uses. Economic theory also holds — in general — that
undistorted price signals established in freely operating markets are the most
effective means of channeling resources into their most productive uses. Thus, price
distortions normally reduce the efficient allocation of resources and accordingly
reduce economic welfare. When applied to insurance, this part of economic theory
indicates that if insurance markets are operating smoothly, the diversion of resources
into catastrophe insurance by means of a tax deduction likely reduces economic
efficiency.
Markets do not, however, always function smoothly, and in the presence of what
economics terms a “market failure,” government intervention — either by a tax
subsidy or other means — may actually improve efficiency by correcting the market
failure. In assessing the proposed deduction for catastrophe reserves, economic
analysis thus asks the following questions:
!would the provision of a tax-deductible catastrophe insurance
reserve constitute preferential tax treatment for catastrophe
insurance?
!if the deduction would pose a tax benefit, are there failures in the
markets for catastrophe insurance that result in its under-provision
without government intervention?
!if there are market failures, would the proposal act to address them?
The analysis here concludes that a tax deduction for reserves is likely a tax
benefit. Beyond this, it does not attempt a final answer regarding market failures or
the efficacy of the proposal in correcting them. It does, however, suggest caution:
there is considerable uncertainty both about whether market failures actually occur
and whether the particular tax proposal in question would address them if they do.
Assuming that the proposal that would be implemented would follow the
general outlines common to each version of the proposal — that is, a tax deduction
for additions to a reserve and taxation of withdrawals — the tax deduction provided
by the proposal would constitute a tax benefit in the form of a deferral or9
postponement of taxes. In general, the federal income tax applies to income as it is


9 Depending on how the proposal is specifically drafted, the exact nature of the benefit could
vary. For example, for an insurance company with stable premiums, the phased-in proposal
in H.R. 2668 appears to provide a 20-year tax deferral for a fraction of premiums collected
on catastrophic risk insurance. With a constant growth in premiums, however, the benefit
could approximate that of an indefinite deferral (i.e., an exemption) of a fraction of
(continued...)

earned and permits business costs and losses to be deducted only when they are
actually incurred. Because the proposal would permit additions to reserves to be
deducted from taxable income before the losses the reserves finance actually occur,
the plan would have the effect of shifting some amount of taxable income towards
the future, permitting firms to delay an amount of tax liability equal to the tax savings
generated by the deduction. A tax deferral, in turn, confers a tax benefit because a
given amount of taxes matters less to a firm the longer its actual payment can be
postponed. In the interim, the funds that would otherwise have to be paid in taxes
can be invested and earn a return, increasing the insurer’s aftertax profits.
The proposal would thus confer a tax benefit; but are there one or more market
failures? Economic analysts of the insurance industry have identified several
theoretical reasons to believe the market failures may exist. First, as described in
other parts of this report, catastrophe insurance, by its nature, sometimes leads
individual insurance firms to seek reinsurance to adequately cover catastrophe risk;
the scope of catastrophes is so broad that, without reinsurance, a single insurance
firm may face claims from a large portion of its policyholders at the same time, thus
increasing its risk of insolvency if either reinsurance or securitization is not
undertaken. Several analysts have argued that the use of reinsurance may, in turn,
increase the likelihood of market failures that result from “asymmetrical information”
— that is, the insured entity knowing more about the size or probability of risk than
the insurer.10
These information related failures are termed “adverse selection” and “moral
hazard.” With adverse selection, reinsurance firms are not as familiar with the risks
they insure as are the individual insurance companies who buy reinsurance (the
primary insurers).11 With moral hazard, the insured entity increases its risky behavior
once it purchases insurance; in the case of catastrophe reinsurance, such risky
behavior consists of a primary insurer increasing the volume of risky policies they
themselves sell to individuals and firms.12 In the case of either failure, reinsurance
firms must increase their premiums to account for the uncertainty, and the volume
of insurance coverage is accordingly lower than it would be without adverse selection
or moral hazard. There are some analysts, however, who doubt that either adverse
selection or moral hazard play a prominent role in the market for catastrophic


9 (...continued)
premiums equal to the growth rate.
10 Other factors that might contribute to catastrophe insurance market failures may be
inefficiencies in the reinsurance market, not discussed here.
11 For a discussion, see David M. Cutler and Richard J. Zeckhauser, Reinsurance for
Catastrophes and Cataclysms, NBER Working Paper 5913 (Cambridge MA: National
Bureau of Economic Research, 1997), pp. 4-5.
12 For a description of how moral hazard may work in reinsurance markets, see Kenneth A.
Froot, The Limited Financing of Catastrophe Risk: An Overview, NBER Working Paper

6025 (Cambridge MA: National Bureau of Economic Research, 1997), p. 13.



insurance,13 so the presence of moral hazard or adverse selection cannot be accepted
as a given.
A second type of market failure that has been suggested as being present with
catastrophe insurance is what is sometimes termed a “principal-agent” problem —
that is, where a principal — in this case, a prospective policyholder — lacks
information about the entity acting as the principal’s agent (in this case, the insurance
firm) and the true value of the risk-sharing instruments sold by the agent. Here, the
potential problem is that the prospective purchaser of insurance lacks the knowledge
or expertise to adequately gauge the chances of an insurance firm becoming insolvent
in the face of a large catastrophe and being incapable of making good on its promises
to reimburse policyholders for their losses. In the case of this type of market failure,
the value of insurance for the buyer is reduced and demand for insurance declines.
As a consequence, less catastrophe insurance is sold in the market than is
economically efficient.14 As with moral hazard and adverse selection, however, some
analysts are skeptical of the existence of a substantial principal/agent problem.
Another possible explanation for a low volume of catastrophe insurance is
government action in the wake of disasters — specifically, the provision of
government disaster assistance at either the state or federal level. Such assistance is
considerable; for example during the 1980s and 1990s, federal disaster assistance
exceeded the average annual loss borne by reinsurers on catastrophe coverage.15 The
assistance may have the effect of reducing demand for catastrophe insurance and thus
reducing its volume: if large numbers of persons believe that disaster assistance will
be forthcoming if a catastrophe strikes, their incentive to buy insurance is reduced.
There is some empirical evidence that is at least consistent with the presence of
market failures. In 2001, a study by Froot found that the volume of catastrophe
insurance is lower and prices are higher than would likely exist in a perfectly
functioning market.16 Here also, however, there is cause for skepticism: the same


13 Jaffee and Russell, for example, are skeptical of the existence of asymmetrical
information: Dwight M. Jaffee and Thomas Russell, “Catastrophe Insurance, Capital
Markets, and Uninsurable Risks,” Journal of Risk and Insurance, vol. 64, June 1997, p. 206.
Froot believes that the design features of catastrophic insurance — high deductible amounts
— mitigate problems of adverse selection and moral hazard. See Froot, Limited Finance
of Catastrophe Risk, p. 14.
14 The principal-agent problem with catastrophe insurance is identified in David M. Cutler
and Richard J. Zeckhauser, Reinsurance for Catastrophes and Cataclysms, NBER Working
Paper 5913 (Cambridge MA: National Bureau of Economic Research, 1997), p. 3.
15 Kenneth A. Froot, The Limited Financing of Catastrophe Risk: An Overview, NBER
Working Paper 6025 (Cambridge MA: National Bureau of Economic Research, 1997), p.
15. See also this paper’s general discussion of the impact of government assistance on the
volume of insurance.
16 Kenneth A. Froot, The Market for Catastrophe Risk: A Clinical Examination (Cambridge
MA: National Bureau of Economic Research, 2001), p. 1. In both this and an earlier study,
however, Froot expresses doubt over adverse selection and moral hazard as explanatory
factors, preferring capital market factors as being a more likely cause. The capital market
(continued...)

study pointed out that there are several other possible explanations for the low
volume and high prices, including the presence of market power on the part of
insurance firms and certain features of capital markets. Froot’s analysis favored
these explanations over the presence of market failures.
Does the possibility of these market failures warrant provision of a tax benefit,
of the type contemplated by the proposed tax deduction for reserves? First, we point
out that the presence of market failures is not certain. While evidence may suggest
that the volume of catastrophe insurance is low and prices high, there may be
alternative explanations other than market failure. For example, costs inherent in
capital markets may drive up prices; also, insurance firms may possess market power,
which might enable them to increase prices and restrict supply. It is thus uncertain
whether there are market failures that can potentially be corrected by the tax benefit.
If there are market failures, would the tax deduction for reserves address the
problem? In principle, if market failures reduce production of a commodity, a tax
benefit could draw resources into the activity, reducing prices and increasing the
quantity sold. In the insurance industry, reserves are amounts collected as premiums
and held by insurance firms (rather than distributed to stockholders) in order to cover
expected losses. In the case of catastrophe insurance, a tax benefit linked to the
establishment of reserves could increase the after-tax return on invested reserves and
reduce the level of premiums insurers must charge to cover expected losses. Key to
the success of such a tax mechanism in increasing supply is whether insurance firms
could be expected to pass the added return on reserves to buyers or whether they
would instead increase distributions to stockholders or expand operations in other
operations not related to catastrophe insurance.
As described above, a firm’s deduction under the proposal would be dependent
on the volume of premiums collected on catastrophe-related insurance lines of
business and would be contingent on creation of catastrophe reserves. Yet even with
this language, whether a price reduction and an increase in the availability of
insurance would occur is not certain. For example, according to a February, 2005,
Government Accountability Office (GAO) report, some analysts have expressed
skepticism about whether tax-deductibility would induce firms to increase reserves
— and, even if the tax benefit induces firms to increase reserves, “some analysts
believe that the reserves would not materially enhance capacity because insurers
might substitute reserves for existing reinsurance coverage, the cost of which is tax
deductible.”17
Ultimately, whether the proposal would expand catastrophe insurance depends
on market conditions. Generally, the more competitive is the market for catastrophe
insurance, the more likely the deduction would expand insurance; on the other hand,
if firms have a fair degree of market power, the impact could be small. Or, factors


16 (...continued)
factors, however, may not be failures per se, but rather higher transactions costs.
17 U.S. Government Accountability Office, Catastrophe Risk: U.S. and European
Approaches to Insure Natural Catastrophe and Terrorism Risks, GAO report GAO-05-199
(Washington: February 2005), pp. 30, 25.

such as the availability of post-disaster government aid could render demand
“inelastic,” thereby limiting the impact of price reductions in expanding insurance,
even if firms were willing to reduce prices.
An analysis of the market for catastrophe insurance is not undertaken here.
Nonetheless, economic analysis indicates there are sufficient theoretical reasons to
recommend caution with respect to the provision of a tax benefit for insurance
reserves. To reiterate, theory indicates that tax benefits such as that recommended
by the reserve proposal improve efficiency only in the presence of insurance market
failures, and the presence of such failures is not certain.



Appendix: Section-by-Section Analysis of the
Policyholder Disaster Protection Act of 2007
(H.R. 164/S. 926)
H.R. 164/S. 926 consists of three major sections.
Section 1: Short Title
The “Policyholder Disaster Protection Act of 2007.”
Section 2: Findings
Congress finds that
(1) the costs of natural disasters are placing an increasing burden on insurers’
ability to pay homeowners’ claims arising from major natural disaster;
(2) present tax laws do not provide incentives for insurers to offer catastrophe
insurance because present tax law requires any surplus assets accumulated to cover
catastrophe losses are derived from after-tax retained earnings (i.e., are not
deductible);
(3) revising tax laws applicable to insurers to permit the accumulation of pre-tax
dollars in separate reserve funds devoted solely to the payment of catastrophe-related
claims will provide incentives for property and casualty insurers to offer catastrophe
insurance, better protect the nation’s homeowners, small businesses, and other
insurance consumers, and help ensure the financial solvency of the insurance system;
and
(4) implementing tax law changes will reduce the possibility of individual
insurer insolvencies in the wake of a major natural disaster, as well as the likelihood
of federal budgetary outlays for disaster relief.
Section 3: Creation of Policyholder Disaster
Protection Funds; Contributions to and Distributions
from Funds; Other Rules
Contributions to Policyholder Disaster Protection Funds. Amends §832
Internal Revenue Code (IRC) (relating to the taxable income of insurance companies
other than life insurance companies) to allow insurers to contribute a proportion of
net written premiums (NWP) for each business line into a tax-deferred policyholder
disaster protection fund (PDPF).
Distribution from Policyholder Disaster Protection Funds. Amends §832(b)
IRC to state the amount of distributions from the PDPF back to the insurer that would
be included in the insurer’s gross income.



Definitions and Other Rules Relating to Policyholder Disaster Protection
Funds (PDPF). Amends §832 IRC (relating to insurance company taxable income)
by adding a new subsection that provide definitions of the rules relating to PDPF.
!Policyholder Disaster Protection Fund (PDPF) — The PDPF is
defined as any custodial account, trust, or any other arrangement or
account that is established to hold money or other assets that are set
aside (on a tax-deferred basis) solely for the payment of qualified
losses. Deductible contributions to the PDPF would be voluntary,
but would be irrevocable once made, except to the extent of
“drawdowns” for actual catastrophic loss events, or drawdowns
otherwise required by state insurance regulators. Requires the assets
in the Fund to be invested by insurers in a manner consistent with
how insurers’ are required to invest under state insurance laws.
Distributions from the funds will be included as income subject to
federal tax liability in the year the funds are withdrawn from the
PDPF.
!Qualified Insurance Company — A qualified insurance company is
any insurer subject to tax under §831(a) IRC.
!Qualified Contribution — A limit is placed on the amount of
deductible contributions made to the PDPF. This limit is based on
the net written premiums in qualified lines of business; it is
generally limited to the excess of the fund’s cap for the year and the
fund’s balance at the end of the preceding year (see below).
!Excess Balance Drawdown Amounts — The excess balance
drawdown amount is the fund balance as of the close of the taxable
year over the fund cap for the following taxable year.
!Catastrophe Drawdown Amount — Specific events, such as
earthquakes, wind, hail, or volcanic eruption, could trigger a
drawdown from the PDPF, but the President of the United States, the
Property Claim Service, or the chief executive official of a state
would have to declare that a catastrophe had occurred. (See
“Qualified Losses.”) The bill establishes certain loss thresholds
which, if exceeded, would authorize the insurer to draw down all or
a portion of the reserves. The insurer may draw funds from the
PDPF to the extent that qualifying losses incurred in the current year
exceed the lesser of the insurer’s prior year Fund Cap or 30% of the
insurer’s prior year surplus, as reflected in the insurers’ annual
statement for the calendar year preceding the taxable year. The
threshold for drawing amounts out of the PDPF is lowered if the
insurer experiences major catastrophe losses in the prior years. In
order to ensure market stability after the major catastrophe, the
standard for drawdown is reduced for three calendar years to
qualifying losses for the current year exceeding the lesser of one
third of the insurer’s prior year reserve cap or 10% of the insurer’s
prior year surplus. The PDPF also may be decreased by the order of



a state insurance regulator to prevent an insurer insolvency (See
“State Required Drawdown Amount”).
!State Required Drawdown Amount — The State Required
Drawdown Amount is the amount that the insurance department for
the insurer’s jurisdiction of domicile requires to be distributed from
the Fund.
!Fund Balance — The fund balance is the sum of all qualified
contributions to the fund in all years, less any net investment loss
and distributions.
!Qualified Losses — Qualified losses are losses and loss adjustment
expenses incurred in the qualified lines of business net of
reinsurance, as reported in the NAIC annual statement for the
taxable year that are attributable to one or more of the qualifying
events, including windstorm (hurricane, cyclone, or tornado),
earthquake, winter catastrophe, fire, tsunami, flood, volcanic
eruption, and hail. The event must be designated as a catastrophe by
the Property Claim Service, declared by the President to be an
emergency or disaster, or declared to be an emergency or disaster by
the chief executive official in the state or jurisdiction in which the
event occurred.
!Fund Cap — The business cap for each separate line of business is
calculated by multiplying the net written premiums by the fund cap
multiplier applicable to such qualified line of business. Qualified
lines of business and their respective fund cap multipliers are: Fire
(0.25); Allied (1.25); Farmowners Multiple Peril (0.25);
Homeowners Multiple Peril (0.75); Commercial Multi-peril — non-
liability portion — (0.50); Earthquake (13.00); and Inland Marine
(0.25). The establishment of the PDPF is phased in gradually at a
rate of 5% per year over a period of 20 years. Thus, the maximum
deductible contribution for the year 2006 would be an amount equal
to 5% of the Fund Cap based on NWP for the year 2005. The
maximum deductible contribution to PDPF would be reached in
2026. Decreases in the amount of an insurer’s Fund Cap could
trigger mandatory “drawdowns” from the PDPF back to the insurer,
an the insurer’s prior tax deduction would be reversed into gross
income to the extent of the amount so distributed.
!Treatment of Investment Income and Gain or Loss — The bill
provides rules for realization of capital gain, and non-recognition of
loss, if an insurer contributes appreciated or depreciated assets other
than cash to the PDPF. If property is transferred to the Fund, it shall
be treated as a sale or exchange of such property for an amount equal
to its fair market value. Similarly, if property is transferred to the
insurer, it shall be treated as a sale or exchange or other disposition
of such property. Investment income derived from the assets held in



the Fund shall be considered items of income, gain, or loss of the
insurer.
!Net Income; Net Investment Loss — Investment income, gain or loss
on the fund balance is taxed in the year earned.
!Annual Statement — The annual statement for PDPF is the same as
set forth in § 846(f)(3) IRC.
!Exclusion of Premiums and Losses on Certain Puerto Rican Risks
— The premiums and losses on risks covered by a catastrophe
reserve established under this law or regulation of the
Commonwealth of Puerto Rico shall not be taken into account when
determining the fund cap or the amount of qualified losses.
!Regulations — The Secretary of the Treasury is authorized to
promulgate regulations to implement the bill.