The perverse effects of subsidized weather insurance.

• Author: Ben-Shahar, Omri
• Position: Introduction through III. The Perverse Effects of Subsidized Weather Insurance A. Distributive Effects 1. Insurance Cross-Subsidies, p. 571-597

Table of Contents Introduction I. Regulation of Weather Risk by Insurance II. Government-Provided Weather Insurance A. The National Flood Insurance Program B. Florida's Citizens Property Insurance Corporation III. The Perverse Effects of Subsidized Weather Insurance A. Distributive Effects 1. Insurance cross-subsidies: who are the beneficiaries? 2. Redistribution under Florida's Citizens Property Insurance Corporation a. Citizens' data and some initial observations b. Empirical analysis c. Discussion 3. Redistribution under the NFIP B. Investment Distortions 1. Regulation of location 2. Regulation of precautions IV. Responding to Concerns About Market (and Government) Failures in Private Weather Insurance Conclusion. Introduction

Catastrophes due to severe weather are perhaps the costliest accidents humanity faces. (1) While we are still a long way from having technologies that would abate the destructive force of storms, there is much to be done to reduce their impact. True, government cannot regulate the weather. But through smart policies and well-designed incentives, it can influence human exposure to the risk of bad weather. Regulation may not be able to control high winds or storm surges, but it can encourage people to build sturdier homes with stronger roofs far from flood plains. We call weather-related catastrophes "natural disasters," but the losses due to severe weather are the result of a combination of natural forces and often imprudent, shortsighted human decisions induced, as this Article will go on to show, by questionable government policies. (2)

Regulating weather risk is an increasingly urgent social issue. There is little doubt that the frequency and magnitude of weather-related disasters are rising over time. (3) Although the precise combination of causes--including emissions of greenhouse gases, natural climatic cycles, and the increased concentration of populations in coastal areas--may be debated, (4) the trend is undisputed. (5) Hurricane Katrina in 2005 and Hurricane Sandy in 2012 brought unprecedented property damage to the Gulf states and to the coastal northeastern states, respectively; (6) and in 2013, Typhoon Haiyan, which devastated the Philippines--eliminating entire villages and killing thousands--may have been the strongest tropical cyclone to hit land in recorded history. (7) Beyond anecdotes, the trend is clear: weather-disaster losses are rising. (8) And the most ominous harbinger of the trend is the expected rise in sea levels. Higher sea levels would lead to a massive increase in the storm surges that accompany severe hurricanes. (9) What in the past would have qualified as "100-year storms" (one percent chance of occurrence annually) would become "10-year storms" (ten percent chance of occurrence annually). (10)

Even as the magnitude and frequency of weather patterns seem to pose a higher risk than ever, a large and growing fraction of humanity's physical assets is located in harm's way. (11) While it is clear that the costs of hurricanes have increased dramatically and are likely to increase even more in the future, much of the upward trend in storm-loss data can be explained not by weather fluctuations but by increased concentration of property in dangerous areas, namely by human decisions to locate more densely in the storms' paths. (12) Thus, the combination of severe natural forces and increased human exposure poses one of the major public policy challenges of our era: how to regulate behavior so as to reduce this risk.

There are many ways that societies can reduce the risk of increasingly large and potentially devastating storms. Our thesis in this Article is simple: the most effective way to prepare for storms is through insurance. But not in the obvious manner, where insurance operates as a form of postdisaster relief. Rather, we argue that insurance contracts should operate as a form of private regulation of safety, incentivizing precautionary behavior prior to the occurrence of losses. We argue that a well-functioning private market for insurance can accomplish two fundamental objectives that are currently ill served. First, private insurance can induce optimal--which is to say, more prudent-investments during the development of communities. Second, private insurance can impose the cost of severe weather on homeowners who live in destruction-prone areas, eliminating an array of subsidies and discounts that they currently enjoy. We demonstrate that these existing subsidies most help those who need them least: namely, affluent homeowners living near the waterfront.

The idea that insurance can create efficient incentives for risk mitigation might surprise some of our readers. Like many, they have been schooled in the paradigm that insurance creates moral hazard. (13) Insurance may be good as a form of postdisaster relief and risk shifting, but the downside, we were taught, is that it dulls the insured party's incentive to mitigate losses. We think, however, that the application of the moral hazard theory to insurance has been overstated. We have written an article dedicated to debunking the myth that insurance necessarily creates moral hazard. (14) Insurers use a variety of contractual tools to prompt policyholders to reduce risks. While it is true that in some settings the presence of insurance coverage can reduce the motivation of policyholders to reduce risk, the opposite can also be true: through powerful incentives provided in the insurance contract, people who purchase insurance often do not fall prey to the moral hazard distortion, and may even take more efficient precautions relative to the uninsured. (15)

The main way in which insurance induces efficient precaution is through graduated or differentiated premiums. If the price of insurance coverage is adjusted according to the riskiness of each individual policyholder, it operates like a private Pigouvian tax, internalizing the costs that might otherwise have been externalized by the activity. (16) Moreover, in contexts in which insurers have better information about risks than their insureds (and they often do because they utilize far more sophisticated data collection and analysis methods), such premium differentials help not only to internalize cost but also to inform insureds of risks that they otherwise might not be aware of or fully appreciate. (17) And insurers have a strong incentive to use such premium differentials because of the power of competition: if one insurer fails to discount its coverage accurately, some other insurer may step in and steal its customers. (18)

The accuracy of risk-adjusted, cost-internalizing insurance premiums in the weather context, as in any other insurance market, affects both the care level and the activity level of policyholders. (19) Such premiums encourage efficient construction methods because sturdier homes are cheaper to insure. And they influence the original locational decisions by signaling to potential buyers the true cost of living in the path of storms. As a result, an entire community's preparedness for severe weather is importantly shaped and potentially improved by the aggregation of insurance contracts held by the community's members.

Unfortunately, in the United States, insurance is denied its potential role as an efficient regulator of prestorm conduct. It does not induce rational precautions by individuals, cost-justified community development by localities, or efficient infrastructure investment. American insurance fails to achieve these straightforward and enormously important roles for a reason that can be stated in one sentence: Insurance policies for extreme weather-related losses--especially for floods and for coastal wind damage associated with hurricanes--are not priced to reflect the real risk. (20) Rather, insurance policies for such risks are sold or subsidized by the government in a way that produces what are often called "cross-subsidies." This means that some parties pay more than their actual risk so that other parties (those who face the greatest and costliest risks) can pay less than their actual risk.

Thus, as a result of government intervention in property insurance markets, through either rate regulation or direct government provision of subsidized insurance, private markets no longer generate price signals regarding the cost of living in severe-weather regions. The cost of insurance for relatively high-risk property owners is suppressed, thus failing to alert private parties who purchase property insurance to the true risk of living dangerously. Such cross-subsidies allow private parties to (rationally) assume excessive risk and dump the cost of living in the path of storms on others. Indeed, much of the development of storm-stricken coastal areas is due to insurance subsidies and would likely not have happened at the same magnitude otherwise. (21)

Public debates over government-subsidized weather insurance often choose to ignore or downplay overdevelopment and excessive risk distortion because they regard government's intervention in weather insurance markets as an important upside that trumps any efficiency distortion. Government intervention in property insurance markets is justified and even necessary because--so...