IF~cdl~1f©ln JJ«~~~1fW~ 15) ~~1t~ t CD) TI § (p\ r.m IF If S\Jl1\ (C 11 ~ CC~ ~

October 16, 1981

Earthquake Economics

Much of the United States is seismically Even well-built structures will be severely active, but the greatest risks damaged by such forces, although much prevail in the western states (see map). depends on the type and quality of con­ Virtually every major metropolitan area west struction. Engineering data suggest probable of the Rockies is in a high-risk zone, facing maximum losses of 7 to 15 percent of value the potential of catastrophic losses. Estimates for typical residences and other wood-frame by the United States Geological Survey construction, 20 to 35 percent for steel-frame (USGS) suggest that a great earthquake in San , and 25 to 85 percent for concrete Francisco today would cause 11,000 deaths, and masonry construction of various qual­ 44,000 hospitalizations and about $40 ities. Loss of life will, in turn, likely be greater billion in propertydamage. Asimilarevent in in the more severely damaged structures. Los Angeles would generate 23,000 deaths, 91,000 hospitalizations and almost $70 But significant are also relatively billion in damage. infrequent. Indeed, there has not been a major earthquake in any densely settled Against this background, it is sometimes region of the country since the great San surprising to learn that the marketplace does Francisco earthquake of 1906. Even in the very I ittle to accommodate earthquake risk. shaky Far West, the probability of a great Despite the fact that over 150 carriers offer quake in any given year is quite low-two to earthquake coverage in California, five percent in the Los Angeles area and one for example, less than 6 percent of the percent in San Francisco, according to USGS households carrying regu­ estimates. larly purchase it. Nor do mortgage lenders, whose assets are secured by Decision-making and risk subject to earthquake risk, routinely protect Earthquakes may thus be characterized as their portfolios with insurance. (Although a low-probability, high-loss events. The fire-coverage requirement is quite standard in conventional economic analysis of decison­ mortgage contracts, no such requirement is making under such conditions employs the widely applied in regard to earthquake risk.) "expected utility" theory that Milton Similarly, real-estate investors and builders Friedman and Leonard Savage developed in often balk at incorporating earthquake­ the 1940's. According to this theory, resistance measures in construction projects. individuals weight the psychic value of alternative outcomes (rather than the Earthquake risk outcome itself) by the probabilities of Decision-making under earthquake risk, of occurence of each outcome. Risk-averse course, is influenced by the tremendously individuals assign a smaller psychic value (or powerful and damaging nature of earth­ utility) to a dollar gained than to a dollar lost, quakes. The great Anchorage (Richter 8.6) wh i Ie the opposite is true for risk-takers. earthquake of 1964, for example, moveda 40,OOO-square-mile chunk of the earth's Expected-utility theory indicates that risk­ crust, thrusting it upward by as much as 30 averse individuals should desire to purchase feet in some places. Even in the compara­ actuarially fair insurance to protect them­ tively moderate (Richter 6.6) San Fernando selves from low-probability catastrophic Valley event of 1971, some structures experi­ losses that they could not easily bear enced ground-motion acceleration in excess themselves. (Actuarially fair insurance is of one "g"-meaning horizontal forces equal insurance which has a premium equal to the to the full force of gravity. probability of the event times the value of the 1F~cdl~1f@n ~~§~ffW~ JTh@\1ffi1k ca) IT § a\1ffi 1F rr @jillCC 11 ~CC (G) . ()pinions expressed in this newsletter do not necessarily reflert the views of the management of the Federal Reserve Bank of San Francisco, or of the Board of Covernors of tht" Federal Reserve Svstem.

loss.) Since most economists view the to protect themselves against the financial household as a risk-averse entity, the lack of effects of massive claims, but they don't have consumer enthusiasm for earthquake enough experience to guide them in this area. insurance (and catastrophe insurance in In addition, U.S. tax laws don't adequately general) thus comes as something of a discriminate between catastrophe reserves surprise. and normal profits, as, for example, the Swedish and Mexican tax codes do. Thus, in The problem may lie with the theory. Indeed, those years when no claims are fi led, insurers economists Danial Kahneman and Amos are unable to set aside reserves for future Tversky have found fault with the Friedman­ contingencies without tax liability. Savage approach, on the basis of evidence from behavioral experiments and markets for Such factors will, of course, affect the cost of other kinds of insurance. In particular, they earthquake coverage. The typical pol icy for a find that people underweight remote events, wood-frame residence in California costs $2 leading to greater risk-taking than would per year for each $1,000 of coverage, with a otherwise be expected. 5-percent and an 80-percent coinsurance provision. The attractiveness of Howard Kunreuther of the Wharton School, this type of policy to individual homeowners in a similar vein, argues that the expected­ depends upon their risk-averseness and the utility theory operates only in an unreal world expected frequency and size of losses. of perfect rationality and information. He Experience suggests that wood-frame res­ argues that people may behave according to idences will suffer earthquake losses, on "bounded rationality," reluctant to purchase average, of 7 percent of value. (That was the insurance or take other preventive steps average near the epicenter of the San Fer­ because of limited knowledge of the nature of nando earthquake, although 50-percent the catastrophe. For example, even the losses were not uncommon and a few total availability of heavily subsidized catastrophe losses were recorded.) Under these con­ insurance has failed in some major instances ditions, the currenttypical premium structure to encourage individuals to protect them­ would be actuarially "fair" for an annual­ selves from risk. Although the residents of event probability of 8 percent. This is higher Rapid City, Iowa, qualified for subsidized than current earthquake-probability esti­ National , only about 30 mates, and thus tends to make coverage less pol icies were in force there when the flood of attractive to all but the most risk-averse 1971 caused $160 million in damage to the households. community. Low utilization of earthquake insurance thus Earthquake insurance seems to be a problem of weak demand at Nonetheless, we should not ignore more current premia rather than lack of awareness conventional explanations for the relatively that coverage is available. Indeed, one in­ low level of private hazard-m itigation efforts. surer mounted a major marketing effort after For example, premiums for earthquake the San Fernando earthquake and sold less i nsu rance may not attract even rational and than 100 policies. On the other hand, in- well-informed individuals. Indeed, catas­ su rers probably cou Id not offer coverage if trophe-insurance coverage poses special people took a sudden interest in earthquake problems. Coverage tends to be concentrated insurance, because existing reserves could in risk-prone areas-three quarters of all not adequately support greatly expanded earthquake-insurance policies are written in coverage. California, for example-so that the insurer cannot reduce the cost of coverage by di­ Disaster aid versifying risk geographically. In theory, Other considerations may also act to dis­ insurers could accumulate sufficient reserves courage private mitigation efforts. For 2 Earthquake ,Risk Map of the United States

Darkest areas Indicate highest risk Source:Applied Technology Council

example, the Federal government provides Angeles County, for example, recently found liberal disaster relief, mostly through the that the cost of bringing 8,000 hazardous provisions of the Federal Disaster Act and the structures up to current standards would be emergency-lending powers of the Small roughly 70 percent of replacement cost. Business Administration (SBA). These Given the low probability of a damaging programs are activated by a Presidential earthquake in anyone location, designation of a region as a disaster area. owners clearly were better off risking the total loss of their . After the Anchorage earthquake of 1964, for example, the SBA loaned funds at 3 percent Even for new construction, the cost of adding both to repair structural damage and to retire earthquake resistance could be a significant outstanding mortgages. Similarly, in 1971, it fraction of the cost of construction -perhaps made special low-interest loans and outright greater than the expected value of the loss (forgiveness) grants to uninsured earthquake that would be incurred without such victims in the San Fernando area. Such protection. Reducing the 20-to-35 percent public-policy conventions, humane as they probable loss to a conventional steel-fr.ame may seem at the time, actually increase highrise, for example, might raise further catstrophe potential by reducing the construction costs by 10 percent or more. attention paid by private parties to risk Given a less than 50-percent probability of exposure. earthquake in a building's lifetime, investors may not consider the additional construction Building economics costs warranted, particularly since the The insurance market is not the only private possible dollar losses may occur well in the activity which pays insufficient attention to future. Much depends, of course, on the type earthquake risk-building design is another. of structure involved and the risk-averseness Since falling structures are the major source ofthe investors. (Construction insurers report, of economic disruption and loss of life in an for example, that Eastern investors are more earthquake, it is interesting to consider the concerned than Western investors about economics of earthquake-resistant design. earthquake damage.) But private incentives to add earthquake resistance generally are Even in the absence of building codes, rather weak. potential liability for injuries and loss of life incurred by buildings' occupants should Conceptually, the burden of coping with encourage architects and builders to employ catastrophes such as earthquakes need not be earthquake resistant design. However, the thrust entirely on the public sector. If courts have seldom found a builder or private-market participants accurately designer liable for casualties arising from the perceive risks and potential losses, and collapse of a structure affected by a perceive that they will bear all associated catastrophic natural force. (In contrast, the costs, they will devote appropriate resources ancient Babylonian held to avoiding or mitigating catastrophic events. the builder liable with his own life!) Unfortunately, public disaster-compensation Considering the uncertainties involved, such systems, limited builder liability, and other a legal position is understandable, but it cited factors may lead the private sector to serves to weaken private incentives to commit insufficient resources to the incorporate special earthquake-resistance mitigation of such hazards. If, in addition, features in new construction. individuals systematically underweight high-loss, low-risk events-as some research Moreover, designers often have no economic suggests-both the public and the rationale for adding earthquake resistance to government may be surprised by the structures to avoid purely structural losses, magnitude of their losses in the next major especially in the case of renovation. Los seismic event. Randall J. Pozdena 3 SS"1::> J.Sl:Il:I

U018U!4SE'M • 4E'ln • uo8aJO • E'pE'l\aN • 04E'PI !!E'MEH • e!UJoJ!IE'::::> euozpv. e>jsE'IV

CG>~<§y~\Ur~Jr~ \ill~CS

'J!le::> IO:lSpueJ:I ues ~ (G) ~ \ill~car ZSL 'ON llW~Bd OIVd @AJr@<§@CQI TI~Jr@1P@~ :J9V lSOd 's'n 11VW SSV1::> lSHI:I :)J \ill ~ \illlJ:)JJJ \2?cd1 ~cm \\{I ~ Jr\2?~~~CQI

BANKING DATA-TWELFTH FEDERAL RESERVE DISTRICT (Dollar amounts in millions) Change from Selected Assets and Liabilities Amount Change year ago large Commercial Banks Outstanding from 9/30/81 9/23/81 Dollar Percent Loans (gross, adjusted) and investments* 153,294 1,400 11,183 7.9 Loans (gross, adjusted) -total# 132,395 1,409 12,261 10.2 Commercial and industrial 40,231 945 4,864 13.8 Real estate 54,575 116 6,087 12.6 Loans to individuals 23,211 172 - 833 3.5 Securities loans 1,532 - 2 591 62.8 U.s. Treasury securities* 5,673 - 42 - 877 - 13.4 Other securities* 15,226 33 - 197 - 1.3 Demand deposits - total # 42,377 3,543 - 4,835 - 10.2 Demand deposits - adjusted 28,630 1,628 - 5,370 - 15.8 Savings deposits - total 29,526 215 - 576 - 1.9 Time deposits - total # 85,200 - 87 20,687 32.1 Individuals, part. & corp. 77,292 110 21,347 38.2 (Large negotiable CD's) 33,858 - 60 9,041 36.4 Weekly Averages Weekended Weekended Comparable of Daily Figures 9/30/81 9/23/81 year-ago period Member Bank Reserve Position Excess Reserves (+ )/Deficiency (-) n/a 59 96 Borrowings 148 53 188 Net free reserves (+ )/Net borrowed( - ) n/a 5 93 * Excludes trading account securities. # Includes items not shown separately. Editorial comments may be addressed to the editor (William Burke) or to the author .... Free copies of this and other Federal Reserve publications can be obtained by calling or writing the Public Infonnation Section, Federal Reserve Bank of San Francisco, P.O. Box 7702, San Francisco 94120. phone (415) 544-2184.