Click for next page ( 146


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 145
Hazard Compensation and Incentive Systems: An Economic Perspective HOWARD C. KUNREUTHER Societal problems involving risk and uncertainty pose special challenges for economics, since standard approaches such as benefit-cost analysis and market-based mechanisms fall short in dealing with them (Arrow, 1983~. This paper examines alternative incentive and compensation systems for dealing with potential hazards from both new and older technologies where one or more of the relevant stakeholders face a low-probability event with a potentially catastrophic outcome. The discussion revolves around two broad areas of current interest: 1. Adopting protective mechanisms: How can individuals be induced to protect themselves against the potential consequences of hazards, for exam- ple, by wearing seat belts when in automobiles or by purchasing insurance against floods or earthquakes? In these cases there are clear benefits to society in having people protect themselves, although few have voluntarily done so: fewer than 15 percent of drivers and passengers in automobiles wear seat belts voluntarily (Arnould and Grabowski, 1981), and few indi- viduals purchased flood insurance despite its being highly subsidized by the federalgovernment(Kunreutheret al., 19854. 2. Siting technologicalfacilities: How can an appropriate location for a new technological plant such as a liquefied natural gas (LNG) terminal or a hazardous waste facility be determined? These facilities frequently evoke objections from local people communities usually favor such a facility only if it is located elsewhere (Popper, 1983~. Can a meaningful compensa- tion mechanism be designed for sharing the gains of potential winners in such situations with people in the same situation who perceive themselves as losing? 145

OCR for page 145
146 HOWARD C. KUNRFUTH~ PROBLEM CHARACTERISTICS Low-probability, high-consequence events have special features that are important in designing alternative compensation and incentive mechanisms for dealing with them. This section discusses five such features: (1) the nature of risk, (2) the existence of externalities, (3) the need to reconcile conflicting preferences, (4) ex ante actions and ex post outcomes, and (5) conflicts between efficiency and equity in designing compensation and incentive programs. By explicitly considering these characteristics, we may be able to recognize the opportunities and limitations of proposing certain courses of action. The first consideration is the nature of the risk. The phenomena that cause catastrophic accidents are often "rare" events for which it is difficult to produce scientifically credible estimates of risk (Weinberg, in this volume). Catastrophic accidents emanating from an LNG plant or environmental pollution from a hazardous waste facility are not sufficiently well under- stood to permit estimates of their probability with any degree of accuracy. The health and environmental consequences of an accident may not emerge for years. This latency period between the occurrence of an event and its first noticeable effects makes it extremely difficult to assign causality to a partic- ular technological facility (Baram, 1982~. For these reasons there are fre- quently disagreements among the experts on both the chances and conse- quences of a particular accident, and there are limited statistical data to settle these differences. The ambiguities associated with specifying probabilities and outcomes therefore raise a series of analytical as well as practical chal- lenges. Even when extensive historical data are used to determine the nature of a risk, such as the frequency and severity of motor vehicle accidents, individ- ual perceptions of risk may- differ from the objective data. Experts and laypersons alike use heuristics, or rules of thumb, in estimating the probabil- ities of certain outcomes, and these rules may lead to large and systematic biases (Slovic et al., 1979~. A second feature of societal problems involving low-probability, high- consequence events is that they create externalities situations in which the actions of one individual affect the welfare of others either positively or negatively. If a community is willing to accept a hazardous waste facility "in its backyard," this creates beneficial effects for an entire region. When a person suffers a severe automobile injury because he or she did not wear a seat belt, there are negative externalities to the general taxpayer because health costs are partially subsidized by the government. These externalities make it difficult to use normal market mechanisms for dealing with risks. For example, unless there is some type of compensation

OCR for page 145
COMPENS'IOlI AND ECONOMIC lNCENTlVES 147 mechanism associated with the siting of a new technological facility, it is unlikely that any community will agree to host it. Similarly, if people know that as victims of a natural disaster they will have their property made whole by liberal government relief, then there is little incentive for them to protect themselves with insurance. A third feature of these societal problems is the many interested parties who have a stake in the final outcome and want to participate in the decision process. Each ofthese stakeholders has goals and objectives that effect his or her preferences. The nature of the risks and the types of externalities associ- ated with these problems make it difficult for standard market mechanisms to reconcile conflicting preferences. Consider, for example, the proposal by a consortium of gas distribution utilities in 1974 to site a liquefied natural gas facility in one ofthree locations in California. Over the next three years, a number of interested parties at the federal, state, and local levels debated the desirability of each of these sites. By 1977 it became clear that none of them would be approved under existing procedures. This impasse led the utility companies and business interests to lobby successfully for a new state siting act. It simplified the licensing process for LNG facilities and stipulated population-density limits that would be acceptable to environmental and public interest groups. To help resolve the safety question, several risk studies were commissioned inde- pendently by some of the interested parties. The various studies' estimates of the risk to population differed enough to cause considerable conflict about whether a proposed facility was "safe enough." Citizen groups sprang up at the proposed sites and used these studies to voice their concerns about having a local LNG terminal. Final approval of Point Conception, a site in an unpopulated area, came at the end of 1982, at which time the utilities decided to defer construction of the terminal until an unspecified date (Kunreutheret al., 19831. A fourth feature of societal problems involving low-probability, high- consequence events is the ex ante-ex post connection. By ex ante is meant actions that are taken in anticipation of a disaster and whose purpose is to mitigate the consequences of a disaster should it occur. For example, a community might demand some type of payment or compensation for agree- ing to host a new technological facility. This could be in the form of taxes paid by the developer or in-kind compensation such as a reduction in resi- dents' electricity rates. Or, ex ante protective actions might be taken by industry- for example, designing a facility to reduce the risk of an accident to an acceptable level. (Buckling one's seat belt is another example of an ex ante action: it reduces the potential consequences of an automobile acci- dent.) By ex post consequences is meant the impact of a disaster or accident on

OCR for page 145
148 HOWARD C. KUNREUTH~ the different stakeholders. Certain actions can be undertaken on an ex ante basis to deal with ex post consequences directly. The classic example is the purchase of insurance, which involves paying a premium today in return for claim payments later. Liability rules that specify who will pay damages after an accident are another example of ex ante contractual arrangements for dealing with ex post outcomes. Government relief and compensation for medical expenses illustrate policy mechanisms that may be instituted only after a disaster occurs. For example, following tropical storm Agnes in 1972, the federal government passed a special disaster relief bill providing for $5,000 grants to victims and 1 percent low-interest loans to cover the . . remalmng c Damage. A fifth feature is the tension between considerations of efficiency and equity in designing meaningful compensation and incentive programs. The economist has traditionally been concerned with efficiency questions, that is, with the allocation of resources to improve the well-being of society. Distributional or equity issues may conflict with this objective when there are many interested parties each focusing on his or her own welfare or status. Policy tools such as compensation may be useful in bringing these two goals closer together, but there are likely to be political and social conflicts that restrict their actual use. Kasperson (in this volume) has clearly outlined these conflicts with respect to the siting of hazardous facilities by developing a set of principles designed to win public trust. He suggests relying on compensation only after it has been determined that certain risks are unavoidable. Rather than focus- ing on unavoidable risks, however, the economist would examine the trade- off between reducing a given risk and providing sufficient benefits, perhaps in the form of in-kind compensation, to make the siting acceptable to nearby residents. These five features of societal problems involving risks and uncertainty challenge the analyst and policymaker to develop strategies for making people aware of the hazards to which they are subject while at the same time inducing them to make benefit-cost trade-offs. Problems involving the adoption of protective mechanisms or the siting of new facilities have the additional complication that individuals must process information or, in other words, come to understand the significance of probabilities and other facts relating to low-probability events. DESIGNING INCENTIVE SYSTEMS FOR PROTECTIVE ACTIVITIES Let us now examine the case where individuals are considering whether to adopt protective actions such as wearing a seat belt or purchasing a car with an air bag. This action involves some ex ante cost ~ C), which represents

OCR for page 145
COMPENS'ION AND ECONOMIC INCENTIVES Amok 149 , U (W-C-L') 6k /~t,q~ ,-`GG~: P) ~ ~cC\~ p - ':/,~ - - FIGURE 1 Decision tree for adoption of a protective activity. ~ U(W-C) U (W- L) - - U (W) either the additional monetary expense of this safety device or the effort and discomfort associated with using it. We will assume that the probability of a disaster is well specified, although potential victims (p) may perceive it differently than the experts (P) do. The potential loss (L) due to a serious injury or health effect can be reduced to L' if the individual takes protective action such as wearing a seat belt or purchasing a car with an air bag. Traditional economic analysis assumes that individuals are rational, that they behave as if they are maximizing their expected utility. In the following example, each potential victim is assumed to have an identical utility func- tion U and initial wealth W so that their choice about whether to adopt a protective activity is depicted by the decision tree in Figure 1 .l,2 For the special case where individuals are risk-neutral, the following simplified decision rule emerges:

OCR for page 145
150 HOWARD C. KUNREUTHER P(L-L') ~ C (Adopt) Otherwise (Do not adopt) This rule states that protective measures should be adopted if the expected benefit [P(L-L')] from those measures exceeds the ex ante cost (C). Rather than actually undertaking some type of benefit-cost comparison as implied by the expected-utility model, individuals appear to ignore the negative consequences of an event until after they personally experience an accident or know someone who has suffered a loss. This behavior can be characterized by a threshold model where the perceived probability (p) is below some critical level (p*) . It illustrates two psychological phenomena to be considered in designing appropriate incentive mechanisms. First, most people tend to assume that a specific chance event, such as an automobile accident, will not happen to them. A study by Svenson (1981) revealed that 80 percent of all drivers believe that they are better than the average driver. This type of behavior can be understood in terms of the theory of cognitive dissonance (Festinger, 1957). People prefer not to acknowledge potential dangers that they believe they cannot avoid, and hence they either underestimate the chances of such a danger's occurrence (Akerlof and Dickens, 1982) or assume that the probability of its happening is so low that it is not worth worrying about (Slovic et al., 1977). Second, when an accident does occur, there is a tendency to overestimate the chances of the event's occurring again in the future, a phenomenon characterized as availability (see Kahneman et al., 1984). People frequently estimate the chances of an event's occurring in the future by the ease with which they can imagine it. In addition, the immediacy of a personal experi- ence can focus attention on the potential for loss in a way that statistical data may not. Given the difficulties of calling people's attention to potential hazards, consideration of the several alternative strategies discussed below . . 1S in orc her. Broadening the Time Horizon If individuals treat the probability of an accident as so small that they do not concern themselves with the consequences, then broadening the relevant time horizon of hazard information may induce people to protect them- selves. For example, if residents fail to buy insurance in a 100-year flood- plain, knowing that the probability of at least one flood during the next 25 years is .22 might induce them to pay attention to the potential consequences and to avail themselves of protective mechanisms. People's reluctance to wear seat belts voluntarily might also be due to the extremely small probability of incurring a fatal accident on a single automo

OCR for page 145
COMPENSATION AND ECONOMIC INCE:ITIVES 151 bile trip (statistical data yield a figure of 1 fatal accident in 3.5 million person trips). However, demonstrating that this low probability compounds to a lifetime probability (based on 40,000 trips) of .01 might put the use of seat belts in a more favorable light. Slovic, Fischhoff, and Lichtenstein (1978) found in controlled laboratory experiments that most people did in fact respond more favorably to this lifetime probability than to a trip-by-trip statement of risk. However, it is not clear whether this way of presenting information would be effective in modifying people's behavior outside of the laboratory. Reducing Insurance Premiums Another way to induce individuals to recognize the potential benefits of protective activities is to focus on the immediate returns ex ante from taking action rather than on the reduction in losses should an accident occur. Insur- ance rates are a particularly appropriate indicator of what actions are worth- while economically. In this spirit, Nationwide Insurance Company reduced its annual insurance premiums by $20 for cars with automatic seat belts. Nordhaus (1984) estimated that the present-value discounted savings from the additional expenditure for en automatic seat belt would be $124.00. It is an open question, however, whether individuals really calculate long- run savings and compare them with short-run expenditures. Automatic seat belts normally cost an extra $100 in a car, so the $20 savings in premiums may look inadequate if consumers think in terms of short-term balance sheets as Thaler (1983) and Kahneman and Tversky (1984) suggest. If this is true, then the obvious alternative is for insurance companies to offer large discounts ($50 to $60) in premiums at the time a car is bought and to reduce the discounts in subsequent years. Penalties and Fines Rather than providing positive incentives for adopting protective mecha- nisms, penalties can be imposed for those who do not make use of protective measures. For example, if residents of hazard-prone areas do not buy insur- ance, they could be held ineligible for federal disaster assistance. On the other hand, those with at least partial coverage could be permitted to take advantage of government aid. With respect to traffic safety problems, Austria has adopted a policy whereby insurance payments for medical expenses are contingent on the use of a seat belt, although it may be difficult to determine whether an accident victim has actually buckled up. This type of conditional ex post compensa- tion will be successful only to the extent that people believe such a determi

OCR for page 145
153 HOWARD C. KUNREUTH~ nation can actually be made after an accident and hence decide to wear seat belts. Rather than linking the penalty to the occurrence of an accident, individ- uals could be fined for not wearing their seat belts. Under a program recently adopted in 15 states, a person may be fined up to $50 for failure to wear a seat belt. More than 30 countries have successfully increased the use of manual seat belts through this mechanism, with between 50 and 80 percent of passengers and drivers buckling up (Graham and Henrion, 1984~. The suc- cess in using penalties to induce protective behavior depends on the individ- ual's perception of the probability of being stopped and on the magnitude of the fine. For example, despite Puerto Rico's introduction of a fine in 1974, seat belt use has not changed appreciably because there has been limited enforcement. All of the incentives mentioned above for the adoption of protective behaviors are designed to force individuals to pay attention to the potential benefits of taking care. They may be beneficial to society if, as controlled experimental evidence and field survey data suggest, individuals are poor processors of information. Not only do people misperceive probabilities, but they use simplified models of choice that frequently lead them to ignore the ex post consequences of their "risky" behavior. COMPENSATION IN SITING TECHNOLOGICAL FACILITIES The siting of new technological facilities poses policy challenges that are different from those of promoting protective activities. The residents of each potential site are likely to raise questions about the safety of the proposed facility and to require sufficient ex ante benefits to compensate for the expected negative impact on their life-styles. For example, even if a waste disposal site is considered to be acceptably safe, it may create additional traffic and noise from trucks transporting materials. There are additional equity arguments for residents demanding some type of compensation for agreeing to local siting of a hazardous facility. If all the surrounding towns in the region benefit from the facility, then they should also be willing to absorb some of the costs. A current problem facing many urban areas is where to locate a trash disposal facility that will serve the needs of the region. By maintaining the status quo, the cost of trash disposal may be considerably higher than if such a facility were established. Hence, some type of benefit-sharing mechanism may be more efficient and equita- ble than the status quo if it can be used both to locate a site and to share gains between winners and losers.

OCR for page 145
COlLIPElYS'ION AND ECONOMIC INCENTIVES Stages in the Siting Process 153 The negotiation process for choosing the site of a hazardous facility con- sists of three distinct stages, reflecting the different ways in which the facility is likely to affect the community. Although each of them is treated separately below for analytic convenience, the three should be viewed as a whole when actually selecting a final site. Stage 1: Building the Facility Unless taxes and employment benefits are significant, some type of ex ante compensation arrangement may be necessary to convince a community that the benefits of a proposed facility are greater than the expected costs. The word compensation is frequently interpreted by the general public to mean bribe or payment, particularly when public health and safety are at stake. Society wants to preserve the belief that life is special; money may cheapen it, while laying bare the inequality of wealth (Calabresi and Bob- bitt, 1978~. For this reason, I recommend changing the term compensation to benefit-sharing, to indicate that interested stakeholders such as the devel- oper, industry, or other communities who benefit from the facility will provide either monetary payments or payments in kind to the host commu- nity. As described below, a number of successful examples of sharing gains between winners and potential losers suggest that this type of benefit sharing may be feasible for future siting decisions. Massachusetts Hazardous Waste Facility Siting Authority Act This leg- islation, passed in 1975, illustrates the use of monetary compensation for siting a regional resource recovery facility. The law provided that a royalty of one dollar per ton would be paid to any community that agreed to host such a facility. The town of Haverhill, Massachusetts, initially agreed, but in the face of opposition, the city council withdrew its offer. The town of North Andover was awarded the site in 1977 after it virtually unanimously approved the project at a town meeting. By early 1981, several communities had committed their wastes to the proposed plant (O'Hare et al., 1983) . Wes-Con, Inc. Wes-Con, Inc., converted two abandoned Titan missile silos in Idaho into small facilities for the disposal of hazardous materials. Although there was no public opposition, the developer offered in-kind compensation in the form of free disposal services, additional fire protec- tion, and medical training (O'Hare et al., 1983~.

OCR for page 145
154 HOWARD C. KUNREUTHF~ Gray Rocks Power Plant Environmental opposition halted construction of Gray Rocks Power Plant, a coal-fired plant in Wyoming, because of the potential damage to the surrounding environment. The suit was settled when the utility company agreed to set up a $7.5-million trust fund to preserve a stretch of the Platte River, the habitat of several species of migratory birds (Lave and Romer, 1983~. The coal plant was completed in 1981 and is fully operational today. LNG Facility in Wilhelmshaven The Federal Republic of Germany pro- vided an additional subsidy for the construction of a recreational facility because of the adverse affects that a liquefied natural gas plant in Wilhelm- shaven were expected to have on the tourist industry in the area (Kunreuther etal., 1983~. Stage 2: Living With the Site An important consideration in a community's decision to accept a site is the impact that the facility will have on future property values and economic development. Residents may be concerned that the presence of a waste disposal facility would discourage others from wanting to live in their com- munity, causing the market price of property to fall and leading to an erosion of the community's tax base. Willingness on the part of the facility's devel- oper to compensate residents forced to sell their houses below market value may allay such economic fears. This arrangement has not been widely implemented, however. One reason may be the difficulty of determining what a fair market price would be in the absence of a waste disposal facility. Regular inspections can be imposed to ensure that a plant is complying with regulations and specific standards. Before a community embraces a new technological disposal facility, residents generally want to be assured that the plant will be forced to meet these regulations or be shut down. The following examples illustrate this type of monitoring and control arrange- ment in practice. Antonell' Corporation The Antonelli Corporation in Providence, Rhode Island, constructed a facility to treat and store electroplating materials. The corporation allowed annual monitoring inspections by the citizens and city officials, and provided additional fire-fighting equipment to deal with any accident (Sanderson, 1984~. Wes-Con, Inc. In negotiation with residents, Wes-Con, Inc., the devel- oper of hazardous materials disposal facilities in Idaho, informally agreed to shut down the facilities permanently if there was ever a fire.

OCR for page 145
COMPENSATION AND ECONOMIC INCENTIVES Stage 3: Accidents 155 Communities are naturally concerned about the risk of an accident and its consequences to residents. There is need for some type of guaranteed ex post compensation by the developer, the insurance industry, or the government to cover, at least partially, the accident costs to victims and to act as an incen- tive for the design of safer facilities. It may be extraordinarily difficult to look toward traditional insurance arrangements to provide this form of com- pensation if the risk is not well specified and if there is a long latency period associated with health consequences (Katzman, 1985~. Today there is great uncertainty about how future court settlements will assign liability for accidents the uncertainties involve both the magnitude of awards and their timing (Huber, in this volume). In addition, chronic health problems have multiple causes and may be influenced by genetic and nutritional factors. Hence, it is difficult to deter- mine whether hazardous substances produce specific effects, and traditional toxic tort law may fail to provide adequate compensation for these damages (Trauberman, 19811. Federal-Private Insurance Some type of federal support for private insurance programs may be necessary to cover the costs of accidents while at the same time offering protection to industry from the consequences of environmental pollution. Specifically, a reinsurance guarantee by the federal government for losses above a certain magnitude may enable the private sector to cover the first layer of losses. The Price-Anderson Act (42 U.S.C. 2210Egl), which provides compensation for damages due to the development or use of nuclear power, authorizes the Nuclear Regulatory Commission to contract with private insurance companies to make payment for damages. Similarly, the National Flood Insurance Program offers some persons in flood-susceptible areas the opportunity to purchase federally subsidized flood insurance through a joint federal- private program. Self-lnsurance Funds If private or government insurance is unavailable, special types of self-insurance arrangements may be necessary to protect industry from large losses due to accidents. For example, the chemical industry may want to administer an insurance fund to pay for specific losses incurred by parties that store or dispose of waste. This concept may be attractive, but it has large-scale administrative problems. Some type of monitoring and control of facilities would be essential in order to base premiums on risk and to ensure that those participating in the process

OCR for page 145
156 HOWARD C. KUNREUTH~ were not behaving carelessly because they knew they were protected from losses. Integrating Stakeholders and Stages The different interested parties and the three stages of the siting process can be integrated through a stakeholder-strategy matnx. Table 1 illustrates the use of the matrix for the hazardous waste siting problem and depicts which interested parties are likely to be involved in a strategy associated with a particular stage in the siting process. For example, monetary or in-kind compensation in Stage 1 should involve the host community, other commu- nities, the developer, the waste generator, and the state siting authority. As the matrix shows, the insurance industry will normally be associated only with Stage 3 of the process. The conditions necessary for determining which sites are feasible can be illustrated by constructing a relatively simple decision tree for any given stage under the assumption that there are n potential candidate sites. Figure 2 depicts the branches of the tree for Stage 1 if there are m different events and event j has a probability Pij that outcome Cij will occur. An example of a particular event in Stage 1 is the construction of a new facility that meets all specifications; another event might be the discovery after the facility is completed that it is on an earthquake fault. We will assume that there are no events that will yield any net benefits from the site itself. For site i to be willing to host the facility, some type of compensation or benefit (Big must be provided if eventj occurs. If a fixed ex ante compensa TABLE 1 Stakeholder-Strategy Matrix for Siting a Hazardous Waste Disposal Facility Stage ~ Locating the Site Monetary or In-Kind Compensation Stage 2 Stage 3 Living With the Site Accident Compensation for Property Value Decreases Insurance Environmental and Ex Post Regulation Compensation Host community Other communities Developer Waste generator State siting authority Federal government agencies (e.g., EPA) Insurance industry . . .. ~ .

OCR for page 145
COMPENS'IOI!I AND ECONOMIC INCENTIVES Site 2 - 157 /P1 2 ~ C1 1 B1 , C1 2 1 P1 m - - \ :;;~ P m1 P \ nm ' ~ FIGURE 2 Decision tree for Stage 1 of siting process. c1 m B. C21 B2 - - ~C2m + B2 ~ -Cm1 + Bn ~Cmn + Bn lion package is given to community i no matter what event occurs, then Bid By for ally. We have adopted this assumption in Figure 2. For Stages 2 and 3, the amount of compensation will depend on specific events. For example, if the developer pays residents for decreases in current property values when they sell their houses, then there will be a probability distribution associated with anticipated changes in market prices of property in the area. = The decision tree in Figure 2 provides a framework for discussing the issues of risk perception and the elicitation of community preferences. Both of these issues are important in developing a practicable program for com- pensation and insurance.

OCR for page 145
158 Perception of Risk HOWARD C. KUNREUTH~ Controlled experiments on public perception of risk suggest that the costs Cij associated with eventj at site i will depend on attributes of the associated risks. Slovic et al. (1980) have analyzed data on individuals' attitudes toward risky activities ranging from skiing to living near a nuclear power plant. Two general characteristics appear to define the public's ranking of the risks associated with these activities: the dread of the risk and whether the risk is knowable. Technologies perceived as uncontrollable, catastrophic, involuntary, and highly risky to future generations score high on dread. Those that are not observable, that are unknown to those exposed, that are delayed, and that are new are viewed as unknown risks. If a proposed technology scores high on both risk characteristics, the required amount of compensation is likely to be extremely high. A further difficulty in siting new technologies is that the relevant stake- holders perceive risks differently. Industry and the developer may feel that the facility does not represent much of a risk because the technology is new to them and considered controllable, whereas the public may have a different view. Von Winterfeldt and Edwards (1984) have developed a taxonomy of technological controversy and investigated the nature of such disputes. They conclude that controversies about threats that may have catastrophic poten- tial, such as Love Canal or LNG siting, are frequently due to differing values among the stakeholders. A related challenge in siting new technologies is that there is likely to be considerable ambiguity associated with probabilities of losses. Ambiguity associated with losses has two principal effects. First, insurance firms have relatively little interest in marketing coverage for these technologies since they are unsure of the chances of incurring specific claims. For example, the insurance industry has opposed providing coverage for losses related to nuclear power, claiming that the risk is not insurable because of the ambigu- ity associated with such losses (Hogarth and Kunreuther, 19851. Second, if there is ambiguity surrounding the probability of specific events, the poten- tial host community may imagine the potential losses in detail rather than focusing on the abstract probability. An advantage of employing some type of benefit-sharing arrangement is that the communities selected as potential sites are forced to concern themselves with trade-offs between obtaining a certain payment now or having to incur possible losses in the future with some unknown probability. Eliciting Preferences for Communities An emerging body of literature in economics attempts to answer the following question: How can one elicit the willingness of an individual or

OCR for page 145
COMPENSATION AND ECONOMIC INCENTIVES 159 community to accept a facility that does potential harm to them but benefits others? If such a mechanism induces truth telling on the community's part, it is called incentive-compatible.3 Economists have searched in vain for incentive-compatible mechanisms that will maximize aggregate net benefits and balance the budget (when the amount paid by developers and other communities covers the amount demanded by the host community). Problems of equity and fairness compli- cate the matter further.4 The institutional arrangements associated with the siting process add a further dimension to designing appropriate compensa- tion mechanisms, as illustrated by the Massachusetts Siting Hazardous Facility Waste Act of 1980. Under this legislation, the developer negotiates a siting agreement with the host community, offering compensation if neces- sa~y to satisfy the residents' concerns. Compensation is also provided to communities likely to be affected by a disposal facility in adjacent districts. Arbitration is offered to break any deadlocks (O'Hare et al., 19831. Under this arrangement the developer has an incentive to offer the lowest possible compensation to the host community as well as to its neighbors, and the communities have an incentive to demand as large a payment as possible. If several communities are vying for the facility, some type of auction might ensure that the community with the lowest bid hosts the facility. Even here, there is no incentive for a community to tell the truth if it knows that its bid is likely to be considerably below those of the other possible sites. To date, no hazardous waste facility has been located under the Siting Act, although the town of Taunton, Massachusetts, was on the verge of accepting in 1984. My colleagues and I (Kunreuther et al., 198515 have been investigating an allocation mechanism whereby each community seeking to host a hazardous facility submits a bid indicating the compensation it would require to provide a site for the facility. The community that submits the lowest bid is the "winner." If there are n candidate sites, then all the other communities would have to pay 11(n-1) oftheirbid. For example, if there are five communities that could host a trash facility, each would submit a sealed bid, and those that were not the lowest bidder would each have to pay a quarter of their bid. This procedure is guaranteed to yield a budget surplus so that the regional siting authority would be able to use any funds not given to the host community to institute improvements benefiting the entire region. Furthermore, the proce- dure induces communities not to overbid, since their tax will be increased proportionately if they are not the low bidder. If there are industries associ- ated with each site, the payments could be made by them rather than using the tax revenues of the communities. One other theoretical advantage of the proposed mechanism is the inability of any community to form a coalition with other candidate sites. One objection to this allocation mechanism is that the poorest communi- ties, which already are well endowed with hazardous facilities, will have the

OCR for page 145
160 HOWARD C. KU~REUTHI~R lowest bids. On equity grounds this is disturbing. There are two ways to alleviate this problem. The communities that would be potential sites could be restricted to areas where there are relatively few hazardous facilities. An alternative approach is to couple this auction procedure with a lottery in which each candidate has an equal chance of being chosen as a site. Kasper- son (in this volume) points out that lotteries are frequently used when fair- ness of selection becomes an overriding requirement, as in the military draft. In the procedure proposed by Kunreuther et al. (1985), communities that are not selected by the lottery still have an opportunity to enter a lower bid than that specified by the one randomly chosen. In this way they can "demand" that they be chosen as the site. Otherwise they would pay 11(n-1) of their initial bid. The Massachusetts Siting Hazardous Facility Waste Act of 1980 reminds us of the importance of integrating compensation and insurance policies into a broader framework. More specifically there is a need at the outset to examine all of the potential sites from both an environmental and an eco- nomic perspective, eliminating any that do not meet specific health and safety criteria or that may be inefficient with respect to the transportation of wastes from industries in the region. Following this step, one must gain an understanding of the values of the interested parties by determining the costs and benefits they perceive at each site. At this point conflicts between the stakeholders will be uncovered and specific benefits and risk-sharing arrangements, such as compensation and insurance, can be explored. Stan- dards for design and operation of a facility can be stipulated through appro- priate monitoring and control procedures by a governmental agency. CONCLUSIONS We have examined the features associated with societal problems involv- ing risk and the cognitive limitations of individuals in collecting and process- ing information related to low-probability, high-consequence events. Incen- tive systems and compensation mechanisms may be useful policy tools for encouraging the adoption of efficient protective activities as well as for improving the siting process for new technologies. In particular, by intro- ducing compensation as an option, the public may begin to evaluate its "willingness to pay" to avoid a net risk in relation to its "willingness to accept" a given net risk. The design and implementation of incentive and compensation programs raise a number of questions regarding the appropriate form, timing, and magnitude of these payments. The following issues are important in this regard:

OCR for page 145
COMPE~S~ION AND ECONOMIC INCENTIVES 161 How do individuals perceive the nature of the risk and the benefit? Who is liable for and should therefore bear the cost of an accident? What is the default option if a proposed alternative is not adopted? What is the perceived role of the public and private sectors in dealing with accident losses? What other policy options could either be combined with or replace compensation (for example, regulations and standards)? These questions pose challenges to both the analyst and the policymaker in educating experts and nonexperts alike. By designing programs that recog- nize the importance of institutional constraints and individuals' decision processes, society may be better able to manage low-probability, high- consequence nsks. NOTES A decision tree is a descriptive model of a problem involving a sequence of decisions ( O. or decision nodes) and chance events (O. or chance nodes). The ultimate outcomes or payoffs for each possible combination of choice and chance are shown at the tips of the branches. When outcomes cannot all be expressed in monetary units, the analyst attempts to quantify the utility (welfare, well-being, or satisfaction) of each outcome. The expected value of an alternative is a long-run average calculated by multiplying the value of each possible outcome at a chance node by its probability and then summing these products. The optimal decision is the one that yields the highest expected value. 2. The same type of analysis could be undertaken by the operator of a technological facility, who can be viewed as a potential victim (from a financial point of view) should there be an accident. A decision about whether to increase the safety of a plant by instituting a specific risk-reduction measure will be determined by comparing the cost (C) with the reduction in loss (L-L') if an accident with probability P occurs. 3. For a detailed discussion of incentive-compatible mechanisms for this problem, see Raiffa, 1982, Ch. 22. 4. There is a large body of recent work designed to elicit willingness to pay and willingness to accept values for public and private goods (see Brookshire et al., 1985, for a summary). Recently, Keny Smith and coworkers (1985) conducted a survey to determine how individuals evaluate the expected reduction of risks due to hazardous materials. On the basis of a field survey in the Boston area, the authors tentatively concluded that one may be able to elicit households' benefits for reduction of risks due to hazardous materials. If this preliminary finding is substantiated, then one may be able to compare these estimates with those obtained by other procedures. 5. The fraction that communities would have to pay could be lower then 1l(n-1) if industry or the state were partially compensating the community. To prevent overbidding, how- ever, it is necessary to impose a tax on communities benefiting from the facility. REFERENCES Akerlof, G., and W. T. Dickens. 1982. The economic consequences of cognitive dissonance. Quarterly Journal of Economics 84:488-500.

OCR for page 145
162 HOlYAIRD C. KUNREUTHER Arnould, R., and H. Grabowski. 1981. Auto safety regulation: An analysis of market failure. Bell Journal of Economics 12:27-48. Arrow, K. 1983. Behavior Under Uncertainty and Its Implications for Policy. Institute for Mathematical Studies in the Social Sciences. Stanford University, Stanford, Calif. Baram, M. 1982. Alternatives to Regulation. Lexington, Mass.: D. C. Heath. Brookshire, D. S., D. L. Coursey, and W. D. Schulze. 1985. Experiments in the solicitation of private and public values: An overview. Paper prepared for the Public Choice Society Meetings, New Orleans, La. Calabresi, G. and P. Bobbitt. 1978. Tragic Choices. New York: Norton. Festinger, L. A. 1957. A Theory of Cognitive Dissonance. Stanford: Stanford University Press. Graham, J., and M. Henrion. 1984. A probabilistic analysis of the passive-restraint question. Risk Analysis 4(1):25-40. Hogarth, R., and H. Kunreuther. 1985. Ambiguity and insurance decisions. American Eco- nomic Review: Papers and Proceedings 95:386-390. Kahneman, D., and A. Tversky. 1984. Choices, values, and frames. American Psychologist 4:341-350. Kahneman, D., P. Slovic, and A. Tversky. 1984. Judgment Under Uncertainty, Heuristics, and Biases. Cambridge, England: Cambridge University Press. Katzman, M. T. 1985. Chemical Catastrophes: Regulating Environmental Risk Through Pollution Liability Insurance. Homewood, Ill.: Richard D. Irwin, Inc. Kunreuther, H., J. Linnerooth, J. Lathrop, H. Atz, S. Macgill, C. Mandl, M. Schwarz, and M. Thompson. 1983. Risk Analysis and Decision Processes: The Siting of Liquefied Energy Gas Facilities in Four Countries. London: Springer-Verlag. Kunreuther, H., P. Kleindorfer, P. Knez, and R. Yaksick. 1985. The Role of Compensation for Siting Hazardous Facilities: Theory and Experimental Design. Working Paper #85-04-03, Department of Decision Sciences, University of Pennsylvania, Philadelphia. Lave, L., and T. Romer. 1983. Specifying risk goals: Inherent problems with democratic institutions. Risk Analysis 3:217-227. Nordhaus, W. 1984. Supplementing notice of proposed rulemaking on federal motor vehicle safety standards. Occupant Crash Protection, Docket #74-14, Notice 35, June 13, Refer- ence 10. O'Hare, M., L. Bacow, and R. Sanderson. 1983. Facility Siting and Public Opposition. New York: Van Nostrand-Reinhold. Popper, F. 1983. LP/HC and LULU's: The political uses of risk analysis in land use planning. Risk Analysis 3:255-263. Raiffa, H. 1982. The Art and Science of Negotiation. Cambridge, Mass.: Belknap Press. Sanderson, D. 1984. Compensation and negotiation primer. Mimeo. Slovic, P., B. Fischhoff, S. Lichtenstein, B. Corrigan, and B. Combs. 1977. Preference for insuring against probable small loss: Implications for the theory and practice of insurance. Journal of Risk and Insurance 44:237-258. Slovic, P., B. Fischhoff, and S. Lichtenstein. 1978. Accident probabilities and seat belt usage: A psychological perspective. Accident Analysis and Prevention 10:281-285. Slovic, P., B. Fischhoff, and S. Lichtenstein. 1979. Rating the risk. Environment 21: 14-38. Slovic, P., B. Fischhoff, and S. Lichtenstein. 1980. Facts vs. Fears: Understanding Perceived Risk in Societal Risk Assessment: How Safe Is Safe Enough? R. Schwing and W. Albers, eds. New York: Plenum. Smith, K., W. H. Desvousges, A. M. Freeman, III. 1985. Valuing Changes in Hazardous Waste Risks: A Contingent Valuation Analysis. Draft Interim Report (2 vols.), prepared for U.S. Environmental Protection Agency. >

OCR for page 145
COMPENSATION AND ECONOMIC INCENTIVES 163 Svenson, O. 1981. Are we all less risky and more skillful than our fellow drivers are? Acta Psychologica47: 143-148. Thaler, R. 1983. Using -mental accounting in theory of consumer behavior. Cornell University Working Paper. Trauberman, J. 1981. Compensating victims of toxic substance pollution: An analysis of existing federal statutes. Harvard Environmental Law Review 5:1-19. von Winterfeldt, D., and W. Edwards. 1984. Patterns of conflict about risky technologies. Risk Analysis 4(1):55-68.

OCR for page 145