Investments in coastal risk reduction measures generate significant benefits to society by reducing damage to buildings and infrastructure from coastal storms and potentially saving lives. Investments in coastal risk reduction also involve significant costs (e.g., construction and maintenance, meeting upgraded building codes, and loss of benefits from restrictions on development in vulnerable areas). Investments in coastal risk reduction may also have impacts on coastal ecosystems, which may generate additional costs (or benefits) through changes in the provision of ecosystem services (see Chapter 3). Investment in coastal risk reduction can take many forms including strategies to reduce the probability of a hazard (e.g., seawalls, surge barriers, dune construction, and marsh restoration), and strategies to reduce the consequences of a storm event (e.g., building codes, zoning requirements, and strategic retreat from vulnerable coastal areas). Different strategies for coastal risk reduction will differ in terms of their risk reduction benefits, costs, and ecosystem impacts. A key question facing society is determining when investments in coastal risk reduction are justified, and if justified, what form they should take. These decisions can be made at national, regional, state, or local levels and involve input from a broad array of stakeholders (those that have an interest in or are affected by decisions regarding coastal risk).
The committee was tasked to address the following questions: How might risk-related principles contribute to the development of design standards for coastal risk reduction projects? What general principles might be used to guide future investments in U.S. coastal risk reduction? (See Chapter 1.) This chapter describes and compares two approaches to de-
termining what investments in coastal risk reduction are worthwhile: (1) a risk-standard approach (sometimes called a “level of protection” approach) and (2) a benefit-cost approach. The risk-standard approach recommends investments in coastal risk reduction measures to achieve an acceptable level of risk reduction, such as reducing the threat of loss of life (e.g., 1 in 1,000 chance annually of more than 50 deaths in a single event) or the probability of severe flooding (e.g., 1 in 200 chance annually of overtopping a levee system). Thus, the risk-standard approach considers a specific consequence and designs cost-effective strategies to alter the probability of that consequence occurring. The benefit-cost approach recommends investment in coastal risk reduction when the benefits of the investment exceed the costs, considering both probability and consequences along a continuum of possible events. Thus, the level of risk reduction provided by projects under a benefit-cost approach is not predetermined but would vary based on the costs and benefits provided.
Differences between the risk-standard and benefit-cost approaches are illustrated in Figure 4-1, which presents a hypothetical plot of risk versus net benefits (benefits minus costs) from risk reduction investments. The benefit-cost approach would advocate investment in risk reduction to the point that maximizes net benefits (Point B). A risk-standard approach requires that investments be made so that risks are reduced to (or below) the acceptable risk (Point C). As drawn, the level of risk reduction that maximizes net benefits, Point B, does not satisfy the acceptable-risk standard. Acceptable-risk standards may thus be considered as a way to constrain benefit-cost outcomes. However, it is also possible that the level of risk reduction that maximizes net benefits is actually well beyond the level needed to meet the acceptable-risk standard (if plotted, Point B would lie to the left of Point C). In urban areas, providing risk reduction measures beyond the acceptable-risk standard could be a wise investment given the large value of property being protected and the potential savings in human lives. For example, even if an acceptable-risk standard is determined to be a 1 percent annual-chance (100-year) event, net benefits in coastal cities might be maximized by providing risk reduction measures designed for a 0.2 or 0.1 percent annual-chance (500-year or 1,000-year) event. For purposes of clear exposition in the sections that follow, these two approaches are treated as separate and distinct. In reality, however, blending elements of each into a hybrid approach may be desirable, as is discussed in more detail in the final section of the chapter.
A risk-standard approach establishes an acceptable risk and makes investments so that risks are reduced below this level. For example, an agency
FIGURE 4-1 Illustration of the risk-standard and benefit-cost approaches. The horizontal axis shows the levels of risk while the vertical axis shows net benefits. Starting from Point A, the status quo, investments in risk reduction initially have positive net benefits as the value of risk reduction exceeds the costs of investment. As risk levels decline toward zero, the costs of making further investments increase faster than the value of risk reduction and net benefits of further risk reduction begin to fall, eventually driving net benefits negative. The optimal risk reduction according to the benefit-cost approach falls at Point B and results in a higher level of risk (in this scenario) than the risk-standard approach, which is constrained by acceptable risk.
may design measures to eliminate or substantially reduce risk for events more frequent than a congressionally mandated level of risk reduction. This can be accomplished either by reducing the probability of the hazard (e.g., by building appropriately sized levees or dunes to substantially reduce risks up to a certain magnitude event) or by eliminating the consequences (e.g., by abandoning an impacted area or by elevating structures above the flood depth). Although risk standards do not consider costs explicitly, the risk-standard approach often implicitly considers costs, and decision makers in collaboration with stakeholders may choose to adjust the risk standard to allow for greater risk when significant costs are involved.
Applying a risk-standard approach requires two things. The first is a risk assessment that analyzes the probabilities and consequences of coastal hazards (see Box 1-2) and evaluates how investments could reduce either, thereby reducing risk. The second requirement for applying an acceptable-risk standard is getting agreement on what is acceptable versus unacceptable risk, which is discussed in more detail in the next section.
Rarely, if ever, is it possible to reduce risk to zero. If some level of risk is unavoidable, or could only be avoided at extreme cost, what is a low-enough level of risk that is satisfactory to stakeholders? How to define acceptable risk in this sense is a challenge. The acceptability of risk is not a technical question—it is a question of politics, economics, values, and ethics.
Standards for acceptable risks from a societal viewpoint have been developed in a number of countries including the United States, the United Kingdom, the Netherlands, Australia, and New Zealand. A prominent example of the application of acceptable-risk standards is in dam safety. Catastrophic failure of a dam can result in fatalities from flooding downstream of the dam. Dam safety programs model risks as a function of probability (in terms of the frequency of dam failures per year) and consequence (in terms of the number of fatalities per event). This approach is colloquially known as an FN curve because it plots the annual frequency (F) on one axis and the number fatalities (N) on the other (Figure 4-2). In the United States, the U.S. Bureau of Reclamation (USBR) led the way in using the FN curve to define the potential fatalities if certain specified dam failures occurred. This approach became an integral part of USBR practice in appraising dam safety risks and in making decisions on remedial actions to reduce safety risks at particular facilities (USBR, 2003). The U.S. Army Corps of Engineers (USACE) has recently adopted similar FN curve criterion for its dam safety program.
Acceptable-risk standards based on protection against loss of life are useful in dam safety but have limitations when considered for coastal risk reduction projects. The first limitation is that such acceptable-risk standards principally relate to catastrophes involving loss of life. Although they can be applied to financial costs, environmental impacts, or other non-loss-of-life consequences, to date this has been less common. Second, the residual risk associated with coastal risk reduction projects in the United States is much greater than residual risk permitted in modern dam safety criteria, in part because dams represent an engineered hazard, whereas coastal risk reduction projects are designed to protect against natural hazards. This
FIGURE 4-2 Dam safety societal risk requirements for new dams and major augmentations showing unacceptable (intolerable) risks (upper right) and acceptable risks that may be mitigated if practicable or are negligible (lower left). Risk is a function of the probability of an event (frequency, F) and its consequence (number of fatalities, N).
SOURCE: Reprinted, with permission, from NSW-DSC (2006).
inconsistency would require resolution before the criteria could reasonably be applied to the evaluation of coastal risk reduction projects.
Acceptable risk in more general contexts beyond evaluating risk to loss of life has proven hard to define. For example, The World Health Organization in addressing standards for water quality provides a number of different concepts that could be used in defining acceptable risk (Fewtrell and Bartram, 2001):
A risk is acceptable when: it falls below an arbitrary defined probability; it falls below some level that is already tolerated; it falls below an arbitrary
defined attributable fraction of total disease burden in the community; the cost of reducing the risk would exceed the costs saved; the cost of reducing the risk would exceed the costs saved when the “costs of suffering” are also factored in; the opportunity costs would be better spent on other, more pressing, public health problems; public health professionals say it is acceptable; the general public say it is acceptable (or more likely, do not say it is not); politicians say it is acceptable.
The concept of acceptable risk is defined by the United Nations (UNISDR, 2009) as “the level of potential losses that a society or community considers acceptable given existing social, economic, political, cultural, technical and environmental conditions.” The United Nations further states:
In engineering terms, acceptable risk is also used to assess and define structural and non-structural measures that are needed in order to reduce possible harm to people, property, services and systems to a chosen tolerated level, according to codes or “accepted practice”’ which are based on known probabilities of hazards and other factors.
Acceptable-risk standards can be set based on some objective quantifiable risk standard. For example, acceptable-risk levels that involve personal injury and death can find guidance in such concepts as the quality-adjusted life-years (Hunter and Fewtrell, 2001). Acceptable risk standards have also been applied in industry. Starting in the 1960s, the United Kingdom’s Health and Safety Executive developed criteria for “broadly acceptable” risks that appear to reflect ambient industrial risks that society willingly accepts, and “tolerable” risks that appear to reflect the highest industrial risks that society accepts if a corresponding benefit is derived (HSE, 2001; Jonkman et al., 2008, 2011). The above acceptable-risk standards have been developed with varying degrees of stakeholder involvement.
One issue with the concept of an acceptable-risk standard is that it singularly highlights one particular level of risk. In reality, less risk is preferable to more risk so that all reductions in risk have value and not just those that reduce risk to the acceptable-risk standard. The benefit-cost approach to risk reduction (discussed later in the chapter) allows for positive marginal benefits of risk reduction over the entire range of risk reduction.
Risk Perception and Setting Acceptable Risk
Setting the level of acceptable risk for any hazard is not a purely scientific or engineering matter but rather involves a societal value judgment. In a democratic society, the involvement of the public is essential for setting acceptable-risk standards. Views on what is acceptable by the
public can then be combined with technical analysis to determine what investments are necessary to meet the threshold levels of acceptable risk. The National Research Council (NRC, 1996) recommended adopting an analytic-deliberative approach with technical experts, public officials, and affected parties taking part in all steps of problem formulation, assessment, and policy recommendation. Hunter and Fewtrell (2001) suggest that the process for assessing acceptable risk should bring together experts with technical knowledge with the affected public. The experts would quantify the impacts of alternatives and present recommendations to be reviewed by all stakeholders. Critics of technocratic or expert judgment point to evidence that individuals and experts do not always agree on what risks are most important to address or to what degree risks should be reduced. Often, rankings of risk by individuals do not align with rankings of risks by experts based on the best evidence of relative risks (Slovic, 2000). If acceptable risk is a level of risk that a “society or community considers acceptable,” then it is a societal or community view of risk rather than expert or technocratic views of risk that are essential.
Social psychologists have provided ample evidence that risk perception varies by individual and circumstance and that “objective” statistics relied on by experts to assess risks only tell part of the story. The U.S. Environmental Protection Agency (EPA, 1988) concluded that “No fixed level of [individual] risk could be identified as acceptable in all cases and under all regulatory programs.” Many factors influence the public’s perception of risk: the voluntary or involuntary nature of the risk, the potential for catastrophe, the degree of familiarity, scientific uncertainty, the sense of dread, inequitable distribution of risks and benefits, and potentially irreversible effects among other things (Allen et al., 1992; Slovic, 2000; NRC, 2012b). For example, acceptable risk standards vary based on the perceived degree of voluntariness (Vrijling et al., 1995, 1998). The National Earthquake Hazard Reduction Program (Building Seismic Safety Council, 1995) established an “acceptable” probability of death for continuous occupancy in an engineered building in a high-seismicity area at 1 in 1 million (1 × 10–6) per person per year (Porter, 2002). This risk can be compared to other risks such as 2 × 10–7 fatal accidents per departure for commercial airline travel1 or the annual risk of dying in a motor vehicle accident of 1.1 × 10–4.2 The much higher accepted risk for motor vehicles noted above is often attributed to the more voluntary, and more routine, nature of the activity and the risk. The public appears willing to accept a risk up to 1,000 times greater for a voluntary risk compared with an involuntary risk (Starr, 1969).
Setting acceptable coastal risk standards would, therefore, be challeng-
ing and would require extensive stakeholder engagement, including members of the public, private interests, and relevant agencies at local, state, and federal levels. All parties would need to collectively consider the risks, societal perceptions of these risks, and the willingness of all parties to pay to reduce those risks.
The 1 Percent Chance (100-Year) Level of Risk Reduction
The 1 percent annual-chance event (see Box 1-3) is a commonly applied level of risk reduction in many inland flood control projects and some coastal risk reduction projects. Until the mid-1970s, Congress supported relatively high levels of risk reduction (e.g., the Standard Project Hurricane [see Box 1-2]). However, the establishment of the 1 percent chance (100-year) event to define the special flood hazard area for the National Flood Insurance Program altered the perception of flood risk. When mandatory flood insurance purchase requirements were waived beginning in the mid-1970s for properties located behind structures designed for a 1 percent chance flood, the 1 percent level of flood risk reduction became a de facto standard for many communities (NRC, 2013). The USACE no longer uses the 1 percent chance event as a standard basis of design for inland or coastal projects (see Chapter 2), although local sponsors often request and fund the additional costs for this level of risk reduction to eliminate flood insurance requirements for residents in flood hazard areas. On some projects, such as the USACE Hurricane Storm Damage Risk Reduction System of New Orleans, this criterion was specified in congressional legislation. Thus, the 1 percent chance event is usually selected as the basis of risk reduction efforts without an explicit calculation of the benefits and costs. Is a 1 percent annual chance of flooding a better choice than say a 2 percent chance or a 0.2 percent chance for coastal risk reduction? Also, why provide the same level of reduction in the probability of flooding to both a densely populated urban area with large immovable structures and a sparsely populated rural area with little in harm’s way? Surely what is at risk should also matter in designing coastal risk reduction investments.
In analyzing whether a given coastal risk reduction investment is worthwhile, the benefit-cost approach assembles evidence on the likely benefits and costs of the investment relative to the status quo (NRC, 2004b). For example, if investing in coastal risk reduction reduces the likelihood of flood damage to properties, the benefits of the investment for these properties can be found by evaluating the reduction in damages from storm events of various magnitudes, and multiplying this by the probability of occurrence of storms of these magnitudes. Consistent with the definition of risk, analysis
of the benefits of coastal risk reduction measures requires an assessment of both the probability of a hazard occurring and the consequence (change in benefits) if it occurs.
Unlike the risk-standard approach, the benefit-cost approach measures the value of risk reduction benefits in monetary terms. Measuring risk reduction in monetary terms is necessary to be able to compare benefits with costs in a common monetary metric. By using a common monetary metric, the benefit-cost approach also allows for incorporation of other costs or benefits associated with coastal risk reduction strategies, such as the value of reduced damages to property, loss of life, or business interruptions, as well as changes in the value of ecosystem services. It can be difficult to quantitatively include other benefits besides risk reduction in the risk-standard approach.
Strict adherence to benefit-cost analysis would recommend funding only those investments where benefits exceed costs. The benefit-cost approach can also be cast in terms of a return-on-investment (ROI) approach, which compares the ratio of the benefits to the cost and recommends investment in coastal risk reduction measures when the benefit-to-cost ratio exceeds a certain threshold. Setting the ROI threshold equal to 1 generates the same outcome as the benefit-cost approach. When faced with a binding budget constraint so that not all projects with positive net benefits can be funded, ROI can be used to set investment priorities. Projects can be ranked by ROI and, starting with the highest ROI, funding can be allocated to the next highest ROI-ranked project until the budget constraint is met.
In discussing environmental, health, and safety regulations, Arrow et al. (1996) state that “benefit-cost analysis can help illuminate the trade-offs involved in making different kinds of social investments. In this regard, it seems almost irresponsible to not conduct such analyses, because they can inform decisions about how scarce resources can be put to the greatest social good.” However, benefit-cost analysis is not the only useful information that should be considered by decision makers. Arrow et al. (1996) also state that benefit-cost analysis is “neither necessary nor sufficient for designing sensible public policy. If properly done, it can be very helpful to agencies in the decision-making process.”
Measuring the Benefits of Coastal Risk Reduction Investments
Although the basic logic of benefit-cost analysis is quite straightforward, there are a number of difficult issues in applying benefit-cost analysis to investments in coastal risk reduction. One of the most difficult issues involves accurately measuring the benefits of investments in coastal risk reduction in monetary terms, as needed in benefit-cost analysis.
Investment in coastal risk reduction can potentially provide a wide array of benefits such as reduced damages to property and infrastructure,
reduction in injury or loss of life, reduced social disruption for coastal communities, and improvement in an array of ecosystem services. Some of these benefits are relatively easy to measure in terms of monetary value, at least in principle. Damage to property and infrastructure can rely on information about loss of property value from storm events or flooding. Other benefits are much more difficult to measure in monetary terms. Valuing reduced disaster-related fatalities, increased socioeconomic stability for coastal communities, or restored ecological functions in monetary terms seems like a tall order. However, over the past half-century economists have developed an array of methods for estimating “nonmarket” value associated with environmental and social benefits that are often thought of as being difficult to impossible to value in monetary terms (Freeman, 2003). For example, economists have developed estimates of the value of clean air, clean water, or access to natural areas by looking at the premium in property values for otherwise similar properties located in areas with different environmental amenities (Harrison and Rubinfeld, 1978; Smith and Huang, 1995).
Even with advances in nonmarket valuation methods and applications, there remain large gaps in the ability to accurately measure benefits. Attempts to measure certain environmental or social benefits in monetary terms remain controversial. When the Exxon Valdez ran aground and spilled oil in Prince William Sound in Alaska, various parties sued Exxon for damages from the oil spill. Courts had to wrestle with questions about how much should Exxon pay to account for damages to the environment and various affected communities. These cases took well over a decade to litigate and spawned heated debate about the legitimacy of various methods to estimate nonmarket values associated with environmental degradation of the Sound (see, e.g., the debate over contingent valuation between Hanemann  and Diamond and Hausman ).
More fundamentally, some critics of benefit-cost approaches claim it is wrong-headed to try to boil down all values into monetary terms (Kelman, 1981; Sagoff, 1988). For example, how can biodiversity or spiritual and cultural values be evaluated in monetary terms? Even trying to do so might change how people think about these values and thereby distort these values. According to these critics, economic accounting should be restricted to market goods and services, and there should be separate consideration of other social and environmental values.
Another critique of benefit-cost analysis revolves around issues of equity and fairness (Ackerman and Heinzerling, 2002). Critics of benefit-cost analysis point out that the rich often get greater weight in benefit-cost analysis simply because they have more money. For example, consider a coastal risk reduction project for a community with 10 homes each worth $1 million versus another coastal risk reduction project for 50 homes each worth $100,000. The first project reduces the risk to property worth
$10 million while the second reduces risk to property worth $5 million. If both projects cost the same amount of money, benefit-cost calculations would favor doing the first project over the second. However, many observers would favor the second project over the first, in part because it affects more people and the people affected may have less ability to cope with loss.
USACE Benefit-Cost Analysis in Coastal Risk Reduction
Benefit-cost analysis has been used widely to evaluate government programs including investments in water projects (Howe, 1971; Brouwer and Pearce, 2005) and investments in environmental improvement under laws such as the Clean Air Act (EPA, 2011). The USACE has a long history of doing benefit-cost analysis dating back to the Rivers and Harbors Act of 1902 (and possibly earlier; Hammond, 1966). By the 1930s, benefit-cost analysis was well established as accepted practice, and the Flood Control Act of 1936 required that benefits exceed costs for USACE project approval. The U.S. Inter-Agency River Basin Committee, Subcommittee on Benefits and Costs (1950) produced a report known as the “Green Book” that attempted to standardize economic evaluation procedures required under the 1936 Act. Although the Green Book was never formally adopted, the Bureau of the Budget built upon the report in the development of Circular A-47 (Executive Office of the President, 1952), which established rigorous standards for evaluating federal water projects.
The 1965 Water Resources Planning Act and the Principles and Standards (WRC, 1973) that resulted from that legislation further shaped benefit-cost analyses for federal water resources project planning. The Principles and Standards required that four accounts be used for evaluating federal water projects—national economic development (NED), regional economic development, environmental quality, and social well-being—accounts that continue to be used in USACE planning today. Environmental quality and NED were originally established as coequal objectives, but this made large, structural engineering projects hard to justify (NRC, 2004b). In 1983, the Principles and Standards were repealed and replaced by the Principles and Guidelines (WRC, 1983; see also Chapter 2), which established a single objective for federal water resources projects “to contribute to national economic development consistent with protecting the Nation’s environment.”
At the time of the writing of the Principles and Guidelines, it was felt that other nonmarket environmental and social benefits could not be accurately evaluated in monetary terms, and these factors continued to be considered in separate accounts for Environmental Quality and Other Social Effects. NED—the increase in the value of marketed goods and services plus
project-related recreation benefits,3 minus construction, operations, and maintenance costs (USACE, 2011a)—became the most important decision criterion in the USACE planning framework. Aside from major adverse environmental impacts, environmental and social effects no longer significantly influenced water resources decisions (see also Chapter 2; a more detailed history is provided in NRC [1999, 2004b]). Although these policies remain in effect, there is ongoing vigorous debate about the principles and procedures governing the use of benefit-cost analyses in federal water resources projects, and as discussed in Chapter 2, Congress in WRDA 2007 directed the administration to revise the Principles and Guidelines.
What Should Count as a Benefit or a Cost?
Questions about whether all of the benefits or costs of investments in coastal risk reduction can be accurately measured in monetary terms raises the issue of whether benefit-cost analysis should attempt to be inclusive of all benefits and costs or whether there should be multiple accounts that are evaluated in different currencies that are not directly comparable (Polasky and Binder, 2012). A fully inclusive approach where everything is measured in a single monetary metric is appealing because it makes it easy for decision makers to compare outcomes and results in a simple and transparent decision rule based on net benefits. But if it is not possible to accurately assess all values in a common metric, then net benefits could systematically under- or overweight some benefits and generate biased decisions. In cases where there are important social, cultural, or ecosystem benefits that are difficult to quantify or monetize, it may be preferable to keep multiple accounts and set standards for acceptable outcomes for each account, or to use some form of multicriteria decision analysis (Keeney and Raiffa, 1993).
By creating separate accounts for environmental and social benefits but focusing on NED as the primary account, the Principles and Guidelines favor projects that score well in terms of value of marketed goods and services while giving inadequate weight to nonmarketed environmental and social benefits. The first part of the congressionally mandated revisions to the Principles and Guidelines—the 2013 Principles and Requirements, developed by the White House Council on Environmental Quality (CEQ, 2013)—give expanded consideration to environmental and social benefits rather than giving primacy to economic development (NED).
The Principles and Requirements summarized the limitations of the earlier approach:
3Under current guidance, recreation benefits may not exceed 50 percent of the overall project benefits.
Heretofore, Federal investments in water resources have been mostly based on economic performance assessments which largely focus on maximizing net economic development gains and typically involve an unduly narrow benefit-cost comparison of the monetized effects. Non-monetized and unquantified effects are often included in the overall analysis process, but are not necessarily weighted as heavily or considered key drivers in the final decision making process. As a result, decision making processes are, at this point in time, unnecessarily biased towards those economic effects that are generally more easily quantified and monetized. A narrow focus on monetized or monetizable effects is no longer reflective of our national needs, and from this point forward, both quantified and unquantified information will form the basis for evaluating and comparing potential Federal investments in water resources to the Federal Objective. This more integrated approach will allow decision makers to view a full range of effects of alternative actions and lead to more socially beneficial investments.
The Principles and Requirements (CEQ, 2013) emphasizes including all benefits and costs in a common framework where feasible, via an ecosystem services approach:
The ecosystems services approach is a way to organize all the potential effects of an action (economic, environmental and social) within a framework that explicitly recognizes their interconnected nature. The services considered under this approach include those flowing directly from the environment and those provided by human actions. Services and effects of potential interest in water resource evaluations could include, but are not limited to: water quality; nutrient regulation; mitigation of floods and droughts; water supply; aquatic and riparian habitat; maintenance of biodiversity; carbon storage; food and agricultural products; raw materials; transportation; public safety; power generation; recreation; aesthetics; and educational and cultural values. Changes in ecosystem services are measured monetarily and non-monetarily, and include quantified and unquantified effects. Existing techniques, including traditional benefit costs analyses, are capable of valuing a subset of the full range of services, and over time, as new methods are developed, it is expected that a more robust ecosystem services based evaluation framework will emerge.
As noted in Chapter 2, these changes are not anticipated to take effect until after revisions to the accompanying detailed interagency guidelines are released, and congressional action has so far blocked USACE implementation of the Principles and Requirements.
In principle, benefit-cost analysis should include all benefits and costs of investments in coastal risk reduction including changes in the value of ecosystem services, the value of reduction in risk of fatalities or injuries, as well as the reduction in losses to property and infrastructure, and the
direct costs of investment. However, it is difficult to quantify or monetize all of the impacts of investment in coastal risk reduction. Some benefits and costs may be relatively small and it might not be worth the investment of resources necessary to analyze these. In cases where benefits or costs are potentially large but it proves too difficult to estimate monetary values, impacts should still be quantified to the extent possible and constraints should be put on what is considered an acceptable outcome.
Valuing Reductions in Potential Loss of Life
With notable exceptions, such as Hurricane Katrina, relatively few people are killed by coastal storms in the United States compared with other natural catastrophes. Normally, transportation infrastructure for moving people from the coastline is sufficient given advanced warning of approaching storms. In fact, a major benefit of investments in advanced warning of approaching storms or improved transportation infrastructure is a reduction in the expected number of fatalities and injuries. Despite the gains in this area, coastal storms still pose a potential for causing fatalities and injuries, and as such are an important consequence of coastal catastrophes that should be included in the accounting of benefits and costs of coastal risk reduction.
Economists estimate the value of the reduction in the risk of fatalities using the concept of the value of a statistical life (VSL). VSL represents a typical person’s willingness to pay to reduce the risk of premature mortality (Mishan, 1971); “for example, a mortality risk of 1/50,000 might be valued at $100, producing a VSL of $5 million” (OMB, 2010). Estimates of VSL can be generated by analyzing wage premiums needed to attract workers to risky jobs or other decisions involving risk of fatality (Mrozek and Taylor, 2002; Viscusi and Aldy, 2003). Estimates of VSL can also be generated by asking people survey questions on how they would choose between risk and income (Krupnick et al., 2002).
The use of VSL in federal decision making, particularly in regulatory applications, is widespread and there is extensive literature on its use (Viscusi, 2004). For example, EPA has long used VSL in evaluating the benefits of the Clean Air Act in reducing mortality due to reduced exposure to air pollution (EPA, 2011). VSL estimates vary depending on methods and data used to construct the estimates as well as by income levels of the populations at risk (Viscusi and Aldy, 2003). The Office of Management and Budget provides guidance to agencies on theory and application of VSL (OMB, 2003) and summarizes the values used by various agencies, noting that these values vary “from roughly $1 million to $10 million per statistical life” but mostly fall in the range above $5 million (OMB, 2010). EPA uses a VSL of $6.3 million (2000 dollars), while the Food and Drug
Administration uses $7.9 million (2010 dollars) and the Department of Transportation uses $6.0 million (2009 dollars) (OMB, 2010).
Rather than deal indirectly with the benefits of reduced fatalities through standard risk criteria, VSL calculations allow risk reductions to be included in benefit-cost analysis along with other benefits and costs measured in monetary terms.
Benefit-Cost Analysis in the Context of Long-Lived Projects
Long-lived investments require an explicit consideration of the future, which necessitates a decision on how to compare present versus future benefits and costs. Investments in coastal risk reduction often generate long-lasting benefits in the form of reduced risk or increases in ecosystem services, or costs in the form of ongoing operation and maintenance expenses or reductions in ecosystem services. Economists discount future values to make them comparable to present values. The typical rationale for discounting is that resources can be invested and earn a positive rate of return (interest) so that receiving an equal amount of money in the future is actually worth less than receiving that amount today.
While discounting is standard practice in most business and economic applications, there are open questions about how to aggregate benefits and costs over time for long-lived projects with significant social or environmental consequences, such as investments in coastal risk reduction. First among these is the proper discount rate to use. OMB recommends a discount rate of 7 percent, but this rate results in greatly reduced benefits or costs beyond a few decades. Others have argued that 7 percent is too high for long-term investments, with longer time horizons and uncertainty about future rates of return leading to lower discount rates (Weitzman, 2001; Arrow et al., 2013). Other debates revolve around whether societal decisions that affect future generations should be treated in the same fashion and use the same discount rate as private investment decisions (see, e.g., the debate around discounting in the context of climate change policy between Nordhaus  and Stern [2007, 2008]). Investments in coastal risk reduction do not raise unique issues in regard to discounting, but evaluating the net present value of such investments requires making potentially difficult or controversial decisions such as determining the proper rate of discount to use in project evaluation.
The long-term nature of investments in coastal risk reduction also means that investment and management will be ongoing rather than a one-time decision. Such recurrent decision making calls for some form of adaptive management (discussed in detail in Box 5-1) in which current investments are evaluated not only with respect to how they affect expected
net benefits but also whether the investment maintains or opens options, and whether the investment allows for greater learning about future conditions or the effectiveness of alternative approaches, thereby improving future decision making. There is value (called option value or quasi-option value in the economics literature) to maintaining flexibility (i.e., preserving the option to be adaptive) in the face of uncertainty about the future (Arrow and Fisher, 1974). There is also a “value of information” from learning about future conditions before committing to irreversible decisions because better information allows decisions that are better matched to likely conditions (Hanemann, 1989).
Investments in coastal risk reduction generate benefits, some of which accrue primarily to those who live in the coastal communities (e.g., projects to reduce the probability of flooding houses and other private property). Some portion of the costs of coastal risk reduction investments is typically paid by federal taxpayers, including those who live far from the coast. The distribution of benefits and costs across different groups in society raises issues about the proper sharing of responsibilities and rewards for risk reduction. Addressing who should pay the costs of coastal risk reduction raises fairness or equity concerns that are not easily answered. Requiring coastal communities to foot the entire bill for investments in risk reduction can place unaffordable burdens on these communities, especially for those with lower or middle incomes. In addition, some benefits from coastal risk reduction generate widespread benefits that go well beyond just the residents of the coastal community being protected, such as recreation or tourism benefits for nonresidents. But having taxpayers elsewhere pay for investments that provide primarily local benefits is also potentially unfair, especially if taxes come from lower- or middle-income taxpayers and go to wealthy coastal communities. Although there are some general principles that can be applied, such as trying to align costs with beneficiaries of coastal risk reduction, there is typically no simple right answer to distributional issues, and often it is up to the political process to sort out competing claims about what is fair.
The preceding two sections have laid out two coherent but different approaches to evaluating investments in coastal risk reduction—a risk-standard approach and a benefit-cost approach. Although each approach
has considerable appeal and numerous examples of application, each approach also has at least one significant weakness. Because the risk-standard approach does not typically factor in benefits other than risk reduction benefits, such as ecosystem services, or explicitly consider costs, this approach may result in choosing investments that yield considerably lower net societal benefits than would alternative investments decisions. The benefit-cost approach, on the other hand, faces the daunting challenge of trying to measure all environmental and social impacts in monetary terms. If such values cannot be accurately measured in monetary terms, then the resulting benefit-cost analysis will be incomplete and misleading.
Given the limitations with each approach, there is an advantage of not rigidly adhering to either approach in its purest form but instead incorporating some elements from each and adopting a hybrid risk-constrained benefit-cost approach. This hybrid approach retains the emphasis on choosing investments that increase net benefits, as in benefit-cost analysis, but puts constraints on what is considered as an acceptable outcome (see, e.g., Figure 4-1). These constraints may arise from societal views on unacceptable risks to which individuals or groups should not be exposed, considerations of equity, or other concerns. Coastal risk planning currently under way in the Netherlands (Box 4-1) represents an example of a hybrid approach that accounts for benefits and costs of investment but adds constraints based on acceptable fatality risk.
A risk-constrained benefit-cost approach is similar in spirit to conclusions of Arrow et al. (1996), who in principle favor the use of benefit-cost analysis in analyzing environmental, health, and safety regulations but are well aware of the practical difficulties of implementing benefit-cost analysis:
Most economists would argue that economic efficiency, measured as the difference between benefits and costs, ought to be one of the fundamental criteria for evaluating proposed environmental, health, safety regulation. Because society has limited resources to spend on regulation, benefit-cost analysis can help illuminate the trade-offs involved in making different kinds of social investments…. In practice, however, the problem is much more difficult, in large part because of inherent problems in measuring marginal benefits and costs…. [N]ot all impacts can be quantified, let alone be given a monetary value. Therefore, care should be taken to assure that quantitative factors do not dominate qualitative factors in decision-making. If an agency wishes to introduce a “margin of safety” into a decision, it should do so explicitly.
This hybrid approach is also similar in spirit to OMB and OSTP guidelines on use of risk analysis in federal agencies (Dudley and Hays, 2007). These guidelines recommended that
The Coastal Risk Approach in the Netherlands: Past, Present, and Future
Until 1953, coastal protection in the Netherlands was in the hands of 2,600 local water boards, which grew out of medieval grassroots democratic organizations. In 1953 a large flood with a return period of about 250 years overwhelmed the coastal levees in the southwest of the Netherlands with more than 1,800 casualties and an economic loss of 10 percent of gross domestic product. As a result of this catastrophe, coastal risk began to be treated in a more rational and uniform way.
The First Delta Commission advised in 1960 that flood protection levels should be determined based on the value of the property to be protected and the cost of protection (i.e., benefit-cost analysis) (Deltacommissie, 1962). This protection level was cast in terms of a probability of flood for different regions in Holland, ranging from a probability of 1/10,000 per year in central Holland (protecting most major cities) to 1/1,250 (river floodplains) (Figure 4-1-1).
The Commission did consider the value of “loss of life” and other difficult-to-measure values, which were added as a multiplication factor over the real estate value. The Commission factored in land subsidence due to oxidation and the effect that closure to the estuaries would have on the surrounding coast, but did not consider sea-level rise because this was not a known issue in the 1960s. The proposed protection levels remain the law of the land.
As of 2013, the floodprone areas, which constitute about 60 percent of the total land surface of the country (34,000 km2), are protected with 95 dike rings with a total length of 3,700 km of dunes and (primary) levees. The protected area includes about 10 million people and 2,000 billion Euros of investments. The 650 km of sea and estuarine coasts is protected by about 15 dike rings, and 27 other “sea defenses,” such as a closure dam, smaller dams, storm surge barriers, and sluices (Kind, 2013; Ministerie van Verkeer en Waterstaat, 2007).With such a large portion of the country, property, and population in floodprone areas, the Dutch flood risk reduction strategy has been to prevent flooding outright, hence the high levels of protection. In fact, evacuation plans for the larger urban areas do not currently exist as they are deemed not viable logistically.
Recent and Future Developments
Because of the combination of sea-level rise, soil subsidence, increased river runoff, and economic and population growth, flood risk is expected to increase quickly in the Netherlands—in some areas by 4 to 8 percent annually. By the end of the century, this amounts to a flood risk increase by a factor of 30 to 700. This means that the protection standards based on the situation of 1960 are no longer tenable. In 1995, high discharges of the Rhine and Meuse Rivers led to evacuation of more than 250,000 people because of fear the levees would break, which was a wake-up call to the nation. To address these issues, the Dutch government commissioned the Second Delta Commission in 2008. Apart from changing hazard conditions and socioeconomic development, the commission also took natural and cultural values into account. On that basis, they advised a 10-fold, across-the-board increase in the protection levels for all dike ring areas (Kind, 2013; Deltacommissie, 2008).
FIGURE.4-1-1 Flood risk criteria for the Netherlands.
SOURCE: Reprinted, with permission, from Van der Most and Wehrung (2005). © 2005 by Natural Hazards.
The Dutch government launched a separate project, Flood Protection for the 21st Century (WV21), which proposed an alternate differentiated approach (Kind, 2013). The WV21 project used benefit-cost analysis and analysis of fatality risk as a basis for proposed new risk reduction standards. Proposed standards are derived on the basis of an optimal dike investment strategy denoting when, where, and how much to invest. The damage cost also includes the cost of human life and other aspects that are difficult to value in monetary terms, with minimal tolerable fatality risks considered separately from total costs and benefits. Additionally, the proposed protection level is no longer cast in terms of a probability of exceedance of the water level but in terms of a probability of flooding, which takes into account potential levee failures.
This proposed risk framework has not yet been adopted and is the focus of ongoing policy deliberations in the Netherlands, but flood protection improvements are under way.
Agencies should set priorities for managing risks so that those actions resulting in the greatest net improvement in societal welfare are taken first, accounting for relevant management and social considerations such as different types of health or environmental impacts; individual preferences; the feasibility of reducing or avoiding risks; quality of life; environmental justice; and the magnitude and distribution of both short and long-term benefits and costs.
This hybrid approach is not entirely dissimilar from the current USACE project planning framework, which is constrained by severe environmental impacts (see Chapter 2). However, aside from this constraint, the USACE planning process largely relegates social and environmental factors to levels that do not influence decision making. The USACE approach could be improved through a broader consideration of benefits and costs (as reflected in the Principles and Requirements), including life-safety, environmental, and societal benefits and costs where feasible.
A major challenge with implementing a risk-constrained benefit-cost approach is deciding what categories of coastal risk reduction benefits and costs should be incorporated directly into the benefit-cost calculation, and what categories are best handled by qualitative or nonmonetary quantitative analysis that are incorporated via constraints on what is acceptable versus unacceptable. When constraints are adopted there is also the difficult decision of what outcome levels are viewed as acceptable versus unacceptable.
Investments in coastal risk reduction generate significant benefits to society by reducing risk to people and property, but they also involve significant costs. Increases in development and population along the coast, along with sea-level rise, only increase the stakes involved in protecting vulnerable coastal areas. This chapter reviewed two approaches for determining what investments in coastal risk reduction are justified: (1) a risk-standard approach and (2) a benefit-cost approach. Although each approach has considerable appeal, each also has at least one significant weakness. In the case of the risk-standard approach, it is difficult to factor in non-risk-related benefits or costs. In the case of the benefit-cost approach, it is difficult to evaluate all environmental and social impacts in monetary terms. Given the limitations with each approach, there are advantages of not rigidly adhering to either approach in its purest form but instead incorporating some elements from each.
Benefit-cost analysis constrained by acceptable risk and social and environmental dimensions provides a reasonable framework for evaluating
coastal risk management investments. Investments in coastal risk reduction should be informed by net benefits, which include traditional risk reduction benefits (e.g., reduced structural damages, reduced economic disruption) and other benefits (e.g., life-safety, social, environmental benefits), minus the costs of investment in risk reduction and environmental costs. However, because it is difficult to quantify and monetize some benefits and costs, it is important to expand the analysis to include considerations of difficult-to-measure benefits or costs through constraints on what is considered acceptable in social, environmental, and risk reduction dimensions. Such unacceptable levels of risk may include a level of individual risk of fatality, the risk of a large number of deaths from a single event, or adverse impacts on social and environmental conditions that may be difficult to quantify in monetary terms. It is difficult, however, to establish societally acceptable risk standards and requires extensive stakeholder engagement. Setting such a standard requires value judgments, on which not all individuals or groups will necessarily agree.
The recently updated federal guidance for water resources planning—the 2013 Principles and Requirements for Federal Investments in Water Resources—provides an effective framework to account for life-safety, social impacts, and environmental costs and benefits in coastal risk reduction decisions. The Principles and Requirements, developed by the White House Council on Environmental Quality in response to a 2007 congressional mandate, represents the first step toward federal water resources policy reform. The document, which applies to water resources investment decision making across the federal government—not just within the USACE—recognizes that water resources investment decisions should also consider social and environmental impacts and not give primacy to benefits or costs that are easily measurable in monetary terms. This represents a significant improvement upon current USACE planning, which uses separate accounts for social and environmental impacts, with largely qualitative measures, effectively relegating such considerations to second-class status behind net economic benefits. Progress has been made on measuring improvements in economic terms and on measuring the value of some ecosystem services and social benefits. For other environmental and social factors that are not easily measured in dollar terms, the Principles and Requirements recognize that these costs and benefits should also be given adequate weight in decision making. The Council on Environmental Quality should expedite efforts to complete the detailed accompanying guidelines for implementing the 2013 Principles and Requirements, which are required before this framework can to be put into action to improve water resources planning and coastal risk management decision making at the federal agency level.
Until the updated guidelines to the Principles and Requirements are finalized, there are steps the USACE could take to improve consideration of
multiple benefits and costs in the current decision process. Specifically, further attempts in the USACE planning process could be made to more quantitatively consider information in the Environmental Quality and Other Social Effects accounts. For example, work that has been done on how to value ecosystem services could be used to value some environmental quality benefits. Once quantified, these costs and benefits should be rigorously considered and clearly communicated to stakeholders. Such an approach could result in different decision outcomes if the additional social and environmental benefits make certain strategies more acceptable to local sponsors and stakeholders than others. However, trying to quantify or monetize social effects and some environmental effects remains challenging. When only some benefits or costs are monetized there is a tendency to overlook or downplay nonmonetized benefits or costs, and additional attention and/ or institutional mechanisms are needed to ensure that these benefits are given adequate weight.
There is no solid basis of evidence to justify a default 1 percent annual-chance (100-year) design level of coastal risk reduction. The 100-year flood criterion used in the National Flood Insurance Program was established for management purposes, not to achieve an optimal balance between risk and benefits. There is also no evidence that reducing risk to a 1 percent annual-chance event is in the best interests of society or that this level is necessarily acceptable to the general public. This level of risk reduction may be appropriate in some settings, unwarranted or excessive in others, and inadequate in highly developed urban areas. Such decisions should, instead, be informed by risk-constrained benefit-cost analyses reflecting site-specific conditions.