Society for Benefit-Cost Analysis

Events available for Registration...

SBCA Webinar Series: All Access
07/02/20 - 12/31/2020
register for this event...

SBCA Webinar Series: Surviving Judicial Review
08/04/20 - 8/4/2020
register for this event...

Society for Benefit-Cost Analysis: European Conference 2020
10/14/20 - 10/14/2020
register for this event...

Events in the month of December 2018

Date

Event

12/8/2018

Roundtable: EPA Transparency in Science and Economics, Chair: Lisa Robinson, Co-sponsored by the Society for Benefit Cost Analysis (SBCA)

Roundtable: EPA Transparency in Science and Economics

Room: Mardi Gras Ballroom D, 3rd Floor 10:30 am–12:00 pm

Chair(s): Lisa Robinson robinson@hsph.harvard.edu

Sponsored by Economics and Benefits Analysis SG & The Society for Benefit Cost Analysis (SBCA)

In April 2018, the U.S. Environmental Protection Agency (EPA) issued a notice of proposed rulemaking entitled “Strengthening Transparency in Regulatory Science.” In June, the Agency then announced that it was considering whether to propose a second rule on “Increasing Consistency and Transparency in Considering Costs and Benefits in the Rulemaking Process.” The titles of each initiative suggest worthy goals, but they have been controversial. One set of concerns relates to the political motivations, given the current Administration’s interest in rolling-back environmental regulations. A second set of concerns relates to the substantive details of each proposal, which raise several complicated and difficult questions.

In this roundtable, we will focus on this latter set of more substantive concerns, bringing together a cross-disciplinary group of experts to discuss the details and implications of the proposals from a variety of perspectives. We will provide an overview of each proposal, comment on the issues raised and on how to best address them, then open the discussion to the audience. The speakers include:

• Lisa A. Robinson, Senior Research Scientist, Center for Health Decision Science and Center for Risk Analysis, Harvard T.H. Chan School of Public Health
• Shaun Goho, Deputy Director, Emmett Environmental Law and Policy Clinic, Harvard Law School
• Anthony (Tony) Cox, Clinical Professor, University of Colorado, and President, Cox Associates; Editor-in-chief, Risk Analysis, and Chair, EPA Clean Air Scientific Advisory Committee
• George Gray, Professor, Milken Institute School of Public Health, George Washington University; former Assistant Administrator, EPA Office of Research and Development and EPA Science Advisor
• Clark Nardinelli, former Chief Economist, U.S. Food and Drug Administration

more info...

12/8/2018

Economics of Food Risk-Benefit Analysis, Chair: Aliya Sassi, Co-sponsored by the Society for Benefit Cost Analysis (SBCA);

Economics of Food Risk-Benefit Analysis

Room: Mardi Gras Ballroom D, 3rd Floor 1:30 pm–3:00 pm

Chair(s): Aliya Sassi aliya.sassi@fda.hhs.gov

Sponsored by Economics and Benefits Analysis SG & The Society for Benefit Cost Analysis (SBCA)

1:30 pm Targeting and Evaluating Investments in Food Safety in Low and Middle Income Countries. Hoffmann S*, Muhammad A, Meade B; USDA Economic Research Service shoffmann@ers.usda.gov

Abstract: In 2016, the WHO released the world’s first estimates of the global incidence and burden foodborne diseases. In low and middle income countries, these regionally specific estimates provide a foundation for analysis that can help target efforts to reduce foodborne illness. This project integrates the authors’ estimates attributing regional foodborne illnesses to specific food exposures and estimates of the influence of income and price trends on national food consumption patterns around the world to show how simulation modeling can be used to help forecast how national food disease and burden patterns might change over time. The modeling uses prior research and sensitivity analysis to explore how patterns of improvement in food safety management and historic data might change with income and price trends.

1:50 pm The Impact of Food Safety Outbreaks on Produce Supply, Imports, and Exports. Astill GA*, Minor T; Economic Research Service, US Department of Agriculture gregory.astill@ers.usda.gov

Abstract: Nearly half of US foodborne illnesses with a known food vehicle are linked to produce commodities. Foodborne illness outbreaks identified to be associated with a produce commody lead to drops in both domestic and foreign demand in addition to possible limits on supply imposed by public health agencies. In the most recent example, 172 people in the US became ill from eating romaine lettuce between March and May 2018. The public health message communicated was “avoid eating raw romaine lettuce”, and there was likely a significant demand response to the outbreak and the media surrounding it. The contaminated romaine lettuce was likely linked to a small number of specific farms or firms and a small portion of the entire romaine lettuce market. However, the inability of public health agencies, consumers, or retailers to rapidly distinguish clean from contaminated product impacted the entire market for the commodity. We examine the impacts of two decades of foodborne illness outbreaks in the US associated with produce commodities on the volume of the commodity sold on the market, the volume exported to foreign markets, and the volume imported to the US domestic market. The market reactions impacting an entire commodity identified in this work negatively impact the earnings of many farms and firms not associated with contamination as well as the benefits of consumers of the product. This work has implications for the development of supply chain traceability by industry, of information sharing among industry and government, and of investigation and dissemination of findings by government.

2:10 pm A Retrospective Analysis of the U.S. Food and Drug Administration’s Seafood HACCP Rule and Program. Sassi A*, Marasteanu J; FDA aliya.sassi@fda.hhs.gov

Abstract: In 1994, the U.S. Food and Drug Administration published the “Procedures for the Safe and Sanitary Processing and Importing of Fish and Fishery Products” proposed rule. It was finalized in 1995 and went into effect in 1997. The aim of this rulemaking was to reduce potential risks to seafood consumers from hazards such as bacteria, viruses, toxic chemicals, natural toxins, parasites and other source and process hazards, and to reduce the number of foodborne illnesses from contaminated seafood. The additional concern was that while more than half of seafood consumed in the U.S. was imported from other countries, some of the countries lacked infrastructures capable of supporting sufficient seafood safety. The rule mandated that all seafood processers and importers must implement a preventative safety system that applies Hazard Analysis Critical Control Point (HACCP) principles, a system that would allow for identifying and minimizing occurrence of hazards in their products. FDA estimated and reported in the original Regulatory Impact Analysis (RIA) for the final seafood HACCP rule that the costs would range from $0.677 to $1.488 billion, while the benefits would range from $1.435 to $2.561 billion. After the 2001 report by the General Accounting Office and the 2000 FDA internal evaluation revealed that certain segments of the seafood industry lagged in compliance, FDA instituted a Mid-Course Correction enforcement initiative. To strengthen its Seafood HACCP program, FDA intensified its focus on seafood products and processors that presented the highest risk to consumers. We now intend to retrospectively examine the impacts of this rule and the Mid-Course Correction on consumers’ health, the seafood industry, and U.S. economy. Using data on foodborne illnesses, imports, exports and industry costs from various sources, including the Centers for Disease Control and Prevention and the FDA, our goal is to retrospectively quantify, describe, and compare these impacts to the original costs and benefits discussed in the RIA for the final Seafood HACCP rule.

2:30 pm Nutritional Risk and Benefit Associated with Red Meat Consumption in France. De Oliveira Mota J*, Tounian P, Guillou S, Pierre F, Membré JM; INRA, Oniris, Route de Gachet, CS 40706, Nantes, France juliana.de-oliveira@oniris-nantes.fr

Abstract: The consumption of red meat has become a public health concern, notably due to its link with colorectal cancer. On this basis, World Health Organization and World Cancer Research Fund have classified red meat as possibly carcinogenic. However, red meat also has beneficial effects by the nutritional contributions, in particular iron, which contributes to the decrease of iron deficiency and, subsequently, anemia. The aim of the study was to quantify the nutritional risk and benefit associated with red meat consumption, and finally to express them in the same metric to enable a comparison of the impact of various scenarios of exposure on the outputs. A probabilistic risk-benefit assessment model was built to quantify the risk of colorectal cancer and the benefit of iron deficiency reduction from red meat consumption. It was established per age class and gender. The effects of variability and uncertainty on the model output were also characterized. Using this model, the burden of diseases of risk/benefit, from the scenario of intake, was calculated. This study will help authorities and consumers to determine the optimal quantity of red meat to consume to take the benefits without the risks. More broadly, results of this study could be compared with additional risks or benefits due to red meat and even with other food substitute health impacts, which is a part of the Risk and Benefit Assessment research area.

more info...

12/9/2018

Improving Risk-Benefit Estimates for Regulatory Analysis

Room: Mardi Gras Ballroom D, 3rd Floor 10:30 am–12:00 pm

Chair(s): Amber Jessup Amber.Jessup@HHS.GOV

Sponsored by Economics and Benefits Analysis SG & The Society for Benefit Cost Analysis (SBCA)

10:30 am Two Far-reaching Changes to How We Quantify and Balance Risks and Regulatory Costs. Finkel AM*; University of Michigan and University of Pennsylvania afinkel@upenn.edu

Abstract: When the current period of disdain for risk-reducing regulation ends, the first question facing a new Executive Branch and Congress will be whether to restore “the old days” of evidence-based regulation, or to consider new ideas for identifying and implementing actions whose health/safety/environmental benefits dwarf their costs. I start from the premise that the 1970-2016 period was not in fact a golden age for cost-benefit-based regulation—risks were often underestimated, costs generally grossly overstated, aspirational goals were set and not tracked or achieved, risk-risk tradeoffs were ignored, and distributional inequities were unexamined. Therefore, I offer two changes to cost-benefit balancing that might swing the pendulum beyond neutral and towards maximally welfare-increasing policies. First, our system’s basic terminology and language places avoiding costs above avoiding risks. For example, a “regulatory budget” (a per-agency amount of regulatory costs that can’t be exceeded, regardless of greater benefits) is a “serious” idea, but no one has proposed the exact obverse, an annual “benefits budget” (a per-agency minimum amount of regulatory benefits that must be met, so long as costs are smaller). Similarly, economists always define “cost-effective” regulatory policies as ones that achieve a given level of benefit at the least cost, rather than the obverse (policies that achieve the maximal amount of benefit for a given cost). Secondly, our method for monetizing risks and comparing them to costs tacitly biases policy towards harms (either environmental/health harms or monetary ones) that affect large populations in very diffuse ways, and away from economic or non-economic suffering that affects small populations intolerably. We could instead construe net benefit not as total population benefit minus total cost, but as the amount of suffering a regulation would avert net of the economic suffering it imposes—shining a light on intolerable individual risks and on unemployment or bankruptcies.

10:50 am How Can We Improve Efficiency of Clinical Trials? Sertkaya A*, Jessup A, Ertis D; Eastern Research Group, Inc. aylin.sertkaya@erg.com

Abstract: Clinical trials are an essential part of bringing new drugs and devices to market, but the cost, duration, and complexity of trials have been steadily increasing. According to one estimate, costs could be as high as $1.4 billion for drug development, and as high as $94 million for development of medical devices that require Premarket Approval (PMA). The increasing cost of clinical research has significant implications for public health, as it affects companies’ willingness to undertake clinical trials. Using information from published studies and expert opinion, we identified several strategies for improving the efficiency of clinical trials for new drug and complex medical device development. We then analyzed the likely potential impact of these strategies on clinical trial costs, duration, and phase transition success probability. We find that electronic health records followed by reduced source data verification and increased use of mobile technologies have the highest potential for improving clinical trial efficiency overall for drugs. For complex medical devices, use of simplified protocols and reduced amendments, followed by use of patient registries and improvements in FDA review efficiency and interactions show the biggest promise among the eleven strategies considered.

11:10 am Discovering Relevant Covariates with Sparse Discrete Choice Models for the Valuation of Environmental Goods. de la Maza C*, Davis A, Azevedo I; Carnegie Mellon University cristobal.delamazag@gmail.com

Abstract: Discrete choice models are widely used in economics and psychology to understand individual preferences. Because of its simplicity, the multinomial logit model has been the workhorse for preference learning for the last 40 years. The model can capture preference heterogeneity for different socioeconomic characteristics of decision makers. In a standard discrete choice experiment, detailed socioeconomic information is requested from participants. Usually in a choice model, depending on the number of socioeconomic covariates, the number of estimated coefficients can be substantial. Parsimonious models can instead identify the stylized elements in a choice task. In this paper, we use tools from machine learning to discover relevant covariates. We develop and test a method for high-dimensional preference modeling, the sparse multinomial logit model, where irrelevant coefficients are set to zero with an adaptive lasso penalty. We illustrate the approach discovering relevant socioeconomic covariates with multinomial logit models both in simulated an empirical data. Our model can systematically recover the true support at the cost of inducing bias in coefficients. In economic and policy analysis, bias in coefficients is a serious concern, as can lead to incorrect conclusions and policy distortions. We further show how to debiased our sparse model results with state-of-the-art convex optimization methods. We test the model on a real data set, discovering relevant socioeconomic covariates in a case study of environmental valuation. We evaluate our model performance with data from a conjoint analysis survey designed to estimate willingness to pay to avoid environmental impacts associated with different electricity generation technologies. The effect on willingness to pay of knowledge, experience, income, family composition, pro-social behavior, trust in public institutions and personal responsibility is uncovered by the model. Taste heterogeneity is fully captured as our model overpowers results from a mixed logit model that capture these differences as random effects.

11:30 am Estimating Social Benefits of N Lives Saved: Preliminary Experiments on Framing Effects and Nonlinear Value Functions. Johnson BB*, Finkel AM; Decision Research; University of Michigan branden@decisionresearch.org

Abstract: All extant estimates of the value of a statistical life (VSL)—the metric that dominates most regulatory benefit estimates, thus determining the stringency of much health-and-safety regulation—may fall short in two fundamental ways. First, VSL is defined as the personal tradeoff between one’s own money versus small incremental probabilities of one’s own mortality, whereas regulations trade off monies amassed from large numbers of consumers versus multiple lives prolonged across large populations. The decisions themselves involve shared national purpose, but the valuation procedures underlying estimated VSL do not, deliberately excluding altruistic (or misanthropic) motivations. Second, the way we use VSL estimates tacitly assumes that benefit is strictly linear as a function of mortality risk, regardless of whether the probability of harm is trivially small or alarmingly large, and assumes the same linearity in estimation of regulatory costs, despite universal acceptance that cost has increasing marginal disutility. Several online experiments elicit estimates of the “social benefit of N lives saved,” probing effects of different framings of the tradeoff questions on responses’ magnitude and interindividual distribution, and whether and how different kinds of altruism (concern for others’ longevity alone, or for their net benefit) may shape reactions. Another experiment estimates parameters of a value function for risk that may reflect sub-linear valuation of de minimus risk reductions and/or supra-linear valuation of “intolerable” risks, and of a similar function for cost (when described as fractions of a household’s income) that may allow for de minimus costs to be discounted and/or “intolerable” costs to warrant special attention. We report here on preliminary results and implications.

more info...

12/9/2018

Economics of Environmental Policy, Chair: Lisa Robinson, Co-sponsored by the Society for Benefit Cost Analysis (SBCA);

Economics of Environmental Policy

Room: Mardi Gras Ballroom D, 3rd Floor 1:30 pm–3:00 pm

Chair(s): Lisa Robinson robinson@hsph.harvard.edu

Sponsored by Economics and Benefits Analysis SG & The Society for Benefit Cost Analysis (SBCA)

1:30 pm Increasing the Benefits of Air Quality Warnings. Robinson LA*, Buonocore J, Hammitt JK, O'Keeffe L; Harvard University robinson@hsph.harvard.edu

Abstract: Although the U.S. has made significant progress in reducing air pollutant emissions, air quality levels continue at times to exceed the levels deemed protective under the National Ambient Air Quality Standards (NAAQS). When such exceedances are forecast, regional and local AQ authorities issue advisories warning individuals (especially those who are particularly vulnerable) to limit their exposures by decreasing their time spent outdoors and taking other actions. The benefits of these warnings depend on several factors, including the accuracy of the air quality forecasts, the effectiveness of the resulting communications, the relationship between the associated reduction in exposure and health effect incidence, and the monetary value of these changes. This project builds on work to improve the accuracy of air quality models used in these forecasts to estimate the possible changes in exposure and the resulting benefits. It also considers ways to increase the effectiveness of the warnings so as to ensure that those targeted take appropriate actions.

1:50 pm Risk of Bias Analysis of Ozone Epidemiological Studies Used in BenMAP Analyses. Sax SN*, Dell L, Mundt KA, Lewis RJ; Ramboll ssax@ramboll.com

Abstract: The Environmental Benefits Mapping and Analysis Program (BenMAP) is a tool used to estimate the health and economic impacts of air pollution. The US Environmental Protection Agency (EPA) uses BenMAP in regulatory impact and risk assessments. Health impacts of reducing air pollution concentrations are calculated using concentration-response functions (CRFs). The CRFs are derived from epidemiological studies that estimate the relationship between ambient air pollution and adverse health effects. We evaluated the study quality of epidemiological studies that are the basis for preloaded CRFs used in BenMAP to estimate adverse health effects from ambient ozone exposure. We conducted a risk of bias analysis using the Office of Health Assessment and Translation (OHAT) Risk of Bias Tool. Overall, we found the risk of bias was high for the studies that form the basis of these preloaded ozone CRFs. We identified two key sources of bias across all studies, confounding and exposure measurement error. Few studies considered co-pollutants as confounders and no studies considered confounding by lifestyle factors. Exposure measurement error was a concern because all studies relied on central-site monitors as surrogates for personal exposures to ozone. Other sources of bias, specific to air pollution studies were also identified and included the exposure metric evaluated, lag times considered, warm season vs. all-year analyses, and model specification. Confounding and exposure measurement error are potentially serious issues in air pollution studies because of the simultaneous presence of various air pollutants and factors that vary daily (e.g., meteorological variables and lifestyle factors). In addition, the preloaded ozone CRFs are dated, although BenMAP offers users the opportunity to load their own CRFs. We recommend that BenMAP users consider the quality of recent epidemiology studies and use updated CRFs to produce more valid estimates of health impacts of ozone.

2:10 pm “Economic Feasibility” Under the Safe Drinking Water Act: Achieving Efficiency, Equity and Equal Protection. Belzer RB*; Independent Consultant rbbelzer@post.harvard.edu

Abstract: The Safe Drinking Water Act of 1974 directed USEPA (and indirectly, the states) to establish standards for drinking water contaminants that are “economically and technologically feasible.” Congress delegated to the Agency the legislative authority to define these terms. Engineering has held sway over the definition of “technological feasibility,” but economics has played a negligible role in defining “economic feasibility.” The Agency’s definition relies on the principle of “affordability,” which it has defined as 2.5% of median household income. The “affordability” principle suffers serious efficiency, equity and philosophical problems. With respect to efficiency, it considers only costs and ignores benefits. As long as a drinking water standard is deemed “affordable” by the USEPA Administrator (or the official implementing a state drinking water program), it does not matter whether the standard delivers commensurate health-risk reduction benefits, or any benefits at all. Households directly pay for drinking water, and this decision-making rule clashes with how they make all other voluntary choices. With respect to equity, the “affordability” principle disproportionately burdens low-income households for whom the sacrifice of any fixed percentage of income for drinking water is greater. Also, when the domain used for calculating median household income includes a wide range of income groups, the “affordability” threshold is sharply magnified for low-income households. Philosophically, fixed standards regardless of income are said to ensure that all households gain equal protection from drinking water contaminants. This is not obviously so, however, because equal residual risk after treatment is achieved by imposing highly unequal prices for risk reduction. This paper sets forth a specific proposal for solving all three of these problems by defining “economic feasibility” using economics. Both efficiency and equity would be improved dramatically.

more info...

12/9/2018

Economics of Low Probability / High Impact Events, Chair: Deborah Aiken, Co-sponsored by the Society for Benefit Cost Analysis (SBCA)

Economics of Low Probability / High Impact Events

Room: Mardi Gras Ballroom D, 3rd Floor 3:30 pm–5:00 pm

Chair(s): Deborah Aiken deborah.aiken@dot.gov

Sponsored by Economics and Benefits Analysis SG & The Society for Benefit Cost Analysis (SBCA)

3:30 pm Treatment of Risks and Benefits in DHS Security Regulations: Theory and Practice. Gungor A*, Aiken D; U.S. Coast Guard and U.S. Department of Transportation seniorgungor@gmail.com

Abstract: Since the attacks of September 11, 2001 (9/11), federal security regulations and investments have increased dramatically. However, benefit-cost analyses of these regulations continue to pose unique challenges. While estimating the costs of these regulations can be difficult, the treatment of security benefits or avoided consequences of the knowable or past security events in the Department of Homeland Security's (DHS) security regulations proves to be much more problematic. DHS must frequently estimate the consequences of highly unlikely or low-probability, high-impact events, such as terrorist attacks in the process of developing security regulations. By definition, these types of terrorist attacks are rare, and historical data is often limited. For some types of incident categories, a truly large event such as 9/11 may never occur again. Nonetheless, the DHS still must design regulations and policies that account for the potential risks of these rare security events, and attempt to first do a comprehensive risk analysis in order to quantify baseline consequences and avoidable costs in order to ultimately monetize their benefits of security rules. This study will provide an overview of various DHS security regulations (such as Transportation Worker's Identification Card rules, and Western Hemisphere Travel Initiative) to help illustrate the challenges the DHS faces in treating the benefits of security regulations. Specifically, it critically evaluates the status-quo practices of benefit estimation, provides an update on DHS's renewed efforts in considering the use of risk analysis techniques as well as other models and tools.

4:10 pm U.S. Tropical Cyclone Flood Insurance Claim Losses: Storm Surge vs. Freshwater. Tonn GL*, Czajkowski JR; University of Pennsylvania gtonn@wharton.upenn.edu

Abstract: Despite persistent record-breaking flood losses from tropical cyclones (TC), the U.S. continues to be inadequately prepared for TC flood events. In this study, we analyze actual residential flood claim data from the National Flood Insurance Program for 28 significant U.S. landfalling TC related flood events from 2001 to 2014. We illustrate key differences between the numbers of claims, paid claim amounts, and damage for freshwater and surge claims, and evaluate differences in flood claims and damage associated with flood zone, geography, TC event, and flood depth. In addition to descriptive data analysis, we perform statistical analysis to better understand the multivariate relationships between the characteristics of communities and properties with paid claim amounts. The findings provide important insights for flood loss and insurance considerations as well as for the mitigation and management of TC flood risk.

4:30 pm Are Elicited Risk Perceptions Reliable Enough to be Used in Decision-makings? Some Evidence from Residential Customers’ Costs of Large Long-duration Power Outages in Regions Facing Significant Risks. Baik S*, Davis AL, Morgan MG; Carnegie Mellon University sunheeb@andrew.cmu.edu

Abstract: American society depends on electric power for many individual, household, and commercial activities, making our individual and collective vulnerability to large power outages of long duration a key question for policy analysis. Given the proliferation of modern "smart" technology and distributed generation, it would be possible, with some modest additional capability, to provide at least limited service to some customers and sustain critical services when large blackouts occur. A key input to assess whether implementing such ability can be justified is understanding how much individuals and organizations value a small amount of electricity during such an outage. However, there are ongoing controversies about the reliability and validity of estimates from contingent valuation method and risk perceptions among the public. Thus, using such estimates as inputs for further decision-making can be controversial. To improve on such estimates, we have conducted two rounds of surveys with residents in the Northeastern United States and elicited their carefully considered value of reliable electric services in the event of large, long-lasting outages. Using the elicited preferences, we explore whether four factors that are known to influence people’s risk perceptions and their risk preferences - respondent’s previous experience, framing of risks, methods of delivery and beneficiaries, and recruitment strategies- affect the respondents’ preferences for a low-amperage backup service during widespread and long-lasting outages. The results suggest that the respondents were more rational and less influenced by behavioral factors and other factors that have caused controversies surrounding contingent valuation studies. The results highlight the importance of helping people think systematically about, and better refine and articulate their values.