2018 Conference - Session 3

2018 Conference - Session 3

SESSION 3 | Thursday, March 15 | 2:00-3:30PM

A.3: New Developments in Regulatory BCA

Chair: Lisa Robinson, Harvard University, Center for Health Decision Science and Center for Risk Analysis

Trump Administration initiatives pose several challenges for the U.S. regulatory development process and for the conduct of regulatory analysis. While many of these challenges were also faced by past administrations, others are new as a result of the substantially increased focus on deregulation. In this roundtable, a group of experts will discuss related issues, such as anticipating the need for retrospective evaluation when designing new regulations, conducting and using the results of retrospective analyses, and improving prospective analyses of regulatory and deregulatory actions. This session builds on a September 2017 discussion hosted by the Society for Benefit-Cost Analysis, the George Washington University Regulatory Studies Center, and the Administrative Conference of the United States, which can be viewed here. In a related panel, “Experience Conducting BCA to Comply with New Regulatory Directives,” Federal regulatory agency staff will be providing their perspectives on these issues.


Joe Aldy, Harvard Kennedy School

Jennifer Baxter, Industrial Economics, Incorporated

Reeve Bull, Administrative Conference of the United States

Cary Coglianese, University of Pennsylvania School of Law

Jonathan Wiener, Duke University

C.3: Recreation, Resources, and Local Economies

Chair: Trudy Ann Cameron, University of Oregon

The first paper in this session will report upon recent analysis of the accumulating data for 2014, 2015 and 2016, from the National Visitor Use Monitoring Program of the U.S. Forest Service. This program is designed to assess visitation, activity participation, demographics, visit duration, measures of satisfaction, and expenditures related to visits to National Forests and Grasslands. This research offers a window on how these public lands contribute to jobs and incomes in the surrounding area. This session also contributes research that has bearing on the national debate about the proper scope of National Monuments, in light of the distributional consequences that stem from their restrictions on resource exploitation and the tradeoffs between jobs in extractive industries and jobs in tourism. The specific application examines the regional impact of Utah's Grand Staircase-Escalante National Monument as a first example. We will also hear about a strategy to measure willingness to pay, in North Carolina, to protect forest resources from a deadly invasive species, focusing on the Great Smoky Mountain National Park, Pisgah National Forest, and Nantahala National Forest. This study utilizes a survey of households and also explores a number of methodological issues relating to the design and analysis of stated preference surveys.


  1. Issues Affecting Recreational Values of national Forests; Don English, U.S. Forest Service
  1. Neither Boon nor Bane: The Economic effects of a Landscape-Scale National Monument; Paul Jakus, Utah State University
  1. Valuing Hemlock Woody Adelgid Control in Public Forests: Scope Effects in Stated Preference Data with Attribute Non-Attendance; Chris Giguere, Appalachian State University

D.3: Time Keeping in BCA

Chair: Joe Devlin, Environment and Climate Change

Discussant: Charles Griffiths, U.S. Environmental Protection Agency


  1. Institutional Economics and the Cost of Capital; Rick Geddes*, Cornell University; Joshua Goldman

Debate regarding the public- versus private-sector's cost of capital is ongoing since at least the mid-1960s. Arrow and Lind's seminal contribution concluded that the social cost of public-sector capital is lower because project risk is more efficiently spread across numerous taxpayers than across relatively concentrated private investors. That issue has gained renewed vigor due to increasing use of public-private partnerships (PPPs) to finance investment in large infrastructure projects. The debate has been largely devoid of institutional considerations, however. Legal structures such as limited liability and transferability of ownership shares have evolved over time to help manage the cost of private-sector risk bearing, but are either weak or unavailable to taxpayers in their role as a public project's residual risk bearers.Differences in institutional and legal arrangements surrounding taxpayer- versus private-investor risk bearing, and their likely impact on the cost of capital have yet to be systematically considered.We analyze the roles of limited liability and transferability of ownership shares, and their effects on the relative social cost of capital.Our analysis of the arrangements surrounding public -versus private-sector risk bearing suggests that a reassessment of Arrow and Lind's conclusions is warranted.

  1. A Framework for Describing and Evaluating Timeframes for Benefit-Cost Analysis; Joe Devlin, Environment and Climate Change

Guidance on benefit-cost analysis typically offers some advice on choosing an appropriate analytical timeframe, but it may not always be clear what this means. In practice, analytical timeframes may not be long enough to completely capture all incremental impacts of a given regulatory proposal, particularly when the timing of these impacts exceeds the duration of our forecasts. Does this matter? This presentation will present a framework for describing analytical timeframes in terms of regulatory actions and consequences. Using this framework we can evaluate the completeness of any given benefit-cost analysis, consider the degree of sufficiency of incomplete timeframes, and better understand the comparability of analyses. Perhaps most importantly, this framework can help to clarify how the choice of timeframe affects the results of the analysis, which should be of interest to decision-makers. Examples of benefit-cost analysis timeframes and applications of the framework will be presented based on published Regulatory Impact Analysis Statements completed by Environment and Climate Change Canada.

  1. Why Discount Nature Differently?; Arjan Ruijs*, PBL Netherlands Environmental Assessment Agency; Mark Koetse; Gusta Renes; Aart de Zeeuw

In the literature, there is no consensus on how to discount the long term effects of e.g. climate change or declining levels of biodiversity or ecosystem services. We investigate the decision of the Dutch government to apply in their cost-benefit analyses a 1% lower discount rate for effects on ecosystem services than for other effects, except for ecosystem services that are easily substitutable. In this paper, we investigate which ecosystem services should be discounted similar to normal consumption goods, for which ecosystem services a lower discount rate is justifiable, and which reduction is appropriate. For this, following Baumgärtner et al. (2015; Env.&Res.Econ.), we derive the Ramsey-rule for a utility function with both consumption of regular goods and of ecosystem goods and services as arguments. According to this rule, the discount rate depends on the difference of the growth rates of both types of goods and the substitutability between regular goods and ecosystem services in the utility function. We investigate on what grounds the different elements of nature should be discounted differently; is it because nature becomes relatively more scarce or because it is difficult to substitute it for other goods. On the basis of empirical evidence for the Netherlands on trends in the production of ecosystem services and an international review of substitution elasticities, we estimate an appropriate discount rate for nature. Finally, we provide practical and easily applicable policy advice on the discount rates for nature that can be used by cost-benefit practitioners.

E.3 Issues in the BCA of Electricital Power Reforms

Chair: R. Jeffrey Lewis, ExxonMobil Biomedical Sciences, Inc.

Discussant: Tim Brennan, University of Maryland, Baltimore County


  1. Parametric and Non-Parametric Models to Estimate Households and Businesses’ Willingness to Pay for Reliable Electricity; Glenn P. Jenkins, Queens University

Access to reliable electricity for increasing the living standards of households and quality service of businesses is crucial in developing countries. However, Nepal suffers from the worst electricity shortages in South Asia, and only half of the demand for electricity can be met by the nation's grid. This results in load shedding of up to 18 hours a day during the dry winter months, when hydropower generation is low. This research examines the applicability of parametric and nonparametric models to estimate households and businesses' willingness to pay (WTP) for improved reliable electricity services in Nepal by using the cheap talk contingent valuation (CV) data from 1800 households and 590 businesses. The parametric models used is logit model, while the Turnbull estimator and the Kriström mean estimation approach were used for the nonparametric estimation. It is interesting to note that the both households and businesses are willing to pay more to get from a 50% reduction in outages to a complete elimination of outages than they are willing to pay to get from their current situation to a 50% reduction in the incidence of electricity outages. The marginal willingness to pay does not decrease as the quality of the service improves. This differential in the estimates of the WTP for these two options is even more pronounced in the case of businesses than for households. This is to be expected because the disruption and coping costs for businesses will not be alleviated to a significant degree by a partial improvement in the electricity service they receive. The major benefit received by businesses is when they have a continuously reliable supply of electricity that they can count on in planning their business activities. These findings have important implications for an investment strategy improving the reliability of electricity service in Nepal. On the benefit side of the cost-benefit analysis, the marginal benefits are likely to be increasing over the range of improvements until full reliability is achieved.

  1. Application of Benefit-Cost Methods for the Evaluation of the Solar PV Net Energy Metering Values in Different Jurisdictions; Scott Bloomberg*, NERA Economic Consulting; Derya Eryilmaz, NERA Economic Consulting

The declining cost of residential solar PV (along with financial incentives) has led to significant increases in residential solar PV installations in states across the U.S. Although residential solar PV can provide environment and economic benefits to solar net energy metering (NEM) customers, the significant increases in residential solar PV have led states to re-evaluate how these customer-generators are compensated. While solar PV NEM customers generate some of their own electricity, they still are connected to the grid, and therefore, still putting demands on the power distribution system.In addition, current NEM policy in most states allows residential solar PV NEM customers to be paid the full retail rate for their generation, raising justified concern that such rates over compensate NEM customers, thus shifting costs on to non-NEM customers. This "cross-subsidy" effect can be important in many regulatory settings, but this equity issue extends beyond the focus of the benefit-cost analysis.The key question facing regulators and stakeholders is, "What is the proper electricity rate that solar PV NEM customers should receive for their generation?And, how can regulators/utilities determine that matches the value that residential solar PV provides?"In this presentation, we will evaluate the various the benefit-cost approaches that are now being used to determine the appropriate value of solar PV NEM customer.Many analyses to date have been focused in areas with either a favorable location for solar PV (e.g., Arizona) and/or heightened social interest in cleaner electricity (e.g., California).In fact, the net economic value of solar is mostly driven by the regional differences and our research demonstrates that simple off-the-shelf approaches may not accurately quantify the benefits and costs of solar PV NEM. This research compares and contrasts the benefit-cost analysis for two different types of jurisdictions as a case study to demonstrate the differences in net benefits of solar:1) high cost electricity/high solar PV penetration, and 2) low cost electricity/low solar PV penetration.The presentation will also highlight the importance of accounting for the synergies between solar and rapidly-developing technologies, such as electricity storage and advanced inverter technologies (and standards) in order to accurately calculate the benefits and costs of solar in different jurisdictions.As these technologies become more prevalent, the economics of renewable integration and the methods to calculate the net benefits are constantly evolving. This study has an objective to address this issue.The analysis will highlight how the benefit-cost analysis will provide potentially very different values depending on the characteristics of region being analyzed.The analysis will also demonstrate that the analyst must keep an open mind since some variables that may be benefits in one region could turn out to be costs in another.

  1. Cost Benefit Analysis of Power Sector Reform in Haiti; Bahman Kashi, Queens University; Juan Belt*, Center for Strategic and International Studies (CSIS); Nicolas Allien; Jay Mackinnon

This paper argues that to improve the power sector in Haiti, which now constitutes a critical constraint to economic growth, it would be necessary to carry out a significant regulatory and utility governance reform; without these reforms, any physical investment program would be ineffective and unsustainable. Haiti has the most underdeveloped and inefficient power sector in the Americas. Numerous past attempts to reform it have failed due to lack of political will. In this paper, we consider a multi-phase program of reform and assess its feasibility. In the first phase, the Government of Haiti (GOH) carries the corporatization of units of Electricité d'Haïti (EDH); establishes cost-reflective tariffs; and introduces management contracts, leases, and concessions, and privatizes EDH units as appropriate. If the first phase succeeds we propose proceeding to later phases that would support EDH. Costs have been estimated based on a similar program implemented in Afghanistan by the United States Agency for International Development (USAID). Our estimation of economic benefits is based on a projected reduction in technical losses, valued at the retail price of electricity for average consumers; net gains in consumer surplus resulting from servicing high value customers are excluded from the model due to lack of reliable data to support a quantitative estimate. Furthermore, the analysis is conducted based on a 50% chance of success for the reform. At these conservative measures of costs and benefits, the project is found to be economically and financially viable (Economic IRR: 15%, Economic NPV: 11 Million 2017 USD, Financial IRR: 28%, Financial NPV: 78 Million 2017 USD).

F.3: BCA of Safety: Cops, Guns, and Maritime Accidents

Chair: Glenn Blomquist, University of Kentucky

Discussant: Mark Dickie, University of Central Florida


  1. Good Cop, Bad Cop: An Analysis of Chicago Civilian Allegations of Police Misconduct; Kyle Rozema*, University of Chicago Law School; Max Schanzenbach

Police officers are tasked with a very complicated undertaking: they must make arrests and prevent crimes without violating the rights of civilians, under subjective standards such as "reasonable force" and "probable cause." Many of their decisions are imbued with a great deal of discretion, which gives rise to opportunities for abuse of power. Those who supervise police officers also confront a complicated undertaking: they must prevent police officer misconduct, while still giving police officers the incentives and flexibility to fight crime. Detection of police officer misconduct by supervisors is difficult because it is costly to monitor police interactions with civilians. To facilitate detection of police misconduct, most large police departments have an administrative process through which civilians can bring allegations of police misconduct.There is little empirical evidence on the viability of civilian allegations to predict police misconduct. Police officer organizations and some criminal justice scholars have questioned the use of civilian allegations, arguing that the rate at which officers receive civilian allegations largely reflects officer productivity (Worden et al., 2012; Lersch, 2002).This article assesses the potential for civilian allegations to predict police officer misconduct using recently released data on over 50,000 civilian allegations of police officer misconduct in Chicago. We use empirical Bayes estimation to construct a "shrunken" measure of officer-level civilian allegations that (1) controls for officer assignment and officer characteristics, and (2) accounts for the reliability of the allegations by shrinking noisier estimates toward zero. This approach has been utilized in other settings, including the teacher value-added literature (e.g., Kane and Staiger, 2008; Chetty et al., 2014, 2015; Bacher-Hicks, et al., 2014).We then test the power of shrunken civilian allegations to predict serious misconduct as measured by civil rights litigation. Because of attorney incentives stemming from contingency fee arrangements and the legal obstacles to civil rights litigation, civil rights litigation filters for the most serious incidents of officer misconduct. We find a strong non-linear relationship between shrunken allegations and future civil rights litigation. For the officers in the bottom 80 percent of shrunken allegations, shrunken allegations are unrelated to litigation. The relationship between shrunken allegations and litigation spikes for worst 5% of officers and, when we consider damages, spikes further for the worst 1% of officers. This non-linear relationship suggests that intervention efforts could be fruitfully concentrated among a relatively small group of officers. Moreover, it suggests that officers with a moderate number of risk-adjusted allegations are at no greater risk of committing serious misconduct than officers who receive no allegations.In a final analysis, we present evidence on the policy debate regarding the requirement that civilians swear an affidavit as part of the investigation process. An affidavit is a sworn statement and those swearing are warned of the consequences of perjury. Roughly 55% of allegations are dismissed for failure to swear an affidavit. We find that allegations lacking an affidavit have the same predictive power as affidavit-based allegations. These findings raise concerns about the value of affidavit requirements.

  1. Benefits and Costs of Confiscating Firearms from Prohibited Persons in California; Chris Mai, Vera Institute

Gun violence continues to be a problem that vexes policymakers at all levels of government. One policy idea that has been tested by both the federal and state governments as a strategy to reduce gun violence is to ensure that certain people, such as those with felony convictions and those with domestic violence misdemeanor convictions, are unable to purchase guns. Federal law ensures that a set of "prohibited" persons are unable to purchase a firearm through a licensed dealer, and several states expand on these laws. But California is the only state that has created an enforcement regime"”the Armed and Prohibited Persons System (APPS)"”to remove firearms from people who purchased them legally but became prohibited from firearm ownership at a later date. Enforcing APPS involves both investigation of state databases to confirm that an individual is both "armed" and "prohibited" and visits to APPS homes to retrieve unlawful firearms.This research will investigate the full range of costs and benefits resulting from APPS. The most obvious potential benefit is a reduction in gun violence by removing firearms from people who are legally prohibited from ownership and who may be at higher risk to commit violence against themselves or others. The evaluation will track the extent to which the intervention reduces gun violence, including gun suicides, in the neighborhoods in which it is implemented, compared to a control group of neighborhoods. The costs include the expense to pay for these enforcement efforts, including personnel costs and travel costs to perform the visits. But there is also an array of intangible costs as well. One potential cost of the intervention is that APPS households may experience harm, such as fear, intimation, or stigma from multiple police officers visiting their home and potentially entering to conduct a search. APPS individuals and households also face a cost from losing access to their firearms, although such access is illegal.This session will present data from a range of sources including the budgetary cost of APPS, extant literature on the cost of gun violence, and research on the impact of police visits to create an original breakeven analysis. Costs and benefits will be monetized, though in practice the monetary valuations of some impacts, such as the police visits, will be speculative. This intervention is unique to California and the evaluation is the first of its kind to investigate the potential reduction in violence and costs of this new type of policy.

  1. Using Near Miss Data to Characterize Baseline Risk in Maritime Regulatory Cost-Benefit Analysis; Kimberly Wilson, U.S. Coast Guard; Ali Gungor, U.S. Coast Guard; Jeffrey Horn, U.S. Coast Guard

To estimate and quantify the benefits of its regulations the U.S. Coast Guard currently relies upon historic maritime incident data, which provides information on incidents involving injuries, fatalities, property damage, and damage to the environment. However, currently Coast Guard benefit analyses do not capture the entire span of risks, including a class of small failures (or near miss incidents) in everyday maritime operations that are unremarked, unreported, and ultimately unexamined. Capturing near miss data highlights unsafe work practices, safety rule violations, unsafe work practices, and potential issues with an entity or corporate safety culture. In addition, near misses are often harbingers of low-probability high-consequence disasters. A better understanding of near misses may shed light on how to avert a catastrophic event, and may also allow the Coast Guard to better capture the baseline risk associated with certain activities in our regulatory analyses. Capturing this baseline risk is becoming increasing important in Coast Guard benefit analyses, as many regulations or policies seek to prevent or mitigate damage from worst case events that may or may not have occurred.The proposed presentation will provide an overview of Coast Guard's work developing a methodology to incorporate near misses into our regulatory analyses. The presentation will provide a summary of how near misses are addressed in a regulatory context, how the Coast Guard addressed issues such defining a near miss and lack of data, and a summary of the approaches currently being developed by the Coast Guard.

G.3: Valuing Health for Applications in BCA

Chair: Sue Hamann, National Institute of Health

Discussant: Elizabeth Quinn, U.S. Food and Drug Administration


  1. Valuing Non-Fatal health Risks: Theory and Empirical Evidence; Daniel Herrera Araujo*, Paris School of Economics; James K. Hammitt; Christoph Rheinberger.

We propose a novel approach to estimating the monetary value of a change in health risk. We start from a general utility function for health, longevity, and wealth and derive theoretically valid lower and upper bounds on the willingness-to-pay for a quality-adjusted life year (QALY). The upper bound matches exactly the value per QALY derived by Hammitt (2013) under the assumption that preferences for health and longevity are consistent with any health-adjusted life year measure. This value is likely to change with current health, wealth, and longevity. We empirically assess the theoretical upper bound derived from the model drawing on a rich stated-preference data set from France. We find an upper bound on the value per QALY gain of approximately euros 90,000, but the value depends on the individual's endowment of health, wealth, and remaining life expectancy.

  1. Health Insurance Price Externalities; Daniel Wilmoth, U.S. Small Business Administration

Analyses of economic efficiency often focus on interactions outside of the market, such as the negative externalities created by pollution.However, interactions between consumers through prices can also affect efficiency.Health insurers often pay prices that are linked to those paid by other purchasers.For example, Medicaid typically receives a rebate of 23.1 percent of the average manufacturer price for brand name prescription drugs. Linked pricing can affect economic efficiency by changing the prices paid by other purchasers.The insured are insensitive to prices, and healthcare providers with market power can increase the linked prices paid for the insured by changing prices in the remaining market.The effects of those changes on deadweight loss can be estimated using basic economic theory and values like quantity, price, and profit margin.A method of estimating effects on deadweight loss is illustrated using the Medicaid price linkage for brand name prescription drugs.

  1. A BCA of Hearing Aids to Reduce the Symptoms of Dementia; Robert J. Brent, Fordham University

Research Question:Dementia is a term used to describe various symptoms of cognitive decline, involving memory, language and thinking that are severe enough to affect daily activities. Global costs of dementia were US$ 818 billion in 2015. Given the prevalence and costs of this disease, it is important that interventions for dementia be identified and evaluated using benefit-cost analysis (BCA) to assess whether they are socially worthwhile. We test to see whether hearing aids reduce dementia symptoms and, if so, estimate whether they are socially worthwhile.The Analysis:In this study we carry out a BCA of hearing aids as a method for reducing the symptoms of dementia.The benefits of hearing aids are obtained by estimating the direct and indirect effects (working through dementia symptoms reductions) of hearing aids on utility and valuing this by the price of a life year attached to the number of life years that an individual has remaining in a particular health state. We provide a novel method for deriving the price of a life year from the value of a statistical life literature.There are two estimation equations: one for estimating the effect of hearing aids on dementia symptoms and a second one for estimating the effect of hearing aids and dementia on utility. We use the Clinical Dementia Rating (CDR) Scale to measure dementia severity and the Geriatric Depression Scale (GDS) to measure utility.The Data and EstimationOur data comes from the National Alzheimer Coordinating Center (NACC) panel of 118,00 patient visits, made up of an average of 3.2 visits over 13 years for over 30,000 participants at 32 US Alzheimer's Disease Centers over the period September 2005 and May 2017.We use a fixed effects model for estimation that controls for time invariant individual variables and test this against a random effects estimator.Main Findings.Hearing aids reduce the symptoms of dementia by about one point (on a scale of 1 to 18). Our methods produce a price of a QALY of $442,000. As hearing aids directly and indirectly increase the utility of a life year by 0.02, the benefits of a hearing aid are $8,840 in 2001 prices. With the cost of a hearing aid being around $3,000, including follow up and maintenance, the conclusion is that hearing aids have a benefit-cost ratio close to 3 and therefore are socially worthwhile and should be covered by Medicare.

  1. Public Safety Under Imperfect Taxation; Yuting Yang*, Toulouse School of Economics; Nicolas Treich, Toulouse School of Economics

In this paper, we examine theoretically the effect of tax system imperfections on the optimal level of public safety. We compare three exogenous tax systems, namely first-best, uniform and income tax. Moreover, we consider several sources of imperfections, namely individuals' heterogeneity in wealth and in risk exposure, and labor supply distortion. We show that the effect of imperfect taxation critically depends on the source of imperfection as well as on the utility and on survival probability functions. Hence, imperfect taxation cannot generically justify more or less public safety. There is thus no fundamental reason to always adjust downwards the value of statistical life (VSL) because of imperfect taxation, nor to assume a marginal cost of public funds systematically greater than one for the benefit-cost analysis of public safety projects.