2016 Conference - Session 6

Session 6 - Friday, March 18, 10:45am - 12:15pm

A.6: The Energy Paradox

Chair: Art Fraas, Resources for the Future

Discussant: Timothy Brennan, UMBC and Resources for the Future

Presentations:

1.     The Energy Efficiency Paradox: Evidence from Three Industry Case Studies, Ann Wolverton,* Heather Klemick and Elisabeth Kopits, U.S. Environmental Protection Agency

Economic theory suggests that profit maximizing firms should have an incentive to incorporate technologies into their products that are cost-effective, absent consideration of environmental externalities. Even in the presence of uncertainty and imperfect information – conditions that hold to some degree in every market – firms are expected to make decisions that are in the best interest of the company owners and/or shareholders. However, simple net present value calculations comparing upfront costs of fuel-saving technologies to future savings suggest this is not always the case. This puzzle has been observed in a variety of contexts and is commonly referred to as the “energy efficiency paradox.” A growing number of empirical studies in the peer-reviewed literature examine why households may under-invest in energy efficiency. To our knowledge, far fewer studies examine whether similar undervaluation occurs on the part of businesses. While a variety of hypotheses could explain this behavior, lack of empirical evidence on why businesses do not always invest in seemingly cost-effective energy saving technologies limits our ability to judge whether and when a given hypothesis is likely to be valid. We investigate capital investment decisions in three different industry contexts - heavy duty trucking, supermarket refrigeration, and data center investments – using a combination of focus groups and interviews. Consistent with the economics literature, in each case we distinguish between market failures, behavioral anomalies, and other factors not accounted for in typical net present value or payback calculations for energy efficient technologies. We then discuss key similarities and differences across the three case studies with regard to the way in which investment are made and the evidence we find of an energy efficiency paradox.

2.     Searching for Hidden Costs: A Technology-Based Approach to the Energy Efficiency Gap in Light-Duty Vehicles, Gloria Helfand,* U.S. Environmental Protection Agency; Jean-Marie Revelt; Lawrence Reichle; Kevin Bolon; Michael McWilliams; Mandy Sha; Amanda Smith; and Robert Beach

The benefit-cost analysis of standards to reduce vehicle greenhouse gas emissions and improve fuel economy by the U.S. Environmental Protection Agency and Department of Transportation displayed large net benefits from fuel savings for new vehicle buyers. This finding pointed to an energy efficiency gap: the amount of energy-saving technology provided in private markets appeared not to include all the technologies that produce net private benefits. The finding of a gap involves three pathways. First, the energy-saving technologies must be effective in achieving fuel reductions. Second, the cost estimates for those technologies must be lower than the present value of fuel reductions. Third, possible “hidden costs” -- undesirable aspects of the new technologies – must not exceed the net financial benefits. This study examines the existence of hidden costs in energy-saving technologies through a content analysis of auto reviews of model-year 2014 vehicles.

Content analysis involves systematic identification in texts of key concepts and coding of those concepts; it makes qualitative assessments available for quantitative analysis. Auto reviewers, as professional evaluators, are likely to be sensitive to the existence of positive and negative characteristics of vehicles. It is unlikely that they would miss important problems, although they may identify negative characteristics that some vehicle owners may not notice.

Results suggest that it is possible to use fuel-saving technologies on vehicles without imposing hidden costs. For each of the technologies examined, the number of reviews that evaluated them positively exceeded the number that spoke negatively. There is scant evidence of a robust relationship between the technologies and vehicles’ operational characteristics, such as handling or acceleration. It seems possible to implement these technologies without adverse effects on vehicle quality; hidden costs do not appear to explain the efficiency gap for vehicle fuel-saving technologies.

3.     Regulating Use of Energy-Saving Technologies: The Case of Aerodynamic Devices on Heavy-Duty Trucks, Randall Lutter,* Batten School, University of Virginia; Art Fraas; Zach Porter; and Alex Wallace

In a 2015 proposal to require heavy-duty vehicle manufacturers to use energy-saving technologies, the US Environmental Protection Agency estimated that the value of such savings greatly exceeds the cost of achieving them. This finding raises questions about cost-minimization in competitive industries.

To address these questions, we collected and analyzed data on aerodynamic energy-saving devices on more than 200 trucks operating on U.S. interstate highways during the summer of 2015. We hypothesize that device use increases with miles per vehicle per year, fleet size, and proximity to California, which has mandated devices on trailers since 2013. We also hypothesize that firms operating at the margin because of limited access to capital markets and/ or various management issues, etc., use aero devices less often. We develop a measure of such issues by constructing an index of noncompliance with federal requirements for hours of service and vehicle maintenance. Using these variables, we model use of aerodynamic devices.

Alcott and Greenstone (2012), Klemick et al., (2015) and EPA (2015) have suggested that adoption of energy-saving technologies may be hindered if owners of capital equipment where such technology is deployed are different from the entities that would enjoy the benefits of such technology. Accordingly, we test whether aerodynamic energy saving devices are less common on trailers towed by tractors with different owners. We find scant evidence of such an effect.

Our findings suggest market failures associated with energy conservation technology may be more limited than claimed by EPA for the trucking industry. More broadly, claims that issues like ownership differences in competitive markets interfere with the adoption of energy-savings technologies merit careful scrutiny. We make practical suggestions about how to conduct economic analysis in instances where benefits to users of new technologies seem to greatly outweigh the costs.

B.6: VSL and Risk Preferences in Public Health and Safety

Chair: Rene Pana-Cryan, National Institute for Occupational Safety and Health

Presentations:

1.     Eliciting Risk Aversion in the Context Of Health, Rebecca McDonald,* University of Warwick; Susan Chilton; Michael Jones-Lee; and Hugh Metcalf

We present a conceptual framework for the elicitation of risk preferences from choices between lotteries whose outcomes are health states. To demonstrate the potential application of this framework in a survey setting, a specific procedure for eliciting health risk preferences is developed and tested in a survey (n=112) and shown to be straightforward to implement. Financial Coefficients of Relative Risk Aversion are also elicited, and correlation between these and the health risk preference measures is shown to be positive and significant, but low in practical terms, casting doubt on the domain generality of risk preferences and strengthening the case for a domain-specific alternative based within our conceptual framework. The specific interpretation of any such risk preference measure in health is also considered. Because there exists no unique, interval or ratio-scale cardinal measure of health, values or utilities are used to measure the health states over which gambles are defined. We show that if the value and/or utility functions for health are non-linear, then the elicited health risk preference coefficient provides a measure of probabilistic risk aversion (if values are used), or risk aversion relative to the average member of the population (if population average utility scores are used).

The implications of our work for BCA and policy are as follows. When individuals’ preferences over health states and fatality risks are elicited, our techniques tend to use risks as the item under valuation (in traditional WTP-based VSL elicitations) or as the response mechanism (in Standard Gamble or Risk-Risk studies). Risk preferences over health and safety are likely to influence the responses in such studies, and subsequent allocation recommendations might be influenced by the level of health-specific risk aversion of the respondents. Understanding the nature of these risk preferences will help to improve the robustness of our policy recommendations.

2.     What Is a Life Year Worth? Exploring the Methodology and Assumptions Behind The Full Income Approach, Angela Chang,* Harvard School of Public Health; Lisa A. Robinson; James K. Hammitt; and Stephen Resch

Background: The 2013 Lancet Commission on Investing in Health (CIH) estimates the value of improved health using a full income approach, adding the value of increased life expectancy to the value of predicted growth in gross domestic product (GDP). This approach captures the intrinsic value of health as well as its effect on economic production. The CIH finds that the value of an additional life year (VLY) averaged 2·3 times GDP per capita in low- and middle-income countries (LMICs), given the increase in life expectancy from 2000 to 2011. Examining related uncertainties provides insights into these findings, as well as options for applying the estimates in other contexts.

Methods: We investigate the sensitivity of the VLY estimates to the underlying assumptions, incorporating recent research, exploring alternative characterizations of the affected population, and examining the sequencing of the calculations. Our analysis addresses the VLY’s relationship to income, age, and life expectancy as well as the effects of adjusting the results for particular age groups.

Results: We find that the VLY estimates are particularly sensitive to the assumptions regarding the effects of income and age-specific survival rates; reasonable alternative assumptions may reduce the estimates significantly. However, the CIH also adjusts the values for young children downwards; eliminating this adjustment increases the estimates. These estimates reflect a specific shift in population life expectancy and may underestimate the value of this shift particularly when health improvements disproportionately accrue at older ages.

Conclusion: Given the lack of primary research on the value that LMIC populations place on a year of life extension, these values must be extrapolated from available research. The CIH develops one such approach. When applying this approach elsewhere, care must be taken to tailor the estimates to the impacts of the intervention and the affected population and to appropriately characterize uncertainty.

3.     Valuing Quality-Adjusted Life Years for Benefit-Cost Analysis, Lisa A. Robinson* and James K. Hammitt; Harvard Centers for Risk Analysis & Health Decision Science

Benefit-cost analysis plays an important role in informing regulatory and other policy decisions, by providing information on how those affected value the benefits they receive in comparison to the costs the policy imposes. However, the usefulness of these analyses is currently hindered by the lack of willingness to pay (WTP) estimates for nonfatal health conditions. As a result, analysts often rely on estimates of quality-adjusted life years (QALYs), valued using a constant WTP per QALY, as a rough proxy. Both theory and empirical research suggest that that this approach is inconsistent with individual preferences: the value per QALY is likely to vary depending on the severity and duration of the condition as well as other characteristics of the risk and the affected individual. Several studies are now available that provide estimates of WTP per QALY for various health conditions. We combine the results of these studies to develop a function that can be used to estimate WTP per QALY, which may depend on the size of the gain. We find that this approach is promising but yields uncertain estimates given the limitations of the available research. Our research has implications for the values used as cost-effectiveness thresholds as well as for benefit-cost analysis, suggesting that these thresholds should be varied for different types of health conditions.

4.     Evaluation of the Distribution of VSL Values by Combining New Vehicle Safety Estimates With a Model of Vehicle Choice, Damien Sheehan-Connor,* Wesleyan University

Many of the studies estimating the value of statistical life (VSL) use labor market estimates that may apply best to a subset of the population that is relatively homogeneous in terms of income and other characteristics. Since many households choose to own automobiles, the safety implications of this choice can be used to estimate the distribution of VSLs and its correlation with income and other demographic variables using broad support in the explanatory variables. A recently developed model of automotive safety (Economic Inquiry 53(3): 1606-29) uses Fatality Analysis Reporting System (FARS) data to estimate the level of safety of vehicles at the model by model year level. Specifically, the probability of someone dying in a particular vehicle over the course of a year is calculated. This probability is a complex function of vehicle weight, class, manufacturer, vehicle age, mean number of vehicle occupants, age of vehicle occupants, and number of miles driven in a year. The safety estimates are combined with information on vehicle costs and data from the National Household Travel Survey (NHTS) to calculate the marginal cost of saving a statistical life for each vehicle type for a particular household. The NHTS provides data from a random selection of households about household characteristics, the vehicle(s) owned by those households, and the way in which these vehicles are used. The safety cost variable is included in a regression model of vehicle choice with a rich set of control variables to impute a VSL for each household in the NHTS. The resulting distribution of VSLs provides evidence about the variation in risk preferences within the population and the correlation of these preferences with income and other demographic variables of interest.

C.6: Methods for Estimating Benefits

Chair: Linda Abbott, US Department of Agriculture

Presentations:

1.     Exploring the Accuracy of Traffic-Noise Benefits Transfers, Henrik Andersson,* Toulouse School of Economics; Jan-Erik Swärdh, VTI, Sweden; and Mikael Ögren, University of Gothenburg, Sweden

One of the most common approaches to conducting benefit-cost analysis (BCA) is to rely on benefits transfers (BT). A benefit transfer takes already existing values from a study case (site) and creates a benefit estimate for a policy case (site) and thereby allows for a BCA to be conducted when direct estimates at the policy case are too time-consuming or expensive to be directly estimated. A recent review of BT studies (Kaul et al., JEEM, 2013) found that only one BT study had been conducted using the hedonic pricing technique. Since the hedonic pricing technique is one of the most influential non-market valuation techniques it is of interest to examine how well this technique works for BT. This study aims to provide evidence of how well the technique performs for BT by using a very rich Swedish data set on property prices and traffic noise. The findings suggest that:

1. The errors in estimates from adopting a BT approach can be significant. For instance the naïve BT in most cases produces non-negligible transfer errors (TE) (> 40%).

2. There is a large variation in how well BT works depending on from which study case values are transferred to which policy case.

3. BT adjusted based on income and the usage of benefit-transfer functions can reduce TE, but not systematically (i.e. TE may also be larger when using more sophisticated methods than the naïve BT method).

Overall, the findings from this study suggest that BT transfers based on estimates from the hedonic regression technique can result in large errors. Hence, policy makers should be cautious when using values from hedonic pricing studies.

2.     Economic Analyses of Benefits And Costs of USDA Conservation Programs: What We Can Do Better? LeRoy Hansen,* USDA Economic Research Service

USDA conservation program expenditures in 2014 were an estimated $5.5 billion. Despite the size of the expenditure, there are few studies that have attempted to estimate the benefits and costs of USDA programs. Economists and other scientists have employed a variety of innovative analytic techniques that have allowed some benefits and costs of some programs in some parts of the country to be estimated.

There are very challenging aspects of USDA conservation program benefit-cost analyses that are probably limiting progress, including 1) a program’s effectiveness at initiating conservation practices differs across the country, 2) the ecological effects of practices vary, and 3) social values of marginal changes in environmental amenities differ spatially. Future benefit-cost analyses can build on the past. Additionally, future analyses will produce more policy-relevant results by incorporating methods and techniques discussed here. The first objective of this paper is to specify and verify the available refinements. Two examples of refinements are 1) results of future research will better support program decision making by following federal guidelines for carbon sequestration benefit analyses and 2) benefits transfer will provide more reliable results when analysts apply (or transfer) marginal-value (not average-value) estimates to marginal changes in environmental amenities. The second objective is to generate estimates of the economic effects that the proposed methods might have.

3.     Lessons in Applied Benefit Transfer Using Meta-Analysis, Patrick Walsh,* Julie Hewitt, Steve Newbold and Matt Massey; US Environmental Protection Agency

This study explores several important issues in the use of meta-analysis in an applied benefit transfer. Although there are many studies on best practices of both conducting a meta-analysis and performing a benefit transfer, there are far less studies on the best practices for using a meta-analysis as a benefit transfer function. On the other hand, federal rules are increasingly using meta-analyses for benefit transfers, especially at the U.S. Environmental Protection Agency (EPA). In order to highlight several important theoretical and empirical issues in using a meta-analysis for benefit transfer, this paper employs data from a past EPA rule. The focus of the meta-analysis is water quality benefits and EPA’s 2003 CAFO rule is used as the example.

The meta-analysis is based on a meta-dataset of 51 stated preference studies, published between 1985 and 2011. Each of these studies used a stated preference approach to elicit survey respondents’ willingness to pay for water quality improvements. We look at several variations in the construction of the meta-analysis function, based on differences in theoretical and empirical assumptions. For example, although Diamond’s “adding-up” property has received a fair bit of attention in the stated preference literature, it has received very little attention in the meta-analysis and benefits transfer literatures, and violations of it can imply strange results. In the context of water quality, a meta-analysis function that violates this property (which most previous meta-analyses do), implies that the WTP for several small rules would be higher than one rule that accomplishes the same change. After applying our functional form variations to the CAFO rule data, our results indicate that the variations in the meta-analysis function explore yield significantly different benefit estimates, and have several important implications for future applications.

4. Cost-Benefit Analysis in the Social Sector, Bahman Kashi,* Queen’s University; Zuzanna Kurzawa, University of British Columbia; and Josh Folkema, World Vision Canada

For a long time, players in the social sector have debated over the use of quantitative tools for measuring the impact of social programs. Advocates maintain that it would facilitate higher levels of professionalism and better use of funds. Opponents however argue that such tools are unsuitable for the social sector. While some newly developed quantitative tools, such as SROI, have gained momentum in recent years, their usefulness for decision making has been subject to criticism. This study has two objectives. The first is to highlight fundamental issues that must be addressed before any quantitative methodology can be effectively applied in the social sector. The second is to provide an abstract framework to overcome such challenges.

D.6: Evaluating the Effects of Regulation on Small Businesses: Practitioner Perspectives (Roundtable Discussion)

Chairs: Patrick Delehanty and Lindsay Scherber, US Small Business Administration

Panelists to Include:

1. Alexei Alexandrov, Consumer Financial Protection Bureau

2. Thomas Henry, U.S. Food and Drug Administration

3. Jonathan Porat, U.S. Small Business Administration

4. Amanda Thomas, Office of Management and Budget

Under Executive Orders 12866 and 13563, and OMB Circular A-4, federal agencies are required to develop regulatory impact analyses (RIAs) for all economically significant rules. As part of the rulemaking process, agencies are also required to evaluate small business impacts under the Regulatory Flexibility Act (RFA), including the identification of compliance costs and less burdensome alternatives.

Featuring economists from across the federal government, this session will highlight how regulatory agencies can integrate small business impact analyses into their broader RIA efforts. As BCA practitioners, panelists will share their strategies for locating high quality small business data, estimating small business impacts during the RIA development stage, and incorporating small business concerns into their broader regulatory analyses and proposals. They will also discuss methods for overcoming analytical challenges unique to the industries they regulate.

E.6: Costs and Benefits of Social Investments

Chair: Lynn Karoly, RAND Corporation

Presentations:

1.     Reasonable Accommodation and Sheltered Workshops for People with Disabilities: Costs and Returns of Investments, Gareth Harper,* Optimity Advisors; Rory Tierney; and Quentin Liger

We conducted an economic analysis in support of an assessment for the European Parliament on programs intended to help disabled people find or remain in employment. As well as synthesizing existing economic literature from the U.S. and Europe and assessing national-level policy initiatives, we conducted cost-benefit analyses of two relevant programs: a Lithuanian program to help people with hearing disabilities find work through the use of recruitment agents who could communicate in sign language, and a Hungarian sheltered workshop for mentally disabled individuals.

Data on costs and effectiveness of these programs was collected through interviews and effects were valued using publicly-available wage, tax and benefits data for each country. Although these data were not complete, by making conservative assumptions around entry into employment, length of employment, and earnings, and by conducting break-even analysis for the Lithuanian program, we were able to calculate benefit-cost ratios for each program from the societal and government perspectives, and net benefit to the participants themselves.

The project provides a good example of how benefit-cost analysis can be used to provide a pragmatic assessment, even when data are limited, and help support recommendations to policymakers (namely that facilitating entry into open labor market employment is likely to be cost-beneficial, while traditional sheltered employment is likely only valuable for those whose disabilities are severe enough that open labor market employment is not possible). Future research into the intangible benefits of employment for disabled individuals, which we were not able to quantify, would provide a useful addition to cost-benefit analysis in this area.

2.     Money Talks: Applying Cost-Benefit Analysis to Policies Combatting Intimate Partner Violence, Nicholas Mastron,* The George Washington University

Gender-based violence is increasingly seen as a problem within America. However, policies combatting this violence often evoke opposing visceral reactions amongst policymakers and citizens alike. Studies thus far primarily employ discourse analysis to assess the ideological themes present. However, this paper argues that cost-benefit analysis (CBA) could theoretically yield a far more comprehensive evaluation of gender violence. Given that gender violence represents a broad range of crimes, the study focuses specifically on heterosexual intimate partner violence (IPV). Therefore, the paper seeks to first examine the existing federal legal framework and the rationales justifying IPV policy creation; then to articulate the applicability of CBA in evaluating IPV policies; and finally to apply its findings to current CBA practices in the IPV field.

Federal regulatory foundations construct the justification basis of enacting IPV policies. The vast majority of pertinent federal rules historically fall under the legal requirement in Executive Order 12866. Regulations such as the Violence Against Women Act require establishing specific female employee protections (i.e. wage equality, discrimination adjudication, etc.) and these protections act as deterrents of IPV. However, E.O. 12866 also supplies an alternative administrative interpretation, specifically providing that “other compelling public need” may justify regulation. Ultimately, this has seen far less IPV application due to its broad potential interpretations.

Next, this paper asserts that cost-benefit analysis can help establish these “public need” interpretations for IPV regulations and also evaluate pre-existing policies. CBA’s use in measuring IPV can be framed by the human capital argument; employees suffering abuse directly correlate with lower productivity outputs, thereby lower revenues. Extending upon productivity concerns, the value of efficiency over time also produces substantive divergence in social cost functions in IPV cost-benefit analyses. Finally, the existing IPV cost models are analyzed using CBA, and these results are summarized.

3.     Can the Ticket to Work Program Be Self-Financing? Craig Thornton,* Mathematica Policy Research

The evaluation of the Social Security Administration’s Ticket to Work (TTW) program provides a convenient way of examining how benefit cost methods can be used to address several challenging issues. The first challenge stemmed from the program’s very nature. It functioned less as a specific program and more as an effort to stimulate the market for employment services in order to help beneficiaries find jobs and become economically self-sufficient. Second, key decision makers in the SSA Office of the Actuary favored analytical approaches that differed from what economists and many other benefit-cost analysts tend to do. Finally, the program was implemented in a way that essentially precluded accurate estimation of program impacts. To address these challenges, the analysis used the available cost information to compute how big impacts would have to be in order to generate net benefits to SSA. It then used the available data and literature to make an assessment of whether impacts that big were at all plausible. The bottom line is that the program is relatively inexpensive and would only need to move a few thousand (out of 10 million) beneficiaries into substantial employment each year to generate net benefits measured just from the government budget perspective. While such an impact is not assured, it was certainly plausible. This approach could be useful in other cases where we lack the impact estimates we would like to have.

4.    The Costs and Benefits of Recycling in New York City, Ken Acks,* Cost-Benefit Group, LLC

On April 22, 2015, NYC Mayor de Blasio declared that by 2030 the city would no longer send any garbage to landfills. New York would join San Francisco, Seattle, and other cities in moving toward a “zero waste” policy.

On October 3, 2015, in “The Reign of Recycling,” the fifth most emailed New York Times article over the past 30 days, John Tierney, reprises a 1996 article and argues that recycling was, and is, wasteful when it comes to the bottom line, both economically and environmentally.

Tierney claims that despite decades of exhortations and mandates, it is still typically more expensive for municipalities to recycle household waste than to send it to a landfill. Prices for recyclables have plummeted; the national rate of recycling has stagnated in recent years and as cities move beyond recycling paper and metals, and into glass, food scraps and assorted plastics, the costs rise sharply while the environmental benefits decline and sometimes vanish. He claims that to offset the greenhouse impact of one passenger’s round-trip coach flight between New York and London, you’d have to recycle roughly 40,000 plastic bottles without counting costs of rinsing. He also claims that all the trash generated by Americans for the next 1,000 years would fit on 0.1 percent of its grazing land, and that landfills are typically covered with grass and converted to parkland, such as the Freshkills Park on Staten Island, once at capacity. Furthermore, washing plastic in water heated by coal-derived electricity results in a net increase in CO2; some landfill operators capture methane for electricity and; modern water-to-electricity incinerators release so few pollutants that they’ve been widely accepted in eco-conscious Europe and Japan.

This paper will utilize a variety of sources to estimate total costs of recycling and alternatives now and in 2030. Costs/Benefits include CO2, methane, air pollution, traffic congestion, water use, energy, transportation, land use, and materials.