Economics, Public Health and the Environment: State of the (Applied) Science

Size: px

Start display at page:

Download "Economics, Public Health and the Environment: State of the (Applied) Science"

Patrick Day
6 years ago
Views:

1 Economics, Public Health and the Environment: State of the (Applied) Science Al McGartland National Center for Environmental Economics National Center for Environmental Economics 1

2 New Release of ACE Soon! National Center for Environmental Economics 2

3 EPA s View Drop terminology Value of Statistical Life (VSL) VMR or VRR Tired of saying we are placing a value on human life. My Standard Benefit-Cost Briefing Lots of criteria for decision making Environmental Justice, Distributional Justice, Benefit-Cost (efficiency), statutory direction, enforcement concerns, potential for innovation. Positive (as opposed to normative) exercise (we don t make up the values) National Center for Environmental Economics 3

4 Economists at EPA Contributions to Risk Assessment Thinking on the Margin Distributional/environmental justice Employment impacts Innovation Policy Design/market incentives for pollution control Sustainability paradigms Social Cost of Carbon

5 Overview Benefit-cost analysis at EPA How EPA conducts a benefits analysis Interactions between science and policy New applications Other tools cost-effectiveness analysis, distributional analysis, Environmental Justice analysis, Health-Wealth tradeoff A new paradigm on macroeconomic performance --the feedback effects of public health on environmental performance Jorgenson and NCEE report Smith et al. look at land reuse and cleanup EPA s Second Prospective study Big gaps in reconciling these paradigms. Science and Data Gaps

6 Executive Orders and Statutes that Require Economic Analysis EO Regulatory and Planning Review (1993); as amended by EO Improving Regulation and Regulatory Review (2011) EO Federal Actions to Address Environmental Justice in Minority and Low- Income Populations (1994) EO Protection of Children from Environmental Health Risks and Safety Risks (1997) E.O Federalism (1999) E.O Consultation and Coordination with Indian Tribal Governments (2000) E.O Actions Concerning Regulations That Significantly Affect EnergySupply, Distribution, or Use (2001) The Regulatory Flexibility Act of 1980; as amended by the Small Business Regulatory Enforcement Fairness Act of 1996 The Unfunded Mandates Reform Act of 1995 The Paperwork Reduction Act of 1995 National Center for Environmental Economics 6

7 Key Steps in the EPA Approach Establish Baseline Conditions (Emissions, Air Quality, Health) Estimate Expected Reductions in Precursor Pollutant Emissions Model Changes in Ambient Concentrations of Ozone and PM Estimate Expected Changes in Human Health Outcomes Estimate Monetary Value of Changes in Human Health Outcomes

8 What Are the Benefits of Reduced Air Pollution? Health Reduced risk of premature death Reduced risk of chronic illness, for example chronic bronchitis Reduced risk of hospitalization Reduced risk of respiratory illnesses and symptoms Reduced sick days Increased productivity for outdoor workers Welfare Visibility improvements Improved agricultural and forest yields Reduced damage to structures Reduced cleaning costs Reduced ecosystem damages

9 How do we measure these benefits? Engineers tell us how much emissions are reduced with specific technologies Air quality models tell us how air quality is expected to change Epidemiology studies give us concentrationresponse relationships to predict how health effects will change Economic studies tell us how much the changes in health and welfare effects are worth

10 How does EPA choose health and welfare effects to include in a benefits analysis? Advice from the Science Advisory Board Consistency with PM and ozone Criteria Documents and Staff Papers Well established concentration-response functions available from the peer-reviewed epidemiological literature No double-counting of benefits Focus on public health impacts rather than physiological responses Focus on well defined impacts that people can value What is the Weight of Evidence for inclusion in benefit-cost analysis? Big Issue!

11 What health effects do we quantify? PM Ozone Mortality Chronic bronchitis Hospital Admissions Asthma ER visits Acute respiratory symptoms Asthma attacks Work loss days Worker productivity Myocardial infarctions School absence rates Cardiovascular ER visits

12 Emerging Public Health Impacts Infant mortality/low birth weight Decreased lung development Cancer Doctor visits New incidence of asthma Mercury neurological impacts on children and cardiac impacts in children and adults ADHD (lead) Not quantified due to Lack of appropriate baseline incidence rates Not enough weight of evidence Not easily monetized or characterized in terms of public health significance

13 So what are the key pieces? Incidence rates Affected populations (prevalence) Estimated dose-response function Modeled changes in ambient air pollution Valuation estimates

14 Premature Mortality Example [ β PM ] Mortality = y ( e ) pop, 0 Key elements: y 0 = county-level all-cause annual death rate per person ages 30 and older β = the pollution effect coefficient (obtained from epidemiological literature) = PM 2.5 = the modeled change in annual mean PM 2.5 concentration pop = total population, 30 and older

15 Key Sources of Uncertainty Projection of inputs and impacts across time and space Uncertainty regarding interpretation of observed data, i.e. potential thresholds in concentration-response functions Use of modeled changes in ambient concentrations of PM and ozone Use of valuation estimates based on similar but not identical health risks

16 When we do not have WTP data: Cost of illness Captures the direct dollar savings to society of reducing a health effect Ignores the value to individuals of reduced pain and suffering Generally a lower bound when no WTP estimates are available

Willingness to Pay Relies on either revealed or stated preferences for risk reductions Revealed preferences from labor market studies provide values for fatal risk

17 Willingness to Pay Relies on either revealed or stated preferences for risk reductions Revealed preferences from labor market studies provide values for fatal risk reductions Stated preferences from contingent valuation studies provide values for chronic illnesses and acute respiratory effects Generally more uncertain than COI

18 Current values for health effects Premature death: $6.8 million Chronic bronchitis: $340,000 Heart attacks: $66,000 - $140,000 Hospital admissions: $6,000 - $18,000 ER visits: $300 Respiratory symptoms $15 - $60 Asthma attacks $40 Work loss days $100 School absences $75

Results of recent analyses Clean Air Interstate Rule Reduction in utility SO2 by 5.4 million tons in 2015. Reductionin utility NOx by 2 million tons in 2015.

19 Results of recent analyses Clean Air Interstate Rule Reduction in utility SO2 by 5.4 million tons in Reductionin utility NOx by 2 million tons in ,000 premature mortalities avoided 8,700 cases of chronic bronchitis avoided 13,000 hospital admissions avoided Millions of respiratory symptoms days avoided Millions of work loss days avoided Valued at $100 billion (relative to $2.6 billion in costs)

of chronic bronchitis avoided 16,000 nonfatal heart attacks avoided Millions of acute

20 Results of recent analyses (cont) Nonroad Diesel Engines By 2030, reduces NOx emissions by over 800,000 tons and diesel PM by over 126,000 tons 9,600 premature mortalities avoided 5,700 cases of chronic bronchitis avoided 16,000 nonfatal heart attacks avoided Millions of acute respiratory symptoms and work loss days avoided Valued at over $80 billion (relative to $2 billion in costs)

21 Where Does Science Play a Role? At each step in the policy analysis process, different scientific disciplines are needed to provide essential inputs The validity of the final analysis rests heavily on the validity of the science used to provide the key inputs

22 Agency Federal Rules using SCC Values Rulemaking Federal Register Date CO 2 benefits as share of total monetized benefits (%)* Final Medium-Heavy Duty Vehicles GHG Emissions EPA/ DOT and CAFE Standards Aug %** Final Light Duty Vehicles GHG Emissions and EPA/ DOT CAFE Standards May %** Proposed (supplemental) NESHAP: Mercury Cell EPA Chlor-Alkali Plants - Amendments Mar % EPA Proposed Sewage Sludge Incinerators NSPS/Emissions Guidelines (CO2 dis-benefits) Oct % EPA Proposed Transport Rule Aug % EPA Final Cement NESHAP/NSPS (CO2 dis-benefits) Sep % DOE Final ECS for Small Electric Motors Mar % DOE Final ECS for Residential Water Heaters Apr % DOE Proposed ECS for Refrigerators Sep % FAA Final Equipage Mandate To Support Air Traffic Control Service May % *Based on the central SCC value (mean, 3% discounting) ($21.4/ton for 2010 reductions (2007$)) **Excludes fuel savings. National Center for Environmental Economics 22

23 Feedback Academic Government Advocacy Industry Research Results Policymakers Public Advocacy Industry Reaction Researchers Interpretation Analysts Advocacy Industry

24 Gaps and Problems Air regulations chase epidemiology no doubt that we look under the lamp post PM mortality Ozone epi (mortality, asthma) Drinking water, pesticides, Toxics, Waste problems all deal with toxic chemicals Primarily utilize animal studies RfC, RfD won t work (Science and Decisions, NAS) New Paradigms to augment Benefit-Cost analysis National Center for Environmental Economics 24

25 Estimating Benefits: Damage Function Approach Emissions Environmental Concentrations Exposure Benefits analysis draws upon results of other disciplines Simplified linear process of benefits analysis, but there may be feedback loops For example, changes in individual behaviors evaluated in the benefits box may have implications for emissions or exposure Effects (dose-response) Benefits ($ for each effect)

26 Issue 1: Need for Dose Response relationships that can be used by economics Source: ENVIRONMENTAL HEALTH RISK MANAGEMENT Stephen McColl, John Hicks, Lorraine Craig and John Shortreed

27 Needs from Risk Assessment Health endpoints that mean something to people Continuous Dose-response functions (not safe levels Can handle uncertainty; need more information on uncertainty

28 Overarching Issue 2 Validity of Stated Preference Methods For most regulations, mortality-related benefits comprise more than 80 percent of total benefits Stated Preference vs. Revealed Preference Three camps: Stated Preference is at least as good as RP Use SP when RP isn t available SP should never be used

29 Other Benefit Transfer Issues with VSL Cancer Premium Altruism Children and VSL Voluntary vs. Involuntary Time between exposure and risks Latency vs. how risks change over time

30 Issue 3: Quantifying Morbidity Benefits Dose-response and valuation

31 Valuing Morbidity Stated preference and averting behavior approaches are often used to value morbidity endpoints For many endpoints, estimates of WTP are not available or are highly uncertain in these cases, cost of Illness (COI) estimates are often used. Potentially a dramatic underestimate of the benefits.

32 Understanding Socio-economic Decision Making Economic decisions are made by all of us that affect both the generation and consequences of environmental pollution. We need to understand how decisions are made. Energy Efficiency Paradox Pollution Prevention technologies

33 New Paradigms Beyond BCA With the high unemployment rate, many have asked how regulations affect the economy Jorgenson and NCEE looked at how the costs and benefits affect the size of the Fewer work loss days Higher IQ, smarter, more productive workers Less demand for health care resources National Center for Environmental Economics 33

34 The Clean Air Act has supported robust economic growth when the benefits are counted Estimated Impacts on GDP of the 1970 CAA % Change from base ~$200 billion net increase in GDP for year 2010 Impact on GDP considering only CAA Costs Net Impact on GDP considering both CAA Costs and Quantified Benefits Source: Jorgenson (2002) 34

35 New Paradigms General equilibrium benefits for environmental improvements: projected ozone reductions under EPA's Prospective Analysis for the Los Angeles air basin (Holger Sieg, V. Kerry Smith, H. Spencer Banzhaf, Randy Walsh) Estimating the General Equilibrium Benefits of Large Changes in Spatially Delineated PublicGoods (Holger Sieg, V. Kerry Smith, H. Spencer Banzhaf, Randy Walsh) National Center for Environmental Economics 35

36 Health-Wealth and Employment Brownfields, land clean up and reuse Putting the chronically unemployed to work Health-wealth tradeoffs Full employment vs. unemployment National Center for Environmental Economics 36

37 Employment Impacts of Regulation There are jobs created by regulations Are they benefits or are they costs? Gross job losses vs. net job loss Morgenstern, Pizer and Shih regulations cause small net increases in employment in the regulated firm (abating pollution is more labor intensive than production) Just as David Michaels talked about shifting costs, regulations can be viewed as shifting employment towards green jobs National Center for Environmental Economics 37

39 Key Features of BenMAP Includes all of the key inputs to a benefits analysis The user only has to provide modeled environmental data or select monitored air quality data for a what if style analysis BenMAP is an integrated GIS mapping, query, and statistics tool Outputs results (exposure, incidence, and valuation) in a variety of formats, including spreadsheets and shape files suitable for use with standard GIS packages such as ArcView

40 Examples of Feedback in Action Epidemiology: The case of fine particulate matter and premature death Economics: The case of the Senior Death Discount

Case Example 1 Research Particulate Matter and Premature Death Results Limited number of statistical studies showing a link between fine particulate matter and premature death, little supporting

41 Case Example 1 Research Particulate Matter and Premature Death Results Limited number of statistical studies showing a link between fine particulate matter and premature death, little supporting clinical or toxicological data Interpretation EPA set new PM2.5 standards in EPA estimated the new standards would result in 15,000 fewer deaths from PM2.5 in Reaction Claims of junk science from industry and conservative stakeholders, demands for release of research data, support from ALA and environmental groups, and hundreds of articles in the press and academic journals

42 Outcomes General Accounting Office Report: Use of Precautionary Assumptions in Health Risk Assessments and Benefits Estimates Found that EPA had not generally adopted precautionary assumptions in estimating benefits of NAAQS HEI Reanalysis costing ~ $2 million Affirmed findings of original studies Suggested areas for additional research Extension of Freedom of Information Act Requires agencies to ensure that all data produced under a [Federal] award will be made available to the public through procedures established under the FOIA. Still causing concerns within scientific community over confidentiality of subject data and propriatary data Appropriations bill language requesting NAS study of benefits analysis methods NAS study completed in 2002 Confirmed that EPA s approach is generally reasonable Recommended enhanced treatment of uncertainty, including that surrounding the PMmortality relationship Suggested using expert judgement to help characterize uncertainty Hundreds of new studies examining the relationship between PM and health effects, sponsored by EPA, HEI, industry, and environmental or public health groups. Scientific foundation upheld New, even tighter standards set in 2006

43 The Senior Death Discount Analytical issue: Most of the premature mortality associated with air pollution occurs in the elderly population, yet the value placed on reductions in premature mortality (the value of a statisticallife or VSL) is based on working age adults. Research: How does the value of reductions in mortality risk (the value of a statistical life) vary with age? Results: Mixed, with some limited evidence from the UK and Canada suggesting individuals over the age of 65 have a VSL around a third lower than individuals aged 40. Research in the U.S. did not seem to support this relationship.

44 Interpretation EPA s Science Advisory Board suggested that if adjustments to VSL for age are made they should be based on age specific willingness topay for mortality risk reductions. Other stakeholder groups suggested that additional adjustments could be made by using the value of life years gained rather than premature deaths avoided. EPA included a sensitivity analysis in an appendix to the RIA for the Heavy Duty Engines rule in 1999 showing the impact of different assumptions about the relationship between age and VSL. In the Clear Skies analysis of 2001, EPA constructed an alternative estimate of benefits that, among other assumptions, reduced the VSL for individuals over 65 by 35 percent. The alternative estimate wasonly 10 percent of the magnitude of the base estimate

Reaction Because of the large difference between the base and alternative estimates, environmental groups became aware of the assumptions being used in the alternative analysis The term Senior Death

45 Reaction Because of the large difference between the base and alternative estimates, environmental groups became aware of the assumptions being used in the alternative analysis The term Senior Death Discount was coined and public outcry over the practice became pronounced. Dozens of articles in major newspapers and journals were published criticizing the use of lower VSL for older individuals. Outcomes In early May, 2003, the EPA administrator announced EPA would not use age-adjusted values in decision making On May 30, 2003, OMB issued a memo directing federal agencies tocease using ageadjusted VSL values in regulatory analysis In July 2003, an amendment to a house appropriations bill was passed that forbid EPA from using different values of VSL for different age groups (sponsored by Rep Allen and Waxman) In November 2003, an amendment to a senate appropriations bill was passed with similar provisions (sponsored by Sens Durbin, Snowe, Jeffords, Boxer, Lautenberg, Cantwell, and Lieberman) Research continues into the relationship between age and VSL The most recent SAB advice is for EPA to use the same VSL for all ages, and to avoid using the VSLY approach

46 So How Do We Improve the Process? Clear from the two examples that we need a better understanding of uncertainty in the translation of scientific results into policy analysis Also need better communication with stakeholder groups to explain our choices of assumptions Moving forward in both areas through tool development

47 Cost-Benefit Analysis with Uncertainty NAS panel recommended moving probabalistic analysis into main body of RIAs OMB s new guidelines require probabilistic analysis for rules costing more than $1 billion Integrated uncertainty assessment requires propogation of uncertainty along the entire pathway of analysis Several methods are available for integrating uncertainty Our current focus has been on using Monte Carlo approach Investigating a number of complementary approaches, including Bayesian model averaging Bayesian meta-analysis Sensitivity analysis Response surface analysis Expert elicitation/judgement

48 New Tools All models are wrong, but some are quite useful W. Edwards Deming Some (hopefully) useful models: BenMAP (Environmental Benefits Mapping and Analysis Program) Air Pollution Response Surface Models

49 Air Pollution Response Surface Models Intended to replicate the response of CMAQ, a complex photochemical air quality model, to changes in emissions inputs Resulting reduced form models can be used to quickly generate policy analyses with confidence in the predicted responses Can be used for Strategy design and assessment Optimization What if? analyses Model sensitivity analysis

50 Response Surface Visual Policy Analysis Tool

51 What About Climate? Climate analysis is even more complicated Current techniques are limited to aggregate impacts --generally global or regional $/ton of CO2 Significant omitted categories and large data gaps Hard to deal with highly uncertain events with large impacts, e.g. collapse of ice shelfs Not clear how to treat global vs domestic impacts Important issue of how to treat intergenerational benefits should they be discounted?

52 Useful Websites EPA s National Center for Environmental Economics BenMAP The Costs and Benefits of the Clean Air Act: 1990 to The Nonroad Diesel Regulatory Impact Analysis EPA Science Advisory Board Health Effects Institute General Accounting Office Report on Use of Precautionary Assumptions in Health Risk Assessments and Benefits Estimates Harvard Center for Risk Analysis Office of Management and Budget Office of Information and Regulatory Affairs

The Role of Health Co-Benefits in EPA s Regulatory Impact Analyses. Scott Bloomberg Vice President

The Role of Health Co-Benefits in EPA s Regulatory Impact Analyses Scott Bloomberg Vice President SBCA 2016 Conference, Washington, DC March 18, 2016 Presentation Outline Introduction Background on Health