Grantee Research Project Results
2002 Progress Report: Not All Deaths are Created Equal: Understanding Individual Preferences for Reductions in Morbidity-Mortality Events
EPA Grant Number: R829485Title: Not All Deaths are Created Equal: Understanding Individual Preferences for Reductions in Morbidity-Mortality Events
Investigators: DeShazo, J. R. , Cameron, Trudy
Institution: University of California - Los Angeles
EPA Project Officer: Hahn, Intaek
Project Period: October 1, 2001 through September 30, 2003 (Extended to May 30, 2006)
Project Period Covered by this Report: October 1, 2001 through September 30, 2002
Project Amount: $360,756
RFA: Decision-Making and Valuation for Environmental Policy (2001) RFA Text | Recipients Lists
Research Category: Environmental Justice
Objective:
Our research generalizes the traditional single-risk, single-period Value of a Statistical Life (VSL) model in several ways. We pursue three major objectives in our effort to generalize the traditional VSL model, all of which involve accommodating various sources of heterogeneity when deriving individuals’ demand for risk reductions. The objectives of this research project are to: (1) develop a more structural model of demand that yields the present discounted value of private risk-reducing programs that yield future reductions in the risk of morbidity and premature mortality; (2) evaluate how the age of the respondent and the age at which risks are reduced affect current period demand for private risk-reducing programs; and (3) evaluate how actual and expected illnesses affect individual’s demand for particular private risk-reduction programs.We also have the corollary objectives of designing and administering three surveys that collect data sets that permit us to empirically estimate individuals’ demand for: (1) private health risk mitigation programs; (2) public risk prevention programs; and (3) public treatment programs that reduce risk.
Progress Summary:
We have completed the first of our objectives. We have developed a more general model. We are endeavoring now to publish our findings. We also have designed and implemented our three surveys. In addition, we have begun to disseminate our results.
Surveys and Collected Data
We have designed and implemented three surveys.
Survey of Demand for Private Risk Reductions Programs
We conducted a national survey of 2,439 U.S. respondents focusing on their demand for private risk reduction programs. We present respondents with an illness-specific health-risk reduction program that involves diagnostic screening, remedial medications, and lifestyle changes that would reduce their probability of experiencing that illness profile. Individuals must pay an annual fee to participate in each risk-reducing program. They are asked to choose between two risk-reducing programs (each associated with a different illness profile) or to reject both programs. An advantage of this choice setting is that the individual faces a portfolio of health risks that resemble those they actually face. Through their choices, individuals reveal trade-offs across specific illnesses and a full continuum of health states of different durations. We also observe them strategically allocating expenditures for risk mitigating programs across the current year and future years of their remaining life. Each health risk in our study is presented as an illness profile that describes a probabilistic time pattern of health states that the individual could experience. Each health profile consists of randomly assigned values for the individual's future age at the time of onset, the severity and duration of treatments and morbidity, the age at recovery (if there is any), and the number of lost life-years (if there are any).
Survey of Demand Public Risk Prevention SurveyTo a nationally representative sample of approximately 1,600 individuals, we administered a stated preference survey for public risk prevention programs. Within a pair-wise choice set of prevention policies, we vary the number of illnesses prevented, the number of deaths avoided, the length of time the policy is in effect, the source of the health threat, the type of disease avoided, the size of the affected population, and several other attributes. We directly evaluate shifts in demand for categorically different types of risk exposures such as air, water, and food contaminants, as well as the risk of highway mortality and morbidity. We also evaluate a comprehensive set of illnesses and targeted groups, including cancer, leukemia, leukemia in children, colon/bladder cancer, asthma, asthma in children, lung cancer, heart disease, heart attack, stroke, respiratory disease, and motor vehicle accidents. Finally, we also elicited individual-specific measures of the incidence of the private benefits of each program as well as several attitudinal measures.
Survey of Demand for Public Mitigation of the Ex post Effects of Health RiskFinally, again to a nationally representative sample of approximately 1,600 individuals, we administered a stated preference survey to elicit demand for publicly available treatment programs that reduce their risk of experiencing time spent in states of illness or of premature death. These interventions would increase the development and adoption of new types of medical treatments for a wide range of illnesses. The availability of these interventions would reduce the risk of death and increase the probability of recovery for those people who already are sick. In addition, treatment-interventions also enable us to examine how respondents demand changes when the treatment benefits specific groups within society. Specifically, we can evaluate how respondent demand shifts for interventions that benefit only children (asthma and leukemia), only adult, only seniors, or any combination of these groups. We also consider treatments that benefit only men (prostate cancer) and only women (breast cancer).
Paper in Progress for Publication
We have produced draft papers for the first three objectives. That is, we have papers that: (1) develop a more general model; (2) evaluate the role of age and latency on demand; (3) evaluate the effects of actual and expected morbidity on demand; and (4) estimate the demand for public risk prevention programs.
We develop a utility-theoretic choice model in which individuals choose among alternative programs to reduce their risk of experiencing future years of illness and/or lost life-years. Unlike previous stated-preference approaches to deriving the VSL, our model is able to produce separate estimates of the marginal utilities of both avoided sick-years and avoided lost life-years. With these marginal utilities, we may infer willingness to pay to avoid a wide range of adverse health profiles over an individual’s future life. Such estimates are particularly important for ex ante benefit-cost analyses of environmental, health, or safety interventions where costs must be incurred now to reduce health risks that will not materialize fully until much later. The model generalizes the single-period, single-risk models (typically used with revealed-preference data to produce single-valued VSL estimates) in that we allow individuals to substitute across health risks with different time profiles. We evaluate our model using data from an extensive national survey that contains a set of randomized choice experiments. The model generalizes the single-period, single-risk models (typically used with revealed-preference data to produce single-valued VSL estimates) in that we allow individuals to substitute across health risks with different time profiles. We evaluate our model using data from an extensive national survey that contains a set of randomized choice experiments.
Correcting for as many sources of sample selection bias is essential to ensure that our estimates of demand for risk reductions truly are representative of the U.S. population. To date, we know of no study that represents the general adult and senior population as we plan to do. The Office of Management and Budget recently has begun to focus extensively on methodological issues such as nonresponse biases in survey data. To address, and preempt, such concerns, we seek to expand our focus on data quality issues.
Researchers frequently acknowledge several reasons for possible nonrepresentativeness in surveys of samples drawn from large consumer panels. We model the selection process for one such panel, starting with a random-digit-dialed set of initial contacts and following these cases through a number of distinct attrition opportunities, ending with one sample drawn for an actual survey and the individuals who chose to respond to it. Using GIS methods, we match more than 525,000 random-digit-dialing addresses or telephone exchanges to the corresponding county and the most appropriate Census tract. We use a set of 15 orthogonal factors based on Census tract characteristics, plus county voting percentages for candidates Gore and Nader in the 2000 presidential election. We find many statistically significant determinants of attrition at our different attrition opportunities. To illustrate the effects of selection, we consider a second subsample where survey respondents expressed their opinions about the proper role of government in terms of environmental, health, and safety regulations. In a formal maximum likelihood selection model, we find some evidence of a slight liberal (proregulation) bias that may stem from nonrandom selection, but the effect is not statistically significant and the hypothesis of “no liberal/conservative bias” cannot be summarily rejected. Less sophisticated models in the class of propensity score corrections show minimal significant effects on selection propensity in the regulatory preference outcome models, but the distortions are quantitatively tiny.
Future Activities:
We will complete the two remaining objectives. We have collected the necessary data to complete these objectives.
Journal Articles:
No journal articles submitted with this report: View all 15 publications for this projectSupplemental Keywords:
air, ambient air, water, drinking water, exposure, risk, risk assessment, health effects, human health, sensitive populations, dose-response, carcinogen, population, children, elderly, stressor, age, race, sex, ethnic groups, susceptibility, life-cycle analysis, decision making, cost benefit, conjoint analysis, nonmarket valuation, contingent valuation, survey, psychological, preferences, public good, Bayesian, socioeconomic, willingness-to-pay, compensation, analytical, surveys, measurement methods,, RFA, Economic, Social, & Behavioral Science Research Program, Health, Scientific Discipline, Health Risk Assessment, Risk Assessments, Susceptibility/Sensitive Population/Genetic Susceptibility, decision-making, Environmental Statistics, genetic susceptability, Sociology, Social Science, Economics & Decision Making, mortality rates, decision making, valuation of mortality, value of statistical life (VSL), environmental policy, health valuation models, mortality studies, models, representativeness, mortality, mortality risks, conjoint analysisProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.