Grantee Research Project Results
Final Report: The Valuation of Mortality Risk Reductions: Application of Two New Survey Instruments
EPA Grant Number: R826608Title: The Valuation of Mortality Risk Reductions: Application of Two New Survey Instruments
Investigators: Krupnick, Alan J. , Cropper, Maureen , Simon, Nathalie , Alberini, Anna
Institution: Resources for the Future
EPA Project Officer: Chung, Serena
Project Period: October 1, 1998 through September 30, 2000
Project Amount: $270,000
RFA: Decision-Making and Valuation for Environmental Policy (1998) RFA Text | Recipients Lists
Research Category: Environmental Justice
Objective:
The overall objectives of this research project were to: (1) develop and test methods for estimating willingness to pay (WTP) for contemporaneous and future mortality risk reductions over a person's life cycle (particularly later in life); and (2) provide the policy community with credible estimates of WTP for these risk reductions in contexts more appropriate for environmental benefits analysis. No estimates of this type currently exist for the United States. To evaluate whether we have been successful in obtaining reliable estimates of WTP, we subjected our results to both internal and external scope tests, examining the following hypotheses drawn from the theoretical literature: (1) WTP should increase with the risk reduction; (2) the farther into the future the change in risk (life expectancy) occurs, the lower WTP is for the change today; (3) WTP for a current risk change should be a hump-shaped function of age, increasing with age over some interval, but eventually decreasing; and (4) the worse a person's health is today, the less his/her WTP for a current risk change. Regarding hypothesis 2, according to the life-cycle consumption model with uncertain lifetime, a person's WTP today for a reduction in his/her risk of dying 10 years hence should equal what he/she would pay for a contemporaneous risk reduction 10 years from now, discounted to the present. A key issue for policy is how much individuals discount future WTP. It also is of interest to see how WTP for a future risk reduction varies with current age and health status.
Much of the justification for environmental rulemaking rests on estimates of the benefits to society of reduced mortality rates. Reductions in risk of death are arguably the most important benefit underlying drinking water regulations, air pollution regulations, and the disposal of hazardous waste.
In both the United States and Canada, estimates of individuals' WTP for mortality risk reductions, and the implied Value of a Statistical Life (VSL), come from revealed preference studies, primarily studies of compensating wage differentials in the labor market, as well as stated preference (contingent valuation) studies. However, the use of figures from these studies to value the lives saved by environmental programs is problematic. Both types of studies have focused on measuring the value that healthy, prime-aged adults place on reducing their risk of dying, whereas the majority of statistical lives saved by environmental programs, according to epidemiological studies, appear to be the lives of older people and people with chronically impaired health (Pope, et al., 1995; Schwartz, 1991; Schwartz, 1993). In addition, these studies have focused on immediate risk reductions, whereas in the case of environmental policies, the benefits of reducing exposure to carcinogens and certain air pollutants are experienced years after the change in exposure.
There are two reasons why older persons likely are to benefit disproportionately from reductions in pollution. First, epidemiological studies typically assume that the effects of a change in exposure are proportional to baseline mortality (Pope, et al., 1995; Morales, et al., 2000). Because persons older than age 65 account for three-quarters of all deaths in the United States and Canada, a larger proportion of statistical lives will be saved among the old rather than among the young. Second, some epidemiological studies have found larger changes in mortality rates for people older rather than younger than age 64 (Schwartz, 1991, 1993). Epidemiological studies also suggest that people with chronic heart or lung conditions likely are to benefit disproportionately from improvements in air quality (Schwartz, 1991; Schwartz and Dockery, 1989; Pope, et al., 1995).
It has been suggested that older people should be willing to pay less for a reduction in their risk of dying rather than younger people on the grounds that they have fewer expected life years remaining. Indeed, some economists have argued that the VSL should be converted to a Value per Statistical Life Year (VSLY), and that lives saved should be valued by multiplying remaining life expectancy by the VSLY. This procedure assumes, implicitly, that each year of life is equally valuable, and that the VSL is strictly proportional to remaining life expectancy. Whether this approach is consistent with welfare economics, however, depends on how, empirically, WTP for reductions in risk of death varies with age.
It has, likewise, been argued that the WTP for people in ill health should be less for a reduction in their risk of dying because their utility from an additional year of life is lower than that of healthy people. In the literature on Quality-Adjusted Life Years (QALYs), saving the life of a person with chronic bronchitis is less valuable than saving the life of a person in good health. This has been used as an argument for assigning a lower VSL to beneficiaries of air pollution control programs (EOP Group, Inc., 1997) than the value currently used by the U.S. Environmental Protection Agency (EPA). However, it is not clear that people with chronic heart and lung disease would pay less than healthier individuals to reduce their risk of dying.
Finally, it has been argued that the current WTP for future risk reductions should be less than the current WTP for current risk reductions. There are two reasons for this. First, people with a positive rate of time preference would prefer to have their benefits closer to the current period rather than later. Second, the risk of dying between the current and future period is greater than zero. Thus, the WTP for future risk reductions may be conditioned on the possibility that the individual would not live to experience them.
This research seeks information on these hypotheses through the development, administration, and analysis of two contingent valuation surveys designed to test the above hypotheses. Survey development was funded through a previous EPA Science to Achieve Results (STAR) grant. One survey was administered to 930 randomly selected residents of Hamilton, Ontario (funded both through this grant and through a grant from Health Canada) and the other, using Knowledge Network Inc.'s pre-selected sample with WEB-TV, to a nationally representative sample of 1,200 U.S. residents. Both surveys elicited respondents' WTP for several reductions in mortality risk. Respondents were limited to persons aged 40 years and older, including those older than 60, to examine the impact of age on WTP. Extensive information was collected about each respondent's health status to see if it systematically influences WTP.
The survey was computerized and self-administered either at a centralized location on a computer or in the home through the Internet. Our survey began with a series of questions about the respondent's health history and the health history of his/her family. This is followed by exercises that acquaint the respondent with the concept of risk and test the respondent's risk comprehension. Respondents are introduced to simple probability concepts using coin tosses and roulette wheels, working up to our standard risk communication device; a 1,000-square grid, in which risks are represented using red squares. To test their comprehension, respondents were asked to compare grids for two hypothetical people (person A and person B) and to determine which of the two has the higher risk of death. They also are asked to select which of the two people they would rather be. The baseline risk of death for a person of the respondent's age and gender is then presented both numerically and graphically.
It is sometimes argued that respondents in contingent valuation surveys find it difficult to report their WTP for a mortality risk reduction because they are not accustomed to trading income for reduced risks. To mitigate this problem, we first acquaint respondents with quantitative risk reductions resulting from medical tests and products that likely are to be familiar to the respondent (e.g., mammograms, colon cancer screening tests, medicine to reduce blood pressure). In doing so, we provide only qualitative cost information for each action or product ("inexpensive," "moderate," and "expensive").
This is followed by the WTP questions. Information about WTP is obtained through a combination of dichotomous choice payment questions with followups and open-ended questions. Respondents are asked an initial dichotomous choice question: would they buy the product at a price randomly chosen from one of four predetermined values? Those respondents who answered "yes" were asked if they would buy the product at a higher price, while those who answered "no" were asked if they would buy the product at a lower price. A final, open-ended question was asked of those respondents who gave "yes-yes" or "no-no" responses.
Each respondent was asked his/her WTP over the next 10 years for three risk reductions. Respondents in part 1 of each survey were asked to value a 5-in-1,000 risk reduction in the next 10 years first, whereas respondents in part 2 were asked to value a 1-in-1,000 risk reduction first. After each question, respondents were asked to indicate their degree of certainty about their WTP responses on a scale from 1 to 7. All respondents younger than 61 received a third WTP question asking for their WTP now for a 5 in 1,000 risk reduction experienced over 10 years, beginning at age 70. Because WTP can be affected by the respondent's understanding of risks and interpretation of the scenario, we included debriefing questions at the end of the questionnaire to identify respondents who had trouble comprehending the survey or who did not accept the risk reduction being valued. The survey ends with questions about the respondent's income, followed by SF-36, a questionnaire commonly used in medical research to measure mental and physical health status (Ware, et al., 1997).
Both the baseline risk of death and the risk reductions are communicated graphically. Baseline risk of death over the next 10 years is represented by coloring in red the appropriate number of squares on a white grid containing 1,000 squares. Reductions in risk of death are shown by turning the appropriate number of red squares to blue.
Summary/Accomplishments (Outputs/Outcomes):
The VSL estimates for Canada for a contemporaneous risk change range from $506,000 and $933,000, when computed using WTP for the 5 in 1,000 risk reduction, compared to $700,000 and $1.54 million for the United States. When based on WTP for the 1 in 1,000 risk reduction, the VSL estimates increased, reaching upwards of $4 million. Because WTP generally is not proportional to the size of the risk change, VSL estimates were larger when calculated using WTP for the smaller risk change. However, when we distinguished respondents by the degree of confidence they have in their answers, median WTP does increase in proportion to the size of the risk reduction. Although the more generous VSL amounts for the United States and Canada appear in line with estimates used in policy assessments in these countries, the results for the 5 in 10,000 (annual) risk reduction were well below the estimates used in these countries. Which results are appropriate to use is a matter for further research and policy debate. However, internal validity tests on the responses to the 1 in 10,000 (annual) risk reductions do much worse in comparison to those for the 5 in 10,000 risk reduction. At the same time, risk reductions delivered by policies were more often in the range of the smaller risk reduction.
Our results provide only weak support for the notion that WTP declines with age; specifically, in our Canadian sample, WTP declines by about 30 percent after age 70, compared to WTP at younger ages. There is no decline, however, in the U.S. sample. We similarly find no support for the idea that people who have chronic heart disease, lung disease, or cancer exhibit less WTP to reduce their risk of dying than people without these illnesses.
Specifically, persons with chronic heart and lung disease are willing to pay at least as much to reduce their risk of dying as persons who do not have these diseases. In our U.S. sample, WTP was significantly greater (holding age, gender, and income constant) for persons who have been hospitalized for a chronic heart or lung condition within the last year and for persons who have high blood pressure. There are no statistically significant differences between the two groups in Canada.
Regarding age, respondents in our Canada sample who were 70 years of age or older were willing to pay about one-third less than their younger counterparts to reduce their risk of dying over the next 10 years. There was, however, no statistically significant impact of age on WTP in the U.S. sample.
Turning to the results for future risk reductions, WTP is significantly lower than that for contemporaneous risk reductions for the 40-60 age group. In addition, WTP depends on expected health status at age 70, and on the subjective probability of surviving until age 70. Discount rates are 0.045 (United States) and 0.080 (Canada). In the Canada study, the discount rate increased with age and decreased with one chronic health measure. However, we do not see statistically significant effects of age or current health status on the discount rate for the U.S. sample.
Our results suggested that VSLs used in the United States may be too large, particularly for future risk reductions. At the same time, our results support current practice regarding treatment of age and health status in both the United States and Canada. The EPA currently uses a central VSL estimate, based primarily on labor market studies, equal to approximately $6 million (1999 U.S. $) for all ages. In contrast, Health Canada employs age-adjusted VSL estimates in its economic assessments, applying a VSL of C$5 million (or U.S. $4 million) to exposed populations under 65 years of age and using an adjustment factor of 0.75 for populations aged 65 years and older. (These adjustments are based on Jones-Lee, Hammerton, and Philips [1985].) Neither agency currently makes adjustments based on health status.
These results stand in sharp contrast to the way in which age and health status are treated in evaluating medical interventions. We believe the comparison is relevant, because it is sometimes suggested that a similar approach be used in benefit-cost analyses of health and safety regulations (U.S. Food and Drug Administration, 1999). The standard approach in the medical literature is to measure life-saving benefits in terms of QALYs. This assumes that the value of lives saved is strictly proportional to remaining life expectancy, and that the value of saving a life-year is less for a person with a chronic disease, such as chronic bronchitis, than for a healthy person, with the exact equivalence determined by QALY weights. Our results do not support either of these assumptions. There is no evidence that the VSL should be equally apportioned over remaining life expectancy, or that the VSL is systematically lower for persons with chronic illness.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
Other project views: | All 30 publications | 3 publications in selected types | All 2 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Alberini A, Cropper M, Krupnick A, Simon NB. Does the value of a statistical life vary with age and health status? Evidence from the U.S. and Canada. Journal of Environmental Economics and Management 2004;48(1):769-792. |
R826608 (Final) |
Exit Exit |
|
Krupnick A, Alberini A, Cropper M, Simon N, O'Brien B, Goeree R, Heintzelman M. Age, health and the willingness to pay for mortality risk reductions: a contingent valuation survey of Ontario residents. Journal of Risk and Uncertainty 2002;24(2):161-186. |
R826608 (Final) |
Exit Exit |
Supplemental Keywords:
cost-benefit, contingent valuation, economics, decisionmaking, willingness to pay, WTP, mortality, contingent valuation, age, health status., RFA, Economic, Social, & Behavioral Science Research Program, Health, Scientific Discipline, PHYSICAL ASPECTS, Geographic Area, Health Risk Assessment, State, Risk Assessments, Physical Processes, decision-making, Ecology and Ecosystems, Economics & Decision Making, Social Science, ecosystem valuation, contingent valuation, social psychology, dichotomous-choice, exposure, decision analysis, life expectancy, valuation of mortality, valuing environmental quality, economic incentives, environmental values, morbidity risks, human exposure, environmental policy, Maryland (MD), psychological attitudes, compliance costs, mortality, mortality risks, willingness to pay, human health risk, economic objectivesRelevant Websites:
Progress 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.