《demand for health risk reductions》由会员分享,可在线阅读,更多相关《demand for health risk reductions(23页珍藏版)》请在金锄头文库上搜索。
1、Demand for health risk reductions$Trudy Ann Camerona,n, J.R. DeShazobaDepartment of Economics, University of Oregon, Eugene, OR 97403-1285, USAbDepartment of Public Policy, School of Public Affairs, 3250 Public Policy Building, UCLA, Los Angeles, CA 90095-1656, USAa r t i c l e i n f oArticle histor
2、y: Received 2 January 2012 Available online 7 August 2012Keywords: Value of a microrisk reduction Value of a statistical life Risk reduction Risk valuation Mortality risks Morbidity risksIllness profilesBenefit-cost analysis Choice model Representative national surveya b s t r a c tA choice model ba
3、sed on utility in a sequence of prospective future health states permits us to generalize the concept of the value of statistical life (VSL). Our representative national survey asks individuals to choose between costly risk-reducing programs and the status quo in randomized stated choice scenarios.
4、Our model allows for separate marginal utilities for discounted net income and avoided illness years, post-illness years, and lost life-years. Our estimates permit calculation of overall willingness to pay to reduce risks fora wide variety of different prospective illness profiles. These can be benc
5、hmarked against the standard VSL as a special case.Murphy and Topel 52).2 Our new approach to measuring the values people assign to health risk reductions represents an improvement over conventional empirical strategies. We begin with a structural model of utility in future periods of an individuals
6、 life as a function of the health status they will experience in those future periods. We differentiate these future health states as current health, sickness, recovered/remission years, and lost life-years. In our stated choice survey (also known as a conjoint analysis or a discrete-choice experime
7、nt), each subject is presented with several opportunities either topurchase one of two illness-specific health-risk reduction programs or to stick with their status-quo health risks.3These risk reduction programs involve diagnostic screening and, when risks are high, medical therapies that would red
8、uce, but not eliminate, the subjects chance of experiencing that particular future illness with its associated pattern of health states. We use the tradeoffs embodied in peoples stated choices to infer their WTP for a given-sized reduction in their baselinerisk of experiencing a specified future ill
9、ness profile. However, these given-sized risk reductions are heterogeneous. The implicit value of an incremental sick year or lost life-year can then be inferred, as in a hedonic model, by taking the derivatives of this overall WTP with respect to the number of sick-years or lost life-years involved
10、.4 Our strategy overcomes several limitations of the conventional VSL approach. These limitations have long been recognized by researchers, but have been unavoidable due to the constraints of existing empirical data and methods. Weintroduce two main innovations. First, we generalize the conventional
11、 strategy by more comprehensively defining the good to be valued. Instead of valuing a single mortality risk reduction in the current period, we value risk reductions for atime profile of possible adverse future health states. Individuals express their WTP to reduce their risks of entire timeprofile
12、s of adverse health states over their remaining lifespans. We do not have to extrapolate these future estimates from only current-period data. Importantly, we can identify inter-temporal substitutability or complementarity among future health states. This is possible because we estimate demands for
13、a much wider range of health risks than usual. Our model subsumes myriad patterns of illness, recovery, and lost life-years across the individuals remaining lifespan. Thisgeneralization is needed because the majority of benefits from many health, environmental and safety policies accrue in future ye
14、ars of the individuals life, as opposed to solely in the current period.5 Second, our structural random utility model for our subjects discrete choices makes it very clear how WTP estimates for reductions in the risks of sick-years and lost life-years depend upon the individuals age, income, margina
15、l utility of other consumption, and discount rate. Informed by the lifecycle model of Ehrlich 27, our structural model also recognizes and builds upon a growing empirical literature which has explored various sources of heterogeneity within traditional VSL estimates.6While we make advances in struct
16、ural modeling in terms of the most important variables in this paper, we cannot comprehensively explore all alternative assumptions or all possible sources of VSL heterogeneity in one paper. For example, we leave to related and future papers a more-detailed exploration of the roles played by, for example, age,current health status, specific-illness effects, subjective risk beliefs, choice set complexity and alternative discounting assumptions. Conceptually, we focu