Keynote Lectures
Intergenerational Ethics under Resource Constraints
When evaluating long-term policies, economists usually suggest to maximize the sum of discounted utilities. On the one hand, discounted utilitarianism was given a solid axiomatic foundation by Koopmans (Econometrica 1960). On the other hand, this criterion has questionable implications when applied to economic models with resource constraints. Based on recent contributions to the literature, I will present various ethical conditions for intergenerational distribution, and use these conditions to illuminate the conflict between equity and efficiency that has been a central theme in this literature.
Valuing Risks to Health and Life: The State of the Art
Many public policies, such as environmental programs, transportation safety laws and workplace safety regulations, save lives by reducing the risk of dying of individuals. These programs differ in the nature and size of the risk reduction, the time at which the risk reduction is experienced (now or in the future), and the population affected. For example, by reducing air pollution, it is possible to save lives in the short term. Epidemiological evidence suggests that the lives saved are primarily those of elderly persons in poor health. Epidemiological studies (Dockery et al., 1993, Pope et al. 1995 and 2002) have also linked sustained exposure to air pollution over a long period of time with shorter life expectancy. Whether pollution to fine particulate matter and aerosols brings forward existing cardiovascular and respiratory conditions, makes them more severe, or causes their onset, is not entirely clear at this time. In hazardous waste and contaminated site cleanup programs, much emphasis is placed on carcinogens, which implies that reducing exposure to these pollutants now reduces the risk of cancer years from now. The population affected is usually comprised of current or future residents of the area in the immediate proximity at the site, or workers at the contaminated site. In some cases, as in pesticide rulemaking, the government agency must decide whether to reduce large risks that affect a small population (pesticide mixers and applicators) or small risks that affect a large population (pesticide residue in produce, all consumers) (Cropper et al., 1992). By contrast, road traffic accident deaths occur primarily among young people. In 2001, in the US motor vehicle accidents were the leading cause of death among children aged 4-15, youths aged 16-20, and adults aged 20-34. Economists would recommend that in designing and implementing such programs, at least some considerations be given to their costs and benefits. Traditionally, the benefits of policies that save lives have been computed as L×VSL, where L is the expected number of lives saved (=population affected×reduction in risk due to the policy) and VSL is the Value of a Statistical Life. The Value of a Statistical Life is defined as the Willingness to pay for a marginal reduction in risk. A major challenge for economists and policymakers is, therefore, how to estimate the VSL. It is, in general, difficult to place a value on risk reductions, because we rarely observe market transactions where individuals buy or sell risk reductions. To circumvent this problem, economists have deployed revealed- and stated-preference non-market valuation techniques. Revealed preference methods that have been used to estimate the VSL include compensating wage studies, hedonic property value studies, and other consumer behavior studies. In this presentation, I discuss the assumptions underlying these approaches, and the data and econometric problems arising in many applications, as well as the appropriateness of the application of their VSL figures to environmental and other safety programs. The method of contingent valuation—an example of stated preference methods—has been frequently applied to value risk reductions. The approach is survey-based and relies on individuals reporting directly their willingness to pay for a reduction in risks. One advantage of using contingent valuation is that this method is flexible and lends itself to a variety of risk contexts. Moreover, it asks respondents to value a risk reduction of a specified nature and size, rather than assuming that people know the risks associated with specific contexts. One problem with contingent valuation, however, is that it places a heavy cognitive burden on the respondents. One measure of the researcher’s success in communicating risks is that WTP satisfies the scope tests, i.e., that it increases with the size of the risk reduction (Hammitt and Graham, 1999), and yet most empirical evidence is that WTP increases only weakly with the size of the risk reduction. Contingent valuation studies have typically resulted in lower estimates of the VSL than those produced by labor market studies, and have sometimes suggested that individuals place the same value on reducing risks in the environmental quality context as they do in the transportation accident context. Much attention is currently being devoted to the issue of whether there should be a unique VSL, or whether there are many VSLs, depending on the size of the risk reduction, on the risk reduction context, the time when the risk reduction is experienced (now or in the future), and on the individual characteristics of the beneficiary of the policy. For example, are the elderly willing to pay less for a reduction in their own risk of dying? How does the willingness to pay—and hence the VSL—change with changes in income? Is it possible to account for growth in income over time when lives are saved over a very long time horizon, as might be the case in the climate change context? How is WTP affected by risk aversion? And how is WTP for a specific risk reduction affected by the presence of high background risks? The latter question is relevant, for example, when we study the WTP of elderly individuals in poor health for a reduction in air pollution. Implicit in the above discussion is the assumption that is appropriate to quantify the physical effects of the policy as the number of lives saved, where the relevant construct is the reduction in risk, which is in turn equal to the increase in the survival function of the affected individual(s). Recently, it has been argued that, when looking at the long-term effects of air pollution, the statistical models estimated by epidemiologists allow one to compute the loss in life expectancy (LLE) due to sustained exposures spread over the population, but not the exact number of lives lost due to pollution (Rabl, 2004). The monetary valuation translation of this claim is that one needs to obtain an estimate of the value of a day/month/6 months/year of life expectancy gain/loss. It is certainly possible to convert the WTP for a specified risk reduction into the Value of a Statistical Life Year (VOLY), but doing so requires making restrictive assumptions about the shape of the survival function after the risk reduction (Alberini et al., 2004). A well designed study would, therefore, directly ask individuals to value a life expectancy gain. We are aware of only few studies that seek to value directly a gain in life expectancy (see, for example, Chilton et al., 2004). Clearly, this is an area where much research is needed. Another area of future research is how to estimate the VSL for children. Some of the earlier studies in this area have attempted to model expenditures on procedures specific to environmental toxicants, e.g., chelation of heavy metals in blood (Agee and Crocker, DATE). More recent theoretical and empirical work (Crocker and Agee, DATE; Dickie and Gerking, 2004) has sought to find a marginal rate of substitution between adult and child values. In the context of the skin cancer, Dickie and Gerking estimate this rate of substitution to be approximately 2, which means that individuals’ WTP for a given risk reduction experienced by their children is twice as much as they would pay for themselves. I conclude by reviewing VSL and VOLY figures explicitly adopted by agencies, as well as the figures implied—by not formally expressed—in their policy decisions (Cropper et al., 1992; Van Houtven and Cropper, 1995; Ashenfelter and Greenstone, 2002).
Space in Unified Models of Economy and Ecology
Economic and ecological systems evolve in time and space. Interactions take place among units occupying distinct spatial points and geographical patterns of production activities, urban concentrations, or species concentrations occur. The emergence of spatial patterns in economics has received relatively little systematic analysis, with the notable exception of the body of knowledge developed in the context of new economic geography. On the other hand, the concept of diffusion has been used in ecological modeling to explain spatial pattern formation in ecological systems.
The purpose of the present lecture is to review and present methods of modeling spatial problems in ecological economics, and to show how these methods can be used to environmental and resource management. In this context modeling methods associated with diffusion processes and "the Turing mechanism" for diffusive instability, which produce steady state spatial patterns, are reviewed.
In particular it is shown how spatial methods can be used to explain the emergence of steady state spatial patterns of harvesting and species concentration. The mechanisms generating spatial pattern formation are also used to explore ways of designing environmental and resource management policies with spatial (e.g. regional) characteristics.