![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081107233008im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
1997 Progress Report: Deriving Biodiversity Option Value Within a Model of Biotechnology Research and Development
EPA Grant Number: R824707Title: Deriving Biodiversity Option Value Within a Model of Biotechnology Research and Development
Investigators: Rausser, Gordon C.
Institution: University of California - Berkeley
EPA Project Officer: Clark, Matthew
Project Period: October 1, 1995 through September 30, 1997
Project Period Covered by this Report: October 1, 1996 through September 30, 1997
Project Amount: $80,000
RFA: Valuation and Environmental Policy (1995)
Research Category: Economics and Decision Sciences
Description:
Objective:The goal of this project is to develop a feasible method for computing the potential value of biodiversity in its role as a source of intellectual property. The focus is on the role of ecological and taxonomic knowledge in the process of biodiversity prospecting. We analyze a sequential-search model of biodiversity prospecting in which genetic materials are usefully differentiated by prior information. The analysis shows that, as prior information allows for the differentiation of biological habitats according to their potential as sources of new drug leads, bioprospecting values increase in some areas, while declining in others. When search procedures are optimized to take account of this information, areas of especial promise may have a high value. Information creates value both by increasing the chance of making a discovery, and by lowering the average cost of conducting searches.
Our work represents a conceptual departure from previous economic models of biodiversity prospecting (e.g., Simpson, Sedjo and Reid, 1996; Polasky and Solow, 1995) in two ways. First, we argue that the proper unit of analysis in such work is not the species, but the physical location. Relevant decisions that bear on bioprospecting ? conservation, fundamental systematics, and goal-driven search projects ? are generally made at the site level. Second, we argue that, once we adopt the site as our unit of analysis, we open the door to the possibility that searches can be guided by observable ecological and taxonomic data. The success of bioprospecting projects depends on the identification of previously unknown complex compounds. Current ecological science suggests that biochemical "creativity" ? the propensity of organisms to generate the complex organic compounds that provide leads for new drugs ? may be correlated with such observables (Dreyfuss and Chapela, 1994). Indeed, Sandoz Pharmaceuticals, Inc., has recently launched a Biolead Project that will attempt to identify observable factors that correlate with microbial creativity.
In the model, a bioprospecting firm conducts a search for a compound that will make possible the development of a lucrative new product. There are a large number N of sites where the compound might be found. Sites are tested sequentially, at a cost c per site. A test of the nth site is treated as a Bernoulli trial with probability pn of scoring a success (or "hit"). The hit probabilities of different sites are assumed to be independent. In order to avoid trivial cases, we assume that no site contains the desired compound with certainty (pn < 1 for all n). Without loss of generality, we can assign labels to sites in order of decreasing hit probability, so that 1 > p1 ? ? ? pN. When a test is successful, a payoff is realized. Multiple hits are redundant. It is shown that the pharmaceutical firm maximizes the payoff of its search program by testing the most promising sites first and, therefore, that the probability ordering (p1, p2,..., pN) is also the order in which sites are examined (up to a permutation of equi-probable sites). Using this principle, a value function is derived that elucidates the expected payoff of the search at each stage, conditional on results at previous stages. Let Vn denote the ex post expected value of continuing the search, after n-1 sites have been tested unsuccessfully. This continuation value is characterized by the recursive relationship:
Vn = pnR+(1 - pn)Vn+1-c , n=1,...,N
where VN+1 = 0. An expression for the expected incremental contribution of the nth site follows:
vn = an[pn(R-Vn+1)-c],
where an = P n-1i=1 is the probability that the search is carried to the nth stage; i.e., the probability of failure in each of the first n-1 tests. Analysis of this formula yields a fundamental insight into the relationship between information resources and the bioprospecting value of genetic resources:
Proposition: Let {pn}Nn=1 be a sequence of hit probabilities on a collection of research leads, index in order of decreasing probability. Let the incremental value vn of the nth lead be defined as above. Then vn can be decomposed into components vn = vnI +vN, where
and where vN = aN(pNR-c) is the value of a marginal lead.
The components vnI and vN are referred to as the information rent and the scarcity rent, respectively, of lead n. Analyzing the model, we find that sites toward the front of the search queue add more to the project's expected return than do those further back. This result is due to a combination of two factors. First, the early, high-probability sites contribute more than the others to the chance of a successful outcome. As repeated failures push investigators to pick through lower-grade ore, it becomes increasingly unlikely that a hit will ever be scored. Second, even if a hit is made eventually, the shift to low-quality sources implies an increase in the expected number of trials required to make the discovery and, therefore, an increase in the expected costs of continuing the search. Since an early success obviates the need for continued (and costly) search, sites toward the front of the queue are valuable for their capacity to reduce total search costs, in expectation. In sum, when search procedures can be optimized to incorporate useful prior information, high-probability sites command information rents associated with their expected contribution to the chance of success and to the avoidance of search costs.
To demonstrate the approach, we apply our formula to Myers' (1988, 1990) data on several biodiversity "hot spots," using the density of endemic species as a proxy for site quality (see Figure 1). Several insights emerge. Information values can be several orders of magnitude larger than the "scarcity value" of the material itself, and can be substantial even when scarcity values are negligible. Indeed, the values associated with the highest-quality sites (on the order of $4,000/hectare in our simulation, for rainforest in Western Ecuador) can be large enough to motivate conservation activities. These results are robust over large ranges of parameter values.
The valuation approach advanced here could be the basis of a technique for assigning an expected bioprospecting value to a habitat or parcel. Such a technique would take advantage of available scientific knowledge, and could be sharpened as new information emerges about relevant relationships. This includes information on the relationship between habitat, ecology, and the creativity of micro-organisms, and on how microbial communities are affected by various forms of environmental disturbance. The expected bioprospecting value of a parcel or region could be incorporated into benefit/cost studies, as an aid to policy decision making for cases in which development could disturb or imperil microbial communities. Future work in this area should include an examination of how bioprospecting values vary with changes in the institutional environment.
Journal Articles:No journal articles submitted with this report: View all 11 publications for this project
Supplemental Keywords:Ecosystem Protection/Environmental Exposure & Risk, Economic, Social, & Behavioral Science Research Program, Scientific Discipline, RFA, Biology, Social Science, decision-making, Economics & Decision Making, Monitoring/Modeling, ecosystem valuation, policy analysis, preference formation, psychological attitudes, public policy, property values, decision analysis, community-based, environmental policy, social resistance, conservation, cost effectiveness, models, valuing environmental quality, genetic materials, biodiversity option values, public values, social psychology, valuation, compensation, environmental assets, biodiversity, economic benefits, economic incentives, environmental values, standards of value
Progress and Final Reports:
Original Abstract
Final Report