efficacy chemicals. Most cases of insecticide resistance are to chlorinated hydrocarbons because these compounds have been in widespread continuous use for many years. Herbicide resistance is becoming more of a problem, and severe impacts can occur when there are no alternative herbicides to control the resistant genotypes or when available alternatives are relatively expensive. By spreading the burden of crop protection over multiple tactics, rather than relying on a single tool, farmers will face less risk of crop loss and lower rates of pest adaptation to control measures (Bessin et al. 1990, Vangessel et al. 1996). Because pests will continue to evolve in response to pest controls, research needs to support development of pest-management tools that reduce selection pressure, delay selection for resistance, and thus increase the life of chemical and other products.
There are several justifications for public engagement in these lines of research. First, there is underinvestment from a social perspective in private-sector research because companies will aim to maximize only what we have called suppliers' surplus (difference between suppliers ' income and their production costs) rather than the social surplus. Companies will compare their expected profits from their patented products resulting from research and will not consider the benefits to consumers and users. For example, the development and release of a self-sustaining biological control agent, such as a parasitoid or entomopathogen, or the development of an integrated crop-rotation program might benefit farmers and food consumers but not provide a marketable commodity to a company. Second, research activities are very risky, and even big corporations might have a higher aversion to risks than society does, and that will lead to underinvestment in risky research activities (Sandmo 1971). Recent developments in computers and medical biotechnology illustrate this point. Many of the breakthrough innovations in these fields were the result of publicly funded university research whose results were patented. The rights to use the patents were sold to new startup companies that in many cases were the result of partnerships between university professors and venture capitalists (Zilberman et al.1998, Parker et al. 1998). That is the case with important companies that include Sun, Cisco, Netscape, Genentech, Chiron, and, in agriculture, Calgene. Major corporations often take over such companies once they are established. Publicly supported university research, through the process of technology transfer, has become a source of economic growth in the United States, and this model is now being imitated elsewhere (Levy 1998). One reason that biotechnology research in agriculture is lagging in its development relative to medi-