More selective than most chemical pesticides in that they ordinarily destroy only unwanted species, biocontrol. agents (such as insects, fungi, and( viruses) eat, infect, or parasitize targeted plant or animal pests. However, biocontrol agents can negatively affect nontarget species by, for example, competing with them for resources: a biocontrol agent might reduce the benefits conferred by a desirable animal species by consuming a plant on which the animal prefers to lay its eggs. Another example of indirect negative. consequcnces occurred in England when a virus introduced to control rabbits reduced the amount of open ground (because large rabbit populations reduce the ground cover), in turn reducing underground ant nests and triggering the extinction of a blue butterfly that had depended on the nests to shelter its offspring. The paucity of known extinctions or disruptions resulting from indirect interactions may reflect not the infrequency of such mishaps but rather the failure to look for or to detect them: most organisms likely to be adversely affected by indirect interactions are of little or no known commercial value and the events linking a biocontrol agent with an adverse effect are often unclear. Moreover, determining the potential risks of biocontrol agents before they are used is difficult, especially when a nonnative agent is introduced, because, unlike a chemical pesticide, a biocontrol agent may adapt in unpredictable ways. so that it can feed on or otherwise harm new hosts.

    Most pre-1990 literature on businesses' use of information technology (IT)-defined as any form of computer based information system-focused on spectacular IT successes and reflected a general optimism concerning IT's potential as a resource for creating competitive advantage. But toward the end of the 1980's, some economists spoke of a"productivity paradox"; despite huge IT investments, most notably in the service sectors, productivity stagnated. In the retail industry, for example, in which IT had been widely adopted during the 1980's, productivity (average output per hour) rose at an average annual rate of 1.1 percent between 1973 and 1989, compared with 2.4 percent in the preceding 25-year period. Proponents of IT argued that it takes both time and a critical mass of investment for IT to yield benefits, and some suggested that growth figures for the 1990's proved these benefits were finally being realized. They also argued that measures of productivity ignore what would have happened without investments in IT productivity gains might have been even lower. There were even claims that IT had improved the performance of the service sector significantly, although macroeconomic measures of productivity did not reflect the improvement.

    But some observers questioned why, if IT had conferred economic value, it did not produce direct competitive advantages for individual firms. Resource-based theory offers an answer, asserting that, in general, firms gain competitive advantages by accumulating resources that are economically valuable, relatively scarce, and not easily replicated. According to a recent study of retail firms, which confirmed that IT has become pervasive and relatively easy to acquire, IT by itself appeared to have conferred little advantage. In fact, though little evidence of any direct effect was found, the frequent negative correlations between IT and performance suggested that IT had probably weakened some firms' competitive positions. However, firms' human resources, in and of themselves, did explain improved performance, and some firms gained IT-related advantages by merging IT with complementary resources, particularly human resources. The findings support the notion, founded in resource-based theory, that competitive advantages do not arise from easily replicated resources, no matter how impressive or economically valuable they may be, but from complex, intangible resources.