requires consideration of this diversity of attributes. Furthermore, some resource systems—such as groundwater basins or airsheds—provide only pure common-pool resources. Others, such as forests, yield some products that are subtractive (e.g., timber) and others that are nonsubtractive (e.g., flood control) (Gibson et al., 2000a). Thus, an analyst trying to understand how institutions affect behavior in regard to forest resources may need to understand which aspects of a forest are common-pool resources and which are public goods. (Subtractive and nonsubtractive products are related, however. For example, cutting timber can reduce a forest’s ability to provide flood control.) We briefly describe three further attributes of resources that may have a major impact on the incentives that individuals face: renewability, scale, and cost of measurement.
Renewability relates to the rate at which resource units that are extracted (or used as a sink) replace themselves over time. The replacement rate over time can take any value between zero (nonrenewable) and one (instantly renewable). Mineral and oil resources are normally considered nonrenewable because once they are extracted from their source, no replacement is generated within a human time frame. Thus, the key problem faced in regulating nonrenewable resources is finding the optimal path toward efficient mining of the resource (Libecap, 1990).
On the other hand, biological species that are harvested for human use regenerate themselves in a cycle that varies from less than one year to decades, assuming the breeding stock and the breeding habitat are protected. Individuals who attempt to achieve sustainable use of such biological resources over time devise rules to limit the number of users; limit the technology, timing, quantity, or location of extraction; and protect the habitat of the species. The costs of designing, implementing, monitoring, and adapting these rules can vary substantially depending on the particular species characteristics, their habitat, the technology used, and the culture of the users. Resources that regenerate slowly are more challenging to manage because overharvesting may not be discovered until recovery of the resource is severely endangered. Fish that tend to cluster in groups are more likely to be destroyed with modern fishing technology because the marginal cost of searching for and harvesting the full extent of the fishery is much lower than for fish that spread out over a larger area (Clark, 1976, 1977).
Some human-made common-pool resources are renewed very rapidly once use has halted or been reduced. Broadcasting bandwidth, for example, is a common-pool resource because it is limited, one person’s use is subtractive, and thus congestion can occur if too many users try to use the same bandwidth at the same time. The resource regenerates immediately, however, when usage declines, so subtractability exists across users, but not across time. Such commons cannot be destroyed permanently by overuse. The type of rules that are effective for regulat-