understanding and modeling of technological innovation could lead to more realistic estimation of costs.

Develop methods for analyzing complex, hybrid policies. Real policies include many complexities that address the needs and concerns of diverse sectors, regions, and interests. As a result, nearly all policies are hybrids that involve some elements of regulation and standards, some aspect of market incentives, and some degree of voluntary action on the part of individuals, firms, governments, and communities. Most existing policy analysis tools, such as benefit-cost analysis, were developed to examine simple policies that use a single modality to influence behavior, what might be termed an “ideal type” or “pure form” policy. However, in reality, many policies are complex. To provide useful analysis to decision makers shaping policy, scientists need to analyze the costs, benefits, and risks associated with complex hybrid forms of policy, which can sometimes be done with extensions of existing quantitative tools but may often require new, more advanced tools. Chapter 4 of the report discusses some of the tools currently available and under development that can be applied to this task.

Further understanding of how institutions interact in the context of multilevel governance and adaptive management. Recent research suggests that polycentric approaches to policy may be well suited to adaptive risk management. However, the interaction of multiple policies with multiple goals and approaches to affecting change, adopted and implemented in multiple contexts, can also lead to less-than-ideal results, and even situations where the outcome is wholly ineffective or even harmful. For example, policies that will change or affect water resource distributions across a multistate watershed (such as the Colorado River) will require enormous coordination among the affected states but will also have important implications for different water-dependent sectors (including agriculture, energy, flood management, industry and urban water users, and ecosystem managers) within any one state. Research is needed to characterize how policies interact across scales and intentions and to diagnose forms of policies that are most and least effective when implemented in the context of other policies. In addition, the problem of effectively linking scientific analysis to public deliberation becomes much more complex when there are multiple stakeholder groups involved, and especially when some of them reflect primarily local interests and perspectives while others are national or even global interests and perspectives (NRC, 2008h).

Develop analytical approaches that examine and evaluate climate policy taking into account its full range of effects including those on human well-being and ecosystems integrity, unintended consequences and equity effects. Policies rarely if ever do only one thing and rarely if ever affect everyone equally. Climate policy