Richard Newell remarked that much of the analytical work that has been done to date has been complicated by the absence of a carbon price, making it difficult to model behavior in the absence of direct empirical evidence of how actors will respond. Thus, a carbon price should also help improve analysts’ ability to model behavior into the future.

Nebojsa Nakicenovic commented that integrated assessment modeling has made huge progress over the last 20 years and has had success in integrating economics with technological perspectives, demographics, and other human dimensions, and then linking all of this to climate models. Where it has been less successful is in folding in impacts and possible adaptation measures. He outlined three areas in need of improvement: (1) dealing with uncertainty, about both technologies and policies—there are no tools to adequately consider low-probability but highly consequential events; (2) analyzing failures, which would provide valuable insight into ways to support R&D efforts—specifically, how does one measure the success of R&D, particularly at the early deployment stage? and (3) heterogeneity of decisionmakers—regionally and sectorally, agents will behave differently, but this heterogeneity is not well reflected in most models.

Finally, as was stressed during the discussion of current analytical capabilities, workshop participants pointed out that there is a pressing need for more regional and household-level data. Regional data is essential to successfully integrating air-quality and land-use models. Recognizing that end users are increasingly requesting detailed outputs (e.g., state-level employment impacts), participants emphasized that the quality and confidence level of such outputs will depend on improved data sets. Participants also noted that international data sets can be of poor quality and difficult to obtain, but nonetheless they are crucial to global modeling efforts.

Institutions and Innovation

David Montgomery remarked that existing models are effective for modeling idealized policies, but these do not reflect the real world. He outlined three shortcomings: failure to consider institutions; grossly underestimating the costs of inefficient nonmarket policy initiatives that are already coming into effect; and not adequately addressing the R&D and innovation process. Processes such as institutional change and innovation are not represented structurally in the models, nor are they generally predictable or controllable, but as one models out through 2050 or 2100, these processes are almost the entire story.

Montgomery contended that modeling global costs requires understanding how the institutional settings in different countries will limit the efficiency of policies and the feasibility of achieving emissions reductions—the rule of law and the existence of economic and political freedom are factors that will have an impact but are not modeled.

On a related point, modeling policies in the United States, Canada, and the European Union requires also being able to model the perverse incentives and unintended consequences of command and control regulations, technology mandates, and targeted subsidies. The field of regulatory economics, however, does have a long and solid history of analyzing the implications of regulatory programs and perverse incentives, and so a dialogue between the modeling community and those who study institutions may be beneficial.

Montgomery also advised that no model will ever capture all of the ways that a smart economic agent can find to circumvent regulations, which can raise costs and diminish a program’s effectiveness. He pointed to a large body of literature (e.g., Cohen and Noll, 1991) dedicated to characterizing the history of R&D and demonstration projects and the role of government. Models do not provide the kind of insight that would inform the design of R&D policies.

Bill Nordhaus stressed that intellectual property rights will be another key component, and reiterated that there are limited instruments (e.g., an aging patent system) to address this. He then raised the question of whether or not climate change is somehow different from other sectors, or is it possible to look to sectors such as health or telecommunications. Marilyn Brown cited a CCTP report (Brown et al., 2006) that analyzed the disciplines that are most critical to the key technology areas looking at climate



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement