Introduction

Many analysts, (including Pizer [Chapter 25], Keller et al. [Chapter 28], Webster [Chapter 29] and Toth [Chapter 30] in this volume, as well as others like Nordhaus and Popp [1997], Tol [1998], Lempert and Schlesinger [2000], Keller et al. [2004] and Yohe et al. [2004]) have begun to frame the debate on climate change mitigation policy in terms of reducing the risk of intolerable impacts. In their own ways, all of these researchers have begun the search for robust strategies that are designed to take advantage of new understanding of the climate systems as it evolves – an approach that is easily motivated by concerns about the possibility of abrupt climate change summarized by, among others, Alley et al. (2002). These concerns take on increased importance when read in the light of recent surveys which suggest that the magnitude of climate impacts (see, for example, Smith and Hitz [2003]) and/or the likelihood of abrupt change (IPCC, 2001; Schneider, 2003; Schlesinger et al., 2005) could increase dramatically if global mean temperatures rose more than 2 or 3 °C above pre-industrial levels. Neither of these suggestions can be advanced with high confidence, of course, but that is the point. Uncertainty about the future in a risk-management context becomes the fundamental reason to contemplate action in the near term even if such action cannot guarantee a positive benefit–cost outcome either in all states of nature or in expected value.