Letter from the Director
Uncertainty and the Burden of Proof
David H. Krantz
I propose that uncertainty about climate change is an argument for vigorous departures from Business-As-Usual. The burden of proof should be shifted: anyone who favors BAU should be obliged to demonstrate that catastrophic climate change is extremely unlikely. Skepticism about climate change has many variants. At an extreme is conspiracy theory: reduction of greenhouse gas emission is a plot to undermine our way of life. Less extreme is the view that environmentalists select and exaggerate evidence of anthropogenic warming in order to lobby for change. Indeed, it is hard for anyone to avoid bias in evidence selection – environmentalists are not excepted. Many skeptics believe that serious consequences will be averted without drastic action – perhaps through scientific breakthroughs, or perhaps simply through massive expenditures on adaptation by future generations. (If economic growth continues, so the argument goes, our heirs will be wealthier than we, and thus both abler and more inclined to spend on environmental amelioration.) Finally, there is skepticism derived from uncertainty. The approximations in climate models lead to uncertainty, model forecasts are intrinsically probabilistic, and climate-impact models are crude; thus, both the future extent and the consequences of global warming are quite uncertain.
Skepticism, however, pales when one properly imagines ecological catastrophes that might affect Homo sapiens. Wally Broecker likens the climate system to an angry beast. We may be uncertain how this complex system will react, if we prod it with a sharp stick; but uncertainty is an argument for avoiding such a prod, not for testing it. The possible consequences of the beast’s reaction to the prod are too severe to run this risk.
Yet it may already be too late. For hundreds of thousands of years, Earth’s atmospheric CO2 has cycled between about 190 and 290 parts per million (by volume), while global mean temperature has co-varied, roughly in phase with CO2, over a range of about 10°C. But in recent years, we have driven CO2 to about 390 ppm in Earth’s atmosphere; and Business As Usual may drive it to double this already highly provocative level.
To help imagine what an ecological catastrophe would be like, one can think about such catastrophes as they affect species other than Homo sapiens. Some populations expand and shrink by a factor of three or more. A well-studied example is the Canadian lynx, whose subpopulations expand unsustainably, in response to easy prey, but then contract drastically. The pain and the intra-specific aggression of starving lynx go unrecorded; but if human population were to shrink by a factor of 3, at least the beginning of that catastrophe would be recorded, and would make the record of human genocide over the past few millennia look like a genteel tea party.
Such a catastrophe is far from certain; but can we rule it out? The burden of proof for BAU would be to show that it is virtually impossible. I don’t think that a credible argument of that sort can be made. Thus, uncertainty makes the case for vigorous departure from BAU.
CRED’s primary funding comes through the National Science Foundation program called DMUU, or Decision Making Under [Climate] Uncertainty. This program itself represents a small, ambivalent departure from BAU. Since climate change was uncertain, the U.S. government invested a little in research, rather than making a commitment to programs that might have strong effects on people’s lives. And over the past 6 years, CRED researchers have begun to understand the complex ways in which uncertainty affects decision making.
One of our themes is strategic use of uncertainty: people use it as an argument for whatever action (or inaction) they already favor for other reasons. A person with strong prevention focus may use uncertainty to favor caution: don’t commit to this romance, or don’t release water from this reservoir. With a promotion focus, the same uncertainty would argue for eagerness: seize the opportunity, it might work out! Discussions of water allocation (CRED field site in Ceará) show clear examples where arguments based on uncertainty follow self-interest. A recent lab result (Min Gong, et al.) shows that lasting effects of subsidies for cooperation can be obtained when there is stochastic uncertainty of outcomes.
A fuller understanding may require us to distinguish carefully among four different sources of uncertainty: (1) stochastic process; (2) psychological distance (temporal, geographic or social); (3) coordination (will enough others make this choice?); and (4) distrust of evidence (e.g., concern about the validity of a climate model or about the motives of a forecaster). Stay tuned!