Summary
Are most climate models too large and complex to allow us to properly explore the worrying tails of the probability distributions of environmental outcomes? Is it time to build a complementary system of models? These models should be simpler, estimated empirically with data where possible and use prior information elicited from experts where the empirical data do not exist.
This is consistent with the recommendation of the renowned epistemologist and polymath Sir Harold Jeffreys: “Physicists always choose the simplest law out of an infinite number that would satisfy the observations equally well because they think that it will maximize the probability of making the correct inference.” In Scientific Inference, Cambridge UP 1973.
In addition, scholars such as Professor Ralph Keeney of Duke University have written that before we look at alternative strategies, we need to think carefully about our values. (Value Focused Thinking. A Path to Creative Decision-making, Harvard University Press, 1992). He claims that thinking about values enables you to: 1) Uncover hidden objectives, 2) Guide information collection, 3) Improve communication, 4) Facilitate involvement in multiple stakeholder decisions, 5) Interconnect decisions,
6) Evaluate alternatives, 7) Create alternatives, 8) Identify decision opportunities and 9) Guide strategic thinking. Guessing first and later interviewing every important nation’s “representative” should identify which values are important and how they are traded off by the nation’s citizens one by one: this information is then summarized in a parameterized utility function. Important nations are not always the most developed; for instance a small nation like Costa Rica has an unique ecosystem.
The model must be dynamic (move through time in appropriately sized steps), it must be stochastic (allow for uncertainties both about parameters and about the future development of the world): and it must be a game [nation i cannot attain the highest level of its utility because every other nation j (j=1,2,…i-1,i+1,….n), for i = 1,2….n ] is trying to attain the highest level of its own utility].
The same model can be solved as a dynamic stochastic optimization, if utility weights are assigned to each nation. The optimal solution will be much more efficient than the game solution. This suggests the interesting possibility that as each decision point is reached, bargains can be reached by winners, if a movement is made from the game solution to the optimal solution – that will make the losers in such a movement better off and there will still be additional “dividends” to be shared!
Finally we note that Einstein, Jeffreys and Zellner all suggested that model simplification is a high priority. The most obvious simplification of this model suggests an interesting hypothesis for audience discussion and debate. Has much of the work summarized in the IPCC documents in the 22 years from 1990 to the present been inappropriately focused?
Biography
Stephen Peck joined the Electric Power Research Institute (EPRI) in 1976 serving in a variety of positions and overseeing work related initially to Utility Planning as well as National Energy Planning and then to Environment. He served finally as Vice President of Environment from 1995 to 2000. From 2002-2006 Peck was President of Fleche, a consulting firm. In this capacity he worked on management of greenhouse gases, on transmission market design and on anti-terrorist strategies for electricity companies. He established new consulting activity in 2012 based in New York, Fort Myers, FL., London and Beijing.
Stephen Peck was educated in England as an engineer (Cambridge, BA, MA) and an economist, (London School of Economics, MSc) and in the United States. He holds an MBA and Ph.D. degree from the University of Chicago’s Graduate School of Business, where he studied with Professors Milton Friedman, Merton Miller, and Hans Theil. He worked most importantly under Professor Arnold Zellner, the Bayesian econometrician who was his thesis advisor. Before joining EPRI, he served as an Assistant Professor in the Economics Department at UC Berkeley, where he taught Price Theory, Macroeconomics, Industrial Organization, Antitrust & Regulation, Monetary Economics and Econometrics.
Stephen Peck has published one book and over seventy five peer-reviewed papers on topics of contemporary energy and environmental interest including:
1) Energy Demand Analysis,
2) Decision Analysis for Large Capital Investments subject to regulatory and other risks,
3) Value of information for an organization like EPRI informing the Environmental Protection Agency, in a Dynamic Stochastic Game Framework, which coincidentally is the Mathematical Theory of Rhetoric, a direct follow-on to The Theory of Inference developed by Sir Harold Jeffreys in Theory of Probability 3-rd Edition, OUP 1936
4) Inter-temporal Cost Benefit Analysis of the Global Climate Change Issue (with T. Teisberg), and
5) Energy & Transmission Pricing for Electricity Networks (with H. Chao) .
His work on the latter two topics is known and admired worldwide. The latter work with Chao is particularly well cited in major journals. A few years ago, it had received the most citations for an article appearing in the Journal of Regulatory Economics, (JRE) (150). The second most cited article in JRE had received 50 citations.
Peck has lectured on a variety of topics in the US and abroad, and has served on NAS, DOE and EPA advisory Committees. He has served as an Associate Editor of the Energy Journal.