EVENTS

Date:	Tuesday, November 16, 2004
Time:	9:00 AM -- 4:30 PM
Location:	Wohlstetter Conference Center, Twelfth Floor, AEI 1150 Seventeenth Street, N.W., Washington, D.C. 20036

November 2004

A Review of the Economic Foundations of Climate Change Models

On November 16, AEI and The Economist co-hosted the Second Annual Climate Policy Conference, convening international panels of experts to examine the controversy over emissions forecasting scenarios and methods. The fundamental input of climate change models is derived from forecasts of future emissions of greenhouse gases, but in recent years there has been considerable discussion and dispute over the economic foundation of these scenarios. The conference reviewed the uncertainties of emissions forecasting with an eye toward highlighting "best practices" that might be used to refine future climate change forecasts.

PANEL I: THE SRES EMISSIONS SCENARIO PROCESS: AN OVERVIEW

Hugh Pitcher
Pacific Northwest National Laboratories

Hugh Pitcher, a member of the Intergovernmental Panel on Climate Change (IPCC) Working Group III, provided an overview of the IPCC's "Special Report on Emissions Scenarios" (SRES), released in 2000 as a part of the IPCC's Third Assessment Report on Climate Change. The SRES contains forty scenarios of greenhouse gas emissions over the course of the twenty-first century that form the primary inputs for the climate models on which warming forecasts are generated.

Emissions estimates over a century time scale create large and novel demands on forecasting techniques. The SRES incorporated numerous emissions scenarios available in existing scholarly literature and generated new scenarios based on population forecasts, expected changes in fertility rates, and estimates of carbon intensity as the world economy grows. Because of the large uncertainty of long-range estimates, the SRES neither assigned probabilities to any of the scenarios nor created a "central-case" scenario. Pitcher took note of the criticisms of the range of the SRES scenarios, observing that some reviewers argue that the lowest emissions scenario is too high while other observers argue that the highest emissions scenario is too low. In addition, one politically necessary assumption of the SRES--that every nation will experience substantial economic growth in this century--may not be realistic. Pitcher emphasized that the process of generating emissions scenarios is extraordinarily difficult and time-consuming; the 2000 SRES took several years to assemble.

PANEL II: CRITIQUES OF EMISSIONS SCENARIOS

David Henderson
Westminster University, London

David Henderson coauthored--with Ian Castles--a critique of the SRES for its use of Market Exchange Rates (MER) rather than Purchasing Power Parity (PPP) as the foundation for comparing inter-country projections of economic growth. This is one reason Henderson and Castles believe the SRES overstates likely emissions growth. Henderson's critique went beyond just the SRES economic methodology, however, as he argued that many economic and statistical issues have been badly mishandled within what he calls "the IPCC milieu."

Letters sent to the chairman of the IPCC aim to trigger debate and actions in the context of the fourth assessment review (AR4). AR4 was launched in November 2003 and will be completed in 2007. The Henderson-Castles correspondence focuses on two points: 1) The SRES should not be taken as the starting point of AR4. There should be a stronger basis. 2) The IPCC process needs broadening through involvement by available national statistics offices and ministries of finance and economics. Because the IPCC will not comment on SRES, it is difficult to tell their position. To improve upon the SRES, there should be a more balanced, professional treatment of economic and statistic aspects of the panel's work. Interior review is not sufficient to be impervious to informed criticism.

Emissions projections can yield different results through different models. The PPP is a better measure for static comparisons of economic wealth and consumption but does not apply to modeling the process of growth. SRES Working Group III reports use MER figures presented by the UN Human Development report. These figures are invalid due to breach of accepted international conventions that require such data to be determined through PPP. The IPCC needs wider official participation and a more informed, professional structure. For example, OECD finance and economic ministries believe that the OECD environmental ministry should look at these issues. 

Warwick McKibbin
Australian National University

Emission projections are critical to the climate debate. The projections drive the climate, the climate drives the impact, and policy ultimately represents all three. As such, projections are useful, but they should be kept to short-term forecasts of no more than thirty years. Patterns are evident in the projections, but because of faulty projection figures and lack of compelling proof, the projections do not necessarily apply to long-term forecasts. Economists should look for such structural changes determining GDP growth and also emissions projections, which are simultaneously determined. The data may look as though carbon emissions are converging, for example, but the data might actually show countries diverging.

McKibbin utilizes G-Cubed--a dynamic, macroeconomic, multi-country model. Its projections show how people respond to changing energy prices. There may be no fixed relationship between carbon emissions and GDP growth. The sector in which each country experiences productivity growth relates closely to global emissions projection. When not considering structural changes, linear trends and constant ratios are likely to be misleading.

The choice of PPP versus MER models makes a big difference in terms of economic growth and carbon emissions projections. The SRES approach requires assumptions of growth and population. Conversions between PPP and MRE can be done, but there is no empirical connection between the two and, therefore, no reason to convert between the two. SRES does not distinguish plausibility or likelihood between storylines and models. We can project up to thirty years into the future; these projections, however, are not accurate enough to drive policy. We must use models that acknowledge uncertainties, encourage investment in alternative technologies, promote self-management of risk, and minimize costs of abatement and adaptation rather than hitting specific targets assumed to hit emissions targets.

LUNCHEON ADDRESS

James Connaughton
Chairman, President's Council on Environmental Quality

Climate policy is not its own exclusive entity; but it must be integrated into an entire agenda of priorities. An emphasis on short-term projects coupled with sustained economic growth is the appropriate and most effective means of addressing environmental issues. Goals have been set to reduce greenhouse gases by 18 percent and ultimately to stop and reverse the growth of emissions. Science, technology, and budgeting, along with international partnerships, will help this process. Carbon capture and storage technology, including Future Gen and carbon sequestration--along with safer nuclear policies and energy, deriving hydrogen from diverse domestic resources, and helping the transition to the global hydrogen economy--will pave the way toward these goals.

Near-term policies include voluntary environmental programs, tax incentives for hybrid vehicles, renewable energy, fuel economy increase for light trucks, farm-bill incentives for carbon sequestration, and international programs. Furthering reductions from major carbon emitters around the globe is a priority. Very recently, the new Methane to Markets objective was initiated to reduce carbon emissions and capture methane for clean energy, pollution reduction, economic growth, and energy security.

PANEL III: REDUCING UNCERTAINTIES IN EMISSIONS FORECASTS

Ross McKitrick
University of Guelph

Mark Strazicich
Appalachian State University

McKitrick and Strazicich proposed to reduce the range of uncertainty of the SRES scenarios by using a per-capita emissions model. Current global CO₂ emissions average about 1.1 tons per person per year and have been roughly the same for decades. Per-capita data varies widely on the national level but not on the global average. National variability cancels out at a global level. Probabilities can be evaluated by looking at the global average emission level because widely variable scenarios are less indicative.

McKitrick and Strazicich explained the technical details of their time-series analysis. If a time series has a stable mean, it has "mean stationarity" and likewise a stable trend in a time series has "trend stationarity." If the series is nonstationary, the variants approach infinity and are therefore not useful in forecasting. "Unit root tests" are experiencing a boom in empirical economics, allowing for two structural breaks in the mean and trend, with such break points being determined internally. If structural breaks are permitted, all countries may be stationary. If not, they are cointegrated, where the global average is stationary.

Working through this analysis, McKitrick and Strazicich conclude that only about seven of the forty SRES scenarios are plausible, and these seven are on the low end of the SRES emission range. The plausible SRES 2050 scenarios show a 9.1 to 11.2 gigaton emission of carbon. Beyond 2050, a population decline is anticipated that could cause an emissions peak. Other scenarios are too improbable to consider. The IPCC should provide an economic explanation for why the global per-capita emissions average has been constant and why the mechanism will stop working, and they should also provide a time-series model so the mechanism shows nonstationary trends. The plausible scenarios indicate figures at the lower end of emissions scenarios.

PANEL IV: ROUNDTABLE DISCUSSION ON DECISION-MAKING UNDER UNCERTAINTY: WHERE DO WE GO FROM HERE? 

Richard N. Cooper
Harvard University

A one-hundred-year outlook is impossible to project and entirely speculative. Recent twenty-five-year projections underestimated growth and overestimated the consumption of materials. As usual, everyone assumed that we would soon run out of oil. The scenario approach is more reasonable than investing in forecasts because the long term cannot be predicted. The only use for PPP is to measure living standards across currency areas, something that MER does not adequately do, whereas PPP cannot measure welfare comparisons. Beyond thirty years ago, reliable PPP figures from outside the United States and Europe do not exist. Purchasing Power Parity brings problems with currency conversion, such as when prices may differ for the same commodity within a country for reasons of economic value--something the PPP does not adequately show. Market exchange rates are the proper form of analysis. Furthermore, the Kyoto Protocol is the wrong way to go about policy. High economic costs are not worthwhile, and environmental quality control requires careful study in intervals.

Nasir Khilji
Energy Information Administration

The publicly funded Energy Information Administration (EIA) projects energy futures through 2025 with a Keynesian large-scale macroeconomic model. The EIA reuses global projections of GDP based on market exchange rates by looking at individual models. Using a macroeconomic framework and market exchange rates allows for broad comparisons. Growth rates and energy forecasts do not change in PPP or MER. Yet when GDPs are aggregated across regions, figures such as per-capita income statements are misleading because the PPP rate shows less emissions intensity. Conversions can always be done with market exchange rates. An international macro model is now being developed with the G-Cubed model as a basis. The EIA will keep looking at macro models to be able to evolve into an integrated macro world-energy model forecasting market exchange rates and price levels.

Steven F. Hayward, AEI's F. K. Weyerhaeuser Fellow, and AEI program assistant Elizabeth White prepared this summary.