Home Page (MVA) > Integrated Assessment Models (IAMs) and Resources > IAMs Thematic Guide
Thematic Guide to Integrated Assessment Modeling
[HOME] [PREVIOUS] [NEXT] [BOTTOM]
Socioeconomic Impacts
Integrated assessment projects ultimately seek to estimate the value of climate-induced changes to people. While describing likely changes in unmanaged ecosystems is clearly an important part of this valuation, most studies concentrate primarily on aspects of impacts that are easier to bound, using one of several approaches.The most basic decision an integrated assessment makes is whether to attempt to disaggregate impacts. Those studies that stop their climate description at a global average change in temperature are constrained to estimate aggregated impacts, since disaggregated impacts require specification of climate changes at regional or national level. No researcher using such aggregated damage functions has claimed any more than illustrative value for them, though most calibrate their functions to pass through one of Nordhaus' estimates of equilibrium three-degree losses for the United States, which are founded on a moderately detailed sectoral analysis of the U.S. economy.
Analyses that disaggregate impacts first split them into market and non-market impacts, since it is plausible to expect these areas to be driven by fundamentally different processes. For example, impacts in market and non-market sectors are likely to depend differently on the timing of impacts and on incomes. They may also depend differently on the level and rate of climate change.
More detailed treatment of market impacts requires disaggregation by sector. The most common approach is to look at sectors with clear climate dependence, such as agriculture, forestry, water resources, energy, transport, and construction. The biggest challenge here is one of matching resolution. Different sectoral impacts depend differently on changes in mean climate and on occurrences of extreme events. While some detailed agricultural models can assess impacts of extended droughts, no firm basis exists yet for projecting extreme climatic events that will change given average climate changes. Where single extreme events such as storms can cause large losses, the basis for projecting such impacts is really not available.
Market impacts estimated by sector must be aggregated to economy-wide impacts. Approaches include simple adding up (which assumes that no effect in one sector passes through to affect any other), input-output modeling, or general equilibrium modeling that can follow adjustments through the economy to calculate aggregate welfare loss.
Impacts that are not mediated by markets are harder still to project and value. These would include all effects people experience as a result of the changed climate and environment: changes in health and well-being; and changes in the pleasure people derive from climate directly and from the pattern of land-use and ecosystems that results from it. Studies so far have used only heuristic, illustrative estimates, and have not attempted to disaggregate various forms of non-market effects. The contingent valuation method, though problematic (National Oceanographic and Atmospheric Administration 1993), would be potentially available to estimate willingness to pay to avoid specified changes. Valuing non-market changes over such a long time is even harder than present applications of contingent valuation, though. Large-scale shifts in the species mix of forests may occur over many decades, but people's feelings about how the forest should look are also likely to adjust over time. Future generations, lacking a memory of how the forests once were, will have even less basis for perceiving the new species mix as a loss.
If either the technical difficulties or the principle objections to aggregating all market and non-market impacts into a single monetary metric are too great, then an alternative is to define a vector of various measures of valued environmental consequences. This can be used in several ways: to present multi-attribute consequence measures to people to gain information about their preferences; to develop multi-attribute utility functions of representative actors, which would ideally be defined over uncertain outcomes as well; or to spell out the effect that a constraint defined in terms of one valued environmental variable would have on the set of feasible combinations of other valued consequences.
The next page is Treatment of Uncertainty.
