The Impacts and Costs of Climate Change
Figure 11. The increase in marginal costs of emissions in future years without post-Kyoto action.
2000 2010 2020 2030 2040 2050 2060
Values are presented as the value in the year of emissions in 2000 prices. They are not discounted back to the year 2000.
The rate of increase for the FUND model also shows rises in future years (albeit at slightly lower
levels than the PAGE model above).
Aggregate models suggest that aggregate impacts of climate change may be positive in the short term
when climate change is still relatively modest, but turn negative for more severe climate change.
Uncertainties also increase rapidly in the longer-term, including the chance of large-scale
discontinuities (thermohaline circulation, West-Antarctic Ice Sheet, loss of biomass carbon from
increased incidence of forest fires or soil drying, greater methane release from boreal ecosystems).
Many of the existing models (at least their best estimates of the social costs) have a time horizon of
2100, thereby excluding these major effects. The impact and costs of climate change are sensitive to
the time horizon chosen, i.e. essentially whether the hazards of the remote future are considered or not,
though these effects are dampened by discounting (especially at higher discount rates).
Some models have now started to look at the effects of time-scale. A longer modelling time-scale
clearly increases the uncertainty, partly because of the uncertainty about the scenarios and partly
because parameter uncertainties accumulate over time. Many models are therefore extending the
lifetime to 2200 or even 2300. What is clear is that the effects of extending the time horizon, even
with discounting, can substantially increase the estimated marginal cost of emissions in the period
2000-2100. The impact from extending the time horizons is increased with a low discount rate, and
with equity weighting.
Reporting of Statistical Data
Both the mean and the median have been used as a measure of central tendency for the social costs of
climate change. Since for skewed distributions they give substantially different results even with the
same underlying data it is important to consider which is appropriate, so that at least consistent
comparisons can be made.
Defining a central value in a data set in the presence of outliers is difficult. The usual measure, the
arithmetic mean or average, is an unbiased measure of the expected value if the data form a
AEA Technology Environment, August 2005
Watkiss, Paul; Downing, Tom; Handley, Claire & Butterfield, Ruth. The Impacts and Costs of Climate Change. Oxford, England. UNT Digital Library. http://digital.library.unt.edu/ark:/67531/metadc29337/. Accessed August 21, 2014.