These are related both to
issues affecting the timing of policies, and to the appropriateness of policy mixes in promoting new
technology development. Each is discussed in turn below.
Wigley et al. (1996) introduced the concept of optimal pathways towards predefined concentration levels.
They concluded that deferring the bulk of abatement efforts would be a cost-effective approach, arguing
that sunk investments, technological change, discounting, and natural decay of atmospheric CO2 would
contribute to making later reductions less expensive that earlier ones.
A number of arguments were made that would modify their initial conclusions, ranging from the
possibility that future climate change “surprises” requires meeting lower concentration levels than
previously anticipated (Ha-Duong et al., 1997) to the role of endogenous technical change (Grubb, 1997).
Furthermore, while Wigley’s model supposed exogenous technical improvements, the role of learning by
doing processes implies a more complex picture. Learning investments may well be necessary in the near
term to make future reductions cheaper.
������ �� (2002) took the opposite view, arguing that, in conjunction with RD&D, timely investment in
new technologies with lower CO2 emissions might be a more cost-effective strategy for reducing global
emissions than postponing investment decisions in the hope that mitigation technologies might somehow
become more attractive through “autonomous” RD&D improvements and cost reductions in step with
natural turnover of capital: “Postponing investment decisions will not by itself bring about the
technological change required to reduce CO2 emissions in a cost-effective way. Even worse, under
unfavourable conditions it might bring about further “lock-in” of energy systems and economic activities
along fossil-intensive development paths”
Such investments may occur through “dissemination subsidies” that may not necessitate more
comprehensive policies to directly or indirectly modify the price system.. As Roehrl and Riahi (2000) put
