Servicio de Información Comunitario sobre Investigación y Desarrollo - CORDIS

Cross-cutting paper, measuring the costs of climate policies: watch out for conceptual traps (Jean-Charles Hourcade and Frederic Ghersi)

Since the late eighties, climate policy debates have made an extensive use of modelling results about the costs of meeting a given climate objective. To what extent this endeavour succeeded in putting some rationale in discussions is not so obvious: first, from COP3 (Kyoto) to COP6 (The Hague), it did not manage to create a common understanding between the optimists and the pessimists about the costs of Kyoto targets; second, and per-haps more importantly, it failed, despite the attempts of the 2nd and 3rd reports of the International Panel on Climate Change (IPCC), to clarify what “good use” could be made of results discrepancies in designing a viable climate regime; third, it proved to be non convincing about any robust and consensual qualitative insights that could be derived from modelling exercises regarding the policy mix most likely to minimize welfare costs. Ultimately, negotiations were conducted under pure diplomatic rhetoric with almost no link to well-grounded, even though controversial, economic analysis.

Reporting the costs of carbon policies is thus currently made under the threat of a sequence of two sources of bias. The first source of bias is the very nature of the model from which the cost estimate is extracted. Three broad modelling paradigms - the disaggregated technology-rich bottom-up model, the multi-sector top-down model (be it based on micro-economics or econometrics) and the single-agent optimal control model - provide indeed three different types of cost, in close connection with the way in which they ac-count for production and end-use technologies. When attempting to compare the meaningfulness or legitimacy of these types of cost, a manner of trade-off appears between “explicitness” (how easy is it to embrace the explicit expenses summed up in these costs?) and “comprehensiveness” (to what extent do these costs express the full economic burden of a carbon policy?), with the necessary conclusion that all three assessment natures should participate to the careful weighing of any policy proposal.

Another source of bias lies in the choice of a reporting template, i.e. in the manner in which modelling results are eventually summed up. The number of dimensions to this choice opens a wide range of possible templates. To any subjectivity, it thus offers the tantalizing possibility to shape a message consistent with its ex-ante conviction, be it optimistic or pessimistic. Clarifying the comparison of two cost assessments expressed in different metrics does not raise conceptual difficulties: it simply requires to focus on one of the two metrics, and to translate the assessment that was not originally expressed following it. However, in most instances, this cannot be done ex-post, i.e. without re-running the model that produced it, an option seldom available.
Under this double threat, communicating the costs of climate policies appears quite a challenge, to both the scientific and decision-making communities indeed. On the one hand, experts should bring out rather than hush the complexities exposed in this paper, obviously taking the greatest care in the act to avoid raising more confusion than they dissipate. On the other hand, the decision-making community should accept that the answer to the question they raise is fundamentally more complex and less straightforward than they would want it to be, on the simple ground, ultimately, that it is manifold in itself.

Información relacionada

Reported by

Campus du Jardin tropical, 45 bis, avenue de la belle Gabrielle
94736 Nogent sur Marne
Síganos en: RSS Facebook Twitter YouTube Gestionado por la Oficina de Publicaciones de la UE Arriba