The Aggregation Dilemma in Climate Change Policy Evaluation

We show that a policy maker who ignores regional data and instead relies on aggregated integrated assessment models is likely underestimating the carbon price and thus the required climate policy. Based on a simple theoretical model we give conditions under which the Aggregation Dilemma is expected to play a role in climate change cost-benefit analysis. We then study the importance of the Aggregation Dilemma with the integrated assessment model RICE (Nordhaus and Boyer 2000). Aggregating all regions of the RICE-99 model into one region yields a 40% lower social cost of carbon than the RICE model itself predicts. Based on extrapolating the results a country-level integrated assessment model would give a more than eight times higher social cost of carbon compared to a fully aggregated model. We suggest that these tentative results require researchers to re-think the aggregation level used in integrated assessment models and to develop models at much lower levels of aggregation than currently available.