We disagree that it is useful to include more behavioural and political science in global integrated assessment models (IAMs; see W. Peng et al. Nature 594, 174–176; 2021). Many of the features the authors suggest have already been reflected in several IAMs (see, for instance, M. G. Morgan and H. Dowlatabadi Climat. Change 34, 337–368; 1996). But high degrees of uncertainty mean that these cannot provide useful quantitative conclusions about policy at the local level.
For instance, in 2000, we used autonomous agents with pro- and anti-climate policies in 13 global regions to study policy formation and dissolution under public pressure from extreme events and high taxes. The key insights were that uncertainties and regional diversity in decision-making and values are too broad and varied for global models to inform specifics in policies.
Regional adherence to global targets requires nuanced strategies at the sub-national scale, reflecting local politics and technical, economic and behavioural aspects of specific problems — just as we need downscaling for detailed climate projections. Consequently, we have spent the past two decades focusing on analysis for specific sector and transition challenges. We urge others to do the same.
Nature 595, 650 (2021)
The authors declare no competing interests.