"Coercive Trade Agreements for Supplying Global Public Goods"
Scott Barrett, Lenfest-Earth Institute Professor of Natural Resource Economics, School of International and Public Affairs, Columbia University.
David Keith, Faculty Associate. Gordon McKay Professor of Applied Physics, School of Engineering and Applied Sciences (SEAS); Professor of Public Policy, Kennedy School of Government, Harvard University.
We develop a model of cooperation on trade and (in keeping with the Paris Agreement) of voluntary cooperation on climate change, and ask whether it is better to keep these issues separate or to link them, by making cooperation on trade conditional on supplying the global public good of reducing greenhouse gas emissions. Nordhaus (2015) has shown that, depending on the social cost of carbon and tariff level, linkage may enable countries to cooperate fully on both trade and climate change, or it may only sustain partial cooperation on both issues, or it may be of no help at all. However, his approach is numerical, and conceals the full nature of the games being played. He also assumes that all countries commit not to retaliate against members of the “Climate Club” that impose punitive tariffs on non-members. Our simple model, by contrast, reveals the full nature of the linked game and the consequences of this assumption about retaliation. We show that if retaliation is prohibited, then linkage gives rise to one of four possible situations (depending on parameter values): a cooperation game, a coordination game, a chicken game, or a prisoners’ dilemma. By contrast, if retaliation is a choice, linkage gives rise to one of only two possible situations (again, depending on parameter values): a coordination game or a prisoners’ dilemma. We find that the conditions that enable linkage to increase cooperation on climate change are more restrictive than suggested by Nordhaus. Moreover, we show that even when linkage could improve climate cooperation, its adoption is not assured. In particular, we show that the payoff dominant Nash equilibrium of the coordination game cannot also be risk dominant, meaning that equilibrium selection is not assured. This in turn means that the procedure for determining whether to link is crucial. In Nordhaus’s paper, the decision to link is made by a computer algorithm. Our paper will study this choice in the lab (our experiment will be conducted soon).