Regional variation, holdouts, and climate treaty negotiations

See Barrett (2005) for a summary of the literature.

We choose to focus on an efficient solution to a global externality, but our model is sufficiently flexible to allow for any level of reduction in the externality.

We abstract from the treaty negotiation process and compliance issues to focus on identifying potential free riders. See Barrett (2005) and Barrett and Stavins (2003) for analysis of those issues.

We put off the form this treaty takes (emissions taxes or permit schemes) till the next section.

These sub-global climate coalitions and the associated free-rider problem is important in the context of international climate change negotiations. See Nagashima and Dellink (2008). This assumption allows us to detect potential free riders by identifying nations with positive net benefits under a climate agreement. If those countries attempt to hold out of an agreement, they will be recognized as free riders, presumably reducing their bargaining power. This assumption allows us to focus on the cost-benefit justification of international environmental agreements while putting aside the issues of external or internal stability that have been debated in the literature.

Linear MAC curves are consistent with the metastudy conducted by Fischer (2006). Ellerman (1998) finds quadratic functions fit most regional MACs very well, but (with the exception of Brazil) the coefficients on the quadratic term are very small.

To see this simply, separate the area of benefits into two pieces, the rectangle below the optimal tax rate (E^max–E^*)*MD_i(E^*) and the remaining area below the marginal damages curve and above the tax rate: Sum these two areas and then combine terms.

This equality holds comes from the definition of the efficient tax, which equates global marginal damages with global marginal costs.

D’Aspremont, Jacquemin, Gabszewicz, and Weymark (1983) first describe coalition stability in collusive price cartels, but the International Environmental Agreement literature has adopted this terminology as well. See Barrett (1994) for an early paper that analyzes coalition stability using a similar cost-benefit-style framework.

See Chander and Tulkens (1992) for an early example of how international transfers can help form cooperative agreements and Rotillon and Tazdat (1996) for an example of the form those transfers might take.

Of course, much of the difficulty in creating efficient international environmental policy is due to the lack of a social planner.

See Eyckmans and Tulkens (2003), Carraro, Eyckmans, and Finus (2006, p. 3), and Carraro and Siniscalco (1993). This ensures that utility is transferable across countries, meaning that transfers are equally weighted. It would be straightforward to extend this analysis to unequal weights using a social welfare matrix that weights transfers. This approach ignores numerous equity and political issues with these transfers that are beyond the scope of this paper.

The construction of the marginal damage and marginal benefits curves ensure that the last unit of pollution abated generates the least net benefits. This means any reduction in the environmental tax will increase the average net benefits of emissions reductions.

See chapter 6 of Stern (2006) for descriptions of the issues and techniques used in Integrated Assessment Modeling and the current state of the IAM literature as well as Tol (2009) for a meta-analysis of recent models.

Detailed regional definitions for the WITCH model are detailed in Table 1.

Climate damages are a function of the stock of pollution, while emissions are directly related to abatement costs. We use the terms abatement and emissions avoided interchangeably.

It is important to note that these estimates are sensitive to the IAMs used to calculate marginal benefits and marginal damages. If these models are incorrect about the distribution of future damages from climate change, that error will be propagated in our estimates.

Because of the relatively consistent emissions reductions estimates, the results are not sensitive to using other plausible measures.

The assumption is conservative in the sense that constant marginal damages minimizes the total benefits of joining a climate change treaty to lower emissions to a given level. Any country that would agree to a climate change treaty under constant marginal damages would agree if its marginal damage function was increasing.

It should be noted that these estimates are based on the conservative assumptions laid out above. As the slope of the marginal damage curve increases, there are thresholds over which both countries become better off under a globally efficient climate treaty. If the damages curve for China is steep enough so that MD_i(E^max) is 10.5 times greater than MD(E^*), then China would be better off under a global climate change treaty even if its share of global damages is only 2.2%. For the US, the comparable figure is 3.9. If the unregulated level of damages exceeds the domestically efficient level by more than 3.9 times, then the US would be better off under a global climate treaty using even the smallest share of global benefits as the basis for our estimate.

See Eyckmans and Tulkens (2003), Carraro et al. (2006), and Germain, Toint, Tulkens, and de Zeeuw (2003), among many others.

For example, marginal damage curves could shift due to better scientific understanding of the damages of climate change or regime shifts due to increasing stock of pollutants. Similarly marginal abatement costs could shift due to technological breakthroughs in carbon capture and sequestration or geoengineering.