The voluntary provision of a pure public good? Another look at CFC emissions and the Montreal Protocol

Based on their finding of a positive and nearly linear relationship between GNP and reductions of chlorofluorocarbon (CFC) emissions in the run-up to the Montreal Protocol, Murdoch and Sandler (1997) have argued that the treaty's initial emission targets were consistent with voluntary provision of abatement rather than cooperative behavior. This paper documents that their analysis relies on largely imputed emission data that overstate emission reductions compared to emission data that countries reported to UNEP. The imputation procedure appears to induce a spurious positive correlation between income and CFC reductions. In a replication of the econometric analysis using UNEP data, the hypothesis of a positive and linear relationship between the two variables is rejected. These findings call for a more cautious interpretation of emission targets set by the Montreal Protocol and alert researchers to important limitations of a data set that has been widely used in empirical studies of this treaty.