An experiment on emissions trading: the effect of different allocation mechanisms

In theory, efficiency and compliance levels induced by an emission trading system should not depend on the initial allocation mechanism for permits in the absence of transaction costs. In a laboratory experiment we investigate this prediction by comparing frequent and infrequent auctioning as well as two different grandfathering schemes under market rules that closely resemble those of the European Union Emission Trading System. Our experimental results suggest that, contrary to theoretical predictions, the initial allocation procedure has the potential to affect efficiency of the final permit allocation. While we do not identify an effect of the initial allocation procedure itself (auction vs. grandfathering), we observe higher final efficiency after infrequent auctioning of permits than for frequent auctioning. Surprisingly, for a grandfathering scheme that distributes permits proportional to expected needs the high initial efficiency is substantially reduced by secondary market trading. An analysis of behavioral patterns shows that permit prices and abatement levels are initially substantially higher if permits are allocated by auction and we also find more over-banking as compared to the grandfathering treatments. Treatment differences diminish in the course of the experiment.