Corporate Average Fuel Economy (CAFE) standards in the United States are designed to increase the fuel economy of cars (targeted characteristic), but they take the size of each car (secondary attribute) into consideration. Firms making larger cars are allowed to have lower average fuel economy. Fuel-economy regulations are attribute-based in the world’s four largest car markets—China, Europe, Japan and the United States.

The goal of this paper is to investigate the welfare consequences of attribute-based regulations (ABR), as opposed to regulations based only on the targeted characteristic. Despite the ubiquity of attribute-based policies, the economics literature has not established theoretical and empirical frameworks for the analysis of this important class of policies. In this paper, we first develop a theoretical model that identifies the key parameters that determine the costs and benefits of attribute basing. We then explore two empirical methods that enable us to estimate those key parameters. Our empirical analysis exploits quasi-experiments in attribute-based Japanese fuel-economy regulations, the features of which provide several empirical advantages for estimating the welfare effects of ABR.

In brief, we conclude that it is unlikely that attribute basing is justified on eciency grounds. We identify conditions under which attribute basing has eciency benefits, but these same benefits could be achieved through compliance trading schemes without incurring distortions associated with attribute basing. Attribute basing is an imperfect substitute for compliance trading, and it is justified only if there is some constraint that prevents trading. Instead, we suspect that many ABR are motivated by distributional considerations. In this case, the distortions induced by attribute-basing represent the cost of achieving desired redistribution.

The Economics of Attribute-Based Regulation: Theory and Evidence from Fuel-Economy Standards
Review of Economics and Statistics, 100 (2): 319-336, 2018. (with Jim Sallee)

For more details, please download Paper and Slides (also browsable below).