Aggregate Emission Intensity Targets: Applications to the Paris Agreement

Abstract

We develop the concept of aggregate emission targets, which are goals for national emissions but do not dictate the forms of regulation used to achieve the goals. We compare aggregate emission intensity, quantity, and price targets adopted at the national level but implemented cost effectively at the firm level. We obtain simple ranking conditions that depend on the slope ratio of marginal emission damage and marginal abatement cost curves, and threshold parameters determined by the variance and covariance of GDP and business-as-usual emission. We apply the ranking conditions to the top 12 CO2 emitters with specific GHG targets in the Paris Agreement, and obtain a robust result that intensity targets dominate quantity targets for most of these nations. JEL: Q58

Price Limits in a Tradable Performance Standard

“…We note that Kellogg (2020) uses a similar model but focused on the TPS versus ETS question (or another type of indexed emission limit). Quirion (2005), Newell and Pizer (2008), Branger and Quirion (2014) and Zhao (2018) are similar in spirit to our paper. They focus on deriving general policy comparisons, including TPSs and feebates, though they focus on a national rather than sectoral level.…”

“…Quirion (2005) takes this ranking idea as far as possible based on theory. Zhao (2018), Branger and Quirion (2014) and Newell and Pizer (2008) provide welfare expressions but again focus on numerical applications that rank policies, rather than estimating the actual welfare difference.…”

Output and attribute-based carbon regulation under uncertainty

“…E now represents total lifetime gallons of gasoline consumed per vehicle, and I continue to measure private welfare B(Q, E, η, F ) as $ per vehicle, normalized to zero for agents' unconstrained choices at each F and η. 34 To calibrate the values of B QQ , B QE , and B EE , I first rewrite B(Q, E, η, F ) as a summation of objects that have empirical counterparts in the literature: the utility U (Q, η) from miles traveled, fuel costs −EF , and the vehicle cost 35 Taking derivatives then yields: 36…”

“…Realized utility for each vehicle will be a function of realized shocks. 35 This formulation for B(Q, E, η, F ) assumes that vehicle depreciation is entirely a function of time rather than mileage. To the extent that depreciation is a function of miles driven (and consumers account for depreciation when making driving choices), then a fuel economy policy behaves more like an emissions cap, and in fact becomes equivalent to an emissions cap in the limit in which depreciation is solely a function of mileage.…”

“…3 This literature, unlike the modeling framework I develop here, models intensity targets such that they do not generate output distortions. Zhao (2018), for instance, models a policy that adjusts an aggregate emissions cap each year in response to GDP shocks but is implemented as a standard cap-and-trade program, so that firms do not have an incentive to distort their output. Such a policy would dominate the distortion-inducing regulations that I model here.…”