Survival of the Cleanest? Evidence from a Plant Level Analysis of Pollutant Emissions in Canadian Pulp

In this paper we capitalize on our access to plant level data in examining the relative changes in emissions of three major pollutants | Biochemical Oxygen Demand (BOD), Total Suspended Solids (TSS), and Greenhouse Gas (GHG) | for Canadian pulp and paper mills covering the period from 2005 to 2013. Over this eight-year period emissions of these three pollutants decreased by 31.0%, 35.5% and 42.5%, respectively, for the plants included in our sample. Access to plant data allows us to investigate the roles played by some specific factors, such as changes in output, emission intensity, allocation of production among surviving plants, and plant closures, in contributing to the abrupt decline in emissions. Information on fairly homogeneous groups of plants, formed on the basis of production processes and output mixes, reveals a diverse picture of the roles played by these factors. For our analytical framework we adapt the factor decomposition technique proposed by Levinson (2015) to plant data. Our findings suggest that output change has been the main factor behind the reduction in emissions at the industry level, and that improvement of emission intensity by surviving plants | the so-called technique effect | brought a small, yet positive contribution. However, production shift among surviving plants and plant closure had almost no effects. Moreover, there are no indications that market operations determining plant output and plant survival led to lower emissions.