Modeling Industrial Sector Greenhouse Gas Emissions Reduction Policies with Plant-Level Data

The industrial sector is an increasingly important area of focus for national-level greenhouse gas (GHG) emissions mitigation policies. Modeling specific industrial sector mitigation options requires data to estimate both benchmark emissions in the absence of the policies and marginal abatement costs.

Major emitting manufacturing industries in the United States include iron and steel, cement, pulp and paper, petroleum refining, organic chemicals, and nitrogenous fertilizer. These industries use complex production processes, manufacturing a wide array of products involving multiple technologies and inputs. Unlike the power sector, where much of the technical and economic information needed to construct credible models is in the public domain, much of the relevant industry data are proprietary, with limited public access. The most comprehensive industrial data sources for the United States are the Economic Censuses conducted every five years, although the publicly available information is limited.

A potential solution to these limitations is via access to the confidential plant-level data collected by the Census Bureau and available to qualified researchers. These data permit examination of the range of emissions from individual plants within an industry as well as changes in those emissions over time. The data from Economic Census years also contain information on physical output quantities for some products manufactured by these industries.

This paper reports on our efforts to develop the databases and estimate separate models for each of the six noted industries and includes both descriptive and econometric analyses, along with simulations of alternative policies.

The descriptive analysis reveals substantial within-industry heterogeneity in GHG emissions intensity, measured as the ratio of GHG emissions to output. This heterogeneity reflects plant-level differences in technologies, fuels, specific products manufactured, and other factors. In the econometric analysis, energy prices affect energy intensity and profits, and the responsiveness varies substantially across industries. For example, the response of GHG intensity (not total emissions reduction)to changes in fuel prices is about twice as large in the petroleum and cement industries as in the other industries studied.

Based on the econometric results, the simulations reveal that carbon prices of $20 per metric ton of GHG reduce emissions by 8–32 percent across the industries considered. These reductions reflect short- and medium-run responses such as input substitution and limited technology adoption. The total emissions reductions vary across industries in accordance with their estimated responsiveness to energy prices; we estimate the largest total emissions reductions for cement and fertilizers and the smallest total emissions reductions for pulp and paper.