The CO2 Content of Consumption Across U.S. Regions: A Multi-Regional Input-Output (MRIO) Approach

Joint Program Reprint • Journal Article
The CO2 Content of Consumption Across U.S. Regions: A Multi-Regional Input-Output (MRIO) Approach
Caron, J., G. E. Metcalf and J. Reilly (2017)
Energy Journal, 38(1): 1-22

Reprint 2017-4 [Download]

Abstract/Summary:

How much will your cost of living rise if a price is put on carbon? The answer may depend on where you live—and how policymakers define who’s ultimately responsible for manmade carbon emissions.On first glance, it might seem intuitive to impose a price on carbon where emissions are generated, from manufacturing facilities to power plants. But none of those point sources would be operating without the end-users of the goods and services that they produce. Consider windshield glass made in Ohio that’s exported to Michigan for assembly into automobiles, which get shipped to New York auto dealers. If responsibility for carbon emissions embodied in that glass—through its manufacture, assembly and transport—is placed on the consumer, then a price on carbon would be imposed in New York.

Using a data-driven carbon emissions accounting strategy that’s far more comprehensive than earlier analyses, former MIT Joint Program postdoc Justin Caron and his coauthors—MIT Joint Program Co-Director John Reilly and Tufts University Professor of Economics Gilbert Metcalf—found that attributing CO2 emissions to states based on consumption rather than production vastly changes the total emissions for which they are responsible. The researchers also determined that the CO2 emissions embodied in the goods consumed by households vary widely among U.S. states and regions. According to the study, the majority of CO2 emissions attributable to households are not due to energy use (e.g., home electricity and heating fuel, gasoline) but rather embodied in all other goods consumed by those households.

Because the producers of imported goods—and hence their embodied emissions—vary considerably from state to state, some states are responsible for much more CO2 emissions than others, and would be more adversely impacted under a uniform carbon pricing policy. The authors determined that some of the largest net importers of embodied carbon are New York, Florida and California, as well as states in the New England and Mid-Atlantic regions, whereas Texas and the south-central and Mountain states are the largest net exporters.The differences in the carbon intensity of production and consumption across regions could impact support for carbon-pricing policies in different states within those regions, and whether those policies will be based on production or consumption. This study’s state-by-state estimates of CO2 emissions production and consumption could inform efforts to ensure that national and regional carbon-pricing policies don’t result in excessive economic hardship for particular states and regions. 

Citation:

Caron, J., G. E. Metcalf and J. Reilly (2017): The CO2 Content of Consumption Across U.S. Regions: A Multi-Regional Input-Output (MRIO) Approach. Energy Journal, 38(1): 1-22 (http://www.iaee.org/en/publications/ejarticle.aspx?id=2850)
  • Joint Program Reprint
  • Journal Article
The CO2 Content of Consumption Across U.S. Regions: A Multi-Regional Input-Output (MRIO) Approach

Caron, J., G. E. Metcalf and J. Reilly

2017-4
38(1): 1-22

Abstract/Summary: 

How much will your cost of living rise if a price is put on carbon? The answer may depend on where you live—and how policymakers define who’s ultimately responsible for manmade carbon emissions.On first glance, it might seem intuitive to impose a price on carbon where emissions are generated, from manufacturing facilities to power plants. But none of those point sources would be operating without the end-users of the goods and services that they produce. Consider windshield glass made in Ohio that’s exported to Michigan for assembly into automobiles, which get shipped to New York auto dealers. If responsibility for carbon emissions embodied in that glass—through its manufacture, assembly and transport—is placed on the consumer, then a price on carbon would be imposed in New York.

Using a data-driven carbon emissions accounting strategy that’s far more comprehensive than earlier analyses, former MIT Joint Program postdoc Justin Caron and his coauthors—MIT Joint Program Co-Director John Reilly and Tufts University Professor of Economics Gilbert Metcalf—found that attributing CO2 emissions to states based on consumption rather than production vastly changes the total emissions for which they are responsible. The researchers also determined that the CO2 emissions embodied in the goods consumed by households vary widely among U.S. states and regions. According to the study, the majority of CO2 emissions attributable to households are not due to energy use (e.g., home electricity and heating fuel, gasoline) but rather embodied in all other goods consumed by those households.

Because the producers of imported goods—and hence their embodied emissions—vary considerably from state to state, some states are responsible for much more CO2 emissions than others, and would be more adversely impacted under a uniform carbon pricing policy. The authors determined that some of the largest net importers of embodied carbon are New York, Florida and California, as well as states in the New England and Mid-Atlantic regions, whereas Texas and the south-central and Mountain states are the largest net exporters.The differences in the carbon intensity of production and consumption across regions could impact support for carbon-pricing policies in different states within those regions, and whether those policies will be based on production or consumption. This study’s state-by-state estimates of CO2 emissions production and consumption could inform efforts to ensure that national and regional carbon-pricing policies don’t result in excessive economic hardship for particular states and regions. 

Supersedes: 

The CO2 Content of Consumption Across US Regions: A Multi-Regional Input-Output (MRIO) Approach