All corrections
LessWrong May 12, 2026 at 01:44 PM

www.lesswrong.com/posts/PmaBozEroBg73a5Di/the-epsilon-fallacy

1 correction found

1
Claim
Practically all of the reduction in US carbon emissions over the past 10 years has come from that shift.
Correction

Coal-to-gas switching was important, but studies and EIA analyses do not show it caused "practically all" of the U.S. emissions decline; other factors such as renewables and slower electricity-demand growth accounted for large shares too.

Full reasoning

This claim overstates the role of coal-to-gas switching.

The post was published on March 17, 2018, so its "past 10 years" claim refers roughly to the late-2000s through 2018. For that period, peer-reviewed and government analyses do not find that natural-gas substitution explains nearly all of the U.S. carbon decline.

  • A 2016 Nature Communications reanalysis of U.S. emissions trends found that fuel switching to natural gas accounted for 28.2% of the overall U.S. emissions reduction from 2007 to 2013 by one method, and 20.4% to 40.7% by another.
  • An EIA analysis published in October 2018 said that from 2005 to 2017, slower electricity-demand growth and changes in the electricity generation mix played nearly equal roles in reducing U.S. power-sector CO2 emissions.
  • Another EIA breakdown published in June 2021 found that, from 2005 to 2019, about 65% of the power-sector CO2 decline came from coal-to-gas switching and about 30% came from increased renewable generation.

So natural gas was a major contributor, but the evidence does not support saying it caused "practically all" of the reduction. A substantial share came from other drivers, especially renewables and slower growth in electricity demand.

3 sources
Model: OPENAI_GPT_5 Prompt: v1.16.0