No, A 2018 Study Doesn’t Show That Animal Products Account for 58 Percent of GHG Emissions

In early June of 2018, the  Guardian covered a recent article by Poore and Nemecek in Science. The Guardian coverage includes a graphic which states that farmed animal products contribute 58 percent of GHG gases. However, what isn’t made clear is that this is 58 percent of the GHG gases that are attributable to the food supply chain. This is made clear in the article by Poore and Nemecek, which states

“In particular, the impacts of animal products can markedly exceed those of vegetable substitutes (Fig. 1), to such a degree that meat, aquaculture,
eggs, and dairy use ~83% of the world’s farmland and contribute 56 to 58% of food’s different emissions, despite providing only 37% of our protein and
18% of our calories.” (pg. 4)

The authors further note

“Today’s food supply chain creates ~13.7 billion metric tons of carbon dioxide equivalents (CO2eq), 26% of anthropogenic GHG emissions.” (pg. 1)

This implies that animal products in the food supply chain account for approximately 15 percent of total anthropogenic emissions.

Cowspiracy’s Misleading “Infographic”

Summary: The Cowspiracy website includes an “Infographic” that provides misleading information. Under the Climate Change section, it indicates that 51 percent of greenhouse gas emissions are from livestock and their byproducts, while 13 percent are from transport. These numbers are clearly meant to be compared, but anyone familiar with how they were obtained will realize that such a comparison is inappropriate. The infographic also fails to clarify that studies looking at dietary changes focus only on the portion of the CO2 footprint that can be attributed to diet (rather than the total).

I’ve already indicated elsewhere on this site why the 51 percent estimate from Goodland and Anhang is meaningless. However, ignoring this issue for a moment, it is worth considering where this number and the estimate for transport come from. The 13 percent figure is from the World Resource Institute. Indeed this percentage can be seen in the linked chart. The total emissions estimate (41,755 MtCO2 equivalent) is also the starting point for the Goodland and Anhang analysis. This estimate, combined with the attribution, implies that the transport sector accounted for approximately 5,600 MtCO2e in the year 2000. Goodland and Anhang go on to argue that the estimate of total emissions undercounts livestock emissions by 22,048 MtCO2e, implying that the actually GHG inventory should be 63,803 MtCO2e.

If someone believes that the Goodland and Anhang analysis is valid, this implies that the share of emissions from the transport sector should be adjusted accordingly. Doing so implies that transport’s share of the total would be 8.8 percent (5,600/63,803 x 100%).

The Cowspiracy Infographic also claims that “A Carbon Based Diet Cuts Your Carbon Footprint by 50%”. The correct claim, which is made elsewhere on the Cowspiracy site, is “A person who follows a vegan diet produces the equivalent of 50% less carbon dioxide, uses 1/11th oil, 1/13th water, and 1/18th land compared to a meat-lover for their food.” However, this qualification is inconsistent with the earlier claim that livestock accounts for 51 percent of total GHG emissions.

Last edited on March 12, 2017.

Kip Andersen of Cowspiracy Gets It Wrong

The website has just published an excerpt from the book, The Sustainability Secret: Rethinking Our Diet to Transform the World. This book, by Keegan Kuhn and Kip Andersen, complements the documentary Cowspiracy. In the excerpt, Kip Andersen writes:

In 2009, Robert Goodland and Jeff Anhang, two environmental advisors to the World Bank Group, released an analysis on human-related greenhouse gases (pdf), concluding that animal agriculture was responsible not for 18 percent as the FAO stated, but was actually responsible for 51 percent of all greenhouse gases. Fifty-one percent. Yet all we hear about is burning fossil fuels.

This difference in the figures is due to factors that the FAO didn’t take into account, such as the massive loss of carbon sinks from clear-cutting rainforests for grazing in addition to the respiration and waste produced by animals. Goodland and Anhang used the Greenhouse Gas Protocol, the global standard for measuring emissions set by the World Resources Institute and World Business Council on Sustainable Development, to reach the figure of 51 percent. According to their calculations, animal agriculture is the number one contributor to human-caused climate change.

On the Cowspiracy website Keegan Kuhn makes the same claim, linking to the GHG Protocol for Cities. The single largest source of “uncounted” emissions, according to Goodland and Anhang, is animal respiration. However, a quick perusal of the GPC Standard reveals the following guidelines, which directly contradict claims that Goodland and Anhang’s analysis is consistent with the Protocol:

CO2 emissions from livestock are not estimated because annual net CO2 emissions are assumed to be zero—the CO2 photosynthesized by plants is returned to the atmosphere as respired CO2. (page 120)

In addition, the Greenhouse Gas Protocol Agricultural Guidance is available online and states:

The carbon incorporated into animal tissues or lost through animal respiration should not be reported in an inventory. (pg.62)

Goodland and Anhang obtain their second largest source of “uncounted” emissions by converting methane from the livestock sector into tons of carbon dioxide equivalent (CO2e) using the 20-year Global Warming Potential (GWP), while continuing to express methane from other sectors in CO2e using the lower 100-year GWPs. On this point, the Protocol indicates:

Individual GHGs should be converted into CO2e by multiplying by the 100-year GWP coefficients in the latest version of the IPCC Guidelines or the version used by the country’s national inventory body. (page 50)

The initial estimate of emissions of 41,755 MTon CO2e (for the year 2000) that is used by Goodland and Anhang is from the World Resources Institute (WRI) and follows IPCC guidelines for national GHG inventories. In this estimate, GHGs are converted to CO2e using 100-year GWPs. The GHG Protocol that Kip Andersen refers to is a standard for organizational and business (not national) inventory reporting. The WRI is one of the organizations that supports the Protocol. The claim that Goodland and Anhang’s analysis complies with the GHG Protocol is not supported by the Protocol’s own documents.

I have previously outlined various problems with the Goodland and Anhang analysis elsewhere on this site.

Appendum The Greenhouse Gas Protcol website summarizes the project as “A series of tools for calculating GHG emissions inventories based on 2006 IPCC Guidelines for National Greenhouse Gas Inventories.” It should not be surprising that the 2006 IPCC Guidelines also explicitly state the respiration should not be included (first paragraph of the introduction to Chapter 10).

Last edited on March 6, 2016 to include mention of the GHG Protocol support for the use of 100-year GWPs.

On Some Cowspiracy “Facts”

Summary: The Cowspiracy fact sheet claims that a vegan diet can reduce dietary emissions by 20 lbs CO2e per day. This figure is higher than recently published estimates and arises because (i) the authors do not replace animal products with vegan alternatives (and include their associated emissions) and (ii) the authors apply the emission factor for cheese, a GHG-intensive product, to all dairy products.

Although I haven’t seen Cowspiracy, coverage of the documentary on the blogosphere made it clear that Goodland and Anhang’s claim that livestock accounts for 51 percent of anthropogenic greenhouse gases (GHGs) emissions is highlighted in the documentary. I have documented problems with this analysis elsewhere. There is a website associated with the documentary that includes a page entitled “The Facts”. In addition to repeating the Goodland and Anhang claim, the page also makes the claim

Each day, a person who eats a vegan diet saves 1,100 gallons of water, 45 pounds of grain, 30 sq ft of forested land, 20 lbs CO2 equivalent, and one animal’s life.

I was interested in where the claim of 20 lbs CO2-equivalent (CO2e) came from, since this number is higher than what has been reported in a number of studies that have been highlighted in the media recently. The Cowspiracy site even cites Scarborough et al.(2014), which looks at UK emissions for different dietary types and concludes that the daily dietary emissions for heavy meat eaters are 7.19 kg CO2e (and 5.63 kg CO2e for medium meat eaters), while for vegans, the corresponding figures is 2.89 kg CO2e. Converting the difference into pounds yields a figure of 9.5 lbs (for medium meat eaters, the difference is 6 lbs). [The site also links to analysis by the Shrink that Footprint site, which uses US consumption data and concludes that compared to a vegan diet, the average American diet results in an extra 1 Ton CO2e/year, or 6.1 lb (2.74 kg) CO2/day and a “meat lover” diet results in an extra 1.8 tCO2e/year, or 10.9 lb (4.9 kg) CO2e/day.] Why is the figure on Cowspiracy twice the figure for heavy meat eaters (and over three times the figure for medium meat eaters)?

It turns out that the figure is not from a peer-reviewed source, but is instead a calculation based on dietary emissions by food type from the Environmental Working Group (EWG) and average consumption figures for the United States. Indeed, in footnote [xiv], the Cowspiracy fact sheet notes:

– CO2 based of feed conversion ratios and the average US meat consumption of 209lbs per year, per person.

Beef is at 22-27 kg CO2 Eq per kg produced/consumed X 2.5 ounces/day=1.75 kg or 3.85 pounds

Cheese/milk is 13.5 kg per kg product X 2 pounds/day=12.15 kg or 12.5 pounds

Pork is 12 kg per kg product X 2 ounces/day=.68 kg or 1.5 pounds

Combination chicken and turkey is 7 kg per kg product X 4.48 ounces/day= .89 kg or 1.96 pounds minimally (using only chicken)

{turkey, for instance, is 11 kg per kg product}

Eggs are at 5 kg per kg product X 2/3 egg per day= (50 g/egg) .55 pounds

— which equals 20.36 pounds of CO2 Eq saved per day.

There are two problems with this analysis. First, they don’t replace the animal products with vegan alternatives (which would have lower carbon emissions), which means even taking their analysis at face value, the saving would be less than they are claiming. A bigger problem has to do with their treatment of dairy products. The EWG provides emissions estimates for four different dairy products, measured in kg of CO2e per kg of product (reported in parentheses following each product) – cheese (13.5), butter (2.5), yogurt (2.2), and 2 percent milk (1.9). The Cowspiracy analysis elects to treat the entire 2 lb of dairy as cheese, which has the highest emission factor. The page doesn’t appear to provide an attribution for where the dairy figures are from, but it’s straightforward to find per capita annual figures on U.S. dairy consumption from the USDA. Figures are available on consumption for individual categories (milk and cream, yogurt, butter, cheese, etc.) and an aggregate measure based on “milk equivalent, milk fat basis”), which was 614 lb in 2012 (close to the 2 lb per day reported on the Cowspiracy fact sheet). In comparison, cheese consumption (American, Other and Cottage) was 35.6 lb (or about a tenth of a pound per day). There doesn’t seem to be any justification for using the emission factor for cheese, other than to inflate the figure. Indeed, the dairy category accounts for over 60 percent of the 20 lbs that they suggest can be saved, which is why their number is double that provided by Scarborough et al.

Edited Sept. 27, 2015 – a short summary was added to the article.

Edited Feb. 17, 2016 – added the emission figures from Shrink That Footprint.