Blogs are opinion pieces and reflect their author’s views

Epicurious Takes a Stand To Stop Global Warming – Where’s the Beef?

Written by: Karen Spencer

Simpson Centre New Growth: Reviews of Current Ag Policy Research and Articles

Source: Science

Summary:    A recent post from the website “Epicurious” outlined their stand on beef, specifically choosing to exclude any new beef recipes on their website due to the relatively high greenhouse gas (GHG) emissions per kilo compared to plant protein sources, and even compared to other meat proteins. In our blog, we dive into the research behind the Epicurious decision, as their post directs the reader to some interesting evidence-based analysis for us to review.


The original Epicurious article (David Tamarkin and Maggie Hoffman, “The Planet on the Plate: Why Epicurious Left Beef Behind,” Epicurious, April 26, 2021. outlines the website’s approach, which has been to reduce its number of beef articles since 2019, and finally now to set that number of new beef recipes to zero. While the website commits to this benchmark, they honor that there is demand for the product, and indeed, their lead article for Memorial Day is “50 Memorial Day Grilling Recipes for 2021,” which includes a number of beef recipes. ( accessed May 29, 2021).

The editors at Epicurious direct the reader to a number of other resources, one of which is a 2019 journal article by Poore and Nemecek (see reference below). This blog summarizes that journal article and its findings. The journal article presents a globally reconciled and methodologically harmonized database on the variation in five environmental impact indicators: life cycle GHG emissions, land use, acidifying emissions, eutrophying emissions, and freshwater withdrawals weighted by local water scarcity. The authors consider these five indicators when analyzing 40 products representing about 90% of global protein and calorie consumption, and begins with primary agriculture inputs, ending at retail, the point of consumer choice. The study’s meta-analyses of other studies include a weighting of data by the share of national production it represents, and each country by its share of global production. Data is validated by comparing to Food and Agriculture Organization (FAO) data.

Today’s agriculture system, feeding 7.6 billion people globally, is resource intensive, covering 43% of the world’s ice- and desert-free land. The authors estimate that two thirds of freshwater withdrawals globally are for irrigation. Food production creates 32% of global terrestrial acidification and 78% of global eutrophication (a dense growth of plant life in a water body due to an increasing amount of fertilizers migrated from runoff, that chokes out other growth through eliminating the oxygen in the environment). The primary food production stage represents 61% of food’s GHG emissions, 79% of acidification and 95% of eutrophication.

The authors group products by their primary dietary roles: protein sources, milks, starches, oils, vegetables, fruits, sugars, alcoholic beverages, chocolate and coffee. They then express environmental impacts per unit of primary nutritional benefit for the 40 products considered.

Major observations include a high variation in impact when comparing products, as expected, but also among producers of a single product. For example, the ninetieth-percentile GHG emissions of beef are 105 kg CO2eq per 100 g of protein, and land use (area multiplied by years occupied) is 370 m2*year. The tenth-percentile GHG emissions for beef are 20 kg CO2eq per 100 g of protein and corresponding land use is 42 m2*year, representing a wide variation. The authors show tenth- and ninetieth-percentile GHG emissions for dairy beef of 9 and 26 kg CO2eq per 100 g of protein, respectively, almost a magnitude less than that of dedicated beef herds. Across other products there are similar ranges, many not as marked as that for beef but still significant. The tenth-percentile and ninetieth-percentile GHG emissions for wheat and rye bread are 0.3 and 0.9 kg CO2eq per one kg, while the comparative figures are 0.4 and 2.4 kg CO2eq per one kg for rice. The tenth-percentile and ninetieth-percentile GHG emissions of peas is 0.3 and 0.8 kg of CO2eq per 100 g of protein.

In case you got lost in all those numbers, the mean GHG emissions for 100 g of protein from beef is 50 kg CO2eq while the same figure for 100 g of protein from peas is 0.4 kg CO2eq. In addition to lower GHG emissions, the 100 g of pea protein also has a smaller land footprint and lower acidification and eutrophication impacts.

The authors note for many products, the impacts are skewed by large producers with high impact value chains. For example, the highest impact 25% of beef herds globally represent 56% of the beef herd’s GHG emissions and the source of 61% of their land use impacts. This can help create a targeted emissions reduction approach.

The authors then review the related data considered, monitoring multiple impacts and seeking relationships with countries and specific products. They note that specific practices and geographical data are required to come directly from producers rather than satellite or census data in order to capture the variation among farms. They conclude setting regional and sector-specific targets will help producers navigate trade-offs and make choices that align with both local and global priorities. The authors also point out that the environmental outcomes of many practices such as conservation agriculture and organic farming are highly variable. They therefore recommend providing producers with multiple ways to reduce their environmental impacts, suggesting many options that specific producers can choose as the best combination for their specific circumstances.

The article notes some changes can be valuable for all producers to pursue. They suggest methane from flooded rice, enteric methane from ruminants, and concentrate feed for pigs and poultry are sizeable globally and represent 30% of food’s GHG emissions. As these are material for all producers, they can be mitigated with actions such as shorter and shallower rice flooding, improving degraded pasture, and improving lifetime animal productivity, developing a “best in class” approach. As discussed earlier, within these specific products there is also a wide range of emissions and impacts, indicating that emissions can be reduced more quickly through targeting higher intensity emitters. Emissions from deforestation and cultivated organic soils contribute a large part of each product’s GHG emissions, so efforts to curb forest loss and cultivation limits on peatlands are also recommended overall.

Impacts further along the supply chain also apply to all products and producers. Processing, more durable packaging, and greater usage of coproducts can reduce food waste. For beef, distribution and retail losses contribute 12 to 15% of emissions, providing more incentive to reduce waste.

Although low-impact products are an aspirational goal, this has limitations. The authors show that despite one-fifth of 2017 palm oil production being certified sustainable, there is virtually no demand for that certified product in China, India and Indonesia. Organic foods pass premiums on to consumers and this limits total market size and widespread practice change. In addition, there is large concentration in marketing and distribution with just ten retailers providing 52% of U.S. grocery sales and 15% of global sales. For change to happen, these global retailers must be a part of the process.

The authors close with a note on the role that consumers can play. If the environmental impacts of the lowest-impact animal products is substantially higher than that of substitute vegetable proteins, then a change in diet demand can help reduce the food system’s environmental impacts.

In a theoretical case, they estimate if all animal products are removed from our diets globally, including all dairy, eggs and seafood, the global food GHG emissions would be reduced by 6.6 billion tonnes CO2eq annually, a 49% reduction in the global food system GHG emissions, plus other major reductions in acidification, eutrophication and freshwater withdrawal impacts. This model assumes new vegetable proteins replace all removed calorie and protein needs, using the estimated impacts for those food sources to replace meat proteins. It assumes the land no longer required for food production could be used to remove tons of CO2eq over the years as well as natural vegetation is re-established and soil carbon re-accumulates. This model assumes much that is theoretical only, and does not include any implications of economic impacts, vast changes in diet in both developing and developed countries, and overall strategies and policies required to move toward this. In Canada, this would imply all livestock farmers, and the entire supply management system, would be replaced with soybean, peas and other lower intensity protein products.

In a second scenario, the authors consider all animal products, and replace the half with above-median GHG emissions with vegetable equivalents in their model. Meat production and consumption is focused on lower emissions strategies and diet habits are changed, but not in a manner as abrupt as the first example case. This case achieves 71% of GHG reductions of the previous scenario with less impact on consumers.

The authors summarize the overall steps they recommend in aiming for lower GHG emissions and other environmental impacts from the global food system. First, farmers need to have a proxy or method of measuring their emissions and other impacts – they need to know where they are starting, and how their actions are making an impact. Second, policy makers should set targets and incentivize producers to change their activities by providing credit, tax breaks or reallocation of agricultural subsidies. And third, an assessment tool has to be used to consider multiple mitigation and productivity and enhancement strategies for producers. Much of the information on impacts would be communicated through the supply chain to consumers. We have good traceability with our livestock at this time, but grains, vegetables and other products are not as transparent. Communication could occur through a combination of environmental labels, taxes or subsidies designed to reflect environmental costs in product prices, and broader education of the true cost of food. The paper does not reflect on strategies to induce consumers to substantially change their diets, other than educating and appealing to the wish to improve their environmental footprint.

Back to the Epicurious article: It appeals to its audience, potentially western consumers, and their interest in changing their environmental footprint in one aspect of their lives. It does not pretend to tackle more complex issues than that, and it backs its claims up with some solid evidence.

For a deep dive into the large amount of reports and corresponding data gathered and synthesized by the authors, please look at the full article, noted below in references.

Dietary change can help in jurisdictions that have above average beef consumption, but in addition mitigation actions are complex due to trade-offs, multiple ways for producers to reduce impacts, and interactions throughout the supply chain. The research suggests a multi-faceted approach where producers monitor their own impacts, flexibly meet environmental targets by choosing from multiple practices, and communicate their impacts to consumers.


Poore, J. and Nemecek, T. “Reducing food’s environmental impacts through producers and consumers.” Science 360 (6392), 987-992, originally published May 31, 2018, Erratum February 22, 2019.

Learn more about The Simpson Centre here