Cowspiracy: fact-checking (2)

With the outlandish statistics and claims made by the documentary Cowspiracy, naturally people would want to know whether the content is scientifically true or a vegan propaganda. This is especially true this year, where the word of the year declared by Oxford dictionary is ‘Post-truth’. Personally, I have to admit nowadays I find it increasingly more difficult to differentiate whether a claim is fuelled with ventured political motives or backed by concrete scientific evidence. However, through deconstructing the so-called 51% GHG emission by agriculture sector statistics, I have partly restored some faith in my ability to critically analyse the validity of the content. Admittedly, have it not been part of an assignment, I would not have the time and effort to look into it and could very well have been persuaded and simply take what the documentary said at face value.

Confronted with the enormity of information and statistics (some of which are contradicting) during my research, whether in mainstream media or peer-reviewed scientific journal articles, I decided to turn to reviews that systematically evaluate the environmental impact of different dietary choices. Particularly I found two papers (Hallstromet al, 2014;) that are very helpful in providing a more systematic assessment of the environmental impact of different dietary choices. They offer a much more holistic view of the impact of different food and their production and consumption cycle and, most importantly, explained the variation of statistics that I have encountered.

Let’s lay down some facts and figures first then. Hallstromet al (2014)’s investigation of 14 peer-reviewed journal articles showed that out of all the different dietary choices, adopting a vegan diet is indeed the best way to reduce carbon footprint. See Figure1 and Table1. According to the researches that are included in this study, there is a potential of reducing the carbon footprint by 20% - 55%. Although huge variation exists, most studies do point into the same direction. In Auestad and Fulgoni (2015), their review also suggested similar result. For example, Berber-Lee et al (2012) suggested that the UK could reduce its food-related emission by almost 25%, along with some potential health benefit if everyone were to adopt a vegetarian or vegan diet. Aston et al (2012) expressed the same view as well. (Phew, I am glad the argument on that is settled, at least in the context of UK.)

However, there are many caveats in the previous paragraph.

Firstly, it is important to note the definition and language used in the studies. For example, the 25% reduction in Berber-Lee et al (2012) is referring to food-related emission, not emission-per-capita and if the two different terms were to used interchangeably, two very different stories would be told. It begs the question of how much carbon emission of each person is attributed to food-related activities. A vegan who overconsumes with many food waste vs an omnivore who eats appropriate amount of calories needed and minimise food waste could make the comparison of carbon footprint a lot more nuance than it first appears. Therefore, it is important to understand the difference in the units used for dietary comparison. 

Secondly, as both reviews rightly pointed out, there are inconsistencies in methods used to determine GHG emission. Although most studies have adopted life cycle assessment (LCA) as the primary method to assess the carbon footprint of different diets, the ‘absence of LCA databases for representative foods in the marketplace’ means methodologic variability exists among different studies. For example, in 22 and 23, their data were extrapolated from other countries as no domestic data available. Additionally, different aspects of food production were included in many studies, leading to different results and interpretation. Should we include the GHGe from only farming itself like in 16, farm-to-retail like in 12 or farm-to-home like in 13 and 27 (i.e. including aspects such as transportation, storage, methods of cooking, and amount of food waste)? Therefore, we have to be very careful when making comparison and not extrapolate one statistic or study done in one geographic location into another.

Thirdly, there are indeed many other aspects that the measure of carbon footprint alone does not encompass. What about land use, water use, chemical pollution, food waste, food miles of different agricultural products? For example, despite Cowspiracy’s claim that cow milk requires more water than almond milk, some argued that as cow milk largely relies on green water (i.e. water from rainfall in the forages that cows are consuming) which cannot be consumed by human directly without treatment, it is actually more sustainable than the former.

Overall, I think it is dangerous to make reductionist assumptions about the benefits or the harm of a particular agricultural product. There are so many different ways of unit representation (which are confusing to geographers/ scientists already, let alone some journalists whose aim is to make sensational headlines and maximise their views as opposed to providing scientifically qualified content), so many aspects of food life cycle and their respective environmental impacts. We ought to be more cautious of the sensational headlines and claims, especially in this day and age where so many fake news prevails.