I’ve got another bad stats (in this case, a strawman hypothesis) take-down on the subject of dark matter and comet showers in the pipeline for next week (or perhaps after Christmas), but another case of bad science of the clickbait variety has bumped itself to top billing for now. In particular, I refer to the study, “Energy use, blue water footprint, and greenhouse gas emissions for current food consumption patterns and dietary recommendations in the US” by Tom, Fischbeck & Hendrickson of Carnegie Mellon, which you may know better from the newspapers as that study that proved eating a vegetarian diet is worse for the environment than eating meat!
Interestingly, the article itself says nothing about aiming to investigate the relative merits of vegetarianism versus meat consumption, rather the authors attempt to cost the impact of three dietary scenarios—all including meat, poultry, fish, dairy, and eggs—two (at different calorie levels) with a food mix in proportion to that recommended by the USDA (see the three scenarios described in their Supplementary Material). It turns out that with the authors’ particular costing estimates (not necessarily applicable outside of a Californian setting) that the recommended ‘healthy’ diets have similar or greater externalities than the current mix with reduced calories.
Unfortunately, with this conclusion the study was only ever likely to pique the interests of fellow scientists in the field, so somewhere along the line someone decided to sex things up a bit. Check out the CMU press release: “VEGETARIAN AND ‘HEALTHY’ DIETS COULD BE MORE HARMFUL TO THE ENVIRONMENT: Carnegie Mellon Study Finds Eating Lettuce Is More Than Three Times Worse in Greenhouse Gas Emissions Than Eating Bacon“. Where the f*#k did that come from? Well, it’s a quote directly from the study’s second author, Paul Fishoeder whose research (ironically) is described on his webpage as focussing on “the quantification and communication of uncertainty”.
Word on the street is that it wasn’t just a press release cock up, but Fishoeder himself who plumped for the clickbait lettuce angle. His additional quote from the Washington Post, “You can’t just assume that a vegetarian diet will reduce your carbon footprint, which is what people think” certainly backs up the idea that he’s personally spouting off about vegetarianism. If he’d done a study comparing vegetarian diets to carnivorous diets then fair enough, but he’s not done a thing like that here: he’s wildly extrapolating from a comparison of three meat based diets. However, you wouldn’t know this from the press release which characterises the ‘healthy’ diets studied as “a mix of fruits, vegetables, dairy and seafood” (conveniently omitting the meat and poultry). With the main paper being behind paywall the average journalist (let alone a clickbait whore) might be excused for taking away the wrong message here!
Even with access to the paper it takes a fair bit of reading to figure out what the actual diets studied were composed of. Figure 1 of the main paper contains a classic ‘lying with graphs’ example (not to suggest that this was the authors’ intent) in which only the relative difference (compared to the status quo) in calorific intake between the food groups in each diet is displayed instead of the total calorific intake by food group which would allow the reader to simultaneously compare diets and understand their absolute composition.
With this tripe being picked up by news papers and websites across the globe and gleefully seized upon by meat eaters reluctant to consider even going a meat-free Monday for the environment one wonders how the scientific community might find a way to censure our clickbait feeding colleagues, or do we just have to wait for them to desist of their own accord having shat all over what remains of public trust in science in the meanwhile?