The Measure is the Message
The Vietnam War and the perils of making food into the next climate change battleground
Go vegan to save the planet! That is the message of countless infographics that circulate social media. Consider this graph from Our World in Data, where beef’s carbon footprint is almost off the chart.
It would seem as if having any environmental commitments at all would leave our hands tied. Shut down the stockyards! Revoke the grazing licenses! Bring on the lab fermented protein!
But the process of picking measures for climate harm (and choosing what goes into those calculations) leaves a lot of room for subjective assumptions and choices. At worst, numbers and metrics are a strategy for laundering values behind the guise of scientific objectivity. Quantitative indexes are politics by other means.
History warns us of the perils of putting excessive faith in what might appear at first to be straightforward quantification. One of the most egregious examples is the Kennedy and Johnson administrations’ approach to the Vietnam War. Secretary of War Robert McNamara was a fan of operations research, having applied mathematical modeling to try to optimize operations as head of the Ford Motor Company.
Although traditional military analysis painted a dismal picture of the prospects for winning, McNamara’s technocrats developed models to suggest the opposite. They pieced together different measures in the effort to develop a “victory index” representing military progress. “Body count” was one of the main data points, and others were fed into the “Hamlet Evaluation System.” Some 90 thousand pages were produced every month, rating thousands of villages on their level of “pacification.”
The embarrassing loss in Vietnam is a story of hubris, illustrating the risks of fetishizing figures and ignoring what numbers can’t tell you. A speechwriter for Kennedy and Johnson later recalled, “No expert on Vietnamese culture sat at the conference table.” An apocryphal joke is often told about the department’s utter derangement, about a computer in the basement of the Pentagon being fed with mountains of data. Leaving the program to predict the end of the war over the weekend, Generals returned on Monday morning to a punch card alerting them to the fact that they had already won in 1965.
The same kind of risk is present as we look to mathematical indexes, models, and scenarios for guidance concerning environmental problems. Looking deeper we may find that the chosen criteria carry with them problematic assumptions, and that ever more complex models tell us what we want to hear, but only because our subjective values have already been baked into the analysis.
If we look back at the chart from Our World in Data, the x-axis immediately stand outs. Does it really make sense to calculate carbon emissions per kilogram of food? It is immediately obvious that a pound of apples isn’t the same as the equal weight of steak.
Other charts will instead graph emissions with respect to grams of protein or calories. That’s a step in the right direction, but all protein isn’t equal. And neither are calories. Animal-based proteins are generally more complete, come with B vitamins and iron, and are more efficiently digested by the body. To calculate a food’s carbon emissions by the kilogram, calories, or even raw protein really just sneaks vegan values in through the backdoor.
A recent paper from a team led by Ryan Katz-Rosene takes another approach, leveling foods according to their levels of essential micronutrients: vitamin A, folate, vitamin B12, etc. While doing so doesn’t totally absolve beef, it does present a very different picture than the Our World in Data figure. The worst offenders are now palm oil, olive oil, and dark chocolate. Beef is high in relative emissions, but not especially so. Most interestingly, cow milk and eggs look pretty good, being roughly as carbon intensive as foods like soymilk, apples, tofu, and oats, which are staples of a plant-based diet.
Of course, Katz-Rosene’s paper raises as many questions as it answers. Which micronutrients should get the most weight? Should we value low-carb foods higher, given that man people seem to better manage their weight by avoiding bread and other starches?
More importantly, which estimates of carbon emissions should we use? Beef production in the United States has a fraction of the carbon footprint of meat produced elsewhere. Some studies suggest that if a person got their beef from a low-impact grazing operations, the carbon footprint is essentially zero (i.e., no different from eating a wild herbivore).
Directly engaging with such questions enables a far more productive political discussion. Assumptions are out in the open, forcing us to more directly reckon with complexities regarding what makes for a “good” human diet.
Environmentalists would do well to learn from McNamara’s mistake, to see the perils of trying to calculate their way to climate victory.
As I have argued elsewhere, progress in solving our pressing political problems will come not from trying to wish away the human values at their foundation but by directly engaging with them. There should be a variety of indexes, ones that reflect different perspectives on the human diet, ones for low-carb eaters, for people on medically necessary or religious diets, and for athletes who have far different nutritional needs than the average desk jockey.
Even then, not everything can be represented in a quantitative index. Much of what people value about food isn’t about the micronutrients, the grams of protein, or even the calories. We eat to bond with other people, to maintain cultural culinary traditions, and to remember times spent with loved ones.
While it is still worthwhile to explore ways to modify our eating habits in order to lessen the impact on the environment, simplistic carbon footprint measures run roughshod over all the important complexities, especially when combined with the fanatical narrative that there is no alternative to giving up some of our favorite foods. Such a technocratic approach is polarizing, and will ultimately backfire. Environmentalists would do well to learn from McNamara’s mistake, to see the perils of trying to calculate their way to climate victory.
Hard agree on this. Scientific studies and quant models are much weaker evidence than people realize. My default stance when evaluating a study is that it merely proves "a group of experts thought this was worth publishing". I can update my credence from there if warranted.
The pendulum is definitely swinging back to more historical, structural, ethnographic methods.