Editor’s Note: This is part of a series of dispatches from the Knight Science Journalism Program’s 2020-21 Project Fellows.
John Sterman, a professor in MIT’s Sloan School of Management students, says his students, who affectionately call him Dr. Doom, generally have no trouble explaining what they do: how they make capital investments, for example, or how they set prices. “What they are terrible at,” he says, “is seeing how all those different pieces fit together to generate the dynamics of the system.” Asking people — even MIT students — to imagine what will happen in a system with more than one feedback loop is apparently a failsafe recipe for disappointment. “Our mental simulation capabilities are terrible,” Sterman told me.
These words came as something of a relief to me, five months into my MIT Knight Science Journalism Project Fellowship. As part of the research for my forthcoming book on refrigeration, I had set myself the task of determining the impact that the ability to keep food cool has had on human health. After all, it seems likely that by transforming how, what, and even when we eat, refrigeration would have an enormous effect on human health — but no one seemed to have even tried to pin down the consequences of the “fridge diet.” How hard could it be?
I spent the fall attempting to enumerate refrigeration’s likely ramifications, and then gathering what little data existed. By February, I was drowning in a sea of bounding variables, confounding factors, correlation, and sheer speculation. The rise of the modern cold chain at the start of the twentieth century might seem to promise a reduction in food poisoning, but how could its effects be untangled from that of municipal water chlorination, an innovation that dates back to the same time period and certainly also transformed the average American’s daily bacterial exposure? And, anyway, wouldn’t that reduction in food poisoning have been cancelled out by the refrigeration-enabled centralization of agriculture, which allows an isolated outbreak of E. coli, say, in the lettuce fields of California’s Salinas Valley, to sicken consumers across the country?
While I went around in circles, I had the good fortune of attending, along with other Project Fellows, a seminar by Sterman, in which the business school professor sought to help us understand how policy decisions affect climate change by using an interactive simulator he calls En-ROADS. I began the session as a skeptic, but was quickly sucked in.
For Sterman, one of the real purposes of the exercise is simply to help people overcome their chronic inability to imagine a system in which the various elements interact with each other, rather than simply adding up.
On Sterman’s screen were a couple of charts: one showing the breakdown of global energy sources over the next century, and another showing net greenhouse gas emissions. Next to them, in extremely large print, was a number: 6.5ºF, or the expected temperature increase by 2100 based on current policies.
Sterman directed our attention to a series of digital levers — policies we could toggle to try to prevent total planetary incineration. “You’ve got to get the warming down to 3.6 degrees Fahrenheit,” he said. “So what would you like to do?”
“What about using more nuclear?” one of us volunteered. Sterman cranked the nuclear subsidy slider. Before long, nuclear power was cheaper than free, at -7 cents per kilowatt-hour, and we’d lowered global warming by a measly tenth of a degree. The graph revealed where we’d gone wrong: New reactors take a long time to design, build, and permit; meanwhile, fossil fuels were still being burned with abandon, and renewable energy investment fell of a cliff as utilities rushed to put their money into nuclear. Subsidizing nuclear power, we quickly realized, is not a high-leverage policy.
Over the course of the next hour and a half, our group eventually managed to reduce projected warming to 3.7ºF. I came away feeling hopeful and empowered — and not just about our potential to avoid climate disaster. Sterman’s model offered an unexpected new way for me to think about the equally complex interaction of refrigeration, diet, and health.
“It’s pretty coarse,” said Sterman, when I followed up after the session to ask him about the assumptions on which the simulation is built. “It doesn’t tell you what happens if I eat an Impossible burger instead of a real burger.” The En-ROADS model is based on existing data, with all its gaps, abstractions, and assumptions — and that’s fine. For Sterman, one of the real purposes of the exercise is simply to help people overcome their chronic inability to imagine a system in which the various elements interact with each other, rather than simply adding up. Nuclear subsidies don’t just lead to an increase in emissions-free power, they also depress investment in renewables; when energy is made artificially cheap, the demand for it also goes up, and fossil fuels will fill some of that gap. Similarly, refrigeration doesn’t simply reduce the growth of bacteria in our food, it reshapes our gut microbiome, centralizes agriculture, and lengthens supply chains. By outlining the contours of the levers, Sterman showed me, we can overcome our system blindness — and begin to see how the pieces fit together.
Nicola Twilley is co-host of the award-winning Gastropod podcast and a frequent contributor to The New Yorker.