I have come to believe very strongly that podcasting is the devil’s work.
Over winter break, Freakonomics Radio host Stephen Dubner updated a story from last January, “Why Is There So Much Fraud in Academia?” It considers the spectacular rise and fall of Dan Ariely and Francesca Gino and investigates how researchers who produce camera-ready findings are rewarded materially and institutionally, even when their work is unsound or entirely fraudulent.
The coverage is acceptable but haunted by its own specter. Not so long ago, Freakonomics themselves breathlessly promoted Ariely and Gino, not to mention other recent fraudsters like Brian Wansink, and I would have liked to see more self-reflection on this account.
So I was intrigued to see a podcast from the other half of Freakonomics’ dynamic duo, UChicago’s own Steven Levitt. Last June, Levitt interviewed Harvard psychology professor Ellen Langer and asked his audience: “How do you know whether you should believe surprising results?”
The results are indeed surprising. Langer’s pet theory, mind-body unity, posits an unbelievably strong placebo effect. In this world, faster perception of time accelerates healing, environments mimicking the past improve hearing and vision in the elderly, and asking nursing home residents some memory questions cuts two-year mortality by over two-thirds.
Those aren’t cherry-picked examples: those are the studies which Levitt and Langer bring up as examples of Langer’s best, most surprising work, which earned her tenure at Harvard and a permanent place in the TED Talk firmament as the “mother of mindfulness.” Even so, other researchers have been showing these findings are invalid for over a decade, though I will leave an in-depth treatment to Columbia’s Andrew Gelman, UPenn’s James Coyne, and DePaul’s David Ramsay.
Langer fits Dubner’s suspect profile to a T: a researcher who reliably produces picture-perfect results which convert into press and popular attention. Armed with honesty and courage, Levitt has an opportunity to reckon with the sloppy science of the past.
He does not take this opportunity.
Despite gesturing towards skepticism, Levitt fails to model it. Encountering unbelievable findings, he takes Langer’s word at face value and moves on. He expresses doubts that such studies could have been successfully replicated and just as quickly insists that they have been replicated “a number of times.” That number is two: one failed to replicate, and the other has not been published. Further, Levitt either did not know or chose to ignore that the mortality results were based on a statistical error. How does he react to such unbelievable results?
LEVITT: You had, through this intervention, seemingly radically changed mortality rates to a degree that I think you would be hard pressed to find any pharmaceutical compound which has had that kind of effect in a controlled study.
The listening experience is excruciating. Before Levitt is a jigsaw which forms a serious challenge to Langer’s signature claim, and for an hour he refuses to assemble it, instead waving each piece in front of the audience like jangling keys, dropped after a moment to pick up the next.
“Why is there so much fraud in academia?” Dubner asks, while his longtime collaborator presents woo to the public without pushback or confrontation. Because pop-science media is a machine that launders bad science into public acclaim and cares not for the truth.
This is not unique to podcasting—print and web media is awash with dreck that misrepresents published research for clicks—but podcasts, in their structure and form of consumption, are considerably more damaging, starting from the difference between reading and listening.
A reader can scan a text for words and topics, focus on and reread individual sentences and phrases, and shift between sections to retrieve lost context. This is a mode that invites, though does not always receive, critical attention.
Listening does not support such engagement. Audio moves at its own pace, giving fluff and critical information the same airtime. Locating a section of audio is effortful and unsupported. Instead of sustained critical attention, it invites passivity, letting the waves flow over you.
This doesn’t mean audio is strictly inferior—I love both audiobooks and the lost art of the audio drama—but I wouldn’t listen to a journal article. The information they’re designed to communicate wouldn’t survive the transition.
Instead, the lengthy podcast lends itself to secondary listening while driving or folding laundry, and the top podcasts cultivate intimacy and parasociality, making the listener feel like an active partner in a fascinating conversation. Bad assumptions pass undetected, hidden in soothing cadences and the listener’s emotional investment in the host.
With all these traits, the podcast is a terrible form for serious science communication… but great for giving the impression of depth and rigor with minimal effort.
LEVITT: So there’s no intervention other than teaching.
LANGER: Exactly.
LEVITT: Nothing can possibly happen.
LANGER: But something happened.
Freakonomics innovated this style, but it was perfected by others; Andrew Huberman, who sits in the top 10 such podcasts globally by scaremongering about cell phones irradiating testicles in between hawking supplements; and the dark priest of anecdata himself, Malcolm Gladwell, whose writing anticipated this new medium.
You may protest that I’m painting with too broad a brush, that the examples above are in opposition; they are not. They target different market segments with the same method. They claim to give the audience tools to answer questions for themselves but provide only the means for complacency and self-deception. Levitt does not model skepticism, rigor, or intellectual responsibility. He only tells listeners to believe this particular set of extraordinary claims based on Langer’s credentials and his own credulity.
We should expect better of our professors than to make these bargains.
Nicolas Posner is a 2024 alum of The College.
