It’s hard to walk a block or finish a morning coffee without encountering some system that attempts to tell you what to read, what to buy, how to lose weight or prevent dementia or tweak your decisions in other specific ways. My watch constantly buzzes with “relax reminders.” The number of calories appears next to every menu item at fast-food restaurants.
These experiences are the result of a concerted scientific effort to understand and adjust human behavior — “nudges,” as the legal scholar Cass Sunstein and the economist Richard Thaler call them, that push us gently to make preferred choices. The “nudge doctrine” the pair developed has led to the creation of hundreds of “nudge units” in governments all over the world (including our own), that seek to put nudges in policies and procedures. Examples of actions that are called nudges include making organ donation opt-out instead of opt-in, and sending information about how much drivers would save by switching to car-pooling and public transit.
There’s just one problem: It’s not clear how effective nudges really are. One recent analysis of a large study of nudging interventions found that, after accounting for the fact that positive results are more likely to get published, the evidence that nudges change the decisions people make in their everyday lives is not particularly strong. The science behind nudging is little more than a thin set of claims about how humans are “predictably irrational,” and our policies and systems should heavily divest from its influence.
The nudge doctrine originated in behavioral economics, a field of applied social science that has deeply influenced public policy and algorithm design. Behavioral economics is based in large part on the Nobel-winning insights of Daniel Kahneman and Amos Tversky, whose groundbreaking experiments in the 1970s showed that humans made systematic errors when reasoning about statistics.
We make everyday predictions, like guesses about the stock market, based not on weighing of evidence, but instead on heuristics like what information happens to be available, or what we have recently been thinking about. This discovery led to the rise of a form of psychology that tested people’s susceptibility to mistakes based on how information was presented, what stories they were told and what stories they invented to make sense of their lives and perceptions. The fundamental premise is that our common-sense intuitions are often irrational, but can be corrected with the help of data.
On paper, that sounds like a worthy goal. Some behavioral interventions do seem to lead to positive changes, such as automatically enrolling children in school free lunch programs or simplifying mortgage information for aspiring homeowners. (Whether one might call such interventions “nudges,” however, is debatable.)
But it’s not clear we need to appeal to psychology studies to make some common-sense changes, especially since the scientific rigor of these studies is shaky at best.
Behavioral economics is at the center of the so-called replication crisis, a euphemism for the uncomfortable fact that the results of a significant percentage of social science experiments can’t be reproduced in subsequent trials. Nudges are related to a larger area of research on “priming,” which tests how behavior changes in response to what we think about or even see without noticing. One of the foundational experiments for priming showed that undergraduates walked out of the lab more slowly after focusing on words associated with old people, like “bingo” and “Florida.” But this key result was not replicated in similar experiments, undermining confidence in a whole area of study. It’s obvious that we do associate old age and slower walking, and we probably do slow down sometimes when thinking about older people. It’s just not clear that that’s a law of the mind.
It also turns out that people without any scientific training are good at correctly guessing or betting on which studies can’t be replicated based only on their descriptions. What some people claim is science might just be common sense dressed up in bad data.
Even Dr. Kahneman is affected by the replication crisis. Outside researchers have found that Dr. Kahneman’s best-selling book “Thinking, Fast and Slow” relies heavily on studies that may not be replicable — troubling given that this widely read book popularized behavioral economics. It’s now everywhere, from what kind of medical care you receive to what your workplace atmosphere is like.
And these attempts to “correct” human behavior are based on tenuous science. The replication crisis doesn’t have a simple solution. Journals have instituted reforms like having scientists preregister their hypotheses to avoid the possibility of results being manipulated during the research. But that doesn’t change how many uncertain results are already out there, with a knock-on effect that ripples through huge segments of quantitative social science. The Johns Hopkins science historian Ruth Leys, author of a forthcoming book on priming research, points out that cognitive science is especially prone to building future studies off disputed results. Despite the replication crisis, these fields are a “train on wheels, the track is laid and almost nothing stops them,” Dr. Leys said.
Sometimes the social science train goes off the rails, as we’ve seen from recent allegations of fraud against the superstar honesty researchers and behavioral economists Dan Ariely and Francesca Gino. Based on their promises to “correct” dishonesty, these figures became international celebrities who jet around the globe and, in Dr. Gino’s case, earn some of the highest salaries at Harvard. If they were indeed cooking their books, that points to deeper problems than a few bad apples. (Both Dr. Ariely and Dr. Gino have denied engaging in research misconduct, and Dr. Gino has filed a defamation lawsuit against Harvard and three researchers who identified data discrepancies in papers she co-wrote.)
These cases result from lax standards around data collection, which will hopefully be corrected. But they also result from strong financial incentives: the possibility of salaries, book deals and speaking and consulting fees that range into the millions. Researchers can get those prizes only if they can show “significant” findings. It is no coincidence that behavioral economics, from Dr. Kahneman to today, tends to be pro-business. Science should be not just reproducible, but also free of obvious ideology.
Technology and modern data science have only further entrenched behavioral economics. Its findings have greatly influenced algorithm design. The collection of personal data about our movements, purchases and preferences inform interventions in our behavior from the grocery store to who is arrested by the police.
Setting people up for safety and success and providing good default options isn’t bad in itself, but there are more sinister uses as well. After all, not everyone who wants to exploit your cognitive biases has your best interests at heart.
Despite all its flaws, behavioral economics continues to drive public policy, market research and the design of digital interfaces. One might think that a kind of moratorium on applying such dubious science would be in order — except that enacting one would be practically impossible. These ideas are so embedded in our institutions and everyday life that a full-scale audit of the behavioral sciences would require bringing much of our society to a standstill. There is no peer review for algorithms that determine entry to a stadium or access to credit. To perform even the most banal, everyday actions, you have to put implicit trust in unverified scientific results.
We can’t afford to defer questions about human nature, and the social and political policies that come from them, to commercialized “research” that is scientifically questionable and driven by ideology Behavioral economics claims that humans aren’t rational. That’s a philosophical claim, not a scientific one, and it should be fought out in a rigorous marketplace of ideas. Instead of unearthing real, valuable knowledge of human nature, behavioral economics gives us “one weird trick” to lose weight or quit smoking.
Humans may not be perfectly rational, but we can do better than the predictably irrational consequences that behavioral economics has left us with today.
Leif Weatherby (@leifweatherby) is an associate professor of German and the director of the Digital Theory Lab at New York University.
The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: [email protected].
Follow The New York Times Opinion section on Facebook, Twitter (@NYTopinion) and Instagram.