Imagine removing a branch of the U.S. government, say the Supreme Court. What are the myriad ways that such an upheaval might reshape people’s lives?
Policy makers and researchers probably would want to have an idea of what those effects might be before erasing the highest court in the land. But “you can’t test deep structural changes like that in an experiment” first, says behavioral decision–making expert David Gal of the University of Illinois Chicago.
Likewise, less wildly hypothetical but perhaps still far-reaching changes to society, such as expanding Social Security or providing universal parental leave, can’t be tested with conventional experiments that include control and experimental groups. As a result, many behavioral scientists today have instead turned to researching “nudges” — smaller interventions that operate within existing policies. Nudges can influence human behavior, research suggests, and can be readily tested using experiments before being applied.
But this recent overreliance on nudges has stifled broader behavioral science research and insights into how to create a better society, Gal and marketing expert Derek Rucker of Northwestern University in Evanston, Ill., contend January 12 in a commentary in Nature Reviews Psychology.
Nudges exploded in popularity in 2008 when economist Richard Thaler of the University of Chicago and law professor Cass Sunstein of Harvard University published a book on the topic. That research netted Thaler a Nobel Prize in Economic Sciences and inspired governments worldwide to set up nudge units to modify or create public policies (SN: 10/9/17; SN: 3/18/17).
Examples of nudges include offering small cash rewards to encourage people to get a new vaccine or sending text reminders about a looming deadline. For instance, researchers recently revamped a court summons form and sent text reminders to get more people to attend mandatory court appointments in New York City. The intervention increased court attendance by roughly 20 percent over previous years, the researchers estimate (SN: 10/08/20).
But such nudges ignore thornier societal problems, such as over-policing in low-income neighborhoods where these summons are typically issued, lawyer and sociologist Issa Kohler-Hausmann of Yale University wrote in a perspective piece that accompanied the research.
“Changing the approach to penal and welfare policy in our country will require interventions that are much more radical than cost-neutral behavioral nudges that everyone can agree on,” Kohler-Hausmann wrote.
Policy makers love nudges, Gal argues. “They don’t have to change anything fundamental.”
As for behavioral scientists, Gal and Rucker attribute nudges’ popularity to the scientists’ desire to mimic the precision of researchers in other fields. Medical researchers, for instance, can test pharmaceutical drugs using randomized controlled trials. With that scientific gold standard, researchers compare outcomes among patients receiving the drugs versus a placebo. Nudge researchers can likewise generate a small change — the drug — and compare outcomes among those who experience the change with those who do not.
“We value experiments because they give us statistically precise estimates,” Gal says.
But nudges that work in the lab often fail in the real world, the authors note. In one analysis, lab studies of 74 nudges and roughly half a million participants increased the desired behavior by an average of 8.7 percentage points, researchers reported in a 2020 working paper in the National Bureau of Economic Research. But similar real-world studies of 243 nudges affecting over 23 million people increased the desired behavior by an average of just 1.4 percentage points.
Rather than chasing statistical precision, Gal would like to see behavioral scientists generate overarching theories that apply beyond a single narrow context. For instance, in the U.S. legal system, juries must reach a unanimous verdict to convict a defendant. But research into conformity suggests that people copy others as a result of social pressure (SN: 8/15/18). Unlike nudges, that research can generate insights into how human behavior interfaces with existing practices, Gal says, and raise crucial questions. In this case, is the push for unanimity preventing jurors from raising valid concerns during deliberations? “Even one dissenter can really reshape the debate and stop this tendency toward conformity,” Gal says.
There’s room for both theoretical and applied behavioral scientists in the field, counters data scientist Kevin Wilson of the Policy Lab, a policy research institute at Brown University in Providence, R.I. “We need people who think about theory, who are really synthesizing these lessons and, as they put it, extrapolating insights. But we also need people who are going to … utilize these insights.”
Right now, nudges are hogging all the attention, Kohler-Hausmann says. Like policy makers, funding agencies and research journal editors seem to prefer the quantifiable results that nudges offer, she says, and that near singular focus has hindered transformational change. “The cost of a narrowly defined intervention is ruling out the study of more compound, complex interventions.”