The Downfall and Consequences of the Cheap-and-Easy 'Behavioral' Dodge in Public Policy

The last decade or so witnessed the emergence of a subfield of social science analysis known as "nudge research."  It became all the rage, and governments around the world (including the US federal government and many state governments) created research policy groups to devise so-called nudges to improve social and economic outcomes.  Want to get people to be more honest in filing their taxes?  Nudge them by putting the "I'm not lying" statement on the top of the form rather than at the bottom, and more people will file honestly (or less dishonestly).  Easy!

That research was based on what had come to be known as behavioral economics, or BE (spawning a nearly identical offshoot in legal scholarship called Behavioral Law & Economics, or BLE).  The idea is that people can be led to make different decisions based on identical information simply by changing the way that the information is presented.  In the case of the tax-honesty pledge, attesting to the same statement would have the same impact or non-impact for a "rational self-interested maximizer" no matter where the oath is placed, because such a person thinks only about the information, not how it is presented.

Even the most basic facts about real life make it clear that people are not rational maximizers.  Why do companies pay millions of dollars for Super Bowl commercials, if not because they believe that people's decisions can be manipulated by high production values and maudlin messages?  Why are "impulse buys" so important that, according to one study in Canada, they "made up anywhere from 40 to 80 per cent of shopper's spending, depending on the store or product category"?

It seems obvious that we should not expect people to be rational maximizers.  Decades of research has explored issues such as "bounded rationality" and other reasons to doubt that people are the data-driven computing machines required in economic models that do not incorporate BE.  But even the most basic observations about people's limited mathematical skills, and simply understanding that rational optimization is itself costly (data being expensive to acquire and then process, at least by taking up people's time if not money), should put the burden of proof on anyone who says that people behave consistently with the rational actor model.

So we ought to do better when we admit that people do not act in ways that orthodox economic models assume.  Nudges are meant to take advantage of these behavioral deviations from pure rationality by getting people to do things voluntarily-but-irrationally rather than forcing them to change involuntarily by, for example, enforcing old-fashioned prohibitions and mandates.

Last year, however, it finally became clear that nudge-based policies were largely a sham.  In a particularly pointed analysis this past November in The Washington Post, a reporter who writes frequently about economics (formerly via a regular column in The New York Times) delivered the surprising news.  Noting that the "move the tax oath" nudge had been inspired by a study of insurance fraud, Eduardo Porter wrote:

In the years after the research was published in 2012, several governments experimented with using honesty-inducing language on tax forms, and the response was a dud. The economists then disavowed their original study. Finally, a couple of years ago, statistical sleuths discovered that some of the data supporting that research had, ironically, been faked.

Yikes.  Maybe that was simply some bad apples?  Porter is unconvinced: "[A] lot of the nudge findings are, in fact, flops."  He conceded that "[m]ost of the research underlying the nudge revolution might not be a fraud," but it is a long distance from not being an outright scam to being honest and reliable:

Much nudge research lives in a gray area between good science and academic fraud. Academics often massage data to make nudges seem significant, and they don’t publish the studies where the nudge didn’t work. A meta-analysis of such studies last year concluded that “after correcting for this bias, no evidence remains that nudges are effective as tools for behaviour change.”

Lots of nudges seem to work only if you squint hard enough. The effects of interventions are either fleeting or so modest that relying on them to address society’s shortcomings — pollution and climate change, discrimination, gun deaths, low retirement savings, you name it — is leading policymakers off-course.

Maybe Porter is just a Gloomy Gus, but Leif Weatherby (who directs the Digital Theory Lab at NYU) wrote a similar piece in The Times later that month that concluded with this: "Humans may not be perfectly rational, but we can do better than the predictably irrational consequences that behavioral economics has left us with today."  Earlier in the piece, he wrote that "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," adding: "What some people claim is science might just be common sense dressed up in bad data."

Double yikes.  Even so, none of this should be particularly surprising.  In 2013, I wrote a skeptical assessment of BE and BLE, asking: "Has Behavioral Law and Economics Jumped the Shark?"  It was obvious by then that a lot of the fizziness around behavioral nudge-y analyses was based on studies that were either non-replicable or were being used opportunistically by motivated, biased analysts.  It turns out that (according to Weatherby's article) some nudge researchers "became international celebrities who jet around the globe" and made huge salaries, and they responded in the way that a simplistic, non-behavioral economic model would predict: They moved up their supply curves, responding to higher pay by doing more of what had brought them fame and fortune.  The word "ironic" is overused, but that is irony.

Even so, when the 2017 faux-Nobel in Economics was awarded to one of the big names in BE/BLE, I put aside my deep skepticism and managed to be mildly enthusiastic: "The short answer [is] that we are better off in a world with behavioral economics than in a world without it."  Looking back at that statement now, I have to ask myself a question in the form that my millennial daughter would use: "But are we?  Are we ... really?!"  And the answer is, probably not.

To be fair to my former self, my grudging support of BE/BLE in 2017 was a lesser-evils conclusion.  That is, I noted that BE had supplanted not some more useful and honest version of social and economic analysis but instead had undermined orthodox (neoclassical) economics.  The latter uses rational actor theory to "prove" that government action is invariably destructive, that regulation is harmful, that fighting inequality is inefficient, and on and on.  Compared to that, BE was a breath of fresh air.  Orthodoxy was not even close to reality and it led to horrible policy choices, so it seemed like an easy call.

Both Porter and Weatherby, however, pointed out that the nudge revolution came at a very steep cost.  When it comes to government policy, Porter wrote that "one of the leading researchers in the field[] had to acknowledge in recent years that nudges are not enough. If there is a case for these interventions, it is not that they can make a sizable difference; it is that they are an inexpensive way of doing little things."

Perhaps that is not a bad approach.  Big changes are often difficult to enact, after all, so maybe we could all simply be pleased when small progress is made at low cost.  Unfortunately, even that is not how it works.  Porter added:

The problem, as Carnegie Mellon behavioral economist George Loewenstein has pointed out, is that nudges’ emphasis on correcting unhelpful psychological biases gives policymakers a cheap way out, an excuse not to do more.

It allows them to believe that obesity will be solved by calorie labels to enhance people’s self-control, rather than by regulations on food companies. That discrimination and segregation will be addressed by making people aware of their implicit biases, not by education, housing or redistribution policies. That they can fix stuff at no cost to taxpayers.

Porter summarizes his column in part by arguing that "[n]udges are at best a distraction, and more likely a smokescreen."  Governments, then, are possibly doing less than they could be doing even under severe political constraints because both Democratic and Republican politicians are drawn to a cheap-and-easy solution that in fact makes matters worse.

Moreover, Weatherby points out that it might be even worse when we consider how BE is used in corporate research: "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."  Because so much market research is driven by nudge-y nonsense, "[t]o perform even the most banal, everyday actions, you have to put implicit trust in unverified scientific results."  He adds this devastating conclusion:

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.

Perhaps we should have known that allowing public policy to be hijacked by a faddish set of ill-supported conclusions would go wrong.  I am not saying that we should go back to relying on orthodox economic analysis.  Far from it.  Instead, we should learn that the failure of two too-easy solutions -- orthodoxy's "leave it to the tender mercies of the Invisible Hand" and BE/BLE's "trick people into making better decisions" -- leave us with the only option that ever made sense: Do the hard work of figuring out how to solve big problems and go big to solve them.  Pragmatism should never go out of style.