Economics teaches us the importance of getting the prices right. But policy based on economics may not effectively confront the big problems facing society—from racism to inequality to climate change. Does incorporating behavioral insights into economics simply offer new tools, such as nudges, or can it help us rethink policy from the ground up?
Over the past several decades, much of the contribution of economists to public policy has been along these lines:
- We should reduce poverty by subsidizing work through the Earned Income Tax Credit.
- We should build competitive markets for health care and encourage people to enter them via a combination of fines and subsidies.
- We should reduce our national carbon footprint by introducing a carbon tax.
This policy agenda is organized around “getting the prices right.” We subsidize desired behaviors, tax undesired ones, and let the market handle the rest. But a group of economists is pushing against this way of translating economics in policy, as outlined in a recent article in Boston Review. They believe that economists should instead create an agenda organized around the concept of inclusive prosperity. They refer to the previous policy agenda as neoliberalism.
“Neoliberalism—or market fundamentalism, market fetishism, etc.—is not the consistent application of modern economics, but its primitive, simplistic perversion.”
They call for a greater pluralism in economics—including the incorporation of insights from the behavioral sciences. But does behavioral economics refute—or merely change—the neoliberal agenda?
Neoclassical economics is dead! Long live neoclassical economics!
In the early days, behavioral economics was often pitched as a radical departure from the mainstream. By calling the rational-actor model into question, it threatened to pull the shaky edifice of economics down altogether.
That didn’t happen.
Today, behavioral economics exists in a friendly détente with the mainstream. Behavioral economists have started to admit a grudging respect for the way economists like Gary Becker or Steven Levitt are able to explain a wide variety of phenomena using incentives. Meanwhile, economists like Raj Chetty have suggested that we can incorporate behavioral facets into our models, much the way we do with other market frictions and wedges. A few years ago, the Boston Federal Reserve shut down its pioneering Research Center for Behavioral Economics—not because it was unsuccessful but because it had become unnecessary.
By calling the rational-actor model into question, behavioral economics threatened to pull the shaky edifice of economics down altogether…That didn’t happen.
Maybe we’ll never all be Keynesians, but it seems that we are all behavioral economists now.
In joining the mainstream, behavioral economics has become an applied science. The success of Cass Sunstein and Richard Thaler’s book Nudge has led to “nudge squads” based out the White House, the United Kingdom cabinet, and the World Bank. “Chief behavioral officers” exist in private companies and tech start-ups.
The work typically done by the nudge squads is decidedly unpolitical. Liberal governments have used it to pursue liberal ends, and conservative government have used it to pursue conservative ends. Ultimately, nudges are presenting state or private entities with a new set of tools beyond the standard carrots and sticks. We can use opt-out default choices to encourage people to save, or text message reminders to ensure people attend checkups.
But behavioral economics can go beyond nudges. Increasingly, behavioral economics is helping incorporate insights from social science into policy more broadly. These insights push against the neoliberal consensus policies that someone who relied on a 101 textbook would recommend.
At the margin, incentives don’t work
Incentives—the standard tool in the economics toolbox—is weaker than economists think. People do not always act on the margin. Incentives do not work when people do not notice them or do not understand them.
More importantly, providing incentives can often have a negative effect. Sam Bowles and Sandra Polanía-Reyes have demonstrated that people use the presence (or absence) or incentives to infer socially desired behavior. Large incentives, even as they activate pecuniary motives, deactivate socially responsible ones. Incentives make us less likely to donate blood, less willing to shoulder common burdens, and more willing to be rude to our associates.
Yet much of policy is based on altering incentives. We add a mandate to encourage people to buy health insurance, we impose a work requirement to stop TANF recipients from loafing. The economic argument for this is clear, but the effect of the policy is far lower than economists assume.
On the other hand, sometimes small incentives can work—even when conventional economics suggests they would not. As Rachel Glennerster and Michael Kremer have pointed out in the Boston Review, a wide variety of trivially small incentives (often nonmonetary) have been effective at changing behavior in education. This might be because these incentives are “just right”—not too big to crowd out social incentives, and big enough to call attention to the desired activity.
Small hassles are big
Economics generally assumes that transaction costs are low. But much recent empirical work has shown this to be false. Bargaining or transportation costs can have large effects. This is especially true when the frictions are one-sided.
Behavioral economics supercharges this finding. Not only do transaction costs matter but they can have enormous effects. A major finding in behavioral economics is that people are present biased—they are impatient in the short term even while being patient in the long term.
As a consequence, small hassles—filling out a form, submitting identification, or mailing a letter—often can be just large enough to deter action altogether. While a standard cost-benefit analysis would suggest that the activity is worth doing—why not take an hour of time to fill out a FAFSA form?—it might not be worth doing today. Unfortunately, the future is a long string of “todays.”
Hassles disproportionately affect the poor. People with good jobs and steady benefits certainly encounter many hassles while filing health insurance claims, but the poor face far more. They have to navigate poorly designed websites to get insurance in the first place, cannot comfortably set their accounts to autopay, and may struggle to follow instructions that are not in their native language. Insofar as economics denies the existence of hassles, it obscures these effects.
We should be less quick to judge
One of the most important biases is the fundamental attribution error. We recognize that our behavior is contingent on specific events: I cut that person off in traffic because I didn’t see them, I snapped at my spouse because I had a late night, I fell behind in my work because I am overloaded.
We rarely grant others the same leeway. Instead, we default to our standing hypothesis: the other guy is a jerk.
We’re generally right to give ourselves the benefit of the doubt and should expand that generosity to others. Much of behavior is driven by contextual and situational factors, not differences in personality.
Ironically, the greatest contribution of behavioral economics may be something economics never managed to do on its own—mount a robust defense of the assumption of homogenous agents.
This matters because policymakers often make policy for an amorphous group of “others” (especially when policymakers do not reflect the people they represent across gender, race, or class). Too often, policymakers fail to think about the specific circumstances that drive their constituents’ behaviors. Are welfare recipients lazy, or do they face a high marginal tax rate? Are public-housing residents careless with apartment maintenance, or are they acting as principal-agent theory predicts? Are African Americans prone to violence, or are they unjustly accosted by the police?
It’s easy to assume that behavior is driven by dispositional elements, but it often leads us in the wrong direction. Instead, we should ask whether people who seem unlike us are more similar than we expect. They may share many of the same goals, values, and beliefs but operate under more challenging circumstances.
This is actually a case where behavioral economics supports classical economics. Mainstream economic models typically assume that people are homogenous and can be modeled by a “representative agent.” Economists typically justify this approach by noting that modeling many different types of people would make the math too complex—a defense that rings hollow.
Ironically, the greatest contribution of behavioral economics may be something economics never managed to do on its own—mount a robust defense of the assumption of homogenous agents.
