Architects of Manipulation

Alexander Hamilton, in the opening essay of The Federalist, famously declared that the American experiment would determine “whether societies of men are really capable of establishing good government from reflection and choice, or whether they are forever destined to depend for their political constitutions on accident and force.” In his new book, Cass R. Sunstein implicitly asks the same question: Are we really capable of governing ourselves through our own reflection and choice? Contra Hamilton, Sunstein concludes that we are destined to rely on accidental circumstances, namely whether “choice architects” create an environment in which we will choose correctly.

Simpler: The Future of Government is one of Cass Sunstein’s better books. Largely unconcerned with constitutional or legal theory, it straightforwardly explains how insights derived from behavioral economics have informed the Obama administration’s revolutionary approach to regulation. (He of course was one of the chief revolutionaries, having served from 2009 to 2012 as head of the White House Office of Information and Regulatory Affairs.)

Those interested in the regulatory state, regardless of their political orientation, will learn much from Simpler. It presents a powerful case for a more nuanced understanding of human behavior than that offered by neoclassical economics, and it connects this new understanding to important regulatory initiatives of recent years. However, in the final analysis, Sunstein overlooks the fundamental questions by neglecting the distinction between “nudging” and shoving and by hiding behind his pragmatist veneer.

Sunstein’s basic purpose here is to advance his project of “libertarian paternalism.” The present book draws heavily upon the concept of Sunstein’s 2008 coauthored book (with Richard H. Thaler), Nudge: Improving Decisions About Health, Wealth, and Happiness. He writes in Simpler that regulation ought to frame choices so that individuals retain the power of choice, but are induced to choose in accordance with the wishes of the bureaucrats who serve as what he calls “choice architects.” Nudges subtly manipulate citizens’ choices. They are, according to him, “approaches that influence decisions while preserving freedom of choice.”

And he believes that nudges are often more effective at changing behavior than regulatory commands because human beings (as he correctly observes) do not simply choose based on rational calculations of their own interest. “The social environment against which we make our decisions” matters much more than our rational calculation of the consequences of our choices. What follows, for him, is that bureaucrats, by manipulating this social environment, can attain the ends they seek without the messy means of mandates. Behavioral economics has a better understanding of human beings than the old neoclassical approach. “At this point,” writes Sunstein, “it is not exactly news that people do not behave like the ‘rational actors’ of economics textbooks. We are Homo sapiens, not Homo economicus.”

He uses an analogy from Daniel Kahneman’s Thinking, Fast and Slow (2011) that conveys the basic reasoning of behavioral economics. By this reasoning, there are two systems in the mind. “System 1 is the automatic system” that relies on instinct and intuition to induce choices. It is a faculty of the mind, writes Sunstein, but “much of the time, it is on automatic pilot.” It does not possess what Aristotle calls practical rationality. That capacity belongs to “System 2,” which is “a bit like a computer or Mr. Spock from Star Trek. It is deliberative. It calculates.” Most human beings employ System 2 infrequently, relying instead on System 1 to make the myriad choices with which they are confronted in everyday life.

The problem is that System 1 acts irrationally. When it is in charge, people make decisions that do not maximize their rational self-interest. As Sunstein points out, people are more likely to make a risky investment on a sunny day, and more likely to give money to charity if the checkbox is green rather than red. They are also loss-averse, which means that they will demand more to give up an item like a lottery ticket than they would be willing to pay for it in the first place. “This finding complicates free market thinking,” Sunstein rightly argues, for the evidence shows that human beings “depart from economic understandings of rationality.”

Armed with their view that what matters most is the environment in which a choice is made, “choice architects” can direct us by changing the environment—by, say, requiring disclosure of information, restructuring default options, automatically enrolling people in certain programs, or presenting citizens with information in a strategic way. By following this model, says Sunstein, government will not only act in a way that is simpler, it will more effectively control citizens.

This argument exploits a massive weakness in neoclassical economic thought. The notion that whenever human beings choose, their decision reflects their true preference misses some very important characteristics of human behavior. The American Founders, of course, did not ignore these characteristics. The Federalist authors’ view of human nature was anything but simplistic. Man’s reason is fallible, James Madison explained in Federalist 10, and this means that his passions and opinions will have a reciprocal effect on each other. “It is a just observation,” Hamilton added in Federalist 71, “that the people commonly intend the public good.” Be that as it may, “their good sense would despise the adulator, who should pretend that they always reason right about the means of promoting it. They know from experience, that they sometimes err; and the wonder is, that they so seldom err as they do.”

As the Founders understood, human beings are not just rational calculators; they are bundles of reason and passions. The key is to have reason inform the passions through good habits, which promote sound practical judgment.

By relying too heavily upon the power of rational self-interest, defenders of neoclassical economics have ceded ground to proponents of behavioral economics, who are not reluctant to manipulate the passions of citizens to influence their choices. Until neoclassical economists return to an understanding of human behavior that recognizes the effects of the passions and of the social environment in which choices are made, they are likely to be frustrated by the advancement of policies based on behavioral law and economics.

But if neoclassical economists wrongly overemphasize the rational faculties, behavioral economists go off track in a different way: They neglect the possibilities of these faculties. What is striking about Simpler is how little Sunstein expects of people, and how his approach to government will ensure that little will be expected of them in the future. Rather than seeking to elevate citizens’ decision-making capacities, Sunstein takes the path of least resistance: let’s just manipulate the environment so that their passions unwittingly produce a good choice.

This is a recipe for an increasingly inept citizenry, devoid of the habits of self-government, as Joshua Wright and Douglas Ginsburg have argued previously on this site. Yet Sunstein is little troubled by nudging’s probable long-term harm to the character of citizens. His lack of unease points to two fundamental problems with his presentation.

The first has to do with what ought to count as a nudge. Sunstein offers many examples that have been pursued by the administration—presumably at his urging—over the past few years:

  • A Department of Transportation requirement that airlines disclose the fees associated with the cost of airline tickets.
  • A Department of Agriculture requirement that all grocery stores disclose the nutrition information of food at their deli counters.
  • A Department of Health and Human Services rule requiring disclosure of all “unreasonable” increases in health-insurance premiums.
  • A Department of Education requirement that educational institutions disclose their costs, their rates of graduation, and other associated statistics.
  • A Federal Reserve rule forbidding banks from charging overdraft fees unless account-holders actively authorize their enrollment in the program.
  • A Food and Drug Administration requirement (struck down in court) that cigarette companies present graphic anti-smoking images on their packaging.

Sunstein calls nudges what are in fact mandates and prohibitions. True, they do not directly restrict the freedom of consumers; but they certainly restrict the freedom of other economic actors. Whether or not one agrees with these requirements (and I would agree with many of them), labeling them as nudges stretches the limits of credulity. The author is keenly aware of the need to make these examples convincing to his readers. But if we employ our “System 2” calculator faculty for a brief moment, it is clear that while we are being nudged, others are being controlled.

Sunstein wants us to think that coercion and constraint are absent from nudging, but the heavy hand of government is still present. Furthermore, as he emphatically affirms—perhaps to appease readers politically to his left—governmental use of nudges, when he was at OIRA, did not entail a corresponding decrease in other kinds of regulation. “We issued a lot of nudges, but we did much more,” he boasts. “If government requires people to make a (tax) payment for failing to purchase health insurance, it is not nudging.” He reminds us that “in the Obama administration, we took big steps to increase the fuel economy of cars, not by nudges but by mandating big increases in the fleet-wide MPG average.” Even by his own spurious definition of nudging, Sunstein acknowledges that much of what he supported at OIRA was traditional command-and-control regulation.

The second problem has to do with Sunstein’s alleged pragmatism, a subject about which I have written elsewhere. In Simpler, as in his other writings, Sunstein attempts to position himself as someone who merely follows where the facts and the data lead. He has consistently pitched his approach as “Regulatory Moneyball”—a reference to Oakland Athletics’ General Manager Billy Beane’s emphasis on sophisticated empirical analysis rather than intuition or feeling in evaluating baseball talent.

But in spite of this attempt to transcend the world of opinion and values, his opinions and his values often creep into the discussion. Organ donation is one example. Sunstein notes that the use of default rules can change behavior dramatically when it comes to designating one’s organs for use upon one’s death. Donation would skyrocket if government were to “presume that people consent, at the time of death, to donate [their organs] but allow them to opt out if they wish.” In other words, the default rule would be to donate, but anyone who wishes to do so could opt out. Now, he quickly adds, “I am not contending that this is right policy; my only claim is that if we want to make more organs available for those who need them, we should be aware that the default rule really matters.”

Another example is preventing obesity, about which the author writes: “When fast food restaurants are located near schools or residences, both children and pregnant women gain a lot of weight. And when unhealthy foods are made even slightly less accessible, their consumption is reduced.”

Sunstein, inexplicably, does not proceed from these “impartial” observations to proposing, say, a default rule in favor of organ donation, or a default rule that fast food restaurants may not be located closer than a half-mile from major residential areas. After all, these would fit Sunstein’s definition of a nudge. If these actions would change behavior in a positive way, why not start nudging?

Especially since bad actors are out there nudging bad things. “Many restaurants ask consumers if they would like to ‘supersize’ their order,” he writes. “What about asking people if they would like to ‘downsize’” and thereby save money? When this was attempted, a dramatic change in consumers’ behavior resulted. So why not mandate a similar script to be read to customers at the fast food window? These would be mandates imposed upon businesses, but as we have seen, for Sunstein this is consistent with nudging as long as consumers’ choices are not mandated.

Sunstein’s refusal, without explanation, to carry the nudge principle into these areas reveals the limits of his alleged pragmatism. He must have recourse to some extra-scientific, value-laden principle that distinguishes why it is okay for government to nudge in some ways to produce better outcomes, but why government can’t employ other nudges that would also produce better outcomes. “Moneyball” will not suffice. The fundamental problem is that the issues in regulatory policy are far more complicated and the objectives are far less explicit and measurable than those in Major League Baseball. It is much easier to operationalize on base percentage than “social welfare,” which (according to Sunstein) is what cost-benefit analysis attempts to measure when it calculates the benefits produced by agencies’ rules. When we hear the government’s rationale for valuing saving a human life at somewhere between $7 million and $9 million, we hardly feel assurance that enlightened social scientists, devoid of the subjectivity of “values,” are at the helm.

In a chapter distinguishing his approach from the “Nanny State,” Sunstein attempts to get out of the problem this way: “The use of paternalism – soft or hard – for illegitimate ends is wrong because the ends are illegitimate.” With this unhelpful tautology, he tries to extricate himself from what he refuses to recognize as serious moral dilemmas. The “choice architects,” by this reasoning, will only use their soft paternalism for legitimate ends, and they will not stray from those legitimate ends. All of this presumes, of course, that we can distinguish legitimate from illegitimate ends – a presupposition Sunstein adopts without defending in order to frame the issue in a palatable way.

Simpler is an important book for framing the strengths and the weaknesses of what Sunstein calls nudging. The view of human nature upon which nudging is based is arguably more robust and accurate than the prevailing view that considers human beings to be rational preference-maximizers. We would do well to take that part of the author’s argument seriously. But he seeks to use choice architecture in place of a political system that helps to cultivate the faculties of practical judgment that are necessary for a well-functioning popular government. Sunstein’s attempt to position himself as a mere pragmatist, and the perils into which he falls as a result, remind us that politics will never be devoid of serious disputes about the nature of justice.