Why Nudge? The Politics of Libertarian Paternalism
Yale University Press, 2014, 208 pp., $25
It seems like only a few years ago that economists in the mode of the recently deceased Gary Becker were deluging policymakers with an image of consumers as hyper-rational utility-maximizers, capable of impressive foresight and calculation. Whether taking drugs, getting married or divorced, or even choosing to commit suicide, individuals could be accurately modeled as altogether rational. Such rational people could be counted on to be effective stewards of their own welfare, and thus the idea that government could improve on voluntary market transactions was a fantasy.
Suddenly, behavioral economists offering a different model of man are everywhere—in top economics departments, on the editorial pages, and increasingly in government itself. They tell us, based largely on laboratory experiments and evidence from psychology, that human beings are predictably self-undermining, and that this impaired decision-making capacity has real consequences.
In order to better act on our self-interest, these economists argue, we need to look to Homer, rather than Becker, as our guide. In Book XII of The Odyssey, our hero Odysseus famously faces the challenge of passing by the Sirens’ isle. Homer states that Odysseus tells his men to seal their ears with wax and then to
take me and bind me to the crosspiece half way up the mast; bind me as I stand upright, with a bond so fast that I cannot possibly break away, and lash the rope’s ends to the mast itself. If I beg and pray you to set me free, then bind me more tightly still.
Thomas Schelling, in his great essay “The Intimate Contest for Self-Command”, took Odysseus’s solution to his problem as a profound challenge to standard liberal and economic theory, which argues that we should defer to the preferences of citizens. But deferring to which preferences, Schelling asked, would be more just? To those of Odysseus at the moment of maximum temptation, or those he held with some rational distance from the decision?
This feature of human nature—that our appetites are nearly irresistible when faced with temptation, but that we are capable of cooler calculations of interest in advance—is not so much a challenge to classical liberalism as a recognition embedded directly in its heart. Liberalism is not rooted in a sunny or calculating view of human nature. One need only consult Hamilton in The Federalist (No. 6), where he observes that “men are ambitious, vindictive, and rapacious.” The purpose of designing markets or constructing constitutions was that men could recognize their own flawed nature and design institutions capable of both restraining their worst impulses and directing them toward virtuous ends. The earliest versions of liberalism differed little from their more repressive predecessors in their understanding of man’s capacity for rationality when faced with the siren’s song; indeed, by stressing the autonomy of decision-making from an overbearing state and an established church, the dilemma inherent in choice within earshot of the Sirens became more problematic. Nonetheless, they believed that, back on shore, mankind was capable of doing better.
This is the challenge that Cass Sunstein’s recent book, Why Nudge? The Politics of Libertarian Paternalism, would seem to address. Behavioral economists have proposed a wide range of social intervention in response to predictable individual decision-making failures. These sorts of interventions are not necessarily outside the scope of a proper republican government. Achieving the basic ends of government requires that the government be granted sufficient power and authority to coordinate the actions of citizens and social organizations and to prevent their incoherence and disintegration, even, in extremis, by force. The danger with all such power, however, was succinctly stated by Hamilton in The Federalist: “In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself.” Having empowered a state to attain our collective ends, how do we ensure that it pursues those ends rather than those that suit the persons given control of governmental power? No theory of government worth its salt can fail to address both questions, and of course it must address them simultaneously.
Libertarian paternalism seeks to authorize a very wide range of tools by which the state might “control the governed.” But how do we ensure that a government so equipped is “oblige(d) to control itself”? Such a concern begs the prior question, too, of who deserves the right to judge individual decisions to be “failures”, or to be suboptimal from a larger social perspective in a political culture that still values individual liberty. That is the real clot of issues involved with the “politics of libertarian paternalism”, and one that Sunstein fails to address adequately. His failure is instructive, however, for it raises serious questions about the future of “big government”—in particular, what sort of big government we might want, or might ruefully think we need.
Sunstein is right to assert that almost all of the great figures of classical liberal thought were, in his words, “choice architects.” But classical liberalism provided a fairly narrow canvas on which choice architects could work in applying their insights into human psychology. The question posed by Sunstein’s work (and that of others) is how far we can go in governing human beings in the name of their own preferences and still claim that we are operating within the context of liberalism?
This is not a mere question of political theory. Various forms of paternalism, justified by psychology and behavioral economics, are all the rage in governments in the United States and beyond. Sunstein observes that in the United Kingdom, “The Behavioural Insights Team has used this research to promote initiatives in numerous areas, including smoking cessation, energy efficiency, organ donation, consumer protection, charitable donation, and compliance strategies in general.” While economists, newly unshackled from their previous insistence on treating human beings as calculating machines, are giddily revealing innumerable deviations from rationality, policymakers across the world are using these findings as justifications for new forms of state action. The presumption here is both clear and arresting: Individuals are not value-maximizers, but governments either are or can be.
What is peculiar is that this newfound mania for “nudging” is happening at the same time as government acting in some of its more traditional forms appears to be shrinking. In the United States, the share of GDP that goes to the domestic discretionary budget—pretty much everything other than entitlements, defense and interest payments—is due to shrink from 3.5 percent of GDP today to about 2.5 percent by 2022. That’s even lower than it was in 1962, before the Great Society. But that does not necessarily mean that government is shrinking. Rather, the effort to squash government appears to have caused it to spread in all sorts of unfamiliar ways. A government that is strapped in its ability to fund what it does itself is also the same government increasingly involved in what its citizens put into their lungs and stomachs. Government is growing in its influence over private spending and conduct, even as its public face as revealed in the Federal budget ebbs.
It would stretch matters only slightly to say that Why Nudge? could serve as a book of common prayer for this new political religion. A government both libertarian in Sunstein’s understanding (because it shies away, for the most part, from direct coercion) and paternalistic (because informed by the ubiquity of self-undermining human irrationality) is one that will be simultaneously nowhere and everywhere. Sunstein’s vision is not absurd, but there are costs to a liberalism more focused on nudging a bit everywhere instead of shoving society in a more circumscribed but definitive way.
The primary mission of Why Nudge? is to establish the argument for paternalism, the “libertarian” modifier being a bit of a dodge. We can imagine two sorts of justification for paternalism. Sunstein primarily associates himself with the claim for paternalism that does not rest on the belief that individuals want the wrong things, but that they are incompetent in attaining their ends without governmental assistance. Sunstein often refers, however, to another justification (he spends little time on it, treating it mainly as a straw man), which argues that the human good can be determined objectively and imposed on individuals regardless of how they understand their own interests.
Sunstein claims that his alternative is “libertarian” because it accepts the relativity of the human good. “At Time 1, the person chose to smoke, to drink, and to eat a lot; at Time 2, the (same) person wishes that she had made none of those choices.” Libertarian paternalism is the theory that prefers the preferences of individuals at Time 2. And, indeed, the best (or at least the most liberal) justification for paternalism is that which Schelling derived from the tale of Ulysses and the Sirens—as a form of self-binding.
Ah, but Ulysses ordered himself bound. What if someone else gives the order? The metaphor is that in a period of relative rationality we can and should restrain our future self from doing things in moments of temptation that, in a later period of repose, we wished we had not. But in a governmental context those who hold power are binding some of us whether we wish to be bound or not. The metaphor is intended to break down our normal ideas of what constitutes consent by complicating, or obfuscating, the agent who can be said to offer consent.
Sunstein makes a second pirouette that he also claims puts him on the “libertarian” side. He distinguishes between soft and hard paternalism, where rigidity is measured by the coerciveness of the measures deployed. Sunstein prefers measures that get individuals to do what it is rational for them to do through devices that shape how decisions are presented to them (such as rules on how potentially harmful goods are displayed), rather than eliminating the possibility that they might consume them entirely. Individuals in Sunstein’s “soft paternalism” world would still be free to engage in self-harm, but they would be fortified in various ways against the temptation to do so thoughtlessly.
But when Sunstein pushes himself, the libertarian components of his argument collapse: “Choice-preserving approaches are usually preferable, but under imaginable assumptions about costs and benefits, the best approach to a palpable neglect of the future might turn out to be a mandate or a ban.” While Sunstein talks a great deal about the justification for paternalism as a means of support for the rationality of “System 2” over the impulsiveness of “System 1”, in the end he’s not willing to defer to individual choice if it leads to a “palpable neglect of the future.” In which case the language about helping people using their higher values over their lower ones seems to be a bit of a smokescreen for an argument designed to legitimate enforcing on an individual what the regulator takes to be superior judgment.
The core of Sunstein’s argument for libertarian paternalism is its inevitability rather than its desirability. He frequently claims that the argument against designing choice architecture depends on the premise that all choices operate in a non-neutral context. Libertarian paternalism merely argues for designing that context consciously based on scientific evidence of what would serve individuals’ own goals, rather than haphazardly or on the basis of corporations’ desire to create contexts in which resisting temptation is nearly impossible.
This is hardly a frivolous argument, but it raises all sorts of problems—problems that Sunstein tries to parse. Much of the argument in Why Nudge? is directed at this level of analysis, which is to say that it is focused on defending libertarian paternalism against arguments that it is somehow illiberal in principle. But the really interesting questions about libertarian paternalism are those hinted at in the subtitle—The Politics of Libertarian Paternalism—but never really addressed squarely in the book.
What is intriguing about Sunstein’s argument is that, while he is enthusiastic about the good that “choice architects” will bring in the future, he seems willfully unaware that he is describing a nascent professional class, one that will inevitably develop its own shared beliefs and interests. Most simply, the supply of people trained as choice architects (and that pretty well describes the Ph.D.s being trained in behavioral economics in some of our best university departments) will be induced to increase demand for their services. They will not just be “in the service” of policymakers, for they will actively try to encourage policymakers to buy what they’re selling. Sunstein’s book is, of course, a part of that project. With the supposed mantle of science on their side, they have a powerful set of tools to get policymakers to either take their advice or to transfer power over to them directly.
Even more problematic, however, is the risk that choice architects will wear away some of the powerful institutional and cultural tools we have to protect ourselves from the instinct to “help” other people through the instrumentalities of the state. Sunstein argues that, “People are sometimes bad choosers, making their lives go worse (and get shorter). In some cases, public officials are in an excellent position to help.” But the human desire to help can very easily shade over into the desire to rule—in other words, to move from benevolence toward social authoritarianism. The Harm Principle, which John Stuart Mill famously laid down in On Liberty, and which Sunstein devotes so much energy to debunking, is one of those tools that we have developed over the years to create a kind of firewall against the urge to help people’s “lives go better.” The Harm Principle does not claim that human beings don’t frequently make errors, simply that correcting those errors is outside of the scope of legitimate state power. Mill’s principle was certainly overdrawn, but in its absolutism it helped buttress those who wanted to push back against social interventions, especially those clad in powerful guises such as science.
More powerfully, the Harm Principle is itself a check on the psychological flaw that is the will to power. In modern societies, the will to power does not primarily manifest itself in the tyrannical exercise of raw coercion. It springs more commonly out of benevolence. Anti-smoking advocates, for example, genuinely believe that they need to protect smokers from the harm they do to themselves, abetted by unscrupulous industries that prey upon their weakness. And there are, indeed, millions of smokers who sincerely want to quit and are manipulated by industries that depend upon their inability to do so. But the desire to help is frequently accompanied by a psychological flaw just as powerful as those Sunstein identifies. The temptation to help by substituting one’s judgment for that of another is so powerful precisely because it is experienced as benevolence, and not as the will to power. Libertarian paternalism, because it combines modern liberalism’s twin principles of science and compassion, can serve to conceal that what it is justifying is the exercise of power. The desire to help, which is a virtue, can become vice when it knows no limit. And libertarian paternalism, at least in Sunstein’s version, shrinks from defining what those limits are or should be.
In a telling passage, Sunstein observes that libertarian paternalism has been characterized as foretelling “the Rise of the Psychological State.” He notes:
This term is not the best advertising, because it seems a bit alarming; no one is likely to vote for a candidate who says that he supports “the Psychological State.” But the term has the virtue of spotlighting efforts, all over the world, to develop sensible, low-cost policies with close reference to how human beings actually think and behave.
The insistence that all that is being proposed are “sensible, low-cost policies” is intended to act as a sedative against overly excitable persons who think that something untoward is going on with libertarian paternalism. But perhaps there is a good reason to be jumpy. State capacity created for seemingly uncontroversial ends is still state capacity. There is an altogether reasonable, non-excitable concern that the power to manipulate incentives and choice architecture in the ways that Sunstein wants for putatively justifiable purposes will make it available for non-justifiable ones as well.
Sunstein makes clear that the machinery of libertarian paternalism is not always likely to be deployed for the subjective ends of individuals themselves. He observes that the purely informational role of government is always likely to be partial (because there is only so much information that can be effectively communicated), and to be driven by the ends of regulators. For example, in one case “government chose a fuel-economy label, rather than an acceleration label or a coolness label, because fuel economy was the feature of cars on which public officials sought to focus consumer attention.” Increasing the efficiency of cars is indeed an important part of the state’s effort to reduce global warming, and may justify increasing its salience through mandatory presentation of information to consumers. But that has nothing to do with enhancing the ability of individuals to attain the ends they choose in moments of cool rationality, and everything to do with advancing collectively determined goals that are seen by regulators as more important than those of individuals.
Most distressing, for a book supposedly on the politics of libertarian paternalism, there is very little here on the core of politics, which is principally about the mobilization of consent. That’s because no such core is anywhere recognized or acknowledged. Sunstein says nothing about how his favored range of interventions can be expected to generate the power necessary to overcome the opposition one would expect from changes sufficient to alter significant social outcomes. Why? Because it is fairly obvious that you will not get masses of ordinary Americans flooding the streets with signs saying, “Nudge Me!” or “Stop Me Before I Snack Again!” The liberalism of nudgers is a recipe for an unmobilized citizenry, not only because these sorts of interventions are unlikely to attract citizens to ensure their enactment, but also because their relatively low-profile, under-the-radar quality will attract, at best, modest interest group support from professional nudgers and those who stand to be economically advantaged by the nudge in question. The more liberalism seeks its ends through nudging, the more it both depends upon and generates a prostrate citizenry—one insufficiently mobilized to actually challenge deep structural inequalities, let alone to explicitly recognize who has leased its consent.
One can sum up what has been said so far into a pretty rigorously libertarian set of objections to Sunstein’s arguments. And in some ways they are such a set. But the real problem with the worldview that comes through in Why Nudge? is not its libertarian pretensions, but rather how it depicts liberalism. It has to do with what sort of big government we want, or may concede that we need. The simplest way to make this distinction is to think about the distinction between a liberalism of nudges and one of shoves.
A government of nudges is opposed to any constitutional principles governing where it may and may not act, because it views all such questions as empirical in nature and thus not subject to being cabined before the fact. It acts, for the most part, in relatively small ways, but over a huge, sprawling canvas. It prefers to be organized technocratically and wants to create islands of expert governance insulated as much as possible from popular control or enthusiasm. It prefers to leave most decisions at least nominally in private hands, but then layers regulation and subsidy on top to push private actors to produce publicly approved patterns of outcomes. So it adds up to a form of micro-stealth paternalism with the potential to produce macro-social consequences.
One can imagine another kind of liberalism. This vision would limit government fairly rigorously, with a very sharp distinction between what government should and should not concern itself with. So while Sunstein and his fellow behavioral economists have spent a great deal of time trying to figure out how to get individuals to save more effectively in private retirement accounts, a liberalism of shoves would increase government’s role in retirement by simply increasing the amount of income replaced by Social Security—by taxing and spending—while eliminating our current regime of regulating and subsidizing IRAs and 401(k)s. Instead of trying to figure out how to get people to more effectively put away money for their children’s college education, it would simply increase Pell Grants. In lieu of using informational campaigns and regulation to reduce global warming, it would impose significant carbon taxes.
Overall, a liberalism of shoves seeks to attain public ends through a transparent application of governmental power. It foreswears Sunstein’s effort to attain liberalism’s ends through fairly low-profile acts of micro-governance in the name of uncontroversial ends, and instead recognizes that most social problems are the result of power imbalances that can only be rectified through shifts in social structure. Where government acts in the name of the liberalism of shoves, it does so by more or less occupying the field, and is legitimated by having secured a clear political mandate to do so. But beyond big interventions with a few big structures, it seeks to bind its own hands against the temptation to act upon everything. A shoving liberalism looks for true torque points; it abjures a capillary-level form of pervasive, statistically based manipulation.
A liberalism of shoves is one that is intensely aware of the injustice that can be produced by the pure operation of individual choice and market competition. But it is equally sensitive to the dangers that come from overburdening government with more discrete tasks than it can effectively administer, or over which it is possible to exercise serious oversight. A government that is everywhere and nowhere—and that is the risk that the liberalism of nudges runs—is one for which oversight will be immensely challenging. By attempting wherever possible to “occupy the field” over which it operates, the liberalism of shoves facilitates oversight and maximizes responsibility; it makes it very difficult to shift blame, because it is clear who has jurisdiction over the outcome.
There will certainly be occasions on which we should rely on nudges more than we already do, and in those circumstances we ought to do so with as much forethought and care as possible. But what is frustrating about Why Nudge? is that it seems so inopportune. On the one hand, where government needs to act with more decisiveness and clarity, as in dealing with our broken health care system, it is, even after the passage of the Affordable Care Act, fractured and weak. But on the other hand, American government is a sprawling mess, incapable of binding its own hands to prevent itself from being exploited by organized interests who use its openness to extract rents from the unorganized. This is a time for focusing government on what only government can do, through the direct application of state power legitimated by democratic politics, and aggressively pruning its spread beyond that. Why Nudge? is not entirely wrong. But it is, at this moment in our history, a series of answers to the wrong question.