Is the Hayekian Response to “Libertarian Paternalism” Sufficient?

In response to: Free to Err?: Behavioral Law and Economics and its Implications for Liberty

Joshua Wright and Judge Ginsburg offer a powerful critique of the Behavioral Law and Economics (BLE) movement on several fronts, but primarily based on the threat BLE poses to individual liberty.  They provide a helpful and straightforward history of the rise of Behavioral Economics, which applies a simple premise that runs contrary to neoclassical economics: people do not always act in their rational self-interest, and cannot reliably be described as utility maximizers.  BLE builds on this insight by applying it to law and public policy.  The result is “libertarian paternalism,” which relies on enlightened experts to frame choices for us so that our preferences align with what we ought to prefer.  Although Wright and Ginsburg’s critique is powerful, it also reveals some of the shortcomings of neoclassical economics, in particular some problems with its conception of liberty.

Wright and Ginsburg identify several practical issues with how BLE is likely to be implemented.  First, there is the problem of identifying irrational biases that need to be rooted out by government experts.  The assumption of neoclassical economics is that a person’s “true” preference is, simply, what that person chooses as his or her preference. (In the words of the authors, the “neoclassical assumption” is that “actual behavior reveals evidence of welfare.”)  Neoclassical economics simplifies the process of determining what promotes a person’s welfare by assuming that the person’s choice reveals what promotes his or her welfare.  By rejecting this premise, behavioral economics is required to replace that assumption with some other method of identifying what will promote a person’s welfare.  Behavioral economics simply replaces the neoclassical assumption with an assertion: “the long-run ex ante preference” is what promotes a person’s true good.  But “[n]othing…in standard economic theory or in behavioral economics justifies this approach to identifying “true” preferences.”  In short, behavioral economics has nothing by which to determine what an individual’s true preferences ought to be, and so there is nothing that helps us determine in what direction to nudge people.

Second, if the purpose of BLE is to frame choices and default options so that people choose rightly, don’t we have to identify the expert who can identify right choices?  And is this a reasonable assumption to make, especially given the very assumptions of BLE, that human beings often behave irrationally?  Don’t experts fall prey to the same problems as normal human beings?

Third, there are costs associated with “nudging” people into making the right choices.  Nudges, in the form of default rules, are still government interventions that distort market outcomes.  By ignoring the costs associated with these interventions, behavioral economists nudge the cost-benefit analysis in favor of interventions rather than letting people choose on their own terms.

However, the central claim Wright and Ginsburg make pertains not to the disadvantage of libertarian paternalism.  Rather, they argue that it is destructive of liberty, and must be abandoned for that reason – regardless of the practical effect BLE has on society.  It is “the significant but underappreciated threat to individual liberty posed by government interventions predicated upon behavioral law and economics” that is the problem.  I agree that the most powerful case to be made against libertarian paternalism is not a utilitarian argument but an argument based on the nature of liberty itself.

And Wright and Ginsburg’s response is compelling in certain respects.  Buttressed by John Stuart Mill, Friedrich Hayek, and Amartya Sen, they assert that making a decision for oneself employs a person’s thinking faculties in a way that delegating the judgment – even if it formally preserves choice – cannot replicate.  In other words, “choice architecture” by experts may hinder citizens’ ability to exercise reasoned judgment.  The argument is very Aristotelian: citizens must learn the habits of choosing that lead to developing the faculty of practical judgment. By essentially treating citizens like children in an elementary school cafeteria where the food choices are pre-selected by wiser adults, a government which nudges infantilizes citizens so that they never develop the ability to judge prudently.  Studies in social science support this claim.  According to the authors, studies in cognitive psychology demonstrate that people raised in a paternalistic state have less well developed decision-making skills as opposed to people raised in a free state.  One is reminded of Alexis de Tocqueville’s warning that governments err when they reduce citizens “to be nothing better than a flock of timid and industrious animals, of which the government is the shepherd.”

But Ginsburg and Wright’s argument provokes two important issues left unaddressed by the authors.  In one striking part of the argument, Wright and Ginsburg concede that “a minor reduction in liberty should not be sufficient to reject an intervention with significant welfare benefits just as an intervention generating only modest welfare benefits is not justified regardless of its negative effect on liberty.  In close cases it will be necessary to consider such trade-offs in order fully to assess the desirability of a proposed policy intervention.”  “[T]he best approach to evaluating these potential tradeoffs,” they conclude, “is to establish a presumption against behaviorist regulation that reduces liberty, rebuttable only by demonstrating that the regulation is likely to generate significant gains in economic welfare.”  This argument seems to resemble the “balancing” of rights that is found so frequently in Supreme Court jurisprudence.  It grants that liberty can be traded off, as long as the net gains in welfare are significant.  But this is a questionable assumption.  Why should we accept policies that infringe upon liberty, even if they produce beneficial results?

A second issue arises when the authors write that BLE “is nearly devoid of thinking about the implications of behavioral law and economics for individual autonomy.”  Strictly speaking, this is inaccurate.  Cass Sunstein, the prominent defender of libertarian paternalism, has a well-developed theory of autonomy that supports libertarian paternalism.  He argues that nudges increase individual autonomy, because autonomy means that a person has been formed properly to make the best choices for himself.  Here, the central divide between libertarian paternalists and actual libertarians is on display.  Sunstein writes, “One goal of a legal system…is not merely to ensure autonomy by allowing satisfaction of preferences, but also and more fundamentally to promote autonomy in the process of preference formation.”  The idea of autonomy, for Sunstein, “will call not merely for the satisfaction of whatever preferences people currently have, but more generally, or instead, for protection of the processes of preference formation,” such as nudges.  In other words, for the libertarian paternalist, one becomes autonomous by being formed in the right preferences, and it is the task of the law – more specifically, of government experts – to form the preferences of citizens.  This difference in the understanding of autonomy is central to the divide between libertarian paternalism and libertarianism.  According to Sunstein, autonomy is actually undermined when people are permitted to choose for themselves without the intervention of a choice architect.

One powerful response to Sunstein’s theory of autonomy was eloquently offered by Friedrich Hayek in The Constitution of Liberty.  There he wrote, “the range of physical possibilities from which a person can choose at a given moment has no direct relevance to freedom.  The rock climber on a difficult pitch who sees only one way out to save his life is unquestionably free, though we would hardly say he has any choice.”  In other words, the preservation of choice is not tantamount to the preservation of freedom.  The theory of libertarian paternalism equates freedom and choice, claiming that by preserving choices for citizens it preserves their freedom.  “The question of how many courses of action are open to a person is, of course, very important,” Hayek conceded.  “But it is a different question from that of how far in acting he can follow his own plans and intentions, to what extent the pattern of his conduct is of his own design, directed toward ends for which he has been persistently striving rather than toward necessities created by others in order to make him do what they want.”  For Hayek, freedom is not defined by the number of choices one has, but by whether one is acting according to one’s own plans, designs, and intentions, or whether one is influenced into pursuing the plans, designs, and intentions of another.

This conception of freedom is why Hayek hinges everything on the meaning of the concept of coercion, which he defines as “such control of the environment or circumstances of a person by another that, in order to avoid greater evil, he is forced to act not according to a coherent plan of his own but to serve the ends of another.”  Liberty, for Hayek, is not reducible to the issue of whether choice exists.  It entails the absence of coercion, that is to say, that the individual is able to act on his own plans and intentions, and his environment is not so deeply controlled that he acts in accordance with the plans of another.  Yet the very existence of “choice architects” in the theory of libertarian paternalism is to achieve this very result.  Libertarian paternalism is fundamentally antithetical to liberty, as long as we understand liberty as something other than the number of choices we are left with.  Proponents of BLE have not offered a good reason why we should prefer their understanding of liberty to Hayek’s.

But this provokes a final, and indeed fundamental question.  A skeptic of libertarian paternalism might not find Hayek’s objection compelling.  He might ask, why should freedom be defined as the ability to follow one’s own plans rather than be influenced into following the plans of others?  Don’t some people have better plans than others?  And shouldn’t society promote a higher freedom for its citizens, in which people still follow their own preferences, but those preferences have indeed been shaped in the right way by law and voluntary associations?

In other words, could one reject Hayek’s definition of freedom in favor of an alternative, without falling into the definition of freedom offered by proponents of BLE?  I believe that the answer is yes.  One might grant, contra Hayek and the neoclassical economists, that one’s true welfare may not be revealed by the choices one makes.  People can choose things contrary to their own true interest.  This does not entail, however, that we need experts trained in social sciences to help us make fully reasonable and hence autonomous choices.  There is a distinction between the public and the private, between government and civil society.  One’s character might be formed properly for exercising freedom in a non-licentious manner by voluntary and private institutions, as opposed to a centralized government filled with social science experts.

Ultimately, neoclassical economics’ critiques of BLE also point to the challenge facing neoclassical economics.  Insofar as neoclassical economics defines a human being’s welfare as equivalent to that person’s preferences, it seems to reduce one’s good to what one immediately desires and plans to pursue.  But this overlooks the important distinction between liberty and license.  Not all freely-made choices are consistent with a person’s welfare.  Hence, the neoclassical economist qua economist is unable to explain how liberty is preserved, even in the face of influences (both public and private) that induce citizens to exercise their liberty in a certain manner.  The neoclassical economist’s understanding of the human person is unable to provide such an explanation.  Yet this does not mean we are left with libertarian paternalism as the means by which we elevate liberty out of license.