Why Libertarians Should Reject State Paternalism

Paternalism often won’t work because it misinterprets why people might engage in risky or harmful activities, and it often treats preferences as more determinate than they are. F. A. Hayek wrote extensively about how choices within complexity and spontaneous orders, not centralized government planning, are what primarily drive successful human societies. The same applies to individuals: imagine a central official trying to plan your personal life. It would be impossible not only because they can’t know all the relevant details about your plans and pursuits, but also because sometimes you don’t know them in advance, only through an experimental discovery process. How could someone try to plan for your own long-​term good when your own good is not only unknown to them, but sometimes unknown—or, maybe, inarticulable—to yourself?

We can ask a smoker why they smoke, and they may be able to respond. Perhaps it’s a soothing habit they enjoy with morning coffee, or they enjoy the nicotine rush, or it helps them think, or they like the flavor, or they enjoy the company of fellow smokers, or it keeps them from gaining weight or drinking to excess or consuming hard drugs. But perhaps they aren’t able to articulate these values clearly; that doesn’t mean they don’t tacitly hold such values, especially as revealed by intelligible actions. Either way, are they acting against their best interests given smoking’s known health risks? It’s not clear how we on the outside can have definitive answers. Without this knowledge, it’s difficult to see how we can say with confidence that interventions to prevent smoking work. They probably work sometimes at getting someone to quit who wants to quit … except for all the times they make someone quit who doesn’t want to.

The Smithian/​Hayekian view that we are unpredictable agents endowed with free choice, often indeterminate preferences, and tacit (sometimes inarticulable) knowledge about ourselves is an important insight for combating views that assume external observers can make more accurate observations about prima facie self-​harmful behavior that may not, on balance, turn out to be against our best interests.

Moreover, we are not bound by our past choices. Defenders of paternalism may fasten onto healthy tendencies and claim they are only trying to correct our anomalous departures from them. However, as I claim in my book: “[T]endencies can form by the very same patterns of individual choice by which they can also erode.”¹ When someone behaves anomalously, in ways paternalists find harmful, they may be deviating briefly from their stated preferences and values, or they may be in the early stages of forming different preferences where the “anomaly” becomes the new normal.

Actions often speak louder than words. More importantly, we, especially as external observers, simply can’t know which way things are going unless we are mind readers, and even mind-​reading assumes there is something definite to read. Efforts to “correct” a person’s actions to be more in alignment with that person’s preferences and values will often fail because the individual often lacks fixed and determinate preferences with regard to such actions, and their values are subject to revision at each moment. Forcing an individual to do what he or she no longer wishes to do—or never wished to do—doesn’t benefit the individual if we construe well-​being, as most modern defenders of paternalism do, in terms of preference satisfaction.

To maintain that we sometimes know the agent’s good better than the agent does is to substitute our judgment for the agent’s, as if our predictions have greater authority over her than she has over her sometimes unpredictable choices.

For some people, their personhood is best respected by leaving them free to choose. Respect doesn’t best track specific value orderings even if paternalists could identify them. That is, respect doesn’t track the content of a person’s preferences and values but rather should acknowledge a person’s jurisdiction over what to do about them, for better or worse. To assume that appeal to this domain of personal choice isn’t enough to defeat appeal to coercive or manipulative interventions itself risks imposing values (the paternalist’s ones) at least on those who endorse exception-​less free choice for whatever reason (e.g., a perfectionist commitment to personal free agency through one’s own efforts and not through any coercive interventions).

A brief aside: Libertarians should avoid trying to defend principled anti-​paternalism in terms of autonomy alone. Autonomy, understood generally as self-​government, is a thicker conception than free agency, and an accomplishment made possible by repeated free choices to develop habits of independent deliberation and actions adhering to those deliberations. Understood thus, certain “perfectionist liberal” paternalists can appeal to autonomy in order to defend state restrictions of free choice that might detract from one’s autonomy. An example is nineteenth century British liberal T. H. Green’s defense of alcohol prohibition in the name of protecting people from immoderation that might detract from the development of their autonomy.

That said, aren’t there sometimes determinate preferences in play, and situations where a person’s words are in fact louder than their actions? The above section assumed that in cases where a person’s actions deviated from their stated preferences, the tiebreak should go to the person’s actions as better revealing their preferences. But surely sometimes a person fails to act in ways they claim to prefer acting due to such factors as weak will. It often takes more effort to do the wise thing by one’s own lights, rather than to succumb to inertia, temptation, or hot emotions. As my previous essay questioned, how is it disrespectful or an imposition to force someone to act in better alignment with their own professed values, at least in those situations where paternalists can assess with confidence that a person is likely deviating from what they really prefer and value? For instance, a smoker spends a lot on efforts to quit but keeps smoking—this suggests an earnest desire to quit as revealed through actions that clash with continued smoking.

A closer look can reveal why, for some people, such interventions are still impositions. Before doing that, let’s consider a “Morality Machine” that not only prevents people from violating each other’s basic rights but also forces everyone to always do the morally proper thing (or, given moral pluralism, at least what each person has decisive moral reason to do given their own values). So, instead of it being your responsibility to choose the right action, the machine is always ready to override any decisions that depart from what’s morally best. I promise to drive a friend to the airport, but when the day comes and I’m feeling too lazy, the machine makes me do it. We can leave aside the fancifulness of the example; the point is whether we’d find this objectionable if it were possible. My sense is that most of us would find such overrides problematic, not because they’re making us do the right thing full stop, but because they’re depriving our human interactions of the proper grounding in good will, intentions of mutual respect and responsibility. Such a world might be more pleasant and less conflictual, but it would come at a deep cost to the quality of our relationships and expectations of one another. Not only would I be perturbed by a machine forcing me to keep a promise I should have kept by my own intention, my friend would likely be averse to benefiting from my forced “good” deed.

If this makes sense for morality, it can also make sense with regard to the wisdom of our self-​regarding choices. Consider a “Prudence Machine” that always forces us to make the most fitting decisions. Apart from the risk of our decision-​making abilities becoming atrophied—issues that concerned thinkers like Tocqueville, Humboldt, and Mill—many of us (but maybe not all!) would anyway rather have the discretion and personal responsibility of making such decisions, not outsourcing them to a device or algorithm. Why? Here’s a possibility: most of the choices comprising our lives are not momentous in and of themselves. They are individually humdrum and banal, and yet exceedingly important when seen as concatenations of actions that originate from our daily efforts. A life well-​chosen deserves many salutes. A day well-​chosen deserves at least a high five. A machine doing it for us deserves nothing.

Choices come in several varieties:

• Willpower choices (“feats of strength”: these involve sometimes gritting your teeth and putting forth more than normal effort on occasion; e.g., following a strict workout regimen, battling withdrawal symptoms, waiting until 5 p.m. to open the wine bottle, writing until you drop

• Humdrum choices (e.g., paying your credit card balance on time—it can take the kind of discipline needed for willpower choices to consistently make humdrum choices for the better. It can also take a little discipline to put yourself in positions where you’re not reliant on willpower choices through adoption of “life hacks”, e.g., avoiding a path that takes you by the liquor store rather than going past it and trying to resist the temptation to purchase booze)

• Momentous choices (e.g., deciding whether to continue law school or pursue a different career. A series of humdrum choices can also add up to momentousness, e.g., a series of many small but wise financial decisions that lead to wealth)

It takes a lot of effort to develop a strong and autonomous character where less effort is then required (e.g., forming good habits). Likewise, it often takes a lot of hard work to be in a position to benefit from good luck. Hockey goals are often a matter of luck in one sense when the puck bounces off another player’s skate and sneaks past the goalie. However, it usually requires a lot of talent to put oneself in the position of taking a shot that results in that luck.

For some, principled freedom of choice, and the moral and personal autonomy it makes possible, are transformative values that make them stronger, and ironically, safer people. Now the fact that something can be transformative does not mean that it transforms everyone, since it often requires a lot of effort. Good things aren’t easily achievable or low-​cost. A healthy regimen is transformative. The fact that many of us aren’t in great physical health doesn’t show that a healthy regimen is a myth. Still, one might think acknowledgement of the need for sustained effort plays into the paternalist’s hands. If the wise life is harder to lead, if it’s costlier in terms of effort, then fewer people are likely to pursue it. However, personal success is a statistical outcome of always free choices to earn one’s well-​being and autonomy, not a reflection that any one person is destined to fail at achieving these. It’s possible that everybody will fail to put forth the needed effort; it’s also possible that nobody will fail. Both of these are statistical outliers, but the outcomes are determined by individuals’ choices nonetheless.

Perhaps not everyone finds ideals of free choice and autonomy attractive. Some people, for reasons such as risk-​aversion or peace of mind, may prefer opting into paternalistic schemes that can protect them from bad decisions. They are free to do that if any are available. But insofar as these measures impose values or reasons on those not committed to endorsing them by opting in, they risk violating these people’s rights of conscience even if they help some other people advance their interests.

The right to a free conscience is less controversial than full self-​ownership rights, but it encompasses the variety and extent of freedoms that libertarians find attractive. This includes widespread freedom of action, not only freedom of thought, expression, and association. As Gerald Gaus writes, “Freedom of thought is also [a] claim to be able to act on one’s beliefs and judgments … [F]or one’s deliberations instruct an agent what to do; to allow her the freedom to deliberate but not to act on her deliberations is not to respect her as an agent, an actor on her beliefs” (Order of Public Reason 354). If one believes that one ought to be free even to make bad self-​regarding decisions, then to interfere with actions reflecting those decisions is to undermine one’s deliberations about choice’s place in one’s set of values.

These matters also apply to people who, while they are not committed to a paternalistic scheme, also lack decisive reason to endorse principled free choice and personal autonomy. Perhaps they never have thought about the matter and so have no well-​formed opinion yet (or ever will). To interfere with these people on paternalistic grounds is to impose reasons that they’ve never taken the time to consider. Surely that interferes with their belief-​formation and their ability to experiment or judge whether it makes sense to opt into a paternalistic scheme. While interference here may not be as obvious or dramatic a violation of free conscience as, say, forced religious conversion, it qualifies as a violation nonetheless.

Hard paternalism differs from many other non-​paternalistic candidates for laws and policies because not everyone considers paternalism an eligible function of government. So, the question is not so much about its liability for abuse as to whether it’s a non-​starter legally speaking. Some people would prefer that there be paternalistic laws, but others may have decisive reasons to prefer none at all and reject such laws, so such laws would not be justified for these people.

By contrast, certain functions of government would be acceptable to almost everyone (e.g., dispute resolution, infrastructure, police, and military). I’m here leaving aside the question of whether such functions are better or more rightfully provided by private institutions; the main issue for present purposes is that these common state and government features are more easily justified than paternalism, if any are to be justified. It would be odd for someone to reject all dispute resolution measures, and settle for the likely resulting chaos, on the grounds that any such measure is subject to the kind of potential abuse that would likely be less harmful than the chaos of no resolution mechanisms at all. But libertarians who object that the private sector can better provide the above functions would likely agree that paternalistic functions are even lesser candidates for state authority, independently of abuse concerns.