Soft Paternalism: Compatible with Libertarianism?

We’ll look at two main candidates for soft paternalistic interventions: (1) competent people acting without sufficient information, such that if they were sufficiently informed, they would likely avoid such actions due to potential harms without compensating benefits; (2) people who lack sufficient competence or for whom we lack sufficient evidence of competence. For (2) I will focus on the claim that people with low general intelligence are apt subjects for certain interventions since they are less able than people of normal intelligence to, say, follow a medical or diet regimen.

For libertarians and liberals, soft paternalism tends to be less objectionable than harder variants, and perhaps unobjectionable altogether, if we infer that almost all people would agree to interventions owing to lack of information or lack of evident competence. Still, objectionability may vary by person and depend on how soft paternalism is communicated and enforced. Warnings are one thing, but coercively imposed warnings raise thornier issues, as we’ll see below.

Hardly anyone wants to consume poisonous food or drive on roads about to collapse. Soft paternalism in these situations may be like John Stuart Mill’s “Bad Bridge” example from On Liberty (1859). Mill argues that if we witness a person, presumably unaware of the danger, about to cross an unsafe bridge, we may temporarily impede his freedom of movement to warn him of the danger. However, if he decides to proceed anyway, we may no longer impede him. We stop the pedestrian not because we think we know better than he does, but because we presume that he doesn’t want to risk falling to his death. This situation describes a one-​on-​one interaction, but what about potential dangers that affect large numbers of people who can’t be expected to have expertise and test everything beforehand? It seems unrealistic at best to presume that each of us is responsible for gathering as much relevant information as possible before assuming any risks.

Perhaps the most general way to frame the question from a libertarian standpoint: How do we best align incentives that preserve maximal freedom while protecting against negligence or corner-​cutting that may cause great harm? One could take “caveat emptor” to the extreme and claim that only the consumer is responsible for ascertaining the safety of his environment, the products he consumes, and the services he employs. Providers of these face no liability for harms that consumers suffer. So even soft paternalistic measures to ensure people are given adequate information would not be (legally) obligatory on this view.

One needn’t abandon one’s libertarian bona fides to find this stance unacceptable. Our rights to associate and engage in free exchange presume that individuals aim for mutual expected benefit. I expect the skydiving course to provide the thrill of freefall and a functional parachute, or else I would not seek their business. I should not be the one who has to hire an engineer to inspect the parachute—by offering the service, the business claims the expertise of promising adequate reassurance of a non-​deadly outcome. If we don’t question their right to collect my money up front for their services, we should also not question my right to get a safe enough service for that payment.

That said, what are the best ways to align incentives? Are government regulations the best way to mitigate or prevent harms from dangerous products? Perhaps, but these bring costs. Enforcing compliance with regulations may require an extensive state apparatus or require businesses to face large if not prohibitive costs through safety mandates. A question becomes whether this can do more harm than good. Couple these matters with the risk of regulatory capture, as personal relationships and revolving-​door mechanisms lead regulators and industry players to collude against the public interest. Merely saying that we need regulations, and then hoping experts can devise optimal ones, doesn’t guarantee they will work well in actual practice. This is not at all to say that every regulation fails or is subject to capture, but it is to highlight the public choice mechanisms often present.

Where relevant, perhaps clear product labeling can help consumers make informed decisions. A challenge arises, however, when the relevant information is complex or hard to understand, requiring many pages of fine print that few people would read entirely. Can the needed information be conveyed clearly and pithily? Perhaps what’s more important is that we can be confident that something is safe without having to know in detail why it’s safe. Few of us know how a modern jet engine works, but most of us who fly aren’t worried the engine will come apart at 35,000 feet. Reputational effects and name brands can offer reliable shortcuts for informed decisions that don’t require expertise or more than a few lines of warnings.

In addition, for-​profit entities such as the Good Housekeeping Institute or Consumer Reports, and the non-​profit Underwriters Laboratory, offer seals of approval and provide additional assurances of quality and safety. These institutions have built reputations as reliable evaluators—their own success, in fact, depends on their reputation for incorruptibility and lack of bias. These also suggest how sometimes voluntary certification can better signal quality and safety than compulsory licensing that may only set minimal thresholds (and again be subject to capture if incentives are misaligned).

We also have legal recourse, such as torts and punitive damages for negligence or defective products. Sometimes harms may have to occur before cases are brought and damages paid, but punitive rulings can be their own deterrent that fosters precompliance and greater safety after the fact. John Hasnas and other libertarian scholars have discussed how the infamous 1994 McDonald’s coffee case shows how torts and class action lawsuits can be effective internal regulators of markets. Any regulations already in place were difficult and costly to enforce, so restaurants faced few incentives at the time to keep their coffee at safer temperatures until it took a customer getting third-​degree burns, and a subsequent large settlement, for compliance to happen.

These measures taken together are not panaceas. A zero-​risk world is impossible. Just as soft paternalism recognizes that our decisions are sometimes couched in imperfect conditions, our world of interactions, despite the many benefits it can provide, is still imperfect too. At the same time, these measures suggest how a libertarian world needn’t be subject to some zero-​liability chaos where businesses can harm consumers with impunity.

Hard paternalists and anti-​paternalist critics focus on people believed to be competent and informed enough; their disagreement arises over whether people should still be allowed to act on certain preferences or values that may set back their own interests. These people’s decision-​making abilities are not questioned so much as the wisdom of certain choices they make. By contrast, the question about less intelligent people alleged to have poorer decision-​making abilities is the province of soft paternalism. Defenders of interventions with lower intelligence adults appeal to the fact that we tend to prevent children from making certain risky decisions until they reach adult status, so how would interventions with lower intelligence adults be different? If age is relevant to soft paternalism, why not also IQ?

These issues may speak to the challenge of how to institute an unavoidably arbitrary line. To paraphrase philosopher David Schmidtz: Sometimes we have to draw arbitrary lines, but it’s not arbitrary that we have to draw them. The reason it may be less arbitrary to draw a line at age versus intelligence is that age is relatively simple. We have good evidence that the normal human brain is not fully developed until around age twenty-​five. Of course, there are significant numbers of younger people who have the maturity of a wise fifty-​something, and there are many fifty-​somethings who act like toddlers. Of this latter group the best that can be said is: “You know better. Consider acting your age.”

By contrast to age, assessments of intelligence and how intelligence relates to decision-​making capacity are fraught with complications. What is sufficient general intelligence, who decides, and how? Is it determined through standardized exams made by a panel of elites, or are there widely recognized and relatively uncontroversial standards? Should we go with a threshold account that regards people as sufficiently competent if their IQ is above a certain number, or should we use a relational account of competence that tends to focus more on percentiles that needn’t reflect a Bell curve distribution? What makes for sufficient competence, especially if “full rationality” is at most a modeling construct and not an accurate reflection of how real choosers decide? Is competence general or domain-​specific? Perhaps most crucially, how does general intelligence relate to people’s preferences?

Let’s stipulate that we can identify a clear class of people (LI) whose low intelligence makes it a challenge for them to follow a doctor’s prescription or maintain a healthy diet. Some defenders of paternalism have proposed supervisors who can ensure that these patients are following instructions and not straying from what is in their best medical or nutritional interest. We then face several issues.

The discussion above focused on people unwittingly making unsafe decisions which, if better informed, hardly anyone would make due to a lack of compensating benefits. Issues of diet and medicine can differ because the relevant decisions draw from a bigger range of potential benefits and harms. If certain diets make food unappealing or more expensive, or if certain medicines have unpleasant side effects without evident compensating benefits, then why assume all LI people would still wish to eat healthy and take their medicine in these circumstances? Aren’t there also significant numbers of normal or high intelligence people who prefer not to eat healthy or take medicine they don’t like, even if they’re more able to follow instructions?

Preferences, not lack of intelligence, could explain many people not going along with some prescribed regimen. But even if supervision is only a matter of helping someone do what they actually want to do, how does this affect consent to the arrangement? Will LI people be allowed to opt out of a supervisory scheme they don’t want, immediately or after a brief trial period? If not, it doesn’t appear that supervision aims to track LI people’s actual preferences. Rather, the paternalist’s inferred preferences win out over the patient’s revealed ones not to follow the regimen and thus not to be supervised. The paternalist may reply that we don’t let children out of guardianship until they reach legal adulthood. Since children don’t get to opt out, why should LI people? A (too brief) response is that adults differ in one key aspect: amount of time to develop preferences. A young child hasn’t had the chance to determine whether a healthy diet and medication are worth it, while an adult has had more time to get feedback on where such priorities fit into his life.

What are we to make of these supervisors? Are they granted uninvited visits to the LI people’s residences to ensure compliance? Are they allowed to spy on them in restaurants and grocery stores, detaining them if needed when they’re not eating the “right” foods? With technological advances, will mandatory tracking devices or AI spies sound an alarm to authorities when LI people stray from the prescribed course? Will they keep detailed records subject to security breaches, intentional or otherwise? One might respond that health is more important than privacy, but this (at best) controversial value judgment isn’t shared by many higher intelligence people either, so why presume all LI people should accept it? Sensitive health information should only be provided by consent, not gathered whether one likes it or not, especially given the potential for abuse like blackmail.

Won’t many LI people be aware that they’ve been selected as unfit to care for themselves along these margins, while others are free to do as they wish? What might this do to their self-​respect, if they’re unequal before the law in the sense of facing coercion and possible civil (even criminal?) penalties for lack of compliance not faced by others, and especially if they don’t prefer what paternalists think they should prefer? While universal paternalism is objectionable, one advantage is that everyone, including alleged experts, is subject to its restrictions, not just a certain class of adults deemed unfit by others. The latter sounds more like a feudalistic arrangement than one consistent with liberty and equality.

What will be the compliance costs of hiring supervisors, letting them monitor patients, making sure supervisors themselves are acting within their bounds, etc.? Might these costs far outweigh whatever social costs arise from unsupervised LI people not acting healthily? What about family and volunteers helping those who seek help, instead of social workers whose dockets may already be full with what are perhaps higher priorities? A better local knowledge of patient needs and priorities may not only help with financial costs but also with the psychic costs of distrust of strangers being in one’s space without invitation.

Depending on how much the supervisors are allowed to do, I could see any number of budget-​expanding government agencies, or private firms that stand to profit, jostling to raise the bar on what constitutes minimal competence. I can also see a slippery slope from LI to people of normal intelligence whom lobbyists claim “should” be supervised to do what they ought even though they have sufficient competence to do so unsupervised. A “Baptists and Bootleggers” dynamic could arise whereby some defenders touting the importance of safety and health team up politically with others seeing greater restrictions and surveillance as helping their bottom line.

While soft paternalism suggests less intrusive and more easily justifiable interventions than hard paternalism, this discussion suggests ways in which libertarians and liberals should still be wary of progressive or conservative efforts to use lack of information or lower intelligence as excuses for more regulatory or legal interference than the situations warrant.