Paternalism and Its Limits: When Is It Justifiable to Decide for Others?

Economic theory has long asked whether it is possible to make decisions on behalf of another person in a way that is genuinely in that person’s own interest. This is the core of paternalism: under what conditions can an intervention in someone else’s decision-making be justified in the name of their welfare? Paternalism can take various forms, from parental care to state intervention. From an analytical perspective, however, the underlying problem remains the same.

“The fundamental question is whether it is even possible to help another person — for example, even a newborn child — by restricting them in some way. For instance, by preventing them from touching a hot stove. If we were to consistently follow the standard economic model, the answer would be NO,” says doctoral student Petr Krautwurm, illustrating the line of reasoning he examined in a scientific article he co-authored.

The Cashew Effect

The branch of economic theory that defends paternalistic interventions is based on the assumption that there may be a discrepancy between what a person does and what they truly want. In other words, a person may, at a given moment, make a decision that does not correspond to their deeper interests.

The American economist Richard Thaler often illustrates this with a bowl of cashew nuts: guests were tempted by freely available cashews placed on the table, thereby spoiling their appetite. When he removed the bowl, he did make a decision on their behalf, but, according to his interpretation, in their overall interest.*

A Reversed Perspective: From Goals to Choices

The authors of the article seek to justify this type of reasoning through a model. They propose a conceptual framework that explains why it makes sense to distinguish between what a person is doing at a given moment and what they want to achieve in the long term. The main idea is simple: our immediate decisions are not always an accurate reflection of our genuine, long-term goals. The authors’ own contribution lies in explaining why it makes sense to consider this distinction in subsequent theoretical debates.

“The alignment between wanting and doing is a fundamental building block of contemporary economic theory. A significant part of welfare analysis, as well as the related analysis of public policies, rests precisely on this assumption. However, if we find a systematic discrepancy between wanting and doing, a substantial part of the key conclusions of economic theory — such as consumer sovereignty — collapses,” explains Petr Krautwurm.

The argument is based on a simple idea. There is usually no direct line between an individual decision and one’s life goals; rather, there is a causal chain. Whereas classical economic literature starts from what people do and infers their goals from that, the authors of the study reverse the procedure: they first assume that people have certain goals, and only then examine how their everyday decisions move them toward those goals. For example, if a person wants to live a long and healthy life, an individual choice such as “having a salad” contributes to that goal only indirectly — through a number of intermediate steps, such as a healthier diet. The salad thus fits into a chain in which a higher intake of vegetables supports healthier eating, which in turn contributes to a long and healthy life. This chain connects everyday action with long-term goals, but at the same time creates room for error, bias, and noise.

According to the authors, this is precisely where the core of the behavioral problem lies: these intermediate steps create space for mistakes. In order to make decisions in everyday life at all, people rely on simpler rules and short-term considerations. In a specific situation, something that makes immediate sense may prevail — such as hunger, habit, or social pressure — but it may correspond less closely to the long-term goal. This is therefore not necessarily irrationality in a strong sense, but rather a consequence of decision-making in a complex world. And it is precisely this mismatch that opens up space for paternalistic intervention.

Paternalism in Practice

Such a justification of paternalism, however, is not without complications. In practice, paternalism would have to be based on a broader pattern of behavior and use that pattern to infer the direction in which a person is heading in the long term, so that it could target the relevant criterion. The key question, then, is how well someone else can correctly identify another person’s true goals.

This line of reasoning underlies some milder interventions, such as so-called “nudges”: arranging a cafeteria so that salad is more visible than cake, or more generally adjusting the environment so that it is easier for people to do what corresponds to their longer-term goals. Similar reasoning can be applied to encouraging retirement savings, mandatory health insurance, or incentives for a healthier lifestyle. In all these cases, the aim is not to fully replace decision-making, but rather to gently correct individual choices that may make sense at a given moment but, from a long-term perspective, may diverge from what the person themselves would want.

The authors of the study emphasize, however, that their model is not a defense of unlimited paternalism. The model provides only a methodological foundation, not a full justification of paternalism in all cases. The greatest problem remains practical: correctly identifying what a person truly wants. This cannot be reliably inferred from a single choice, nor from what a person says. Rather, long-term observation is needed. And even if the goals of a particular individual can be identified, the issue is still not resolved.

“People have different goals and assign them different weights, so general interventions may benefit some while harming others. Thus, while we can most likely help our child toward a better life by preventing them from touching a hot stove, we cannot be equally certain of this in the case of an adult whom we have seen for the first time,” adds Petr Krautwurm.

Effective paternalism may be theoretically possible, but in practice it is far more fragile and problematic. From this perspective, parental paternalism directed at a specific individual has a stronger justification than state paternalism aimed at the average individual.

*Richard Thaler regards this episode quite literally as a revelation — to such an extent that he devoted not only the opening part of his Nobel Prize lecture to it, but also the title of the lecture itself: https://www.nobelprize.org/prizes/economic-sciences/2017/thaler/lecture/