What Do You Do With Stupid People?
Cipolla, Institutional Stupidity, and Rational Paternalism
There is an uncomfortable problem at the center of every philosophy that takes human autonomy seriously: what should be done when people exercise that autonomy in ways that are demonstrably irrational, destructive, and harmful not only to themselves but to others? The issue is not simply that people make mistakes, miscalculate risk, act emotionally, or choose values that others would reject. A serious theory of autonomy must distinguish between legitimate differences in preference and conduct that systematically produces losses without any corresponding benefit. Liberalism has traditionally been uneasy with this question because the power to identify irrationality can quickly become the power to suppress disagreement. Traditional paternalism, by contrast, is too willing to conclude that the wiser person should intervene for the benefit of the less wise. Libertarianism often moves in the opposite direction and treats the right to make mistakes as nearly absolute. None of these approaches is entirely satisfactory once one person’s exercise of judgment creates substantial consequences for people who did not choose to bear them.
Carlo M. Cipolla provides a surprisingly useful way into this problem. In The Basic Laws of Human Stupidity, Cipolla argued, partly satirically and partly in earnest, that stupidity should not be understood as a simple absence of intelligence. A person need not be ignorant, uneducated, unsophisticated, or cognitively limited to behave stupidly. He may possess impressive academic credentials, occupy a prestigious position, control substantial resources, or exercise significant social authority. Cipolla’s definition is behavioral rather than psychometric. The stupid actor is the person whose conduct imposes losses upon others while producing no meaningful benefit to himself and, in many cases, producing losses for himself as well.
This distinction separates stupidity from ordinary selfishness. A thief who takes $100 from another person and becomes $100 richer has harmed the victim, but the transaction remains intelligible. The thief has an incentive, and the incentive provides a framework for prediction, deterrence, negotiation, punishment, or prevention. Cipolla calls this person a bandit because the bandit benefits himself at another person’s expense. The stupid actor is more dangerous because the transaction does not create even that crude rationality. He may cost another person one hundred dollars while losing fifty dollars himself. In that circumstance there is no economic surplus to negotiate, no rational incentive to redirect, and no obvious strategic bargain available. Value is not merely transferred from one person to another; it is destroyed.
Cipolla’s four-part classification captures this idea elegantly. The intelligent person creates value for himself and for others. The bandit benefits himself while imposing losses on others. The helpless or naïve person benefits others while suffering a loss himself. The stupid person creates losses for others without creating a meaningful compensating gain for himself. The terminology is deliberately provocative, but the model is useful because it forces attention away from stated intentions, social status, credentials, and moral self-description and toward the consequences of conduct.
For Rational Paternalism, Cipolla’s model exposes a difficult question that professional ethics often avoids. Most ethical systems assume that people can be influenced by information, incentives, education, persuasion, or rules. They assume that if the person is shown the consequences of a decision, he will reconsider it, or that if the incentives are structured properly, his behavior will improve. But what happens when these mechanisms fail? What happens when a person entrusted with discretion over the interests of others repeatedly makes decisions that harm the client, harm the institution, harm the profession, and ultimately harm himself, even after the consequences have been explained? At that point the problem is no longer mere ignorance, inadequate disclosure, or misaligned incentives. The problem is defective judgment.
That problem appears throughout professional life. Financial advisors encounter it when clients or colleagues insist upon strategies that contradict their own stated objectives. Attorneys encounter it when clients pursue litigation that destroys more value than it can possibly recover. Physicians encounter it when patients or administrators choose courses of action that undermine the very goals they claim to value. Trustees, directors, accountants, regulators, judges, executives, and managers encounter the same problem in different forms. Every profession based on delegated judgment must eventually answer the same question: how much discretion should a person be permitted to exercise when the consequences of that discretion are borne partly by others?
Rational Paternalism cannot answer by declaring autonomy absolute, because autonomy does not eliminate external consequences. Nor can it grant the professional unlimited authority to decide which decisions are rational and which are not. The theory must therefore navigate between two dangers: irrational discretion on one side and paternalistic arrogance on the other. Cipolla is useful precisely because he forces us to confront both.
The first task is to distinguish stupidity from ignorance, incompetence, malice, and ordinary disagreement. An ignorant person lacks information, and that problem can often be corrected. A client who does not understand the difference between term and permanent life insurance is not stupid merely because he lacks specialized knowledge. The professional exists precisely because specialized knowledge is unevenly distributed. An incompetent person is different. He may possess information but lack the skill to apply it reliably. That deficiency may be corrected through education, supervision, experience, or a narrower scope of responsibility. A newly licensed professional who misapplies a technical rule because he has never encountered a complex case may need mentoring rather than condemnation.
A malicious person presents another category altogether. He may understand the facts perfectly well and deliberately harm another because the harm advances his own interest. The advisor who knowingly recommends an inferior product because it pays a higher commission may be unethical, but he is not necessarily stupid in Cipolla’s sense. He benefits while the client loses. He is behaving like Cipolla’s bandit.
Cipollian stupidity becomes relevant when the losses cease to make rational sense. An advisor recommends something harmful to the client, exposes himself to litigation, creates compliance problems for his firm, damages his reputation, receives little or no economic benefit, and nevertheless persists. An executive destroys a profitable commercial relationship because of ego or internal politics. A regulator imposes a burdensome procedure that consumes industry resources without producing a measurable protective benefit and then defends the procedure because abandoning it would imply that the original decision was mistaken. A bureaucracy maintains a policy that frustrates employees, increases costs, reduces productivity, and provides no meaningful benefit to the people it was supposedly designed to protect. In these situations, reason is no longer functioning as the governing mechanism of conduct.
Even then, Rational Paternalism should be reluctant to describe the person himself as stupid. Human beings are not fixed categories. Intelligent people make stupid decisions, sometimes repeatedly, and ordinary people may exercise excellent judgment in areas in which highly educated experts fail. A person can be brilliant in one domain and irrational in another. Expertise itself can become dangerous because competence in one field often creates unjustified confidence in unrelated fields. A gifted surgeon may be financially reckless, while a highly successful entrepreneur may exercise disastrous judgment in litigation or succession planning. For that reason, the proper object of analysis is not the inherent worth or intelligence of the person but the conduct, the decision-making process, the pattern of consequences, and the willingness to revise judgment in response to evidence.
This distinction is essential because Rational Paternalism would otherwise become an intellectual license for arrogance. A professional could simply declare that his position is rational and that disagreement is therefore stupidity. That would eliminate the epistemic humility necessary for any serious theory of professional judgment. The central difficulty is not proving that stupid decisions exist; it is determining who is entitled to identify them.
A physician may believe that a patient is foolish for rejecting a treatment, while the patient may place greater value on quality of life than longevity. An advisor may believe that a client is irrational for maintaining excessive liquidity, while the client may value psychological security more highly than expected return. An attorney may believe that settlement is economically optimal, while the client may care more about vindication than money. None of these disagreements necessarily constitutes stupidity because rationality must always be assessed relative to the person’s legitimate objectives.
This is where Rational Paternalism differs from crude paternalism. The professional should not begin by saying, “I know what is best for you.” The professional should ask what the client is trying to accomplish and whether the chosen course is logically capable of accomplishing it. The distinction is fundamental. Rational Paternalism does not require agreement with the professional’s values. It requires coherence between the client’s objectives and the means chosen to pursue them.
If a competent adult understands the consequences of an action and knowingly chooses an outcome the professional would not choose, autonomy should ordinarily prevail. A client may legitimately prefer certainty over expected value, current consumption over inheritance, liquidity over yield, or philanthropy over personal wealth accumulation. The professional’s responsibility is not to substitute his own preferences for the client’s.
The problem becomes different when the client states a clear objective and insists upon a strategy that demonstrably frustrates that objective without producing any countervailing benefit he values. If a client says that maximizing the inheritance for his children is his primary objective but repeatedly insists upon a strategy that materially reduces that inheritance, the professional is no longer dealing merely with a different preference. He is confronting a breakdown between ends and means. Rational Paternalism is strongest when it intervenes not because it disapproves of the client’s goals, but because reason demonstrates that the chosen means are inconsistent with those goals.
This creates an important epistemic safeguard. The professional must be able to demonstrate the inconsistency rather than merely assert superior judgment. He must explain alternatives, acknowledge uncertainty, identify conflicts, distinguish facts from opinions, and remain open to the possibility that the client understands something the professional has overlooked. Any philosophy built around rational judgment must preserve the possibility that the supposedly rational advisor may turn out to be the one behaving stupidly.
Cipolla’s model also clarifies a distinction that discussions of autonomy frequently blur: the difference between freedom and discretion over the interests of others. An adult should possess broad freedom to make decisions that primarily affect his own life. He may spend too much, invest too conservatively, choose an impractical career, buy an expensive automobile, donate an inheritance, or otherwise make decisions that outsiders consider foolish. A liberal society cannot protect autonomy while simultaneously eliminating the right to make mistakes.
Professional authority is different because a professional is not acting solely with his own resources. A surgeon operating on a patient, a trustee managing trust assets, a financial advisor recommending the disposition of retirement wealth, or a director allocating shareholder capital exercises delegated authority. The moment one person acquires meaningful control over another person’s welfare, competence becomes morally relevant to the amount of discretion that should be permitted.
This leads to a central proposition of Rational Paternalism: freedom should be presumed, discretion should be earned, and authority over others should require competence. The proposition is not authoritarian. It limits the professional rather than the client. The client retains broad presumptive freedom over his own life, while the professional does not possess an equivalent right to exercise unlimited discretion over the client merely because he has credentials or occupies a professional position. Authority must be justified by competence, responsibility, informed consent, and the legitimate scope of the professional relationship.
Many modern regulatory systems often invert this logic. They impose extensive procedural burdens upon clients and professionals while devoting less attention to the quality of actual judgment. They measure whether a disclosure was delivered, a form signed, a box checked, or a specific sentence included in the documentation. These requirements may serve legitimate purposes, but they do not establish that the advice itself was intelligent. Compliance can document a process. It cannot manufacture judgment.
Once stupidity is understood behaviorally rather than insultingly, Rational Paternalism calls for a graduated response rather than immediate coercion. The first response should be clarification because apparent stupidity may be nothing more than misunderstanding. The professional may not fully appreciate the individual’s objectives, information, incentives, fears, time horizon, or personal values. Before correcting another person’s reasoning, the professional should make sure he has understood it. This is particularly important because expertise creates a recurring cognitive hazard: the temptation to confuse technical knowledge with comprehensive wisdom.
When the problem is informational, education should follow. The professional should explain relevant facts, probabilities, costs, alternatives, and consequences. Education is the least paternalistic form of intervention because it expands rather than restricts autonomy. The better informed the decision-maker becomes, the more genuinely autonomous the decision can be.
If the person understands the information but demonstrates unreliable judgment in implementing decisions, structure may be appropriate. Cooling-off periods, written decision criteria, second opinions, dual approvals, investment policy statements, independent review, staged commitments, governance rules, or other decision architecture can reduce the consequences of predictable error without eliminating autonomy. The objective is not to control the person but to create conditions in which impulsive or inconsistent conduct is less likely to produce irreversible harm.
When the pattern becomes persistent and the person’s decisions materially endanger others, containment becomes justified. Containment does not necessarily require coercion. It may mean reducing the scope of delegated discretion, requiring supervision, limiting authority, or creating independent review for high-risk decisions. A junior professional may require closer oversight. A trustee’s investment authority may be narrowed. A committee may require a second approval for consequential transactions. A firm may refuse to allow one employee to create unlimited exposure simply because he occupies a senior position.
If the risk cannot be adequately contained, exclusion may become necessary. A professional who repeatedly demonstrates dangerous judgment should not retain authority over the interests of others merely because restricting or removing him seems harsh. Professionalism is not a social entitlement. It is a role that carries obligations to those whose interests are entrusted to the professional. When competence fails persistently, the authority attached to that role may legitimately be withdrawn.
The same principle applies in the opposite direction. Sometimes the client cannot be persuaded, educated, or protected through reasonable structure. He may insist upon a course that the professional regards as fundamentally incompatible with competent practice. Rational Paternalism does not necessarily justify forcing the client to accept the professional’s judgment. It may instead justify the professional’s refusal to participate. There is a critical ethical difference between telling a client that he is prohibited from acting and telling him that he remains free to act but that the professional will not use his credentials or authority to facilitate the action.
Adults must retain a meaningful right to be wrong. Without that right, autonomy becomes an empty concept because freedom would mean little more than the freedom to choose the option already approved by the expert. Rational Paternalism therefore requires a stopping rule. The professional may educate, challenge, recommend, document, and refuse participation, but when a competent client understands the material facts, appreciates the consequences, is not being coerced, and chooses an option that primarily affects his own interests, the professional must ultimately accept the decision or terminate the relationship. A duty to advise does not create a right to command.
The right to be wrong becomes narrower when decisions impose substantial involuntary costs upon others. A trustee cannot invoke autonomy to gamble with beneficiary assets. A corporate officer cannot defend reckless conduct affecting shareholders by claiming that everyone has a right to make mistakes. A licensed professional cannot invoke personal freedom to disregard professional standards while exercising delegated authority. The relevant ethical variable is therefore not stupidity in isolation but stupidity combined with power.
A foolish individual controlling only his own affairs may create limited damage. A foolish person who controls a large portfolio, an operating room, a corporation, a regulatory agency, or a public institution can magnify the consequences dramatically. The central concern of Rational Paternalism should therefore be not the distribution of stupidity, which may be impossible to change, but the distribution of unchecked discretion.
Cipolla’s intelligent quadrant also has a significant relationship to Objectivist ethics and enlightened self-interest. The intelligent transaction produces a gain for the actor and for others. It is not sacrificial. A competent professional earns a fee, the client receives valuable advice, the institution becomes stronger, and trust in the profession may increase. There is no moral contradiction between the professional benefiting and the client benefiting. In fact, a sustainable professional relationship ordinarily requires that both receive value.
This is one reason Rational Paternalism rejects the false dichotomy between selfish exploitation and altruistic self-erasure. Professional ethics is often described as though virtue requires the professional to subordinate his interests completely to those of others. Yet a professional who systematically destroys his own economic viability cannot continue serving clients indefinitely. A firm that cannot earn a return eventually disappears. A physician who is expected to provide unlimited uncompensated services cannot maintain a practice. An advisor whose ethical framework requires him to ignore his own legitimate interests will eventually leave the profession, violate the standard, or find indirect ways to recover the value being denied. Enlightened self-interest provides a more coherent model because the professional benefits by creating genuine value for others.
Cipolla’s bandit quadrant is equally revealing because much of consumer-protection regulation is designed around the assumption that professionals may exploit clients for personal gain. That concern is legitimate. Rules restricting abusive commissions, undisclosed conflicts, misrepresentation, unsuitable recommendations, and predatory practices attempt to reduce the ability of one party to benefit at another’s expense. However, a regulatory system designed primarily to prevent banditry may still do very little to prevent stupidity. A rule may limit compensation without improving judgment. A disclosure may reveal a conflict without producing competence. A fiduciary label may impose a legal obligation without creating wisdom. One can be impeccably compliant and disastrously wrong.
This is why Rational Paternalism treats competence as an ethical issue rather than a mere technical qualification. The law is reasonably good at identifying prohibited incentives. It is much less effective at creating judgment.
The problem becomes even more significant when the analysis moves from individual stupidity to institutional stupidity. An institution does not literally think or fail to think in the same way a person does, but it can generate outcomes that closely resemble Cipolla’s definition. An institution can impose losses on outsiders, impose losses on itself, consume resources, and produce little or no compensating benefit. The paradox is that such outcomes may result from the actions of highly intelligent individuals who are behaving rationally within the narrow incentives of their respective roles.
This is what makes institutional stupidity so dangerous. Individual stupidity is usually constrained by the person’s own resources and authority. Institutional stupidity can be financed by shareholders, taxpayers, clients, donors, or employees. It can be enforced by compliance departments, protected by legal counsel, legitimized by policy, and reproduced across thousands of employees who have no personal incentive to challenge it.
Institutional stupidity often begins when means become detached from ends. A legitimate problem is identified, a policy is created to address it, and the policy gradually generates procedures, metrics, reporting systems, training programs, consultants, departments, compliance functions, budgets, and careers. Over time, the organization may begin measuring compliance with the mechanism rather than success against the original objective. The instrument becomes more important than the purpose for which it was created.
At that point the relevant question is no longer whether the original objective was worthy. The relevant question is whether the institutional mechanism still works.
This distinction is essential when examining controversial contemporary examples such as DEI, ESG, corporate social responsibility, compliance systems, and zero-risk organizational cultures. Rational Paternalism should not classify these ideas as stupid merely because they are politically contested. The appropriate analysis is empirical and teleological. What problem was the institution attempting to solve? What measurable benefits resulted? What costs were imposed? Who received the benefits and who bore the costs? Were unintended consequences considered? Was contrary evidence allowed to alter the policy? Could the institution reverse course without imposing severe reputational costs on the people who created the program?
The last question is especially important because institutions frequently develop policies that become difficult to abandon precisely because careers, identities, budgets, and reputations become invested in their continuation. Once that happens, institutional self-preservation can replace the original objective.
DEI provides one example. The original propositions underlying diversity, equity, and inclusion are not inherently irrational. An organization that arbitrarily excludes talented people because of characteristics irrelevant to performance reduces its access to human capital and may damage itself economically. Removing irrational barriers is entirely consistent with enlightened self-interest and Rational Paternalism because the organization should want the most competent people available to it.
The problem begins when the corrective mechanism becomes detached from that purpose. If demographic representation itself becomes the primary measure of justice or effectiveness, the metric can begin replacing the underlying objective. Hiring, promotion, contracting, admission, or compensation decisions may gradually be evaluated by demographic outcomes rather than individual capability. Managers may discover that satisfying representation targets is institutionally safer than exercising independent judgment, while employees may learn that questioning the effectiveness of the program creates greater career risk than quietly complying with it.
At that point an organization can incur substantial administrative costs, consume managerial attention, damage internal trust, create legal or reputational exposure, and cause employees to question whether merit is genuinely recognized, while the benefits of the program become increasingly difficult to identify independently of the metrics the program itself created. The rational question is not whether diversity is good or bad, because that formulation is too crude to be analytically useful. The relevant question is whether a particular institutional mechanism improves the organization’s ability to identify talent, treat individuals fairly, and accomplish its legitimate purpose. If it does, it should be retained. If it does not, it should be modified or abandoned. Institutional stupidity begins when asking that question becomes morally suspect.
ESG creates a similar problem. Environmental, social, and governance analysis can serve a legitimate economic function. Environmental liabilities, governance failures, regulatory exposure, labor practices, litigation risk, and energy-transition risks can materially affect enterprise value. Ignoring such factors when they are financially relevant would itself be irrational. The difficulty begins when ESG ceases to function as risk analysis and becomes an independent system of corporate morality.
When that occurs, managers may pursue objectives whose relationship to shareholder welfare, customer welfare, employee welfare, or long-term enterprise value is unclear. The problem becomes more complicated because environmental, social, and governance concerns involve different objectives, methodologies, time horizons, and tradeoffs. Combining them into a unified institutional mandate can create the appearance of analytical rigor while concealing normative decisions about whose interests should prevail and at what cost.
From the standpoint of Rational Paternalism, the central question is not whether environmental or social concerns matter. The question is who authorized management to make the relevant tradeoffs and according to what standard those decisions should be judged. If reducing an environmental exposure improves long-term enterprise value, the decision can be defended within ordinary corporate rationality. If stronger governance reduces the probability of fraud or capital destruction, the connection is equally clear. The problem arises when management voluntarily sacrifices substantial economic value for a political or social objective that shareholders never authorized and customers do not value. At that point the issue becomes one of delegated authority.
Institutional stupidity becomes possible when corporate leaders spend other people’s capital to obtain moral prestige for themselves while neither the owners of the capital nor the intended beneficiaries receive a commensurate benefit. The agency problem becomes even more serious when abandoning an ineffective ESG commitment would impose greater reputational costs upon the executive personally than continuing it would impose upon shareholders economically. In such a circumstance, the executive’s incentives and the institution’s interests have diverged.
Corporate social responsibility raises the same question in another form. The phrase sounds almost impossible to oppose because the apparent alternative is corporate irresponsibility. Yet the concept is meaningful only if responsibility can be connected to identifiable obligations. A corporation should honor commitments to employees, customers, owners, creditors, suppliers, communities, and governments. It should obey the law, refrain from fraud, deliver what it promises, and bear responsibility for harms it causes. None of that requires an open-ended theory of social obligation.
The difficulty begins when responsibility expands into a vague obligation to remedy social problems unrelated to the corporation’s competence or purpose. A corporate executive may spend shareholder money on a favored social cause and describe the expenditure as responsibility even though the money is not his. A board may announce commitments to broad objectives whose success cannot be measured meaningfully. If the initiative fails, responsibility is difficult to assign because the objective itself was vague.
The result can become responsibility without accountability. Rational Paternalism requires the opposite. The greater the authority over other people’s resources, the more specific the responsibility should become. An executive controlling shareholder capital should be able to explain why a particular use of that capital advances a legitimate organizational purpose or satisfies a clearly articulated obligation. Corporate philanthropy may be entirely rational when it strengthens communities, enhances reputation, improves employee loyalty, or reflects values legitimately endorsed by owners. The problem arises when managers use delegated capital to display personal virtue. In such cases, what appears to be altruism may simply be agency cost expressed in moral language.
Compliance systems offer perhaps the clearest example because the financial services industry has developed increasingly elaborate structures intended to prevent misconduct. Many of these safeguards are necessary. Disclosure requirements, supervisory systems, suitability analysis, anti-fraud rules, recordkeeping, and conflict controls serve legitimate purposes. However, compliance possesses a built-in tendency toward expansion because its success is difficult to measure. When misconduct occurs, the organization asks why another safeguard was not added. When misconduct does not occur, it is often impossible to determine whether existing safeguards prevented the problem or whether the problem would not have occurred anyway.
The bureaucratically safest response therefore becomes another form, another approval, another disclosure, another training module, or another supervisory layer. Eventually the marginal compliance requirement may impose more cost than the risk it mitigates. Professionals spend more time documenting advice and less time thinking about it. Clients receive disclosures so lengthy that meaningful information becomes buried in legal text. Employees learn that following procedure is institutionally safer than exercising judgment. A transaction clearly in a client’s interest may become difficult to complete, while an economically inferior transaction proceeds smoothly because it fits the approved workflow.
The irony is that every participant may be acting rationally within his narrow institutional role. The compliance officer minimizes regulatory exposure, the lawyer minimizes litigation exposure, the manager minimizes career exposure, and the advisor minimizes disciplinary exposure. The client, whose welfare ostensibly justified the entire structure, may become the only participant whose interests are not optimized. This is why Rational Paternalism insists that a procedure deserves respect because it serves a legitimate purpose, not merely because it exists.
The pursuit of zero risk provides another example that cuts across government and private enterprise. No sophisticated institution openly claims that all risk can be eliminated, yet many behave as though every preventable adverse event proves that the control system failed. This creates a powerful asymmetry. The cost of excessive caution is widely distributed and often invisible, while the cost of one visible failure is concentrated and reputationally dramatic. Institutions therefore have a persistent incentive to overcorrect.
A hospital introduces another protocol after an adverse event. A school eliminates an activity because someone may be injured. A financial institution declines a legitimate transaction because approving it creates identifiable compliance risk. A corporation adds approval layers until meaningful discretion disappears below senior management. Each individual decision may be defensible in isolation, yet the cumulative result may be paralysis. The organization eventually spends more effort protecting itself from the possibility of error than accomplishing the purpose for which it exists.
This example matters especially for Rational Paternalism because RP itself could become vulnerable to the same error. If intervention is justified whenever it might prevent a bad decision, paternalism will expand indefinitely because another possible mistake can always be imagined. Rational Paternalism therefore requires proportionality. Intervention must itself be evaluated as a source of risk, cost, and error. The decision to prevent someone from deciding is itself a decision, and the rule intended to prevent stupidity can itself become stupid.
Across these examples, the anatomy of institutional stupidity is remarkably consistent. A legitimate concern is identified, an institutional mechanism is created to address it, the mechanism generates metrics and professional constituencies, and those constituencies gradually acquire an interest in the mechanism’s continuation. Evidence of failure then becomes threatening because it challenges not merely a policy but the budgets, status, careers, and moral identities attached to the policy. Eventually the institution measures conformity to the mechanism rather than success against the original purpose.
This is the institutional form of Cipolla’s stupidity because the organization may impose losses on nearly everyone involved without any participant deliberately intending the aggregate result. Shareholders lose capital, customers pay higher prices, employees lose productive time, managers lose flexibility, professionals lose discretion, and the institution loses credibility. Yet the process continues because every person operating one part of the machinery has a locally rational reason not to stop it. Intelligent people can therefore manufacture stupid institutions.
This is also where Cipolla connects with the broader idea of the banality of ethical failure. Institutional stupidity does not require villains, and it may not even require fools. It requires only enough people who perform their assigned functions without accepting responsibility for the aggregate outcome. The employee did not design the rule, the manager inherited it, compliance insists upon consistency, legal advises against deviation, consultants validate the methodology, and executives fear reputational risk. Everyone can explain why his own conduct was reasonable, while the overall result remains absurd.
The ethical failure lies in the fragmentation of responsibility. When responsibility is divided among enough actors, no single actor feels accountable for stopping an irrational process. The organization becomes capable of producing outcomes that no rational individual would consciously design. This is why institutional stupidity can be more dangerous than individual stupidity. A foolish person is limited by his own resources, while an institution can mobilize capital, authority, bureaucracy, legitimacy, and sometimes law behind an irrational idea.
The most dangerous moment occurs when the institution begins to punish the people who question the mechanism. At that point the institution is not merely making a bad decision; it is damaging its own capacity to discover that the decision is bad. The problem becomes epistemological before it becomes ideological. The institution no longer knows how to know that it is wrong.
Rational Paternalism provides an institutional corrective by insisting that authority remain connected to purpose. Every institutional intervention should be judged continuously against the problem it was created to solve. The institution should be able to identify what success would look like, what evidence would constitute failure, what costs are being imposed, who receives the benefits, who bears the burden, whether the policy can be reversed, and whether dissent is allowed to alter the analysis. These are elementary questions, but institutions frequently abandon them once a policy acquires moral or bureaucratic status.
A Rational Paternalist institution must therefore preserve dissent because dissent supplies information. Criticism of a particular DEI mechanism does not imply support for discrimination. Questioning an ESG metric does not imply hostility to environmental stewardship. Challenging a corporate social program does not imply indifference to social consequences. Criticizing a compliance requirement does not imply support for misconduct. When an institution becomes incapable of making these distinctions, ideology has displaced analysis.
The same logic applies to professional credentials. Cipolla’s insight is disturbing precisely because education, prestige, and professional status do not eliminate stupidity. Credentials may actually magnify the damage because they create trust and authority. A foolish stranger has little influence over another person’s retirement assets, while a foolish professional with respected credentials may persuade a client to move millions of dollars. A foolish layperson cannot perform surgery, but a foolish surgeon can. A foolish citizen may hold an absurd legal opinion, but a foolish judge may enforce one. A foolish observer may misunderstand risk, while a foolish regulator may impose that misunderstanding upon an entire industry.
The higher the delegated authority, therefore, the greater the ethical importance of demonstrated judgment. Professions often emphasize entry credentials and continuing education while paying much less attention to the quality of actual reasoning. Passing an examination demonstrates knowledge. Completing continuing education demonstrates participation. Following prescribed procedures demonstrates compliance. None of these necessarily demonstrates wisdom.
Professional competence requires something more demanding. It requires the disciplined application of knowledge to reality, the ability to identify relevant facts, understand the purpose of rules, recognize uncertainty, compare alternatives, evaluate consequences, acknowledge conflicts, and revise conclusions when the evidence changes. The greatest danger may therefore not be the openly ignorant person who knows that he is ignorant, but the credentialed person who possesses enough knowledge to acquire authority and insufficient judgment to question himself.
Cipolla’s stupid actor becomes especially dangerous when he has a license, a title, a budget, a bureaucracy, and confidence.
Rational Paternalism ultimately requires the professional to possess the ability to say no. Not every client request should be fulfilled, not every profitable transaction should be accepted, and not every technically legal strategy should be recommended. A professional who recognizes serious harm and participates anyway cannot always hide behind autonomy or procedure. The assertion that the client requested the action does not automatically eliminate professional responsibility, just as the assertion that a rule permitted the action does not make it ethically sound.
The difficult question is determining when refusal becomes appropriate. The professional should possess genuine subject-matter competence, the consequences should be sufficiently serious, the reasoning should be explainable, reasonable alternatives should have been considered, the client’s objectives should be understood, and relevant conflicts should be disclosed. Only then does refusal acquire genuine moral force. Even then, refusal should usually mean withdrawal from participation rather than coercive control over the client.
This balance preserves both autonomy and professional integrity. The client cannot demand professional validation for irrational conduct merely because he is paying the fee, while the professional cannot seize the client’s autonomy merely because he possesses expertise. Both remain moral agents responsible for their own decisions.
The most dangerous combination in Cipolla’s framework is therefore poor judgment joined with substantial power and insulation from consequences. Individuals who bear the full cost of their own foolishness often learn quickly because markets, relationships, reputational damage, and financial loss provide feedback. Institutions weaken this mechanism because the decision-maker may impose costs upon clients, shareholders, taxpayers, employees, or future generations while personally bearing only a fraction of the downside. Bureaucracy further obscures accountability because it becomes difficult to identify who actually owns the decision.
For that reason, Rational Paternalism should concern itself not merely with controlling behavior but with designing feedback systems. Authority should be connected to consequences, decisions should be measurable where possible, reversal should remain available, dissent should be permitted, and the person exercising discretion should receive information about the real-world results of his actions. Systems that suppress negative feedback institutionalize stupidity.
The solution is therefore not simply more rules. Another rule may simply add another layer to the problem. The deeper solution is to reconnect authority with responsibility, discretion with competence, and decisions with consequences.
We can now answer the deliberately impolite question with which this chapter began. What should be done with stupid people? The answer is not persecution, humiliation, or the elimination of ordinary freedom. Nor should society create an intellectual aristocracy empowered to declare dissenters irrational. The first obligation is to determine whether the apparent stupidity is actually misunderstanding, missing information, a legitimate difference in values, or a different assessment of risk. If the problem is informational, education is appropriate. If judgment is unreliable but improvable, structure may reduce the consequences of error. If persistent conduct creates substantial risks for others, discretion may need to be constrained. If the person cannot responsibly exercise delegated authority, that authority may need to be withdrawn. If, however, a competent adult understands the consequences and chooses a foolish course primarily with his own resources, autonomy ultimately requires allowing him to be wrong.
The same logic applies to institutions. A policy should not be condemned merely because it is controversial. Its purpose should be identified, its benefits and costs examined, its incentives understood, its unintended consequences considered, and its capacity for revision preserved. An institution that can no longer explain why a procedure exists except by pointing to the fact that it already exists has stopped reasoning.
The governing principle of Rational Paternalism is therefore proportionality. The greater the external consequences of a decision, the stronger the justification for demanding competence from the person or institution exercising authority. Human beings are neither perfectly rational nor uniformly irrational, and institutions can either correct those weaknesses or magnify them. A serious ethical framework cannot be built upon the fantasy that everyone will make intelligent decisions if merely left alone, nor can it be built upon the equally dangerous fantasy that experts, corporations, regulators, or institutions are automatically qualified to make those decisions for others.
Rational Paternalism occupies the space between those errors. It recognizes freedom without romanticizing judgment, expertise without worshiping experts, intervention without converting it into domination, and institutions without assuming that institutional authority creates institutional wisdom. Professionalism, in this framework, requires something more demanding than benevolent intentions and something far more difficult than procedural compliance. It requires the disciplined exercise of reason under conditions of uncertainty when the consequences are borne by other human beings.
Cipolla’s satire therefore ends at a surprisingly serious destination. The greatest danger posed by stupidity is not simply that foolish people exist, because they always have and always will. The greater danger arises when foolish conduct acquires authority, when authority becomes insulated from consequences, when institutions reward conformity more heavily than judgment, and when compliance with a process is mistaken for achievement of a purpose.
A rational society cannot eliminate stupidity, and neither can a rational profession or a rational institution. What each can do is refuse to give stupidity unlimited leverage. Rational Paternalism therefore arrives at a broader principle: freedom should be presumed, authority should require competence, discretion should be proportionate to demonstrated judgment, institutions should remain accountable to the purposes used to justify their power, dissent should remain possible, and the right to be wrong should be preserved so long as the costs of being wrong are not involuntarily imposed upon others.
That is not contempt for human beings. It is respect for reality.