Exclusion as Justice
Toward a More Moral End
Deportation is my favorite legal remedy.
To clarify: I am deeply pro-immigration. I believe in the free movement of people across borders as a matter of moral principle and economic efficiency. Using deportation to punish people whose only crime is being present without government permission is one of the more self-defeating and morally confused applications of state power in contemporary American life.
And yet, when a non-citizen causes serious harm to another person—assaults them, defrauds them, violates the terms under which they were welcomed—deportation strikes me as a cleaner, more humane, and more philosophically coherent remedy than imprisonment. It removes the offender from the jurisdiction. It does not cage them. It does not assert total dominion over their body and time. It says: you have forfeited your access here.
That intuition—that the most direct response to a violation is revocation of access rather than infliction of suffering—is the seed of a theory of justice I think deserves serious consideration. My first year of law school introduced me to three theories of criminal justice. Retributivism holds that wrongdoers deserve to suffer proportionally to their offense. Utilitarianism holds that punishment is justified when it produces better aggregate outcomes—deterrence, incapacitation, rehabilitation. Restorative justice holds that the goal is healing broken relationships between offenders, victims, and communities. In my experience, most law students are presented these as if they exhaust the space of serious options.
They do not.
I. The Undergirding Principle
The framework I want to defend begins with Murray Rothbard’s non-aggression principle (NAP): the initiation of force against another person or their property is never morally justified. Force is only legitimate in response to force—as defense or remedy.
This principle establishes a moral architecture in which the legitimate scope of any justice system is tightly bounded. You may respond to aggression proportionally. You may not punish people to satisfy a moral intuition about desert. You may not harm individuals because doing so produces aggregate social benefits. You may only respond, and only to the extent necessary.
Under the NAP, the minimum necessary response to most violations is exclusion: the withdrawal of access that the offending party forfeited by breaching the terms under which it was granted.
II. The Core Claim
Every human interaction takes place within a structure of boundaries. Some are physical—your body, your home, your business. Others are contractual—agreements about what you will and will not do in exchange for access to a service, a community, or a relationship. These boundaries are expressions of individual sovereignty. To respect another person is, at minimum, to respect the conditions they have set for your presence in their life or on their property.
A violation of those conditions is a breach of a specific access agreement. The most direct remedy for that breach is revocation.
If you assault a fellow tenant, the landlord’s most legitimate response is to terminate your lease. If you defraud a business partner, the most direct remedy is dissolution of the partnership and denial of future dealing. If a non-citizen causes serious harm, deportation is the revocation of access to a jurisdiction—coherent, proportionate, and requiring no cage.
The consent that legitimizes any enforcement action matters enormously here. The consent I have in mind is real consent, and it is narrow. The fiction endemic to modern state systems—that being born within a jurisdiction and lacking the resources to leave constitutes agreement to whatever rules that jurisdiction imposes—is not consent. It is captivity dressed in the language of social contract. This is manufactured consent, and I reject it as a basis for legitimate enforcement.
The only consent that legitimizes enforcement in this framework is explicit consent via signed agreement, or the implicit consent of someone who voluntarily enters property they do not own. Walk into my home and you are bound by my rules. Subscribe to a service and you are bound by its terms of conduct. Oral agreements and implied contracts, where both parties conduct themselves as explicitly bound parties typically do, can fill the remaining gaps. Parties can opt into systems that provide for such gap filling rules or offer other ways forward. That is the full scope of legitimate implied agreement. Everything else requires explicit contract.
I envision a legal system grounded in this. Private property owners set the rules on their property. Service providers codify terms of conduct. Communities adopt governance frameworks—drafted by specialized private legislators competing for adoption on the merits of their rules—that residents voluntarily choose to live under. Rights precede government. The right to set terms of access on your property, within your community, and in your contracts is a natural extension of self-ownership.
III. The Moral and Consequentialist Case
The United States incarcerates nearly 2 million people, and its incarceration rate remains among the highest in the world. If that figure reflected a system producing safety and order, a consequentialist might defend it. It does not.
Utilitarians typically defend incarceration on three grounds: deterrence, rehabilitation, and incapacitation. In the American system, the first two rest on weak ground. The Bureau of Justice Statistics found that 67.8% of released state prisoners were rearrested within 3 years and 83.1% within 9 years, and the National Academies concluded that increasing sentence length has no significant deterrent effect on crime rates. Rehabilitation programs can help, but their effects are uneven and usually modest. In practice, that leaves incapacitation as the main argument left standing.
Incapacitation is the argument that we simply need to remove dangerous people from society and hold them there. This is a more honest position than pretending the system rehabilitates.
The El Salvador model is the sharpest version of this argument. President Bukele imprisoned roughly 2% of the adult population beginning in 2022. Homicides fell from 18 per 100,000 that year to under 3 in 2023. The results are real and a utilitarian incapacitator will cite them. But the model also produced documented mass detention without due process, reports of torture and deaths in custody, and a meaningful fiscal commitment to incarceration. El Salvador has not been running this experiment long enough to know whether it produces durable safety or a pressure cooker. History suggests the latter. And the incapacitation model requires that you trust a government with nearly unlimited detention authority to use it accurately. That trust is hard to justify when wrongful convictions are so devastating and impossible to prevent: the Innocence Project’s DNA exoneration page lists 375 DNA exonerees, with an average of 14 years served, and 69% of those cases involved eyewitness misidentification. A trade of liberty for security is a Faustian bargain. History gives little reason to assume that a state this powerful will always identify the right people, for the right reasons, for long enough to make the trade worthwhile.
Beyond individual wrongful conviction cases, prisons often breed criminality. When you pack damaged people into violent, overcrowded institutions and call it justice, you should not be surprised when they come out worse. Prison can harden the very men it claims to reform, and research has found that exposure to prison can increase later criminal behavior among low-risk offenders who otherwise might have desisted. Norway understands this. Its prisons are built on normality, not humiliation; on reintegration, not permanent degradation. And the results tell the story: about 20% of released prisoners are reconvicted within two years, versus roughly 40% to 45% in the United States over a comparable window. That 2X gap is not a mystery of national virtue. It is the consequence of choosing a system designed to produce despair and then pretending despair will somehow produce safety. The American system does not merely fail to rehabilitate. It actively manufactures more of what it claims to address.
My faith adds another layer to the moral claim against indefinite and indiscriminate incapacitation. The notion that any person is categorically beyond growth and redemption is untenable in a tradition that holds repentance and moral development as central to the purpose of mortal life. Giving up on human beings is not a morally neutral position. But neither is ignoring crime and failing to defend the innocent.
Exclusion offers a different logic. For the victim, it provides targeted and direct relief in the actual context of the harm. The neighbor who assaulted you is removed from your building. The business partner who defrauded you is excluded from your industry. The remedy attaches to the relationship that was violated. For society, it creates no pressure cooker, no criminal graduate school, no festering environment that concentrates desperation and models anti-social behavior. And for the offender, it preserves the one thing incarceration reliably destroys: the possibility of a future.
IV. A Theological Pattern
For readers with a religious frame—particularly within Latter-day Saint theology—the structure of exclusionary justice may feel familiar. While violence, coercion, and force certainly exist in the Biblical corpus, Restoration theology and the revealed nature of God in Christ both often point toward a different form of redress.
In the War in Heaven, Lucifer rebels. He is not destroyed. He is cast out—excluded from the divine community, denied continued access to the presence of God.
In the Garden of Eden, Adam and Eve transgress. They are not imprisoned within the garden under tighter surveillance. They are expelled—sent into a different environment where they retain the freedom to act, choose, and grow under new conditions.
In the final judgment as understood in LDS theology, individuals are not sorted into a hierarchy of punishment and suffering. They choose different kingdoms of glory, each corresponding to the laws they are willing and able to abide. The principle is alignment between person and environment, administered through access.
Temple participation follows the same logic. Covenants are required. Those who cannot or will not keep them are not coerced into compliance. They are simply excluded from that space.
Whether or not one accepts the theology, the structural pattern is worth observing: exclusion—the calibration of access to the laws one is willing to live by—appears to be God’s preferred instrument of justice when He acts directly. It is worth asking why.
V. Against the Three Standard Theories
Retributivism holds that wrongdoers deserve to suffer proportionally to their offense. Morally, it treats punishment as an end in itself—suffering is warranted because it is deserved, full stop. This collides directly with the NAP: inflicting pain on someone who no longer poses a threat is itself an initiation of force, a new aggression rather than a response to one. I do not view suffering as morally purifying. It may hold some utility as a means, but it is a grotesque candidate for an end. And retributivism is explicitly indifferent to outcomes—it offers no theory of social improvement whatsoever. Whatever the moral intuitions driving it, they do not belong at the center of a justice system’s operating logic.
Utilitarianism is more consequentially serious, and I have more respect for its intellectual ambitions. Morally, however, it is willing in principle to impose severe harm on an individual whenever doing so benefits the collective. It treats persons as resources to be allocated. The NAP forecloses this entirely: the individual is not available for sacrifice on the altar of aggregate welfare. Consequentially, as established above, the American system fails even on utilitarian grounds—deterrence effects are negligible, rehabilitation has not scaled, and the incapacitation model produces as many problems as it solves.
Restorative justice may be my closest intellectual cousin, and I respect it more than the other two. It centers victims, emphasizes repair, and resists the punitive instinct. Where it falls short morally is in underweighting firm boundaries. Healing a relationship does not obligate the wronged party to continue granting access. Restoration and exclusion are compatible—but restorative frameworks often treat exclusion as a failure rather than a legitimate and humane outcome. Consequentially, restorative programs show real promise in limited settings and have not demonstrated efficacy at scale with serious violent offenses.
My framework differs on every axis. Morally, it is grounded in the NAP, centered on sovereignty, and treats both victim and offender as agents. Consequentially, it delivers targeted relief to victims, avoids the contagious criminality dynamic of mass incarceration, and creates market incentives for enforcement systems to produce results rather than simply produce more inmates.
VI. The Hard Cases
I have wrestled with two objections.
The first is the violent offender. What happens when someone poses an ongoing physical threat—a murderer, a serial predator, someone who has demonstrated they will harm others given access?
The NAP is clarifying here. Force in response to ongoing force is legitimate. Temporary incapacitation of someone who poses a direct and continuing physical threat is not a violation of their sovereignty—it is the minimum necessary response to their ongoing violation of others’. A temporary holding cell used to prevent force and violence does not violate the NAP prima facie. But it may only be deployed so long as is necessary to make the changes needed to prevent access to all the spaces the offender has lost access to via breach of contract. The edge case justifies the edge-case response. It does not justify a carceral system applied indiscriminately to the full range of human conduct. According to the Bureau of Justice Statistics, nearly half of all state prisoners are serving time for nonviolent offenses. And it is reasonable to assume that many nonviolent offenders acquire violent and anti-social behaviors inside prisons they did not exhibit before entering.
My proposed system would require the offender to be released to either property he owns or to the property of someone who is willing to allow his presence the moment exclusion can be guaranteed. It may seem dystopian to have areas where the most anti-social, violent offenders live together or alone, sequestered from the rest of the world. But that is exactly what our current system is, only my proposed one would include a far smaller population and retain some agency, sovereignty, and human dignity. This hard case does not redeem the current system.
The second objection follows from the moral necessity of releasing a person in holding. It is the propertyless person—someone with no property, no contract standing, no market access to any property anywhere. In a world organized around property rights and voluntary agreement, where does such a person go when universally excluded?
Two answers.
First, the comparative point: the poor are not well served by the current system. They are disproportionately incarcerated, disproportionately victimized, and disproportionately governed by rules they had no hand in designing. Formal democratic voice has not translated into meaningful protection for economically marginal people. A system of competing private enforcement bodies and private legislators with market incentives to attract the residency and business of the poor is not obviously worse.
Second: the right response to the hardest exclusion cases is to increase options rather than mandate cages. Consider the voluntary structured facility—what looks on the surface like a prison but is entered by contract. An individual with nowhere to go could voluntarily admit themselves to a facility providing room, board, and structured programming in exchange for labor or participation. Such institutions could offer market-recognized rehabilitation credentials that restore access elsewhere. What makes involuntary incarceration monstrous is the coercion, not the structure. A voluntary arrangement with genuine exit rights is morally different in kind from forced imprisonment. It treats a desperate person as an agent capable of making choices. It creates market incentives for facilities to actually rehabilitate, because people with real exit options leave bad ones. Is this also dystopian? Perhaps it sounds that way. But again, is it more dystopian than what we currently operate? My goal is not to design the full system here—it is to suggest that consent-based solutions exist that our current imagination, stunted by the assumption that cages are the only option, has not seriously explored.
VII. No Perfect System
Justice exists because human beings are flawed. Any system designed to address that reality inherits the imperfection of the problem it tries to solve. I am not arguing that exclusionary justice is perfectly moral, perfectly effective, or free of hard cases.
I am arguing that it outperforms the alternatives—on moral grounds, by centering sovereignty and avoiding domination; on consequentialist grounds, by delivering direct relief to victims and escaping the recidivism trap that renders the current system an embarrassment.
Every community, from a household to a nation to a kingdom of glory, ultimately answers the same question: who is allowed in, and on what terms? That question is not a failure of justice. It is justice. The framework we choose to answer it—and how much coercion we are willing to authorize in the process—is among the most important choices a society makes.
We may never fully eliminate force from the administration of justice. But we can decide whether it is the foundation of our system, or its last resort.
"A system of competing private enforcement bodies and private legislators with market incentives to attract the residency and business of the poor is not obviously worse."
Please elaborate.