The year was 2016. I was deputy commander of a Military Police battalion in one of the most dangerous neighborhoods of the capital of Pará—one of the poorest and most violent states in Brazil. At 3:40 in the morning, the radio broke into an intense exchange of dispatches, unusual for a Wednesday. The operations center gives me the picture: a sequestro relâmpago in progress, the “lightning kidnapping”: criminals take over the victim’s car at a traffic light and force cash withdrawals from ATMs. I head to the scene. It looked like a movie—negotiators, press, ambulances. The negotiation ran past sunrise, and at six o’clock the two kidnappers surrendered without hurting anyone. Both were 19 years old.
That same afternoon—to absolutely nobody’s surprise—they were already out. At the custody hearing, a procedure whose underlying goal, in practice, is to decarcerate (an ideology deeply rooted in the Brazilian judiciary), the judge basically asks whether the detainee was mistreated by the officers. No victim or police officer is heard. The accused walk free. Next case. The victims’ money, the trauma, the cost society paid to arrest the criminals safely? None of it matters.
That closed one more successful episode for the state: the police showed efficiency, the judge proved worthy of the robe, the press got its footage, the criminals gained notoriety with organized crime. The victims—that middle-aged couple coming home from their daughter’s birthday party—never received so much as a visit from a state psychologist.
I spent 28 years watching episodes of this macabre series called crime in Brazil. What I was watching has a name. I found it, years later, in the work of a French anthropologist. Before we get to him, though, two economists tried to solve this puzzle, without success.
Two Tests, One Failing Grade
Gary Becker, in his 1968 essay “Crime and Punishment: An Economic Approach,” treated the criminal as a rational agent and asked what the “optimal” amount of crime for a society would be, weighing enforcement costs against the value taken. A prisoner produces nothing: imprisoning criminals is expensive and repairs none of the victims’ losses. A fine at least transfers resources to someone instead of burning them in costly prisons. That arithmetic belongs to the Chicago School, not to me.
Murray Rothbard rejects Becker’s utilitarian framework entirely. In The Ethics of Liberty, crime is aggression against someone’s just property, and the only justice is double restitution to the victim: give back in full what you took, and pay the same amount again for having taken it. Prison, in the Rothbardian view, is also useless: the victim receives nothing and is then “robbed” once more through taxes to feed and house the criminal.
A utilitarian and a natural-rights theorist diverge on method and converge on the conclusion: prison fails. What Becker calls waste, Rothbard calls injustice, and these two are almost never on the same side of anything. A policy that fails both tests should be fragile; in practice, in Brazil and elsewhere, the opposite wins: “more cages, more votes” keeps setting electoral agendas around the world.
The Standard Answer, and Why It Is Not Enough
Public choice has a ready answer: prison guard unions, construction lobbies, prosecutors and police building political careers on conviction counts—concentrated benefits, dispersed costs.
Fine. That explains who profits from the appetite for punishment, but not the appetite itself. Even the voter in a gated community, who will hardly ever cross paths with a street criminal, keeps voting for the “more cages” candidate and shrugs at anyone talking about compensating the victim. The lobby surfs this appetite. But who created it?
The Sacrifice Factor
The French anthropologist René Girard spent a career troubled by that question. In Violence and the Sacred and The Scapegoat, he argues that archaic societies contained the tension born of imitative rivalry through a sacrificial mechanism: choose a victim, load society’s guilt onto it, kill or banish it, feel the peace return. The Judeo-Christian tradition exposed the victim’s innocence; after that, nobody sacrifices with a clean conscience, and the mechanism operates in disguise.
Girard has passed through this site before: David Gornoski, on the Human Action Podcast, cast the state itself as sacrificer. What I have not seen anyone do is point the lens where I watched this mechanism run for 28 years: criminal justice.
Understand the penal system as a sacrificial ritual: the defendant carries far more than his own act; he carries the entire tension of a society that needs to expiate its sins and contained anger through a scapegoat. Prison is that altar. What it produces is not only justice: it is public catharsis. That explains why neither Becker nor Rothbard gets traction with a society that applauds candidates shouting that “a good criminal is a dead criminal,” a slogan that wins elections in Brazil. Those voters are buying exemplary punishment as the sine qua non of social peace: society left the archaic period; the rituals were merely replaced. Calling the new ritual “justice” is the modern version of calling the sacrifice the will of the gods.
The Ritual Seen from Inside
Go back to that morning in 2016. There are details in it that took me years to understand. That day, the ritual worked halfway. The staging of the capture played its part: the pre-dawn cordon, the negotiation, the surrender, the press hungry for footage. That was the day’s catharsis. But the release by the courts prevented full consummation, and Girard has a name for that: sacrificial crisis.
When the rite fails to discharge society’s tension, the violence does not disappear: it spreads without form. A country that treats sacrifice as mere spectacle and fails to punish its scapegoats reaps both plagues at once: the victim goes without the restitution preached by Rothbard, and the crowd goes without the catharsis described by Girard. So the crowd seeks the sacrifice on its own: lynching, vigilante killings, expanding militias, a cult of police violence.
Before anyone concludes that the answer is to consummate the sacrifice properly, look at the other laboratory. The United States consummated it like no one else: it holds the largest prison population on the planet. Did it buy peace? No. Did it restitute one cent to the victims? Also no. Impunity and mass incarceration are not opposites; they are the two failure modes of the same sacrificial game.
The social tension of crime is born of an open account. Someone took and did not pay. Punishment closes that account symbolically, and that is why it must repeat forever: the audience is never sated. Restitution closes the account for real: the offender himself pays his own victim, in public, which carries within it the ritual aspect of sacrifice. Before the state monopolized justice, the legal orders that contained vengeance were restitution systems: the Anglo-Saxon weregild, the customary law Bruce Benson documents in To Serve and Protect. Public compensatory payment was the meeting point between Rothbard’s restitutive thesis and Girard’s sacrificial one, replacing blood with money to end a feud. Vengeance thrives where the account does not close. Punitivism burns the debtor; impunity tears up the debt; the victim is left with the loss either way, and that feeling of injustice is the perfect recipe for violence and populist votes.
And I owe Rothbard’s readers an honest objection, straight from the street. Restitution is elegant on paper. Most criminals I arrested could not have restituted so much as a bus fare to their victims. A nineteen-year-old with no assets does not create assets because a court ordered him to pay double. I carried that objection for years, and it made me see libertarians as dreamers who had never personally met cruel criminals.
I discovered later that the objection was aimed at the wrong target. It does not disqualify restitution; it rested on the fantasy that restitution means writing the victim a check. Rothbard admitted that restitution could be paid off with work over time, and Bruce Benson documented practical mechanisms for it in “Restitution in Theory and Practice.” Hard to implement? Yes, but do the math nobody does. We already spend fortunes to cage thousands of insolvent criminals for years. A society willing to fund that fortune in cages would also be willing to fund supervised work that pays down the debt to the victims. The obstacle is not fiscal, it is liturgical.
What This Means for Austrians
The defense of a victim-centered legal order keeps losing, and it does not lose for lack of a better argument. It loses to a ritual need, the oldest social technology our species owns, and will keep losing, election after election, until someone names the mechanism out loud: the real product of the penal system is collective catharsis, and the bill is still paid by the victim.
Becker proved the altar is expensive. Rothbard proved it is unjust. Girard explains why it is still standing. Restitution as public ritual is the only proposal on the table that passes all three tests, and it comes from the Austrian School: cheap enough for Becker, just enough for Rothbard, and the only one able to satisfy the sacrificial desire the altar never sates. And somewhere in Brazil, a couple who was driving home from their daughter’s birthday party still waits for some institution of their country to treat what was taken from them as the point of it all.
