In early 1993 a group announced itself in Medellín under a name formed from the initials of “persecuted by Pablo Escobar”. Within months it had killed lawyers, accountants, relatives and associates of the organization it opposed, burned properties and destroyed the network that had made the cartel's offensive possible. It was, by the only measure it proposed for itself, extremely effective: the organization it targeted collapsed within the year. It also contained several men who would go on to command the paramilitary confederation responsible for a large share of the massacres documented by Colombia's Truth Commission. Both halves of that sentence are the file.
// The theoryOrder first, and the criminology that said so
Right realism emerged in the 1970s and 1980s as a reaction against a criminology its proponents considered self-indulgent. James Q. Wilson's Thinking About Crime (1975) made the founding argument: root-cause explanations may be true and are largely useless to policy, because governments cannot rebuild families or abolish inequality on the timescale in which people are being victimized. What governments can do is alter the immediate situation — raise the probability of apprehension, incapacitate frequent offenders, and above all maintain visible order, which is the thesis Wilson and Kelling developed as broken windows. Alongside it ran a harder strand: Wilson and Herrnstein's Crime and Human Nature on individual differences, and Charles Murray's underclass thesis, which attributed rising crime to welfare-induced dependency and family breakdown. The tradition is easy to caricature and its central observation is not silly: where order visibly fails, the population does not become philosophical about root causes, it arms itself or submits to whoever is armed. That is a description of Medellín in 1992, and it is why this case belongs to right realism rather than to the critical tradition that would ordinarily claim a paramilitary file.
The state's monopoly is not a formality
What this case tests is the one premise right realism shares with every other tradition and never examines, because in its home jurisdictions it was never at risk: that the provider of order is the state. Remove that and the rest of the argument survives intact and produces something else entirely. Los Pepes supplied exactly what the framework recommends — certainty of consequence, rapid response, the incapacitation of frequent offenders, visible demonstration that predation would be answered — and they supplied it without warrant, without evidentiary standard, without appeal and without any mechanism by which a mistake could be identified, let alone corrected. Max Weber's definition of the state as the holder of the monopoly on legitimate violence is usually taught as a piece of political sociology; this is the case that shows why it is a criminological proposition. An order-provider that is not accountable does not merely risk abuse: it has no way of distinguishing between an enemy and an inconvenience, and no incentive to acquire one. The documented trajectory follows directly — the same networks, financing and personnel scaled into a confederation whose victims were overwhelmingly civilians with no connection to trafficking at all.
// The subjectA real grievance, an effective method, and what it built
The biography matters because it explains the recruitment logic. Castaño's brother was kidnapped and killed by a guerrilla group in the early 1980s, and the account he gave of his own formation always began there — a private grievance against an armed actor the state had failed to control. That is the standard entry route the research on paramilitary recruitment documents: not ideology first, but a specific unanswered harm, followed by an organization that offers the answer. The trajectory afterwards is equally documented. The alliance with the state's pursuit of the Medellín organization gave the group operational cover and legitimacy; the money and land accumulated in the process gave it independence; and the confederation that emerged in the later 1990s pursued objectives — territorial control, land appropriation, the elimination of civilian populations suspected of sympathy — that had nothing to do with the original grievance.
On sources, the same caution as every real case here. The dramatization compresses and invents; the record is in the judicial files, in the Justicia y Paz testimony and in the Truth Commission's final report, and the victim numbers in this file come from the last of those.
Carlos Castaño
Get the map of theories for free
Subscribe to the criminal archive —one case a week, no spam— and I'll send you The map of theories: the great currents of criminology on a single sheet.
// The crackThe tradition's own evidence, and what it will not look at
Right realism has four problems, and this case is the fourth. First, its flagship policy does not hold: broken windows produced the most cited empirical claim in the tradition, and Harcourt and Ludwig's reanalysis found the New York effect was mean reversion, while the Moving to Opportunity data gave no support for the disorder-to-crime pathway. Second, the underclass thesis was ideology ahead of evidence — the predicted relationship between welfare provision and crime runs the wrong way in cross-national comparison, where the states with the most generous provision have the least serious violence. Third, its enforcement-first prescriptions produced stop-and-frisk, mass incarceration and a documented collapse of trust in the communities most exposed to them, which by the tradition's own logic is self-defeating, since the informal control it wants depends on that trust. And fourth, the premise this file is about: the framework assumes an accountable provider and has nothing to say when the provider is not. That is not a marginal scenario. It is the situation of a large share of the world's population, and the research on private order — Gambetta on mafia protection, the literature on vigilantism and lynching, the Colombian paramilitary record — converges on one finding: private order is real, it is often effective against the specific threat that produced it, and it reliably becomes a predatory institution because nothing external constrains what it does next. The part of right realism that survives is its starting observation, and it is worth keeping: an absent order is not neutral, and somebody will supply it.
Order without accountability
Castaño commanded and financed structures responsible for massacres of civilians, and confessed to a great many of them himself. The grievance that began it was genuine, the cartel his group destroyed had done everything it was accused of, and neither fact touches the account. This archive files the case under right realism precisely because the tradition is the one that took the demand for order seriously, and because the answer it never had to give is the one this case forces: order supplied by an unaccountable provider is not a lesser form of the state's order. It is a different institution with its own interests, and it does not stop when the original threat does.
Three key ideas
- Right realism argued that root-cause explanations are useless to policy on the timescale in which people are victimized, and that the state's first duty is visible order. Its founding observation holds: where order visibly fails, people arm themselves or submit to whoever is armed.
- The premise it never examined is that the order-provider is the state. Los Pepes supplied exactly what the framework recommends — certainty, speed, incapacitation, visible response — without warrant, evidentiary standard, appeal or any mechanism for identifying a mistake.
- Its problems: broken windows does not survive reanalysis, the underclass thesis runs backwards in cross-national data, enforcement-first policy destroyed the trust its own theory depends on, and private order reliably becomes predatory because nothing constrains what it does next.
Sources · to keep reading
- Wilson, J. Q. (1975). Thinking About Crime. Basic Books.
- Wilson, J. Q. y Herrnstein, R. J. (1985). Crime and Human Nature. Simon & Schuster.
- Harcourt, B. E. y Ludwig, J. (2006). “Broken Windows: New Evidence from New York City”, University of Chicago Law Review, 73(1).
- Comisión de la Verdad de Colombia (2022). Hay futuro si hay verdad. Informe Final.
- Romero, M. (2003). Paramilitares y autodefensas, 1982-2003. IEPRI.
- Gambetta, D. (1993). The Sicilian Mafia: The Business of Private Protection. Harvard University Press.
Case-file questionsFrequently asked questions
Who is Carlos Castaño?
Carlos Castaño (“Los Pepes”) is a villain from Narcos, analysed on KRIMINIS through the lens of criminology. His argument was that somebody had to answer.
Why does Carlos Castaño become a villain?
Carlos Castaño, Los Pepes and right realism: the private provision of order where the state does not reach. The file on the tradition that put order first — and on what happens when somebody other than the state does it.
Which criminological theory explains Carlos Castaño?
Carlos Castaño's case is analysed through Right realism. Right realism — Wilson, Kelling, Herrnstein, Murray — reoriented Anglo-American criminal policy with a simple thesis: a state's first duty is order, criminology had spent decades explaining crime instead of preventing it, and an authority that does not protect produces predation. What the tradition underestimated is what happens when that vacuum is filled by somebody other than the state. Los Pepes, the group organized explicitly in 1993 to attack the Medellín cartel's structure, are that experiment. They worked, they were effective, and out of them came a paramilitary structure that years later was committing massacres on a scale the cartel had never reached.
