Prison, Violence and Giving: Why Our Instincts Mislead Us and What We Can Do

In 1978, an American documentary convinced thousands of families, elected officials and educators that they had finally found the answer to juvenile delinquency: take at-risk teenagers to visit a prison and leave them face to face with inmates describing their lives. The programme was called Scared Straight. It won an Oscar. And decades of rigorous research showed that it increased delinquency. Not that it was ineffective: that it made things worse.

Romain Barbe

Founder · Mieux Donner · Reading time: 10 min

This is a central problem with interventions designed to help others: our instincts mislead us, often profoundly. This article explores data, prioritisation frameworks, what you can do if you want to act on violence and incarceration with your money, and more broadly, where your generosity might have the greatest impact.

Why we get prison and violence wrong

Punitive instincts and the reality of the evidence

When faced with violence, our most spontaneous responses tend to be the same: punish more harshly, incarcerate for longer, show offenders what awaits them. These instincts are understandable. They reflect a legitimate need for justice and security. But they do not hold up well under scrutiny.

The idea that harsher sentences deter more crime seems logical: the higher the risk, the less a rational individual should take it. The problem is that most crimes are not committed in that state of rational calculation. They occur under the influence of substances, in the heat of a conflict, or by people who do not believe, at the time, that they will be caught.

Sentence severity and reoffending: what the research says

Scientific research has confirmed this for several decades. A review of the literature published in the Journal of Economic Literature by Chalfin and McCrary (2017), covering forty years of econometric studies, concludes that there is little evidence that crime responds to the severity of criminal sanctions.[2] Two factors, by contrast, have well-documented effects: police presence, which increases the perceived probability of being arrested, and access to stable, legal economic alternatives — employment, training, sufficient income. The latter point is significant: part of criminal activity is explained by the absence of options perceived as viable outside the informal economy. Improving those options also reduces the motivations to cross the line.

This finding is not an invitation to systematic leniency: there are situations where incarceration protects society by temporarily incapacitating dangerous individuals. But it does challenge a central assumption of many criminal justice policies, and raises a question for any donor: if the punitive instinct is often wrong, what does the research tell us about interventions that actually work?

Scared Straight: the textbook case of good intentions making things worse

It is hard to find a more striking example of the gap that can exist between intention and outcome than a programme called Scared Straight. The idea seemed obvious: take at-risk teenagers into a prison, confront them with inmates who describe their lives, and send them home frightened. This preventive shock was meant to steer them away from delinquency.

The programme was born in New Jersey in the 1970s. A documentary filmed on site in 1978 won an Oscar and convinced policymakers, educators and families across the country. Dozens of variations sprang up. In the 2010s, a television series gave it a massive new audience.

An Oscar-winning programme, a measured failure

But at the very moment the series was breaking viewing records, research was accumulating a contrary verdict. A meta-analysis published in the Cochrane Database of Systematic Reviews by Petrosino and colleagues, pooling nine randomised trials conducted between 1967 and 1992, is unambiguous:

The authors' conclusion is direct: they cannot recommend this programme as a crime prevention strategy.[5] Sensitivity analyses confirm the robustness of the result: even when excluding studies with the most questionable protocols, the effect remains negative.

Why the programme continued despite the evidence

Several mechanisms have been proposed to explain this paradoxical effect: exposure to the prison environment can normalise incarceration in the eyes of some young people, or reinforce a delinquent identity rather than discouraging it. Presentations by inmates, however intense, can fascinate more than frighten. What intuition framed as a deterrent shock functioned, for some young people, as an introduction.

This case illustrates something important for any donor: a donation can also fail to help, and sometimes make things worse. That is not a reason to stop acting; it is a reason to look for the evidence before acting.

Open Philanthropy and criminal justice reform: an ambitious bet, a valuable lesson

Understanding effective giving and the prioritisation framework

Open Philanthropy is one of the most significant philanthropic organisations in the world, founded by Dustin Moskovitz (co-founder of Facebook) and Cari Tuna.[1] It is also one of the rare organisations to make its strategic thinking public, including its mistakes. It has since changed its name to Coefficient Giving.

Its approach rests on the principles of effective giving: identifying the causes where each dollar invested can produce the greatest impact, by evaluating three main criteria. The scale of the problem: how many people are affected, and how severely? The tractability: are there effective interventions capable of meaningfully solving the problem? And the neglectedness: does the cause lack funding, meaning that each additional dollar has a stronger marginal impact?

It was through this logic that Open Philanthropy identified reform of the American prison system as a priority cause in the mid-2010s. The United States incarcerates a higher proportion of its population than almost any other country in the world. The human and economic costs are considerable. And the research, notably the report Reasonable Doubt produced by David Roodman for Open Philanthropy,[7] showed that decarceration did not appear to increase crime. The problem seemed simultaneously serious, under-funded, and potentially solvable.

$200 million and mixed results

Between 2015 and 2021, Open Philanthropy awarded more than $130 million in direct grants to organisations working on criminal justice reform in the United States. In 2021, when transferring the programme to an independent organisation called Just Impact, endowed with $50 million in initial funding, the organisation was transparent about its reasons for doing so.[6]

In its official statement, Coefficient Giving wrote that the top charities recommended by GiveWell, particularly those distributing malaria bed nets or funding vitamin A supplementation in low-income countries, offer greater cost-effectiveness than what they estimate they can achieve by funding programmes that benefit citizens of wealthy countries, including criminal justice reform.

This is a rare and valuable admission in the philanthropic world. Not an admission of absolute failure: the organisation believes it influenced the priorities of the field and contributed to several policy wins. But an admission that, viewed against other opportunities, donations would probably have had greater impact elsewhere.

Opportunity cost is a central concept in thinking about the impact of donations: choosing one cause also means forgoing all others. And when you compare the impact of a dollar spent reducing incarceration in the United States with that of a dollar spent distributing bed nets in sub-Saharan Africa, where malaria kills more than 1,100 children every day, the gap can be considerable. This is precisely what counterfactual reasoning allows: asking what else could have been done with the same resources.

What actually works: the promising approaches

Cognitive behavioural therapy: a validated approach

If punitive interventions disappoint researchers, other approaches prove more robust. Cognitive behavioural therapy (CBT) adapted for crime prevention is the most thoroughly documented example.

The underlying principle differs fundamentally from Scared Straight: rather than inducing fear, CBT seeks to modify the automatic thought patterns that precede violent or criminal acts. By working on how individuals interpret situations, manage their emotions, and anticipate the consequences of their actions, these programmes aim for a lasting transformation of behaviour. More than fifty high-quality randomised trials show that CBT-based approaches reduce criminal reoffending by 25 to 50 %.[10]

Charity Entrepreneurship and ACTRA

This is the context in which ACTRA was incubated by Charity Entrepreneurship, the organisation behind Mieux Donner. A word on that organisation, since it is central to understanding what follows.

Charity Entrepreneurship (now renamed Ambitious Impact) is a British NGO founded in 2016 whose mission is to identify and then create the most effective organisations in the world in still-neglected areas. Its process begins with a deep research phase: over several months, teams analyse hundreds of potential causes using a rigorous framework, assessing the scale of the problem, the potential for improvement through evidence-based interventions, and the neglectedness of the cause. The highest-ranked ideas are then subjected to detailed studies before being presented to prospective founders during an intensive incubation programme. It was through this process that Mieux Donner was created in 2024, identified as one of the most promising opportunities in the field of effective giving awareness.

ACTRA (Acción Transformadora) was born through the same process, selected during the cohort of late 2024. Latin America concentrates some of the highest rates of violence in the world, and only 8 to 10 % of public prevention programmes in the region are evidence-based.[8] ACTRA seeks to close this gap: it adapts culturally translated group CBT programmes targeting at-risk young men in Colombia, with the ambition of scaling through government funding once the evidence of effectiveness has been established.

What makes ACTRA interesting is the combination of a well-documented approach (CBT), a geographic context with high unmet need, and a potential for public funding that could multiply the impact of private donations.

What can an individual do?

You have reached the end of an article that, in large part, describes things that did not work. That is deliberate: understanding why our instincts mislead us is the first step toward acting more effectively. Here are some concrete reference points.

In all cases, the most useful approach is to ask the opportunity cost question. Not to become discouraged, but so that every donated dollar goes where it can genuinely make the greatest difference.

Scared Straight taught us that instinct can mislead us. Open Philanthropy taught us that even the best reasoning can lead to suboptimal choices, and that honesty about those limitations is a rare and precious form of philanthropic integrity. ACTRA and CBT programmes remind us that evidence-based solutions do exist, even if they require more patience than spectacular ones.

Frequently asked questions

Published 4 May 2026