The Logic That Feels Unassailable
Take a troubled teenager to a maximum-security prison. Let convicted lifers describe rape, assault, and decades of regret in unflinching detail. Watch the fear register. Send the kid home. Wait for that fear to keep them straight.
This is the premise behind "Scared Straight," a juvenile deterrence model that has operated in some form across more than 30 US states since 1978. The original program at Rahway State Prison in New Jersey became a national phenomenon after a 1979 television documentary reported a 94% success rate among its participants. The documentary neglected to include a control group. Congress held special hearings anyway. The model spread nationwide. A&E later turned it into a hit reality series, "Beyond Scared Straight," which became the network's highest-rated show.
The success claims were never tested against the most basic scientific standard: what would have happened to those kids without the program?
What Nine Experiments Actually Found
Anthony Petrosino of WestEd, along with Carolyn Turpin-Petrosino at Bridgewater State University, Meghan Hollis-Peel, and Julia Lavenberg, conducted a systematic review spanning every randomized controlled trial ever run on these programs. Published through both the Campbell Collaboration and the Cochrane Library (the gold standard for evidence synthesis in medicine and social policy), their review identified nine qualifying trials. All were conducted in the United States between 1967 and 1992, across eight states. A total of 946 juveniles participated, almost all male, with average ages between 15 and 17.
Seven of the nine trials reported reoffending data from official records: arrests, convictions, or police contacts. The meta-analysis produced an odds ratio of 1.68 under a fixed-effect model (95% confidence interval: 1.20 to 2.36). Under random effects, the result was nearly identical: OR 1.72 (95% CI: 1.13 to 2.62). Juveniles who attended Scared Straight and similar programs were 68 to 72 percent more likely to reoffend than those who received no intervention at all.
Petrosino tested the finding's robustness by removing studies one at a time. Drop the trial with questionable randomization: OR 1.47, still statistically significant. Drop the trial with high attrition: OR 1.96, meaning the effect grew stronger. Drop both: OR 1.68, unchanged. No single study drove the result.
A Running Tab Nobody Kept
The review quantified the harm within its nine trials. What it did not calculate was the cumulative cost across four decades of national implementation.
Consider a conservative scenario. If 10,000 at-risk juveniles per year passed through Scared Straight-type programs nationally, a modest estimate given operations across 30+ states, and the baseline reoffending rate for comparable youth is roughly 30%, the arithmetic is direct. An odds ratio of 1.68 converts a 30% base rate to approximately 42% (odds of 0.429 ร 1.68 = 0.720, probability = 41.9%). That is 1,200 additional juvenile offenses per year attributable to the intervention. Over the program's 48-year lifespan from 1978 to 2026, the total reaches an estimated 57,600 excess offenses. Each represents a victim, a court proceeding, and a young person pushed deeper into the system that was supposed to scare them away from it.
Even halving the enrollment estimate yields 28,800 excess offenses. Even quartering it leaves 14,400 crimes that would not have occurred if the government had simply done nothing.
Why Fear Backfires on Adolescents
Several mechanisms explain why these programs produce the opposite of their intended effect. Labeling theory, proposed by Howard Becker in 1963, holds that treating someone as a criminal can cause them to internalize that identity. A juvenile marched into a prison and told "this is where you're headed" may absorb the message not as a warning but as a prophecy.
Adolescent reactance compounds the problem. Psychological research consistently shows that high-pressure, fear-based persuasion triggers resistance in teenagers. Jack Brehm's reactance theory predicts that when people perceive a threat to their behavioral freedom, they become more motivated to engage in the restricted behavior. Telling a 16-year-old "you'd better not" is an invitation to prove otherwise.
There is also the exposure effect. Some programs locked juveniles in cells with parolees for over an hour. Rather than deterring criminal behavior, sustained contact with incarcerated adults can normalize it. The inmates become familiar figures rather than cautionary ones. And the group experience itself creates peer bonds among the at-risk youth, the very social networks that predict future offending.
The Strongest Case for Continuing
The most defensible counterargument comes from researchers who distinguish between confrontational and nonconfrontational program designs. Klenowski, Bell, and Dodson published a 2010 evaluation in the Journal of Offender Rehabilitation that separated programs by approach. Purely confrontational methods โ the yelling, the intimidation, the graphic descriptions of sexual violence โ were consistently ineffective. But nonconfrontational programs featuring educational discussion and inmate mentorship showed modest, ambiguous promise.
This argument has some empirical grounding. Petrosino's nine trials did not all test identical interventions, and programs have evolved since the last randomized trial in 1992. A program built around mentorship differs from one designed to terrify.
Two facts limit this defense. First, the Campbell review included both confrontational and less confrontational variants, and the pooled effect was still harmful. The OR of 1.68 is not driven by the most extreme programs alone. Second, the A&E series that aired from 2011 to 2015 documented programs that closely resembled the original Rahway model: inmates screaming, juveniles weeping, fear dispensed as medicine. The confrontational version never disappeared. It simply got better ratings.
What We Didn't Prove
All nine trials were conducted in the United States. Whether the finding generalizes to programs in Australia, the United Kingdom, Norway, or Germany, all of which operate variants, remains untested. The trials span 1967 to 1992; no randomized trial has been published since. Programs meaningfully redesigned after 2000 might produce different results, though the burden of proof should fall on advocates given the existing evidence base.
Individual trial sample sizes were modest (the largest enrolled 176 participants), which is precisely why systematic review matters: pooling nine independent experiments provides statistical power that no single trial achieved alone. The review could not disaggregate effects by race, socioeconomic status, or type of prior offense because the underlying trials rarely reported these breakdowns. The 57,600-offense estimate relies on assumptions about national enrollment that could be high or low by a factor of two; it is illustrative, not precise.
The Bottom Line
Scared Straight programs do not reduce juvenile crime. They increase it. Nine randomized controlled trials, pooled through the most rigorous evidence synthesis methodology available, consistently found that juveniles exposed to prison-visit programs were more likely to reoffend than those who received no intervention. The finding is statistically significant, robust across every analytical approach applied, and has not been overturned by any subsequent trial. These programs continue to operate because they satisfy an intuition about how deterrence should work. The data says the intuition is wrong.
What You Can Do
If you are a parent whose child has been referred to a Scared Straight program, you now have grounds to refuse. Cite the Cochrane review (DOI: 10.1002/14651858.CD002796.pub2) by name. If you work in juvenile justice, advocate for evidence-based alternatives: Functional Family Therapy and Multisystemic Therapy both show consistent reductions in juvenile reoffending across randomized trials. They cost more than a dollar-per-participant prison tour, but they produce results in the right direction. If you vote, pay attention to what your local jurisdiction funds. As of the review's most recent update, governments continued authorizing and funding these programs despite two decades of evidence documenting harm. Demand that any juvenile intervention carry a published evaluation before enrollment begins. The kids cannot advocate for themselves. The evidence can.