---
title: Failure Lab
group: Research
order: 8
summary: Build a source-linked postmortem, test possible failure mechanisms and track a corrective action and its follow-up.
status: live
statusNote: Live case browsing and postmortem software are available. Archive entries remain drafts; cross-case findings and independent field investigations still need researchers and evidence.
conditions: [silence-about-failure]
firstMilestone: Fifty documented cases in the Failure Archive, each with public sources, and a first published analysis of the harm mechanisms they share.
help:
  - Add a documented failure to the Failure Archive by pull request, with public sources and without blaming anyone.
  - Publish a failure from your own organization and link it in an issue labeled program.
  - Help sort archived cases by harm mechanism for Open Problem 20.
related: [/record/failure-archive/, /open-problems/when-charity-harms/, /open-problems/the-failure-commons/, /programs/the-charitology-prize/]
sources:
  - title: "‘Scared Straight’ and other juvenile awareness programs for preventing juvenile delinquency"
    publisher: Anthony Petrosino, Carolyn Turpin-Petrosino, Meghan E. Hollis-Peel and Julia G. Lavenberg, Cochrane Database of Systematic Reviews
    year: 2013
    url: https://pmc.ncbi.nlm.nih.gov/articles/PMC11973542/
  - title: 'Failure Report 2010: Learning From Our Mistakes'
    publisher: Engineers Without Borders Canada, via ALNAP
    year: 2011
    url: https://alnap.org/help-library/resources/failure-report-2010-learning-from-our-mistakes/
---

## Use the live research desk

Choose a documented case as a starting point or open a blank [postmortem](#workspace-failure-lab). Separate intentions, observations and explanations. Record the proposed mechanism, its evidence label, a corrective action and how you will check it. Study the system and published actions, without judging people.

Your workspace stays in this browser. Save records, edit them or fork frozen versions, export JSON, a research memo or CSV, and import a backup with an explicit replacement step. Deleting or clearing offers an undo while the page remains open. No notes are submitted to Charitology. The institutional goals below remain the program’s longer-term charter.

## What it is

Every field learns from what went wrong, if it gets to see it. The Failure Lab studies programs that didn’t work and programs that did harm, and it publishes what they teach in the [Failure Archive](/record/failure-archive/).

## How it works

Each case in the archive should be documented from public sources: what was tried, what happened, how anyone found out and what changed afterward. The lab would then look across cases for mechanisms that repeat, such as aid that undercuts local markets, programs that keep running after the evidence turns against them and ideas that worked in one place and failed in another. Mapping those mechanisms is [Open Problem 20, When Charity Harms](/open-problems/when-charity-harms/).

Some lessons are stark. Programs that take young people on organized prison visits to deter them from crime, known as “Scared Straight,” were tested in nine randomized trials in the United States between 1967 and 1992. A Cochrane review of those trials concluded that the programs increase offending compared with doing nothing, and noted that such programs remained in use.

## Why it matters for the system

The [Silence About Failure](/diagnosis/silence-about-failure/) costs the field its most valuable data. When a failure goes unpublished, the next group to try the same idea starts from zero. The lab treats that condition by turning failure into a shared resource instead of a private embarrassment, and by showing that a well-documented failure is a contribution. Some organizations have already shown the way: Engineers Without Borders Canada has published yearly reports of its own failures.

## Risks and limits

Naming failures can punish the very honesty the lab wants to reward. So the archive draws on public sources or on cases organizations choose to share, and it describes mechanisms rather than blaming people. An organization that publishes its own failure is credited as a contributor. No case is published with personal details about the people a program served.

## What success looks like

Within five years, the archive holds enough cases to show which harm mechanisms recur and under what conditions, and organizations consult it when they design new programs.