Evidence about what did not work can improve later decisions. Charitology’s working hypothesis in the Silence About Failure asks whether the system makes those lessons harder to share. This archive collects documented failures and mixed results, each written as a lesson for the system.
The entries examine published actions and results, without inferring anyone’s motives. Where an organization has responded publicly, its response is included.
Use this reference base
Filter by recorded result, place, depth or review status. Inspect the original case and its sources before drawing a lesson. Select several cases to assemble a learning dossier that keeps the context and uncertainty attached to each one.
Coverage and review
The archive reuses documented cases from the historical collection. It is not a complete record of failure, a causal model of every failure or a claim about anyone’s motives. Mixed results remain labeled as mixed.
These are published failures and mixed results, not a list of bad organizations. Result labels summarize a case, not everyone involved. Read the setting, evidence and response before transferring a lesson elsewhere.
0 selected. Choices stay in this browser tab, including after a reload. Use Clear list to remove them. If tab storage is blocked, choices last only for this page.
No entries match these filters. Try fewer words or reset the filters.
D.A.R.E.’s original curriculum
Police officers teaching children to resist drugs became the most widely used school drug prevention program in the United States. Rigorous evaluations of the original curriculum found little or no effect on drug use, and the program eventually replaced it.
- Recorded result
- What failed
- Place
- United States
Inspect the recorded fields
- Period
- 1983–2013
- Depth
- 2 · Prevention
- Lessons stated in the entry
-
- Popularity and evidence can run on separate tracks for decades. The program stayed dominant while rigorous studies found small or no effects.
- How a lesson is taught can matter as much as what it says. In a 1994 comparison, interactive programs that built social skills did better than D.A.R.E.’s officer-led lessons.
- An evaluation changes nothing until it reaches the people who decide what gets funded.
- An organization can change course on evidence and say so. D.A.R.E. now states on its own history page that the original curriculum had little long-term effect.
Read the source list (4)
- How effective is drug abuse resistance education? A meta-analysis of Project DARE outcome evaluations — Susan T. Ennett, Nancy S. Tobler, Christopher L. Ringwalt and Robert L. Flewelling, American Journal of Public Health 84(9), 1994
- Project DARE: no effects at 10-year follow-up — Donald R. Lynam and colleagues, Journal of Consulting and Clinical Psychology 67(4), 1999
- Project D.A.R.E. Outcome Effectiveness Revisited — Steven L. West and Keri K. O’Neal, American Journal of Public Health 94(6), 2004
- History — D.A.R.E. America
One Laptop per Child
A nonprofit started by MIT Media Lab faculty set out to give schoolchildren around the world low-cost laptops of their own, and governments bought hundreds of thousands. In Peru children used computers far more but did no better in math or language, and in Uruguay reading and math scores did not improve.
- Recorded result
- Mixed result
- Place
- Peru, Uruguay and other countries
Inspect the recorded fields
- Period
- 2005–2014
- Depth
- 3 · Structure
- Lessons stated in the entry
-
- Delivering a thing is not the same as delivering what the thing is for. The laptops reached children, and better test scores did not follow.
- Test the theory of change before scaling it. The big evaluations reported after governments had already bought laptops for whole school systems.
- What you measure shapes what you learn. Peru’s evaluation found no effect on test scores and inconclusive gains in general thinking skills, and observers disagreed about which outcomes mattered most.
- Randomizing during a national rollout is what made these lessons knowable. That choice deserves as much attention as the results.
Read the source list (4)
- Annan presents prototype $100 laptop at World Summit on Information Society — MIT News, 2005
- Technology and Child Development: Evidence from the One Laptop per Child Program — Julian Cristia, Pablo Ibarrarán, Santiago Cueto, Ana Santiago and Eugenio Severín, American Economic Journal, Applied Economics 9(3), 2017
- The Impact of a One Laptop per Child Program on Learning: Evidence from Uruguay — Gioia De Melo, Alina Machado and Alfonso Miranda, IZA Discussion Paper 8489, 2014
- Evaluating One Laptop Per Child (OLPC) in Peru — Michael Trucano, World Bank Blogs, 2012
PlayPumps
A merry-go-round that pumped water as children played drew $16.4 million in pledges and wide celebrity support. Field reports later found many pumps broken, unused or slower than the hand pumps they replaced, and the backers stopped the rollout and said so publicly.
- Recorded result
- What failed
- Place
- Southern Africa
Inspect the recorded fields
- Period
- 2005–2010
- Depth
- 2 · Prevention
- Lessons stated in the entry
-
- A story can outrun the evidence. The picture of children playing while water flowed raised millions and a 4,000-pump target before independent field assessments came in.
- Bad news traveled slowly. Field partners saw breakdowns early and a 2007 UNICEF assessment raised concerns, but the report stayed internal, and the first public acknowledgment came in 2009.
- When the people who will use a technology don’t choose it, mismatches follow. In some places a PlayPump replaced a hand pump that delivered more water per hour.
- Stopping a program and explaining why, in public, is how the field learns. It works best when it happens early.
Read the source list (3)
- Southern Africa: Troubled Water — Amy Costello, FRONTLINE/World, PBS, 2010
- Timeline: The PlayPump Trail
— FRONTLINE/World, PBS, 2010
Summarizes the 2007 UNICEF assessment, the 2008 report for Mozambique’s government, PlayPumps International’s 2009 letters and Jean Case’s 2010 letter.
- The PlayPump: What Went Wrong? — Daniel Stellar, State of the Planet, Columbia Climate School, 2010
Rebuilding Haiti after the 2010 earthquake
Billions of dollars poured into Haiti after the 2010 earthquake, almost all of the relief aid through international organizations. In 2015 ProPublica and NPR reported that the American Red Cross had raised nearly half a billion dollars and built six permanent homes. The Red Cross said it had helped millions of Haitians in other ways.
- Recorded result
- What failed
- Place
- Haiti
Inspect the recorded fields
- Period
- 2010–2015
- Depth
- 1 · Repair
- Lessons stated in the entry
-
- Giving that follows the news arrives in a burst, mostly without research, and attention fades long before rebuilding is done.
- Money raised for one disaster is, in practice, tied to that place and to the group that raised it. When land titles, partners and government capacity take years to sort out, a large earmarked sum can outrun the ability to spend it well.
- Aid that goes around local institutions leaves them no stronger. Almost all relief aid went to international agencies and organizations rather than through Haitian ones.
- Counting people helped is not the same as measuring what changed. Without project-level spending and results, big claims can’t be checked by anyone.
Read the source list (5)
- How the Red Cross Raised Half a Billion Dollars for Haiti and Built Six Homes — Justin Elliott (ProPublica) and Laura Sullivan (NPR), 2015
- American Red Cross Responds to Latest ProPublica and NPR Coverage — American Red Cross, 2015
- Give aid directly to Haitian institutions to build back better, says UN report — UN News, on a report by the Office of the UN Special Envoy for Haiti, 2011
- Haiti: Where Has All the Money Gone? — Vijaya Ramachandran and Julie Walz, Center for Global Development, 2012
- Real Time Charitable Giving — Aaron Smith, Pew Research Center, 2012
Scared Straight
Programs that brought at-risk teenagers into prisons to be confronted by inmates spread across the United States on the strength of a documentary. Randomized trials found they increased later offending, yet versions of the programs carried on for decades.
- Recorded result
- What failed
- Place
- United States and other countries
Inspect the recorded fields
- Period
- 1970s–2010s
- Depth
- 2 · Prevention
- Lessons stated in the entry
-
- Success stories without a comparison group can’t show that a program works. The original documentary reported that 16 of 17 teenagers stayed out of trouble for three months, and compared them with no one.
- A program can feel right, cost almost nothing and still cause harm. Only randomized comparisons revealed it.
- Evidence of harm does not stop a program by itself. In one case, negative results ended the evaluation instead of the program.
- The more dramatic, cheap and intuitive a program is, the faster it spreads, so those are the programs most in need of testing.
Read the source list (2)
- Scared Straight and Other Juvenile Awareness Programs for Preventing Juvenile Delinquency, a systematic review
— Anthony Petrosino, Carolyn Turpin-Petrosino, Meghan E. Hollis-Peel and Julia G. Lavenberg, Campbell Systematic Reviews 2013:5, 2013
A free copy is held by the University of Tübingen repository at https://publikationen.uni-tuebingen.de/xmlui/handle/10900/64695
- Justice Department Discourages the Use of “Scared Straight” Programs — Office of Juvenile Justice and Delinquency Prevention, OJJDP News @ a Glance, 2011
The ALS Ice Bucket Challenge
In the summer of 2014 a viral video challenge raised $115 million for the ALS Association, which turned the money into research grants, clinics and advocacy. It is a case of attention-driven giving that funded lasting work, and a reminder that attention picks causes by what spreads, not by need.
- Recorded result
- Mixed result
- Place
- United States and online worldwide
Inspect the recorded fields
- Period
- 2014–2024
- Depth
- 1 · Repair
- Lessons stated in the entry
-
- Attention can raise a fortune in a single summer, but it chooses causes by what spreads, not by the size or neglect of the need.
- A windfall does the most good when it is turned into things that keep working, such as research teams, clinics and trained people.
- Self-reported impact needs independent checks. The 10-year evaluation leaned on a grantee survey that only about a quarter of grantees answered.
- A campaign total can’t say whether the money was new giving or giving moved from other causes. That counterfactual is the question that matters most and is the hardest to answer.
Read the source list (2)
- The ALS Ice Bucket Challenge — The ALS Association
- ALS Association Ice Bucket Challenge Impact Evaluation Report — Joshua Schnell and colleagues, RTI International, 2024
The campaign to eradicate polio
Since 1988 a global partnership has cut wild polio cases by more than 99 percent and eradicated two of the three wild strains. The original deadline was the year 2000, and the last stretch has proved the hardest.
- Recorded result
- Mixed result
- Place
- Worldwide
Inspect the recorded fields
- Period
- 1985–present
- Depth
- 4 · Source
- Lessons stated in the entry
-
- A precise goal with a shared scoreboard, zero cases, has held governments, agencies and a volunteer movement together for decades.
- A civic network can be funder, fundraiser and advocate at once. Rotary helped start the effort, and its members have contributed more than $2.7 billion.
- The last 1 percent can be harder than the first 99. A deadline set at the start rarely survives the hardest places.
- Program data can look better than reality. Independent audits are how a campaign finds out what its own numbers miss.
Read the source list (5)
- Global eradication of poliomyelitis by the year 2000 (resolution WHA41.28) — World Health Assembly, 1988
- Poliomyelitis, fact sheet — World Health Organization, 2025
- History of Polio: Key Milestones & Global Eradication — Global Polio Eradication Initiative
- Partners — Global Polio Eradication Initiative
- Update on the ongoing public health emergencies of international concern: poliomyelitis (EB/SCHEPPR/8/2) — World Health Organization, report by the Director-General, 2026
The Green Revolution
Foundation-funded crop research gave farmers high-yield wheat and rice, and cereal harvests in the developing world roughly tripled. The gains were uneven, and some of the costs, from falling water tables to narrower diets, showed up decades later.
- Recorded result
- Mixed result
- Place
- Mexico, South and Southeast Asia, and other developing regions
Inspect the recorded fields
- Period
- 1943–2000
- Depth
- 3 · Structure
- Lessons stated in the entry
-
- Research that every farmer can use is a public good. Funding the institutions that keep producing it can change a whole food system, not one farm at a time.
- Measuring success in one unit, tons of grain, hid costs that were counted in other units, such as water, soil, diets and who was left out.
- Many of the harms grew from the rules around the technology, such as subsidies that encouraged overuse of inputs. The same tool does more or less damage depending on the policies it arrives with.
- Costs that arrive decades later are easy to leave out of a success story. Long-horizon work needs long-horizon accounting.
Read the source list (4)
- Green Revolution: Impacts, limits, and the path ahead — Prabhu L. Pingali, Proceedings of the National Academy of Sciences 109(31), 2012
- Photo Essay: Mexico and the Launch of the Green Revolution — Liesel Vink, Rockefeller Archive Center, 2019
- Norman Borlaug, Facts — NobelPrize.org
- India groundwater
— University of California, Irvine, on Rodell, Velicogna and Famiglietti in Nature 460, 2009
Reports a satellite study finding that northern India lost more than 109 cubic kilometers of groundwater between 2002 and 2008, mainly to irrigation pumping.