https://www.experimentsociety.org/dispatch Quarterly notes on civic experimentation — what works, what fails, and why it matters. en-us hello@experimentsociety.org (The Experiment Society) Thu, 09 Jul 2026 19:20:43 GMT https://www.experimentsociety.org/dispatch/the-commitment-device https://www.experimentsociety.org/dispatch/the-commitment-device In 2004, researchers at Green Bank of Caraga in the Philippines offered customers a savings account with an unusual feature: once you deposited money, you couldn't take it out until you reached your goal. Twenty-eight percent of people offered the account opened one. Savings balances increased by 81% for those who did. The finding raised a question that is still working its way through public policy: if people will voluntarily restrict their own future behavior, what does that mean for how we design the systems they interact with? Fri, 19 Jun 2026 00:00:00 GMT Behavioral economics, Savings, Self-control https://www.experimentsociety.org/dispatch/what-summer-does-to-reading https://www.experimentsociety.org/dispatch/what-summer-does-to-reading Every fall, something predictable happens in American schools. Students from higher-income families return from summer having held steady or improved. Students from lower-income families return having lost ground — on average, two to two and a half months of reading achievement. The gap is not about intelligence or teachers or schools. It is about access to books. A randomized trial across three cities tested whether giving children twelve books of their choosing could prevent that loss. It could. The effect was as large as summer school. The cost was about fifty dollars. Sun, 14 Jun 2026 00:00:00 GMT Education, Libraries, Inequality https://www.experimentsociety.org/dispatch/the-land-question https://www.experimentsociety.org/dispatch/the-land-question One of the central claims of development economics is that insecure property rights inhibit investment. If you don't know whether the land will still be yours next year, you won't plant trees or build irrigation or take the long view. The solution, in theory, is formalization — giving people legal title to what they already possess. Rwanda ran the largest systematic land registration program in Sub-Saharan Africa and measured what happened. The results confirmed the theory in some ways and complicated it in others. Tue, 09 Jun 2026 00:00:00 GMT Property rights, International development, Land tenure https://www.experimentsociety.org/dispatch/what-stockton-showed https://www.experimentsociety.org/dispatch/what-stockton-showed In February 2019, the city of Stockton, California began sending $500 a month to 125 randomly selected residents. No conditions. No work requirements. No documentation of how the money was spent. The pilot, called SEED, ran for 24 months. It was designed from the beginning as a randomized controlled trial, with a matched control group and pre-registered outcomes. What it found was more specific — and more useful — than the debate that surrounded it. Thu, 04 Jun 2026 00:00:00 GMT Guaranteed income, Economic policy, Field experiments https://www.experimentsociety.org/dispatch/the-oregon-health-lottery https://www.experimentsociety.org/dispatch/the-oregon-health-lottery In early 2008, the state of Oregon did something unusual. It wanted to expand its Medicaid program to cover more low-income adults — but it only had enough funding for about 10,000 new enrollees, and more than 90,000 people had signed up for the waiting list. Rather than prioritize by income, geography, or need, Oregon held a lottery. Names were drawn. Winners got insurance. Losers didn't. The study that followed became one of the most cited, most debated, and most misrepresented pieces of health research in recent decades. Sat, 23 May 2026 00:00:00 GMT Healthcare, Medicaid, Natural experiments https://www.experimentsociety.org/dispatch/campbell-and-the-experimenting-society https://www.experimentsociety.org/dispatch/campbell-and-the-experimenting-society In 1969, a psychologist named Donald T. Campbell published a paper titled 'Reforms as Experiments.' It was not a study of any particular policy. It was an argument about what governments should become. Mon, 18 May 2026 00:00:00 GMT Donald Campbell, Evaluation theory, Intellectual history https://www.experimentsociety.org/dispatch/the-nudge-unit https://www.experimentsociety.org/dispatch/the-nudge-unit In 2010, the UK Cabinet Office created a small team with an improbable mandate: apply behavioral economics to government, and prove the results through randomized controlled trials. What happened over the next decade reshaped how governments around the world think about experimentation. Wed, 13 May 2026 00:00:00 GMT Behavioural Insights Team, Behavioral science, Government innovation https://www.experimentsociety.org/dispatch/ostrom-and-the-commons https://www.experimentsociety.org/dispatch/ostrom-and-the-commons In 1968, Garrett Hardin argued that shared resources would inevitably be destroyed by self-interest. The person who showed why he was wrong became the first woman to win the Nobel Prize in Economics. Fri, 08 May 2026 00:00:00 GMT Elinor Ostrom, Polycentric governance, Institutional theory https://www.experimentsociety.org/dispatch/why-job-training-fails https://www.experimentsociety.org/dispatch/why-job-training-fails In 1994, the largest randomized evaluation of job training ever conducted found no significant effect on earnings. Twenty years later, a different study found a 30% increase. Both were rigorous. Both measured job training. The difference is what the training is connected to. Sun, 03 May 2026 00:00:00 GMT Job Training, Human capital, Labor markets https://www.experimentsociety.org/dispatch/the-mailer-that-changed-political-science https://www.experimentsociety.org/dispatch/the-mailer-that-changed-political-science In 2006, researchers mailed 344,000 Michigan voters a simple letter showing their voting history alongside their neighbors'. Turnout increased 8.1 percentage points — the largest effect ever documented in a randomized experiment. Then the backlash started. Tue, 28 Apr 2026 00:00:00 GMT Voter Engagement, Social norms, Behavioral science https://www.experimentsociety.org/dispatch/federalism-as-experiment https://www.experimentsociety.org/dispatch/federalism-as-experiment In 1932, Justice Brandeis wrote that a state may 'serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.' The idea is more precise — and more interesting — than either political camp tends to acknowledge. Thu, 23 Apr 2026 00:00:00 GMT Federalism, Constitutional law, Intellectual history https://www.experimentsociety.org/dispatch/friction-is-policy https://www.experimentsociety.org/dispatch/friction-is-policy Across twelve countries and four policy domains, one finding recurs: when you reduce the steps required to comply with a government process, compliance goes up — sometimes dramatically. Administrative burden is usually not a policy decision. It is a policy default. Sat, 18 Apr 2026 00:00:00 GMT Simplification, Tax & Revenue, Benefits Enrollment https://www.experimentsociety.org/dispatch/what-hot-spots-policing-actually-showed https://www.experimentsociety.org/dispatch/what-hot-spots-policing-actually-showed The Kansas City Gun Experiment is widely cited as the origin of hot spots policing. The finding has been replicated more than 50 times. It is also routinely misapplied — because the effective component of an intervention is usually narrower than the label on the box. Mon, 13 Apr 2026 00:00:00 GMT Public Safety, Targeting, Replication https://www.experimentsociety.org/dispatch/the-null-result-problem https://www.experimentsociety.org/dispatch/the-null-result-problem Roughly 60% of randomized controlled trials in public policy show no statistically significant effect. In published literature, that share drops to about 30%. The gap is not random — and it produces systematic overconfidence in what works. Wed, 08 Apr 2026 00:00:00 GMT Evidence quality, Publication bias, Registry https://www.experimentsociety.org/dispatch/how-conditional-cash-transfers-work https://www.experimentsociety.org/dispatch/how-conditional-cash-transfers-work Mexico's Progresa was one of the first large-scale conditional cash transfer programs. More than 200 papers have been published on its effects. What does three decades of evidence actually show? The conditions are probably not the mechanism. Fri, 03 Apr 2026 00:00:00 GMT Cash Transfers, International Development, Education