Press & Media
The infrastructure for government that learns.
The Experiment Society is a nonprofit building the tools, evidence base, and institutional relationships that make randomized experimentation a routine part of local government. We are available for interviews, background conversations, and expert referrals.
116
Experiments in the registry
22+
Countries represented
16
Policy areas covered
11
Dispatch issues published
Who we are
Making the experimenting society real.
In 1969, Donald Campbell argued that governments should treat every major reform as an experiment — designed from the beginning to generate reliable knowledge about whether it works. More than fifty years later, the institutions he imagined are still largely hypothetical.
The Experiment Society exists to close that gap. We partner directly with city departments to run low-cost randomized pilots, maintain a public registry of what has been tested and what was found, and publish rigorous accessible writing about evidence from government experiments.
Our standard: if you can't replicate it, we won't publish it. If the result was null, that's required — not optional.
Story angles
Evidence worth writing about.
The registry documents experiments across 16 policy areas and 22 countries. A few patterns that recur and are worth covering:
The Oregon Health Insurance Study
When Oregon couldn't afford to expand Medicaid to everyone on the waiting list, they held a lottery — accidentally creating the largest randomized trial of health insurance in US history. The results were inconvenient for everyone.
Friction is policy.
Across 12 countries, one finding recurs: reducing the steps required to comply with a government process increases compliance — sometimes dramatically. Administrative burden is usually not a policy decision. It's a default.
The null result problem.
60% of randomized policy trials show no significant effect. In published literature, that share drops to 30%. The gap isn't random — and it produces systematic overconfidence in what works. The registry requires null results.
Laboratories of democracy — what Brandeis actually meant.
The phrase is invoked constantly by both parties. The idea is more precise — and more interesting — than either camp acknowledges. Brandeis was describing a specific epistemic claim about how knowledge accumulates in a federal system.
Background
For reporters on deadline.
What is The Experiment Society?
A nonprofit organization dedicated to making randomized experimentation an ordinary part of local government decision-making. We build and maintain a public registry of civic experiments, partner with city departments to design and run low-cost randomized pilots, and publish the Dispatch — a newsletter covering evidence from government experiments worldwide.
Who runs it?
The Experiment Society is an independent nonprofit. Our Technical Advisory Network includes statisticians, economists, and policy evaluators who provide pro bono study design support to city partners.
What is the Public Registry?
A curated database of 116 randomized and quasi-experimental studies conducted by or for government agencies. Each entry includes the intervention, assignment method, sample size, effect estimate, and a link to the primary source. Null and negative results are required — not optional.
How does a city partnership work?
We handle study design, power analysis, randomization, and write-up. City departments implement the intervention. Small pilots (under 2,000 participants) are typically free. The typical timeline from first call to randomization is 8–14 weeks.
Get in touch
We respond to every press inquiry.
Available for interviews, background conversations, data requests, and expert referrals to researchers in the registry. Typical response within one business day.
Email the press team →