experimentsociety.org
hello@experimentsociety.org
experimentsociety.org/get-involved
We help cities run rigorous experiments on their own programs — and build the evidence base that lets other cities learn from what they find.
Most government programs are implemented without any serious attempt to test whether they work. Evaluation — when it happens at all — is often done by advocates, after the fact, and designed to confirm success rather than discover truth. The result is systematic overestimation of what works — and billions spent on programs that don't.
We partner with city departments to design and run low-cost randomized pilots — handling study design, power analysis, randomization, and write-up while city staff implement. We also maintain a public registry of 116 civic experiments across 22 countries, so cities don't have to start from zero.
01
Identify a question
What program, process, or message are you uncertain about?
02
Design together
We handle power analysis, randomization design, and IRB prep.
03
Your team runs it
You implement. We support. Timeline: 8–14 weeks to launch.
04
Shared findings
Results go into the public registry under your city's name.
116
Experiments in registry
22+
Countries represented
Free
For pilots under 2,000 participants
8–14 wks
First call to randomization
Chicago, IL
A plain-English letter reduced library fine balances by 34%, converting inactive cardholders.
New York City
Streamlined rental assistance applications cut processing time by 40%, reducing eviction filings.
UK (national)
Pension auto-enrollment raised participation from 55% to 83% — the default was the intervention.
Philadelphia, PA
Right-to-counsel in eviction cases reduced removals by 77% among represented tenants.
Ready to start a conversation?
Email hello@experimentsociety.org or visit experimentsociety.org/get-involved
The Experiment Society · 2026