ExperimentSociety.org
Modern governments face a paradox. The problems they attempt to solve are increasingly complex, local, and uncertain. Yet public policy often remains centralized, slow-moving, and optimized around persuasion rather than learning.
The Experiment Society helps governments and communities adopt a different operating model: small experiments, transparent measurement, continuous learning, and local adaptation.
Core beliefs
Knowledge is local
People closest to problems often possess information unavailable to centralized planners.
Variation is valuable
Progress emerges from trying multiple approaches. Uniformity reduces learning.
Evidence should outrank ideology
Political disagreement is unavoidable. Empirical feedback should narrow uncertainty.
Small experiments outperform large irreversible mistakes
Change should begin small. Successful ideas should earn expansion.
Failure is information
Negative results should be published. What did not work is as valuable as what did.
From the Registry
158 documented experiments across 28 countries and 70+ jurisdictions. Positive, null, and negative results — published equally.
Social pressure arm: +8.1 percentage points vs. control; self-disclosure: +4.9 pp; Hawthorne: +2.5 pp; civic duty: +1.8 pp. Social pressure arm produced the largest GOTV effect ever recorded in a randomized study at that time.
SMarT participants: savings rate increased from 3.5% to 13.6% over 4 years; immediate advice group: from 4.4% to 8.8%; advice refusers: 6.6% (no change). SMarT had 78% retention rate vs. 26% for immediate increase group.
Quarterly earnings: +$1,000/quarter (+30%) at 3-year follow-up; annual earnings advantage: ~$4,500; employment rate: +11 pp; effect sustained through year 5 in extended follow-up
Youth homicides: −63% in 24 months (from 3.8 to 1.4 per 100,000); gun assault incidents: −25%; effect attributable to intervention not broader national trend through synthetic control analysis
Not partisan
We do not advocate for specific policies or outcomes.
Not coercive
No one is assigned to a pilot without awareness. Opt-out is always available.
Not extractive
Findings are published publicly under CC-BY. Cities own their results.
Not paternalistic
Communities define what outcomes matter. We measure; we do not prescribe.
Mission
ES is not a partisan policy organization. It does not tell communities what outcomes they must value. Its role is to help institutions ask better questions, design safer pilots, measure honestly, publish what was learned, and make local learning useful to others.
What Makes This Different
Traditional evidence-based policy finds the best intervention and recommends it. That model produces good papers. It rarely produces learning institutions. The Experiment Society is building the infrastructure for governments to run experiments themselves — continuously, locally, and without waiting for academic intermediaries.
| Question | External Research Model | Laboratories of Democracy |
|---|---|---|
| Goal | Find the best policy | Build learning institutions |
| Output | Publish studies | Run experiments |
| Scale | Centrally | Adapt locally |
| Answer | One answer | Continuous discovery |
| Who leads | External researchers | City practitioners |
| Timeline | Years to publication | 90-day pilots |
Programs
ES supports municipalities, schools, agencies, and civic institutions in designing, running, measuring, and sharing policy experiments.
Program 1
90-day experiments with public institutions. Define a question, run a pilot, measure outcomes, publish findings.
Components
Program 2
An open repository of every civic experiment — including null and negative results. Cumulative public knowledge.
Components
Program 3
Train local leaders in causal inference, experimentation, ethics, measurement, and decision science.
Components
Program 4
Shared infrastructure: experiment templates, dashboards, analysis tools, and implementation guides.
Components
Theory of Change
Condition
Institutions can experiment safely and transparently
Mechanism
They learn faster from evidence
Outcome
Better policy
Goal
Improved human flourishing
Core Beliefs
01
People closest to problems often possess information unavailable to centralized planners.
02
Progress emerges from trying multiple approaches. Uniformity reduces learning.
03
Political disagreement is unavoidable. Empirical feedback should narrow uncertainty.
04
Change should begin small. Successful ideas should earn expansion.
05
Negative results should be published. What did not work is as valuable as what did.
06
Governments should improve not only outcomes but their ability to improve outcomes.
Intellectual Lineage
The Experiment Society draws on a rich tradition of thinkers who recognized that decentralized, empirical, iterative approaches to governance outperform centralized certainty.
→Federalism
James Madison
Multiple centers of governance capable of trying different approaches. Variation creates comparison. Comparison creates learning.
→Local Institutions
Alexis de Tocqueville
Democratic life flourishes through local participation. Citizens learn judgment through engagement.
→Dispersed Knowledge
Friedrich Hayek
No central authority possesses sufficient information to optimize society globally. Progress requires decentralized experimentation.
→Piecemeal Social Engineering
Karl Popper
Social improvement through small interventions, observable outcomes, reversibility. Good institutions discover error early.
→Successive Approximation
Charles Lindblom
Policymakers cannot optimize globally. Progress happens through incremental adjustment, each step informed by the last.
→Experimental Society
Donald Campbell
Policy should be hypothesis-driven. Programs should be designed, measured, compared, and revised.
→Polycentric Governance
Elinor Ostrom
Multiple overlapping decision centers create resilience. Successful approaches spread voluntarily.
→Democracy as Inquiry
John Dewey
Democracy is collective problem solving. Communities identify problems, test solutions, revise beliefs, and improve.
From the Registry
Michigan, USA · 2006
Michigan Social Pressure Voting Mailer
The Michigan social pressure experiment produced the largest voter mobilization effect ever document…
Positive result
United States (multi-site) · 2012
Opower Home Energy Reports
Telling households how their energy use compared to similar neighbors—and providing a simple smiley …
Positive result
Rural Mexico · 1997
PROGRESA Conditional Cash Transfer
PROGRESA is the most influential anti-poverty experiment in modern policy history. Its village-level…
Positive result
Stockton, CA, USA · 2019
Stockton SEED — Guaranteed Income Pilot
The Stockton SEED experiment directly refuted the assumption that unconditional cash transfers reduc…
Positive result
Tennessee, USA · 1985
Tennessee STAR — Student/Teacher Achievement Ratio
The Tennessee STAR experiment is the most influential randomized trial in the history of American ed…
Positive result
Baltimore, Boston, Chicago, Los Angeles, New York · 1994
Moving to Opportunity Housing Vouchers
Moving to Opportunity is one of the most consequential policy RCTs ever conducted. The initial evalu…
Mixed result
Get Involved
We are looking for municipalities, civic institutions, and local leaders ready to run their first experiment. Phase 0 is open: one signed pilot, one public result.