Action Labs · The Governance Collective

India's governance system learns.
It just doesn't remember.

We embed inside government at every level, surface solutions that work, ensure the learning doesn't die with the posting, and open that knowledge so any office can integrate it fast — until the system no longer needs us to.

● Active — Bihar Scouting — 7 states State mandate required
How Action Labs work The Governance Collective
2–3
Year average DC tenure — the window in which institutional knowledge is built and then lost
700+
Districts across India, each solving problems their neighbours have already solved
Zero
Organisations systematically converting what government learns at any level into open, validated, replicable knowledge — until now

The Problem We Are Solving

India does not have a governance knowledge problem.
It has a governance memory problem.

Every year, across India's districts, blocks, departments, and secretariats, capable officers figure out how to make things work. How to make a nutrition protocol reach the last mile. How to include communities that the scheme was never designed to reach. How to build accountability into a system that had none.

And then the posting order comes.

The officer transfers — every two to three years on average. The knowledge, the relationships, the operational intelligence they built leaves with them. The next person starts from scratch. The problem that was being solved begins to unsolvable itself again.

This is not limited to the district. It happens at every level — in the secretariat, in the department head's office, in the state review room. The system learns constantly. It almost never remembers.

2–3 yrs
Average District Collector tenure — the window in which most district-level institutional knowledge is built, and then lost to the next posting order
700+
Districts across India, each largely operating in isolation — solving problems their neighbours have already solved, making mistakes that have been made before
₹0
Currently invested in governance memory infrastructure across India's 28 states — no organisation systematically converts what government learns into something the system knows permanently

How We Work

Two interlocking pillars.
One goal: a government that learns from itself.

Action Labs are where the problem-solving happens — embedded, mandate-driven, ground-level. The Governance Collective is where the learning stays — curated, validated, openly accessible to any office facing the same problem. Neither works without the other.

Micro — The field work

Action Labs

Embedded. Mandate-driven. Problem-specific. Designed to become unnecessary.

  • State mandate required before any engagement begins — named problem, named outcome, named accountable officer
  • Chronic, specific, failure-mode problems — not general governance improvement
  • Five-stage discipline: Diagnose → Design → Embed → Measure → Replicate
  • Works at any level — district, block, department, secretariat — wherever the problem sits
  • 90-day decision points: scale, adapt, or pause. Government holds all the cards.

Macro — The knowledge commons

Governance Collective

Curated. Practitioner-led. Open to any office in India.

  • Curated playbooks from Action Labs — nine-section, evidence-labelled, peer-reviewed
  • Curated external solutions from NGOs, research, social enterprises — vetted, not a directory
  • Requests board — officers post real problems, get real responses from practitioners
  • Practitioner directory — searchable by sector, geography, methodology, evidence quality
  • Failures documented as rigorously as successes. Editorial standard enforced.
The loop: Action Labs generate evidence → evidence enters the Collective → the Collective enables the next office to start from where the last one finished → the system builds capacity → eventually, it does this on its own. That is the design. We commit to staying until the loop runs without us — eight to ten years per state.

Pillar One — Action Labs

The five-stage discipline.
No stage is skipped. No design before diagnosis.

Entry condition: We do not enter without a state-level mandate — a named problem, a named outcome target, an accountable officer, and senior leadership prepared to review evidence and make decisions at 90-day intervals. The first conversation is always with the CS office or equivalent.
01
Diagnose
Understand before designing
  • Positive Deviance scan
  • 5-Why root cause analysis
  • District data audit
  • Ethnographic field work
  • COM-B behaviour diagnosis
02
Design
Minimum viable intervention
  • Design space crawl — find before creating
  • Actor mapping
  • Two interventions: system + behaviour
  • Four stress-tests
  • 90-day pilot scope agreed with DC
03
Embed
Fellow as problem-owner
  • Named outcome target from Day 1
  • Practitioner mentor + decision toolkit
  • Weekly cadence — non-negotiable
  • Accountability chain to CS office
  • 80% of time on mandate — protected
04
Measure
Learn, not just report
  • Contribution tracing, not attribution
  • Max 6 dashboard metrics
  • Pre-commitment learning log
  • Searchframe, not logframe
  • Cross-lab learning (NIC model)
05
Replicate
Knowledge stays in the system
  • Nine-section playbook at 90 days
  • Peer diffusion — Rogers Early Adopter logic
  • ATI module development
  • Governance Collective curation
  • Designed for irrelevance

Pillar Two — The Governance Collective

What works in one office should be available to every office.

India's first curated, practitioner-led open knowledge platform built for government — not a database, not a research library, but a living platform with a specific editorial standard and a specific design principle: every entry must enable someone who was not part of the original work to implement the approach faster and with less trial-and-error.

📋
Playbooks
Action Lab documentation
Nine-section, evidence-labelled playbooks from Action Labs and practitioner contributions. Evidence labels: Pilot (tested once) → Replicated (tested across contexts) → Systemic (embedded in the system). Every entry peer-reviewed by someone who was not part of the original work. Failures documented as rigorously as successes.
✦
Curated Solutions
External vetted approaches
Approaches from NGOs, social enterprises, and research institutions that government can integrate — not a directory of everyone who claims to work on a problem, but a vetted catalogue with a minimum evidence bar and contextual fit assessment. Every entry specifies the conditions under which it works and the conditions under which it doesn't.
◈
Requests Board
Real problems, real responses
Officers, fellows, and organisations post specific governance problems and connect with practitioners who have relevant experience. Not generic networking — problem-driven matching. A CDPO who needs a protocol for unannounced monitoring. A DC who needs a convergence approach adapted for a mining district. Real requests, real responses.
⌘
People & Organisations
The practitioner network
Searchable directory of practitioners, fellows, and organisations by sector, geography, methodology, and evidence quality. For district-to-practitioner matching, peer connection, and building the relationships that are the actual connective tissue of governance improvement across India.
Editorial standard: The Collective's value is entirely dependent on the honesty of what it contains. A platform that records only successes is not a knowledge platform — it is marketing. The most valuable entries will be the ones that say clearly: we tried this, it didn't work as designed, here is what we changed and why, and here is what a future district should check before trying it themselves. We enforce this through practitioner peer review and evidence labelling that distinguishes between what has been tested once, tested across contexts, and embedded in how a system operates.

Where We Work

Starting in Bihar. Building toward a national commons.

Active Engagement — Bihar
  • CM Fellowship Scheme (CMFS) — cabinet-approved September 2025
  • 121 experienced professionals deployed to district collectorates
  • IIM Bodh Gaya as academic partner
  • CS Pratyaya Amrit mandate — full state, not a pilot
  • 38+ districts. Day 1 full-state engagement.
Scouting — 7 states
Jharkhand Odisha Chhattisgarh Karnataka Telangana Tamil Nadu Himachal Pradesh

Active engagement scouting where state mandate conditions exist and the governance reform conversation is live.

Year 3 Targets
3
States operational
100
Action Labs completed
80+
Collective entries
25
Team members

The long-term ambition

A national Governance Collective that any officer, at any level, in any state can access — not because Governance Initiative built it, but because practitioners across India contributed to it, states invested in it, and it is genuinely more useful than starting from scratch. The Collective outlasts us. That is the point.

Who We Are

We have been inside this system.
Not studying it from outside — inside it.

A
Ankush Singh
Co-Founder

Years inside Bihar's hardest administrative contexts — PMRDF in LWE-affected districts, tribal school construction, I-PAC Core Team across 41 UP constituencies, Reliance Foundation, Vayam. Has sat in anganwadis at 10am on a Tuesday and seen what the register says versus what is actually there. Governance Initiative is built from that specific knowledge — not from a study of governance failure, but from being inside it long enough to see what the fix would need to look like.

Chevening Scholar University of Sussex · IDS TISS · ISB DLabs PMRDF Bihar I-PAC · Reliance Foundation
M
Mayank Lodha
Co-Founder

At the institutional level — building the partnerships between government, philanthropy, and the social sector that make district-level work sustainable and credible. PMRDF, Teach For India, state government institutional partnerships, the philanthropy ecosystem. The Governance Collective is the infrastructure he has been trying to find for years. The practitioner knowledge in India's governance ecosystem is extraordinary. What is missing is the architecture that makes it openly available.

Acumen Fellow FMS Delhi PMRDF Teach For India Philanthropy & ecosystem

Work With Us

There are three ways to be part of what we are building.

For Government

We enter where there is a mandate — not a letter of support, but a named problem, a named outcome target, an accountable person, and senior leadership prepared to review evidence and make decisions at 90-day intervals. If you are a Chief Secretary's office, a Principal Secretary, or a Commissioner who is serious about a chronic governance failure your system has not solved — we want to talk.

Start a conversation
For Funders

We are building the infrastructure that no one else is building — because it does not produce a beneficiary count, it produces a government that is better at solving its own problems. That requires patient, honest capital. We are raising our first round. We are not asking for programme funding — we are asking for the capital to establish Action Labs at state scale and build the Governance Collective's first fifty entries.

Request a briefing
For Practitioners

If you are a PMRDF or CMFS fellow, a district officer, a block-level practitioner, an NGO with evidence of what works, or a researcher whose work sits in journals that no DC will ever read — the Governance Collective is built for you. If you have field experience worth documenting, a solution worth validating, or a request worth posting — we want you contributing, reviewing, and using what is being built.

Join the Collective