Sethu Nguna.
AidData · Presidential Precinct · 2024–2026

Data Command

Delivered to three cohorts

A decision simulation for emerging leaders, run with three fellowship cohorts and rebuilt each time. The 2026 version puts a newly elected government in front of two readiness reports that disagree, 72 hours before a donor board decides the future of a health system serving 3.2 million people.

← Explore all projects
Learners
Mandela Washington Fellows (2024, 2025) and Global Democracy Fellows (2026)
Setting
In person, within the Presidential Precinct’s fellowship programmes
Format (2026)
Facilitated; three hours; four institutional teams; two rounds and a mid-session revision
Research base
AidData’s From Pork to Performance, on open government and service delivery in the Philippines
Global Democracy Fellows working through the Data Command simulation
Global Democracy Fellows in the Data Command session, 2026. Photo: Presidential Precinct, via AidData.
The design problem

The challenge

AidData’s study of open government in the Philippines found that publishing data doesn’t by itself improve services or accountability. That takes coordination between agencies, timely reporting, frontline feedback and institutions able to respond. Leaders need practice governing when the evidence is contested, not just reading it.

The logic

Give each team a different institutional mandate and the same contested evidence. Set a deadline that makes delay costly. Never say which report is right. The simulation doesn’t reward finding the answer. It rewards a position the group can defend in public.

What teams produce

One joint public statement, reached from four separate institutional positions, that weighs legal commitments, service continuity, fiscal limits, public accountability and donor relations.

My role

I designed and facilitated all three iterations. For 2026 I wrote the scenario, team briefs and evidence, built the facilitator deck and participant instruments, and produced the intro film and the mid-session inject video.

Three iterations

2024 · Mandela Washington Fellows

Fellows took roles as government officials, community leaders and NGO representatives in Firania, a fictional nation under compound pressure. Teams analysed development data, negotiated across competing interests and allocated resources against the clock, then adapted when injected crises, a natural disaster and a corruption scandal, overturned their plans.

2025 · Mandela Washington Fellows

I reworked the roles around the incoming cohort’s professional profiles, so that government, civil society, media and social innovation fellows each played close to their own work. The scenario centred on Firania’s bid for a $1 billion regional development grant.

Mandela Washington Fellows working through Data Command around a table with briefing papers and laptops
Mandela Washington Fellows in a Data Command session.

2026 · Global Democracy Fellows

Rebuilt around one contested decision and grounded in AidData research. The Republic of Aruvia, four years into its first democratically elected government, must take over a donor-run health system that covers about 40% of the country’s primary care. The evidence on whether it’s ready doesn’t agree.

How the 2026 session runs

Position, revise, negotiate, commit.

Three hours, four teams, and one decision the group can’t postpone.

01

Brief

An intro film sets out Aruvia’s situation. Teams receive their briefs as the Office of the Permanent Secretary, the External Financing Secretariat, the Office of the Transition Counsel and the Directorate of Service Delivery.

02

Round 1 · Position

Each team builds its institutional position, choosing which evidence to bring from a limited set of evidence cards.

03

Revision notice

An inject video, then a read-aloud notice: the official readiness figure is corrected from 76% to 61%, and new financing conditions split the teams.

04

Round 2 · Negotiate

Teams negotiate across their mandates toward a single position, with the donor board’s vote still fixed.

05

Commit

The group issues one joint public statement it can defend, with every institution’s name on it.

Design decisions

Every mechanic has a job.

The 2026 design rehearses a way of deciding that holds up under stress. Each feature makes participants practise one part of it.

DecisionWhat it trainsIn the simulation
Two reports that measure different things Asking what a figure actually measures before deciding whether to trust it. An official readiness report says 76%. A leaked internal assessment says 41%. One tracks progress against the ministry’s own targets, the other whether the government can keep services running after handover.
A deadline that can’t move Recognising that delay has its own cost. The donor board votes in 72 hours: proceed, add conditions, delay or withdraw. The handover date was fixed by an agreement the previous government signed.
Four institutions, one problem Negotiating with colleagues who read the same evidence differently. Each team holds a different mandate: the Permanent Secretary, the Financing Secretariat, the Transition Counsel and Service Delivery.
Evidence that moves mid-session Updating a position honestly when the numbers change. The revision notice corrects the official figure from 76% to 61% and adds financing conditions that pull the teams apart.
No right answer Judgment instead of answer-hunting. Nothing in the materials says which report to trust or which policy to adopt.
Build the research into the scenario Seeing why open data alone doesn’t produce accountability. The research finding that donors fund outputs while leaving gaps in workforce, financial controls, supply chains and accountability shapes the roles, evidence and constraints.
A public commitment as the output Making commitments that hold up under scrutiny. The session ends in one joint statement, not a private recommendation.
Evidence

What’s documented

Data Command ran with two Mandela Washington Fellows cohorts and one Global Democracy Fellows cohort. AidData published an account of the 2026 session, in which the Fellows moved from separate institutional positions to one public statement they could defend.

Limits

What the evidence can’t show

No outcome data has been published for these sessions, so the evidence is of delivery and of what participants produced. The design aims for Fellows to keep using these techniques without a facilitator. Whether they do isn’t something the current evidence can show.

Supporting artifacts

Explore the work.

Start a conversation

Build the capability.
Make it work in practice.

Learning strategy, institutional systems and decision practice.

sethnguna@gmail.com ↗