From a public problem to a decision that can be demonstrated
GOVERNA reads the problem, gathers the evidence, separates the causes, makes the hypotheses compete — and ends in a possible decision, with its provenance in plain sight.
Public problems don't need more diagnoses: they need to learn what works, and to leave memory of what was learned.
Rural school dropout
What do we do now?
Evidence doesn't dictate action: it frames it. With the same reading, three paths are legitimate — and GOVERNA says which one each state of the evidence opens, instead of recommending blind.
Demonstration environment over a real MEN figure. The recommendation is withheld while the evidence that justifies it does not exist: to recommend without it would be the error.
What the system lets you do
Compete hypotheses
A problem can have more than one explanation: the system makes them compete with for-and-against, not settle on the first.
Tell level from dynamic
The number today and its movement are two different questions. Reacting to one, understanding the other.
Separate fact from inference
Every claim carries its source and provenance. A correlation isn't promoted to a cause.
Derive the design from evidence
Not a recipe: an intervention whose evidence can be traced, or a decision withheld until it exists.
Claim impact only with a counterfactual
«It worked» isn't evidence. Impact is proven against a valid counterfactual before scaling.
From a problem to a decision, without skipping a step
- Data
- Context
- Analysis
- Scenarios
- Decision
The loop
- Problem
- Evidence
- Causes
- Hypotheses
- Intervention
- Experiment
- Result
- Learning
- Memory
Learning returns to the system
Problem N+1 doesn't start from zero
The core records what was decided, what resulted and what was learned, with its provenance. A decision that leaves no memory gets made from scratch again; a result isn't promoted to a general rule without evidence. Memory accumulates — and the next decision starts with what was learned.
A diagnosis would have repeated the premise. GOVERNA challenged it
«Rural dropout is rising» is not confirmed: it fell from the 2022 peak. Telling level from dynamic is the difference between reacting to a number and understanding a movement.
Level (5.01%) and dynamic (declining) are two different questions. A dashboard that shows only the level invites reaction; the full series lets you understand.
When the decision has to stand on evidence
- Governments and agenciesDesign and test interventions before scaling
- NGOs and foundationsTurn a problem into evidence to decide and fund
- CooperationEvaluate contribution with counterfactual discipline
- Academia and observatoriesAccumulate institutional learning with provenance