Good Policy Problem Definition
A public policy problem is not yet a problem unless you can name at least three mutually exclusive solutions to address it. Until then, you are working on a solution looking for a justification, not a problem looking for a solution.
Origin
The framework comes from development economist Lant Pritchett, whose lectures on problem-driven iterative adaptation (PDIA) stress that reform must start with a problem, not a preferred answer. The version used here was introduced in Anticipating the Unintended.
What it says
The most common mistake in policy analysis is to define a problem as the absence of your preferred solution.
- Weak framing: “Government school teachers are paid less than they should be.” This smuggles in a single solution (raise salaries) and forecloses every other path.
- Strong framing: “Only half the children in class 5 can read a class 2 textbook.” This opens up teacher training, learning materials, accountability reforms, parental engagement, structured pedagogy — all of them in scope.
A good problem definition has three properties:
- Outcome-oriented. It describes the state of the world, not the absence of an intervention.
- Specific enough to be measured. Vague problem statements cannot be tracked.
- Broad enough to admit multiple solutions. If there is only one conceivable fix, you have stated a solution, not a problem.
Applied
India’s primary education debate has historically been hijacked by solution-first framing. Teacher unions demand higher pay; civil society demands smaller class sizes; technocrats demand more testing. Each defines the problem as the absence of their remedy. ASER data — showing that half of class 5 students cannot read class 2 text — reframes the issue around learning outcomes, forcing all sides to argue why their proposal would actually move that number.
The same logic applies to air pollution. “We need more electric buses” is a solution masquerading as a problem. “Delhi’s winter PM2.5 exceeds WHO limits by a factor of ten” is a problem statement that invites stubble-burning controls, industrial emission norms, dust management, and transport reform alike.
When it falls short
The “three solutions” heuristic is arbitrary. Some genuine problems — say, a collapsed banking system — may have a narrow range of viable responses. The framework can also be gamed: analysts can generate three token alternatives while pushing their favourite. Finally, it is silent on who gets to define the problem, which is often the most political step of all.
Further reading
- Head, B. W. (2008). Wicked Problems in Public Policy.
Originally explored in A Framework a Week on Anticipating the Unintended.