A familiar sequence has developed in health policy. A genuine harm is identified in a minority of people. The response is then designed for everyone.
The logic is rarely stated outright, because stated outright it sounds strange. It runs roughly as follows: because some proportion of the population is harmed by a product or behaviour, and because that proportion cannot be identified in advance, the population as a whole becomes the unit of intervention. What began as a case for helping people at risk becomes a case for managing the conduct of people who are not.
The move from treatment to management
Public health has always operated on populations; that is what distinguishes it from clinical medicine, and it is why it has been effective against infectious disease. The shift worth noticing is not the population frame itself but the widening of what the frame is applied to — from transmissible risk, where one person’s exposure genuinely determines another’s, to individual consumption, where it largely does not.
Once that widening is accepted, there is no natural stopping point built into the argument. Any behaviour with a distribution of outcomes contains some people at the bad end of it. The case for intervening in one is structurally the case for intervening in all.
What would count as evidence
The question CGHA is interested in is narrower and more answerable than the philosophical one. When a measure of this kind is introduced, does the harm it was justified by actually fall? Not whether exposure fell, or compliance rose, or the measure was implemented as designed — whether the outcome that motivated it moved.
- What specific harm was the measure justified by?
- What was that harm doing before the measure was introduced?
- What has it done since?
- What else changed at the same time that could account for it?
These are not hostile questions. They are the questions any measure confident in its own effect should welcome.
