Making Decisions with Uncertain Science

We rarely get to wait for perfect science before deciding. Pandemics, new technologies, and environmental risks force choices while the evidence is still incomplete and experts still disagree. Knowing how to act wisely under uncertainty — neither frozen nor reckless — is one of the most valuable products of a scientific education.

Deciding without certainty

Waiting for certainty is itself a decision, often a costly one. The realistic goal is to act on the best available evidence while staying ready to update. Good decisions weigh the probabilities and the stakes: how likely is each outcome, and how bad or good would it be?

The precautionary principle

The precautionary principle says that when an action might cause serious or irreversible harm, a lack of full scientific certainty should not be a reason to delay protective measures. It is valuable for high-stakes, irreversible risks — but it has limits: applied too broadly, it can block beneficial innovation, and inaction also carries risks. Precaution must be weighed, not applied absolutely.

Reversibility and expert disagreement

Two practical guides help. First, favour reversible steps when uncertain, so mistakes can be corrected. Second, when experts disagree, look at why: is it genuine frontier uncertainty, differing values, or a fringe view against a strong consensus? Disagreement at the edges does not erase agreement at the core.

A new chemical might be harmful, but studies are inconclusive. A reasonable approach: assess the plausibility and severity of harm, prefer reversible precautions (limit exposure while studying it) over either banning it outright or ignoring the risk, and revisit as evidence accumulates. The decision is provisional and proportionate — matched to the stakes and the uncertainty.

Both extremes fail. Demanding absolute proof before acting can allow real harm to accumulate (as happened historically with tobacco and asbestos, where certainty came late). But applying maximum precaution to every speculative risk would halt medicine, technology, and daily life. And treating any expert disagreement as "nobody knows" hands undue weight to fringe views. Wisdom lives between paralysis and recklessness.

A landmark example of deciding well under uncertainty is the response to the ozone layer. In the 1970s and 1980s, scientists warned that certain human-made chemicals (CFCs) were damaging the stratospheric ozone that shields life from harmful ultraviolet radiation. The science was strong but not complete, and the economic stakes were large. Rather than wait for absolute certainty, the world's governments agreed to the Montreal Protocol in 1987 to phase out the chemicals — and as evidence continued to firm up, they strengthened it. Decades later, the ozone layer is slowly recovering. The episode is widely cited as a success of acting on strong-but-incomplete science with proportionate, revisable measures — precaution applied wisely, guided by evidence and updated over time.

• "We must be certain before acting." Waiting is itself a choice with its own risks.
• "The precautionary principle means always choose the safest-sounding option." Inaction carries risk too; precaution must be weighed.
• "Any expert disagreement means the science is unsettled." Frontier disagreement can coexist with a firm core consensus.
• "Acting under uncertainty means guessing." It means weighing probabilities and stakes and staying ready to update.

Take a real risk decision (personal or public). List: how likely the harm is, how severe, whether the response is reversible, and whether waiting is safe. What does a proportionate response look like?

Think Like a Maester: When certainty is out of reach, act on the best evidence, prefer reversible steps, and keep updating — that is courage guided by science.

1. Why is waiting for certainty itself a decision?
2. What does the precautionary principle say, and what are its limits?
3. Why favour reversible steps under uncertainty?
4. How should you interpret disagreement among experts?
5. Why is the Montreal Protocol (1987) cited as deciding well under uncertainty?

We usually must decide before the science is complete, so the goal is to act on the best available evidence, weigh probabilities and stakes, prefer reversible steps, and update as we learn. The precautionary principle helps for serious, irreversible risks but must be weighed against the risks of inaction and lost benefits. The 1987 Montreal Protocol shows the ideal: acting on strong-but-incomplete science with proportionate, revisable measures — between paralysis and recklessness.

• The Limits and Frontiers of Science
• Science for a Better Future