In ordinary scenarios, we maximize expected utility (EU Max). However in precautionary scenarios, we want to avoid catastrophic harm (Precautionary Principle)
Tripod Framework
Manson’s Framework for when to use the precautionary principle
- Serious damage (damage condition): some outcomes exceed a threshold of badness and are considered catastrophic
- Thresholds for levels of harm: serious, catastrophic, irreversible, civilization collapse
- Thresholds for time scales: present, pre-2100 (current population), post-2100 (future generations)
- Scope: local, global
- Uncertainty (knowledge condition): some outcomes meet a required threshold of evidence to count as a serious threat
- Probability: possible, non-negligible probability, predicted by a model, predicted by a well-understood scientific mechanism
- Proportional remedy: response should correspond to the plausibility and severity of the threat
- Prohibition/ban, further research, moratorium, mitigation/restriction
However, could potentially lead to irrational outcomes or outcomes that are contradictory. e.g.
- Greenhouse gases (GHG) may cause catastrophic temperature rise ⇒ Ban GHG at once
- Banning GHG may lead to economic crisis, world war and nuclear holocaust ⇒ Don’t ban GHG
Versions of PP
- If the activity meets the damage condition, and
- If the link between the e-activity and the e-effect meets the knowledge condition,
- Then apply the e-remedy.
But, PP versions must be internally consistent; it cannot ban its own remedies
Catastrophe Principle
If an effect is catastrophic, and there is a possible link between the activity and the effect, then the remedy is to ban the activity completely.
For example, ban nukes because their possible use has a catastrophic effect