Risk and Uncertainty

1. Risk, Uncertainty, and the
Precautionary Principle

2. Types of Probability
●a priori probability: known outcomes.
–ex. rolling a dice, roulette wheel
●Statistical probability: Observed frequencies used to
predict outcomes.
–ex. odds of being killed on a single airline
flight are 1/29 million
●Estimated probability (uncertainty)
–Most common, demands judgment

3. Risk vs Uncertainty
●Risk: possible outcomes are known, as are their
probabilities of occurring. Tangible and quantified.
●Uncertainty: outcomes and/or probabilities are
either unknown or are estimated with low precision

4. Risk vs Uncertainty cont'd...
●Whether a decision is made under risk or uncertainty
depends on your confidence in the reliability of the
probability estimate.
●Real-life situations usually involve uncertainty
without exact probabilities available. So strictly
speaking, almost all decisions are made “under
uncertainty.”
●There are methods for trying to turn uncertainty into
risk and manage that risk.

5. Bayesian Probability
●Assigns a definite probability value to every
statement about the world.
–Almost nothing (apart from axioms) is fully believed.
●Unfeasible: Doesn't help us achieve a practical
belief system because we have limited cognitive
capacities.
–In order to grasp complex situations (and make
decisions), we need to reduce uncertainty to de facto
belief rather than just probabilities.

6. Unknown Possibilities
●Sometimes we don't have a complete list of the
alternatives or consequences that should be taken
into account.
–Creators of first atomic bomb worried that detonation
might consume the atmosphere, kill all life on Earth.
●Still, no scientific calculation can remove
apprehension about the possibility of something that
nobody has been able to think of.
●But taking this logic to extreme paralyzes decision-
making.

7. Unknown Possibilities cont'd...
●When to account for unknown possibilities, when to
ignore them?
●Novelty: new and untested phenomena should be treated
with special care.
●System limitations: cautiousness important when potential
impacts may have unlimited or very long lasting
consequences.
●Complex systems: ecosystems and the atmospheric system
may be impossible to restore after a major disturbance.

8. Responding to Risk and
Uncertainty
●Expected utility can be used to make decisions
–ex. betting $10 on red in Roulette pays 1:1
●[($10)*(0.47)] + [(- $10)*(0.53)] = - $0.60 EV
●But the morally relevant aspects of situations of risk
and uncertainty go far beyond the impersonal sets of
consequences that expected utility operates on
●Proponents of nuclear energy emphasize how small
the risks are (noun), whereas opponents question the
very act of risking (verb) improbable but potentially
devastating accidents.

9. The Precautionary Principle
●Designed to ensure that the absence of scientific
certainty isn't used as a reason for postponing
actions that could protect people and environment
when there's a credible threat of serious or
irreversible harm.
●Maximin approach: choose the alternative that
maximizes the minimum possible outcome.
–Useful when the negative outcome is ruinous. But
ignores the probability of the ruinous outcome as
well as potentially forsaken benefits.

10. Principle of Bounded
Subadditivity
●An event has more psychological impact when it
turns impossibility into possibility, or possibility
into certainty, than when it merely makes a
possibility more likely.
–ex. Greater impact of changing from 0.9 to 1 or from
0 to 0.1 than changing the probability from 0.3 to
0.4
●Biopsy example
–People value the elimination of a hazard more than a
comparable reduction in its likelihood

11. Applications for Climate
Policy

12. The Cascade of Uncertainty

13. Approaches to acceptable-risk
decisions
●Formal Analysis: cost-benefit analysis and decision analysis
–Formalized prescriptive procedure
–Complex problems decomposed into more manageable
components to be studied
●Bootstrapping: use tried and tested methods
–Historical experience and standards prescribe future action
●Professional Judgment: alternatives emerge from decisions
of qualified experts
–They may use formal analysis, but their own “best judgment”
is final arbiter of whether or not to accept a given risk

14. Iterative Risk Management Process

15. Framing and Decision Processes
●“Predict-then-act” (aka “top-down”, science-first,
scenario approach)
–Impact uncertainty described independently of other
parts of the decision problem
–Probability estimates followed by impact projections
●“Assess-risk-of-policy” (aka “bottom-up”, context-
first, vulnerability approach)
–Starts with decision-making context
–Uncertainty description customized to focus on key
factors and goals as decided by policy-makers

16. 7 Criteria for evaluating
approaches to acceptable-risk
●Comprehensive
●Logically sound
●Practical
●Open to evaluation
●Politically acceptable
●Compatible with institutions
●Conducive to learning