Interview
#2 - Prof David Spiegelhalter on risk, stats and improving understanding of science
80,000 HoursProf David Spiegelhalter, Robert Wiblin, Jess Whittlestone, David Spiegelhalter, Keiran Harris
- David Spiegelhalter serves as the Winton Professor for the Public Understanding of Risk at Cambridge University, funded by a hedge fund (Winton Capital Management) to focus on improving public comprehension of statistics and risk.
- His professional focus centers on "everyday" risks (e.g., diet, exercise, cancer screening) rather than abstract global existential threats, aiming to correct public biases regarding probability estimation.
- Spiegelhalter notes that the public frequently overestimates salient, media-driven risks (e.g., terrorism, Ebola, plane crashes) due to the "availability heuristic" while underestimating chronic, familiar threats like heart disease and cancer caused by lifestyle factors.
- He cites the redesign of UK cancer screening leaflets as a key success story, where they present balanced benefits and harms without issuing recommendations, allowing individuals to make informed choices.
- Spiegelhalter advocates for public engagement as a professional duty for publicly funded academics, supporting career structures that incentivize outreach and the development of "statistics ambassadors."
- He describes academia as increasingly difficult due to pressure to publish, yet valuable for building transferable skills vital for global initiatives like the Millennium and Sustainable Development Goals.
- Regarding the criticism that quantifying impact is impossible, he argues that while measurement is never perfect, it is superior to "gut feeling," citing the UK's National Institute for Health and Care Excellence (NICE) as a successful model.
- NICE prioritizes health interventions using Quality-Adjusted Life Years (QALYs), typically accepting treatments costing under £20,000 per QALY and negotiating prices for those between £20,000 and £30,000.
- Current NHS policy applies a discount rate of 3.5% to future health outcomes, valuing a life year in 25 years at roughly half its present value; Spiegelhalter notes this rate is appropriate for healthcare but potentially too high for long-term climate change policy.
- For decision-making under "deep uncertainty" (where options and outcomes are unknown), he recommends strategies prioritizing flexibility, adaptability, and resilience over optimization.
- To build resilience, individuals and organizations should avoid putting "all eggs in one basket" and cultivate transferable skills that allow for career pivots in response to unforeseen events.
- Spiegelhalter suggests a mixed-strategy approach for altruism, balancing investments in predictable, safe-return charities with speculative, high-impact interventions to manage risk.
- He invented the "microlife," a unit representing a 30-minute change in life expectancy, to quantify and compare chronic risks like smoking, diet, and exercise.
- The "micromort" (a 1-in-1-million chance of death) was used to quantify acute risks, such as the ~7 micromorts associated with skydiving.
- He discusses the potential for new metrics like a "micro-happy" to measure improvements in subjective well-being (e.g., moving a person from a 6 to a 7 on a 1-10 scale).
- Well-being scores vary significantly by culture, with residents of South America typically reporting higher scores (around 8) and those in Eastern Europe, Asia, and Africa reporting lower scores (sometimes as low as 4), despite external life conditions.
- Spiegelhalter warns against "metricizing" pitfalls where organizations obsess over narrow indicators (e.g., academic impact factors) to the detriment of unmeasured but valuable activities.
- He argues that statistical literacy requires understanding both the strengths and frailties of numbers to avoid the extremes of blind acceptance or total rejection of quantitative evidence.