Hay Festival Special

Episode Summary

Executive Summary: Recorded live at Hay Festival, this More or Less episode examines how statistics can be misleading or revealing depending on what they measure. It contrasts the UK's status as the fifth largest economy with its much lower GDP per capita ranking, explains why child poverty doesn't track national wealth automatically, debunks the advice to always trust your first answer on multiple-choice tests, and features David Spiegelhalter on how statistical monitoring could have exposed Harold Shipman earlier.

Main Topics: GDP vs. “richest country” claims (Priority: 5/5): The show contrasts total GDP with GDP per capita, showing that the UK is fifth in overall economic size but much lower when wealth is measured per person. Child poverty and relative measurement (Priority: 5/5): Using OECD child poverty definitions, the episode shows that richer economies can still have worse child poverty rates because poverty is measured relative to median income. Playing multiple-choice strategy (Priority: 4/5): A live audience quiz is used to explore whether people should trust first instincts or second thoughts when changing answers on tests. The first-instinct fallacy (Priority: 5/5): Research cited by the programme suggests that changing answers is more often beneficial than harmful, contradicting the common belief that first answers should be preserved. Dunning-Kruger and overconfidence (Priority: 3/5): The episode connects test-taking overconfidence with the Dunning-Kruger effect, illustrating how people outside their expertise often misjudge their competence. Statistics as detective work: Harold Shipman (Priority: 5/5): David Spiegelhalter explains how statistical analysis of death patterns could have flagged Shipman's murders years earlier and potentially saved lives.

Key Arguments: The UK is correctly described as the fifth largest economy by GDP, but not the fifth richest by GDP per capita. A large economy does not automatically mean lower child poverty, because poverty is defined relative to median household income. Among the UK’s comparable large economies, the US, China, Japan, and sometimes Germany all have higher child poverty rates. GDP per capita gives a more meaningful picture of living standards than total GDP for judging national wealth. Most people believe they should stick with first instincts on multiple-choice questions, but research shows answer changes usually improve scores. The emotional discomfort of switching answers biases people toward sticking, even though data shows switching is often beneficial. The Dunning-Kruger effect illustrates how people can be unaware of their own lack of expertise. Statistical monitoring could have identified Harold Shipman’s unusually high mortality patterns much earlier than 1998. A GP with unusually high mortality is not necessarily culpable, but the data should trigger closer investigation rather than final judgment. Statistics are most powerful when they are used to ask the right questions of the right data, not just to produce averages.

Data Points: UK GDP ranking: 5th largest in the world - Total size of the UK economy in the GDP discussion UK GDP per capita ranking: 20th - IMF list cited in the live quiz Four largest economies by GDP: US, China, Japan, Germany - Audience quiz on countries ahead of the UK in total GDP Countries with higher child poverty than the UK: US, China, Japan; Germany sometimes depending on year - OECD relative child poverty comparison Poverty threshold: Below 50% of median household income - Standard OECD child poverty measure explained on stage EU countries with higher GDP per capita than the UK: 10 - Audience quiz on EU wealth rankings Richest countries by GDP per capita: Luxembourg, Switzerland, Norway, Ireland - Final live quiz on per-capita wealth Ireland vs UK on GNI: 30% richer per person - Alternative measure mentioned after GDP per capita Typical student belief about answer changing: 75% say keep first answer - Research summary on test-taking behavior College instructors agreeing with first-answer rule: 55% - Survey result mentioned during discussion of answer switching Shipman victims: More than 200 over 20 years - Background on Harold Shipman Shipman inquiry warning point: About 1985 - Statistical monitoring estimate for when an alert could have been triggered Lives potentially saved: More than 150 - Estimated impact if Shipman had been detected earlier Victims in one cluster: Nine victims in seven weeks, including two in one day - Public inquiry database described by Spiegelhalter

Pivotal Quotes: "We have statistics, and they're dreadful. They're positively Dickensian." — Ruth Alexander as Emma Thompson: Used to frame the political claim about child poverty and UK wealth "I think by most standards of richest, that's not right. Fifth largest, not fifth richest." — Tim Harford: Clarifying the difference between total GDP and wealth per person "when an answer is changed, it is more usually changed from wrong to right." — Tim Harford: Summarizing research on multiple-choice answer switching

Implications: Listeners are reminded to scrutinize what a statistic actually measures before repeating it. For policymakers and audiences, the episode shows that rankings, poverty, and risk can be distorted unless adjusted for context and the right denominator.