Radio 4: Royal Twins & Autumn Statement

Episode Summary

Executive Summary: This episode of More or Less examines how numbers can mislead when stripped of context: from morning sickness and royal twin speculation, to interpreting the Autumn Statement, to the underreported flaws in African GDP statistics, a Telegraph headline error, and mathematical oddities like zero being even. The thread running through all segments is that relative figures, selective comparisons, and poorly defined baselines can create false certainty.

Main Topics: Morning sickness and the odds of twins (Priority: 5/5): The show tests media claims that severe pregnancy sickness indicates twins, using large Scandinavian birth datasets to show the risk rises only modestly in absolute terms. Reading the Autumn Statement correctly (Priority: 5/5): Tim Harford and guests unpack political claims about tax shares and public investment, stressing how selective time periods can make ordinary fiscal figures look more dramatic than they are. African GDP statistics and outdated base years (Priority: 5/5): A longer segment explains how Ghana’s and Nigeria’s GDP can be badly mismeasured when economic structures change faster than statistical methods and base years are not updated. Telegraph headline error on primary school arrests (Priority: 4/5): The program debunks a misleading headline that confused arrests of under-18s with primary school pupils, showing how a simple category mistake can produce absurd conclusions. Lego and the nature of pure mathematics (Priority: 3/5): A mathematician uses Lego as an analogy for pure maths, arguing that mathematics is built from simple blocks and rules that can be recombined creatively. Zero as an even number (Priority: 3/5): The episode closes with a mathematical clarification that zero is even, illustrating how public intuition can differ from formal definitions.

Key Arguments: Severe morning sickness is associated with twins, but the absolute probability remains low; the media and bookmakers overstated what the data imply. Political claims about who pays more tax need precise definitions: the relevant group is top-income taxpayers, not vague notions of “the rich.” George Osborne’s claim that the richest paid a greater share of income tax in every coalition year was broadly, but not perfectly, supported by the data for the top 5% and 1%. Comparisons of public investment across governments can be misleading because spending depends heavily on the economic cycle, not just party politics. African GDP figures can become wildly inaccurate when economies diversify and statistical base years remain outdated for years or decades. Bad headline framing can transform a statistic about juvenile arrests into a nonsensical claim about primary-school children. Mathematics can be understood as combining basic elements with logic, much like building with Lego. Zero satisfies standard definitions of evenness, but many non-mathematicians still find the idea counterintuitive.

Data Points: Chance of twins in the UK: about 1.5% to 1.6% - Baseline probability discussed in the morning sickness segment Adjusted chance of twins with acute morning sickness: about 2.4% to 2.5% - Estimated absolute risk after severe morning sickness Relative increase in twin probability: 50% to 100% increase - Association found in large Scandinavian datasets Twin rate in the UK: 1.5% - Used by Yan Wong to emphasize rarity of twins Hospital-admission-level morning sickness rate: about 1% - Shows why large datasets are needed to detect correlation Norwegian birth dataset: almost 1 million births, 1967-2005 - Evidence base for the morning sickness/twins analysis Top 10% income tax threshold: £48,600 a year or more - Defined by Ruth Alexander as the relevant income-tax group Top 5% income tax threshold: more than £66,100 - Used to check Treasury claim about tax share Top 1% income tax threshold: £153,000 - Used to check Treasury claim about tax share Next year GDP growth forecast: 1.2% - OBR forecast cited in the Autumn Statement 2014 GDP growth forecast: 2% - OBR forecast cited in the Autumn Statement Deficit this year: 7.7% of GDP - Autumn Statement figure mentioned early in the fiscal discussion Deficit last year: 7.9% of GDP - Used to show only a modest near-term fall Forecast UK deficit: £108 billion - Described as uncomfortably above OECD average UK net debt this year: 74.7% of GDP - Autumn Statement forecast UK net debt next year: 76.8% of GDP - Autumn Statement forecast UK net debt in 2014: 79% of GDP - Autumn Statement forecast Projected total UK debt: almost £1,200 billion - Expected by the end of the tax year Ghana GDP revision: 60% overnight growth - 2010 statistical rebasing example Ghana base year: 1993 - Explains why GDP estimates became outdated Potential error in Ghana GDP figures: plus 10-20% - Estimate given by a Ghanaian analyst for informal economy undercounting Primary-age pupil arrests wrongly implied by headline: 200,000+ - Telegraph headline misframed arrests as if all were primary school pupils Arrests under 18 in England and Wales: 209,000 - Actual figure from Howard League FOI requests Arrests of primary-age pupils: 2,117 - Actual subgroup figure cited in the correction Under-11 arrests among that group: 2,117 are under age 11 - Shows the absurdity of the mistaken headline Arrests of primary pupils over 11: fewer than 100 pupils at primary school age over 11 - Highlights why the headline could not be correct Crushed Lego brick stack: 375,000 bricks - Previous experiment referenced in the Lego segment Inverted pyramid Lego tower: 380,380 bricks, 104 layers - Suggested alternative construction, about one metre tall

Pivotal Quotes: "Double what chance? Treble what chance? Without knowing the underlying probabilities, we're not really learning a lot." — Tim Harford: Explaining why relative-risk claims are misleading without a baseline "In many countries, the statistical office is like an orphan." — Shanta Devarajan: Describing why African statistical systems are weak and underfunded "I think that pure maths is really easy, because it's just like playing around with Lego." — Dr Eugenia Chang: Using Lego to explain the construction of mathematical ideas

Implications: Listeners should be wary of headlines and political claims that cite percentages without baselines, dates, or definitions. Better statistics require context, updated methods, and careful category checks—especially in economics and public debate.