Gender Pay Gap

Episode Summary

Executive Summary: This episode examines how statistics can mislead when definitions and contexts differ. It compares rival claims about the gender pay gap, critiques politicians’ use of “hospital equivalents” to frame spending, audits a partisan fact-checking site about Jeremy Corbyn, and explains why death penalty statistics hide complex realities. The show ends with upbeat UK tourism, beach, and restaurant numbers.

Main Topics: Gender pay gap: different measures, different conclusions (Priority: 5/5): The programme compares an IFS analysis showing persistent gender pay gaps with Fraser Nelson’s claim that women born after 1975 face no gap. It explains that the answer depends on whether one includes part-time workers and which cohort is measured. “Hospitals as a unit” for political spending claims (Priority: 5/5): The show argues that converting large sums into numbers of hospitals is a poor metric because hospital costs vary wildly by type, location, and project scope, and because NHS care is not organized around hospitals alone. Fact-checking fact-checkers: CorbynFacts under scrutiny (Priority: 5/5): More or Less tests claims on a partisan fact-checking website, showing that one “122 speeches” claim actually came from media mentions, and that electoral and polling claims about Labour’s performance and popularity are selective or misleading. Death penalty statistics and hidden complexity (Priority: 5/5): The episode revisits global death penalty decline statistics and shows that ‘de facto abolition’ often still includes ongoing death sentences and death row populations, meaning the headline numbers obscure ongoing punishment and legal limbo. The limits of statistical labels and categories (Priority: 4/5): Across all segments, the programme emphasizes that numbers only make sense when definitions are clear; otherwise, the same data can support contradictory narratives. Good news metrics: tourism, beaches, and Michelin stars (Priority: 3/5): The closing segment uses positive indicators to show improvement in UK tourism, beach cleanliness, and restaurant quality, framing them as evidence of long-term progress.

Key Arguments: The gender pay gap and equal pay are not the same thing; equal pay is illegal to measure directly, while the gender pay gap captures broader structural differences in pay. Including part-time workers significantly widens the measured gender pay gap because women are much more likely to work part-time and part-time hourly pay is lower. Fraser Nelson’s claim can appear true when focusing only on full-time workers born after 1975, while the IFS uses a broader measure including part-timers and finds a remaining gap. Using “one hospital” as a spending unit is misleading because hospital projects range from small units to billion-pound rebuilds, and operating costs also vary greatly. Politicians favor hospitals as a rhetorical device because buildings are tangible, but health spending is often about staff, drugs, GP care, and social care rather than new hospitals. The CorbynFacts website selectively framed a media-coverage statistic as “speeches,” demonstrating how wording can distort evidence. Claims that Labour matched Ed Miliband’s 2012 electoral performance are incomplete because 2016 saw seat losses and a lower projected national vote share. Polls did not show Labour neck-and-neck with Conservatives for a sustained period; any brief parity was short-lived and later reversed. Death penalty categories such as “de facto abolition” do not mean the punishment has disappeared; many countries still impose death sentences and maintain large death rows. Legal advocacy and court appeals have been instrumental in reducing executions, but political reluctance and deterrence arguments keep the death penalty on the books in some countries.

Data Points: Women part-time workforce share vs men: Women are three times more likely to be part-time workers - Used to explain why including part-time workers widens the gender pay gap Part-time vs full-time pay gap among women: 32% less per hour - Female part-timers compared with female full-timers Vote Leave hospital cost claim: £350 million a week - Converted into “one new NHS hospital every week” Huffington Post hospital cost estimate: £178 million - Implied cost per hospital in a political savings argument Daniel Hannan hospital cost estimate: £700 million - Derived from claimed misspending enough to build 10 hospitals Cornwall hospital project cost: £7 million - Example of a very small hospital project, a mental health unit St Bartholomew’s Hospital reconstruction cost: £1.1 billion - Example of a very large hospital project MRI scanner cost: £895,000 - Suggested as a more fixed and comparable healthcare unit Hip operation cost: £5,000 to £6,000 - Suggested alternative metric for spending claims Tourists visiting the UK in 2015: 36 million - Used in the good news segment about growth in tourism Tourists visiting the UK in 1980: 2.9 million - Historical comparison for tourism growth Tourist nights stayed in the UK in 2015: 271 million nights - Evidence of increased tourism volume Tourist spending in 2015: £22 billion - Additional tourism growth indicator Clean English beaches and bathing sites in 2015: 97% - Compared with much poorer conditions 20 years earlier Unclean English beaches and bathing sites 20 years earlier: 45% not clean enough - Shows improvement in water quality Michelin-starred restaurants in the UK: 169 - Used as a sign of improvement in British cooking Michelin-starred restaurants in 1980: 20 - Historical baseline for restaurant quality Countries abolishing the death penalty in 1998: 35 - Earlier global abolition figure Countries abolishing the death penalty today: 107 - Current abolition figure cited in the episode Countries that are de facto abolitionist: 52 - Countries that have not executed anyone in at least 10 years Countries executing anyone in the past 10 years: 39 - Shows how few countries actually used the death penalty recently Kenya death row population: Over 4,000 people - Example of a de facto abolitionist country still imposing death sentences Commonwealth Caribbean countries with people on death row within 10 years: 12 of 12 - Shows the breadth of ongoing death-row sentencing in the region Countries that passed at least one death sentence in 2015: Almost half - Indicates continued use of the sentence in de facto abolitionist countries Labour projected national share in local elections 2015: 29% - BBC PNS comparison under Corbyn-related claims Labour projected national share in local elections 2016: 31% - Shows modest improvement year over year Ed Miliband Labour PNS in 2012: 38% - Used to challenge claims that Corbyn matched Miliband’s local election performance Corbyn’s comparable 2016 PNS: 31% - Lower than Miliband’s 2012 result Polls monitored by David Cowling: About 75 to 80 polls - Basis for judging claims that Labour was neck and neck with Conservatives Periods when Labour was ahead in polls: A handful of polls, by 1-2 points or even - Temporary rather than sustained lead 2016 local-election seat result cited on social media: 47% of seats - Used to compare Corbyn with Miliband’s 48%, but criticized as misleading Ed Miliband’s comparable seat result: 48% of seats - Social-media comparison point

Pivotal Quotes: "The number you get depends on the question you ask." — Tim Harford: Summing up the gender pay gap discussion and the importance of definitions "In fact, the more we talk about this, the more I'm absolutely in agreement that it's actually completely the wrong measure because it varies so much." — John Appleby: Explaining why hospital costs are a poor spending metric "It reminds us that you look at a statistic ... but unless you really understand what's being counted, you haven't really learned anything at all." — Tim Harford: Conclusion of the death penalty segment

Implications: Listeners should treat headline statistics and political comparisons cautiously, ask what exactly is being counted, and prefer metrics that are comparable and well-defined. Partisan or simplified numbers often obscure more than they reveal.