More or Less: Behind the Stats

Episode: Are Afghan nationals more likely to be convicted of sexual offences?

Host: Tim Harford, BBC Radio 4
Date: September 10, 2025

Episode Overview

This episode of “More or Less” investigates several high-profile statistical claims circulating in UK public debate. The team, led by Tim Harford, scrutinizes recent news stories, including:

• The true cost – and meaning – of the UK’s “national debt”
• The accuracy and context behind claims that Afghan migrants are far more likely to be convicted of sexual offences than British nationals
• The claim that 1 in 10 women leave work due to menopause
• A playful examination of how the Russian army's progress in Ukraine compares to the pace of a snail

Key Discussion Points & Insights

1. How Costly Is the UK’s National Debt?

With Economist Duncan Weldon
[00:00 – 08:48]

• June’s Per Capita Interest Payment: In June, the UK paid about £240 per person in debt interest, but this was unusually high due to the timing of annual interest payments. The average monthly figure is closer to £135 per person.
• Gold-Edged Bonds (Gilts): The UK borrows via “gilts” (bonds)—currency jargon with no impact on the fundamental borrowing process.
• International Comparisons: Comparing UK and German borrowing costs is tricky due to different currencies and economic contexts; German bond yields are lower mainly because of persistent lower growth and inflation.
• Trust & “Moron Premium”: The idea that the UK pays a ‘premium’ due to government policy decisions, especially post-2022 (Liz Truss era), with borrowing costs rising more compared to France or Germany. Inflation and certain government choices increase UK costs.

Notable Quotes:

• “Interest rates in general are much lower in the Eurozone ... because growth is weak, because inflation is weak, interest rates [are expected] to remain very low over the longer term.”
  — Duncan Weldon [04:28]
• “It might be unfair to call it a ‘moron premium,’ but ... some of this is to do with the actions the government has taken.”
  — Duncan Weldon [08:28]

2. Are Afghan Nationals 22 Times More Likely to be Convicted of Sexual Offences?

With Madeleine Sumption, Director, Migration Observatory, Oxford
[08:48 – 17:11]

• The Claim: A commentator on BBC’s Today programme claimed, “Afghan nationals are 22 times more likely to be convicted of a sex offence” than UK nationals, citing government data.
• Data Source & Method: The claim is based on Ministry of Justice figures obtained via FOI request—dividing recorded convictions for Afghan nationals by census-based estimates of Afghan residents, and comparing rates with UK nationals.
  • Potential Data Issues: Nationality reporting is self-declared and contains outdated or incorrect country names; small errors can be significant due to small sample size.
  • Actual Conviction Numbers: 2021: 20 convictions; 2022: 26; 2023: 31—totalling 77 over three years.
  • Not Individuals: Number counts convictions, not convicted people; repeat offenders count multiple times.
• Population Uncertainty: Unreliable estimates exist for the number of Afghan nationals, especially post-2021 (after Kabul’s fall and increased asylum applications).
• Ratio Uncertainty: Different denominators (population figures) produce ratios anywhere from 14 to 22 times higher.
  • Madeleine Sumption’s estimate: 14.5 times, not 22.
• Small Numbers, Big Variability: With just a handful of cases, small data errors can dramatically shift headline ratios.

Comparability & Demographics:

• Younger Population Skews: Migrant groups tend to be younger (20–40), the group most likely to be convicted of any crime. Comparing all ages between migrant groups and national population is misleading.
• When Adjusted for Age:
  • Overall, migrants are not overrepresented among prisoners, and may even be underrepresented for their age bracket.
  • Specific nationalities do show variation. Some Eastern European and Albanian nationals have higher conviction rates; Indian and Chinese nationals are far lower than British rates.

Notable Quotes:

• “It’s really hard to say. We don’t have a precise estimate of how many Afghan citizens live in the UK ... so you can make different decisions and come up with different numbers.”
  — Tim Harford [13:28]
• “When we look at these larger populations, it is clear that the rates at which people are convicted of crimes or incarcerated does vary quite a lot by nationality...India has less than half the rate as British citizens.”
  — Madeleine Sumption [17:11]

3. Do 1 in 10 Women Leave Work Due to Menopause?

With Dr. Gillian Paul, London School of Economics
[17:11 – 23:33]

• The Claim: News outlets and the Equality and Human Rights Commission stated that “1 in 10 women have left work due to menopause,” based on a Fawcett Society survey.
• Survey Flaw #1: Sample Bias: Only women experiencing menopause symptoms (often severe) were surveyed, leading to overrepresentation of those adversely affected.
• Survey Flaw #2: Ambiguous Question: “Have you left a job due to menopause symptoms?” does not specify leaving the workforce altogether—could include simply changing jobs.
• Stronger Data: Only 3% reported “retiring early” (i.e., leaving all work) due to menopause symptoms. Even this is likely an overestimate due to sample selection.
• Labour Force Comparison: More robust data suggest about 7% of women aged 45 in paid work are out of work at 56 due to any long-term sickness or disability—not just menopause.

Notable Quotes:

• “Compared to other surveys, these women seem to have a particularly high number of symptoms that they're experiencing. So it's suggesting not only is this a sample of women who are experiencing symptoms, but they seem to be particularly adversely affected.”
  — Dr. Gillian Paul [19:11]
• “There’s no way of knowing how much of that 10% of survey respondents were simply moving between jobs or leaving paid work altogether. And that’s quite a big difference.”
  — Dr. Gillian Paul [20:09]
• “So we've got very good reason to believe that it's not the case that 1 in 10 women leave work permanently because of menopause symptoms.”
  — Tim Harford [22:41]

4. Is Russia’s Advance in Ukraine Slower Than a Snail?

With Lizzie McNeil and Prof. Angus Davidson
[23:33 – 27:11]

• The Claim: Ex-Chief of Defence Staff Sir Tony Radekin claimed that a literal snail, leaving Russia’s Rostov-on-Don at war’s start, would have reached Poland by now—implying Russia’s advance is just as slow.
  • Fact-Check: For a snail to “make it”—1,500km from Rostov to Poland in about 1,300 days—it would need to move 50 metres per hour non-stop.
  • Reality Check: Fastest recorded snail covers about 10 metres per hour at a sprint; snails are not continuous travelers and rarely exceed 5 metres in a straight line.
• Bottom Line: The metaphor is greatly exaggerated, though the underlying point—that Russia’s advance is glacial—is valid.
• Wartime Reality: Recent Russian advances (Kupiansk, 2024–2025) average about 50 metres per day, slower than at the Battle of the Somme.

Notable Quotes:

• “The general's imaginary snail is five times faster than the fastest actual snail at a sprint.”
  — Tim Harford [26:11]
• “The recent Russian advances since January 2024 have been some of the slowest that we have identified since World War I.”
  — Seth Jones (cited) [27:01]

Key Timestamps & Segments

• [00:00 – 08:48]: National debt, bonds, and cross-country interest comparisons
• [08:48 – 17:11]: Analysis of sexual offence conviction rates among Afghan nationals vs. British nationals
• [17:11 – 23:33]: Claim that 1 in 10 women leave work due to menopause symptoms—scrutinized
• [23:33 – 27:11]: Comparing Russia’s Ukraine advance to a snail's pace

Memorable Moments

• The “moron premium” and its political baggage around UK borrowing costs.
• The distinction between convicted offences and convicted people in Ministry of Justice data—crucial for interpretation.
• Humorous yet factual debunking of the snail metaphor by actual snail-racing data.
• The repeated reminder that ambiguous survey phrasing leads to misleading headlines.

Summary: Facts Behind the Numbers

This episode typifies More or Less’s commitment to demystifying statistics driving headlines. Across topics—from government debt, crime rates, workplace gender claims, and even snails—the team demonstrates:

• Numbers may be accurate, but interpretation and context are everything.
• Small datasets and unclear denominators can produce wildly exaggerated ratios.
• Seemingly robust “data-driven” headlines often mask more nuance (or ambiguity) than they reveal.

Whether you care about public debt, immigration policy, women’s workplace experience, or the fortunes of slow-moving molluscs, this episode offers a masterclass in skeptical, compassionate statistical thinking.