Deaths at Home, Supermarket Infections and the Cobra Effect

Episode Summary

Executive Summary: This episode fact-checks several pandemic-related claims and one drug-deaths claim. It explains why daily COVID death headlines can be misleading, examines evidence that deaths shifted from hospitals to homes, evaluates a proposed self-isolation payment for possible 'cobra effect' risks, debunks a claim that supermarkets were a major source of infection, and corrects misinformation about Scottish drug-death statistics.

Main Topics: COVID death reporting and headline accuracy (Priority: 5/5): David Spiegelhalter explains that reported daily deaths reflect registration/reporting dates, not necessarily the day people died, making headline comparisons misleading. Deaths shifting from hospitals to homes (Priority: 5/5): ONS and Scottish data suggest many deaths moved location from hospitals to homes during the pandemic, raising questions about end-of-life care and whether people were avoiding hospitals. Self-isolation payments and the 'cobra effect' (Priority: 4/5): An economist assesses whether extending £500 self-isolation payments could perversely incentivize people to seek positive tests, while likely still producing a net public-health benefit. Where people catch COVID-19 (Priority: 5/5): A supermarket transmission claim is debunked; the program explains that contact-tracing tables were misread and that available evidence points more strongly to household and close social contact transmission. Scottish drug-deaths statistics and misinformation (Priority: 4/5): Social media claims that Scotland’s drug-death figures are artificially inflated by including unrelated deaths are shown to be false; the broader definition still leaves Scotland with exceptionally high rates.

Key Arguments: Daily COVID death figures are often reported by date of registration, not date of death, so single-day comparisons can be deeply misleading. The apparent surge in January headline deaths did not mean 1,820 people died that day; provisional data suggested closer to 1,028, possibly around 1,100. Comparisons with the first wave are distorted because testing was much more limited then, causing undercounting in the dashboard figures. Data from England, Wales, and Scotland suggest a substantial migration of deaths from hospitals to homes, especially for cancer, heart disease, and stroke. The shift in place of death does not automatically mean a rise in non-COVID deaths; many people may have died where they otherwise would have died, just not in hospital. Self-isolation payments can help people comply with quarantine, and existing evidence on sick pay generally suggests a net positive public-health effect. The 'cobra effect' is a real incentive-design risk, but no empirical evidence was presented showing that £500 payments would significantly encourage people to seek infection. The supermarket claim was based on a misunderstanding of Public Health England tables showing places visited by cases, not places where infection was acquired. Available contact-tracing evidence suggests household and close social contacts are the highest-risk transmission settings, while location-specific evidence for supermarkets is limited. Scottish drug-death figures do not include car crashes, drownings, assaults, or other unrelated deaths merely because drugs were present. Even if Scottish statistics differ slightly from some other countries, the scale of Scotland’s drug-death rate remains far higher than UK and European comparisons.

Data Points: Reported UK COVID deaths on 20 January: 1,820 - Headline figure discussed as a report count, not necessarily deaths occurring that day. Provisional actual deaths on that day: at least 1,028 - David Spiegelhalter said the true number for that day was far lower than the headline. Possible revised actual deaths: about 1,100 - Spiegelhalter estimated the total may rise modestly as more deaths are registered. First-wave dashboard peak deaths: 1,073 - UK dashboard figure for 8 April in the first wave. First-wave ONS-adjusted peak deaths: 1,456 - Shows undercounting in first-wave dashboard comparisons due to limited testing. England and Wales excess deaths in homes: 42,000 - Excess deaths registered at home up to 8 January, with only a small portion recorded as COVID. COVID deaths among home excess deaths: 4,000 - Of the 42,000 excess home deaths in England and Wales, only about 4,000 were COVID. Fewer non-COVID deaths than normal in hospitals: 42,000 - Mirrors the excess-home-death pattern, suggesting migration of deaths from hospitals to homes. Scottish extra deaths in 2020: about 7,000 - Total extra deaths in Scotland during 2020. Scottish COVID deaths in 2020: about 6,000 - Roughly matches the excess-death total in Scotland. Extra cancer deaths at home in Scotland: 1,900 - Part of the rise in home deaths during 2020. Extra heart disease and stroke deaths at home in Scotland: about 1,200 - Another key contributor to increased home deaths. Observed quarantine adherence: 11% - Government-backed study from September 2020 found only this share stayed indoors for the full required quarantine period. Self-isolation payment: £500 - Payment introduced for some people asked to self-isolate and later discussed for wider extension. Supermarket figure misread in media: 20% - Actually the share of locations in a table where multiple confirmed cases had both been, not the share of cases infected at supermarkets. Scotland drug deaths in 2019: 1,264 - Highest number in 25 years, 6% higher than the previous year. Increase in Scottish drug deaths over a decade: 132% - Long-term rise highlighted in the report. Scotland drug deaths per million: 295 - 2018 comparable rate used in European comparisons. Sweden drug deaths per million: 81 - Nearest high comparator in the European table cited. UK drug deaths per million: 76 - Used for comparison with Scotland. EU drug deaths per million: 24 - European average cited for comparison.

Pivotal Quotes: "these numbers depend so much on the day of the week, for example. Wednesday is almost always a high day" — David Spiegelhalter: Explaining why daily reported COVID death totals can be misleading. "The Cobra effect is a term that basically refers to a scenario where the incentives designed to solve a problem not only don't work, but they actually backfire completely" — Abby Adams Prassell: Introducing the risk that self-isolation payments could create perverse incentives. "we just don't know that" — Charlotte MacDonald: On claims that supermarkets are where people catch COVID-19, after explaining the data do not support that conclusion.

Implications: Listeners should treat daily pandemic figures and viral social-media claims cautiously. Better interpretation of reporting dates, transmission data, and drug-death definitions is essential before drawing policy conclusions or making comparisons.