When a prominent UK politician announced on live TV that the Government had hit its target of 100,000 coronavirus tests a day by the end of April, on the very last day of that month no less, journalists faced a challenge.
Two hours earlier, specialist publication Health Service Journal had revealed that the figures had been fudged: instead of counting the numbers of tests that had been conducted on samples, a source informed them, the Government had quietly changed its own metric so that a test that had been sent out in the post — and not returned or tested — could now be added to the figures.
40,000 tests were then sent out in one day.
By any reasonable understanding, a test sent was not the same thing as a test done, as a raft of jokes — from people saying they had marked their students’ homework by sticking it in the mail, or paid their tax by receiving a letter from the taxman — pointed out.
And yet there was the Government making its claim — at length and without question, on the national broadcaster, and on the websites of national news organisations.
It was 20 minutes before the claim was queried by a reporter, by which time many viewers had switched off.
How journalists responded to this announcement — in different ways, at different times, and in different places — provides a valuable case study for anyone dealing with numbers and the claims that politicans make about them.
Live factchecking, tickers, notifications and alerts
In the US a debate has been raging for weeks about how news media should approach the country’s own daily coronavirus briefings, where falsehoods and health-endangering statements are common.
Suggestions have included broadcasting only part of the briefing, to adding a time delay to allow narrators to add factchecking context, to not broadcasting it at all.
In the UK, where no such discussion has taken place, it’s fair to say that national broadcasters and news organisations have been unprepared to check statements as they go out live.
Should they have been? Certainly the Health Secretary — at one point expected to be made the ‘fall guy’ over the Government’s handling of coronavirus — was not expected to hit the target, at least without some fudging.
And with the Health Service Journal article having been live for two hours by the point the announcement was made, there’s a strong argument that news organisations should have better anticipated the dubious claim that was about to be made using their channels.
Notably, as the claim went out live on the BBC its own health correspondent Hugh Pym — along with other journalists — used Twitter to provide some factual context about how the total was calculated (it’s not clear why he didn’t link to HSJ or specify the scale of the numbers involved).
But the same information wasn’t incorporated into the tickers that accompanied the live broadcast.
The BBC’s news ticker, for example, quickly updated to say that 122,000 tests had been “carried out” — words that were, it turned out, demonstrably false.
In fact those listening carefully to the speech might have noticed that the passage containing the claim avoided using any reference to tests being carried out at all: “The number of tests, yesterday, on the last day of April, was 122,347.”
The only claim that was being made for the tests on that day, in other words, was that they merely existed.
Elsewhere, text alerts and news app notifications immediately alerted audiences of the announcement that the target had been “hit”.
But the rush to notify left vital context behind. Were those writing the notifications unaware of the Health Service Journal story? Or did it complicate the story too much to fit into a push notification?
Whichever it was, notifications in this context turned news into stenography: or worse, like the BBC news ticker, misrepresented the claim in the process of rewriting it.
Meanwhile, 20 minutes passed, and as the daily briefing moved on to questions from the news media, it quickly became clear that the game was up. Sky’s Sam Coates asked about the Health Service Journal exclusive and, after his question wasn’t directly answered, Channel 4’s Victoria MacDonald followed it up, and the devil in the detail was exposed.
But by then it was too late — the alerts had gone out.
Writing the headlines, writing the scripts
The announcement challenged editors across news organisations not to mislead in their headlines.
Some fell back on “he said, she said” approaches: the FT, Evening Standard and Birmingham Live, for example, reported that the Government “claimed” that the target had been reached.
The Guardian went further, headlining both on Matt Hancock “saying” that the target had been hit and “claims” that the numbers had been artificially boosted.
Broadcasters’ websites handled the tension differently, Sky News‘s headline perhaps best of all: stating simply and factually that it had altered the counting criteria.
The BBC News website also stated factually that a third of tests were sent in the post, but relegated that fact to a secondary position underneath the headline heralding that the target had been passed.
But it was on air where broadcasters took a particularly interesting decision: to move past the figures and focus on what was actually happening on the ground.
ITV News anchor Tom Bradby was frank in reporting the announcement: it had been achieved by “fiddling the figures”, he said, but then noted the achievement of the increase to the 80,000 tests that had actually been completed. Then, instead of bogging down valuable airtime on discussing semantics, the programme simply moved on to whether tests were reaching the right people.
Over on the BBC, a similar — albeit more cautious — approach was being adopted. “That includes nearly 40,000 test kits posted out, which may not yet have been used or returned to laboratories,” the presenter said of the target being “hit”.
“Some frontline medical staff,” she continued, “say they are still struggling to get access to tests,” a factor fleshed out in the report that followed.
The story wasn’t about the numbers, it wasn’t about targets or political survival (stories which could be picked up later) — it was about people and whether things were working as they should.
UPDATE (May 14 2020): Research by Stephen Cushion and colleagues sheds further light on how consumers reacted to the day’s coverage, and adds detail on how the announcemet was reported in other outlets.
A great Google Sheet with the test numbers and results: https://sourceful.co.uk/doc/1/uk-coronaviruscovid-19-figures-officially-confirmed-by-dhsc-and-phe-on-twittergovuk
Pingback: 3 more angles most often used to tell data stories: explorers, relationships and bad data stories | Online Journalism Blog
Pingback: 3 more angles most often used to tell data stories: explorers, relationships and bad data stories – Blu-Ray Films
Pingback: 3 more angles most often used to tell data stories: explorers, relationships and bad data stories – Totally news
Pingback: 3 more angles most often used to tell data stories: explorers, relationships and bad data stories – BUZZINCLICK
Pingback: 3 more angles most often used to tell data stories: explorers, relationships and bad data stories – WordPress Website
Pingback: 3 more angles most often used to tell data stories: explorers, relationships and bad data stories – Media