Thread by @EdConwaySky

"I don't recognise those figures".
This is odd given the figs @TrevorPTweets quoted to @MattHancock are govt's own figs from its NHS Test & Trace spreadsheets. They show it waited far longer to put India on the red list than Pakistan after its Covid positivity hit a certain level

https://twitter.com/ridgeonsunday/status/1401477808774320134

Let me take you (& @matthancock) through it.
This may seem like water under the bridge but it MATTERS because:
1 we now have more of this Delta/Indian variant than almost any other country outside Asia
2 we still have little clarity on how govt decides on red/amber/green lists

It's worth saying at the outset that @MattHancock is absolutely right about one thing: when taking these decisions it's far better to focus on data collected through our testing system on incoming travellers than data you get from, for instance @OurWorldInData. So let's do that.

Potted backstory: on Apr 2 the govt put Pakistan and Bangladesh on the red list. @MattHancock told us a few weeks ago he didn't put India on the list at the same time because the % of passengers coming back from 🇵🇰 & 🇧🇩 testing positive was 3x higher than from India. This was odd

Because the data publicly released by the NHS on traveller positivity suggested that in the period when the decision was announced, levels of covid among passengers from India were higher than Bangladesh and only slightly below Pakistan. I tweeted abt this a week or two back

https://twitter.com/edconwaysky/status/1394361567786016768

Mr Hancock appeared in the Commons a few days after that to reveal that this decision was in fact based on internal, unpublished figures. Presumably @DHSCgovuk is getting daily numbers while the rest of us have to rely on a spreadsheet which shows data for periods of time. Fine.

https://twitter.com/edconwaysky/status/1395021287253286920

Except while @matthancock might have been able to provide figures to prove his point about "three times higher", in so doing he revealed the trigger point for test positivity which compelled govt to put Pakistan on the red list: 4.6%. Remember that number...

Because we know from the govt's own data that by 7 Apr test positivity among passengers from India was at least 5.1% (I say at least because it's unclear whether the govt data on this is an avg for the period or the end date fig) gov.uk/government/col…

We're left with a question, the one @TrevorPTweets asked @matthancock today:
Why, when he put Pakistan on the list after test positivity hit 4.6%, did he not put India on when its positivity hit 5.1%?
Because the govt then waited not a few days but two WEEKS before doing so.

It's a q @MattHancock has yet to answer. But it matters enormously because we know now that the variant gained a foothold during that period.
Sure, it prob would have got here in the end. But time matters. Acting fast matters. In this case, govt paused in spite of worrying data

It matters because govt is incredibly opaque about these kinds of decisions. On lists. On lockdowns.
It's meaningless to say you're "following the data" if you don't then explain what level of data will provoke a given reaction. In fact it's worth than meaningless. It's damaging

Because in much the same way as the credibility of evidence-based policy was damaged by the Blair govt, which frequently ignored evidence if it suited its policy, the credibility of data-based policy is damaged by behaviour like this.