The picture isn’t as clear-cut as one might think, so here’s a short thread on the key details:
At first glance, it seems pretty clear that London should go into T3. Overall, London case rates are currently roughly at the T3 average, with several boroughs showing rates that would be high even in the top tier.
but ...
... the question is not necessarily how case rates compare between the tiers *now*, but how current rates compare to *when tiers were introduced*
On that basis, London as a whole is still at very low end of T3 when it came in (grey circles), and most boroughs exceptionally low
In addition, there’s a clear divide within London. Wealthier Boroughs like Camden, Islington & Richmond already had low rates, and have fallen further, whereas places like Havering, Newham & Enfield had higher rates and have risen in the last two weeks.
Splitting London across tiers would be very difficult, but to the extent that London should go up a tier, there is limited evidence pointing in that direction for some boroughs, but zero for others.
Finally, a positive side-note:
The grey circles (case rates before tiers came in) are almost all higher than coloured circles (latest data), showing that new infections have fallen considerably since restrictions came in, especially in Tier 3 areas.
Tiers appear to be working.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
NEW: the latest UK data is out, and it’s not good.
London is into Tier 3, but that’s only a small part of the wider story of UK’s looming Covid winter.
Case rates rising fast in London, SE & E, but also now rising in Midlands, NW and SW, and decline has halted in NE.
Regrettably, it’s also clear only tough restrictions suppress transmission.
During national lockdown, case rates either went from rising to falling, or from falling slowly to fast, in all English regions.
Since restrictions eased, all declines have either flattened or reversed.
The picture in Wales is even more stark:
During the 17 days of their "fire-break" lockdown, the weekly case rate fell from 295 per 100k to 184. In the first 17 days after restrictions eased, it rebounded to 301.
In the 6 days since then, the rate has increased by 45% to 440.
NEW: it’s expected (but not yet confirmed) London will go into Tier 3 when England’s Covid restrictions are reassessed next week, so what do the data show?
Let’s dig into all 6 metrics the govt is using.
First, new case rates:
• High in much of London, well into old T3 levels
But with cases, *growth rate* is also a key consideration, and this could be the big one for London.
Rates are rising in the vast majority of London boroughs, increasing by 50% per week in many places, and rising faster than almost all current Tier 3 areas.
Sticking to cases, government is also looking at rates specifically among those aged 60 and over.
As with the pattern in the general population, London boroughs send that Tier 2 slope steeply upwards on the left, with case rates among the 60+ well into former Tier 3 territory.
Let’s set aside that ONS series, and instead look at:
• Imperial REACT study
• Covid symptom tracker
• ONS’s model-free weighted estimates (not subject to revisions)
• The dashboard
Each as it stood when lockdown was announced.
All pointed to rapidly increasing infections
tl;dr
The way ONS models incidence can a) give a misleading shape to recent days of the outbreak, and b) cause confusion where estimated prevalence is retrospectively altered.
But there’s no doubt the virus was spreading rapidly in late October when lockdown was announced
Our job is to communicate clearly to people. If they are confused, that's on us much more than on them.
And no, this doesn't mean we should change our definition of things like unemployment and GDP to match what people [mis]understand, but it does mean we should provide explainers if we're using these terms when speaking to a mass audience.
It's the same in #dataviz. If someone doesn't understand my chart, that's mainly on me.
This is why we always included log scale explainers when sharing our trajectory trackers, for example.
We're hear to communicate, not to make pronouncements from on high and then walk away.
NEW: here’s the definitive chart on which parts of England have been hard-done-by or "let off" by the new tiers.
This one combines all the metrics the govt says it’s using:
• Cases (overall rate, rise or fall, and over-60s)
• Hospital occupancy & admissions
• Positivity rate
In summary:
• Tiers seem correct for majority of places. All high-risk areas are in highest tier
• "Harsh" decisions like Stratford typically areas whose neighbours have high risk
• Outer London has been "let off", but difficult to have outer & inner London in different tiers
The question some might ask is:
If places like Stratford were bumped up a tier because of high risk in surrounding areas, why was inner London not bumped up to tier 3 where parts of outer London appear to belong?
I think there are valid reasons, but it’s worth pondering.