Covid-19 Is the UK Response Improving?


The previous blog Covid-19 How is the UK doing? asked the question and sadly came up with the answer that the UK, and in particular England, has had the worst response in Europe. Post-publication a new, interesting, but depressing paper has appeared: Measuring excess mortality: England is the European outlier in the Covid-19 pandemic, which supports this diagnosis.

How have things developed? One hopes that even though the UK had a slow start, there would be the capacity, desire and ability to improve its response. A Dunkirk to D-Day transformation? It has indeed been promised that by Monday (1st June), for example, that UK will not only have a fit for purpose Test, Track and Isolate (TTI) system in place, but it would be “world-leading.”

The fact that lockdown is being eased more rapidly in England than any other part of these islands suggests that Covid-19 in England, in particular, is under control, and the ability of primary schools to reopen on Monday 1st June surely gives confidence that the worst is over?

Alas, the reality is very different. HMG’s line seem to be increasingly detached from actual reality evidenced by official figures. So much so, in fact, as with the Dominic Cummings affair, it is increasingly impossible to trust this government or even to take it seriously.

This piece will look at the most recent numbers, EU comparisons, the response to lock down, and the Cummings affair. Analysis of why Cummings is so key is presented and conclusions drawn.

More Recent Numbers

As discussed in the last Covid blog, the ONS Mortality figures published every Tuesday are spin-free and can be taken at face value. The most recent ones to Week 20 (to 15th May) are plotted in Fig. 1. Deaths are counted on the day of registration. The uptick in Wk20 over Wk19 is due to the VE day bank holiday causing a delay in registration. The trend is still down. Excess deaths from Wk10-20 are determined using the 10yr average (2010-9).

Fig.1 Deaths and Excess deaths from the ONS by date of registration.

The deaths for Scotland an NI can be added and these numbers can be projected forward using hospital deaths and community estimates. Jamie Jenkins, @statsjamie (a former statistician at the @ONS), who adjusts to the actual day of death and uses a 5yr average (2015-19) has estimated the number of excess deaths at 65,880 as of the 25th May, as illustrated in Fig. 2. Jamie has also noticed that the latest ONS release includes a death where Covid was mentioned on the death certificate in the week ending 14/02/2020.

Fig. 2 Estimate of the number of death to 25th May from Jamie Jenkins.

There is good news here. Even though the numbers of excess deaths are still very high they are coming down rapidly. The lockdown measures have worked extremely well. With a few more weeks of lockdown, excess deaths should hopefully decrease further and indeed may be below zero (fewer people dying than average) as seen in many European countries. Sadly they still remain very high in England.

Fig.3 Preliminary z-scores for Week 20 from Euromomo

Fig.3 shows the preliminary z-scores for Week 20. England is very much the centre of the pandemic in Europe, being the only country with a very high excess of deaths. Sweden and Belgium have moderate excess. In Belgium, it seems the pandemic is finally under control. Sweden is pursuing a controversial “herd immunity” strategy.

England was still in a bad place in Week 20. This is now Week 22 and more recent data may be available to HMG, but easing lockdown still looks premature and a political decision rather than a public health one. Despite the fact that Scotland is showing low excess and Wales and Ireland (both North and South) show no excess, all are taking a far more cautious approach.

Detailed European Comparisons

In this previous blog the peak z-score was used to measure the severity of the pandemic in the various countries in these islands and beyond: England 42.75, Wales 19.71, NI 8.90, Scotland 7.03 and Ireland 3.95 all in week 15.

Tony B made an interesting remark on the previous blog re Wales:

[The] Welsh z-score for week 15 was inflated by catch up of one area that had non-reported, a local failure. If that data is returned to earlier weeks to reflect the true excess deaths with time then the Welsh z-score would peak below 19 and perhaps follow Scotland’s trend line.

I still think, at present, the z-score is the best metric for international comparison, but it is overall deaths throughout the duration of the pandemic, rather than the peak, which is more important. This can be captured by measuring the area under the curve in z-score weeks. This area, however, will be proportional to the average z-score value, which is probably an easier number to understand, and that is what is presented here. The numbers used throughout this section are those published on the Euromomo site. It was decided a priori to use Weeks 10 through to 20 for the computation of the average z-score.

Caveat: the numbers for weeks 18-20 are provisional and are likely to change.

Fig. 4 Excess Deaths (Z-Score) for England, Scotland, Wales, NI and Ireland Wk 10-20 2020.

In Fig. 4 excess z-scores for Britain and Ireland are plotted. The average z-score from week 10-20 are:

  • England 19.89
  • Scotland 7.63
  • Wales 7.07
  • Northern Ireland 3.31
  • Ireland 0.84

England does extremely badly on this measure, and Wales looks slightly better than Scotland. Northern Ireland does well, but the clear leader using this metric is Ireland. The English figures are more than twice as bad as Scotland and Wales and far worse than NI or IE.

It is interesting also that IE is so much better on this metric than NI. Other analyses put IE ahead of NI, but nowhere near the factor of about four that this indicates. It is possible as the Wk18-20 data is provisional and the IE numbers lower than NI in this period, adjustments may close the gap.

It’s worth noting that outside the main cities of Dublin and Cork the highest numbers of Covid cases in Ireland occurred in the border region, notably in County Cavan, suggesting that Ireland’s figures may include adverse effects of having two different public health regimes with different guidance (e.g., 7 and 14 day isolation periods in NI and IE respectively).

Fig.5 Excess deaths for England and chosen European countries.

The z-score for England and some chosen European countries are plotted in Fig. 5. Three of the countries Belgium (BE), Spain (ES) and Italy (IT) have been chosen as they are the worst in the EU. Sweden (SE) because of its controversial “herd immunity” strategy and Norway (NO) as being one of the best in class and a good comparator for Sweden. The average z-scores from Wk10-20 are:

  • England 19.89
  • Spain 11.9000
  • Belgium 11.2700
  • Italy 8.5300
  • Sweden 6.9000
  • Norway 0.1900

Spain and Belgium are more or less tied at 2nd place. England once again is a clear “winner”, approaching a score double the worst in the EU.

The contrast between Sweden and Norway is very stark. It seems Sweden is taking a gamble, let’s hope it pays off.

Ending Lockdown and the Cummings Affair

On the basis of the Week 20 estimates, England is the last country in Europe that should be easing lockdown. Worryingly, it is the country in these islands which is easing lockdown fastest. Scotland, Wales NI and Ireland are taking a much more cautious approach with the aim of avoiding or minimizing subsequent waves or outbreaks. Ireland, unlike the UK, has had a working Test, Track and Isolate system operation for the past two months with active public cooperation with all measures. Attitudes in the UK seem less supportive:

There have been widely publicised indications that in certain parts of England social distancing guidelines are not being strictly adhered to, both before (e.g. during the VE day celebrations) and after the Dominic Cummings affair.

The Cummings affair has been so shocking that it has cut through to the general English public. Two of the very best English language journalists writing today, Marina Hyde and Fintan O’Toole have written excellent pieces: The truth about why Cummings hasn’t gone: Johnson is too terrified to sack him and Cummings’ contempt for lockdown rules makes the public feel like fools respectively. Prof Chris Grey of the excellent BrexitBlog has also produced an analysis.

Fig 6. from @jonworth‘s Twitter Feed

I won’t attempt to compete with Hyde and O’Toole, their articles need to be read in full, but Fig 6. shows in visual form why the Cummings breach was on a totally different level to the infractions of Calderwood and Ferguson.

Compounding this has been the laughable defence of his actions given by Cummings in the Downing St. Rose Garden and the complete lack of contrition. The closing of Tory ranks to defend the plainly indefensible is extraordinary. Given that industrial-scale dishonesty has been the Tory norm since the Brexit referendum (with honourable exceptions) and tribal loyalty more important than truth for Tory supporters, Cummings may survive, for now.

This survival, however, will come at great cost.

Lockdown has worked very well. However, the feeling that we are all in it together is essential and as Karl Sharro succinctly puts it in a tweet:

It seems there is one law for the plebs and a totally different one for the elite. This has caused justifiable anger on the part of most who complied with the rules and is seen as a perfect opportunity to break lockdown rules by others.

Why is Cummings Being Defended?

Given the anger, even from natural Tory supporters, including even The Daily Mail, there must be a very good reason to keep Cummings on. The reason, of course, is that he is central to the entire project. Electoral success, the fact that Cummings and Johnson are essentially a double act, and delivering for the shadowy right-wing Tory paymasters are key. More sinister possibilities have, of course, been suggested.

Electoral success has been achieved, but at horrific cost. In The Resistable Rise of Cum-Jo (a Brechtian analysis) Martin Agombar argues:

Cummings uses the rule of thirds. It goes something like this. One third of the people you must twist up into a storm of righteous anger. You do this by convincing them they have been betrayed. They are your third. The other two thirds are your enemies and must be driven apart. To do this, make them confused. Destroy their will to fight. And do not forget to keep your third on the boil. They will respond with unblinking loyalty. Cummings and Johnson are the classic double act. Johnson distracts, so the real business of corruption can go on.

For it to work, we need the carefully cultivated third angry. How? By pushing their buttons. They are huge and well-defined buttons in England. Just lift the safety guard and press. They are marked Xenophobia, Patriotism, and Exceptionalism. Press, press, press!

This is made far easier in the UK because of the archaic FPTP system. The Tories achieved an 80 seat majority with only 43.6% of the vote. Past performance is no guarantee of future success but Cummings is perceived very much by the Tories as having an electoral Midas touch and is sold to many backbenchers as such.

Regarding delivery to the Tory paymasters, Carole Cadwalladr probably has the best analysis on Cummings and there is further discussion in TruePublica:

The real story is the huge transfer of government assets to private companies that he is overseeing under cover of a pandemic while restructuring the Cabinet Office into his & Michael Gove’s private fiefdom. We have no insight into these contracts which did not go to tender but to Cummings’s friends & associates. Including a massive transfer of our most private, intimate NHS data. The privatisation of essential state functions is happening in real time in almost complete darkness. Cummings made absolutely clear his conditions for working in government were smashing civil service and rebuilding in his image and creating British ‘ARPA’ – an AI-driven research facility. That’s exactly what he’s doing right now in Cabinet Office with Gove his long-time co-conspirator.

This is uniquely possible in the UK because of the Elective Dictatorship nature of the unwritten constitution. It seems the Tory paymasters are indeed being rewarded very handsomely.


England is by no means out of the woods regarding Covid. At the daily Downing Street briefing (28th May), chief scientific adviser Sir Patrick Vallance stressed “there is not a lot of room” to ease measures, with new Covid-19 cases still around 8,000 per day. He warned that with the rate of transmission, or R-value, “close to one”, the test and trace system must be effective in order to prevent exponential spread returning.

The continuing defence of Cummings doesn’t help. As Chris Grey has put it:

[…] With every ministerial announcement that he was within his rights to make his own interpretation of the rules comes the erosion of the possibility of the population at large adhering to them.

The TTI system is untested. The possibility of a return to exponential growth of the pandemic is very real. So too is the possibility of the UK, perhaps even just England, needing to be placed into international quarantine.

Another consequence is the further shredding of the UK’s international reputation. Brexiters imagine the call for Cummings to be sacked is entirely some kind of revenge by Remain voters and believe that his departure would adversely affect the UK’s continuing negotiations with the EU. In reality his retention signals that the UK cannot be assumed to be trustworthy, reliable and willing to act in good faith. Accordingly, the likelihood of a no-deal Brexit seems ever greater. This may well be what Cummings and Tory paymasters want. Agreeing to the EU’s “Level Playing Field” demands for single market access would stymie their deregulation agenda and with it their ability to pollute, lower workers rights, maximise profits and tie the UK into the orbit of the US.

Una Mullally writes in the Irish Times:

Other countries need to pay attention to modern Britain. What has happened is instructive. Accepting plummeting standards in media, possessing a pathological deference to posh bullies, being ignorant of one’s history, accepting inequality as some kind of natural law, celebrating buffoons, and a blithe and arrogant exceptionalism is the real downfall, not Dominic Cummings’ lies.

On a more optimistic note, many are now seeing what Fintan O’Toole this week called “the unpardonable snigger of elite condescension”. Fintan also notes that fall from power of the Catholic Church in Ireland was for similar reasons. The deference to the “Elite” in England, like that to clergy in Ireland once, is baffling to an outsider.

A no deal Brexit, if it happens, may turn out to be a “Black Monday” moment when the UK spectacularly crashed out of the ERM and the Tory poll lead never recovered. It may well be that the Tories reputation may be trashed for at least a generation. The worry is that with a five-year fixed-term parliament and a large majority, the damage may be so great that it may take decades to recover.

Una Mullally again:

The good news is, there is a way out. Throughout history, empires fall, societies implode. But the process of rebuilding can only happen if there is a period of profound self-reflection, a real and honest acknowledgement of who you are, and why you are. So stop the “shock” at the latest scandal, and start looking in the mirror.

Is it really necessary for the UK to experience at home the neoliberalism it inflicted on others in order to see itself clearly in the mirror, as it did when abolishing slavery?


  1. Peter May -

    Appalling entitled sham science from Wakefield’s father. The sense of entitlement beggars belief. Much the same from Johnson’s father of course, who seemed to think people weren’t entitled to be critical because they couldn’t even spell Pinnochio – indeed perhaps I can’t.
    But I’m afraid we are beginning to discover that these people just aren’t very bright which is why people are beginning to call Johnson and his advisor Dim and Dom.
    Which would be much funnier if so many people were not losing their lives. And now they are reducing lockdown with insane limitless travel measures when the R is, broadly, one.
    I fear it can now only go up.

    1. Samuel Johnson -

      Couldn’t agree more. It really speaks to the inequality in the UK and in the US that we have seen breathtaking entitlement from people who spout about their “good genes” and imagine that they are inherently superior to others and correspondingly deserving of their inherited privilege. They notably include Trump, whom the New York Times revealed, in its investigation of the family business, was a made millionaire as a child, accumulating $200,000 a year in today’s money before he was out of nappies.

      The most eloquently scathing writer who has skewered the egotistical arrogance of the successful in business who automatically conflate their success exclusively with their talent, instead of crediting exogenous factors that can include luck, mistakes by competitors, accidents of timing etc., is Nicholas Taleb. Ironically, perhaps, he is a monstrous egotist himself, nevertheless his demolition jobs in “Fooled by Randomness” and other books are both worthwhile and entertaining–because he is both brilliant and vicious, as well as convincing.

      Trump, Cummings and Johnson all appear to have had an early conviction that rules are for other other people (Trump’s placement in a remedial school disguised as a military academy because of his challenging behaviour as a child; Johnson’s well known earliest school report; Cummings’ undergraduate behaviour, and, notoriously, that the truth is whatever they wish it to be.

      Such people exist in every society. However, the extent of the passive social toleration of both vast inequality; indulgence of vulgar, unfounded conceits of innate superiority; as well as the open gaming of society to sustain immense hereditary privilege, seems simply feudal. Where else are private schools established as tax-deductible charities, at the expense of society as a whole? Good genes or a society rigged for the advantage of privileged brats? It’s a tough call.

    2. Simon Nicol -

      Oddly, the video of Wakefield spouting his entitled wisdom “cannot be played” at the time of writing (June 5th). I think I may be allowed to catch the drift, as it were, without seeing it and still not be labelled gullible.

      1. Sean Danaher -

        It very much emphasised that Wakefield very much believes in the upper class having superior genetics to the lower orders and should intermarry between themselves.

  2. Charles Adams -

    Thanks Sean, The z score is the number of standard deviations away from the mean, so if for some reason the standard deviation in the death rate is particularly small then we might see a much higher Covid z score. I wonder if there is any correlation between the standard deviation and other parameters, e.g. total population or median age, obesity smoking rates, etc? Tricky business statistics!

    1. Samuel Johsnon -

      There are known correlations with factors such as diabetes, obesity (correlated in turn with income and education), age, gender, social factors relating to how people are cared for or not in their old age. Nevertheless, it’s abundantly clear that what really matters when it comes to death rates is the quality of the governance of different societies.

      The experience of places like Kerala have shown clearly that mitigating the impact of the virus is not simply a question of resources, having hospitals, ventilators and trained staff.

      Contrasts are very unlikely to involve genetics at a population level. Higher death rates among BAME communities is likely a function of employment related exposure to risk.

      The US illustrates the point. It should have been uniquely well prepared and resourced and yet it has experienced a catastrophe that is far from over. The reasons aren’t complicated and are well documented (in particular in The Atlantic); examples

      The differences made by factors such as smoking surely cannot explain the difference between the UK and Germany or the UK and Ireland. As Sean has documented the UK, as a life science hub in Europe, with a strong public health tradition, would be expected to do much better than it has done. No doubt there’ll be a public inquiry eventually, but I’ll be very surprised if the conclusions don’t point to governance and political choices as being key determinants of the UK’s comparative high mortality.

    2. Sean Danaher -

      Agree. Tricky business indeed. The starkest ratio I think is the Eng/IE one which at over 20 is incredible in both senses of the word. I have no doubt at all IE is doing much better than England but by a factor of 20?

      Having a very smart GP as Taoiseach who is gay with an Indian father and has achieved his position entirely through merit is very different to the UK. Johnson is a mediocrity only there because of birth, Eton, Oxford, the English elite looking after its own and the cap tipping deference afforded by many of the English working class.

      Leadership matters.

  3. Charles Adams -

    Re. Privatisation. On the ability of outsourcing and privatised enterprises to mess up, the Horizon IT scandel is worth studying. Also featured in a current Radio 4 documentary.

    1. Sean Danaher -

      Thanks, Charles
      I don’t have time to look tonight but I suspect part of the disastrous UK response was an overreliance on private firms, TTI being a prime example.

  4. Boris Barbour -

    As mentioned on Twitter, using the z-score is confusing and misleading. The z-score will automatically be larger for countries or regions with larger populations, even if the per capita mortality is the same. In the present case, use of z-scores exaggerates the already awful UK performance.

    If you want per capita excess mortality, then show that. All of the z-score stuff is totally uninterpretable for the reader.

    This description is hopelessly wrong and confusing:

    “… the z-score is the best metric for international comparison, but it is overall deaths throughout the duration of the pandemic, rather than the peak, which is more important. This can be captured by measuring the area under the curve in z-score weeks. This area, however, will be proportional to the average z-score value, which is probably an easier number to understand, and that is what is presented here.”

    1. Sean Danaher -

      Thanks, Boris

      I think I would put it rather differently, but I think your criticism is valid none the less.

      Charles Adams has already commented that statistics is a tricky business and there has been some discussion as to other parameters that will skew the data.

      Standard measurement theory says that in the presence of a Gaussian White Noise background the quality of your measurement increases by the square root of the number of samples (N). Your measurement increases with N but noise by the square root of N. Your signal to noise ratio increases with the square root of N.

      In this context, smaller populations will likely have larger standard deviations as the sample size is lower and random fluctuations more important. As England has a large population in the European context, all other things being equal, it will have a more stable death rate and in general a lower standard deviation than smaller countries.

      One of the starkest z-score differences is the Eng, IE one at 19.89 and 0.84, respectively. This gives a ratio of 23.68. Normalising by the square root of the population ratio this reduces by sqrt(56/4.9) to around 7.

  5. Pingback: Is it really time to re-open England?
  6. David Jackson -


    Thanks for this article which I initially saw on the Tax Research site.

    1. The issue about the uncompetitive award of recent contracts to private sector firms in which friends and associates of Cummings have an interest. There is a real risk that the pandemic will provide a clear opportunity for these firms to become entrenched for the longer term in the provision of services on behalf of HMG. For example in a contract awarded uncompetitively for 5 years, what is the likelihood that a new contract will be awarded to another firm when it is retendered?

    What alternative provider is going to be equipped with the scale, expertise and willingness to undertake a procurement process with probably little chance of success.

    Quite apart from the fact that due to a variety of factors there is no certainty that it will go out to competition 4 to 5 years from now.

    2. Coronavirus on the island of Ireland and reasons for differences in infection rates North/South. I came across the high infection rate in County Cavan a few weeks ago and Irish Ministers & Officials were quick to discount proximity to the North as being a siginificant factor. Also, the areas of Northern Ireland that are contiguous to Cavan have relatively low infection rates compared to Greater Belfast.

    1. Sean Danaher -

      Thanks David

      I worry too about contracts. The pandemic has a long way to run yet in the UK I fear and I do not doubt that many of the shady Tory backers are disaster capitalists who will use the mayhem and suffering to enrich themselves. I fear much worse is to come.

      The Cavan figures are interesting, and you may very well be correct. Monaghan and Louth are also border counties and closer to greater Belfast and other higher infection rate areas.

      I’ve seen an Rz table and the high areas all seem to be towards the east of NI. The Causeway Coast and Glens seems to be the area with the highest Rz in the UK.
      The high Rz tables which seem to indicate the more Unionist the area is, the higher the infection rate.

    2. Samuel Johnson -

      The reasons for Cavan having an apparently anomalous rate of infection compared to other counties, which show a clear link with distance from concentrated urban populations, cannot be known for certain. It’s not inconceivable that Irish ministers could have political reasons for discounting proximity to NI as a reason, given the difficulties experienced on both sides with policy coordination. We simply don’t know. And the far lower levels of testing in NI don’t help elucidate matters. There have, not unexpectedly been reciprocal concerns and assurances in NI (see e.g.,–1919600/).

      However, the anomaly draws attention to the undesirability of having two different public health regimes and policies in operation on a small island. This has been graphically illustrated with tweets of satellite images showing schools a few miles apart on either side of the border with annotations indicating whether they were open or closed and the very different durations of recommended self-isolation periods (7 days in NI, 14 in Ireland). In addition, it’s been noted that the limits on distance traveled by citizens of either jurisdiction having crossed the border into the other have not been enforceable. The island has common sanitary and phytosanitary regulations, a common travel area, shared cross-border health services permitting sick people to be treated in the nearest hospital. The logic of a single island-wide approach has been noted by numerous public health experts including Prof Gabriel Scally, but as long NI unionists see anything like this is an infringement on their Britishness (they don’t mind their cows being Irish beef) difficulties will remain.

      It may be worth noting that the concerns are not all unidirectional. Earlier this month photos of a crowded Aer Lingus flight to London, with few if any passengers wearing masks, trended on Twitter, generating a lot of outraged comments about Covid-19 being flown into Britain from abroad, the lack of social distancing etc. The story was also reported more widely (see e.g., Many commentators failed to notice that the flight originated in Belfast, part of the UK, or to consider that in the absence of or reduced services from other airlines operating from Belfast people in NI have little choice but to fly Aer Lingus. Not unexpectedly some NI politicians were exercised about crowding on a “foreign” airline over which they have no control. I say “foreign” as Aer Lingus is a privately owned company in the same group as British Airways and is simply an Irish brand operating from the island of Ireland.

      The first confirmed case of Covid-19 in NI involved someone who had arrived on the island via Dublin and then traveled north. The number of international airports on the island that will remain viable in the near future is unclear. A year ago Dublin airport passed 30m on an annualised basis, compared to 2.4m for Belfast (in 2019), which has now lost the airline that carried most of them: Flybe. Whatever about Cavan and the land border, it’s not hard to imagine that the East-West common travel area will be a focus of some concerns and, possibly some new controls at Irish ports, in the event of a significant 2nd wave of Covid-19 in Britain.

Comments are closed.