Show Hide image

The Covid modeller

Why Neil Ferguson of Imperial College London – the so-called Professor Lockdown – believes he has been vindicated and his team’s work has saved many lives during the coronavirus pandemic. But he still has serious regrets.

Sign Up

Get the New Statesman’s Morning Call email.

On 16 March, Neil Ferguson, a leading epidemiologist at Imperial College London and then a member of the Scientific Advisory Group for Emergencies (Sage), published a now much-contested report that is widely believed to have panicked the government into changing its coronavirus strategy. The response to the report hastened the United Kingdom into lockdown and the economy into recession, and it would also change Ferguson’s life in ways he’s still struggling to understand.

Before Ferguson’s intervention, which subsequently made him a global public figure of hate on the libertarian right and earned him the sobriquet “Professor Lockdown”, the government had sought to slow or mitigate the spread of coronavirus rather than suppress it, as had happened in China and was happening in Italy and elsewhere in Europe. In this early phase of the pandemic, President Macron had described the Johnson government’s approach as one of “benign neglect”.

From the middle of January, Ferguson’s infectious disease modelling team at Imperial had warned about the lethal threat posed by the discovery and spread of a novel coronavirus related to Severe Acute Respiratory Syndrome (Sars) in Wuhan, China. Ferguson’s initial modelling was based on flu transmission but he refined it as more precise data became available, notably to take account of the catastrophe unfolding in Lombardy, where local health systems collapsed because of the effects of the pandemic. By mid-March his co-authored report stated that if unchecked the virus could kill more than 500,000 people in the United Kingdom and more than 2 million in the United States. Even with mitigation, the Imperial report said, 250,000 could die in the UK and up to 1.2 million in the US. “Suppression is the only viable strategy at the current time,” the report concluded. The media seized on it and the headlines were lurid.  

The British government responded by abruptly changing course, from herd immunity to lockdown. On the evening of Monday 23 March, in a televised address, Boris Johnson, an instinctive libertarian, declared a national emergency, having just weeks earlier joked about shaking hands with coronavirus patients. We had entered lockdown and, in the days that followed, a strange, surreal silence fell upon the land as we braced ourselves for what seemed like the terror of mass death and sickness.

But should we have eschewed traditional public health methods in preference for policy founded on the predictions of models that are exquisitely sensitive to even small errors in what is inevitably imprecise input data? This is the fundamental question and it is central to understanding the controversy surrounding Professor Ferguson and his team at Imperial.   

Reflecting now on the 16 March report and its consequences, Ferguson told me when we spoke via Zoom on 27 July that the figure of 500,000 deaths was “never really going to happen”. But the “number that really did influence things” was that “if we did everything short of a lockdown, we would end up with 250,000 deaths, and with a level of NHS demand which would be unsustainable. Those are the numbers which really changed policy.”

His greatest regret now is not having been more emphatic and less equivocal in Sage meetings in February and early March. “I could have talked as I’m talking to you, rather than saying, ‘These are the policy options, these are the pluses and minuses.’ I never said, ‘And in my view we should do this.’ I wonder if I should have done. And we got close to doing it in the 16 March report… It’s not a moral judgement: it’s a functional judgement. If you want a functioning health system, this is the only policy option open to you.”

Later in our conversation, he said: “Pandemics are not like hurricanes. You don’t hunker down, weather the storm, and then everything goes back to normal. It’s a dynamical system, it’s a virus spreading in the human population: what happens to it depends on how we behave, how we interact.”

He warns against complacency: people should understand that we are still in the “early stages” of the pandemic, he said. The virus was seeded later in the the global south and populations there are younger, and yet it is spreading alarmingly fast not just in the US, Brazil and Mexico, but especially in India and sub-Saharan Africa. In some countries – notably India, Nigeria and South Africa – mortality rates are believed to have been radically under-reported or, perhaps more accurately, under-ascertained.

In the US, where the official death toll from Covid-19 has reached 150,000, Ferguson told me the mortality rate could double by the end of the year - but added that the Trump administration had “not completely” lost control. “If you look at the rate of increase of cases now, it’s a lot lower than it was in March, but yes, the R [reproduction number] is above 1 in most US states. But they’re not in the two million [deaths] scenario: they have controls in place.”

He paused, as if carefully considering his next sentence. “We said that if they just went for a mitigation scenario they could be facing a million deaths or so. But they could be facing the same number of deaths again for the rest of this year as they’ve seen so far this year. What we’ve learned of this virus is that health systems cannot cope with a surge, so even the most Trumpian right-wing nut is going to end up having to put in some controls to stop that happening. They can’t tolerate having all the hospitals full.”

In Europe cases are rising again, notably among young adults – Boris Johnson spoke this week of a “second wave” breaking in certain countries. Citing Imperial’s serological survey, which aligns with antibody data from the Office of National Statistics, Ferguson estimates that only 7 per cent of the population of England has been infected by the virus; the figure is higher in London, around 15 per cent. Cases are decreasing but “decreasing less slowly” and we are seeing localised spikes with worse to come. “It’s going to be more difficult moving into the autumn. We’re not going to get back to anything close to normal life for the foreseeable future – not until we have a vaccine.”

He is particularly anxious about the return of schools in September. “Primary schools are less of a risk for reasons we don’t fully understand. But teenagers – my best estimate is that they behave a lot like adults in terms of transmission, even if they don’t really get very sick. And that poses a real risk. High schools are large places… and they connect a lot of households together, because society is a social network. All our modelling suggests that this will lead to an increase in transmission and the [reproduction rate] going above R-1.”

By early October, he expects to see a steady increase in “infection incidents”, as he put it, not least because most respiratory viruses transmit more effectively in late autumn and winter. “We will have to roll back on forms of contact; reverse some of the relaxation. Either that, or we’ll get better at understanding which types of contact pose risks.”

He does not expect another nationwide lockdown but nor is he confident that, without some further curtailment of our liberties, the NHS can avoid being plunged into a winter crisis. “We don’t currently know whether we can both prevent the NHS being overloaded and not go back into something closer to the lockdown. We may have to go back to the situation we were in in May and early June.”

But here’s something to consider: will continuing social distancing, self-quarantining of people with viral symptoms of any sort, and widespread mask wearing - plus the enhanced seasonal flu vaccination campaign - actually protect the NHS from a winter crisis better than it has been protected before, in spite of Covid-19? Could it be that our anti-Covid measures might choke off so many of the other viruses that contribute to the standard NHS winter crisis? We will know before too long.

***

Somewhat surprising for someone who has had such a transformative effect on government policy, Ferguson reveals that he has never spoken to the Prime Minister. “I never met Johnson, I never had one-on-ones, it’s not the way science advice works in the UK. It all goes through Sage and then the scientific consensus is communicated to Johnson by Chris Whitty and Patrick Vallance.”

This was a fascinating revelation - because there is no such thing as a scientific “consensus” around something like Covid-19, and our system seems to have given rise to the idea that there is. 

Why didn’t Johnson seek a direct opinion from Ferguson and his team? “We did have some conversations on the margins of Sage meetings with No 10 special advisers, but that was the closest I got to any politician.”

In early March, Ferguson showed his modelling results to a colleague in France, Simon Cauchemez from the Institut Pasteur, who, in turn, showed them directly to President Macron. It was, Ferguson said, “one of the three pieces of evidence that led Macron to locking down France, apparently”.

I asked him what Britain got wrong in March as the virus was spreading exponentially – or, as the Sunday Times Insight team has put it, as the country was “sleepwalking” towards disaster. “There’s some interesting reflection to be had on whether having a very nuanced, reflective, risk-adverse scientific advisory system going into policy gives you the most agile system,” he said. “In other European countries, like France, a small group of scientists – experts on the disease - directly talked to the politicians. Here, we have this interface layer, which does stop charismatic individuals unduly influencing policy.”

But surely we want these charismatic individuals arguing their positions in the presence of key decision-makers, who then have to choose what seems to them to be the emerging “best” approach. The politicians decide, after all – or they should. In any event, isn’t Ferguson an example of one such charismatic individual who directly influenced or subverted policy? He became a ubiquitous presence in the media, making his case. “Exactly!” he replied. “But there are plenty of people in the country who would say that I’ve already influenced policy unduly. What I mean is that I don’t have a clear view on what the perfect system is.”

Ferguson has discussed the government’s slow (others would say negligent) response in March with former Sage colleagues. “Paradoxically the government probably had a more nuanced and detailed understanding of what we knew and what we didn’t about the virus than any other government – I’ve dealt with a lot of governments in this crisis - and yet we were too slow, to be blunt.”

The fatal mistake, he believes, was failing “to get surveillance in place fast enough” as the virus was spreading in hospitals, in care homes and in the community. By surveillance, he means widespread testing for Covid-19.

The initial focus in the UK was on keeping cases out of the country: ramping up border testing and border controls. “But these measures aren’t terribly effective,” Ferguson says now. “It gives a false sense of security. Because there was such limited testing, the testing was very focused on travellers coming in. What we should have been doing – I and John Edmunds [of the London School of Hygiene & Tropical Medicine] banged on about this – is starting to get representative surveillance within the community, particularly targeting hospitals, but also GP practices, to get a sense of how much infection had slipped through the net and had got into the country.”

You could turn up at a hospital in late February with viral pneumonia and you would not be tested for Covid-19. “What we know now is there was a lot of Covid in the country. People were going into hospitals and dying at the time with Covid who never got tested.”

Testing should have been scaled up and the right people should have been tested, he told me. Were the wrong people tested? “It’s about the use of tests. And we were using them all on testing people coming into the country… In early March, France, Spain and Italy had the highest infection rates, particularly around their major cities, and that had seeded a lot of infection in London and around the country. Had we just turned on that surveillance a week or two earlier, we would have recognised that and would, I suspect, have acted much earlier.”

I mentioned Ferguson’s explanation about what he believed had gone wrong to Dr Phil Whitaker, a GP and the New Statesman’s health columnist. “I agree we got it wrong focusing on travel from Wuhan and few high-risk locations, and we were blind to the seeding from Europe, but was Ferguson really strongly arguing for massive TTI [test, trace, isolate]? He might have been, but I just didn't get that feel at the time. And where in the Sage process were the other voices that would have spoken emphatically in favour of that ‘traditional’ public health approach? I agree it was bonkers that people with viral pneumonia in late February didn’t qualify for testing unless they'd been to a few specific countries or had contact with known Covid-19 case - madness.”

***

Neil Ferguson, who was born in 1968, has tightly cropped, receding hair, dark-framed spectacles and mournful eyes. When we spoke, he was wearing a low-necked T-shirt and casual, unzipped hooded top. He grew up in Wales – his father was an educational psychologist and his mother an Anglican priest – and studied physics at Oxford before he switched to mathematical biology after completing a PhD. He has been entangled in public controversy before – his modelling work on both swine flu in 2009 and, before that, the foot-and-mouth outbreak in 2001, was described by some opponents as “overzealous” (one of his models predicted that in the worst-case scenario 65,000 people in the UK could die of swine flu — the actual figure was no greater than 500). One Devon vet called the contiguous cull of cattle and sheep in 2001, in which as many as six million were killed, “carnage by computer”.

But nothing had prepared Ferguson for becoming one of the public faces of the lockdown. He has been viciously denounced and abused, particularly by the libertarian right in the US, which, like Donald Trump, casually dismissed the lethality of the virus and agitated to keep society open. They are rather less belligerent now that even Trump has worn a face mask. The reliability of the Imperial modelling has been repeatedly challenged as well – both the assumptions that inform it and the opaque codes used - notably in a striking exchange with the Conservative peer Matt Ridley during House of Lords select committee on science and technology, to which Ferguson gave evidence. 

I asked Ferguson about how he felt about the abuse. “From March onwards, I’d been getting nice emails from people, but I was also receiving – and young colleagues of mine, too, which is very distressing, but particularly myself - dozens of really quite unpleasant emails a day, accusing me of destroying Western society, never mind the bizarre conspiracies around George Soros, Bill Gates and myself. Not many, but some death threats. And you develop a thick skin, but it still has some impact personally. I’ll give you an indication: Imperial has had to put in place very specific cybersecurity around my email accounts, all our servers. At one point, my email account was being spammed with over 10 million emails a day.”

He smiled sadly and looked directly at me. “So, yes, I’m smiling about it, because you kind of laugh or cry, really. Funnily enough, that kind of died down with the May thing” – we will come to that later - “to some extent, and it’s dying down more now. As we’re seeing a resurgence of transmission in the United States, a lot of the scepticism – not all of it, but a lot of the scepticism really originated on the alt-right in parts of the US. As they see case numbers and mortality rise again, it’s becoming harder and harder – I mean, they’re not stopping, but they’re becoming more marginalised. You can see that just in terms of even the president’s recent statements”.

He believes the hatred directed at him - as well as at colleagues and his friend the virologist Christian Drosten in Germany (“There’s a sceptic movement in Germany as well”) - is motivated by a refusal to accept the role of chance and randomness in our lives. “I kind of understand it in a very academic, psychological sense, but it has been disappointing and distressing. I understand the desire for people to explain, not really accept that just bad things happen by chance: this is a chance event, the emergence of a lethal pandemic virus, and it is very bad luck for the world, but some people do not like having their lives governed by chance, they much prefer to see some sort of dark hand behind everything, hence the popularity of conspiracy theories of all types.”

***

Now to the “May thing”. On 5 May Neil Ferguson, who by this time had succumbed to but recovered from coronavirus after self-isolating for two weeks, resigned from Sage after the Daily Telegraph revealed that his girlfriend (who is in a polyamorous marriage) had at least twice visited him at his London flat, thus breaking the rules of lockdown. Ferguson was accused of hypocrisy and, on the evening of his resignation, was shamed and vilified on social media, with the American billionaire Elon Musk tweeting: “This guy [Ferguson] has caused massive strife to the world with his absurdly fake ‘science’.”

It must have been a profoundly painful experience. Ferguson remains married, he and his wife, though not in a “traditional relationship”, as he puts it, are “close friends”, and they have a son. He does not want his wife or partner to be damaged by what he considers to be an error of judgement. “I look back and think, ‘How could I have been so naive?’”

What happened to him, he told me, “talks to the kind of compulsion versus voluntary measures. I have always been somewhat sceptical myself and made risk-based decisions in terms – not always accurate ones, I should say – in terms of living my own life, and I suppose I carried on, which was the wrong thing to do. I wasn’t aware quite how public a figure I’d become.”

But you were Professor Lockdown?

“Yeah, I hated it! I also got kind of bizarrely categorised as being almost a semi-official government spokesman. If you go back to the Sage meetings, I was never sceptical about lockdown, but I was always very, very conscious of the economic and social impact it would have, and the fact that it wasn’t a long-term solution: basically, if we adopted it, we would be in this position forever. I concluded reluctantly it was what we had to do, but it isn’t a perfect solution by any means. But once we adopted it, I suppose I felt a need to explain it to people, communicate why it was necessary.”

And yet, despite his prominence and the contested influence of the 16 March report, he did not self-identity as a public figure. “You have a professional side, you give interviews to Today via FaceTime in the morning, and it didn’t really impinge on my personal life in that strange cognitive disconnect between yourself as a personal individual with relationships – a complicated personal life – and my professional life. And the two came crashing together. I was stupid and naive, and I regret it. Not just regret it a lot personally, because as well as myself it was very traumatic for my wife and son. They had the tabloids camped outside for nearly a week. But the experience does make you humble, though I’ve never tried to portray myself as a perfect individual, but you do learn lessons quite quickly.”

Because of these experiences, I wondered what he thought of the Dominic Cummings case: the dash by Johnson’s chief aide, when his wife had coronavirus symptoms, to his parents’ estate in the north-east; the trip to Barnard Castle on his wife’s birthday to test his eyesight; the impact it had on people’s attitude to lockdown. “I still don’t think what I did posed any risk, but – well, it was one rule for us and one rule for them, and it did smack of elitism. I did understand why people got angry. I understand how and why in both cases – mine and Cummings’ - the public responded. In my case, it was my own subjective, probably invalid, judgement of risk, but I never denied that I didn’t break the lockdown rules. In Cummings’ case, he made what I thought was a slightly torturous case that somehow he didn’t break the lockdown rules, which I don’t think was shared by many other people. I don’t think the two cases were comparable, because I stepped down, I accepted responsibility quite quickly.” The subtext is clear: he resigned; Cummings did not.

***

The pandemic is changing society – the way we live, work, consume, socialise, travel, communicate – in ways we never imagined possible before it. We are facing an autumn of recession and the probable return of mass unemployment. For Ferguson personally, 2020 has been extraordinarily turbulent and exacting. He considers the pandemic to be the “big one”: a once-in-a-century shock to the world order, exposing our vulnerabilities, testing our national resilience, reminding us of the fragility of our global interconnections, and intensifying our geopolitical rivalries. “If we re-ran the 1918 pandemic today, it would probably cause no more deaths than we’re seeing from coronavirus – because the world has moved on, especially nutrition, better medical treatment. You can imagine worse things, but in some sense, this is the big one.”

(The Spanish flu pandemic caused an estimated 50 million deaths in three waves from 1918-20. But many of the deaths were caused not by the flu itself but from bacterial pneumonia that developed on top of it. Today most of these cases would be treated successfully by antibiotics.)

Ferguson expects the UK will have an effective and scalable vaccine by the middle of next year, perhaps as early as the spring. The Oxford project, which is seeking to produce a recombinant vaccine, has entered phase three trials. The genome of a different, attenuated virus - adenovirus – is edited so that when it is introduced, the body starts making the coronavirus spike protein, triggering an immune response. There are no human vaccines yet licensed using this technology, although it has been shown to be effective in animal studies. There is, for instance, an animal rabies vaccine produced commercially using the technology.

“I would give it probably a 50/50 chance that we will have demonstrated some degree of efficacy and safety of a vaccine candidate by mid-November to early December,” Ferguson says. “The benefit of the Oxford vaccine is that it builds on what’s a known vaccine platform; it’s a technology that has been used before. But there are dozens and dozens of vaccine candidates and developments. We will get a vaccine. It will not be a perfect vaccine, probably, but it will probably be good enough. Best case, the UK population will have some vaccine available to target high-risk people by February, March. More realistically, this time next year.”

Some good news, then. But other research scientists I know, such as the infection biologist Michael Barrett, a professor at Glasgow University, have been more sceptical about the prospects of finding a vaccine. But, Barrett says now, “I’m growing more confident of a vaccine that will at least diminish virulence if not sterilise against the virus.”

Suppose, then, we have one: will societies experience a kind of post-traumatic growth as we emerge from the crisis? Will we live and work better after the pandemic? Ferguson is encouraged that the crisis has seemingly provoked greater awareness of the global environmental emergency, but doubts persist. “It is depressing how, at least in the political and economic sense, this pandemic has seen countries become more insular and less cooperative,” he said.

***

What will Neil Ferguson do next? In August he intends to take some holiday and, beyond this, he is attracted by policymaking - not giving up academic research entirely, but broadening his portfolio. “I’d never want to enter politics, but sort of advisory positions within either government or international organisations. I’m not sure whether this year has improved or reduced my chances of doing that!” He smiled again.

He seemed in good spirits during our long conversation, so I asked about his morale. “Generally, up and down. A lot of people in our team are feeling burned out right now, and May, from that point of view, did set me back, because I’d been working 16-, 18-hour days. Now I work more like ten-hour days – partly because of that [“the May thing”] knocking one back psychologically, but also partly just because it’s not sustainable long-term to work at that level. We’re trying to move towards a more sustainable work stream. Because this isn’t going away: the demands on us, we’re working with dozens of countries around the world, trying to support many more through work with the WHO and the like. We need to do this from a kind of long-term perspective, at least another year or more. But I would say I’m in better spirits now than I was, say, six to eight weeks ago – it still can be challenging.

Despite the doubts expressed about the accuracy of his modelling, Ferguson takes pride in the work of his team – and believes that it has been vindicated. “Overall, we haven’t been perfect and we’ve made mistakes, personal and scientific, but I think we have made a difference globally. I’m proud of what the team has done, and I should emphasise, it’s not just me; it really is a team effort, we’ve had up to 50 of us working on this at Imperial. We alerted the world to the scale of the potential problem earlier than any other group, we came up with the first estimates of severity, some of the first estimates of transmissibility, and I think those have proven to be roughly right – not perfect, not exact estimates. I think our early assessment gave the world the warning it needed. Yes, not everybody listened. There was a lot of uncertainty and people can be justifiably sceptical, but I would stand by the great majority of – no, I stand by all the work we’ve done, and we’ve got a more nuanced understanding. Some of the precise numbers have changed, but basically I think we got it right.”

Perhaps he would agree with the aphorism, attributed to the British statistician George Box, “All models are wrong but some are useful.”

Usefully wrong or usefully right, in his case?

***

For Dr Phil Whitaker, who has been working on the front line treating patients with the virus, lockdown was the correct policy even if it was imposed too late. “As it is, we have somewhere north of 50,000 fatalities, and tens if not hundreds of thousands with ongoing life-altering disabilities or illness,” he told me.

But Whitaker is also sceptical of Ferguson’s modelling. “Where I think the debate on accuracy is important is that, in late March, based on Ferguson’s modelling, we were being shown various scenarios of the future few weeks based on the degree to which the population complied with lockdown measures. In every lockdown scenario the NHS’s breakneck-speed expanded ICU capacity was going to be massively outstripped - it was merely a matter of how many fold the capacity was going to be outstripped. So, either the modelling overestimated the size of the surge, or underestimated the efficacy of lockdown measures, or both.”

I asked why this mattered if the right policy was implemented. He said it mattered because it created the “chilling prospect” that innumerable sick patients would not receive the life-support they needed, as had happened in northern Italy. “That led to the sheer panic that in turn lead to the Nightingale effort to build new hospitals. This was not a bad thing but - this is really important – the panic led to the policy of evacuating medically fit inpatients at top speed, with no regard to Covid-19 positivity, into nursing homes. The balance of risk from not doing that mass evacuation looked huge - even fewer acute hospital beds available for the predicted hordes of dying patients. But as the reality unfolded, the modelling had vastly over-predicted numbers; the policy looked increasingly negligent.”

For his part, Neil Ferguson is convinced that he and his team at Imperial saved many lives. “From a global perspective, yes, we did make a difference in that regard [saving lives]. But from a UK perspective, I wonder if I could have done more, whether I could have pushed harder for testing… I wonder if I’d adopted a more subjective view, whether things might have happened faster. But that maybe shows a lack of humility on my part!”

Jason Cowley is editor of the New Statesman. He has been the editor of Granta, a senior editor at the Observer and a staff writer at the Times.

This article appears in the 14 August 2020 issue of the New Statesman, This house must fall