Article sponsored by

Covid-Delta, Science, and the Problem of Known Unknowns – Analysis by Keith Rankin.

It’s a known known that the late Donald Rumsfeld’s principal legacy to the world is the following quote:

“As we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don’t know we don’t know.”

I might say that there are also unknown knowns, nuggets of truth buried in archives and barely read books and articles.

When it comes to known unknowns, there are two basic types. The first are questions posed for which we do not yet have any plausible answers. This could be due to technical (eg measurement) difficulties, or economic difficulties (the expected cost of finding answers being too high). The other type is because at least some people don’t want the questions answered (and may even place embargos on finding answers), or because the possible answers simply do not fit the prevailing filter through which the questions are framed in public discourse.

Re unknown unknowns there are also multiple categories. First there are cases where questions have never had reason to be posed, cases beyond the prevailing human imagination. These are genuine unknown unknowns; things that are true but that none of us had the capacity to imagine. I can offer no present examples. A past example is the electric light bulb, which could not have been imagined even by Julius Caesar. (Although the light bulb could have been imagined well before it happened; for example, after the scientific work of Michael Faraday.)

Then there are questions of wilful blindness, questions which verge on the final category of known unknowns. Third are the black swans, events that happen ‘out of the blue’ or ‘out of left-field’; but which were predictable ‘with hindsight’! Black swans represent something akin to wilful blindness; they are events that were genuinely unexpected, and for which the precise form of their manifestation could never have been predicted. Covid19 is a black swan. (Note that ‘black swan’ is an unfortunately ironic term. It was either first coined – as a metaphor – by someone who did not know that black swan birds exist, and are the normal type of swan in some parts of the world, including ours. Or it was coined because the ‘coiner’ believed that, for almost everybody except themself, actual black swans were unknown. In reality, black swans are not black swans, and they have been known in Europe since the seventeenth century. Nor are the black-white swans I saw on the island of Chiloe; I was not expecting to see them, but one reason people travel is to discover things that other people already know.)

(That the Taliban would rule over the whole of Afghanistan as soon as July 2021 was a black swan as recently as June 2021. It was not conceived of as even a possibility, except – maybe – by the Taliban themselves.)

Related to these unknowns are the unseen – or ignored – red flags. It is missed red flags that create black swans, such as Covid19. And – I would argue – our failure to adopt joined-up-scientific thinking re the known facts about Covid19 and related viral diseases, means that there are more black swans lurking. The most important red flag that I would mention in this case, is the present apparent loss of natural immunity to respiratory viruses; loss of what I will call ‘community respiratory viral immunity’ (CRVI). CRVI is not a binary concept; it’s a level of community immunity that, like the economy, has for the most part grown over time. Natural variation of immunity rises (or diminishes) through the changes in the pathogenic environment, rather than through interventions such as vaccination. (The economic growth analogy of a temporary loss of community immunity is a recession.)

The problem in New Zealand at present is epitomised by the way we report about ‘Delta’, a highly transmissible variant of the SARS-COV2 virus that causes Covid19; this ‘comms’ problem is perpetrated in particular by our technocrats, our bureaucrats, our ‘policrats’ (narrative-framing politicians), and our mediacrats.

Before discussing this further, I need to emphasise that New Zealand’s present ‘Level 4 lockdown’ is absolutely the correct emergency policy measure for the present outbreak in New Zealand of Covid19.

The Delta-Bogey and the Missing Science and CW2 (Covid-War 2)

The dominant narrative in Aotearoa New Zealand is that Covid19 – which we (the New Zealand ‘team of five million’) defeated in battle in 2020 – has morphed, like some demon – into Delta. And that Delta is a seriously mean beast. In creating Demon Delta, we implicitly treat its predecessors as comparatively harmless. Yet by far the majority of Covid19 deaths in the world have been caused by non-Delta variants. (If this latest outbreak had been identified as non-Delta, we should have been more – not less – alarmed; it would probably have come from South America, where Covid19 has most affected, and where Delta has been least present.) In this narrative, populations have four weapons at their disposal: macro barrier methods (lockdowns and quarantines), micro barrier methods (hygiene practices, including facemask wearing), contact tracing, and ‘silver bullet’ vaccinations. (By ‘silver bullet’ vaccinations, we mean that – after a course of vaccinations – a person may be classified as ‘immunised’; this is how we understand, for example, measles vaccinations.)

In this narrative, the implicit counterfactual is that the adverse consequences of a Delta outbreak are much greater than of an outbreak of the South American versions of Covid19, or of the original Wuhan version. While we have heard much about the greater transmissibility of Delta, I have heard of no scientific studies that compare Delta and non-Delta strains in fully comparable populations. (All scientific pharmaceutical trials require that drugs be tested alongside ‘control’ treatments.)

This predominant narrative may be called Hypothesis One.

There is a second obvious hypothesis (Hypothesis Two): that (i) Delta is substantially the same as previous versions of Covid19, though just enough more transmissible to displace other variants in circulation in the same environments (like grey squirrels displacing red squirrels), and (ii) from late 2020 – and especially in 2021 – populations have reduced CRVI (community respiratory viral immunity). (There is another possibility, an in-between hypothesis, that Delta is significantly more harmful than other variants, and that its impact is exacerbated through many current host populations having reduced CRVI.)

An extension of Hypothesis Two is that immunisation by vaccination may not be permanent. (It is a known known that measles immunisation is permanent, but that immunisation against influenza is temporary.)

In summary, Hypothesis One is that the major problem leading to ongoing mortality and morbidity is the more aggressive behaviour of the enemy (of Delta). Hypothesis Two is that the major problem leading to ongoing mortality and morbidity is the reduced CRVI of the population.

The counterfactual to the first hypothesis is that the Covid19 pandemic would be in its endgame, were it not for Delta. The counterfactual to the second hypothesis is that the outcome of New Zealand’s August 2021 outbreak of Covid19 (and recent Asian outbreaks) would be much the same – serious – whether or not this was the Delta strain. Both hypotheses predict that – without appropriate policy responses – there will be problematic levels of mortality and illness.

Policy Implications

Both hypotheses require policies of ’emergency lockdown’ and, if available, ‘vaccination’. (Fortunately, we do have available effective vaccinations which target the SARS-COV2 virus; had these vaccinations proven to be ineffective, then we would further emphasise ’emergency immunity management’ policies such as lockdowns.

There are two important policy differences, however, depending on which hypothesis is more true. The first policy difference relates to how populations should behave outside of periods of emergency lockdown. The second is about the ongoing vaccination programme.

In Hypothesis One, Delta is the problem, and success is the ‘elimination’ of SARS-COV2 (whereby SARS-COV2 goes to the same place that SARS-COV1 – in 2003 – went to), aided by the immunisation by vaccination of the population.

In Hypothesis Two, reduced CRVI is the problem, and success is a level of community immunity that would tolerate SARS-COV2 circulating in future as another seasonal ‘common cold’ coronavirus. And success means adopting practices – including regular vaccinations – that extend CRVI levels in the population. (The good news here is that regular Covid19 vaccinations should reduce illness from other endemic viruses by facilitating high CRVI levels. Good for labour productivity as well as for general wellness.)


What exactly is community respiratory viral immunity? It’s probably not quantifiable as a precise metric, but is a real-world parameter that rises or falls under different conditions; and it’s a community attribute that, ideally, should be optimised but not necessarily maximised.

The key idea is that it is a measure of general immunity to an important class of pathogenic diseases, and not immunity to a specific respiratory virus. And it should be understood as a population measure, rather than a measure of an individual person’s immunity.

CRVI increases with exposure to respiratory viruses in aggressive or attenuated form. It relates to what might be called the ‘common’ classes of community viruses: influenzas, coronaviruses, rhinoviruses, and other similar viruses such as respiratory syncytial virus (RSV).

Novel viruses (such as coronavirus SARS-COV2) can be classed as ‘aggressive’ (mainly because they are unknown to our immune systems), though some may be more aggressive than others (eg SARS-COV1 was more lethal than SARS-COV2). Attenuated viruses can be classed as those which have evolved to be less aggressive, forming equilibriums with populations with given levels of community immunity. And the label ‘attenuated’ can be used to describe the deactivated viral sequences that constitute the active ingredients of our vaccines. Vaccination against community viruses is a relatively recent episode in the wider history of vaccination; until 2020, only influenza vaccinations were in place for this viral class, and even they are comparatively recent (ie, in practice, influenza vaccinations are twentyfirst century interventions).

Over the history of humanity, CRVI has increased, and necessarily so. As more community viruses circulate within human populations (ie become endemic to humans) – typically viruses passed to humans from other species for which they were already endemic – then CRVI levels increase due to accumulated exposure to ever-greater-numbers of these virus species. Thus, in 2019, CRVI levels in human populations throughout the world were probably at the highest level that they had ever been in human history. Indeed, the main driver of rising CRVI levels in recent decades has been the decreasing cost and increasing convenience of international air travel. Another important driver has been the introduction of annual influenza vaccinations.

CRVI levels are generally higher in urban populations, and highest of all in the world’s metropolitan cities; cities which are both densely populated and within close reach to international airports. One of the most important unknown knowns (or at least ‘little known’ knowns) in this regard is the difference in community immunity levels between different homeplaces of young men called into the United States’ military in 1917, the year that a new H1N1 influenza virus started to circulate in the United States. The weaker – indeed ‘weedier’ – city boys proved to be significantly more resistant to the virus than the muscular young men from the farms and the provincial towns. (Refer to The Pandemic Century (2019), by Mark Honigsbaum.)

An important feature of CRVI is that it wanes when not fortified by ongoing exposure to community viruses. CRVI is nuanced, in that if fortified mainly by rhinoviruses in one year, then populations become a little more susceptible to serious illnesses from influenza viruses in the following year. Nevertheless, exposure to one class of community viruses probably gives some degree of resistance to other classes of community viruses.

So, under Hypothesis Two, in early 2020, global population CRVI levels were very high. The result was that Covid19 illness – caused by the then novel coronavirus SARS-COV2 – was resisted by the younger infected population, including the middle-aged-populations which represented the majority of airline passengers. Thus the major health consequences were faced by the older and comorbid populations who were less able to mount the requisite immunity responses.

However – and under Hypothesis Two – the important but not understood story of 2020 was the unusually rapid waning of CRVI levels in (now largely physically disconnected) human populations. This waning was a result of restrictive behaviours, mandated and unmandated. Restrictive behaviours include both mandated isolations, and personal barrier restrictions (such as physical distancing and the widespread use of facemasks).

Under emergency conditions, a loss of CRVI is the necessary price we must pay in order to minimise – if not eliminate – a dangerous pathogen. This elimination was achieved with SARS-COV1. The under-recognised challenge is to – as best as possible – start to restore CRVI levels as soon as emergency conditions are lifted (and, as part of this, to fully lift domestic emergency mandates as soon as a novel virus has been suppressed).

Part of the CRVI restoration process is of course vaccination, and it is probable that booster influenza vaccinations did to some extent increase our abilities to resist new outbreaks of Covid19. Of course vaccines that specifically target coronaviruses – and SARS-COV2 in particular – would have much more impact during a coronavirus pandemic; and the beneficial side-effect of coronavirus vaccines is that they most likely reduce populations’ susceptibility to the other community viruses that give us colds and influenzas.

It is now possible to talk of the ‘benefits of complacency’. ‘Complacent’ barrier behaviour – though not so much complacency towards contact tracing – helps to restore CRVI, and prepares populations for the next (or next wave of) community viral infection.

Hypothesis Two states that the major single factor in the severity of outbreaks of Covid19 since the middle of 2020 has been the loss of CRVI, and not the increased virulence of the evolving viral agents.

1917-19 Influenza Pandemic

It is worth digressing here to note the epidemiology of the H1N1 influenza pandemic of 1917-19; the pandemic best called the ‘black flu’, though more commonly (and inappropriately) called the ‘Spanish flu’.

This pandemic essentially hit the world in three waves, with the second wave being the most severe. In New Zealand the probable fatality rate was about 0.8% of the population, though ‘officially’ it was more like 0.4%. (We note that the present official fatality rate of Covid19 in the United States state of New Jersey has, just this month, surpassed 0.3%.)

The first wave of this pandemic appears to have begun in the United States, and its spread was almost certainly facilitated by the mobilisation of conscipted troops, as the United States entered World War One (WW1). However, another variant of H1N1 influenza had been emerging in China, and it seems that, on the Western Front, the two versions fused into a new super H1N1 variant, the ‘second-wave’ variant that was brought to New Zealand by returning soldiers. (The best source for New Zealand epidemiological information is the second edition of Geoffrey Rice’s book Black November. And we should note that evolution – of viruses as well as larger creatures – is about hybridisation [fusion] as well as through descendant mutation [fission].)

The earlier variant had however circulated in New Zealand in the late winter of 1918, with some severe health outcomes, but also raising the effective CRVI in those parts of New Zealand that were affected. When the big second wave hit in November 1918, two groups suffered least. First were those – such as Ngāti Porou – who implemented local quarantines. Second, were those in the places most affected by the first wave. After the short emergency period (essentially the month of November 1918) people reverted to normal life – or as near to normal life as possible in the month that WW1 ended. CRVI levels were clearly very high by New Year 1919, so when the slightly attenuated third wave hit in 1919, New Zealand was barely affected. Australia – which had imposed a full quarantine in November 1918 – suffered much worse in 1919 than New Zealand, though not as badly as New Zealand had done in 1918. Clearly, New Zealand had – for that time in history – very high levels of CRVI in 1919; ‘herd immunity’ to influenza, and most likely a higher than normal immunity to other community respiratory infections.

The ‘black flu’ pandemic was an event that featured both a more virulent muted version of the H1N1 influenza virus, and significantly varying levels of community immunity to respiratory viruses.

We note that in the present pandemic, both Hypothesis One (a very lethal variant of the virus) and Hypothesis Two (waning CRVI levels) may contribute to the story. In 2021, however, the Hypothesis One story (the ‘delta’ story) seems less convincing; I suspect, because the newer more aggressive variant is a descendant (fission) variant, not a fusion of two already aggressive variants (as was the case in 1918).

Hypotheses One and Two: the Evidence

To start with, we need to look at the big European second wave of Covid19, in the northern autumn of 2020. By then, there was increased knowledge of Covid19, improved testing and contact tracing, and renewed use of barrier interventions to viral circulation; all of these should have reduced the impact of the second wave if the virulence of the virus was the main determinant of the level of deaths and serious illnesses. But none of the new ‘more transmissible’ variants were present at that time; Covid19 was not more virulent then. (‘Alpha’ was the ‘Kent variant’ that surfaced in England in about December 2020.)

Instead, what happened was that, in late 2020 in most West European countries, the death rates were similar to those of the preceding spring wave. Spain was different; its fatality rates were significantly lower. Of most importance for this analysis was East Europe, within the European Union. There, where, in the spring, barrier methods had largely kept Covid19 out, fatality rates soared in the autumn to levels much higher than in West Europe. It was the same virus in both parts of the European Union. This picture negates Hypothesis One, and strongly supports Hypothesis Two. The major determinants of Covid19 death in Europe in late 2020 would have been varying CRVI levels, lowest in the east due to its successful earlier precautions, highest in Spain. Whereas summer complacency in the Czech Republic (where CRVI had become dangerously low) undoubtedly contributed to the problem, summer complacency in Spain most likely contributed to the solution, by boosting CRVI there. We also note that, for the most part, younger people were more likely to die from Covid19 in East Europe. This is consistent with lower CRVI levels there and then, rather than greater levels of complacency (unlikely) being the problem in East Europe. By September 2020, Covid19 was a known known, no longer a ‘black swan’.

In the Americas, throughout the pandemic, piecemeal barrier protection almost certainly reduced the peaks of the outbreaks, but also brought about depressed CRVI levels. We see that, in the United States, the timing of outbreaks in the ‘blue’ (Democratic) states (where barrier controls were most followed) and the ‘red’ (Republican) states (where barrier controls were most resisted). In general, the new outbreaks started in ‘blue’ states (with less CRVI), and eventually moved on to red states (with higher CRVI than blue states, but less CRVI than in 2019). In the very latest outbreak, though, the blue states were saved through higher CRVI arising from much higher vaccination rates; the present outbreak is accentuated in the red states.

Hypothesis Two predicted that, in 2021, Asia (which had imposed the most effective barriers in 2020) would be very vulnerable. That has come to pass. And – as in Indonesia today – the age profile of fatalities has been coming down, suggesting that levels of CRVI in Asia in 2021 are even lower than in Eastern Europe in late 2020 and early 2021. The tragedy of Indonesia is that even very young children are dying.

Further, in Asia in 2021, those countries unable to implement sufficient barrier protections (such as India), have seen short (but severe) outbreaks of Covid19, this time with the Delta variant of SARS-COV2 featuring as a circulating virus. An extreme case of this is Afghanistan, already in political turmoil when Covid19 hit in June this year. Briefly, Afghanistan in June – as Nepal in May – was amongst the worst affected countries in the world. But now, in August and with even greater political turmoil, Covid19 seems to have largely disappeared. It looks like Afghanistan has experienced a dramatic boost to its CRVI status.

The present outbreaks in Australia are proceeding differently from those of 2020. The popular narrative is that of Hypothesis One – that the people are now up against a more vicious foe, a devil called Delta. But the manner of the more lethal spread of Covid19 in the young population is more suggestive of low CRVI levels, as in East Europe late in 2020. The (counterfactual) control here is the United Kingdom, and West Europe. In these places CRVI has largely been restored (though, as in Israel, may be waning due to the earliness of its jabs). There, Delta has behaved more like a pussy cat than a devil, infectious but not lethal. In the United Kingdom, CRVI was largely restored by vaccination, but the removal of mandated barrier protections will be ensuring that vaccination-induced CRVI is being enhanced by renewed community circulation of seasonal (non-novel) respiratory viruses. Australia – especially young Australians – have substantially less CRVI protection from serious illness.

Eastern Europe is an interesting test case; it seems to have been immune from Delta so far. However, waning community immunity may see it vulnerable this coming northern winter.

Here in Aotearoa New Zealand, the rapidly imposed emergency measures should – after a few weeks – repel the current outbreak. The challenge will be for us to substantially restore CRVI levels, in time for the southern winter of 2022. While vaccinations this year – and 2022 booster vaccinations for the more vulnerable – will represent the main part of meeting this challenge, a high dose of Level 1 barrier-complacency this summer (but not QI-code complacency, for contact tracing) should help to keep the unvaccinated somewhat safe, the rest of us safe from them, and should help the vaccinated to hold on to raised CRVI levels through next autumn. It means that, once back to Level 1 (no community presence of Covid19), we should be encouraged to remove our masks – and to enjoy mixing and mingling – at least until another border-infringement outbreak occurs. And, when international travel is once again opened up, our priority should be to maintain – and extend – high levels of CRVI (community respiratory viral immunity). Low CRVI means lots of infections, many serious, and not all Delta Covid19.


The evidence, at least as I have presented it, comes closer to disproving Hypothesis One (the aggressive Delta hypothesis) than to disproving Hypothesis Two (the deficient CRVI hypothesis). I would like to see, in the media, a proper scrutiny of both hypotheses. Until this happens, the attention that community respiratory viral immunity requires will be negligently missed. The likely truth that is Hypothesis Two will remain a known unknown.

Barrier methods – macro and micro – work in emergency contexts, much as cortisol reduces stress and anxiety in these contexts. But, out of these acute situations, excess cortisol becomes a source of ill health. Barrier infection blocks all of the community viruses that support CRVI levels, making us over time more vulnerable to community infections, and making those infections more dangerous. On the other hand, annual vaccinations for influenza and coronavirus will substantially extend CRVI levels, making us generally more healthy with respect to both influenza and common cold viruses. In the United Kingdom and West Europe, Delta Covid19 shows all the signs of becoming – in a few years – another common cold coronavirus.


Keith Rankin, trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

contact: keith at rankin dot nz



64 − = 59

This site uses Akismet to reduce spam. Learn how your comment data is processed.