• Melinda Mills

The Human Cost of Natural Herd Immunity

Jennifer Beam Dowd1,2*, Per Block1,2, Malia Jones3, Melinda C. Mills1,2

1 Leverhulme Centre for Demographic Science, University of Oxford

2 Nuffield College, University of Oxford

3 Applied Population Laboratory, University of Wisconsin-Madison

*Corresponding author for questions & interviews: jennifer.dowd[at]

A spectre is haunting Europe—the spectre of herd immunity.

In response to the second wave of COVID-19, discussion of natural herd immunity as a strategy to manage the pandemic has emerged. The aim of natural herd immunity is to allow uncontrolled transmission to infect the ‘low risk’ population and to ‘shield the vulnerable’ [1]. While this approach may sound appealing for hastening a return to economic and social activity, to date proponents of the strategy have provided few details of the logistics of such ‘focused protection’ of the vulnerable or the expected morbidity and mortality toll needed to achieve this. To transparently and explicitly clarify the true costs of adopting the natural herd immunity approach, we calculate how this strategy would play out using a range of scenarios derived from current estimates of COVID-19 age-specific infection fatality rates (IFRs) and population estimates from the UK.

The stated goal of ‘shielding the vulnerable’ is to build population immunity amongst the remainder of the population by allowing them to be infected. Eventually enough people would become infected that the virus would run out of susceptible hosts and the spread of the disease would slow. This would ultimately protect those who are most vulnerable to death from COVID-19 by ensuring that those around them are already immune. Setting aside the impracticality of what it would take to isolate over 18% of the UK population over age 64 (as well as other vulnerable groups, all the members of their households, and their close contacts), COVID-19 mortality does not fall into neat categories of ‘vulnerable’ and ‘non-vulnerable.’ This means that we can expect substantial mortality even if the ‘shielding’ version of natural herd immunity were somehow possible and perfectly executed.

To assess the likely death toll of reaching natural herd immunity in the UK, we calculate the expected fatalities dependent on the proportion of population infected to reach herd immunity and ability to shield the vulnerable population over the age of 65 for three scenarios of optimistic, intermediate and pessimistic. We use current estimates of IFRs by age [2] and the population structure of the UK for our calculations. Results are displayed in Table 1 and visualized in Figure 1, across our three scenarios.

Figure 1. Total number of deaths by proportion of population infected for three levels of shielding success

The first optimistic scenario (blue line) simulates very successful “focused protection” of the vulnerable. Here we assume that the odds of infection (i.e., the probability of being infected versus the probability of no infection) for individuals aged 65+ is 50 times lower compared to those under 65 who roam freely. Assuming that 67% of the population will need to have immunity to COVID-19 to reach herd immunity, and immunity is unequally distributed between people under versus at least age 65, this translates to around 80% of those under 65 being infected (odds of 4:1), and ~7.4% of the over 65 population infected (odds 1:12.5). Given the current estimates of age-specific infection fatality ratios, this optimistic scenario would result in 134,900 deaths, of which approximately half would be above the age of 65 and slightly more than half under 65.

In the most pessimistic scenario (red line), we make more realistic assumptions about how the ‘shielded’ population interacts with personal caregivers and intergenerational households.[3] We assume that the infection rate of those over 65 would be 10 times lower than for those under 65, with further assumptions outlined in Table 1. This scenario would result in 458,200 deaths in the UK, of which 381,900 would be above 65 and 76,300 under the age of 65.

Our intermediate scenario (green line) assumes that the odds of infection for the elderly are 20:1 compared to the under 65 group, resulting in an estimated 288,700 deaths, of which 214,100 would be 65 or older, and 74,600 under 65.

Thus, even under our most optimistic scenario, we estimate that the natural herd immunity strategy will mean at least 135,000 deaths in the UK, and as many as 450,000 in more pessimistic scenarios.

If we are optimistic, adopting the natural herd immunity strategy would mean asking the public to sacrifice substantially more fellow humans than the number of British civilians killed during World War II. In the most pessimistic scenario, we would be asking them to surrender 458,200, slightly more than the combined number of British military and civilian casualties during World War II.

Given estimates of a roughly 6% infection rate in England and Wales during the first wave, the vast majority of these deaths have not yet been realized [4].

In addition to deaths, we must also account for the cost of achieving herd immunity through natural infection in terms of hospitalizations, ICU admissions, and unknowns about long-term effects of infection on survivors. The number of hospitalizations and ICU admissions would be many multiples of the number of COVID-19 deaths. Recent modeling shows that even in the best-case ‘shielding’ scenario, the pace of hospitalizations in the ‘non-vulnerable’ would quickly overwhelm the NHS and require renewed social distancing measures to carefully balance the pace of infections with hospital capacity [5]. Without careful control of the pace of new infections, our hospitals would become unable to treat all COVID-19 patients, leading to sharp increases in the infection fatality ratios, and many more deaths. In addition, our hospitals do routinely provide other life-saving treatments, which would become unavailable if they became overwhelmed, leading to increased indirect deaths from lack of treatment for other conditions. This is far from the clear and easy path to herd immunity implied by the focused protection strategy.

On top of the human cost, the natural herd immunity proposition does not account for essential unknowns about duration of immunity and possibility for re-infection [6]. If immunity is short-lived—as it is for other coronaviruses—waning immunity would lead to a continuous supply of people returning to the ‘susceptible’ pool, leaving protective herd immunity perpetually out of reach. Vaccines, on the other hand, can induce more robust and durable immunity than natural infection, and can also be safely re-administered periodically, preventing people from returning to the susceptible pool [7,8].

Further, the natural herd immunity strategy does not acknowledge any risks of COVID-19 to younger people beyond death. But COVID-19 affects the lungs, heart, kidneys, blood vessels, and possibly the brain [9]. While we do not yet have good estimates of the prevalence of Long Covid, mounting evidence of long-term health effects of infection in some survivors should give us pause before actively seeking out widespread infection among the young.

If 80% of young people aged 0-34 were infected and only 5% of them experienced chronic Long Covid, this would be roughly 1.1 million young people in the UK with potentially debilitating symptoms requiring continued health care and hindering their ability to work and participate in society for months, years, or even life.

Finally, it is a straw man argument to claim that our COVID options are ‘majority back to normal’ versus ‘lockdown.’ While life is far from what it was prior to March, many activities have resumed, better treatments against COVID have been developed and people continue to take voluntary precautions (face coverings, social distancing). With the virus running amok and hospitalization and deaths surging even while shielding the vulnerable, it is unlikely that a large fraction of people would rush back to public spaces, meaning the strategy will fail to achieve its goal of resuming normal economic and social activity. Economists continuously emphasize this false trade-off between public health and the economy and the fact that economic activity will not return to normal in the presence of high transmission and overrun hospitals [10].

Natural herd immunity advocates create a sunny picture where those who are not vulnerable would immediately be allowed to resume life as normal. As we have shown, this false dichotomy bakes in substantial morbidity and mortality among the allegedly ‘non vulnerable’ who are expected to sacrifice themselves at the altar of population immunity.

The additional mortality from the inevitable spillover of infections to vulnerable groups combined with uncertainty about re-infection and duration of immunity and Long Covid clearly undermine the fantasy that this is the fastest way to bring society back to normal.


1 Alwan NA, Burgess RA, Ashworth S, et al. Scientific consensus on the COVID-19 pandemic: we need to act now. Lancet Published Online First: October 2020. doi:10.1016/S0140-6736(20)32153-X

2 Levin AT, Hanage, Willam P., Owusu-Boaitey N, et al. Assessing the age specificity of Infection Fatality Rates for COVID-19: Systematic review, meta-analysis and public policy implications. medRxiv Published Online First: 2020. doi:10.1101/2020.07.23.20160895

3 Block P, Hoffman M, Raabe IJ, et al. Social network-based distancing strategies to flatten the COVID-19 curve in a post-lockdown world. Nat Hum Behav 2020;4:588–96. doi:10.1038/s41562-020-0898-6

4 Ward H, et al. Antibody prevalence for SARS-CoV-2 in England following first peak of the pandemic: REACT2 study in 100,000 adults. medRxiv Published Online First: 2020.

5 Brett TS, Rohani P. Transmission dynamics reveal the impracticality of COVID-19 herd immunity strategies. Proc Natl Acad Sci 2020;117:25897–903. doi:10.1073/pnas.2008087117

6 Iwasaki A. What reinfections mean for COVID-19. Lancet Infect Dis Published Online First: October 2020. doi:10.1016/S1473-3099(20)30783-0

7 Anderson RM, May RM. Infectious Diseases of Humans: Dynamics and Control. Oxford: : Oxford University Press

8 Keech C, et al. First-in-Human Trial of a SARS CoV 2 Recombinant Spike Protein Nanoparticle Vaccine. medRxiv Published Online First: 2020.

9 Marshall M. The lasting misery of coronavirus long-haulers. Nature 2020;585:339–41. doi:10.1038/d41586-020-02598-6

10 Open Letter Swiss-based Economists to Federal Government. Saving the Economy Requires Controlling the Pandemic. Open Lett. 2020. (accessed 3 Nov 2020).